The Biology of Acinetobacter Taxonomy, Clinical Importance, Molecular Biology, Physiology, Industrial Relevance by K. J. Towner, E. Bergogne-Bérézin, C. A. Fewson (Auth.), K. J. Towner, E. Bergogne-B

The Biology of
A cineto bacter
Taxonomy, Clinical Importance,
Molecular Biology, Physiology,
Industrial Relevance
FEDERATION OF EUROPEAN MICROBIOLOGICAL SOCIETIES
SYMPOSIUM SERIES
Recent FEMS Symposium volumes published by Plenum Press
1990 • MOLECULAR BIOLOGY OF MEMBRANE-BOUND COMPLEXES IN
PHOTOTROPHIC BACTERIA
Edited by Gerhart Drews and Edwin A. Dawes
(FEMS Symposium No. 53)
1990 • MICROBIOLOGY AND BIOCHEMISTRY OF STRICT ANAEROBES

INVOLVED IN INTERSPECIES HYDROGEN TRANSFER
Edited by Jean-Pierre Belaich, Mireille Bruschi, and Jean-Louis Garcia
1990 • DENITRIFICATION IN SOIL AND SEDIMENT

Edited by Niels Peter Revsbech andJan ScDrensen
1991 • CANDIDA AND CANDIDAMYCOSIS

Edited by Emel Tiimbay, Heinz P. R. Seeliger, and Ozdem Ang
1991 • MICROBIAL SURFACE COMPONENTS AND TOXINS

IN RELATION TO PATHOGENESIS
Edited by Eliora Z. Ron and Shlomo Rottem
1991 • GENETICS AND PRODUCT FORMATION IN STREPTOMYCES

Edited by Simon Baumberg, Hans Kriigel, and Dieter Noack
1991 • THE BIOLOGY OFACINETOBACTER: Taxonomy, Clinical Importance,

Molecular Biology, Physiology, Industrial Relevance
Edited by K.J. Towner, E. Bergogne-Berezin, and C. A. Fewson
A Continuation Order Plan is available for this series. A continuation order will bring delivery of each new
volume immediately upon publication. Volumes are billed only upon actuaI'shipment. For further informa-
tion please contact the publisher.
The Biology of
Acinetobacter
Taxonomy, Clinical Importance,
Molecular Biology, Physiology,
Industrial Relevance
Edited by
K.}. Towner
Department of Microbiology and PHLS Laboratory
University Hospital
Nottingham, United Kingdom
E. Bergogne-Berezin
Departement de Microbiologie, CHU Bichat
Paris , France
and
C. A. Fewson
Department of Biochemistry
University of Glasgow
Glasgow, United Kingdom
Springer Science+Business Media, LLC

Library of Congress Cataloging in Publication Data
The Biology of Acinetobacter: Taxonomy, Clinical Impottance, Molecular Biology,
Physiology, Industrial Relevance / edited by K. J. Towner, E. Bergogne-Bérézin, and
C. A. Fewson.
p. cm. — (FEMS symposium; no. 57)
Based on the proceedings of the 2nd International Workshop on Acinetobacter held
under the auspices of the Federation of European Microbiological Societies, on Sept.
6-7, 1990 in Paris, France.
Includes bibliographical references and index.
ISBN 978-1-4899-3555-7
1. Acinetobacter —Congresses. I. Towner, K.J. II. Bergogne-Bérézin, E. III. Fewson,
C. A. IV Federation of European Microbiological Societies. V. International Workshop
on Acinetobacter (2nd: 1990: Paris, France) VI. Series.
QR82.N4B56 1991 91-12826
589.9'5-dc2O CIP
Based on the proceedings of a symposium held under the auspices of the

Federation of European Microbiological Societies,
on September 6-7, 1990, in Paris, France
ISBN 978-1-4899-3555-7 ISBN 978-1-4899-3553-3 (eBook)

DOI 10.1007/978-1-4899-3553-3
©1991Springer Science+Business Media New York

Originally published by Plenum Press, New York in 1991
Softcover reprint of the hardcover 1st edition 1991
All rights reserved

No part of this book may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical, photocopying, microfilming,
recording, or otherwise, without written permission from the Publisher
PREFACE
The 1st International Workshop on Acinetobacter was held on 6th

September, 1986, in Manchester, England, in association with the 14th
International Congress of Microbiology. That occasion was so well attended
and productive that there were soon discussions about how, when and where
the next meeting should be held. This time, however, there was sufficient
confidence to think of a more substantial meeting and to plan for the
proceedings to be published. It emerged that there was wide agreement that
the time was ripe to take stock of the entire biology of Acinetobacter: its
occurrence and taxonomy; its molecular biology, biochemistry and
physiology; its clinical importance and its industrial and commercial
applications. The 2nd International Workshop on Acinetobacter took
place from 6th to 7th September, 1990, at the Institut Pasteur, Paris, and was
sponsored by the Federation of European Microbiological Societies. There
were about 100 participants from 19 countries. The backbone of the meeting
consisted of 23 plenary lectures. There were 28 posters and the meeting
closed with a general discussion which went on long after the official
finishing time despite all the counter-attractions of a sunny Parisian Friday
afternoon. Indeed discussions continued while cruising along the Seine and
while dining at the top of the Tour Montparnasse. However, the vitality and
usefulness of even the most successful meeting is difficult to transmit by the
printed word. The editors decided, therefore, not simply to publish the
proceedings of the Workshop, but rather to commission a set of review
articles with the aim of giving an up-to-date and rounded picture of our
current knowledge of the genus Acinetobacter. These reviews, together
with a number of articles that expand on research talks presented at the
Workshop, form this book, the timely production of which owes much to the
expert assistance of Linda Bowering, Gerry Holmes and Gillian Johnston, to
whom the editors express their grateful thanks.
In addition to sponsorship by the Federation of European

Microbiological Societies, the Workshop was supported by financial
contributions from Beecham, Boots, Bristol-Meyers, Glaxo, ICI, Lilly, Merck
Sharpe & Dohme, Pharmuka, Roussel and Unilever, to all of whom the
organisers and participants are greatly indebted.
K.J. Towner
E. Bergogne-Berezin
C.A. Fewson
v
CONTENTS
Acinetobacter: Portrait of a Genus 1

K.J.Towner, E.Bergogne-Ben!zin, and C.A.Fewson
Taxonomy of Acinetobacter 25
P.A.D.Grimont and P.J.M.Bouvet
Typing of Acinetobacter 37
P.J.M.Bouvet
Hospital Epidemiology of Acinetobacter Infection 53

W.C.Noble
Epidemiology of Acinetobacter Strains Isolated from

Nosocomial Infections in France 63
M.L.J oly-Guillou, E.Bergogne-Berezin,
and J.F. Vieu
Molecular Epidemiological Analysis of Nosocomial

Acinetobacter baumannii Isolates 69
J.Vila, M.A.Canales, M.A.Marcos, R.Gomez-Lus,
and M.T.Jimenez de Anta
Factors Influencing the Virulence of Acinetobacter 77

J.-L.Avril and R.Mesnard
Antibiotic Resistance Mechanisms in Acinetobacter 83

E.Bergogne-Berezin and M.L.J oly-Guillou
The Chromosomal B-Lactamases of the Genus Acinetobacter:

Enzymes which Challenge Our Imagination 117
J.Hood and S.G.B.Amyes
The Use of Molecular Techniques for the Location and

Characterisation of Antibiotic Resistance Genes
in Clinical Isolates of Acinetobacter 133
B.G.Elisha and L.M.Steyn
vii
Plasmid and Transposon Behaviour in Acinetobacter
K.J.Towner 149
Plasmid Stability and the Expression and Regulation

of a Marker Gene in Acinetobacter and Other
Gram-negative Hosts
C.Winstanley, J.A.W.Morgan, J.R.Saunders, 169
and R.W.Pickup
Random Insertional Mutagenesis in Acinetobacter

B.Vosman, R.Kok, and H.J.Hellingwerf 183
Genetic Organisation of Acinetobacter

A.Vivian 191
Evolution of Genes for the B-Ketoadipate Pathway

in Acinetobacter calcoaceticus
L.N.Ornston and E.L.Neidle 201
Organisation, Potential Regulatory Elements and

Evolution of Trp Genes in Acinetobacter
G.Haspel, V.Kishan, and W.Hillen 239
Codon Usage in Acinetobacter Structural Genes

P.J.White, I.S.Hunter, and C.A.Fewson 251
The Outer Membrane of Acinetobacter: Structure-

Function Relationships
P.Borneleit and H.-P.Kleber 259
Energy Reserves in Acinetobacter

L.M.Fixter and M.K.Sherwani 273
Energy Generation and the Glucose Dehydrogenase

Pathway in Acinetobacter
J.A.Duine 295
Acinetobacter - Citric Acid Cyclist with a Difference

P.D.J.Weitzman 313
Metabolism of Alkanes by Acinetobacter

O.Asperger and H.-P.Kleber 323
Metabolism of Aromatic Compounds by Acinetobacter

C.A.Fewson 351
Metabolism of Cyclohexane and Related Alicyclic

Compounds by Acinetobacter
P.W.Trudgill 391
viii
Metabolism of Trimethylammonium Compounds
by Acinetobacter 403
H.-P.Kleber
Applications of Acinetobacter as an Industrial

Microorganism 411
D.L.Gutnick, R.AHon, C.Levy, R.Petter,
and W.Minas
Index 443
ix
ACINETOBACTER: PORTRAIT OF A GENUS
K.J.Towneru, E.Bergogne-Berezinband C.A.Fewson'
'Department of Microbiology & PHLS Laboratory

University Hospital
Nottingham NG7 2UH, UK
bLaboratoire de Microbiologie
CHU Bichat
75877 Paris Cedex 18, France
'Department of Biochemistry
Glasgow G12 8QQ, UK
INTRODUCTION
The genus Acinetobacter originally proposed by Brisou and Prevot

(1954) comprised a heterogenous collection of non-motile Gram-negative
organisms which could be distinguished from other similar organisms by
their lack of pigmentation (Ingram and Shewan, 1960). Now classified in the
family N eisseriaceae, the present generic description (Juni, 1984) allows
unambiguous identification of strains to the genus level. A difficulty arises in
that many current members of the genus have been classified previously
under a variety of different names (reviewed by Henriksen, 1973), and much
of the early literature concerning this group of organisms is, therefore,
difficult to interpret owing to confusion over nomenclature and the lack of a
widely-accepted classification scheme. Delineation of species within the
genus is still the subject of much research. The latest developments are
described in detail elsewhere (Grimont and Bouvet, this volume), but the
reader should bear in mind that many of the Acinetobacter strains
described in this volume still occupy a somewhat nebulous taxonomic
position within the genus.
Why the interest in this genus? Acinetobacters are ubiquitous

organisms in the environment and have formed the subject of several
previous reviews (e.g. Henriksen, 1973, 1976; Juni, 1978). They have been
implicated in a variety of food spoilage and human disease processes. They
are often resistant to commonly used antibiotics and may form a reservoir of
antibiotic resistance genes, particularly in hospital environments. They are
easy to isolate and can be grown in the laboratory in a simple mineral
medium containing a single carbon and energy source; indeed the vast
majority of isolates resemble saprophytic pseudomonads in being able to use
anyone of a large number of organic compounds as a carbon and energy
source. Although the utilisation of carbohydrates is relatively uncommon, a
major biochemical feature of the genus is the range of compounds that can
be metabolised. This range encompasses aliphatic alcohols (including 2,3-
butanediol), amino acids, dicarboxylic and fatty acids, n-alkanes, alicyclic
compounds, and many aromatic compounds. Members of the genus are,
therefore, particularly suitable organisms for studying unusual peripheral
biochemical pathways, and may also have an important role to play in the
biodegradation of a variety of hazardous and unpleasant aromatic pollutants
and industrial residues. All three of the major modes of gene transfer are
known to occur in Acinetobacter, and it is already clear from detailed
studies of gene structure, organisation and genetic regulation that
Acinetobacter has some unusual and exciting features that may help in
gaining an understanding of important processes that occur during
prokaryotic evolution in general. The aim of this introductory article is to
present a broad overview of the genus Acinetobacter and to highlight the
most important features, many of which are described in detail in the
remainder of this volume.
IDENTIFICATION AND TAXONOMY
Morphological and Biochemical Identification
Members of the genus Acinetobacter are short, plump, Gram-

negative rods that often become more coccoid in the stationary phase. Cells
commonly occur in pairs or chains of variable length that are occasionally
difficult to destain. No spores are formed. Flagella are absent but, although
generally considered to be non-motile, "twitching" and "gliding" motility has
been reported, particularly on semi-solid media (Barker and Maxsted, 1975;
Henrichsen, 1975; Mukherji and Bhopale, 1983). Many strains are
encapsulated. All strains of Acinetobacter are non-fastidious strict aerobes
that are catalase-positive and oxidase-negative. Most strains can grow in a
simple minimal medium containing a single carbon and energy source at a
temperature between 20 and 30·C. The GC content is 38-45 mol %. There is
no single biochemical test that enables differentiation of this genus, but they
can be readily distinguished from similar bacteria by means of a combination
of standard nutritional tests for non-fastidious, non-fermentative organisms.
Unambiguous attribution of strains to the genus Acinetobacter is achieved
by means of the DNA transformation test described by Juni (1972).
2
Classification and Nomenclature
Members of the genus Acinetobacter have suffered a long history of

extensive taxonomic changes which has inhibited a proper comparative study
of their properties. At least 15 different "generic" names have been used to
describe these organisms (Baumann et aI., 1968), the most common of which
are Bacterium anitratum (Shaub and Hauber, 1948), H erellea
vaginicola and Mima polymorpha (Debord, 1939), Micrococcus
calcoaceticus and "B5W" (Juni, 1978), and M oraxella glucidolytica and
M oraxella lwoffii (Piechaud et aI., 1956; Brisou, 1957). The genus
Acinetobacter, as conceived by Brisou (1957) and Piechaud et al. (1956),
included both oxidase-positive (M oraxella) and oxidase-negative strains. In
1971 the Subcommittee on the Taxonomy of Moraxella and Allied Bacteria
proposed that the genus Acinetobacter should include only the oxidase-
negative strains. Classified in the family Neisseriaceae, the genus
Acinetobacter originally comprised only one species, A. calcoaceticus,
with, according to the Approved Lists of Bacterial Names (Skermann et aI.,
1980), two varieties: var. anitratus (formerly H erellea vaginicola) and
var. lwoffii (formerly Mima polymorpha). The latest definition of
Acinetobacter (Bouvet and Grimont, 1987) was arrived at by using a
combination of phenotypic properties (including carbon source utilisation
tests) and identification of genotypic species by means of modern taxonomic
methods (genetic transformation, DNA hybridisation and RNA sequence
comparison). This new definition of Acinetobacter has resulted in the
identification of 17 genomic species and 19 biotypes. The major genomic
species comprise A. baumannii (with nine biotypes), Acinetobacter sp. 3
(six biotypes), A. haemolyticus, A. johnsonii, A. junii, A. lwoffii, A.
calcoaceticus, Acinetobacter sp. 6 and Acinetobacter sp. 11 (for
further details see Grimont and Bouvet, this volume). Simultaneously,
independent studies performed by Tjernberg and Ursing (1989) and
Nishimura et a!. (1988b) have provided data compatible with the groupings
outlined above. Phenotypic group 5 has been named A. radioresistens
(Nishimura et al. 1988a; Ino and Nishimura, 1990). Thus a satisfactory
classification of the genus Acinetobacter is now at the disposal of
microbiologists, and Acinetobacter is recognised as a group clearly distinct
from M oraxella and Neisseria.
OCCURRENCE AND ISOLATION
Strains of Acinetobacter can be obtained easily from soil, water and

sewage with appropriate enrichment techniques (Baumann, 1968), and
various methods have been developed for their enumeration (LaCroix and
Cabelli, 1982; Bifulco et aI., 1989). It has been estimated that acinetobacters
may constitute 0.001 % of the total heterotrophic aerobic population of soil
and water (Baumann, 1968). They have been found to be common in
American rural drinking water supplies (Bifulco et a!., 1989) and to be
present at densities exceeding 10 4 organisms /100 ml in freshwater
ecosystems and 106 organisms /100 ml in raw sewage (LaCroix and Cabelli,
3
1982). Although they can be isolated from heavily polluted water, they are
found more frequently near the surface of freshwater and where freshwater
flows into the sea (Droop and Jannasch, 1977). Acinetobacters are also
found in a variety of foodstuffs, including milk products, fresh meat,
eviscerated chicken carcasses and other poultry meats (Koburger, 1964;
Barnes and Thornley, 1966; Eribo and Jay, 1985). Members of the genus are
responsible for spoilage of a number of foods, including chickens, eggs,
bacon and fish, even when stored under refrigerated conditions (Shewan et
al., 1960; Thornley et a!., 1960; Gardner, 1971; Jay, 1982) or following
irradiation (Firstenherg-Eden et al., 1980). Acinetobacters are normal
inhabitants of human skin and have consequently been implicated as
presumed causal or contributory agents in a variety of infectious disease
processes (see below). They also have some veterinary significance (e.g.
Dickie, 1978; Gibson and Eaves, 1981). Acinetohacters have also heen
isolated from a bizarre range of other sources, including laboratory dental
pumice (Williams et aI., 1983), eye shadows (Dawson and Reinhardt, 1981)
and stored buffalo semen (Ramachandra et a!., 1981).
Isolation of acinetobacters can be achieved on standard laboratory

media such as nutrient agar or trypticase soy agar. Most strains have an
optimum growth temperature of 30-35° C, and grow well at 37" C, but some
isolates are mesophilic and may be unable to grow at 37" C (Breuil and
Kushner, 1975). Baumann (1968) described an enrichment culture
procedure for isolating members of the genus from soil and water which
involves the use of a simple mineral medium with acetatp. as the sole carbon
and energy source, and Holton (1983) descrihed a selective differential
medium for the isolation of acinetobacters from clinical specimens.
Acinetobacters have a slightly acid pH optimum for growth, and vigorous
aeration at a pH of 5.5-6.0 favours their enrichment.
CLINICAL IMPORTANCE OF ACINETOBACTER SPP.
Extent of the Problem
Strains of Acinetobacter are distributed widely in the hospital

environment and have been isolated from various inanimate sources such as
ventilatory equipment, blood collection tubes, mattresses and plastic urinals.
Nevertheless, it seems that the skin of patients and staff is the likely source
for most outbreaks of nosocomial Acinetobacter infection (Bergogne-
Berezin et a!., 1987; Noble, this volume). It has been shown that 29% of
nurses who had contact with infected patients carried epidemic strains
(French et aI., 1980), but Acinetobacter spp. may become established in
hospitals as part of the resident skin flora in addition to being simply
transient contaminants on the hands of staff and patients. A. baumannii is
the main species associated with nosocomial infection (Bouvet and Grimont,
1987), whereas other species such as A. haemolyticus, A. junii, A.
4
johnsonii or A. lwoffii seem to be isolated primarily from environmental
sources and only occasionally from clinical samples.
The pathogenic role of Acinetobacter is limited to nosocomial
infections, but a dramatic increase in the incidence of this organism has been
noted in many reports during the last few years, particularly when compared
with other common organisms causing nosocomial infection. Joly-Guillou et
aI. (1990b) reported that Acinetobacter strains comprised 1-9% of all
bacterial species ,isolated from clinical specimens between 1984 and 1988.
The major sites of infection do not differ from those of other nosocomial
Gram-negative pathogens, with the lower respiratory tract and the urinary
tract accounting for 15-28% of total Aci n e to b ac te r infections (Retailliau et
al.,1979; Joly-Guillou et al., 1990b). Nosocomial pneumonia caused by
Acinetobacter spp. is a particular problem in intensive care units (Cunha et
aI., 1980; Vieu et aI., 1980; Muller-Serieys et aI., 1989), and often involves
patients with severe underlying disease requiring assisted mechanical
ventilation with equipment that is capable of aerosolising the organism.
Some isolates of Acinetobacter may reflect colonisation rather than
infection (French et aI., 1980), but the presence of pneumonia with a lobar
distribution, purulent secretions, and isolation of the organism as the sole or
predominant pathogen from bronchial brushings confirms the pathogenic
role of Acinetobacter. Nosocomial urinary tract infections caused by
Acinetobacter spp. occur at rates varying from 2-61%, depending on local
outbreaks (Hoffmann et aI., 1982; Mayer and Zinner, 1985). Other
nosocomial infections occasionally reported in the medical literature include
meningitis, skin and wound infections, burn wound superinfections and
bacteraemia (Bergogne-Berezin et aI., 1987).
The severity and difficulties associated with Acinetobacter

nosocomial infection are partly related to the high rates of resistance to most
major antibiotics found in Acinetobacter. The evolution of antibiotic
resistance in Acinetobacter spp. over the last 20 years has resulted in a
progressive decrease in susceptibility to commonly used antimicrobial drugs,
to the extent that acinetobacters now exhibit one of the most impressive
patterns of antibiotic resistance found in nosocomial bacteria.
Acinetobacters are frequently resistant to most available antibacterial drugs
of the B-Iactam and aminoglycoside classes, with increasing proportions of
clinical isolates becoming resistant to the initially effective
fluoroquinolones. Even resistance to imipenem, which was for several years
the most effective drug in treating Acinetobacter nosocomial infections,
has now been described in several reports. Detailed analysis of the unusual
large molecular mass chromosomal B-Iactamases produced by members of
the genus has raised the exciting hypothesis that these enzymes could
represent an intermediate evolutionary step between "normal" B-Iactamases
and penicillin binding proteins (Hood and Amyes, this volume). A full
description of the different antibiotic resistance mechanisms in
Acinetobacter can be found in the review by Bergogne-Berezin and Joly-
Guillou (this volume).
5
Pathogenicity and Virulence Factors
The virulence of Acinetobacter has been studied primarily in animal

models involving mice (Lutz et ai., 1956; Daly et ai., 1962; Obana et aI.,
1985; Obana, 1986; reviewed by Avril, this volume) and rats (Loeffelholz and
Modrzakowski, 1988). Strains of A. baumannii have limited virulence in
mice (LDso values of 106 -108 cfu/mouse) when inoculated intraperitoneally,
even in neutropenic mice treated with cyclophosphamide. In guinea pigs,
the virulence of Acinetobacter strains has been shown to be even more
limited following sub-cutaneous inoculation (Lutz et aI., 1956; Von
Graevenitz, 1983). Analysis of possible antigenic determinants has shown
that Acinetobacter produces a capsular polysaccharide formed of L-
rhamnose and D-glucose, or galactose (Juni, 1978), but apart from its
possible role in mediating adhesion to cells (Rosenberg and Rosenberg,
1981), the capsule does not seem to play an important virulence role. In
contrast, the production of "slime", a mucous extravascular substance, does
seem to result in enhanced virulence of Acinetobacter (Obana, 1986;
Avril, this volume). Lipopolysaccharide from Acinetobacter has certain
unusual structural features (Borneleit and Kleber, this volume) and may also
have pathogenic properties.
Typing and Epidemiology
Analysis of the cell structure of acinetobacters has also provided the

basis for certain typing systems. Hospital outbreaks of Acinetobacter
infection require suitable typing methods for epidemiological studies. Novel
systems that have beeq investigated in recent years include fluorescent
antibody serotyping to detect capsular antigens, phage-typing (Bergogne-
Bef(~zin et aI., 1987), bacteriocin-typing, cell envelope protein patterns
(Dijkshoorn et ai., 1987) and plasmid analysis (Gerner-Smidt, 1989; Harstein
et aI., 1990; Joly-Guillou et aI., 1990a; Vila et ai., this volume).
Determination of antibiograms and biotypes has also proved to be useful, but
most investigators have concluded that, at present, a single reliable typing
system for Acinetobacter does not exist, and that epidemiological studies
require a combination of several different approaches. The current situation
regarding typing systems is reviewed by Bouvet (this volume), and examples
of the application of some of these typing systems to outbreaks of nosocomial
Acinetobacter infection are described by Joly-Guillou et al. (this volume)
and Vila et al. (this volume).
GENETICS AND MOLECULAR BIOLOGY
Plasmid and Transposon Behaviour
There is still a relative paucity of information about the detailed role

of plasmids and transposons in the overall biology of Acinetobacter. The
clinical importance of antibiotic resistance in nosocomial isolates of
Acinetobacter has already been alluded to above. A variety of plasmids and
6
transposons conferring antibiotic resistance can be transferred to
Acinetobacter (Towner, this volume), where they may form an important
hospital reservoir of antibiotic resistance genes. Although much is known
about the biochemical mechanisms of antibiotic resistance in
Acinetobacter (Bergogne-Bert!zin and Joly-Guillou, this volume), little is
known about the structure and regulation of these genes in Acinetobacter
at the molecular level. The complex nature of the problem and the efforts
needed to unravel it are well-illustrated by Elisha and Stern (this volume).
Plasmids also seem to be involved in determining a range of other phenotypic
properties in Acinetobacter (Towner, this volume). Again, there is a
relative lack of information, but it seems clear that while some plasmids can
transfer readily between Acinetobacter and other genera, there is also a
pool of plasmid-mediated genetic information which is confined largely to
Acinetobacter and not readily mobilis able to other organisms. This seems
to result,however, from a barrier to transfer rather than expression (Towner,
this volume). Further information on plasmid stability and expression of
plasmid-encoded genes in Ac ine to b a c te r is provided by Winstanley et aI.
(this volume), with particular reference to the release of genetically
engineered microorganisms (GEMs) into the natural environment.
Chromosome Organisation and Gene Regulation
Acinetobacter has been shown to have a circular chromosomal

linkage map (Towner, 1978), and a total of 29 genetic loci have now been
mapped (Vivian, this volume). It is not possible, as yet, to discern any overall
features in the genome arrangement, but all three major modes of gene
transfer (transformation, conjugation and transduction) have been
demonstrated to occur in Acinetobacter (Vivian, this volume), and so the
genetic tools for future work aimed at defining the gross topological
structure of the chromosome are readily applicable. Methods for random
insertional mutagenesis, marker rescue and the cloning of Acinetobacter
genes in appropriate vectors are also available (Goosen et aI., 1987; Hunger
et aI., 1990; articles by Gutnick et al.; Ornston and Neidle; Vosman et aI.,
this volume).
So far as fine structure of the chromosome is concerned, particular

attention has focused on the organisation and regulation of genes concerned
with tryptophan biosynthesis (Haspel et aI., this volume) and the B-
ketoadipate degradative pathway (Ornston and Neidle, this volume). With
regard to tryptophan biosynthesis, striking similarities exist with respect to
the organisation and regulation of the trp genes from A. calcoaceticus and
Pseudomonas putida, but the complete regulatory circuits for this pathway
have yet to be fully resolved in Acinetobacter (Haspel et aI., this volume).
Studies of Acinetobacter genes encoding the B-ketoadipate pathway have
indicated homologies between the Acinetobacter and Pseudomonas
systems, but with different forms of transcriptional regulation. Although the
GC content of the structural genes for the various enzymes differs by about
20%, there is a striking level of amino acid sequence identity. The difference
in GC content is reflected in distinctive patterns of codon usage (articles by
7
Haspel et al.; Omston and Neidle; White et aI., this volume). Studies by
L.N.Ornston and colleagues have demonstrated that considerable gene
rearrangements must have occurred as the B-ketoadipate pathway diverged in
Acinetobacter and Pseudomonas, and their results have generated the
exciting hypothesis that sequence exchange between DNA slippage
structures causes mutations leading to genetic divergence, while repair
events involving sequence exchange between slipped DNA strands may
contribute to conservation of the evolved sequences. The experiments and
discoveries leading to these stimulating concepts are described in detail by
Omston and Neidle (this volume).
BIOCHEMISTRY AND PHYSIOLOGY
Energetics oj Growth, Electron TransJer and Oxidative

Phosphorylation, and Transport Mechanisms
There have been systematic examinations of the effects of variables

such as substrate, temperature, oxygen status, growth rate and nutrient
limitation on the yields and composition of several strains of Acinetobacter
(Abbott et aI., 1974; Kleber et aI., 1975; Du Preez and Toerien, 1978; Du
Preez et a!., 1981, 1984; Fewson, 1985; Hardy and Dawes, 1985), and in
some cases substrate inhibition has been found to he important (Abbott,
1973; Jones et aI., 1973; Fewson, 1985). Growth yields and respiratory
efficiencies of A. c:alcoac:eticus NCIB8250 have been measured in the
course of hatch and continuous culture experiments (Fewson, 1985; Hardy
and Dawes, 1985). Growth yields were generally found to be rather low,
probably because the effective P/O ratio is only about one under most
conditions, even though there may be two or more sites of oxidative
phosphorylation (Fewson, 1985). Failure to make use of all the potential
synthesis of ATP is consistent with the conclusions that the redox chain of A.
calcoac:eticus H01-N is divided into two equivalent sites of energy
conservation (Ensley et aI., 1981) and that proton translocating loops 0, 1
and 2 are active in A. twoffii 4B (Meyer and Jones, 1973; Jones et aI.,
1977a). Results of experiments to determine H + /0 quotients all seem to
suggest that there are at least two sites of oxidative phosphorylation (Meyer
and Jones, 1973; Jones et aI., 1977a; Hardy and Dawes, 19R5). There is
evidence that some strains of Acinetobacter which cannot grow on glucose
as sole source of carbon and energy can nevertheless use the
dehydrogenation of glucose to energise active transport systems and to
provide ATP for biosynthesis (Kitagawa et a!., 1986a,b). However, the
observation that glucose can increase the molar growth yield on acetate
(Muller and Babel, 1986; van Schie et aI., 1987) may have to he subjected to
more detailed interpretation in order fully to understand its quantitative
significance (Duine, this volume).
The nature of the electron transfer chain in acinetohacters is not

entirely clear despite many year's work (e.g. Whittaker, 1971; Meyer and
Jones,1973; Jones et aI., 1977a; Ensley and Finnerty, 1980; Ensley et aI.,
8
1981; Hardy and Dawes, 1985; Asperger and Kleber, this volume). In
general, under conditions of high aeration a cytochrome o-containing
oxidase and a cytochrome b554 are the predominant species, whereas under
oxygen-limited conditions a cytochrome d-containing oxidase is predominant
(Ensley and Finnerty, 1980; Hardy and Dawes, 1985; Geerlof et al., 1989).
In recent work with A. calcoaceticus strain LMD 79.41, cytochrome 0
oxidase, cytochrome d oxidase and cytochrome b S54 have all been partially
purified and characterised (Geerlof et al., 1989). Cytochrome b S62 is
involved in linking glucose dehydrogenase to the main electron transfer
chain (Beardmore-Gray and Anthony, 1986; Dokter et al., 1988; Geerlof et
al., 1989; Duine, this volume). There is no evidence that Acinetobacter
spp. contain a cytochrome c, and this accords with their status as oxidase-
negative organisms. Several strains contain cytochrome P450 and this is
reviewed by Asperger and Kleber (this volume). An NADH dehydrogenase
connected with the respiratory chain has been solubilised, purified and
partially characterised (Borneleit and Kleber, 1983). Ubiquinones, but not
menaquinone, have been found in acinetobacters, including the neotype
strain ATCC 23055. The number of isoprene units attached to the
benzoquinone nucleus is typically nine, but small amounts of the next higher
and lower isoprenologues are usually found (Whittaker, 1971; Collins and
Jones, 1981; Nishimura et al., 1986, 1988a; L.M. Fixter, unpublished
results); however, some Acinetobacter strains appear to contain either
approximately equal amounts of CoOs and Co0 9 , or just CoOs (Denis et al.,
1975).
It is clear that acinetobacters contain active transport systems for

several substrates, and that in some strains these can be energised by glucose
dehydrogenation (van Schie et al., 1985; Kitagawa et al., 1986 a,b). There is
also evidence that some aromatic compounds are taken up by specific
transport systems (Fewson, this volume). The cytoplasmic membrane of A.
calcoaceticus NCIB8250 appears to be impermeable to monosaccharides
(Cook and Fewson, 1973), but Midgley et al. (1986a,b) have presented
evidence that the apparent impermeability of strain NCIB8250 to low
concentrations of methyltriphenylphosphonium ion is partly the consequence
of an efflux system. Some strains of Acinetobacter produce siderophores
and iron-repressible outer membrane proteins (Smith et al., 1990;
P.Williams and K.J.Towner, unpublished results), and these observations
could have implications for explaining the organism's virulence in relation to
the availability of iron. The entire question of uptake of substances into
acinetobacters, and efflux systems from them, requires a great deal more
effort if we are to understand their physiology and pathogenicity adequately.
Metabolism and Enzymology
Early work on the metabolism of Acinetob acter was fully reviewed by

Juni (1978); since then there has been a steady flow of papers describing
enzymes, co-factors, metabolic pathways and products, as well as other
aspects of the biochemistry and physiology of the genus. In general, these
have confirmed the view that Acinetobacter is typical of other Gram-
9
negative eubacteria, but that it has distinctive metabolic features which are
well-adapted to its lifestyle and support its nutritional versatility.
Metabolism centres on a Krebs tricarboxylic acid cycle which has a

rather unusual pattern of regulation. The activities of pyruvate
dehydrogenase, citrate synthase, isocitrate dehydrogenase and 2-
oxoglutarate dehydrogenase are all modulated by AMP, and this may act in a
concerted manner to direct metabolite flux through the cycle. This
distinctive multi-point control of the cycle may be a significant feature in the
regulation of energy metabolism in Acinetobacter (Weitzman, this
volume). There is a repressible glyoxylate bypass (references in Juni, 1978;
Asperger and Kleber, this volume). Phosphoenolpyruvate carboxylase
serves an anaplerotic function under tight control, especially by nucleotides.
Again, AMP appears to play an important role. The activation of
phosphoenolpyruvate carboxylase by AMP may be physiologically
advantageous hecause, under such conditions, there may be a high demand
for oxaloacetate by the tricarboxylic acid cycle (Richter and Kleber, 1980).
Oxaloacetate can be degraded by a coupled reaction involving malic enzyme
and NAD-linked malate dehydrogenase (Juni, 1978). Many acinetobacters
can grow in butane 2,3-diol and convert it into acetate by a cyclic mechanism
(Juni, 1978). A plethora of complex catabolic pathways for the degradation
of aromatic and alicyclic compounds, as well as alkanes, amino acids and
related compounds, feed into the tricarboxylic acid cycle (articles by
Asperger and Kleber; Fewson; Ornston and Neidle; Trudgill, this volume).
Biosynthetic pathways, such as those involved in gluconeogenesis and amino
acid synthesis, are subtly influenced by the need to integrate with the overall
metabolism of the organism. This requirement shows up with respect to the
details of regulation of gene expression as well as the regulation of enzyme
activity by allosteric and other mechanisms (e.g. Mukkada and Bell, 1971;
Juni, 1978; Haspel et al., this volume). There is no good evidence for the
operation in Acinetobacter spp. of global metabolic regulatory systems,
such as those involving cyclic AMP, but presumably they exist, and it has
been suggested that there is a Pho regulon which is controlled by a single
protein analogous to the product of the phoB gene of E. coli (Yashphe et
al., 1990).
Many strains of Acinetobacter accumulate poly-B-hydroxybutyrate

and wax esters as energy reserves, and some strains accumulate
polyphosphate, either as an energy reserve or as a phosphate reserve (Fixter
and Sherwani, this volume). The mechanisms involved in polyphosphate
metabolism provide a particular challenge because our present knowledge of
the enzymes that might be involved is insufficient to provide a coherent
explanation for the synthesis and degradation of polyphosphate in the genus
as a whole.
Amongst the more special features of the genus is the ability of at least
one strain to grow on malonate, apparently by converting it into malonyl-
CoA followed by decarboxylation to acetyl-CoA (Kim and Kim 1985; Bahk
10
and Kim, 1987; Lim and Kim, 1987). The acetate kinase of this strain
appears to be quite different from the equivalent enzyme in E. coli (Kim and
Park, 1988). Strain JCl is a carboxydobacterium, and its inducible carbon
monoxide dehydrogenase has been isolated and partially characterised (Kim
and Kim, 1989; Kim et ai., 1989). There has also been a report that some
acinetobacters, isolated from the surfaces of soybean nodules (where there
may be appreciable conccntrations of hydrogen because of the action of
nitrogenase), can oxidise hydrogen (Wong et ai., 1986).
Despite the name 'anitratus', many strains of Acinetobacter can

assimilate nitrate, presumably without appreciable accumulation of nitrite
(Villalobo et aI., 1977; Juni, 1978). Ammonia appears to be assimilated by
an NADP-dependent glutamate dehydrogenase, at least in strain NCIB8250.
When glutamate is used as the nitrogen source for growth, the initial step
involves transamination to form aspartate; if, in addition, the glutamate is
serving as the source of carbon and energy, then aspartase is induced and
releases surplus ammonia (A.C. Borthwick, L.M. Fixter and C.A. Fewson,
unpublished results). A sulphur transferase and a rhodanase have both been
isolated from acinetobacters (Vandenbergh and Berk, 1980; Layh et al.,
1982; Aird ct al., 1987). Acinetobacter twoffii GB1 can volatilise
mercury from media containing HgCl 2 (Bhattacharyya and MandaI, 1987),
and enzymic reduction of HgCl 2 has been demonstrated in cell extracts
(Chandhuri et aI., 1990).
The enzymes of nucleic acid metabolism in Acinetobacter have not

been much studied at thc molecular lcvel, although the overall pattern was
established some time ago (Juni, 1978). A. calcoaceticus has at least
three restriction endonucleases, Acel, AcelI and AccIII, whose recognition
sequences and cleavage sites make them useful for genetic recombination
work (Kita et aI., 1985).
Acinetobacters seem to be quite adept at producing extracellular

enzymes (Juni, 1978); examples include lipases and esterases (Breuil and
Kushner, 1975), amylase (Onishi and Hidaka, 1978), collagenase (Monboisse
et aI., 1979), endo-B-1,6-glucanase (Kotohda et ai., 1979) and B-lactamase
(Blechschmidt ct ai., 1989). There are several interesting periplasmic
enzymes (Borneleit and Kleber, this volume), including an insulin-cleaving
metalloprotease (Fricke and Aurich, 1989), and membranc-bound enzymes
in abundance, including an NAD-independent malate dehydrogenase (Jones
et al., 1977b), proteases (Fricke et ai., 1987), leucine aminopeptidase
(Ludewig et al., 1987), alanine aminopeptidase (Jahreis et al., 1989) and
NAD-indepcndent aldehyde, alcohol, lactate and mandelate dehydrogenases
(references in Asperger and Kle ber, this volume; F ewson, this volume).
So far as co-factors are concerned, acinetobacters have played a major

role in the elucidation of the biological role of pyrroloquinoline quinone
(PQQ), particularly in respect of its function in glucose dehydrogenase
(Jongejan and Duine, 1989; Duine, this volume).
11
INDUSTRIAL, COMMERCIAL, CLINICAL AND
OTHER APPLICATIONS
A quite astonishing range of uses for Acinptobacter and its

products can be gleaned from the scientific and patent literature (Table 1),
although for one reason or another, not all of them may have proved to be
successful. Around 200 processes involving Acinetobacter had been
patented, or patents had been applied for, by mid-1990. In some cases the
patents gave almost world-wide coverage, such as those for extracellular
emulsifiers of oils and hydrocarbons in the European Community, the United
States, Canada, Japan, Norway and Denmark. Perhaps the most novel and
surprising of all the suggestions for using Acinetobacter is that for the
enhancement of the clarity of fingerprints. The organisms proliferate on the
material present in the deposit of barely detectable prints; they then multiply
and form colonies on the ridges and, as a result, the characteristic fingerprint
pattern becomes visible (Harper et al., 1987).
Products
The biochemistry, molecular biology and applications of Acineto-

bacter bioemulsifying agents are extensively reviewed by Gutnick et al. (this
volume), and these compounds appear likely to bring the organism to the
attention of a new audience. Exploitation of the ability of acinetobacters to
accumulate simple wax esters (Neidleman and Geigert, 1984; Fixter and
Sherwani, this volume) will largely depend on the economics of the process,
particularly the cost of feedstocks. These waxes might perhaps be used for
the production of scarce alcohols. It is clear that the composition of the
waxes, in terms of chain length and degree of unsaturation, can be strongly
influenced by the conditions under which the organism is grown (M.J.
McAvoy, L.M. Fixter and C.A. Fewson, unpublished results), and these
observations might be worthy of industrial application.
Pollution Control
The involvement of Acinetobacter strains in the removal of

phosphate from waste-water is reviewed by Gutnick et al. (this volume), and
this process, which is important in minimising eutrophication, depends on
the ability of the organisms to accumulate polyphosphate, which is in turn
discussed by Fixter and Sherwani (this volume). The degradation of
pollutant aromatic and alicyclic compounds and hydrocarbons which enable
Acinetobacter to be used for effluent and other waste treatment has also
been reviewed (articles by Asperger and Kleber; Fewson, this volume), and
there have been attempts to increase the degradative capability of
acinetobacters by genetic engineering (e.g. Liaw and Srinivasan, 1990).
BiotransJ ormati ons
The general versatility of acinetobacters with regard to their ability to

metabolise aromatic, alicyclic and a great range of other compounds
12
underlies some of the possible commercial biotransformation processes
(articles by Asperger and Kleber; Fewson; Trudgill, this volume), including
the production of L-carnitine (Kleber, this volume). Production of gluconic
acid depends on the possession of glucose dehydrogenase, a fascinating
Table 1. Industrial, commercial. clinical and other applications of

Acinetobacter
Production of emulsifying agents, e.g. Emulsan, Biodispersan

-to aid oil biodegradation
-to stabilise heavy oil in water emulsions and coal in water slurries (in
order to facilitate transportation by reduction in viscosity and to
increase the efficiency of combustion)
-to help prevent dental plaque
-to facilitate grinding of limestone for use in paper making
Waste and effluent treatment
-phosphate removal from waste-water
-effluent purification in the petrochemical industry
-degradation of cyclohexane, acrylonitrile, polyacrylate, acrylamide,
polyethylene, poly(vinyl alcohol) resins, linear alkylbenzene
sulphonates, aromatic hydrocarbons, cresols, phthalic acid esters
Anti-tumour and anti-leukaemia agents, e.g. asparaginase-glutaminase
Production of optically-active carboxylic acids and their enantiomeric
esters, gluconic acid, racemic thiol-containing esters, amino acids
and related compounds, L-carnitine, unsaturated organic acids
Components of cosmetics, detergents and shampoos, e. g. fragrant
compositions, bioemulsifiers, moisturisers
Wax esters, e.g. for the production of medium-chain alcohols
Enzymes, e.g. Accl restriction endonuclease, mutarotase, lipases, glucose
dehydrogenase, aspartylglycosylamine amidohydrolase
Coal and petroleum desulphurisation
Manganese leaching from ores
Production of immune adjuvants
Decaffeination processes and debittering of orange juice
Production of protein-rich biomass
Development of latent fingerprints
enzyme in its own right (Duine, this volume) and one which may have
applications in the construction of glucose sensors (D'Costa et aI., 1986) and
in the determination of POO (B.W. Groen and J.A. Duine, unpublished
results).
13
Biomass Production
Production of single cell protein for animal feed is often limited by

costs of possible feedstocks. Economically-viable processes must be based
on raw materials that are cheap, plentiful and in constant supply. Two
possible applications of Ac in e to b act e r for single cell protein production
involve growth on palm oil (Koh et aI., 19H5, 19H7) or the effluent from the
industrial production of oil from coal (Du Preez et aI., 1985). In addition,
ethanol (Abbott et aI., 1973) or alkanes (Asperger and Kle ber, this volume)
might, at some time in certain parts of the world, become cheap enough to
support protein production in the form of Acinetobacter biomass.
G lu taminase -Asparaginase
The possible use of glutaminase-asparaginase as an anti-cancer agent

has led to this being perhaps the best-studied of all Acinetobacter enzymes.
The logic of the approach is that cancerous cells may lack the enzymic
capability to synthesise adequate amounts of certain non-essential amino
acids, and if the circulatory level of these amino acids is reduced the
cancerous cells will be selectively killed. Parenteral administration of L-
asparaginase has caused regression of lymphomas and leukaemias in
experimental animals and humans. The E. coli enzyme has been used in the
treatment of acute lymphoblastic leukaemia for many years, and other
amidohydrolases have been used in clinical trials (Ammon et aI., 1988).
glutaminase-asparaginase has been purified from A. calcoaceticus (Joner
et aI., 1973; Joner, 1976) and from A. "glutaminasificans" (Holcenberg,
1985). The latter enzyme uses both L-glutamine and L-asparagine as
substrates and is a tetramer of identical subunits (M, = 35,500; 331 residues)
(Tanaka et aI., 1988). Detailed studies of its kinetic properties and substrate
binding site have been undertaken (Holcenberg et aI., 1978; Steckel et aI.,
1983; Holcenberg, 1985). It has been crystallised (Wlodawer et aI., 1977;
Ammon et aI., 1983), and a preliminary quaternary structure has been
reported showing that each subunit folds into two domains: the amino
terminal domain contains a five-stranded B-sheet surrounded by five a-
helices, while the carboxyl terminal domain contains three a-helices (Ammon
et aI., 1988). Various studies have been aimed at increasing the circulatory
life of the enzyme and at reducing immunological complications; these have
included chemical modifications such as succinylation and glycosylation (e.g.
Holcenberg et aI., 1975, 1979; Blazek and Benbollgh, 1981), and
immobilisation in microspheres (e.g. Edman et aI., 1987). There is an
exhaustive literature on the possible application of Acinetobacter
glutaminase-asparaginase in cancer therapy (e.g. Kien et a!., 1985; Ammon
et aI., 1988).
ENVOI
Acinetobacter is a genus that has, for many years, been somewhat

ignored by the main body of microbiologists. However, rapidly dawning
14
interest in the genus resulted in the meeting which formed the basis for this
book, and it is hoped that this brief introductory article will encourage the
casual reader to delve more deeply into the wealth of scientific material
which follows. Members of the genus Acinetobacter have a plethora of
interesting facets to their lifestyles, some of which are unique, while others
provide us with exciting insights into prokaryotic behaviour in general. We
firmly believe that scientists, of whatever microbiological speciality, cannot
fail to be stimulated by a closer acquaintance with this fascinating group of
organisms.
REFERENCES
Abbott, B.l., 1973, Ethanol inhibition of a bacterium

(Acinetobacter calcoaceticus) in chemostat cultures, 1. Gen.
Microbiol., 75:383.
Abbott, B.l., Laskin, A.I., and McCoy, C.l., 1973, Growth of
Acinetobacter calcoaceticus on ethanol, Appl. Microbio!.,
25:787.
Abbott, B.l., Laskin, A.l., and McCoy, C.l., 1974, Effect of growth rate and
nutrient limitation on the composition and biomass yield of
Acinetobacter calcoaceticus, Appl. Microbiol., 28:58.
Aird, B.A., Heinrikson, R.L., and Westley, 1., 1987, Isolation and
characterization of a prokaryotic sulfurtransferase, 1. BioI.
Chem., 262:17327.
Ammon, H.L., Murphy, K.C., Sjolin, L., Wlodawer, A., Holcenberg, 1.S., and
Roberts, 1., 1983, The molecular symmetry of glutaminase-
asparaginases: rotation function studies of the Pseudomonas 7A
and Acinetobacter enzymes, Acta Cystallog., B39:250.
Ammon, H.L., Weber, LT., Wlodawer, A., Harrison, R.W., Gilliland, G.L.,
Murphy, K.C., Sjolin, L., and Roberts, 1., 1988, Preliminary crystal
studies of Acinetobacter glutaminasificans glutaminase-
asparaginase, 1. BioI. Chem., 263:150.
Bahk, Y.Y., and Kim, Y.S., 1987, Acidic isocitrate lyase from
Acinetobacter calcoaceticus grown on malonate, Han'guk
S aenghwa H akhaechi, 20:191.
Barker, 1., and Maxsted, H., 1975, Observations on the growth and movement
of Acinetobacter on semisolid media, 1. Med. Microbio!.,
8:443.
Barnes, E.M., and Thornley, MJ., 1966, The spoilage of eviscerated chickens
stored at different temperatures, 1. Food Technol., 1:113.
Baumann, P., 1968, Isolation of Acinetobacter from soil and water, 1.
Bacterio!',96:39.
Baumann, P., Doudoroff, M., and Stanier, R.Y., 1968, A study of the
Moraxella group II. Oxidase negative species (genus
Acinetabacter), 1. Bacterial., 95:1520.
15
Beardmore-Gray, M., and Anthony, c., 1986, The oxidation of glucose by
Acinetabacter calcaaceticus: interaction of the quinoprotein
glucose dehydrogenase with the electron transport chain, i. Gen.
Microbial., 132:1257.
Bergogne-Berezin, E., loly-Guillou, M.L., and Vieu, 1.F., 1987,
Epidemiology of nosocomial infections due to Acinetobacter
calcoaceticus, i.Hosp. Infect., 10:105.
Bhattacharyya, G., and Mandel, A., 1987, Volatization of mercury by a soil
bacterium Acinetobacter lwoffi GBl, Ind. i. Exp. BioI.,
25:347.
Bifulco, J.M., Shirey, J.J., and Bissonnette, G.K., 1989, Detection of
Acinetobacter spp. in rural drinking water supplies, Appl.
Environ. Micro bioi., 55:2214.
Blazek, R., and Benbough, J.E., 1981, Improvement in the persistence of
microbial asparaginase and glutaminase in the circulation of the
rat by chemical modifications, Biochim. Biophys. Acta, 677:220.
Blechschmidt, B., Borneleit, P., Emanoulidis, I., Lehman, W., and Kleber,
H.-P., 1989, Extracellular location of a B-Iactamase produced by
Acinetobacter calcoaceticus, Appl. Microbiol. Biotech.,
32:85.
Borneleit, P., and Kleber, H.-P., 1983, Purification and properties of the
membrane-bound NADH dehydrogenase from hydrocarbon-grown
Acinetobacter calcoaceticus, Biochim. Biophys. Acta,
722:94.
Bouvet, P.l.M., and Grimont, P.A.D., 1987, Identification and biotyping of
clinical isolates of Acinetobacter, Ann. Inst. Past.
Breuil, c., and Kushner, D.J., 1975, Lipase and esterase formation by
psychrophilic and mesophilic Acinetobacter species, Can. i.
M i crob ial., 21 :423.
Brisou, J., 1957, Contribution a l'etude des Pseudomonadaceae.
Precisions taxonomiques sur Ie genre Acinetabacter, Ann. Inst.
Past., 92:134.
Brisou, J., and Prevot, A.R., 1954, Etudes de systematique bacterienne. X.
Revision des especes reunies dans Ie genre Achramobacter,
Ann. Inst. Past., 86:722.
Chandhuri, J., Bhattacharyya, G., Ghosh, D.K., and MandaI, A., 1990,
Enzymatic reduction of mercuric chloride by cell-free extract of a
mercury-resistant strain of Acinetabacter lwoffi, Ind. i. Exp.
BioI., 28:138.
Collins, M.D., and Jones, D., 1981, Distribution of isoprenoid quinone
structural types in bacteria and their taxonomic implications,
Microbial. Rev., 45:316.
Cook, A.M., and Fewson, C.A., 1973, Role of carbohydrates in the
metabolism of Acinetabacter calcaaceticus, Biachim.
Biophys. Acta, 320:214.
Cunha, B.A., Klimek, J.J., Gracewski, J., McLaughlin, J.c., and Quintiliani,
R., 1980, A common source outbreak of Acinetobacter
pulmonary infections traced to Wright respirometers, Pastgrad.
M ed. i., 56:169.
16
Daly, A.K., Postic, B., and Kass, E.H., 1962, Infections due to organisms of
the genus H erellea, Arch. Intern. M ed., 110:580.
Dawson, N.L., and Reinhardt, D.l., 1981, Microbial flora of in-use, display
eye shadow testers and bacterial challenges of unused eye shadows,
Appl. Environ. Microbiol., 42:297.
D'Costa, E.J., Higgins, 1.J., and Turner, A.P.F., 1986, Quinoprotein glucose
dehydrogenase and its application in an amperometric glucose
sensor, Biosensors, 2:7l.
Debord, G.G., 1939, Organisms invalidating the diagnosis of gonorrhea by
the smear method, 1. Bacteriol., 38: 119.
Denis, F.A., D'Oultremont, P.A., Debacq, J.J., Cherel, J.M., and Brisou, J.,
1975, Distributions des ubiquinones (coenzyme Q) chez les
bacilles a Gram negatif, Compt. Rend. Seanc. Soc. Bioi.,
169:380.
Dickie, C. W., 1978, Equine myositis and septicemia caused by
Acinetobacter calcoaceticus infection, f. Amer. Vet. M ed.
Assoc., 57:530.
Dijkshoorn, L., van Vianen, W., Degener, J.E., and Michel, M.F., 1987,
Typing of Acinetobacter calcoaceticus strains isolated from
hospital patients by cell envelope protein profiles, Epidemiol.
Infect., 99:659.
Dokter, P., van Wielink, J.E., van Kleef, M.A.G., and Duine, J.A., 1988,
Cytochrome b-562 from Acinetobacter calcoaceticus LMD
79.41. Its characteristics and role as electron acceptor for
quinoprotein genome dehydrogenase, Biochem. 1., 254: 131.
Droop, M.R., and lannasch, H.W., 1977, Bacterial indication of water
pollution, Adv. Aquat. M icrobiol., 1:346.
Du Preez, J.e., and Toerien, D.F., 1978, The effect of temperature on the
growth of Acinetobacter calcoaceticus, Water SA, 4:10.
Du Preez, J.e., Toerien, D.P., and Lategan, P.M., 1981, Growth parameters
of Acinetobacter calcoaceticus on acetate and ethanol, Eur.
f. Appl. Microbiol. Biotech., 13:45.
Du Preez, J.e., Lategan, P.M., and Toerien, D.F., 1984, Influence of the
growth rate on the macromolecular composltlOn of
Acinetobacter calcoaceticus in carbon-limited chemostat
culture, FEMS Microbiol. Lett., 23:71.
Du Preez, J.e., Lategan, P.M., and Toerien, D.F., 1985, Utilization of short
chain monocarboxylic acids in an effluent of a petrochemical
industry by Acinetobacter cal coaceticus, Biotech. Bi oeng.,
27: 128.
Edman, P., Artursson, P., Bjork, E., and Davidson, B., 1987, Immobilized L-
asparaginase-L-glutaminase from Acinetobacter glutamin-
asificans in microspheres: some properties in vivo and in an
extracorporeal system, Int. 1. Pharmaceut., 34:225.
Ensley, B.D., and Finnerty, W.R., 1980, Influences of growth substrates and
oxygen on the electron transport system of Acinetobacter sp.
H01-N, 1. Bacteriol., 142:859.
Ensley, B.D., Irwin, R.M., Carreira, L.A., Hoffman, P.S., Morgan, P.S.,
Morgan, T.V., and Finnerty, W.R., 1981, Effects of growth
17
substrates and respiratory chain cOmpOSItIOn on bioenergetics in
Acinetobacter sp. strain HOI-N,.I. Bacteriol .. 14(l:508.
Eribo, B.E., and Jay, J.M., 1985, Incidence of Acinetobacter species and
other Gram-negative, oxidase-negative bacteria in fresh and
spoiled ground beef, Appl. Environ. M icrobiol., 49:256.
Fewson, C.A., 1985, Growth yields and respiratory efficiency of
Acinetobacter calcoaceticus, .f. Gen. M icrobiol., 131:865.
Fir,tenberg-Eden, R., Rowley, D.B., and Shattuck, G.E., 1980, Factors
affecting inactivation of Moraxella - Acinetobacter cells in an
irradiation process, Appl. Environ. M i c rob i 01., 40:480.
French, G.L., Casewell, M.W., Roncoroni, A.J., Knight, S., and Phillips, I.,
19(10, A hospital outbreak of antibiotic-resistant Acinetobacter
anitratus, epidemiology and control,.I. Hosp. infect., 1:125.
Fricke, B., and Aurich, H., I l)(ll), Characterization of a periplasmic insulin-
cleaving metalloproteinase from Acinetobacter calcoaceticus,
Biomed. Biochim. Acta, 9:661.
Fricke, B., Jahreis, G., Sorger, H., and Aurich, H., 19H7, Proteases in
different membrane fractions of Acinetobacter calcoaceticus,
.T. Basic Microbiol., 27:75.
Gardner, G.A., 1971, Microbiological and chemical changes in lean
Wiltshire bacon during aerobic storage, .f. App!. Bacteriol.,
34:465.
Geerlof. A., Dokter, P., van Wielink, J.E., and Duinc, l.A., 19H1), Haem-
containing protein complexes of Acinetobacter calcoaceticus
as secondary electron acceptors for quinoprotein glucose
dehydrogenase, Ant. v. Leeuw., 56:gl.
Gerner-Smidt, P., 19H9, Frequency of plasmids in strains of Acinetobacter
calcoaceticus,.T. Hosp. Infect., 14:23.
Gibson, J.A., and Eaves, L.E., 1981, Isolation of Acinetobacter
caleoaceticlls from an aborted equine foetus, Aust. Vet . .f.,
57:530.
Goosen, N., Vermaas, D.A., and van de Putte, P., 1997, Cloning of the gencs
involved in synthesis of coenzyme pyrroloquinolone-quinone from
Acinetobacter calcoaceticus, .I. Bacteriol., 16<):303.
Hardy, G.A., and Dawes, E.A., 19H5, Effect of oxygen concentration on the
growth and respiratory efficiency of Acinetobacter
calcoaceticus, .I.Cen. Microbiol., 131:855.
Harper, D.R., Clare, C.M., Heaps, C.D., Brennan, J., and Hussain, J., 1997, A
bacteriological technique for the development of latent
fingerprints, Forens. Sci. Internat., 33:209.
Harstein, A.l., Morthland, V.H., Rourke, J.W., Freeman, J., Garber, S.,
Sykes, R., and Rashad, A.L., 1990, Plasmid DNA fingerprinting of
Acinetobacter calcoaceticus subspecies anitratus from
intubated and mechanically ventilated patients, Infect. Control
Hosp. Epidemiol., 11:53l.
Henrichsen, J., 1975, The occurrence of twitching motility among Gram-
negative bacteria, Acta Path. Microbiol. Scand., B83:171.
Henriksen, S.D., 1973, Moraxella, Acinetobacter and the Mimiae,
Bacteriol. Rev., 37:522.
18
Henriksen, S.D., 1976, Moraxella, Neisseria. BranhameLla and
Acinetobacter, Ann. Rev. Microbiol., 30:63.
Hoffman, S., Mabeck, C.E., and Vejsgaard, R., 1982, Bacteriuria caused by
Acinetobacter calcoaceticus biovars in a normal population
and in general practice, J. Clin. Microbiol .. 16:443.
Holcenberg, 1.S., 1985, Glutaminase from Acinetobacter glutaminas-
ificans, Meth. Enzymol., 113:257.
Holcenberg, 1.S., Schmer, G., Teller, D.C., and Roberts, 1., 1975, Biologic
and physical properties of succinylated and glycosylated
Acinetobacter glutaminase-asparaginase, .I. BioI. Chem.,
250:4165.
Holcenberg, 1.S., Ericsson, L., and Roberts, 1., 1978, Amino acid sequence of
the diazooxonorleucine binding site of Acinetobacter and
Pseudomonas 7A glutaminase-asparaginase enzymes,
Biochemistry, 17:411.
Holcenberg, 1.S., Camitta, B.M., Borella, L.D., and Ring, B.J., 1979, Phase 1
study of succinylated Acinetobacter L-glutaminase-L-
asparaginase, Cancer Treatment Reports, 63: 1025.
Holton, J., 1983, A note on the preparation and use of a selective differential
medium for the isolation of Acinetobacter spp. from clinical
sources, J. Appl. Bacteriol., 54:141.
Hunger, M., Schmucker, R., Kishan, V., and Hillen, W., 1990, Analysis and
nucleotide sequence of an origin of DNA replication in
Acinetobacter calcoaceticus and its use for Escherichia coli
shuttle vectors, Gene, 87:45.
Ingram, M., and Shewan, J. W., 1960, Introductory reflections on the
Pseudomonas-Achromobacter group, .I. Appl. Bacteriol.,
23:373.
Ino, T., and Nishimura, Y., 1990, Radiation sensitivity of Acinetobacter
species, J. Gen. Appl. Microbiol., 36:55.
lahreis, G., Sorger, H., and Aurich, H., 1989, Eine membrangebundene
Alaninaminopeptidase aus Acinetobacter calcoaceticLis. I.
Isolierung und Reinigung des Enzymes, Biomed. Biochim. Acta,
48:617.
lay, J.M., 1982, Modern Food Microbiology, 3rd edn., Van Nostrand, New
York.
loly-Guillou, M.L., Bergogne-Berezin, E., and Vieu, J.F., 1990a, A study of
the relationships between antibiotic resistance phenotypes, phage-
typing and biotyping of 117 clinical isolates of Acinetobacter
spp.,J. Hosp.infect., 16:49.
loly-Guillou, M.L., Bergogne-Berezin, E., and Vieu, J.F., 1990b,
Epidemiologie et resistance aux antibiotiques des Acinetobacter
en milieu hospitalier. Bilan de 5 annees, Press. M ed., 19:357.
Joner, P.E., 1976, Purification and properties of L-asparaginase B from
Acinetobacter calcoaceticus, Biochim. Biophys. Acta,
438:287.
loner, P.E., Kristiansen, T., and Einarsson, M., 1973, Purification and
properties of L-asparaginase A from Acinetobacter
calcoaceticus, Biochim. Biophys. Acta, 327:146.
19
Jones, G.L., Jansen, F., and McKay, A.J., 1973. Substrate inhibition of the
growth of bacterium NCIB8250 by phenol, J. Gen. Microbiol.,
74:139.
Jones, C.W., Brice, J.M., and Edwards, c., 1977a, The effect of respiratory
chain composition on the growth efficiencies of aerobic bacteria,
Arch. Microbio!., 115:85.
Jones, M., Fernandes, R., and King, H.K., 1977b, Solubilization studies and
properties of particulate NAD-independent malate dehydrogenase
of Acinetobacter calcoaceticus (NCIB8250), Biochem. Soc.
Trans., 5:258.
Jongejan, J.A., and Duine, J.A., eds., 1989, PQQ and Quinoproteins, Kluger,
Dordrecht.
Juni, E., 1972, Interspecies transformation of Acinetobacter: genetic
evidence for a ubiquitous genus,.I. Bacteriol., 112:917.
Juni, E., 1978, Genetics and physiology of Acinetobacter, Ann. Rev.
Microbiol., 32:349.
Juni, E., 1984, Genus III. Acinetobacter. Brisou and Prevot 1954, in
"Bergey's Manual of Systematic Bacteriology, vol. 1," p.303,
N.R.Grieg and J.G.Holt, eds., Williams and Wilkins, Baltimore.
Kien, c.L., Anderson, A.J., and Holcenberg, J .S., 1985, Tissue nitrogen-
sparing effect of high protein diet in mice with or without ascites
tumour treated with Acinetobacter glutaminase-asparaginase,
Cancer Res., 45:4876.
Kim, S.J., and Kim, Y.S., 1985, Isolation of a malonate-utilizing
Acinetobacter calcoaceticus from soil, M isaengmul
H akhoechi, 23:230.
Kim, Y.J., and Kim, Y.M., 1989, Induction of carbon monoxide dehydrog-
enase during heterotrophic growth of Acinetobacter sp. strain
JCl DSM3803 in the presence of carbon monoxide, F EM S
Microbiol. Lett., 59:207.
Kim, Y.S., and Park, c., 1988, Inactivation of Acinetobacter
calcoaceticus acetate kinase by diethylpyrocarbonate, Biochim.
Kim, K.S., Ro, Y.T., and Kim, Y.M., 1989, Purification and some properties
of carbon monoxide dehydrogenase from Acinetobacter sp.
strain JCl DSM3803,.I. Bacteriol., 171:958.
Kita, K., Hiraoka, N., Oshima, A., Kadonishi, S., and Obayashi, A., 1985,
AccII, a new restriction endonuclease from Acinetobacter
calcoaceticus, Nucl. Acids Res., 13:8685.
Kitagawa, K., Tateishi, A., Nakano, F., Matumoto, 1., Morohoshi, '1'., Tanino,
T., and Osui, T., 1986a, Generation of energy coupled with
membrane-bound glucose dehydrogenase in Acinetobacter
calcoaceticus, Agric. Bioi. Chem., 50:1453.
Kitagawa, K., Tateishi, A., Nakano, F., Matsumoto, T., Morohoshi, T.,
Tanino, T., and Osui, T., 1986b, Sources of energy and energy
coupling reactions of the active transport systems in
Acinetobacter calcoaceticus, Agric. Bioi. Chem., 50:2939.
Kleber, H.-P., Frikke, B., Klossek, P., and Aurich, H., 1975, Effect of growing
conditions on protein content and amino acid composition of
Acinetobacter calcoaceticus, Ukran. Biochem. J., 47:422.
20
Koburger, S.A., 1964, Isolation of Mima polymorpha from dairy products,
1. Dairy Sci., 47:646.
Koh, J-S., Yamakawa, T., Kodama, T., and Minoda, Y., 1985, Rapid and
dense culture of Acinetobacter calcoaceticus on palm oil,
Agric. Bioi. Chem., 49:1411.
Koh, J-S., Yamakawa, T., Kodama, T., and Minoda, Y., 1987, Cell mass
production of Acinetabacter calcoaceticus in palm oil-limited
chemostat culture, 1. Ferment. Tech., 65:229.
Kotohda, S., Suzuki, F., Katsuki, S., and Sato, T., 1979, Purification and some
properties of endo-B-1,6-gluconase from Acinetabacter sp.,
Agric. Bioi. Chem., 43:2029.
LaCroix, S.J., and Cabelli, V.J., 1982, Membrane filter method for
enumeration of Acinetobacter calcoaceticus from
environmental waters, Appl. Environ. Microbial., 43:90.
Layh, G., Eberspaecher, J., and Lingens, F., 1982, Rhodanase in chloridazon-
degrading bacteria, FEMS Microbial. Lett., 15:23.
Liaw, H.J., and Srinivasan, V.R., 1990, Expression of an Erwini a sp. gene
encoding diphenyl ether cleavage in Escherichia coli and an
isolated Acinetobacter strain PE 7, Appl. Microbiol.
Biotech., 32:686.
Lim, K.J., and Kim, Y.S., 1987, Malonate uptake by Acinetobacter
calcoaceticus, Han 'guk S aengh wa H akhoechi, 20:286.
Loeffelholz, M.J., and Modrzakowski, M.C., 1988. Antimicrobial
mechanisms against Acinetobacter calcoaceticus of rat
polymorphonuclear leukocyte granule extracts, infect. Immun.,
56:552.
Ludewig, M., Fricke, B., and Aurich, H .• 1987, Leucine aminopeptidase in
intracytoplasmic membranes of Acinetobacter calcaaceticus,
1. Basic Microbiol., 27:557.
Lutz, A., Grooten, 0., Velu, H., and Velu, M., 1956, Role pathogene et
frequence des bacteries du type Bacterium anitratum, Ann.
inst. Past., 91:110.
Mayer, K.H., and Zinner, S.H., 1985, Bacterial pathogens of increasing
significance in hospital-acquired infections, Rev. Infect., Dis.,
7(suppl. 3):S371.
Meyer, D.l., and Jones, C.W., 1973, Distribution of cytochromes in bacteria:
relationship to general physiology, Int. J. Syst. Bacterial.,
23:459.
Midgley, M., Iscandar, N.S., and Dawes, E.A., 1986a, The interaction of
phosphonium ions with Acinetabacter calcoaceticus: evidence
for the operation of an efflux system, Biochim. Biophys. Acta,
856:45.
Midgley, M., Iscandar. N.S., Parish, M., and Dawes, E.A., 1986b, A
fluorescent probe for a phosphonium ion efflux system in bacteria,
FEMS Microbiol. Lett., 34:187.
Monboisse, J-c., Labadie, J., and Gouet, P., 1979, Induction et repression de
la synthase de collagenase chez Acinetabacter sp., Can. 1.
Microbiol., 25:611.
Mukherji, S., and Bhopale, N., 1983, Gliding motility of Acinetobacter
anitratus, 1. Clin. Pathol., 36:484.
21
Mukkada, A.J., and Bell, E.J., 1971, Partial purification and properties of the
fructose-1,6-diphosphatase from Acinetobacter lwoffi, Arch.
Biochem. Biophys., 142:22.
Miiller, R.H., and Babel, W., 1986, Glucose as an energy donor in acetate
growing Acinetobacter calcoaceticus, Arch. Microbiol.,
144:62.
Muller-Serieys, c., Lesquoy, J.B., Perez, E., Fichelle, A., Boujeois, B., Joly-
Guillou, M.L., and Bergogne-Berezin, E., 1989, Infections
nosocomiales aAcinetobacter, Press. Med., 18:107.
Neidleman, S.L., and Geigert, J., 1984, Biotechnology and oleochemicals:
changing patterns, 1. Amer. Oil Chem. Soc., 61:290.
Nishimura, Y., Kanbe, K., and Iizuka, H., 1986, Taxonomic studies of aerobic
coccobacilli from seawater, 1. Gen. Appl. Microbiol., 32:1.
Nishimura, Y., Ino, T., and lizuka, H., 1988a, Acinetobacter
radioresistans sp. nov. isolated from cotton and soil, Int. 1.
Syst. Bacteriol., 38:209.
Nishimura, Y., Kanzaki, H., and lizuka, H., 1988b, Taxonomic studies of
Acinetobacter species based on the electrophoretic analysis of
enzymes, 1. Basic M icrobiol., 6:363.
Obana, Y., 1986, Pathogenic significance of A. calcoaceticus. Analysis of
experimental infection in mice, M icrob iol. immunol., 30:645.
Obana, Y., Nishino, T., and Tanino, T., 1985, In vitro and in vivo activities
of antimicrobial agents against Acinetobacter calcoaceticus,
1. Antimicrob. Chemother., 15:441.
Onishi, H., and Hidaka, 0., 1978, Purification and properties of amylase
produced by a moderately halophilic Acinetobacter sp., Can. 1.
Microbiol., 24:1017.
Ph!chaud, D., Piechaud, M., and Second, L., 1956, Varietes proteolytiques de
Moraxella lwoffi et de Moraxella glucidolytica, Ann. Inst.
Past., 90:517.
Ramachandra, R.N., Rao, M.S., Raghavan, R., and Keshavamurthy, B.S.,
1981, Isolation of Acinetobacter calcoaceticus from extended
frozen buffalo semen used for artificial insemination, Curro Sci.,
50:1066.
Retailliau, H.F., Hightower, W.A., Dixon, R.E., and Allen, J.R., 1979,
Acinetobacter calcoaceticus: a nosocomial pathogen with an
unusual seasonal pattern, 1. Infect. Dis., 139:371.
Richter, K., and Kleber, H.-P., 1980, Regulation of phosphoenolpyruvate
carboxylase from Acinetobacter calcoaceticus by nucleotides,
FEBS Lett., 109:202.
Rosenberg, M., and Rosenberg, E., 1981, Role of adherence in growth of
Acinetobacter calcoaceticus RAGI on hexadecane, 1.
Bacteriol., 148:51.
Shaub, I.G., and Hauber, F.D., 1948, A biochemical and serological study of
a group of identical unidentifiable Gram-negative bacilli in human
sources, 1. Bacteriol., 56:379.
Shewan, J.W., Hobbs, G., and Hodgkiss, W., 1960, The Pseudomonas and
Achromobacter groups of bacteria in the spoilage of marine
white fish, 1. Appl. Bacteriol., 23:468.
22
Skerman, V.B.D., McGowan, V., and Sneath, P.A., eds., 1980, Approved lists
of bacterial names, Int. 1. Syst. Bacterio!., 30:225.
Smith, A.W., Freeman, S., Minett, W.G., and Lambert, P.A., 1990,
Characterization of a siderophore from Acinetobacter
calcoaceticus, FEMS Microbiol. Lett., 70:29.
Steckel, J., Roberts, J., Phillips, F.S., and Chou, T-C., 1983, Kinetic
properties and inhibition of Acinetobacter glutaminase-
asparaginase, Biochem. Pharmaco!., 32:971.
Tanaka, S., Robinson, E.A., Appella, E., Miller, M., Ammon, R.L., Roberts,
J., Weber, LT., and Wlodawer, A., 1988, Structures of
amidohydrolases. Amino acid sequence of a glutaminase-
asparaginase from Acinetobacter glutaminasificans and
preliminary crystallographic data for an asparaginase from
Erwinia chrysanthemi, 1. Bioi. Chem., 263:8583.
Thornley, M.J., Ingram, M., and Barnes, E.M., 1960, The effects of
antibiotics and irradiation on the Pseudomonas-
Achromobacter flora of chilled poultry, 1. Appl. Bacteriol.,
23:487.
Tjernberg, 1., and Vrsing, J., 1989, Clinical strains of Acinetobacter
classified by DNA-DNA hybridization, APM I S, 1987:595.
Towner, K.J., 1978, Chromosome mapping in Acinetobacter
cal coaceticus, J. Gen. M icrob io!., 104: 175.
van Schie, B.J., Hellingwerf, K.J., van Dijken, J.P., Elferinck, M.G.L., van
Dijl, J.M., Kuenen, J.G., and Konings, W.N., 1985, Energy
transduction by electron transfer via a pyrrolo-quinoline quinone
dependent glucose dehydrogenase in Escherichia coli,
Pseudomonas aeruginosa and Acinetobacter calcoaceticus
(var. lwoffi), J. Bacteriol., 163:493.
van Schie, B.J., Rouwenhorst, R.J., de Bont, J.A.M., van Dijken, J.P., and
Kuenen, J.G., 1987, An in vivo analysis of the energetics of aldose
oxidation by Acinetobacter calcoaceticus, Appl. Microbiol.
Biotech., 26:560.
Vandenbergh, P.A., and Berk, R.S., 1980, Purification and characterization
of rhodenase from Acinetobacter calcoaceticus, Can. J.
Microbiol., 26:281.
Vieu, J.F., Bergogne-BeH!zin, E., Joly, M.L., Berthelot, G., and Fichelle, A.,
1980, Epidemiologie d'Acinetobacter calcoaceticus, Press.
Med.,9:355l.
Villalobo, A., Roldan, J.M., Rivas, J., and Cardenas, J., 1977, Assimilatory
nitrate reductase from Acinetobacter calcoaceticus, Arch.
Von Graevenitz, A., 1983, Acinetobacter calcoaceticus var. anitratus:
okologie, isolierung und Pathogenitat, in "Neus Von Alten
Erregen und nene Erreges," p.112, C.Kraseman, ed., de Gruyter,
Berlin.
Whittaker, P.A., 1971, Terminal respiration in Moraxella lwoffi
(NCIB8250), Microbios, 4:65.
Williams, R.N., Falkler, W.A., and Hasler, J.F., 1983,
Acinetobacter contamination of laboratory dental pumice, 1.
Dent. Res., 62:1073.
23
Wlodawer, A., Roberts, J., and Holcenberg, J.S., 1977, Characterization of
crystals of L-glutaminase-asparaginase from Acinetobacter
glutaminasificans and Pseudomonas 7A, 1. M oZ. BioI.,
112:515.
Wong, T.Y., Graham, L., O'Hara, E., and Maier, R.J., 1986, Enrichment for
hydrogen-oxidising Acinetobacter spp. in the rhizosphere of
hydrogen-evolving soybean root nodules, Appl. Environ.
Yashphe, J., Chikarmane, H., Iranzo, M., and Halvorson, H.O., 1990,
Phosphatases of Acinetabacter lwaffi. Localization and
regulation of synthesis by orthophosphate, Curro Microbial.,
20:273.
24
TAXONOMY OF ACINETOBACTER
P.A.D. Grimont and P.J.M. Bouvet
Unite des Enterobacteries

INSERM Unite 199
Institut Pasteur
75724 Paris Cedex 15
France
INTRODUCTION
The history of the oxidase-negative, non-motile, Gram-negative

diplobacilli now constituting the genus Acinetobacter has been confused
for many years (Henriksen, 1973, 1976). These bacteria were originally
classified into various genera including "Bacterium ", Neisseria,
Alcaligenes, "Mima ", "H erellea ", "Achromobacter" and M oraxella.
For some time this group of bacteria was referred to as the oxidase-negative
M oraxell a. The application of modern methods of taxonomy, including
carbon source utilisation tests, genetic transformation, DNA hybridisation,
and rRNA sequence comparison (by hybridisation or direct sequencing), has
now resolved the earlier confusion. Progress in Acinetobacter taxonomy
has now provided a new insight into Acinetobacter ecology and
epidemiology.
DELINEATION OF THE GENUS ACINETOBACTER
The nutritional studies of Baumann et al. (1968) clearly showed that

oxidase-negative strains differed from oxidase-positive strains of the genus
Moraxella. Baumann et al. (1968), in search of a generic name, exhumed
the name Acinetobacter, originally coined by Brisou and Prevot (1954), to
accommodate non-motile "Achromobacter" strains. DNA from oxidase-
negative strains of Moraxella (Acinetobacter spp.) was found to
transform a competent strain (strain BD413trpE27) of that group,whereas
DNA from oxidase-positive strains failed to transform this strain (Juni,
1972). Acinetobacter strains were therefore recognised as a natural group
25
distinct from the genera M oraxella and Neisseria. The current definition
of the genus Acinetobacter is as follows (adapted from Juni, 19R4): strictly
aerobic, non-motile, oxidase-negative coccobacilli; Gram-negative, but
sometimes difficult to destain; grows well on complex media between 20 and
30°C without growth factor requirements; nitrates arc rarely reduced;
cxtracted DNA is able to transform strain BD413trpE27; the G + C content
of the DNA is between 39 and 47 mol %.
The genus Acinetobacter constitutes a discrete phylogenetic branch

in superfamily II (Van Landschoot et al., 1986) of the Proteobacteria. The
closest genera are Moraxella, Pseudomonas (fluorescent group), and
X anthomonas (Woese et al., 1985; Van Landschoot et al., 1(86).
DELINEATION OF SPECIES IN THE GENUS ACINETOBACTER
Only two species, A. calcoaceticus and A. twoffii were listed in the

Approved Lists of bacterial names (Skerman, 19RO), and only one species was
given in Bergey's Manual of Systematic Bacteriology (Juni, 19R4), although
early DNA hybridisation experiments with a nitroccllulose filter method had
shown the genus Acinetohacter to be heterogeneous (Johnson et al., 1970).
Using a cut-off value of 50% relatedness to delineate DNA groups, Johnson
et al. (1970) found five such groups in the collection of strains they studied.
There is presently a general agreement to define a genomic species as

a group of strains which are at least 70% related by DNA hybridisation with
divergence values (!.H'm) below 5°C (Wayne et al., 1987). Genomic species
which can be identified with phenotypic tests can be given a formal species
name (Wayne et al., 1987). Using the above criteria to interpret the results
of DNA relatedness data, Bouvet and Grimont (19R6) delineatcd 12 gcnomic
species (numbered 1 to 12) among 85 strains, although the separation
between species 8 and 9 was uncertain. A total of ten genomic species could
be identified by phenotypic tests. The type strains of A. calcoaceticus and
A. twofIii fell in genomic species 1 and R respectively. Since a reference
strain of "Achromobac:ter haemolyticus" (Stenzel and Mannheim, 19(3)
fell in genomic species 4, the epithet was revived in a new combination,
Acinetobacter haemolyticus. Genomic species 2, 5 and 7 were named A.
baumannii, A. junii and A. johnsonii respectively. The other genomic
species, which contained less than ten strains or which could not be
identified unambiguously using phenotypic tests, were not named. This work
was later expanded to include 27 proteolytic strains which did not fit into any
of the 12 genomic species described prcviously (Bouvet and Jeanjean, 1989).
Five additional genomic species were delineated and numbered 13 to 17. In
spite of these additional efforts, several strains remained ungrouped.
Independent work, carried out by Tjernberg and Ursing (1989)

(abbreviated as "T&U" study) on 168 consecutive clinical isolates and 30
reference strains, showed a good correlation with the results of Bouvet and
Grimont (1986) (abbreviated as "B&G" study). Ten genomic species were
26
common to both studies. Three additional genomic species (numbered 13 to
15) were described in the T&U study. Genomic species 13 (T&U) was
represented by a single strain in the B&G study. Group 14 (T&U) contained
a reference strain of "Achromobacter conjunctivae" (Stenzel and
Mannheim, 1963) which was ungrouped in the B&G study. Genomic species
14 (T&U) corresponded to group 13 of Bouvet and Jeanjean (1989).
Genomic species 15 (T&U) was not represented in the studies of Bouvet and
Grimont (1986) and Bouvet and Jeanjean (1989). Thus, the data obtained in
two different laboratories were fully compatible.
Nishimura et al. (1987) delineated five DNA relatedness groups

among 31 strains. The reference strains included in the study of Nishimura
et al. (1987) allowed them to identify groups 2 and 3 as A. twoffii and A.
johnsonii, respectively. Group 4 corresponded to both A. haemolyticus
and genomic species 6 (B&G). However, a careful examination of the data
from Nishimura et al. (1987) shows that the reference strain of genomic
species 6 was only 65% related by the filter hybridisation method to the
reference strain of A. haemolyticus. It has been shown (Bouvet and
Grimont, 1986) that a DNA relatedness value of 65% obtained by the filter
method corresponds to about 44% relatedness as determined by the S1
nuclease method (Le. the method used by Bouvet and Grimont, 1986). The
data obtained in Tokyo and Paris are thus not contradictory and only the
interpretation differs. A more serious discrepancy occurred with group 1
(Nishimura et al., 1987), which corresponded to both A. calcoaceticus and
A. baumannii. The type strains of both species showed 98% relatedness
(Nishimura et al., 1987). When strains from both laboratories were
exchanged, the strain used as the type strain of A. calcoaceticus in Tokyo
(lAM 12087) was found to be a typical A. baumannii, both by phenotypic
tests and DNA relatedness. Group 5 contained radiation-resistant
Acinetobacter strains isolated from soil and cotton samples (Nishimura et
al., 1987). This group was later named A. radioresistens (Nishimura et
al., 1988a). Tjernberg and Ursing (1989) showed that A. radioresistens
corresponded with B&G genomic species 12. Although this species was
poorly represented in the B&G study, it represented the second largest group
(31 isolates from pathological samples) in the T&U study. Radiation
resistance should be studied in clinical isolates contained in this group to
determine whether this property can define the species.
Table 1 compares and summarises the species designations and DNA

groups delineated in the different laboratories.
CHARACTERISATION OF ACJNETOBACTER SPECIES
Protein Electrophoresis Patterns
Polyacrylamide gel electrophoresis (PAGE) of whole cell proteins is

increasingly used for the characterisation of bacteria. Whole cell protein
electrophoresis of Acinetobacter strains yields a number of patterns
27
N
00
Table 1. A comparison of the DNA groups delineated in different laboratories.
Reference strain DNA relatedness group according to Present

nomenclature
Bouvet and Grimont,1986; Tjernberg and Nishimura et al.,
Bouvet and Jeanjean,1989 Ursing, 1989 1987, 1988a
ATCC 230~5t;IAM120~7t 1 1 A.calcoaceticus

CIP70.34 ;IAM12088 2 2 1 A. baumannii
CIP70.29;ATCC19004 3 3 nt Not named
CIP70.11;ATCC17903 ug 13 nt Not named
CIP64.3 t ;ATCC17906 t 4 4 4 A.haemolyticus
CIP64.5t;ATCC1790S t 5 5 nt A.junii
CIPA165;ATCC17979 6 6 4 Not named
CIP64.6 t =ATCC17909 t 7 7 3 A. johnsonii
ATCC15309 t =CIP64.10 t S 8 2 A.lwoffii
CIP70.31=ATCC9957 9 8 nt Not named
CIP70.12;ATCC17924 10 10 ug Not named
CIP63.46=NCIB8250
11 11 ug Not named
;ATCCl1171
strain FO-1 t =IAM13186 t (12)a 12 5 A.radioresistens
CIP64.2=ATCC17905 13 14 nt Not named
Bouvet 382 14 nl nt Not named
Bouvet 240 15 nt nt Not named
CIP70.18=ATCC17988 16 ug nt Not named
Bouvet 942 17 nt nt Not named
Tjernberg ISla nt 15 nt Not named
aunpublished result; ttype strain.

nt, not tested; ug, ungrouped.
(Alexander et aI., 1984, 1988). Unfortunately these data have not yet been
integrated in the newer taxonomic scheme outlined above and it is not yet
known whether whole cell protein patterns can differentiate the species
delineated by DNA hybridisation.
Electrophoresis of outer membrane proteins was applied to the strains

previously characterised with DNA hybridisation by Ino and Nishimura
(1989). Patterns generally contained 20-25 distinct protein bands including
one or two major proteins with an apparent molecular size of 20,000-50,000.
Correspondence between outer membrane protein patterns and some DNA
relatedness groups was fairly good (Ino and Nishimura, 1989).
Cell envelope protein patterns of a number of strains were determined

by Oijkshoorn et aI. (1987a,b). These strains were later identified using the
latest taxonomic scheme (Bouvet et aI., 1990). With the exception of A.
haemolyticus (E patterns) and Acinetobacter sp. 3 (0 patterns), in which
strains showed related patterns within a species, all other studied species
showed a broad range of patterns. Cell envelope protein patterns may thus
be more useful for typing isolates than for species identification.
Enzyme Electrophoresis
Nishimura et aI. (1988b) used PAGE to examine 22 strains from

culture collections and 14 isolates from soil and cotton samples for the
presence of twelve enzymes. The strains divided into four clusters (Z-1, Z-2,
Z-3 and Z-4). Cluster Z-l further divided into three subclusters (Z-la, Z-
1b, Z-1c). Cluster Z-1 included four different Acinetobacter spp. (A.
calcoaceticus, A. baumannii, A. haemolyticus, and Acinetobacter sp.
6); cluster Z-2 included two species (A. johnsonii and Acinetobacter sp.
11); cluster Z-3 corresponded to A. radioresistens (equivalent to B&G
Acinetobacter sp. 12); cluster Z-4 corresponded to A. lwoffii.
Picard et aI. (1989) examined the electrophoretic polymorphism of

malate and glutamate dehydrogenase and diverse esterases produced by
Acinetobacter strains belonging to known genomic species. Following
treatment of the data by correspondence analysis (a type of factor analysis),
the results were found to correlate with the distribution of strains into
genomic species.
Serological Cross-Reactions
Traub (1989) has developed a taxonomic scheme for typing A.

baumannii, the most prevalent species in hospitals. Twenty serovars have
been delineated. With one exception, the sera directed against the 20
serovars of A. baumannii did not cross-react with seven serovars of
Acinetobacter sp.3. This supports the separation of sp. 3 from A.
baumannii, which otherwise can only be separated by the upper limit to its
growth temperature (see below).
29
rRNA Gene Restriction Patterns
When total bacterial DNA is cleaved by a restriction endonuclease and

the DNA fragments separated by agarose gel electrophoresis, a large number
of DNA fragments can be visualised after staining with ethidium bromide.
Although restriction patterns obtained from two strains can be directly
compared, the number of fragments is too large for a pattern to be recorded
for later comparison. If the DNA fragments in the agarose gel are
transferred to a nylon membrane and then hybridised with a universal probe
containing labelled Escherichia coli 16+23S rRNA, a simpler pattern of
hybridised fragments (rRNA gene restriction pattern) can be visualised
(Grimont and Grimont, 1986). Depending upon the bacterial group studied
and, to a lesser extent, upon the restriction endonuclease chosen, the level of
identification provided by the method is at or below the species level. This
method is currently being applied to Acinetobacter strains (F. Grimont, S.
Jeanjean, P. Bouvet, and P.A.D. Grimont, manuscript in preparation). It has
been found that rRNA gene restriction patterns generated by endonuclease
C I aI contain between four and 14 fragments. A single pattern was observed
with four strains of genomic species 3. A total of 13 different patterns were
given by 20 A. baumannii strains. In other genomic species it was found
that almost every strain gave a unique pattern.
Physiological and Biochemical Characterisation
Optimal temperature for growth is around 30°C; however, the upper

growth temperature differs between species. A. johnsonii strains cannot
grow at 37°C, whereas only strains of A. baumannii can grow at 44°C (Table
2).
The ability to produce acid from glucose has long been used to
differentiate A. calcoaceticus sensu lato from A. Iwoffii sensu lato.
Acid production with D-glucose results from the oxidation of glucose to
gluconic acid (Aubert et aI., 1952). The glucose dehydrogenase involved is
actually a non-specific aldose dehydrogenase that can also oxidise a number
of different aldoses (Hauge, 1960). The phenons delineated by Baumann et
al. (1968) often contained both glucose-oxidisers and non-oxidisers. Duine
et al. (1979) demonstrated that glucose dehydrogenase was a quinoprotein.
Glucose dehydrogenase is an inactive apoenzyme and pyrroloquinoline
quinone is a co-factor required for enzymic activity. Strains producing the
apoenzyme, but not the co-factor, will not oxidise glucose unless the co-
factor is added to the medium (Duine, this volume). Bouvet and Bouvet
(1989) found that the following percentages of strains (in parentheses)
produced acid from glucose without added pyrroloquinoline quinone: A.
calcoaceticus (100), A. baumannii (99), sp. 3 (100), A. haemolyticus
(63), sp. 6 (50), A. Iwoffii (5), sp. 10 (100), and sp. 12 (40). No strains of A.
junii, A. johnsonii or sp. 11 produced acid from glucose under these
conditions. When pyrroloquinoline quinone was added, all genomic species
contained strains producing acid from glucose (percentages of strains in
parentheses): A. baumannii (100), A. haemolyticus (69), A. junii (72),
30
Table 2. Current identification scheme for AcinetobBcter genomic species
DNA relatedness groupa

2 3 4 5 6 7 8/9 10 11 12 13 14 15 16 17
Growth at: 44°C +

4PC + + 90
3PC + + + + + + + + + + 89 + + + 75
Acid from D-glucose + 95 + 60 50 6 + 40 + +
Gelatin hydrolysis 96 + + + + + +
Utilisation of:
DL-Lactate + + + + + 99 + + + + + + + +
DL-4-Aminobutyrate + + + + 90 35 40 + + + 11 + 25 +
trans-Aconitate + 99 + 52 11 67 50
Citrate + + + 90 82 + 98 + + + + + + +
Glutarate + + + + + + + +
Aspartate + + + 64 40 66 61 + + +
Azelate + 90 + + 50 25 + +
/3-Alanine + 95 95 + + + 75 +
L-Histidine + 98 94 96 + + + + + + + + +
D-Malate 98 + 96 + 66 20 60 + + + + + + +
Malonate + 98 85 15 + 11 + 50 50
Histamine 75 +
L-Pheny1alanine + 87 66 + + + + + +
Phenylacetate + 87 66 94 25 50 + + + + + +
anumbering according to Bouvet and Grimont (1986); Bouvet and Jeanjean (1989). Named species are:
1, A. cal coacet icus ; 2, A. baumannii; 4, A. haemolyticus ; 5, A. junii; 7, A. johnsonii; 8/9, A. lwoffii;
12. A. radioresistens.
+ ~ all strains positive; all strains negative. Numbers are percentages of strains positive for a
particular character.
w
sp. 6 (SO), A. johnsonii (97), A. lwoffii (90), sp. 11 (100), and sp. 12 (l00)
(Bouvet and Bouvet, 1989). Since glucose does not enter Acinetobacter
cells, only the rare strains which can both utilise gluconate and oxidise
glucose can utilise glucose in a minimal medium (Baumann et aI., 1968;
Bouvet and Grimont, 1986). The taxonomic significance of acid production
from carbohydrates in Acinetobacter taxonomy is thus of less importance
than was once thought.
The nutritional abilities (growth in a defined medium with ammonia as

nitrogen source and a given carbon source) of strains are essential properties
to consider in studying the taxonomy of the genus Acinetobacter (Baumann
et ai. 1968; Bouvet and Grimont, 1986). A simplified identification scheme
for named and unnamed Acinetobacter species has been published (Bouvet
and Grimont, 1987) and is updated in Table 2.
CLINICAL AND ECOLOGICAL SIGNIFICANCE OF NEW TAXONOMIC

CHANGES
Progress in taxonomy does not always meet with warm acceptance from
those who use identification schemes or simply refer to bacteria by scientific
names in their routine work. It is thus important for the taxonomist to
indicate how a new taxonomic scheme can generate new concepts in the
users' fields. It seems that different Ac i ne t 0 b act e r species may have
different habitats, and these are outlined below. More complete accounts of
the occurrence of the different Acinetobacter species in clinical situations
have been given elsewhere (Bergogne-Berezin and Joly-Guillou, this
volume) and the interested reader should refer to this review for full details.
The habitat of A. calcoaceticus sensu stricto is where Beijerinck

originally found this bacterium, i.e. in the soil (Beijerinck, 1911). This
species is very easy to isolate from soil by enrichment using a minimal liquid
medium with ammonium ions as nitrogen source, acetate as carbon source,
and aerobic conditions (Baumann et aI., 1968). A. calcoaceticus is seldom
associated with human infection; indeed, only three cases, from wounds or
sputum, are known to us (Tjernberg and Ursing, 1989), and it is noteworthy
that these three isolates were somewhat divergent from the type strain,
showing T m values ranging from 3.0 to S.SoC (rounded to the nearest O.S°C)
in hybridisation experiments (Tjernberg and Ursing, 1989); such a level of
divergence can indicate a relationship at the sub-specific level (Grimont,
1988).
A. baumannii has, to date, only been isolated from humans. It is the

species of Acinetobacter most commonly associated with nosocomial
infections (Bergogne-Berezin and Joly-Guillou, this volume). Of 291
isolates recovered from 264 patients, this species comprised 244 isolates
32
(84%) (Bouvet and Grimont, 1987). No habitat, other than man, is known
for this species. Thus, its presence on inanimate objects in the hospital
environment (Bouvet and Grimont, 1987) can be interpreted as
contamination from an infected patient. Since A. calcoaceticus and A.
baumannii are similar in many metabolic properties, biotechnological
applications should use A. calcoaceticus rather than A. baumannii
wherever possible.
Acinetobacter sp. 3 has been found in soil and clinical samples.

Although infrequent in French studies (Bouvet and Grimont, 1987), this
species constituted the largest group of clinical isolates studied by Tjernberg
and Ursing (1989).
A. haemolyticus has been isolated occasionally from patients (about

3% of total Acinetobacter isolates), the hospital environment, and
activated sludge.
A. junii has been isolated both from clinical samples and the
environment.
A. j ohnsonii has been isolated from the environment (soil, activated

sludge, river sediment), animals or animal products (chicken, raw milk), the
skin of healthy people, and rarely from clinical samples (cerebrospinal
fluid). This was the most frequent Acinetobacter species isolated from
activated sludge in Australia (Duncan et aI., 1988). It seems to be a common
contaminant of refrigerated eviscerated chicken (Bouvet and Grimont,
1987). This species is commonly found on the skin (hands) of nurses, but is
also found on the skin of people not working in a hospital environment
(Bouvet and Grimont, 1987). Cc?mbined with the fact that this species cannot
grow at 37°C, knowledge of this habitat requires that the clinical significance
of an isolate (especially from cerebrospinal fluid) should be critically
evaluated, with a strong suspicion of hand-borne contamination. In an
epidemiological investigation, the presence of Acinetobacter strains on the
hands of personnel is meaningless unless the identification procedure
differentiates A. baumannii from A. johnsonii and other species.
A. lwoffii has been isolated from human patients (about 3 to 6% of

total Acinetobacter isolates), hands of uninfected personnel, eviscerated
chickens and activated sludge (Bouvet and Grimont, 1987; Duncan et aI.,
1988).
A. radioresistens, originally found in association with cotton, has

also been isolated from patients (Tjernberg and Ursing, 1989).
The habitats of the other genomic species are, as yet, poorly

understood.
33
CONCLUSIONS
Accurate identification to the species level of a clinical isolate

belonging to the genus Acinetobacter is essential since: (i) identification
is a primary epidemiological marker; (ii) habitats may differ between
species; (iii) the antibiotic susceptibility of A. baumannii compared with
other species is quite different (Freney et al., 1989). Further work is
required to better delineate, characterise, and name unnamed species, and
to make such identification readily available to more laboratories. More
precise identification should improve our knowledge of the ecology of the
different Acinetobacter species. The numerous biochemical studies
performed with Acinetobacter strains will have a significantly greater
scientific impact when the strains used in such studies have been precisely
identified.
REFERENCES
Alexander, M., Ismail, F., Jackman, P.J.H., and Nobel, W.c., 1984,
Fingerprinting Acinetobacter strains from clinical sources by
numerical analysis of electrophoretic protein patterns,
1. M ed. Microbiol., 18:55.
Alexander, M., Rahman, M., Taylor, M., and Noble, W.c., 1988, A study of
the value of electrophoretic and other techniques for typing
Acinetobacter calcoaceticus, 1. Hosp. Infect., 12:273.
Aubert, J.P., Milhaud, G., and Gavard, R., 1952, Etude preliminaire d'un
systeme enzymatique bacterien oxydant Ie glucose en acide
gluconique, Compt. Rend. Seances H eb. Acad. Sci., 235:1165.
Moraxella group. II. Oxidative-negative species (genus
Acinetobacter), 1. Bacteriol., 95:1520.
Beijerinck, M.W., 1911, Uber Pigmentbildung bei Essigbakterien,
Proc. Royal Acad. Sci. (Amsterdam), 13:1066.
Bouvet, P.J.M., and Bouvet, O.M.M., 1989, Glucose dehydrogenase activity
in Acinetobacter species, Res. M icrobiol., 140:531.
Bouvet, P.J.M., and Grimont, P.A.D., 1986, Taxonomy of the genus
Acinetobacter with the recognition of Acinetobacter
baumannii sp. nov., Acinetobacter haemolyticus sp. nov.,
Acinetobacter johnsonii sp. nov., and Acinetobacter junii sp.
nov., and emended descriptions of Acinetobacter
calcoaceticus and Acinetobacter lwoffii, Int. 1. Syst.
Bacteriol.,36:228.
Bouvet, P.J.M., and Grimont, P.A.D., 1987, Identification and biotyping of
Bouvet, P.J.M., and Jeanjean, S., 1989, Delineation of new proteolytic
genomic species in the genus Acinetobacter, Res. Microbiol.,
140:291.
34
Bouvet, P.J.M., Jeanjean, S., Vieu, J.F., and Dijkshoorn, L., 1990, Species,
biotype, and bacteriophage type determinations compared with
cell envelope protein profiles for typing Acinetobacter strains,
1. Clin. Microbiol., 28:170.
Brisou, J., and Prevot, A.R., 1954, Etudes de systematique bacterienne. X.
Revision des especes reunies dans Ie 'genre Achromobacter,
Dijkshoorn, L., Michel; M.F., and Degener, J.E., 1987a, Cell envelope
protein profiles of Acinetobacter calcoaceticus strains
isolated in hospitals, 1. M ed. Microbial., 23:313.
Dijkshoorn, L., van Vianen, W., Degener, J.E., and Michel, M.F., 1987b,
hospital patients by cell envelope protein profiles,
Epidemiol. Infect., 99:659.
Duine, J.A., Frank, J., and van Zeeland, K., 1979, Glucose dehydrogenase
from Acinetobacter calcoaceticus: a "quinoprotein", FEBS
Lett., 108:443.
Duncan, A., Vasiliadis, G.E., Bayly, R.e., and May, J.W., 1988, Genospecies
of Acinetobacter isolated from activated sludge showing
enhanced removal of phosphate during pilot-scale treatment of
sewage, Biotech. Lett., 10:831.
Freney, J., Bouvet, P.J.M., and Tixier, c., 1989, Identification et
determination de la sensibilite aux antibiotiques de 31 souches
cliniques d'Acinetobacter autres que A. baumannii,
Ann. Bioi. Clin., 47:41.
Grimont, P.A.D., 1988, Use of DNA reassociation in bacterial classification,
Can. 1. Microbial., 34:541.
Grimont, F., and Grimont, P.A.D., 1986, Ribosomal ribonucleic acid gene
restriction patterns as potential taxonomic tools, Ann. Inst. Past.
Microbial., 137B:165.
Hauge, J.G., 1960, Kinetics and specificity of glucose dehydrogenase from
Bacterium anitratum, Biochim. Biophys. Acta, 45:263.
Henriksen, S.D., 1973, Moraxella, Acinetobacter and the Mimae,
Bacterial. Rev., 37:522.
Henriksen, S.D., 1976, Moraxella, Neisseria, Branhamella and
Acinetobacter, Ann. Rev. Microbial., 30:63.
Ino, T., and Nishimura, Y., 1989, Taxonomic studies of Acinetobacter
species based on outer membrane protein patterns, 1. Gen. Appl.
Microbial., 35:213.
Johnson, J.L., Anderson, R.S., and Ordal, E.J., 1970, Nucleic acid
homolpgies among oxidase-negative M 0 raxe II a species, 1.
Bacterial., 101:568.
evidence for a ubiquitous genus, 1. Bacteriol., 112:917.
Juni, E., 1984, Genus III. Acinetobacter Brisou et Prevot 1954, in
"Bergey's Manual of Systematic Bacteriology, vol. 1," p.303, N.R.
Grieg and J.G. Holt, eds. Williams and Wilkins, Baltimore.
35
Nishimura, Y., Kano, M., Ino, T., Iizuka, H., Kosako, Y., and Kaneko,T.,
1987, Deoxyribonucleic acid relationship among the radiation-
resistant Acinetobacter and other Acinetobacter, 1. Gen.
Appl. Microbio!., 33:371.
Nishimura, Y., Ino, T., and Iizuka, H., 1988a, Acinetobacter
radioresistens sp. nov. isolated from cotton and soil, Int. 1.
Syst. Bacterio!., 38:209.
Nishimura, Y., Kanzaki, H., and Iizuka, H., 1988b, Taxonomic studies of
Acinetobacter species based on the electrophoretic analysis of
enzymes, 1. Basic Microbiol., 6:363.
Picard, B., Goullet, P., Bouvet, P.l.M., Decoux, G., and Denis, 1.B., 1989,
Characterization of bacterial genospecies by computer-assisted
statistical analysis of enzyme electrophoretic data,
Electrophoresis, 10:680.
Skerman, V.B.D., McGowan, V., and Sneath, P.H.A., eds., 1980, Approved
lists of bacterial names, Int. 1. Syst. Bacteriol., 30:225.
Stenzel, W., and Mannheim, W., 1963, On the classification and
nomenclature of some nonmotile and coccoid diplobacteria,
exhibiting the properties of Achromobacteriaceae, Int. Bull.
Bacteriol. Nomen. Taxon., 13:195.
Tjernberg, I., and Ursing, 1., 1989, Clinical strains of Acinetobacter
classified by DNA-DNA hybridization, AP MIS 97:595.
Traub, W.H., 1989, Acinetobacter baumannii serotyping for delineation
of outbreaks of nosocomial cross-infection, 1. Clin. Microbio!.,
27:2713.
Van Landschoot, A., Rossau, R., and De Ley, 1., 1986, Intra- and intergeneric
similarities of the ribosomal ribonucleic acid cistrons of
Acinetobacter, Int. 1. Syst. Bacterio!., 36:150.
Wayne, L.G., Brenner, D.l., Colwell, R.R., Grimont, P.A.D., Kandler, 0.,
Krichevsky, M.l., Moore, L.H., Moore, W.E.C., Murray, R.G.E.,
Stackebrandt, E., Starr, M.P., and Truper, H.G., 1987, Report of
the ad hoc committee on reconciliation of approaches to bacterial
systematics, Int. 1. Syst. Bacterio!., 37:463.
Woese, C.R., Weisburg, W.G., Hahn, C.M., Paster, B.l., Zablen, L.B., Lewis,
B.l., Macke, T.l., Ludwig, W., and Stackebrandt, E., 1985, The
phylogeny of purple bacteria: the gamma subdivision, Syst. Appl.
M i crob io!., 6:25.
36
TYPING OF ACINETOBACTER
P.J.M. Bouvet
Unite des Enterobacteries

INSERM Unite 199
Institut Pasteur
France
INTRODUCTION
Need for a Typing System in the Genus Acinetobacter
In the last 25 years, outbreaks of Acinetobacter infection have been

a growing concern in hospitals (Bergogne-Benhin et aI., 1987, this volume;
Noble, this volume). These infections can be difficult to treat since strains
are commonly resistant to multiple antimicrobial agents. In 1986 the
taxonomy of the genus Acinetobacterwas changed extensively. At least 19
DNA hybridisation groups have now been delineated (Bouvet and Grimont,
1986; Nishimura et aI., 1988; Bouvet and Jeanjean, 1989; Tjernberg and
Ursing, 1989) and five new species have been described: A. baumannii, A.
haemotyticus, A. junii, A. johnsonii and A. radioresistens. In
addition, the original descriptions of A. catcoaceticus and A. twoffii have
been emended (Grimont, this volume).
Preliminary results published in 1987 (Bouvet and Grimont, 1987)

showed that A. baumannii was the most prevalent Acinetobacter species
isolated from clinical specimens, with 253 of 299 (84.6%) nosocomial
isolates being identified as A. baumannii. Additional supporting data are
now available (Freney et aI., 1987; Giammanco et aI., 1989; Hercouet et aI.,
1989; Traub, 1989). In contrast, Tjernberg and Ursing (1989) studied 168
strains isolated consecutively during one year from hospitals and outpatient
clinics in the city of Malmo, Sweden. Only six strains (3.6%) were identified
as A. baumannii. The most prevalent Acinetobacter species in this study
were A. twoffii (22.6%), A.radioresistens (17.9%), Acinetobacter sp. 3
(16.7%) and A. junii (10.7%). Although the ratio of hospital to outpatient
37
isolates was not disclosed, this statistic may strongly influence the proportion
of isolates obtained. A previous study (Gerner-Smidt et aI., 1985) showed
that most Acinetobacter isolates (153 of 185) obtained in general practice
were "A. calcoaceticus biovar lwoffii" (possibly A. lwoffii, A. junii, A.
johnson ii, or Acinetobacter spp. 11, 12 ... etc.), whereas 93 of 143 strains
(65%) isolated in hospitals were of "biovar calcoaceticus" (most probably
A. baumannii). Thus, there is an acute need for a typing system allowing
differentiation of A. baumannii strains.
Several methods, such as bacteriophage typing or cell envelope

protein profiles, which were originally developed for epidemiological
purposes before 1986, should be reassessed in light of the new taxonomic
knowledge. Since 1986, biotyping (typing by biochemical characteristics)
and serotyping methods have been developed for Acinetobacter strains
identified to the species level according to the new classification. Molecular
methods such as low-frequency-deavage restriction endonuclease digest
analysis, enzyme electrophoresis and rRNA gene restriction patterns are also
potential tools for studying epidemiology. A few examples of potential
typing methods will be reviewed below.
BIOTYPING
Early attempts at biotyping Acinetobacter isolates relied on

gelatinase production, haemolysis, and acid production from glucose
(Gilardi, 1978). Within the newly described Acinetobacter species
(Bouvet and Grimont, 1986; Bouvet and Jeanjean, 1989), gelatinase
production is restricted to very few species, i.e. A. haemolyticus,
Acinetobacter sp. 6 and Acinetobacter DNA groups 13 to 17. These
species have never been implicated in outbreaks of Acinetobacter
infection.
For many decades D-glucose oxidation has been used to distinguish

Acinetobacter strains or "species" (calcoaceticus and lwoffii). We have
recently published a study of D-glucose oxidation by 1134 strains of
Acinetobacter (Bouvet and Bouvet, 1989). Four species, i.e. A.
calcoaceticus sensu stricto, A. baumannii (with a few exceptions),
Acinetobacter sp. 3 and Acinetobacter sp. 10, were represented solely by
glucose-oxidising strains. A. lwoffii (with a few exceptions), A. junii, A.
johnsonii and Acinetobacter sp. 11 were represented solely by glucose
non-oxidising strains. Other species, i.e. A. haemolyticus,
Acinetobacter sp. 6 and Acinetobacter sp. 12, contained both glucose-
oxidising and glucose non-oxidising strains. These results confirmed that
the former taxonomic scheme for the genus Acinetobacter (two species or
biovars differing only by glucose oxidation) was untenable and that
identification of Acinetobacter strains to the species level should be
performed using better methods such as carbon source utilisation tests.
38
Table 1. Scheme for identifying biotypes of Acinetobacter baumannii
Biotype
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Utilisation of:
1 evu1inate + + + + - - - + + + +
citraconate - + + + + + - +
L-pheny1alanine + + + - + + + + + + + + - +
and pheny1acetate
4-hydroxybenzoate + + + + - + + + + + + - + +
L-tartrate + - + - - + - + + + +
+ - growth after 2 days; - - no growth after 2 days
Further attempts at biotyping were published by Towner and Chopade

(1987) using a commercially available identification system. This 20-test
system (mainly carbon source utilisation tests) allowed the differentiation of
209 possible biotypes. A total of 31 different biotypes were identified among
clinical isolates from Nottingham hospitals. This method, if used alone, still
does not allow the identification of Acinetobacter strains to the species
level.
More recently, a biotyping system has been devised to type isolates of

A. baumannii, the species usually implicated in nosocomial outbreaks of
infection. This system was based on the utilisation of six carbon sources
(levulinate, citraconate, L-phenylalanine, phenylacetate, 4-hydroxybenzoate
and L-tartrate), and allowed the recognition of 19 biotypes (Table 1), of
which biotypes 1, 2, 6 and 9 were most frequently encountered. This
biotyping system has, in some instances, been valuable in typing nosocomial
strains (Giammanco et al., 1989; Hercouet et al., 1989; Buisson et al., 1990).
The following sections will compare biotyping data with results obtained by
other typing methods such as bacteriophage typing or cell envelope protein
profiles.
PHAGE TYPING
A phage typing system developed ten years ago at the Institut Pasteur
(Vieu et al., 1979; Santos-Ferreira~~l., 1984) has been found to be useful
for the study of hospital outbreaks (Vieu et al., 1980). Unfortunately, this
39
phage typing system was developed in the absence of a well-defined
taxonomic structure and has only been used in a single laboratory in the
world. Briefly, the bacteriophage typing system is based on the use of two
complementary sets of bacteriophages isolated from sewage (Santos-
Ferreira et at, 1984). The first set included 21 bacteriophages and
distinguished 116 phage types (Vieu et aI., 1979). A further nine new phage
types have been added subsequently. The proportion of untypable strains,
due to their insensitivity to this set of bacteriophages, was 30% (Santos-
Ferreira et aI., 1984). A second set of 14 phages was introduced to sub-
divide the untypable strains. A total of 21 phage types (called sub-types) was
recognised by this second set of phages, and the overall proportion of
untypable strains was reduced to about 20% (Santos-Ferreira et at, 1984).
Typing of strains isolated from countries other than France has sometimes
proved to be difficult. In the study of Giammanco et aI. (1989), performed in
Palermo, Italy, only 19 of 54 A. baumannii strains examined (35%) were
correctly typed, while six strains had a new phage type and 25 strains were
resistant to all phages tested.
Data on species, biotypes and bacteriophage types of 528 hospital

isolates from published and unpublished results are available for
comparison. In all, 355 isolates were from 16 French hospitals, 68 strains
were from an Italian hospital, and 105 strains were from 11 hospitals in the
Netherlands. Data from 444 A. baumannii strains were examined in order
to correlate biotypes and phage types. A total of 380 strains (85.5%) were
typable, of which 331 strains belonging to 11 biotypes were typable using the
first phage set, while 49 strains belonging to three biotypes were typable
using the second set of phages. The results presented in Tables 2 and 3 list
the number of typable strains for each biotype that were susceptible to the
different phages of the first phage set (Table 2) or the second phage set
(Table 3). In the first system, phages 2, 5, 6, 11, 14-18,20 and 21 were nearly
always inactive on the 331 strains tested. Most biotype 1 and biotype 2
strains were susceptible to phages 9 and 12. Most biotype 1 strains were
susceptible to phage 10, whereas biotype 2 strains were not. Most strains
(76%) of biotype 2 were susceptible to phage 7. Phages 1 and 13 seemed to
be specific for biotypes 6 and 9. All strains of biotypes 6 and 9 were typable
with the first system, compared with 55% and 84% of the strains in biotypes 1
and 2 respectively. Acinetobacter sp. 3 strains, which are phenotypically
and genetically related to A. baumannii, were also typable (25 of 29 strains
studied; data not shown). Among strains belonging to species other than A.
baumannii or Acinetobacter sp. 3, only seven of 59 strains (12%) were
typable. The typable strains belonged to A. haemolyticus (two strains), A.
johnsonii (three strains), A. lwoffii (one strain), and Acinetobacter sp.
10 (one strain).
Evidence from some studies has cast doubt upon the reproducibility of
this method. In the study of Buisson et aI. (1990), of 24 A. baumannii
biotype 6 strains studied, 23 belonged to phage type 17 and one strain to
phage type 16. Phage type 17 (susceptibility to phages 1, 12 and 13) differed
from phage type 16 (susceptibility to phages 1, 12 and 13) by susceptibility to
40
Table 2. Susceptibility to the 21 phages of the first phage typing system of 331 typable A.baumannii strains
Number of strains susceptible to phage no.

Biotype No. strains Total
No. typable strains 234 5 6 8 9 10 11 12 13 14 15 16 17 18 19 20 21
by this set typable
36 66 276 3 3 6 8 23 23 o 22 o o o o o o 14 o o
2 88 105 00330 o 67 44 56 o o 62 o o o o o o 2 o 2
6 136 136 134 6 9 39 o 2 33 o 5 132 4 3 o o o o o
....(
9 54 54 40 4 7 2 o 6 2 12 13 o 16 38 o 2 o o o o o
8 I, 4 00000 o o o 2 o 2 o o o o o o
10 o 0 o o o o o o o o o o o o o o o o o
11 6 6 452 3 2 o o 4 o o o o o o o o o o
14 2 4 o o o o o 2 2 o o o o o o o o o o
17 2 2 o 2 o o o o o o o o o o o o o o o
18 o 0 o 0 o o o o o o o o o o o o o o o o
19 o o o o o o o o o o o o o o o o
~'
..
One strain was typable by both phage typing systems.
~
.j>.
N
Table 3. Susceptibility to the 14 phages of the second phage typing system of

49 typable A. baumannil strains
Number of strains susceptible to phage no.

Biotype No. strains Total
No. typable strains 2 3 4 5 6 7 8 9 10 11 12 13 14
by this set typable
30 66 0 30 30 29 4 ) 0 0 0 0 13
2 17 105 0 17 17 17 4 0 0 0 0 0 0 3
14 2 4 0 0 2 2 2 0 0 0 0 0 0 0 2
a single phage. In the same study, ten biotype 1 strains belonged to phage
type 86 (susceptibility to phage 9), phage type 88 (susceptibility to phages 9,
10 and 12) and phage type 89 (susceptibility to phages 10 and 12). It
therefore seems that the phage typing system is of limited value in its present
form and should be extended with other phages in order to increase the
percentage of typable strains. Further studies should be done with strains
that are well-defined epidemiologically in order to evaluate the
reproducibility and reliability of this phage typing method.
SEROTYPING
There have been numerous attempts to type Acinetobacter strains by

serological reactions (see the review by Henriksen, 1973; Adam, 1979; Das
and Ayliffe, 1984). Limited success has been obtained and most schemes
have been rendered obsolete by taxonomic developments. Recently, Traub
(1989) studied 152 clinical isolates of A. baumannii from 152 patients and
developed a serotyping system for this species using polyclonal rabbit
immune sera. A total of 20 serovars were delineated. The sera were fairly
specific for A. baumannii. Most nosocomially significant A. baumannii
isolates belonged to serovar 4. Several outbreaks of nosocomial cross-
infection caused by serovars 4 and 10 were identified (Traub, 1989).
CELL ENVELOPE PROTEIN PA TTERNS
Cell envelope protein profiles have been shown to be a valuable tool in

typing Acinetobacter strains (Alexander et aI., 1984; Dijkshoorn et aI.,
1987a,b, 1989; Crombach et aI., 1989). Cell envelope protein patterns
allowed differentiation among strains associated with epidemiological
outbreaks (Dijkshoorn et aI., 1987a) or cross-infection (Dijkshoorn et aI.,
1987b). Bouvet et al. (1990) subsequently identified, biotyped, and phage-
typed 120 strains for which cell envelope protein patterns had been
determined by Dijkshoorn et al. (1987a,b) These strains could be classified
into 15 protein patterns (designated A to Hand J to P). Patterns that were
similar, but not identical, were designated with a capital letter followed by a
number (BI-4, Dl-3, EI-4). The correlation between identification, biotype
determination and cell envelope protein profile is shown in Table 4. A total
of 90 strains belonging to A. baumannii (nine biotypes), Acinetobacter
sp. 3 (six biotypes) and A. calcoaceticus could be assigned to 16 different
protein patterns A, Bl, B2, B4, C, Dl, D2, D3, G, H, J, K, L, M, N. Among
these strains, 17 had unique protein profiles.
A. baumannii biotypes land 2 were indistinguishable by protein

profile (protein pattern A), but could of course be differentiated by
biotyping. Most biotype 6 strains had related type B protein patterns, but
two strains had pattern J, one strain pattern M, and one strain a unique
pattern. A. haemolyticus strains had related type E profiles. Most strains
belonging to other species had unrelated unique patterns. Multiple repeat
43
~
~
Table 4. Correlation between identification, biotyping and cell envelope protein patternsa(data ,from Bouvet et a1.,1990)
Species A Bl B2 B3 B4 C Dl D2 D3 E F G H K L M N 0 P Unique
A. baumannii
biotype 1 11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
biotype 2 17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2
biotype 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
biotype 6 0 7 5 1 0 0 0 0 0 0 0 0 0 2 0 0 1 0 0 0 1
biotype 8 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 2
biotype 9 0 0 0 0 0 2 0 0 0 0 0 4 0 0 0 0 0 0 0 0 5
biotype 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0
biotype 16 0 0 0 0 0 0 0, 0 0 0 0 0 5 0 0 0 0 0 0 0 5
biotype 19 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Acin~toba~t~r sp. 3
biotype 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0
biotype 8 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 2
biotype 9 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1
biotype 11 0 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0
biotype 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0
biotype 18 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
A. ha~mol~ticus 0 0 0 0 0 0 0 0 0 6 0 0 0 0 0 0 0 0 0 0 1
A. johnsonii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1
A.junii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2
A.lwoffii 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 10
A. ~al~Qa~ticus 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 3
Acin~tot!a~tgr liP.6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
Acinetobacter sp.ll 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
Ungrouped strains 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 3
a
a total of 118 different strains were studied.
isolates from the same patient were indistinguishable by either typing
method. Strains with profiles G or L (Dijkshoorn et al., 1987b; Bouvet et
al., 1990) which were demonstrated to be responsible for cross-infection
could be easily detected by using identification and biotyping and, to a lesser
extent, phage typing. However, when strains from other origins were
examined, this correlation could not be reproduced; indeed a study by
Giammanco et al. (1989) led to an opposite conclusion. Analysis by
polyacrylamide gel electrophoresis of the cell envelope protein failed to
show any diversity, not only within, but also between some of the biotypes of
A. baumannii. Alexander et al. (1988), in a study comparing the value of
electrophoretic and other techniques (plasmid analysis, antibiograms and
biochemical tests) for typing Acinetobacter, also pointed out the
reproductibility problems associated with this method.
PLASMID PROFILE TYPING AND PLASMID FINGERPRINTING
With other bacterial species, determination of plasmid profiles has

proved to be an inexpensive and easily performed epidemiological tool with
good discriminatory power, especially when used in conjunction with other
typing methods (Mayer, 1988). Few studies have been performed with the
genus Acinetobacter, although Gerner-Smidt (1989) investigated the
plasmid content of 93 clinical isolates of Acinetobacter. Unfortunately,
strain identification followed the older taxonomic schemes for the genus.
Plasmids were found in 75 of 93 (81%) strains studied, including 56 of 71
epidemiologically unrelated strains. The number of plasmids found varied
between strains (range 0-20 with a median of two). A single plasmid was
found in 13% of strains. Among the glucose-oxidising strains, 77% had
plasmids, compared with 82.8% of glucose non-oxidising strains. Three
strains from the Neurosurgical Department and six strains from the Intensive
Care Unit were identified by Gerner-Smidt (1989) as atypical A. baumannii
biotype 9 (malonate not utilised) and all of these strains contained a plasmid
of the same size. Two additional strains from the Intensive Care Unit were
found to contain two plasmids and were also identified as A. baumannii
biotype 9. The finding of only one or two plasmids in these strains illustrates
the major limitation in the use of plasmid profiles for epidemiological
purposes. It is well-known that microorganisms may gain or lose plasmids in
hospitals and that repeated culturing and storing may result in loss of
plasmids. A separate study by Vila et al. (1989) used plasmid profiles and
restriction endonuclease digestion (plasmid fingerprinting) to examine 25
clinical glucose-oxidising strains (one strain from each patient). The strains
could be divided into two groups on the basis of assimilation of two
substrates (phenylacetate and adipate), the number of plasmid bands (one or
two), and restriction endonuclease digestion profiles (identical for strains
within a group and different between groups.) Determination of plasmid
profiles seems to be a simple and reproducible tool for use with
Acinetobacter strains. The method should, however, be used in
conjunction with other typing methods because of the risk of plasmid
instability and because 20% of strains do not contain detectable plasmid
45
DNA. Plasmid profiles have been used to either verify the identity of
bacterial isolates or show that they are different. It should, however, be
noted that while in some cases plasmid profile analysis seems to be the most
discriminating typing method, in other cases plasmid profiles have been of no
value. Restriction endonuclease digestion of plasmid DNA may be of
additional value when only a single plasmid can be identified or when similar
plasmid profiles occur in strains which cannot be differentiated by othe'r
methods.
ANTIBIOTIC RESISTANCE PATTERNS
Antibiotic resistance patterns have been used to differentiate strains

involved in outbreaks of infection, especially when a new resistance
phenotype emerges (Castle et aI., 1978; French et aI., 1980; Devaud et aI.,
1982; Holton, 1982; Gerner-Smidt et aI., 1985; Allen and Green, 1987;
Crombach et aI., 1989). Since October 1984, outbreaks of infections caused
by strains resistant to kanamycin and structurally related aminoglycosides,
including amikacin, have occurred in France. Resistance to aminoglycosides
was demonstrated to result from synthesis of a new type of 3'-aminoglycoside
phosphotransferase [APH(3')] (Lambert et aI., 1988). This resistance was
carried in the particular A. baumannii strain investigated by a 63 kb
plasmid, self-transferable to other A ci n e to b act e r species such as A.
haemolyticus or A. lwoffii, but not to Escherichia coli. This new
kanamycin resistance gene (aphA-6) has been sequenced by Martin et aI.
(1988). Buisson et al. (1990) studied amikacin-resistant, tobramycin-
sensitive Ac i ne to b ac ter strains isolated in a hospital and demonstrated,
using dot blot hybridisation with an aphA - 6 -intragenic probe, that the gene
coding for amikacin resistance had disseminated among different
Acinetobacter species (A. baumanni, A. haemolyticus, A. junii, A.
johnsonii, A. lwoffii, Acinetobacter sp. 3 and sp. 11) and different
biotypes of A. baumannii. In addition, this phenotype was encountered in
saprophytic A. johnsonii isolates recovered from the hands of 11 healthy
workers among the medical staff. The finding of this type of resistance in A.
johnsonii, which is a normal saprophyte of the human skin, and its isolation
from the hands of medical staff, suggests that this species may have played a
role in the transmission and dissemination by hand carriage of the amikacin
resistance gene to more pathogenic species. Typing of Acinetobacter
strains by antimicrobial resistance pattern must, therefore, be interpreted
with care and only after identification of the strain to the species level.
LOW-FREQUENCY -CLEAVAGE RESTRICTION

ENDONUCLEASE ANALYSIS
Analysis of bacterial restriction endonuclease digests provides a direct

and sensitive means of detecting minor genomic differences between
microorganisms. Restriction endonucleases specifical\y cleave DNA into
different lengths. depending on the number and position of the individual
46
recognition sequences, provided that they have not been modified in any way.
A DNA polymorphism involves a change in the size of a restriction fragment.
If a change occurs in a genomic DNA sequence, this can delete or create a
new recognition site and result in the generation of a restriction fragment
length polymorphism (RFLP). The DNA fragments generated by restriction
enzyme digestion are separated according to size by electrophoresis in
agarose gels to give a pattern of bands. The usefulness of such patterns as
diagnostic tools is limited by their complexity; indeed they may comprise
> 50 bands of various sizes, depending on the cutting frequency of the
restriction endonuclease used and the size of the organism's genome. Large
DNA molecules can now be separated by pulsed-field gel electrophoresis,
making it possible to carry out DNA analysis with low-frequency-cleavage
restriction endonucleases. These produce fewer fragments, usually < 40,
and electrophoretic patterns can be compared more precisely and easily.
This technique was used by Allardet-Servent et al. (1989) to investigate an
outbreak of A. calcoaceticus infection (although the species was not
precisely determined by these authors) in a urologic department. Two series
of strains were studied. The first (eight strains) consisted of isolates
obtained during three weeks from blood or urine cultures of patients
admitted to the urologic department. The second group (five strains)
comprised occasional isolates from different departments of the same
hospital. The eight strains of the first group were of the same phage type
and all but three had different antibiotypes. Following digestion with SmaI
and separation of the resulting DNA fragments with pulsed-field
electrophoresis (over a 48 h period), all these isolates presented the same
pattern of about 30 distinct bands. The patterns of the strains belonging to
the second group were all different and distinct from that of the former
group. The authors concluded that low-frequency-cleavage restriction
endonuclease analysis of genomic DNA is a versatile and precise method for
epidemiological investigation.
ELECTROPHORETIC ANALYSIS OF ISO ENZYMES
Electrophoretic analysis of isoenzymes has been suggested as a

method for studying the genetic diversity of bacteria (Selander et ai., 1986,
1987) and has been used for the phenotypic characterisation of bacterial
species defined by DNA-DNA hybridisation (Goullet and Picard, 1986).
Electrophoretic analysis of esterases has also been used as an
epidemiological typing method (Picard and Goullet, 1987). Electrophoretic
analysis (Picard et aI., 1989) of 27 esterases and two dehydrogenases from 81
genetically characterised A c i net 0 b act e r strains (Bouvet and Grimon t,
1986) enabl'ed clear computerised differentiation of the 12 genospecies.
The number of electrophoretic variants for each enzyme among the
Acinetobacter strains varied between two and 17 (with a mean of 6.5 for
each enzyme). Preliminary analysis of 40 unrelated A. baumannii strains
belonging to 14 biotypes demonstrated that all patterns were dissimilar.
Further studies on epidemiologically related strains should be carried out in
order to properly assess the utility of this method for typing Acinetobacter
isolates.
47
CONCLUSIONS
As an initial step in typing, it is necessary to establish which particular

Acinetobacter species are of clinical an'd epidemiological relevance. Once
this is known, new isolates should first be identified to the species level in
order to trace epidemic strains.
Bacteriophage typing of Acinetobacter is used in only one laboratory

in the world. Furthermore, its apparent lack of reproducibility and the
relatively high percentage of untypable strains (particularly when strains
isolated from countries other than France are investigated) makes its use as
a typing method difficult. However, when strains are typable, the method
can successfully sub-divide strains within a biotype or protein profile.
The recently developed serotyping system for A. baumannii and

Acinetobacter sp. 3, should become a rapid and accurate tool in the
surveillance of nosocomial outbreaks. Comparison of cell envelope protein
profiles also seems to be a precise method for the relative classification of
Acinetobacter. Its feasibility depends mainly on the epidemiological
context. If clinical isolates from patients in a ward or small area are to be
typed, the method allows particular strains to be identified. In contrast, if
large numbers of isolates from the hospital environment or the skin of
healthy people need to be examined, comparison of protein profiles may be
difficult and laborious and will not identify species.
Identification to the species level and biotyping using physiological

and nutritional properties are easy and accurate methods for screening
strains. When used together, identification/biotyping and cell envelope
protein profile typing has enabled all the strains so far investigated to be
typed or fingerprinted. At the present time it seems that a combination of
these two methods may be the best method for performing epidemiological
studies. Newer methods, such as analysis following treatment with low-
frequency-cleavage restriction endonucleases, electrophoretic analysis of
isoenzymes, or restriction endonuclease analysis of rRNA genes, still require
further evaluation using strains that are well-defined both taxonomically and
epidemiologically.
REFERENCES
Adam, M.M., 1979. Classification antigenique des "Acinetobacter",

Ann. Microbiol. (Inst.Past.), 130A:404.
Alexander, M., IsmaiL F., Jackman, P.J.H., and Noble, W.e., 1984.
numerical analysis of electrophoretic protein patterns.
1. Med. Microbiol .. 18:55.
Alexander, M., Rahman. M .. Taylor. M., and Noble, W.e.. 1988, A study of
Acinetobacter calcoaceticus. 1. Hosp. Infect., 12:273.
48
Allardet-Servent, A., Bouziges, N., Carles-Nurit, M.J., Bourg, G., Gouby, A.,
and Ramuz, M., 1989, Use of low-frequency-cleavage restriction
endonucleases for DNA analysis in epidemiological investigations
of nosocomial bacterial infections, 1. Clin. Microbiol., 27:2057.
Allen, K.D., and Green, H.T., 1987, Hospital outbreak of multi-resistant
Acinetobacter anitratus: an airborne mode of spread,
1. Hosp. Infect., 9:110.
Bergogne-Berezin, E., J oly-Guillou, M.L.. and Vieu, J.F., 1987,
Epidemiology of nosocomial infections due to Acinetobacter,
Bouvet, P.J.M., and Bouvet, O.M.M., 1989. Glucose dehydrogenase activity
inAcinetobacterspecies, Res. Microbiol., 140:531.
nov., and emended descriptions of Acinetobacter calco-
aceticus and Acinetobacter twoffii, Int. 1. Syst. Bacteriol.,
36:228.
Microbiol.,138:569.
Bouvet, P.J.M., and Jeanjean, S., 1989, Delineation of new proteolytic
genomic species in the genus Acinetobacter, Res. Microbiol.,
140:291.
Bouvet, P.J.M., Jeanjean, S., Vieu, J.F., and Dijkshoorn, L., 1990, Species,
biotype, and bacteriophage type determinations compared with
cell envelope protein profiles for typing Acinetobacter strains,
1. Clin. Microbiol., 28:170.
Buisson, Y., Tran Van Nhieu, G., Ginot, L., Bouvet, P.J.M., Schill, H., Driot,
L., and Meyran, M., 1990, Nosocomial outbreaks due to amikacin-
resistant tobramycin-sensitive Acinetobacter species:
correlation with amikacin usage, 1. H osp. Infect., 15:83.
Castle, M., Tenney, J.R., Weinstein, M.P., and Eickhoff, T.e., 1978,
Outbreak of a multiply-resistant Acinetobacter in a surgical
intensive-care unit: epidemiology and control, Heart Lung,
65:641.
Crombach, W.R.J., Dijkshoorn, L., van Noort-Klaassen, M., Niessen, J., and
van Knippenberg-Gordebeke, G., 1989, Control of an epidemic
spread of a multi-resistant strain of Acinetobacter
calcoaceticus in a hospital, Intens. Care M ed., 15:166.
Das, B.e., and Ayliffe, G.A.J., 1984, Serotyping of Acinetob acter
calcoaceticus, J. Clin. Pathol., 37:1388.
Devaud, M., Kayser, F.H., and Bachi, B., 1982, Transposon-mediated
multiple antibiotic resistance in Acinetobacter strains,
Antimicrob. Agents. Chemother., 22:323.
Dijkshoorn, L., Michel, M.F., and Degener, J.E., 1987a, Cell envelope
protein profiles of Acinetobacter calcoaceticus strains
isolated in hospitals, J. M ed. Microbiol., 23:313.
49
Dijkshoorn, L., Van Vianen, W., Degener, l.E .. and Michel, lvI.F., 19~7b,
hospital patients by cell envelope protein profiles,
Epidemiol. Infect., 99:659.
Dijkshoorn, L., Wubbels, l.L., Beunders, A.J., Degener, J.E., Boks, A.L., and
Michel, M.F., 1989, Use of protein profiles to identify
Acinetobacter calcoaceticus in a respiratory care unit,
.I. Clin. Pathol., 42:853.
French, G.L., Casewell, M.W., Roncoroni, A.l., Knight, S., and Phillips, I.,
1980, A hospital outbreak of antibiotic-resistant Acinetobacter
anitratus: epidemiology and control,.1. H osp. Infect., 1:125.
Freney, J., Bouvet, P.l.M., and Tixier, c., 1989, Identification et
cliniques d'Acinetobacter autres que A. baumannii,
Ann. Bioi. Clin .., 47:41.
Gerner-Smidt, P., 1989, Frequency of plasmids in strains of Acinetobacter
caLcoaceticus,.I. Hosp. Infect., 14:23.
Gerner-Smidt, P., Hansen, L., Knudsen, A., Siboni, K., and Sogaard, 1., 1985,
Epidemic spread of Acinetobacter calcoaceticus in a
neurosurgical department analyzed by electronic data processing,
.1. Hosp. Infect., 6:166.
Giammanco, A., Vieu, J.F., Bouvet, P.J.M., Sarzana, A., and Sinatra, A.,
1989, A comparative assay of epidemiological markers for
Acinetobacter strains isolated in a hospital, Zbl. Bakt. 272:231.
Gilardi, G.L., 1978, Identification of miscellaneous glucose non-fermenting
Gram-negative bacteria, in "Glucose Nonfermenting Gram-
negative Bacteria in Clinical Microbiology," p.45, G.L. Gilardi,
ed., CRC Press, West Palm Beach.
Goullet, P., and Picard, B., 191\6, Comparative esterase electrophoretic
polymorphism of Escherichia coli isolates obtained from animal
and human sources,.1. Gen. lVlicrobiol., 132:1843.
Henriksen, S.D., 1973, Moraxella, Acinetobacter and the i\llimae,
Hercouet, H., Bousser, J., Donnio, P.Y., and Avril, J.L., 1989, Activite in
vitro des antibiotiques sur les souches hospitalieres de
Acinetobacter baumannii, Pathol. BioI., 37:612.
Holton, J., 1982, A report of a further hospital outbreak caused by a multi-
resistantAcinetobacter anitratus,.I. Hosp. Infect., 3:305.
Lambert, T., Gerbaud, G., and Courvalin, P., 1988, Transferable amikacin
resistance in Acinetobacter spp. due to a new type of 3'-
aminoglycoside phosphotransferase, Antimicrob. Agents
Chemother., 32: 15.
Martin, P., Jullien, E., and Courvalin, P., 1988, Nucleotide sequence of
AC'inetobacter baumanni aphA-6 gene: evolutionary and
functional implications of sequence homologies with nucleotide-
binding proteins, kinases, and other aminoglycoside-modifying
enzymes, Mol. Microbio!., 2:615.
Mayer, L.W., 1988, Use of plasmid profiles in epidemiological surveillance
of disease outbreaks and in tracing the transmission of antibiotic
resistance, Clin. Microbiol. Rev., 1:228.
50
Nishimura, Y., Ino, T., and Iizuka, H., 1988, Acinetobacter radioresistens
sp. nov. isolated from cotton and soil, Int. 1. Syst. Bacteriol.,
38:209.
Picard, B., and Goullet, P., 1987, Epidemiological complexity of hospital
aero monas infections revealed by electrophoretic typing of
esterases, Epidemiol. Infect., 98:5.
Picard, B., Goullet, P., Bouvet, P.J.M., Decoux, G., and Denis, J.B., 1989,
Characterisation of bacterial genospecies by computer-assisted
statistical analysis of enzyme electrophoretic data,
Electrophoresis, 10:680.
Santos-Ferreira, M.O., Vieu, J.F., and Klein, B., 1984, Phage-types and
susceptibility to 26 antibiotics of nosocomial strains of
Acinetobacter isolated in Portugal, 1. Internat. Med.Res.,
12:364.
Selander, R.K., Caugant, D.A., Ochman, H., Musser, J.M., Gilmour, M.N.,
and Whittman, T.S., 1986, Methods of multilocus enzyme
electrophoresis for bacterial popUlation genetics and systematics,
Appl. Environ. Microbio!., 51:873.
Selander, R.K., Musser, J.M., Caugant, D.A., Gilmour, M.N., and Whittam,
T.S., 1987, Population genetics of pathogenic bacteria,
Microb. Pathogen., 3:1.
Tjernberg, I., and Ursing, J., 1989, Clinical strains of Acinetobacter
classified by DNA-DNA hybridization, AP MIS, 97:595.
Towner, K.J., and Chopade, B.A., 1987, Biotyping of Acinetobacter
calcoaceticus using the API 20NE system, 1. Hosp. Infect.,
10:145.
Traub, W.H., 1989, Acinetobacter baumannii serotyping for
delineation of outbreaks of nosocomial cross-infection, 1. C lin.
Vieu, J.F., Minck, R., and Bergogne-Berezin, E., 1979, Bacteriophages et
lysotypie de Acinetobacter, Ann. Microbiol. (Inst. Past.),
130A:405.
Vieu, J.F., Bergogne-Ben:;zin, E., Joly, M.L., Berthelot, G., and Fichelle, A.,
1980, Epidemiologie d'Acinetobacter calcaaceticus,
Press. M ed., 9:3351.
Vila, J., Almela, M., and Jimenez de Anta, M.T., 1989, Laboratory
investigation of hospital outbreak caused by two different
multiresistant Acinetabacter calcaaceticus subsp. anitratus
strains, 1. Clin. Microbial., 27:1086.
51
HOSPITAL EPIDEMIOLOGY OF ACINETOBACTER INFECTION
W.e. Nohle
Department of Microhial Diseases

Institute of Dermatology
United Medical and Dental Schools
St Thomas' Hospital
Lambeth Palace Road
London SEI 7EH
UK
INTRODUCTION
Any consideration of hospital infection must take account of the

hackground, or community, level of infection on which hospital experience is
superimposed. It must also take account of the availability of typing
schemes, since these determine the level of discrimination possible; these
are discussed elsewhere in this volume. Hospital isolates of Acinetobacter
are generally resistant to many antibiotics (see for example Garcia et aI.,
1983; Obana et al., 1985), although this is more often a characteristic of
A. calcoaceticus var.anitratus than of var.lwoffii. It may be that the
predominance of var.anitratus in hospitals is a result of this resistance,
rather than a greater virulence, and also that it is the resistance which draws
attention to the spread of the strain. Gerner-Smidt et al. (1985) successfully
used antibiotic resistance alone to detect outbreaks of infection, but over-
reliance on antibiotic resistance may be misleading. Carlquist et al. (1982),
in a study of infection in 37 patients over five months in an intensive care unit
(ICU), found a gradual shift in zone size to three aminoglycosides and
carbenicillin. They interpreted this as changes in the resistance of a single
strain, although no other typing techniques were used to confirm this. No
changes in resistance were seen in Pseudomonas aeruginosa isolates from
the ICU in this period, nor in other non-ICU acinetobacters in the same
hospital. However, an increase in resistance was also recorded hy Stone and
Das (1985) in a single strain additionally characterised by serotype and
bacteriocin type. The most acceptable approach in hospital epidemiology
would therefore seem to be the use of several typing methods, obtaining a
53
concensus on identity with data from several sources (Alexander et aI., 1988).
In this way simultaneous outbreaks of infection by two strains of
A. calcoaceticus var.anitratus have been detected (Vila et aI., 1989).
Various aspects of Acinetobacter infection have been previously reviewed
by Vivian et ai. (1981), Bergogne-Berezin and Joly-Guillou (1985) and
Bergogne-Ben!zin et ai. (1987).
NORMAL OCCURRENCE OF ACINETOBACTER
Both A.calcoaceticus var.lwoffii and A.calcoaceticus

var.anitratus are normal inhabitants of human skin and are found in at least
the axillae, groin, toewebs and antecubital fossa of about 25% of the
population (Taplin et aI., 1963; Somerville and Noble, 1970). Patients with
skin disease, such as eczema, frequently carry Acinetobacter on their
lesions, although primary infection of the skin is rare (Glew et aI., 1977;
Gaughan et aI., 1979). As a result of observed colonisation, skin has been
suggested as a source for infection/contamination of the blood stream (AI-
Khoja and Darrell, 1979; Hoppe et aI., 1983). Al Khoja and Darrell (1979)
reported that up to half of patients with kidney disease might be colonised
and that persistent carriage could last for up to at least 9 weeks. In countries
where there is a well-defined summer and winter with a marked temperature
swing, acinetobacters are more commonly recovered from the skin in the
summer months (Kloos and Musselwhite, 1975). This is presumably the
result of greater sweating in the summer, especially in males. Higher skin
colonisation rates in summer are reflected in higher prevalence rates of
endemic infection with a marked excess of young males. A survey of
infection in 81 hospitals in the USA (Retailliau et aI., 1979) found a
pronounced summer peak of infection for A. calcoaceticus var.anitratus.
Patients aged 21-30 years were in marked excess, although traces of this
could also be seen in infection with other organisms (Table 1). Seasonal
effects have also been recorded in single hospitals (Holton, 1982; Smego,
1985).
Buisson et ai. (1990) have observed amikacin-resistant, tobramycin-

sensitive strains of A. baumannii infecting patients together with
A.johnsonii strains with the same resistance pattern on the skin. They
speculated that exchange of resistance plasmids between strains of
Acinetobacter may occur on the skin, but were unable to substantiate this,
although such mechanisms are well established for other members of the skin
flora such as the staphylococci.
ENDEMIC INFECTION
The endemic rate of infection is hard to determine SlDce it

often fails to arouse interest. Occasional infections are reported
from almost all sites on the body, but it is rare for the number of
patients at risk or other denominator to be given. Galvao et al.
54
Table 1. Age distribution of endemic infection due to ACinetobacter,
Pseudomonas and other organisms
Percentage Distribution
Age Acinetobacter Pseudomonas Other
(Years) Male Female Male Female Male Female
<1 l.9 4.1 2.8 2.2 8.5 4.4

1-10 4.9 2.2 2.1 2.0 2.3 l.4
11-20 9.6 7.9 6.4 4.0 5.0 7.1
21-30 15.2 10.7 10.4 6.8 7.5 13.1
31-40 6.3 9.4 6.6 6.5 6.1 9.3
41-50 9.3 13.5 7.4 8.0 8.3 10.1
51-60 13.8 14.2 13.4 13.2 15.2 12.7
61-70 17 .8 13.8 22.0 20.1 21.1 16.2
71-80 15.9 15.1 20.2 2l. 2 18.0 15.6
>80 5.4 9.1 8.7 16.0 8.0 10.3
Total 100.1 100 100 100 100 100.2
Based on Retai11iau et a1. (1979)
denominator to be given. Galvao et al. (1989) observed 23 episodes of

infection in 450 patient years of chronic ambulatory peritoneal dialysis
(CAPD), making acinetobacters the second most frequent Gram-negative
bacillary infection in CAPD. There were eight var.anitratus and three
var.lwoffii in the strains speciated. Endemic infection in a respiratory care
unit has been reported by Dijkshoorn et al. (1989), who used protein gel
electrophoresis patterns to identify and observe the spread of several
different strains; perhaps significantly, environmental strains sometimes
differed from those recovered from the patient. A similar study over 3.5
years has been reported, based on antibiotic resistance patterns (Sakata et
aI., 1989), again with a number of strains evident. In a study in a
neurosurgical unit, also over a period of 3.5 years, Gerner-Smidt et al. (1985)
found that isolations from sputum were more frequent than from any other
sites. Although the isolates were sub-divided on the sole basis of antibiotic
sensitivity patterns, it was clear that small epidemics, each of a few cases,
were superimposed on the endemic pattern. A similar mixed pattern was
reported by Larson (1984) in her 10 year study of infection in a single
hospital.
55
EPIDEMIC SPREAD AND SOURCES OF INFECTION
Contaminated Apparatus
Most reports of epidemic spread seem to have been associated with the
use of contaminated apparatus. Thus, over a period of six months,
contaminated resuscitators were associated with infection due to a single
Acinetobacter type in nine of ten babies in a special care baby unit (Stone
and Das, 1985). Similarly, contaminated Wright respirometers resulted in
eight patients developing pneumonia and a further nine being colonised in
the upper respiratory tract over a period of about 30 days (Cunha et al.,
1980). A single serotype and antibiogram again accounted for all
infections/ colonisations and no other source was found. Cold water room
humidifiers were reported by Smith and Massonari (1977) as the source of
infection for 24 patients, 21 of whom had positive blood cultures and fever.
Large numbers of Acinetobacter were easily demonstrated up to 2 m from
the humidifiers and it seems probable that skin contamination by the
airborne route subsequently resulted in infection via the indwelling venous
catheters possessed by 22 of the 24 patients. In a similar incident (Snydman
et al., 1977), contamination of the water reservoir in mist tents resulted in
false-positive blood cultures in 11 children. Since the nose and skin of the
children was heavily contaminated with Acinetobacter, it seems probable
that contamination of the needle occurred during vene-puncture, resulting in
a contaminated sample. Indeed, over a period of several years in a single
hospital, Hoppe et al. (1983) were able to distinguish both genuine blood
stream infection and contamination which they attributed to skin as a source.
Other Types of Environmental Contamination
Two other outbreaks, associated with contamination of mattresses and

peritoneal dialysis bags respectively, have been reported. In the first, an
outbreak of infection in burns patients lasted 21 months, during which time
63 of 103 patients were colonised by Acinetobacter (Sheretz and Sullivan,
1985). Infections occurred in 43 of the 63 patients, with the burn itself
accounting for 34, the urinary tract 16, the blood stream seven and the lower
respiratory tract five patients. Wet mattresses were found to harbour
Acinetobacter and the epidemic ended when mattresses were removed as
the patient was discharged. In the second outbreak, Abrutyn et al. (1978)
found that the water used to warm peritoneal dialysis fluid was contaminated
with Acinetobacter, as a result of which 13 patients suffered peritonitis.
Contamination of the rubber bung is thought to have resulted in
contamination of the dialysis fluid when organisms were carried into the bag
on the prong of the administration set.
In a further (hopefully unique) incident, contamination of intrathecal

methotrexate for leukemia patients resulted in meningitis caused by A.
calcoaceticus var.anitratus in eight patients and death in three (Kelkar et
56
aI.,1989). The outbreak was associated with the re-use of needles, which no
doubt resulted in contamination of the drug during reconstitution. An
unusual outbreak of infection due to A. calcoaceticus var.lwoffii has also
been associated with contaminated parenteral nutrition fluid. Although this
source was not proven, nine of 28 babies in an intensive care unit were
receiving parenteral nutrition and seven became infected with
Acinetobacter, with signs of severe clinical infection and septic shock (Ng
et aI., 1989).
Skin Carriage
Many authors comment on the contamination or colonisation of

patients' or staff skin and the spread of Acinetobacter which can occur as a
result. For example, infection by a single biotype and antibiogram of
Acinetobacter in intensive care patients was found to be strongly
associated with intubation and skin carriage amongst the hospital staff
(Buxton et aI., 1978). One therapist with chronic dermatitis was found to be
a semi-permanent carrier and to have contaminated the apparatus used for
the intubation. Similarly, Allen and Green (1987) reported wide spread of a
multi-resistant A. calcoaceticus var.anitratus strain which colonised 10
and infected 18 of 38 patients, probably by a mixture of hand contact and
airborne spread. Staff members also become colonised as a result of
working in wards during an epidemic. French et ai. (1980) reported that 11
of 38 staff in a urological ward became carriers of a gentamicin-resistant
strain during an outbreak, as did two microbiologists investigating the
outbreak. A. calcoaceticus var.anitratus survives better on the skin than
does var.lwoffii (Musa et aI., 1990), another factor which may bias the
distribution of the species in hospitals.
Airborne Spread
Dissemination of acinetobacters via the air is repeatedly mentioned in

association with spread from environmental sources and as one mode of
contamination of the skin. Airborne dispersal of Gram-negative bacilli is
relatively rare (Ayliffe and Lowbury, 1982); however, acinetobacters are
known to survive drying better than most Gram-negative bacteria and will
thus tend to survive aerosolisation. Once dried, Acinetobacter may persist
for at least a week on dried cloth (Buxton et aI., 1978) or at least five days on
formica surfaces (Musa et aI., 1990).
Cordes et al. (1981) reported community-acquired airborne infection

in foundry workers. Three men developed pneumonia due to
A. calcoaceticus var.anitratus serotype 71. Two of the men died and, at
post-mortem examination, were found to have pneumoconiosis and iron
particles in the lung. Acinetobacter, of unknown origin, were recovered
from the air of the foundry and 15% of the 167 workers there, but only 2% of
93 local community residents had significantly high serum titres to serotype
71. Contamination of the skin was found in only one of 107 samples and
from none of 106 throat swabs in men at the foundry.
57
Table 2. Percentage distribution of infections due to
Acinetobacter calcoaceticus
(1) (2) (3)
all Acinetobacter
isolates only
Respiratory tract 29 22 42 15
Urinary tract 27 36 10 43
Surgical wound 21. 5 12 9 27
Blood stream 10.5 20 18 15
Peritoneum 1 1 17 0
Other 11 8 5 0
Total patients
with Acinetobacter 1372 612 85 33
Data taken from (1) Retaill iau et a1. , 1979 ; (2) Larson, 1984;
(3) Muller-Serieys et al., 1989.
Urinary Tract I nf ecti on
The second, and about equally numerous, group of infections is

urinary tract infection. Hoffman et al. (1982) studied urinary tract infection
in the community and reported 14 isolations of A.calcoaceticus
vaLanitratus and 93 of vaLlwoffii from the 97 persons participating in the
study. Normal as well as particularly susceptible individuals were at risk of
infection. Similarly, French et al. (1980) studied an epidemic of infection in
a teaching hospital, caused by a single strain of gentamicin-resistant
A. calcoaceticus var.anitratus, and found that the index case and 25 of the
40 infected patients were from a urological ward. Twenty-five of the 31
patients with infected urinary tracts had been catheterised.
RATES OF INFECTION
It is difficult to determine rates of infection since these clearly differ

according to local circumstances; differences in reporting strategies also
obscure any comparison. Retailliau et al. (1979) reported that
58
Acinetobacter infection occurred in 3.11 per 10,000 patients discharged,
whilst Larson (1984) gave Acinetobacter as representing 1.4% of 6115
hospital infections, with no increase over a ten year period. According to
Hoffman et aI. (1982), Acinetobacter cultivars formed 4% of community-
acquired urinary tract infection, of which only 13% were A. calcoaceticus
var.anitratus. In contrast, community-acquired Acinetobacter
pneumonia is a rare event with only about 30 reported cases (Gottlieb and
Barnes, 1989). Table 2 shows the distribution of sites of infection in three
published series. In a briefly reported outbreak (LeMaistre et aI., 1985) in a
surgical intensive care unit, 37 patients were recorded as colonised, of whom
16 had overt clinical infection with a highly-resistant A. cal co acetic us
var.anitratus strain. The distribution of sites colonised/infected in this
one outbreak closely followed the general trend in consisting of 22
respiratory tract, 13 urinary tract, seven blood stream, six surgical wounds
and three other infections. A generally similar distribution was also found
during an outbreak in a neurosurgical ward (Allen and Green, 1987),
although the most common infections with clinical significance were those of
the respiratory tract.
CONCLUSIONS
Acinetobacter, usually A. calcoaceticus var.anitratus, is not a

frequent cause of hospital infection, but may be intractable because of
multiple antibiotic resistance. Infection may occur at any site, but is most
frequent in the respiratory tract, the urinary tract and surgical wounds. Both
varieties of Acinetobacter are found on human skin as normal residents;
however, especially when exposed to aerosols of Acinetobacter, skin may
become heavily colonised and subsequently act as the major source for
contamination or clinically evident infection of deeper tissue.
REFERENCES
Abrutyn, E., Goodhart, G.L., Roo's, K., Anderson R., and Buxton A., 1978,
Acinetobacter cal co acetic us outbreak associated with
peritoneal dialysis, Amer . .I. Epidemiol., 107:328.
Alexander, M., Rahman, M., Taylor, M., and Noble W.e., 1988, A study of
Acinetobacter calcoaceticus,.I. H osp. Infect., 12:273.
AI-Khoja, M.S., and Darrell, J.H., 1979, The skin as a source of
Acinetobacter and Moraxella species occurring in blood cultures,
.I. Clin. Pathol., 32:497.
Allen, K.D., and Green, H.T., 1987, Hospital outbreak of multiresistant
Acinetobacter anitratus, an airborne mode of spread'?,
.I. Hosp. Infect., 9:110.
Ayliffe, G.A.J., and Lowbury, E.J.L., 1982, Airborne infection in hospital,
59
Bergogne-Berezin, E., and Joly-Guillou, M.L., 1985, An underestimated
nosocomial pathogen Acinetobacter calcoaceticus,
I. Antimicrob. Chemother., 16:535.
Bergogne-Berezin, E., Joly-Guillou, M.L., and Vieu, J.F. 1987, Epidemiology
of nosocomial infections due to Acinetobacter calcoaceticus,
I. Hasp. Infect., 10:105.
Buisson, Y., Tran van Nhien, G., Ginot, L., Bouvet B., Schill, H., Driot, L.,
and Meyran, M., 1990, Nosocomial outbreaks due to amikacin-
resistant tobramycin-sensitive Acinetobacter species: correlation
with amikacin usage, I. Hasp. Infect., 15:88.
Buxton, A.E., Anderson, R.L., Werdegar, D., and Atlas, E., 1978,
Nosocomial respiratory tract infection and colonization with
Acinetobacter calcoaceticus. Epidemiologic characteristics,
Amer. I. Med., 65:507.
Carl quist, J.F., Conti, M., and Burke J.P., 1982, Progressive resistance in a
single strain of Acinetobacter calcoaceticus recovered during
a nosocomial outbreak, Amer. I. Infect. Control, 10:43.
Cordes, L.G., Brink, E.W., Checko, P.J., Lentnek, A., Lyons, R.W., Hayes,
P.S., Wu, T.e., Tharr D.G., and Fraser, D.W., 1981, A cluster of
Acinetobacter pneumonia in foundry workers, Ann. Int. M ed.,
95:688.
Cunha, B.A., Klimek, J.J., Gracewski, J., McLaughlin, J.e., and Quintiliani,
R., 1980, A common source outbreak of Acinetobacter pulmonary
infection traced to Wright respirometers, Postgrad. M ed. I.,
56:169.
Dijkshoorn, L., Wubbels, J.L., Beunders, A.J., Degener, J.E., Boks, A.L., and
Michel, M.F., 1989, Use of protein profiles to identify
Acinetobacter calcoaceticus in a respiratory unit,
I. Clin. Pathol., 42:853.
anitratus: epidemiology and control, I. Hasp. Infect., 1:125.
Galvao, e., Swartz, R., Rocher, L., Reynolds, J., Starmann, B., and Wilson,
D., 1989, Acinetobacter peritonitis during chronic peritoneal
dialysis, Amer . .I. Kidney Dis., 14:101.
Garcia, I., Fainstein, V., Le Blanc, B., and Bodey G.P., 1983, In vivo activities
of new B-lactam antibiotics against Acinetobacter spp.,
Antimicrob. Agents Chemother., 24:297.
Gaughan, M., White, P.M., and Noble, W.e., 1979, Skin as a source of
Acinetobacter/Moraxella species,.I. Clin. Pathol., 32:1193.
Gerner-Smidt, P., Hansen, L., Knudsen, A., Siboni, K., and Sogaard, I., 1985,
Epidemic spread of Acinetobacter calcoaceticus in a
neurological department analysed by electronic data processing,
Glew, R.H., Moellering, R.C., and Kunz, L.J., 1977, Infection with
Acinetobacter calcoaceticus (Herellea vaginicola): clinical
and laboratory studies, Medicine, 56:79.
Gottlieb, T., and Barnes, 0.1., 1989, Community acquired acinetobacter
pneumonia, Aust. NZ . .I. Med., 19:259.
60
Hoffman, S., Mabeck, C.E., and Vejlsgaard, R., 1982, Bactiuria caused by
Acinetobacter calcoaceticus biovars in a normal population
and in general practice, l. Clin. Microbiol., 16:443.
Holton, J., 1982, A report of a further hospital outbreak caused by a multi
resistantAcinetobacter anitratus, l. Hosp. Infect., 3:305.
Hoppe, M., Potel, J., and Malottke, R., 1983, Clinical importance of
A. calcoaceticus isolations from blood cultures and vein
catheters, Zent. Bakt. Hyg. A., 256:80.
Kelkar, R., Gordon, S.M., Giri, N., Rao, K., Ramakrishnan, G., Saikia, T.,
Nair, C.N., Kurkure, R.A., Pai, S.K., Jarvis, W.R., and Advani,
S.H., 1989, Epidemic iatrogenic Acinetobacter spp. meningitis
following administration of intrathecal methotrexate,
l. Hosp. Infect., 14:233.
Kloos, W.E., and Musselwhite, M.S., 1975, Distribution and persistence of
Staphylocccus and Micrococcus species and other aerobic
bacteria on human skin, Appl. Microbio!., 30:38l.
Larson, E., 1984, A decade of nosocomial acinetobacter,
Amer. l. Infect. Control, 12:14.
LeMaistre, A., Mortensen, J., La Rocco, M., and Robinson, A., 1985, An
outbreak of multiply resistant Acinetobacter anitratus in an
intensive care unit, Clin. Microbiol. Newslett., 7:102.
Muller-Serieys, c., Lesquoy, J .B., Perez, E., Fichelle, A., Boujeois, B., Joly-
Guillou, M.L., and Bergogne-Berezin, E., 1989, Nosocomial
Acinetobacter infections. Epidemiology and therapeutic
problems, Press. M ed., 18: 107.
Musa, E.K., Desai, N., and Casewell, M.W., 1990, The survival of
Acinetobacter calcoaceticus inoculated on fingertips and on
formica, l. Hosp. Infect., 15:219.
Ng, P.c., Herrington, R.A., Beane, C.A., Ghonheim, A.T.M., and Dear,
P.R.F., 1989, An outbreak of acinetobacter septicaemia in a
neonatal intensive care unit, l. H osp. Infect., 14:363.
Obana, Y., Nishino, T., and Tanino T., 1985, I n vitro and in vivo activities
l. Antimicrob. Chemother., 15:441.
Retailliau, H.F., Hightower, A.W., Dixon, R.E., and Allen, J.R., 1979,
unusual seasonal pattern, l. Infect. Dis., 139:371.
Sakata, H., Fujita, K., Maruyama, S., Kakehashi, H., Mori, Y., and Yoshioka,
H., 1989, Acinetobacter calcoaceticus biovar anitratus
septicaemia in a neonatal intensive care unit: epidemiology and
control,l. Hosp. Infect., 14:15.
Sheretz, R.J., and Sullivan, M.L., 1985, An outbreak of infection in burn
patients: contamination of patients' mattresses, l. Infect. Dis.,
151:252.
Smego, R.A., 1985, Endemic nosocomial Acinetobacter calcoaceticus
bacteraemia. Clinical significance, treatment and prognosis,
Arch. Int. Med., 145:2174.
Smith, P.W., and Massonari, R.M., 1977, Room humidifiers as the source of
Acinetobacter infections, l. Amer. M ed. Assoc., 237:795.
61
Snydman, D.R., Maloy, M.F., Brock, S.M., Lyons, R.W., and Rubin, S.J.,
1977, Pseudobacteraemia: false-positive blood cultures from mist
tent contamination, Amer. 1. Epidemiol., 106:154.
Somerville, D.A., and Noble, W.c., 1970, A note on the Gram-negative
bacilli of human skin, Eur. 1. Clin. Biol. Res., 40:669.
Stone, l.W., and Das, B.C., 1985, Investigation of an outbreak of infection
with Acinetobacter calcoaceticus in a special care baby unit,
Taplin, D., Rebell, G., and Zaias, N., 1963, The human skin as a source of
Mima-Herella infections,.I. Amer. M ed. Assoc., 186:952.
Vila, J., Almela, M., and Jiminez de Anta, M.I., 1989, Laboratory
multiresistant Acinetobacter calcoaceticus subsp.anitratus
strains,.I. Clin. Microbio!., 27:1086.
Vivian, A., Hinchcliffe, E., and Fewson, C.A., 1981,
Acinetobacter calcoaceticus: some approaches to a problem,
62
EPIDEMIOLOGY OF ACINETOBACTER STRAiNS ISOLATED
FROM NOSOCOMIAL INFECTIONS IN FRANCE
M.L. Joly-Guillou", E. Bergogne-Berezin'and J.F. Vieub
"Laboratoire de Microbiologie, CHU Bichat

46 rue Huchard, 75877 Paris Cedex 18, France
bInstitut Pasteur, rue du Dr.Roux

75724 Paris Cedex 15, France
INTRODUCTION
Acinetobacter has emerged as an important nosocomial pathogen

during the last 20 years and, as described by Noble (this volume), is now
frequently involved in outbreaks of hospital infections or colonisations. The
increasing importance of infections caused by Acinetobacter, and the
multi-resistance of the strains involved (Bergogne-Berezin and Joly-Guillou,
this volume), suggests that studies of the epidemiology and resistance
phenotypes will be required for the prevention of further infections with this
organism. Nosocomial infections with Acinetobacter have been a
particular problem in France. This article reviews our experience to date
with this organism.
EPIDEMIOLOGY OF HOSPITAL INFECTIONS CAUSED BY

ACINETOBACTER
Taxonomy of Acinetobacter Isolates from French Hospitals
The current taxonomic situation of the genus Acinetobacter has

been reviewed by Grimont (this volume). Using the taxonomic scheme
proposed by Bouvet and Grimont (1986), 84-90% of Acinetobacter isolates
from patients in French hospitals can currently be identified as
A. baumannii. Remaining isolates consist primarily of A. haemolyticus,
A.twoffii, A.johnsonii, A.junii, or unnamed species, especially
genospecies 3 (Bouvet and Grimont, 1987; Joly-Guillou and Bergogne-
BeH!zin, 1990a).
63
Environmental Sources and Human Carriage
Strains of A.baumannii are widespread in nature. They can be

isolated from soil and water samples (Baumann, 1968), and also form part of
the bacterial flora found on the skin of healthy adults; indeed, up to 25% of
normal adults may carry Acinetobacter spp. on their skins (Rosenthal and
Tager, 1975). The human skin is thus an important reservoir in hospitals for
Acinetobacter and has been implicated in several outbreaks of hospital
infection (French et aI., 1980; Holton, 1982). Other outbreaks of nosocomial
infection (reviewed by Noble, this volume) have often occurred in connection
with the use of respiratory equipment. Acinetobacter has frequently been
associated with nosocomial respiratory tract infection in France and
represented 18% of the organisms isolated following fibre-optic
bronchoscopy in French hospitals (unpublished data). Several studies
(Buxton et aI., 1973; Cunha et aI.. 1980; Hartstein et aI., 1988) and our own
data (Vieu et aI., 1980) have also implicated ventilatory equipment as an
environmental reservoir. Thus in 1980, six patients in a surgical intensive
care unit at the Bichat Hospital (Paris) acquired acinetobacter pneumonia
following contamination of a valve respirator.
Acinetobacter spp. other than A.baumannii are normally not

responsible for infections in patients, but are frequently found in the
hospital environment. They are often found on human skin and are
frequently carried by staff. These species are generally more susceptible to
antibiotics (Freney et a!., 1989).
Acinetobacter Infections
Although most isolates of Acinetobacter from clinical specimens

reflect colonisation rather than infection (Rosenthal, 1974), this bacterium
has been isolated from various types of opportunistic infection (reviewed by
Noble, this volume). Various factors predisposing to severe infections, such
as underlying disease, malignancy. burns or major surgery, have been
identified and are common to many other pathogens. In a French hospital
(Bichat Hospital, Paris), acinetobacters represented 1.85% of bacteria
isolated from hospitalised patients. The distribution by site of
Acinetobacter infections did not differ from other nosocomial Gram-
negative bacteria. Thus, among 1,800 acinetobacters isolated between 1980
and 1989 in France, 63% were from intensive care units (surgical, medical
and burn units), while 32% came from wounds, 30% from urines, 18% from
respiratory tract infections and 13% from blood cultures (Joly-Guillou and
Bergogne-Berezin, 1985, 1990a).
Nosocomial respiratory tract infection with Acinetobacter spp.

seems to be one of the most frequent clinical manifestations of
Acinetobacter infection in France, particularly in medical intensive care
units. In a study of nosocomial pneumonia in patients receiving continuous
mechanical ventilation in a medical ICU of Bichat Hospital, Acinetobacter
and Pseudomonas aeruginosa were the most frequently isolated Gram-
64
negative bacilli (15 and 31 % respectively) from protected pulmonary brushes
(Fagon et a!., 1989).
TYPING OF ACINETOBACTER IN FRENCH HOSPITALS
The various typing methods available for Acinetobacter have been

reviewed by Bouvet (this volume). This section describes the application of
some of these methods to isolates of Acinetob ac ter from French hospitals.
Antibiogram Typing
Antibiogram typing has proved useful in differentiating outbreak

strains when there are marked similarities in resistance patterns (Bartoli et
a!., 1986). A recent phenotypic analysis of the distribution of fl-Iactamases
in Acinetobacter strains (Joly-Guillou et a!., 1988) has been applied as an
epidemiological marker to define outbreaks of nosocomial cross infections.
However, the main limitation of this method as an epidemiological tool is its
lack of specificity, as several strains with the same resistance pattern appear
to be different. A progressive increase in resistance to commonly used
antimicrobial drugs has been observed during the last few years in French
hospitals (Godineau-Gauthey et a!., 1988; Joly-Guillou et a!., 1990a), and
has also been reported in other countries (Castle et a!., 1978; Murray and
Moellering, 1979; Vila et aI., 1989). In a recent study, 34% of the strains
tested were resistant to all fl-Iactams, with the exception of imipenem,
together with the aminoglycoside and quinolone antibiotics commonly used
in hospitals. Thus, while antibiogram typing is important to detect new
resistance phenotypes, it is not a reliable epidemiological marker and
utilisation of phenotypic analysis requires a complementary typing system.
Biotyping
The phage-typing system developed at the Pasteur Institute by Vieu et

a!. (1970) has been modified so that it currently recognises 125 phage-types
and 25 sub-types. Outbreaks of Acinetobacter infection seem to be
associated with a limited number of phage-types, whereas in the environment
a large number of phage-types are present. In a recent investigation of 117
acinetobacters isolated from outbreaks of infection in intensive care units of
a French Hospital (Bichat Hospital, Paris) during 1987 and 1988, phage-
types 17 and 124 seemed to be particularly important. Although these phage-
types were not distributed as a function of geographical origin or site of
infection, a significant difference in their distribution between 1987 and
1988 was observed. Phage-types such as phage-type 17 seem to be ubiquitous
and are found in various countries (Santos-Ferreira et aI., 1984). Other
phage-types seem to be limited to a single region or hospital, as seems to be
the case for phage-type 124 at Bichat Hospital. Using a combination of
biotyping and phage-typing, we have been able to establish the importance of
three sets of markers: biotype 9 / phage-type 124 (30%); biotype 6 / phage-
type 17 (26%); biotype 6 / phage-type 124 (15%) (Joly-Guillou et a!., 1990b).
65
These sets of markers were, generally, also associated with multi-resistant
phenotypes. The combination of these typing systems has demonstrated that
patients can be colonised with several different strains, but that only a few
strains are endemic and regularly isolated from infections.
Other Typing Systems
Plasmid analysis. This method has, to date, not been widely used in
French hospitals. A study in progress (M.L. Joly-Guillou, unpublished
results) has demonstrated the presence of six different plasmid profiles
among 15 Acinetabacter isolates, with plasmids between 15 and 80 Md in
size, but further evaluation of this technique will be required.
Bacteriocin typing. Andrews (1986) has described a technique for

typing acinetobacters by bacteriocin production. An investigation of 138
Acinetabacter strains isolated during 1987 at Bichat Hospital, Paris (A.
Bauernfeind, unpublished results), provided rather inconclusive results. Of
the nine bacteriocin types investigated, seven (numbers, 1,3,4,5,6,7 and 8)
were produced by 15-46% of the strains, while the remaining two types were
produced by 2% and 3% of the strains respectively. 6.5% of the strains were
untypable. The strains produced between one and eight bacteriocin types,
and a total of 31 patterns were identified. Several predominant patterns (4,
7,4+6,3+4+6 and 1 +5 +8) represented 60% of the strains.
Other suggested typing systems (reviewed by Bouvet, this volume)

have, to our knowledge, yet to be evaluated in French hospitals.
CONCLUSIONS
A number of epidemic outbreaks of infection caused by

Acinetobacter spp. have occurred during recent years in French hospitals,
particularly among immunocompromised patients (e.g. those suffering from
burns) and in intensive care units. Epidemiological investigations have been
hampered by the lack of a convenient typing scheme. From our experience it
seems that a combination of the various typing methods described above will
be required for epidemiological purposes. Biotyping and antibiogram typing
are methods that can easily be employed in routine microbiological
laboratories. Other possible typing systems (plasmid profiles, envelope
proteins, bacteriocins, serotypes etc.) do not seem, at the present time, to be
suitable for routine use in hospital laboratories.
REFERENCES
Andrews, H.J., 1986, Acinetobacter bacteriocin typing, J. Hasp. Infect.,

7:169.
Bartoli, M., Leguenedal, R., Hengy, c., Courrier, P., Thonnon, J., and
Courtois, D., 1986, Recherche de l'identite des souches
66
d'Acinetobacter calcoaceticus responsables d'infections
hospitalieres par antibiotypie a l'aide du traitement informatique,
Med. Mal. Infect., 1:46.
Baumann, P., 1968, Isolation of Acinetobacter from soil and water, 1.
Bacteriol., 96:39.
calcoaceticus and Acinetobacter lwoffii, Int . .1. Syst.
Bacteriol.,36:228.
clinical isolates of Acinetobacter, Ann. Inst. Past. Microbiol.,
138:569.
Buxton, A.E., Anderson, R.L., Werdegar, D. and Atlas, E., 1973, Nosocomial
respiratory tract infection and colonization with Acinetobacter
calcoaceticus. Epidemiologic characteristics, Amer . .1. M ed.,
65:507.
Castle, M., Tenney, J.H., Weinstein, M.P., and Eickhoff, T.C., 1978,
Outbreak of a multiply-resistant Acinetobacter in a surgical
intensive-care unit: Epidemiology and control, Heart Lung,
7:641.
Cunha, B.A., Klimeck, J.J., Gracewski, J., McLaughlin, J.c., and Quintiliani,
R., 1980, A common source outbreak of Acinetobacter
pulmonary infections traced to Wright respirometers, Postgrad.
M ed. 1.,56: 169.
Fagon, J.Y., Chastre, 1., Domart, Y., Trouillet, J.L., Pierre, 1., Dame, c.,
and Gibert, c., 1989, Nosocomial pneumonia in patients receiving
continuous mechanical ventilation, Amer. Rev. Resp. Dis.,
139:877.
French, G.L., Casewell, M.W., Roncoroni, A.J., Knight, S. and Phillips, I.,
1980, A hospital outbreak of antibiotic resistant Acinetobacter
anitratus: epidemiology and control,.1. H osp. Infect., 1:125.
Freney, J., Bouvet, P.J.M., and Tixier, C., 1989, Identification et
cliniques d'Acinetobacter autres que A. baumannii, Ann. BioI.
elin., 47:41.
Godineau-Gauthey, N., Lesage, D., Tessier, F., Kollia, D., and Daguet, G.L.,
1988, Acinetobacter calcoaceticus variete anitratus ou
A c i net 0 b act e r b a u man n ii, e t u de del a sen sib iii tea ux
antibiotiques de 65 souches hospitalieres, Med. Mal. Infect., 2
bis: 124.
Hartstein, A.I., Rashad, A.L., Liebler, J.M., Actis, L.A., Freeman, J.,
Rourke, J.W., Stibolt, T.B., Tomalsky, M.E., Ellis, G.R., and
Crosa, J.H., 1988, Multiple intensive care unit outbreak of
Acinetobacter calcoaceticus subspecies anitratus respiratory
infection with contaminated, reusable ventilator circuits and
resuscitation bags, Amer. 1. M ed., 85:624.
67
resistant Acinetobacter anitratus, 1. Hosp. Inject., 3:305.
Joly-Guillou, M.L., and Bergogne-BeH!zin, E., 1985, Evolution
d'Acinetobacter calcoaceticus en milieu hospitalier de 1971 11
1984,Press. M ed., 14:2331.
Joly-Guillou, M.L., Vallee, E., Bergogne-Ben!zin, E., and Philippon, A.,
1988, Distribution of betalactamases and phenotype analysis in
clinical strains of Acinetobacter, 1. Antimicrob. Chemother.,
22:597.
J oly-Guillou, M.L., Bergogne-Berezin, E., and Vieu, J.F., 1990a,
Epidemiologie et resistance des Acinetobacter en milieu
hospitalier. Bilan de 5 annees, Press.M ed., 8:358.
Joly-Guillou, M.L., Bergogne-Berezin, E., and Vieu, J.F. 1990b, A study of
relationships between antibiotic resistance phenotypes, phage-
typing and biotyping of 117 clinical isolates of Acinetobacter
spp., 1. Hosp. Inject., in press.
Murray, B.E., and Moellering, R.C., 1979, Aminoglycoside-modifying
enzymes among clinical isolates of Acinetobacter
calcoaceticus subsp. anitratus (H erellea vaginicola):
explanation for high level aminoglycoside resistance,
Rosenthal, S., 1974, Sources of Pseudomonas and Acinetobacter
species found in human culture materials, Amer. 1. Clin. Path.,
62:807.
Rosenthal, S., and Tager, I.B., 1975, Prevalence of Gram-negative rods in the
normal pharyngeal flora, Ann. Int. Med., 83:355.
Santos-Ferreira, M.O., Vieu, J.F., and Klein, B., 1984, Phage-types and
susceptibility to 26 antibiotics of nosocomial strains of
Acinetobacter isolated in Portugal, 1. Intern. Med. Res.,
12:364.
lysotypie d'Acinetobacter, Ann. Microbiol. Inst. Past.,
130A:403.
Vieu, J.F., Bergogne-Berezin, E., Joly, M.L., Berthelot, G., and Fichelle, A.,
1980, Epidemiologie d'Acinetobacter calcoaceticus, N ouv.
Press. M ed., 46:3551.
Vila, J., Almela, M., and Jimenez de Anta, T., 1989, Laboratory
multiresistant Acinetobacter calcoaceticus subsp. anitratus
strains,l. Clin. Microbiol., 27:1086.
68
MOLECULAR EPIDEMIOLOGICAL ANALYSIS OF
NOSOCOMIAL ACINETOBACTER BAUMANNII ISOLATES
J. Vila', M.A. Canales', M.A. Marcos", R. Gomez-Lus"

and M.T. Jimenez de Anta'
'Departmento de Microbiologia
Hospital Clinic de Barcelona
Universidad de Barcelona
Barcelona, Spain
bDepartmento de Microbiologia
Universidad de Zaragoza
Zaragoza, Spain
INTRODUCTION
The incidence of nosocomial infections caused by Acinetobacter

calcoaceticus var. anitratus, now designated as A. baumannii, has been
steadily rising in recent years (Mayer and Zinner, 1985; Bergogne-Berezin et
aI., 1987). Several outbreaks of nosocomial infections caused by multiply-
resistant strains of A. baumannii have been documented (e.g. Castle et aI.,
1978; Holton, 1982; Vila et aI., 1989). Many of these outbreaks have
occurred in intensive care units where extensive use of antibiotics can select
for the emergence of multiply-resistant strains (Bergogne-Berezin and J oly-
Guillou, this volume; Noble, this volume). The various methods which are
available for typing acinetobacters have been reviewed elsewhere (Bouvet,
this volume). The purpose of this short report is to describe and compare
the results obtained with some of these methods when used to determine the
molecular epidemiology of A. baumannii isolates endemic in a Spanish
hospital.
BIOTYPING
The 49 strains included in this study were isolated in our laboratory

during the period September 1988 to August 1989 from a wide variety of
clinical specimens obtained from patients admitted to the hospital. Repeat
isolates from the same patient were excluded.
69
Table 1. Biochemical characteristics of the two A. baumannii groups
Test Group One Group Two
Morphology CC CC
Growth at:
44°C + +
41°C + +
37°C + +
Glucose fermentation
Glucose oxidation + +
Haemolysis
Nitrate reduction
Indole
Arginine dihydrolase
Urea, Christensen
Esculin hydrolysis
Gelatin hydrolysis
B-galactosidase
Assimilation of:
Glucose
Arabinose + +
Mannose
Mannitol
N-Acetylglucosamine
Maltose
Gluconate
Caprate + +
Adipate +
Malate + +
Citrate + +
Phenylacetate +
Catalase + +
Oxidase
CC, coccobacillus; - negative reaction; +, positive reaction
Presumptive identification of the genus Acinetobacter was on the

basis of the following characteristics: Gram-negative coccobacilli; strictly
aerobic; non-motile; catalase-positive; oxidase-negative. The commercial
API 20NE microsystem was used for further identification. Supplementary
tests to identify A. baumannii were as follows (Bouvet and Grimont, 1986):
growth at 37o C, 41 C and 44 o C; lack of haemolysis.
0
The API 20NE system identified all 49 isolates as A. calcoaceticus

var. anitratus, with two different analytical profile index numbers: 0001051
and 0001073. Further investigations (Table 1) showed that all the strains
70
were able to grow at 41° and 44°C, demonstrating that both groups were A.
baumannii. Thus two biotypes (Groups One and Two) could be defined on
the basis of these biochemical studies.
SDS-PAGE PROTEIN PATTERNS
A loopful of overnight growth from a MacConkey agar plate was

suspended in 10 ml of brain heart infusion broth and incubated at 37°C for
18 h on an orbital shaker. Cells were harvested by centrifugation and
processed with a slight modification of the procedure described by Alexander
et al. (1988). SDS-PAGE was performed by the method described by
Laemmli (1970), with a concentration of 12% acrylamide in the running gel.
Boiled samples (5 Itl) were applied to the gel and electrophoresed at a
constant current of 30 rnA in a Mini-Protean gel apparatus (Bio-Rad). Gels
were stained, either with Coomassie brilliant blue or by the silver stain
method. One-dimensional SDS-PAGE of whole-cell protein extracts from
the 49 A. baumannii strains produced patterns containing 30-40 discrete
bands with molecular sizes varying from 15 to 100 kd. Proteins < 15 kd in
size were not resolved by the electrophoretic conditions used in this study.
Fig. 1 illustrates the profiles obtained with 11 of the 49 strains. The total
protein profiles were similar when the gel was stained with Coomassie
brilliant blue, whereas a protein band with an apparent molecular size of 70
kd could be seen in all the strains belonging to Group Two when the gel was
stained by the silver stain method. Similar results were obtained with the
remaining 38 strains not shown in Fig. 1. Thus PAGE revealed two distinct
protein profiles that were related to the groups defined by biotyping.
A B
5 1 2 3 4 5 6 7 8 9 10 11 5 1 2 3 4 5 6 7 8 9 10 11
43-
31_
21-
Fig. 1. Electrophoretic protein patterns. Lanes 1,2,5,7,8,9 and 11

represent strains of A. baumannii belonging to Group One; lanes 3,4,6
and 10 represent strains of A. baumannii belonging to Group Two. Gels
were stained either with Coomassie brilliant blue (A) or by the silver
stain method (B). Molecular size standards are shown on lane S.
71
1 2 3 4 5 6 1 2 3 4 5 6
Fig. 2. Analysis of plasmid DNA. Plasmid DNA was separated by

electrophoresis in a 0.7% agarose gel at 30 V for 18 h (A) or 30 h (B).
Lanes: 1, HindIII-digested lambda DNA molecular size markers (kb); 2,
pBR322; 3, example of a strain from Group One (plasmid isolated using
large scale preparation method); 4, A. baumannii strain isolated in
another hospital; 5, example of a strain from Group Two; 6, example of
a strain from Group One (plasmid isolated using small scale preparation
method.
PLASMID PROFILES AND RESTRICTION DIGESTS OF TOTAL DNA
Plasmid DNA or total cellular DNA was isolated essentially as

described by Birnboim and Doly (1979) or Roussel and Chabbert (1978)
respectively. Plasmid profile analysis revealed two groups of A.
baumannii, with the patterns shown in Fig. 2. Four major plasmid bands
were identified in Group One strains when the plasmids were isolated by a
large scale preparation method (500 ml), while six major plasmid bands were
observed following isolation with a.mini-scale preparation (10 ml). No
plasmid bands were observed in Group Two strains by either method.
Restriction endonuclease digestion (HindIII, EeaRY or BamHI) of

genomic A. baumannii DNA yielded complex patterns that required careful
analysis. While similarity could be detected between strains of the same
group, the fingerprints were clearly different between the A. baumannii
groups defined previously by biotyping, PAGE/SDS whole-cell protein
patterns and plasmid profile analysis.
ANTIMICROBIAL SUSCEPTIBILITIES
Susceptibility to antimicrobial agents was determined by two methods:

(i) the disc diffusion method with Mueller-Hinton medium, as advocated by
the National Committee for Clinical Laboratory Standards (1984); (ii) MICs
72
Table 2. Antimicrobial susceptibility patterns of the two A. baumannii
groups
MICs (mg/l)
Antimicrobial agent
Group One Group Two
Ampicillin >32(R) >32(R)

Mezlocillin >256(R) 64(MS)
Piperacillin 256(R) 64(MS)
Ticarcillin >256(R) <16(S)
Azlocillin >256(R) 64(MS)
Amoxicillin/Clavulanic acid >16/8(R) 16/8(R)
Cephalothin >32(R) >32(R)
Cefamandole >32(R) >32(R)
Cefuroxime >32(R) >32(R)
Cefoxitin >32(R) >32(R)
Moxalactam >32(R) >32(R)
Cefotaxime >32(R) <8(S)
Cefoperazone 128(R) 16(S)
Ceftazidime 32(R) <8(S)
Cefsulodin >32(R) 32(R)
Ceftizoxime 64(R) <8(S)
Ceftriaxone 64(R) <8(S)
Chloramphenicol >16(R) >16(R)
Gentamicin >16(R) 8(MS)
Amikacin 64(R) <16(S)
Tobramycin >16(R) <4(S)
Netilmicin <l(S) >16(R)
Trimethoprim-sulphamethoxazole >16/304(R) >16/304(R)
Imipenem <0.5(S) <0.5(S)
Colistin <2(S) <2(S)
Ciprofloxacin <0.5(S) <0.5(S)
Ofloxacin l(S) <0.5(S)
R, resistance; S, susceptibility; MS,moderate susceptibility
were determined with the Sceptor system (BBL Microbiology Systems,

Cockeysville, Md.) as described previously (Vila et aI., 1989). The results
(Table 2) allowed the 49 strains to be clearly separated into two groups. The
more resistant group coincided with Group One established by the molecular
epidemiological markers described above. In this group, of the 27
antibiotics tested, only colistin. netilmicin. imipenem and the new
fluoroquinolones had some inhibitory activity. The antibiotic susceptibility
pattern of the Group Two isolates consisted of resistance to ampicillin,
cephalothin, cefamandole, cefuroxime, cefsulodin, chloramphenicol,
73
netilmicin and trimethoprim-sulphamethoxazole; moderate susceptibility to
mezlocillin, piperacillin and gentamicin, and susceptibility to the other
antibiotics tested.
DISCUSSION
The data generated with the different techniques for epidemiological

typing allowed us to sort the strains included in this study into two distinct
groups. Firstly, when the API ZONE system was used as a biotyping method,
two distinct biotypes were found. Growth at 41°C and 44°C showed that both
biotypes were A. baumannii. A biotyping system described by Bouvet and
Grimont (1987) has differentiated 17 biotypes for this species and we are
now trying to correlate these biotypes with Groups One and Two.
The use of SDS-PAGE to analyse either whole cell proteins or cell

envelope proteins of A. calcoaceticus has been described previously
(Alexander et aI., 1984; Dijkshoorn et aI., 1987; Alexander et aI., 1988).
Whereas biotyping was able to sub-divide the isolates into two groups, no
substantial difference in protein patterns was revealed when gels were
stained with Coomassie brilliant blue. However, two groups could be
differentiated when the silver stain method was used. Both groups coincided
with those previously identified by the API ZONE system. The same two
groups were also identified by plasmid profile analysis and antibiotic
susceptibility profiles. Finally, Allardet-Servent et al. (1989) have described
previously how fragments of DNA generated by restriction endonucleases
which cleave infrequently can be used for epidemiological investigations of
A. calcoaceticus outbreaks following separation by pulsed field gel
electrophoresis. In the present study, cleavage with three restriction
endonl!cleases, followed by separation of the DNA fragments on
conventional agarose gels was used. The fingerprints obtained were clearly
different between the two groups.
Despite its low virulence, A. calcoaceticus has often been described

as a cause of hospital infections (Noble, this volume). We have described
previously (Vila et al., 1989) an outbreak in our hospital caused by two
different multi-resistant A. calcoaceticus var. anitratus strains, which
diminished in importance during the period that the respiratory intensive
care unit was closed for repairs. The number of isolates of A.
calcoaceticus var. anitratus subsequently increased in number and
stimulated the investigations outlined in this study. The Group One strains
coincided with one of the groups described in our original study, whereas the
Group Two strains were totally different in terms of genetic and phenotypic
markers (data not shown). Similar isolates were found throughout the
hospital, and attempts are currently being made to identify the nosocomial
reservoirs.
There is no universally accepted method for the epidemiological

typing of A. baumannii outbreaks. A serotyping method has recently been
used for delineation of outbreaks caused by A. baumannii (Traub, 1989),
74
but serotyping demands a large number of antisera which are difficult and
expensive to prepare and standardise. In our study we have tried to perform
an epidemiological analysis which avoids cumbersome procedures and does
not require sophisticated equipment (e.g. an ultracentrifuge or pulsed field
gel electrophoresis apparatus). We have found that the combination of
genotypic and phenotypic methods described in this report is useful for
epidemiological studies of A. baumannii and believe that the methods can
be readily applied in any routine clinical microbiology laboratory.
ACKNOWLEDGEMENT
This research was supported by grant 88/0206 from DGCYT-Spain.
REFERENCES
Alexander, M., Ismail, F., lackman, P.l.H., and Noble, W.e. 1984,
numerical analysis of electrophoretic protein patterns, 1. M ed.
Microbiol., 18:55.
Alexander, M., Rahman, M., Taylor, M., and Noble, W.e., 1988, A study of
Acinetobacter calcoaceticus, 1. Hosp. Infect., 12:273.
Allardet-Servent, A., Bouziges, N., Carles-Nurit, M.l., Bourg, G., Gouby, A.,
and Ramuz, M., 1989, Use of low frequency cleavage restriction
endonucleases for DNA analysis in epidemiological investigations
of nosocomial bacterial infections, 1. Clin. Microbiol., 27:2057.
Bergogne-BeH:zin, E., loly-Guillou, M.L., and Vieu, 1.F., 1987,
calcoaceticus, 1. Hosp. Infect., 10:105.
Birnboim, H.e., and Doly, 1., 1979, A rapid alkaline extraction procedure for
screening recombinant plasmid DNA, Nucl. Acid Res., 7:1513.
Bouvet, P.l.M., and Grimont, P.A.D., 1986, Taxonomy of the genus
Acinetobacter johnsonii sp. nov., and Acinetobacter junii
sp. nov. ~nd emended description of Acinetobacter
calcoaceticus and Acinetobacter lwoffii, Int. 1. Syst.
Bact., 36:228.
Bouvet, P.l.M., and Grimont, P.A.D., 1987, Identification and biotyping of
Castle, M., Tenney, 1.M., Weinstein, M.P., and Eickhoff, T.C., 1978,
Outbreak of a multiply resistant A.cinetobacter in a surgical
intensive care unit. Epidemiology and control, Heart Lung,
7:641.
Dijkshoorn, L., Michel, M.F., and Degener, 1.E., 1987, Cell envelope protein
profiles of Acinetobacter calcoaceticus isolated in hospitals,
1. M ed. Microbiol., 23:313.
75
resistant Acinetobacter anitratus, J. Hasp. Infect., 3:305.
Laemmli, V.K., 1970, Cleavage of structural proteins during the assembly of
the head of bacteriophage T4, Nature, 227:680.
Mayer, K.H., and Zinner, J.H., 1985, Bacterial pathogens of increasing
significance in hospital-acquired infections, Rev. I nf ect. Dis.,
7 (SuppJ.3) :5371.
National Committee for Clinical Laboratory Standards, 1984, "Performance
Standards for Antimicrobial Disk Susceptibility Test, 3rd edn.,"
National Committ~e for Clinical Laboratory Standards, Villanova,
Pa.
Roussel, A.F., and Chabbert, Y.A., 1978, Taxonomy and epidemiology of
Gram-negative bacterial plasmids studied by DNA-DNA filter
hybridisation in formamide, J. Gen. Microbiol., 104:269.
Traub, W.H., 1989, Acinetobacter baumannii serotyping for delineation
of outbreaks of nosocomial cross-infection, J. Clin. Microbiol.,
27:2713.
Vila, J., Almela, M., and Jimenez de Anta, M.T., 1989, Laboratory
multiresistant Acinetabacter calcaaceticus subsp. anitratus
strains, J. Clin. Microbial., 27:1086.
76
FACTORS INFLUENCING THE VIRULENCE OF ACINETOBACTER
1.-L. Avril and R. Mesnard
Laboratoire de Bacteriologie-Virologie
Faculte de Medecine
Universite de Rennes
Avenue du ProLL.Bernard
35043 Rennes Cedex
France
INTRODUCTION
Members of the genus Acinetobacter can frequently be isolated from

healthy people and are also commonly present in soil and water as free-living
saprophytes. In spite of an increasing number of reports that
Acinetobacter induces respiratory and urinary tract infections and
septicaemia (Bergogne-Berezin et aI., 1987), there have been few
investigations of the nature of its pathogenicity. Although only one species,
A. calcoaceticus, is described in the latest edition of Bergey's Manual of
Systematic Bacteriology (Juni, 1984) the taxonomy of Acinetobacter has
recently been extensively revised (Grimont and Bouvet, this volume). A.
baumannii is now the most prevalent Acinetobacter species isolated from
clinical specimens, but for historical reasons this review will refer to
Acinetobacterwithout detailed consideration of individual species.
Virulent organisms exhibit pathogenicity when introduced into the

host in very small numbers. This property can be subdivided into
invasiveness, including the ability to enter, multiply and spread within the
host tissues, and toxigenicity, including the ability to produce toxic
substances.
This article considers the possible factors which may contribute to the
virulence of Acinetobacter.
77
INVASIVENESS
Polysaccharide Capsule
Invasiveness of bacteria may be associated with surface components

that protect the bacteria from phagocytosis. By comparing two sets of
mutant Acinetobacter strains with their corresponding parent strains, an
enzymic decapsulation study has demonstrated that polysaccharide capsules
can decrease the ability of Acinetobacter to adhere to hydrocarbon
(Rosenberg et ai., 1983a,b). This was presumably the result of making the
outer bacterial surface more hydrophilic. Similarly, the surface
hydrophobicity of human isolates of Acinetobacter also appears to play an
important role in their attachment properties. A recent study has shown that
hydrophobicity is higher in A. baumannii strains isolated from infected
catheters and tracheal devices than in environmental isolates (Boujaafar et
ai., 1990).
Adherence to Epithelial Cells and to Hydrocarbons
Acinetobacter RAG-I, a strain isolated following growth selection

on hydrocarbon, adhered to human epithelial cells when grown under
conditions which promoted its adherence to hexadecane (Rosenberg and
Rosenberg, 1981; Rosenberg et ai., 1981), but showed only poor adherence
to epithelial cells when grown under conditions which resulted in low affinity
towards hexadecane. A mutant derived from RAG-1 that was deficient in its
ability to adhere to hydrocarbon was similarly unable to adhere to epithelial
cells.
The production of fimbriae by strains of Acinetobacter was reported

by Henricksen and Blom (1975). These authors found two major types of
fimbriae: "thin" fimbriae, ca. 3 nm in diameter, and "thick" fimbriae, ca. 5 nm
in diameter. Presence of the thick fimbriae in Acinetobacter strains was
shown to correlate with twitching motility, but no physiological role was
proposed for the thin fimbriae. Acinetobacter RAG-1, a hydrocarbon-
degrading strain which adheres avidly to hydrocarbons, possesses numerous
thin fimbriae, ca. 3.5 nm in diameter, on the cell surface. A non-adherent
mutant of this strain lacks these fimbriae (Rosenberg et aI., 1982).
Adherence of wild-type RAG-1 cells to hexadecane was considerably
reduced after shearing treatment. In addition, RAG-l cells were
agglutinated in the presence of specific antibody, whereas the non-adherent
cells were not.
Enzymic Activities
Invasiveness may be aided by enzymes that favour the spread of

bacteria. Poh and Loh (1985) determined the enzymic profiles of clinical
isolates of Acinetobacter by conventional plate tests and the rapid API-
ZYM system. No activity could be detected for DNase, elastase, haemolysin,
protease or gelatinase in the majority of strains tested. These enzymes are
78
therefore unlikely to be involved in the virulence of Acinetobacter. In
contrast, significant levels of butyrate esterase (C-4), caprylate esterase (C-
8) and leucine arylamidase were detected in all isolates tested. These
enzymes hydrolyse short chain fatty acids and may therefore be involved in
causing damage to tissue lipids. No trypsin, chymotrypsin, alkaline-
phosphatase or glucosidase activities were present.
Iron Uptake
The ability of bacteria to assimilate iron appears to be related to

invasiveness. Some virulent strains of Enterobacteriaceae are known to
synthesise an hydroxamate type of iron chelator termed aerobactin.
However, in a study of 50 Acinetobacter clinical isolates (Martinez et aI.,
1987), no aerobactin-producing strai ns were detected.
TOXIGENICITY
The virulence of clinical isolates of Acinetobacter has been studied

in mice by Obana et al. (1985). Acute systemic infection was induced in mice
by intraperitoneal injection of Acinetobacter. The virulence of 40
Acinetobacter clinical isolates was expressed in terms of the number of
viable cells of the test strains required to kill 50% of the mice challenged.
Following intraperitoneal challenge, the LD,o values ranged from 103 .to
> 10 6 viable cells/mouse, but for most of the strains were> 10 6 viable
cells/mouse. However, the virulence of a few Acinetobacter strains was
found to be equivalent to the virulence of Escherichia coli, Serratia
marcescens and Pseudomonas aeruginosa. A particularly interesting
observation was that a considerable enhancement of virulence was obtained
if strains were injected as suspensions in 3% hog gastric mucin.
Slime
The virulence of five slime-producing strains was compared with the

virulence of three non-slime-producing strains in mice by Obana (1986).
However, no difference was observed between the two groups (LD50s of 107 -
108viable cells/mouse) and no correlation could be made between virulence
and the amount of slime produced.
Disease is as dependent on the host as on the infecting microorganism.

Infections caused by Acinetobacter are often nosocomial. The
experiments described above were therefore also performed with mice that
had been immunosuppressed with cyclophosphamide, but no significant
increase in virulence was observed.
Obana (1986) also studied the mortality of mice resulting from mixed
infection. Mortality was higher in the case of mixed infection with P.
aeruginosa and a slime-producing strain of Acinetobacter than in the case
of single infection with a variety of inoculum sizes. Even in a mixture with a
79
non-slime-producing strain, the virulence of P. aeruginosa was greater
than in single infection. Similar results were observed for mixed infections
with either E. coli or S. marcescens.
The chemical composition of the slime produced by Acinetobacter

was similar for two strains tested. In each case there was a high sugar
content and a substance giving a positive orcinol reaction was also present.
About 30% of the total composition remained to be determined, but the
DNA content was approximately 3%, a value which is significantly different
from the slime of P. aeruginosa in which the main component is DNA.
Nevertheless, it is known that Acinetobacter slime exhibits lethal activity
in mice, although intraperitoneal inoculation of mice with slime obtained
from Acinetobacter TMS 262 and Acinetobacter TMS 266 resulted in
LDso values of 1.8 and 0.09 mg/mouse respectively, thereby showing a
considerable difference between the two strains.
To study the effect of Acinetobacter slime on the virulence of

various other Gram-negative bacilli in mice, concurrent intraperitoneal
inoculations of bacteria and increasing doses of slime have been carried out.
Enhancement of virulence was proportional to the dose of slime inoculated,
with the slime of Acinetobacter causing a ten-fold or greater enhancement
in the lethal activity of E. coli, S. marcescens or P. aeruginosa.
The virulence-enhancing effect of slime has also been demonstrated by

intraperitoneal inoculation into mice of a mixture of a sub-lethal dose of
Acinetobacter slime and the Gram-negative species mentioned above.
The bacterial counts in the peritoneal exudate showed a gradual decrease in
the absence of slime, but increased in the presence of slime with subsequent
death of the mice. This enhancement of virulence apparently resulted from
the cytoxicity of slime for mouse neutrophils and the inhibition of the
cellular migration of neutrophils into the peritoneal exudate of mice.
Toxins
Virtually all species of Gram-negative bacteria produce

lipopolysaccharide (LPS) components of their cell walls that can function as
endotoxins, of which lipid A is the active principle. Brade and Galanos
(1983) compared the biological activities of the LPS and lipid A from
Acinetobacter with those from Salmonella. The results showed that the
LPS and lipid A from Acinetobacter had toxic and other biological
properties similar to those found with LPS and lipid A from
Enterobacteriaceae. These properties included lethal toxicity in mice,
pyrogenicity in rabbits, complement inactivation in vitro, a local
Shwartzmann reaction, mitogenicity for mouse-spleen B lymphocytes and a
positive limulus amoebocytes lysate test. It therefore seems that the in-vivo
production of endotoxin is probably responsible for the characteristic signs
of disease and death observed following Acinetobacter septicaemia. No
protein toxins have so far been described in Acinetobacter.
80
ROLE OF HOST ANTIMICROBIAL SYSTEMS
During phagocytosis, ingested bacterial cells are exposed to high

concentrations of various bactericidal proteins released from
polymorphonuclear leukocyte (PMN) granules into phagolysosomes. Studies
of acid extracts from PMN granules, including fractionation of such extracts
by gel permeation chromatography, indicate that the bactericidal activity of
the various fractions depends on the particular organism tested. The acid
extract fraction from PMN granules with the lowest molecular size (peak D)
is particularly active against Acinetobacter (Greenwald and Ganz, 1987).
This fraction consisted of a mixture of three human defensin peptides:
HNPl, HNP2 and HNP3. Purified defensins reproduced the bactericidal
activity, suggesting that defensins could playa major role in the killing of
Acinetobacter by human neutrophils.
Because of its readily alterable membrane, Acinetobacter has been

used in studies designed to elucidate the role of bacterial cell envelope
structure in interactions with PMN s (Loeffelholz and Modrzakowski, 1987).
An alkane-induced increase in the outer membrane permeability of
Ac in e t 0 b act e r was concomitant with an enhanced susceptibi Ii ty to the
oxygen-independent antimicrobial activity of extracted contents from rat
PMN granules. The change in outer membrane permeability was not
associated with an alteration to LPS O-antigen. A further study by
Loeffelholz and Modrzakowski (1988) demonstrated that rat PMN granule
extract reduced the viability of Acinetobacter by inhibiting the transport
and incorporation of macromolecule precursors. The granule extract activity
that increased outer membrane permeability did not appear to be directly
responsible for the observed decrease in viability.
CONCLUSIONS
Knowledge of the factors influencing the virulence of Acinetobacter

is still at an elementary stage. Several of the characteristics considered in
this review seem to playa role in enhancing the virulence of particular
strains. However, on the basis of the available evidence, it can be suggested
that the observed pathogenicity of Acinetobacter does not normally result
from the simple presence of the organism, but is often related to the clinical
status of the infected patient. Thus virulence is greatly enhanced in patients
who are either compromised hosts or who have polymicrobial infections.
REFERENCES
Bergogne-Berezin, E., Joly-Guillou, M.L., and Vieu, l.F., 1987,

calcoaceticus, 1. Hosp. Inject., 10:105.
Boujaafar, N., Freney, l., Bouvet, P.l.M., and leddi, M., 1990, Cell surface
hydrophobicity of 88 clinical strains of Acinetobacter baumannii,
Res. Microbiol., 141:477
81
Brade, H., and Galanos, e., 1983, Biological actIvItIes of the iipopoly-
saccharide and lipid A from Acinetobacter calcoaceticus, 1. M ed.
Microbial., 16:211.
Greenwald, G.I., and Ganz, T., 1987, Defensins mediate the microbiocidal
activity of human neutrophil granule extract against Acinetobacter
calcoaceticus, Infect. Immun., 55:1365.
Henricksen, J., and Blom, J., 1975, Correlation between twitching motility
and possession of polar fimbriae in Acinetobacter calcoaceticus,
Acta Path. Microbial. Scand., B83:103.
"Bergey's Manual of Systematic Bacteriology, vol. 1," p.303, N.R. Grieg
and J.G. Holt, eds., Williams and Wilkins, Baltimore.
Loeffelholz, M.J., and Modrzakowski, M.e., 1987, Outer membrane
permeability of Acinetobacter calcoaceticus mediates
susceptibility to rat polymorphonuclear leukocyte granule contents,
Infect. Immun., 55:2296.
Loeffelholz, M.J., and Modrzakowski, M.C., 1988, Antimicrobial
mechanisms against Acinetobacter calcoaceticus of rat
polymorphonuclear leukocyte granule extract, I nf ec t. I mmun.,
56:522.
Martinez, J.L., Cercenado, E., Baquero, F., Perez-Diaz, J.e., and Delgado-
Tribarren, A., 1987, Incidence of aerobactin production in Gram-
negative hospital isolates, FEMS Microbial. Lett., 43:351.
Obana, Y., 1986, Pathogenic significance of Acinetobacter
calcoaceticus: analysis of experimental infection in mice,
Microbial. Immunol., 30:645.
Obana, Y., Nishino, T., and Tanino, T., 1985, In vitro and in vivo activities of
antimicrobial agents against Acinetobacter calcoaceticus, J.
Antimicrob. Chemother., 15:441.
Poh, e.L., and Loh, G.K., 1985, Enzymatic profile of clinical isolates of
Acinetobacter calcoaceticus, Med. Microbial. Immunol.,
174:29.
Rosenberg, M., and Rosenberg, E., 1981, Role of adherence in growth of
Acinetobacter calcoaceticus RAG-Ion hexadecane, 1.
Bacterial., 148:51.
Rosenberg, M., Perry, A., Bayer, E.A., Gutnick, D.L., Rosenberg, E., and
Ofek, T., 1981, Adherence of Acinetobacter calcoaceticus RAG-1
to human epithelial cells and to hexadecane, I nf ect. I mmun., 33:29.
Rosenberg, M., Bayer, E.A., Delarea, J., and Rosenberg, E., 1982, Role of
thin fimbriae in adherence and growth of Acinetobacter
calcoaceticus RAG-Ion hexadecane, Appl. Environ. Microbial.,
44:929.
Rosenberg, E., Gottlieb, A., and Rosenberg, M., 1983a, Inhibition of
bacterial adherence to hydrocarbons and epithelial cells by emulsan,
Infect. Immun., 39:1024.
Rosenberg, E., Kaplan, N., Pines, 0., Rosenberg, M., and Gutnick, D., 1983b,
Capsular polysaccharides interfere with adherence of Acinetobacter
calcoaceticus to hydrocarbon, FEMS Microbial. Lett., 17:157.
82
ANTIBIOTIC RESISTANCE MECHANISMS IN ACINETOBACTER
E. Bergogne-Berezin and M.L. Joly-Guillou
Laboratoire de Microbiologie
CHU Bichat
46 rue Huchard
France
INTRODUCTION
Antibiotic therapy for infectious diseases is now over 40 years

old. In the early years of the antibiotic era, Staphylococcus aureus,
Streptococcus pneumoniae and Il-haemolytic streptococci wcre the
predominant organisms involved in hospital-acquired infections. Since 1960
the continuous development and therapeutic introduction of new potent
antimicrobial agents has resulted in significant changes in the character of
nosocomial infections (Finland, 1979). In the early 1960s ~taphylococci
subsided in importance as the major cause of nosocomial infection and were
replaced by enteric Gram-negative bacilli, enterococci and fungi (Mayer and
Zinner, 1985). Since 1980, increased numbers of aerobic Gram-negative
bacilli, including Acinetobacter spp., have heen noted as major bacterial
agents of hospital infections (Glew et aI., 1977; Retailliau et aI., 1979;
Bergognc-Bcrezin and Joly-Guillou, 1985). The pathogenic role of
Acinetobacter is limited to nosocomial infections (Glew et aI., 1977;
Retailliau et al., 1979; Holton, 1982; Joly-Guillou and Bergogne-Berezin,
1985; Bergogne-Berezin et al., 1987) and its role in community-acquired
infection has only seldom been mentioned in the medical literature (Rudin
et al., 1979). The changing profile of nosocomial infections is not only
related to the sequential introduction of new antimicrobial agents, hut also
reflects various ecological factors, especially major changes in medical and
surgical technologies in intensive care units. In this context, one of the most
difficult problems confronting the clinician who deals with nosocomial
infections is that of microbial resistance to antibiotics (Finland, 1979;
McGowan,1983). Acinetobacter is a genus that has developed one of the
most impressive patterns of antibiotic resistance during the last 20 years
83
(Bergogne-Berezin et al., 1971; Pinter and Kantor, 1971; Devaud et a1.,
1982; Holton, 1982; Sanders, 1983; 10ly-Guillou and Bergogne-Berezin,
1985b; Shannon et aI., 1986; 10ly-Guillou et aI., 1987; Godineau-Gauthey
et aI., 1988; Freney et aI., 1989). Nosocomial isolates of Acinetobacter
currently exhibit antibiotic resistance profiles which result in infections that
are extremely difficult to treat. An increasing number of reports in the
medical literature document the high rates of resistance in Acinetobacter,
although there are few studies of the precise mechanisms of resistance to B-
lactams, aminoglycosides, f1uoroquinolones and the other major antibiotic
classes. Several studies of the genetic determinants of resistance in
Acinetobacter have been published, but current knowledge of transferable
resistance plasmids in this genus is still at an early stage. This article deals
with antibiotic resistance in,Acinetobacter spp.. A review of past and
current resistance profiles is followed by a description of known mechanisms
of resistance and current information on the genetic determinants of
resistance in Acinetobacter.
EVOLUTION OF RESISTANCE TO ANTIBIOTICS IN

ACINETOBACTER SPP
Overall Susceptibilities and Resistance in Acinetobacter spp
In the historical development of drug resistance among Gram-negative

bacilli, there has been a progressive decrease in susceptibility to commonly
used antimicrobial drugs among strains of Enterobacteriaceae and
Pseudomonas (McGowan, 1983; Mayer and Zinner, 1985); this has also
occurred with Acinetobacter (Bergogne-Berezin and 10Iy-Guillou, 1985).
In the early 19705, Acinetobacter infections were treated successfully with
gentamicin, minocycline, nalidixic acid, ampicillin or carbenicillin, either as
single agents or in combined therapy. In early in-vitro studies, the majority
of Acinetobacter clinical isolates were susceptible to ampicillin (61-70%),
gentamicin (92.5%), chloramphenicol (57%), or nalidixic acid (97.8%)
(Bergogne-Berezin et aI., 1971). Of 86 Acinetobacter strains isolated in
1969, only 3.4% were resistant to carbenicillin, which in those days was a new
B-Iactam under investigation (Bergogne-Berezin et aI., 1971). In the same
study, a comparison between strains isolated from medical wards (120
isolates) and those from surgery or intensive care units (120 isolates) showed
a significant difference in susceptibilities; the former showed higher
proportions of isolates susceptible to ampicillin (80% vs. 41.6%),
cephaloridine (65.8% vs. 29.1%) and streptomycin (84.1% vs. 42.5%). The
difference in gentamicin susceptibility between the two groups of strains was
more limited; indeed the vast majority of Acinetobacter isolates were
susceptible to gentamicin (95.8% of strains from medical wards; 87.5% of
those from intensive care units). Strains of A. calcoaceticus var.
anitratus, as identified before the current taxonomy became available
(Bergogne-Berezin et aI., 1987; Freney et a1., 1989), were significantly more
resistant than var. lwoffii strains in early studies (Bergogne-Berezin et al.,
1971; Pinter and Kantor, 1971), as well as in more recent investigations
84
(Bergogne-Bc€dzin and Joly-Guillou, 1985, 1989; Joly-Guillou and
Bergogne-Berezin, 1985b; Godineau-Gauthey et al., 1988; Freney et al.,
1989). A change in the isolation frequency of var. anitratus (A.
baumannii according to the new Acinetobacter species definition)
(Freney et al., 1989) has been noted, rising from 77.5% of the strains isolated
in 1971-1980 to 94.5% in 1984-85 (Joly-Guillou and Bergogne-Berezin,
1985b). Among various factors associated with increasing resistance of
Acinetobacter to antibiotics, the increased incidence of the more resistant
var. anitratus may also have changed the reported overall antibiotic
resistance of clinical isolates of Acinetobacter. All other named or
unnamed Acinetobacter spp. (A. lwoffii, A. johnsonii, A.
haemolyticus, A. junii) (Freney et al., 1989), which may be isolated from
the hospital environment, but are seldom involved in nosocomial infections,
are much more susceptible to most antibacterial agents (Joly-Guillou and
Bergogne-Berezin, 1985b; Bergogne-Berezin et al., 1987; Freney et al.,
1989). Thus most studies focus on the susceptibility or resistance of A.
calcoaceticusvar. anitratus (A. baumannii) strains that are responsible
for the majority of outbreaks of nosocomial infections associated with this
genus.
Current Resistance to Antibiotics in Acinetobacter spp
The association of increased antibiotic usage in hospitals with

generalised increases in antibiotic resistance is an accepted "fact" of the
current therapeutic era (McGowan, 1983; Mayer and Zinner, 1985).
Successive increases in the resistance of clinical strains of Acinetobacter
have been reported periodically (Bergogne-Berezin et al., 1971; Dowding,
1979; Joly-Guillou and Bergogne-Berezin, 1985b; Godineau-Gauthey et al.,
1988). High proportions of strains have become resistant to older
antibiotics and, at the present time, certain Acinetobacter strains are
resistant to most commonly used antibacterial drugs, including
aminopenicillins, ureidopenicillins, cephalosporins of the first (cephalothin)
and second (cefamandole) generation (Morohoshi and Saito, 1977; Joly-
Guillou and Bergogne-Berezin, 1985b). cephamycins such as cefoxitin
(Garcia et al., 1983; Joly-Guillou and Bergogne-Berezin, 1985b), most
aminoglycosides-aminocyclitols (Dowding, 1979; Devaud et al., 1982;
Goldstein et aI., 1983; Medeiros, 1984; Joly-Guillou andBergogne-Berezin,
1985b; Joly-Guillou et al., 1987) chloramphenicol and tetracyclines.
Contrasting data are sparsely distributed in the literature for some
antibiotics: minocycline and doxycycline were cited in one study as the most
active of 21 antibiotics tested, including ten B-Iactams, three aminoglycosides
and nalidixic acid (Obana et al., 1985), while piperacillin and
aminoglycosides were effective in a study of the in-vitro activity of 25
antibacterial agents (Garcia et al., 1983). These differences in the
susceptibilities of Acinetobacter isolates might be attributable to
environmental factors such as the different patterns of antibiotic usage that
exist in Japan (Obana et aI., 1985), the USA (Garcia et al.. 1983) or France
(Joly-Guillou and Bergogne-Berezin, 1985b). Table 1 summarises the
susceptibility of A. calcoaceticus var. anitratus (A. baumannii) to
85
Table 1. Susceptibility of A.calcoaceticus var. anitratus
to antimicrobial agents (MICs in mg/1)
Antibiotic No of MIC range MIC 50 MIC 90 Year Ref.

isolates
Cephalothin 51 l. 56->100 50 >100 1983 (1)

100 >128 >128 >128 1988 (2)
Cefoxitin 51 6.25->100 100 >100 1983 (1)

100 0.12->128 128 >128 1988 (2)
Cefuroxime 51 3.12->100 25 50 1983 (1)
Cefamando1e 51 3.12->100 100 >100 1983 (1)
Cefoperazone 51 6.25->100 100 >100 1983 (1)

100 4->128 44 >128 1988 (2)
Cefotaxime 40 0.12-128 l.0 16 1981 (3)

116 0.12-128 7.9 22 1986 (4)
100 0.12->128 16 100 1988 (2)
Moxa1actam 40 0.5-64 4.0 32 1981 (3)

51 l.56-100 50 100 1983 (1)
100 2.0->128 50 120 1988 (2)
Ceftazidime 40 0.25-32 l.0 8.0 1981 (3)

51 l.56-25 6.25 12.5 1983 (1)
116 0.12-32 4.4 10 1986 (4)
100 0.12-128 6.5 22 1988 (2)
Aztreonam 51 6.25->100 100 >100 1983 (1)

100 0.12->128 18 87 1988 (2)
Temocil1in 51 3.12->100 >100 >100 1983 (1)

100 >128 >128 >128 1988 (2)
Ticarcil1in 51 1.56->100 25 50 1983 (1)

100 l. 0->256 >256 >256 1988 (2)
Piperacil1in 51 3.12->100 12.5 50 1983 (1)

100 l.0->128 100 >128 1988 (2)
continued
86
Table 1.(continued)
Mezlocillin 51 3.12->100 2~ 100 1983 (ll
100 1.0->128 >128 >128 1988 (2)
Imipenem 40 0.01-0.25 0.12 0.25 1981 (3)

51 0.05-12.5 0.20 0.39 1983 (1)
134 0.01-4.0 0.28 0.61 1986 (4)
100 0.12->128 0.37 0.60 1988 (2)
54 0.5-4.0 2.0 2.0 1988 (5)
Erythromycin 51 12.5->100 >100 >100 1983 (1)

84 a 6.25 12.5 b 1985 (6)
Tetracycl ine 51 0.78->100 25 100 1983 (1)

84 a 3.13 6.25 b 1985 (6 )
Doxycycline 84 a <0.19 O.39 b 1985 (6)
Minocycline 84 a <0.1 <0.19 b 1985 (6)
Gentamicin 51 0.10-~0 0.39 3.12 1983 (1 )

54 1.0-128 64 128 1988 (5)
Tobramycin 51 0.10-32 0.39 0.78 1983 (1)

96 0.12->128 3.2 44 1986 (7)
54 1.0-32 16 32 1988 (5)
Amikacin 51 0.20-12.5 1. 56 3.12 1983 (1 )

96 0.25->128 4.6 74 1986 (7)
54 1. 0- 64 32 64 1988 (5)
a strains classified as "A.calcoaceticus"

b MIC concentration required to inhibit 80% of strains
80
References: (1) Garcia et al., 1983; (2) Jo1y-Guillou et a1., 1988;

( 3 ) Phillips et al. , 1981 (4)Bergogne-Berezin and Joly-Guil1ou, 1986;
(5) Chow et al., 1988; (6) Obana et al., 1985; (7) personal unpublished
data.
various antimicrobial agents. Combined data for each compound confirm

the general trend towards decreased susceptibility of strains as a function of
time. With some major antibiotics, such as the third generation
cephalosporins (e.g. cefotaxime or ceftazidime), imipenem, tobramycin and
amikacin, the data show clearly that the minimal inhibitory concentrations
(MICs) have increased between 1981 and 1988. These data will be analysed
in a subsequent section on mechanisms of resistance.
87
Multiresistance in Acinetobacter spp
Multiresistance has been documented since the first studies of

outbreaks of Acinetobacter infections. In our earliest experience
(Bergogne-Ben!zin et a!., 1971), several multiresistant strains were isolated
in a surgical ward prior to any initiation of antibiotic therapy. One strain was
resistant to streptomycin and tetracycline, while a second isolate was
resistant to ampicillin, cephaloridine and chloramphenicol. Other reports
of multiresistant strains can be found in the medical literature. Retailliau et
al. (1979) described nosocomial Acinetobacter strains that were resistant
to ampicillin, cephalosporins and chloramphenicol. French et al. (1980)
described an outbreak caused by a strain of A. anitratus in which the isolate
was resistant to 18 antibiotics. Devaud et al. (1982) characterised
multiresistant isolates of A. calcoaceticus which were resistant
simultaneously to ampicillin, gentamicin, kanamycin, streptomycin,
tetracycline, chloramphenicol and sulphamethoxazole, and were able to
subsequently identify the genetic markers and transfer the resistance
determinants to Escherichia coli K12. Similarly, in a strain of A.
calcoaceticus resistant to ampicillin, aminoglycosides, chloramphenicol,
sulphonamides and trimethoprim, the resistance genes could be transferred
to E. coli by conjugation (Goldstein et a!., 1983). More recently,
multiresistant A. calcoaceticus var. anitratus strains isolated from a
hospital outbreak of infection were resistant to 14 antibiotics and were
inhibited only by netilmicin and imipenem (Vila et aI., 1989). In our most
recent study (MuJler-Serieys et aI., 1989), the predominant phenotypes of
multiresistant Acinetobacter strains included resistance to penicillin:"
cephalosporins, gentamicin, amikacin (but not tobramycin), nalidixic acid,
trimethoprim and sulphamethoxazole. Strains with such a resistance pattern
were isolated in intensive care units, from respiratory nosocomial infections,
urinary tract infections and septicaemia. More details of the mechanisms of
resistance in these multiresistant strains will be provided in a subsequent
section.
MECHANISMS OF RESISTANCE TO B-LACTAMS IN

ACINETOBACTER SPP
After the first recognised outbreaks of Acinetobacter infections

(Bergogne-Berezin et aI., 1971; Pinter and Kantor, 1971), it was noted that
isolates became rapidly resistant to aminopenicillins and cephalosporins of
the first (cephalothin, cephaloridine) and second (cefuroxime, cefoxitin,
cefamandole) generations (Retailliau et aI., 1979; Garcia et aI., 1983;
Bergogne-Ben!zin and loly-Guillou, 1985). Indeed, it has been stated that
Acinetobacter is "intrinsically insensitive" (Goldstein et aI., 1983) to most
B-Iactams, especially ampicillin and cephalosporins. However, as mentioned
above, early studies with ampicillin and carbenicillin (Bergogne-Ben!zin et
aI., 1971), and later with ureidopenicillins such as piperacillin (Philippon et
a!., 1980; Devaud et a!., 1982; loly-Guillou and Bergogne-B€n!zin, 1985b),
showed that these compounds were active against 70-80% of Acinetobacter
88
strains. At a later stage, when Acinetobacter became resistant to early B-
lactams, the third generation cephalosporins were initially active,
particularly cefotaxime (Phillips et aI., 1981) and ceftazidime (Philipps et aI.,
1981; Bergogne-Berezin and Joly-Guillou, 1986). However, within a few
years, a rapid increase in resistance to carboxypenicillins and, to a lesser
degree, ureidopenicillins and third generation cephalosporins, led several
authors to investigate the possible production of inactivating enzymes by
Acinetobacter strains.
Penicillinases in Acinetobacter spp
The commonest mechanism of resistance to penicillins is inactivation·

of these agents by B-Iactamases encoded either by the chromosome or by
plasmids (Medeiros, 1984; Bauernfeind, 1986). 13-Lactamases have been
detected in almost all Gram-positive and Gram-negative bacteria, including
Acinetobacter spp. Resistance of Acinetobacter to ampicillin,
carboxypenicillins and ureidopenicillins has been attributed to the presence
of a penicillinase. Both TEM-l (Philippon et aI., 1980; Goldstein et aI.,
1983) and TEM-2 (Devaud et aI., 1982) 13-Iactamases have been found in
representative isolates of epidemic strains that also produced
aminoglycoside-modifying enzymes. The genes encoding resistance to
penicillins (as well as those encoding resistance to aminoglycosides,
tetracyclines, chloramphenicol and sulphonamides) have been identified on
plasmids (see subsequent section on genetic determinants). In our own
study (Joly-Guillou et aI., 1988), 100 Acinetobacter strains selected from
clinical strains isolated in our hospital between 1980 and 1986 (including five
strains of var. twoffii and three ATCC strains) were analysed for their 13-
lactamase activities. B-Lactamase activity was detected in crude extracts by
the gel iodometric method (Labia and Barthelemy, 1979) in the presence of
substrate inhibitors. Two 13-Iactamase inhibitors were used to determine the
inhibitory profiles of 13-Iactamase specific activities (Philipp on et aI., 1980):
clavulanic acid, which specifically inhibits penicillinases of the TEM type,
and cloxacillin, which inhibits cephalosporinases. Isoelectric focusing of
crude enzyme extracts was performed according to a standard procedure
(Matthew and Harris, 1976): analytical isoelectric focusing was carried out
on sheets of polyacrylamide gel at 200-400 V (for about 18 h), with 30 JLI
samples (extracts) applied near the anode on blotting paper discs (for more
details see Matthew and Harris, 1976; Joly-Guillou et aI., 1988). The
enzymes were detected with benzylpenicillin as a substrate by the iodometric
overlay method; their isoelectric point (pI) was determined by comparison
with enzymes of known pI, as described previously (Joly-Guillou et aI., 1988).
The results showed that the 100 Acinetobacter strains divided into two
populations with different ticarcillin susceptibilities: 24 susceptible strains
(geometric mean MIC < 32 mg/I) and 76 resistant strains (geometric mean
MIC > 256 mg/I). Analysis of B-Iactamase activities showed the presence of
a penicillinase activity in 41 % of resistant strains (Table 2). A TEM-l type
enzyme (pI 5.4) was identified in 34% of the strains, while 7% of the strains
produced a 13-Iactamase of pI 6.3, characterised subsequently as a CARB-
type enzyme, which inactivated ampicillin and carbenicillin, but not
89
CD
o
Table 2. B-lactamase distribution in clinical strains of A.calcoaceticus
Inhibited by
Enzyme clavulanic acid cloxacillin sulbactam i. strains
None detected 18
Penicillinase 41
TEM type (pI 5.4) + + (34)
CARE type (pI 6.3) + + (7)
Cephalosporinase + + 9
(pI >8)
Penicillinase +
cephalosporinase 32
TEM type + Ca + + + (30)
CARE type + Ca + + + (2)
ahigh pI (>8.0) enzyme assumed to be a chromosomal cephalosporinase

data taken from Joly-Guillou et al. (1988)
methicillin or cloxacillin, and had a relatively low activity against
cephalosporins. Subsequent analysis of the characteristics of this enzyme
was carried out with either cell-free extracts of a selected strain that was
highly resistant to ticarcillin or with purified enzyme obtained from a sonic
lysate of cells from a 4 I culture of the selected strain (Paul et aI., 1989). The
substrate profile of the purified enzyme was characterised by preferential
hydrolysis of penicillins (ampicillin, carbenicillin, azlocillin), while
cloxacillin, cephalothin and cephaloridine were very poorly hydrolysed. The
enzyme was inhibited by p-chloromercuribenzoate (PCMB), which inhibits
many enzymes containing cysteine residues (Bauernfeind, 1986), and also by
cloxacillin and the B-lactamase inhibitors clavulanic acid and sulbactam, but
not by chloride ions (100 mM). Immunological studies were performed with
several antisera prepared against purified enzymes (CARB-3, TEM-1, OXA-
2, OXA-4 and OXA-6), but only antiserum to CARB-3 neutralised the
enzyme activity and no cross-neutralisation was obtained with other antisera.
Thus this new enzyme appeared to be a CARB-type enzyme, but since it did
not co-focus with CARB-3 (pI = S.7), it was designated CARB-S. Its
molecular mass was estimated from gel filtration (on Ultrogel AcAS4) to be
28,000, similar to other CARB enzymes described previously (Philippon et
aI., 1980; Medeiros, 1984). This was the first demonstration of a CARB-
type enzyme in strains of A. calcoaceticus var. anitratus. Substrate and
inhibition profiles of this novel B-Iactamase from A. calcoaceticus are
given in Table 3. This novel enzyme belongs to the carbenicillinase group of
transferable B-Iactamases. This is a homogenous group of enzymes which is
also characterised by a high level of enzyme production in the host strains.
The penicillinases described above (TEM-1, TEM-2 and CARB-S) belong to
the group of constitutive plasmid-determined B-lactamases described by
Bauernfeind (1986) and Medeiros (1984).
Cephalosporinases in Acinetobacter spp
The older cephalosporins, such as cephalothin, cephaloridine or

cefazoline, can also be inactivated by plasmid-mediated B-Iactamases. The
third generation cephalosporins (e.g. cefotaxime, ceftazidime, ceftriaxone,
or moxalactam), characterised by two substituents on the C-7 side cha,in
(methoxyimino and aminothiazole groups), exhibit a particularly high B-
lactamase stability (Knothe and Dette, 1983), but most of the
chromosomally-determined B-Iactamases preferentially hydrolyse
cephalosporins, including many of the newer B-Iactams which resist
hydrolysis by the plasmid-determined B-lactamases (Medeiros, 1984). As
described earlier (Table 1), less than SO% of Acinetobacter strains are
inhibited at acceptable concentrations of cefotaxime, moxalactam, or even
ceftazidime, which was initially the most active drug (Phillips et aI., 1981;
Bergogne-Bcf(:!zin and Joly-Guillou, 1986). Almost all Gram-negative
bacilli produce a chromosomally-determined B-lactamase that is species-
specific (Medeiros, 1984). The rapid increase of resistance to third
generation cephalosporins (Joly-Guillou et aI., 1987) has led several authors
to investigate the possible presence of an inducible cephalosporinase in
Acinetobacter spp. Before the third generation of cephalosporins became
91
Table 3. Substrate and inhibition profiles of the novel fl-lactamase
from A. calcoaceticus strains A85-l35 (Paul et al., 1989)
Cell free Purified

extract fraction
Substrate profile a
Ampicillin 86 80
Carbenicillin 76 61
Azlocillin 57 61
Methicillin 1 16
Oxacillin 1 3
Cloxacillin 1 2
Cephalothin 10 4
Cephaloridine 19 8
Cefotaxime <1 <0.5
Cefsulodin <1 <0.5
Imipenem <1 <0.5
Inhibition profi1e b
Cl - (100 roM) o o
CLAV (1 ttM) 58 94.5
SUL (1 ttM) 9l. 3
Cloxacillin (1 roM) >90
pCMB (0.5 roM) >90
Antiserum:
anti-CARB-3 >80
anti-TEM-1 50
anti-OXA-2 <30
anti-OXA-4 <30
anti-OXA-6 <30
aV max values, expressed relative to the Vmax for benzyl-

penicillin (100%).
bpercentage of neutralised activity at the indicated inhibitor
concentration in the presence of benzylpenicillin. The
inhibitory effects of clavulanic acid (CLAV) and sulbactam (SUL)
were determined after preincubation with the enzyme for 10 min
at 37°C.
92
available in 1977, an initial study (Morohoshi and Saito, 1977) had
demonstrated production of an inducible class I type 13-Iactamase
(cephalosporinase) in Acinetobacter. The enzymedosely resembled the
enzymes produced by Proteus, Citrobacter and Pseudomonas spp. in
terms of molecular mass (30,000), sensitivity to inhibitors and inducibility.
Further studies of resistance mechanisms in Acinetob acter spp. (Devaud et
al., 1982; Goldstein et al., 1983) cited the presence of a cephalosporinase,
but this was not clearly demonstrated since the reported experiments focused
on plasmid-mediated enzymes. In describing the two major groups of
cephalosporinases outlined by Sawai et al., (1982), Medeiros (1984)
described the enzyme produced by Acinetobacter strains as belonging to
the "typical cephalosporinase group", which includes inducible enzymes such
as those produced by Enterobacter aerogenes, Ent. cloacae, Serratia
marcescens and Ps. aeruginosa. One of the main characteristics of
"typical cephalosporinases" is their narrow spectrum of substrate specificity;
they have little or no activity against benzylpenicillin, ampicillin and
carbenicillin, but render bacteria resistant to various cephalosporins. The
contribution of such cephalosporinases to 13-Iactam resistance can be based
on at least two major mechanisms (Piddock and Wise, 1985): (i) hydrolysis of
older cephalosporins, which requires high enzyme affinity (KJ for the
antibiotic (Medeiros, 1984) - this is the case with the Acinetobacter
cephalosporinase; indeed the initial resistance of Acinetobacter to most
first and second generation cephalosporins, as described above, was
associated with enzymic hydrolysis of antibiotics (Joly-Guillou and
Bergogne-Ben!zin, 1985b); (ii) although third generation cephalosporins
are normally resistant to hydrolysis by 13-Iactamases, the amount of 13-
lactamase (cephalosporinase) prod uced in the periplasmic space of the cell
increases with exposure to .B-Iactams such as cefoxitin or ceftazidime
(induction), and it has been postulated (Sanders, 1983) that the 13-Iactamase
protects the bacterial cell by binding the antibiotic molecules rather than
splitting them - thus the enzyme acts as a peri plasmic "sponge" ("sponge
effect" or "trapping") which keeps the antibiotic from reaching its target site
on the cytoplasmic membrane. Such a mechanism may occur in
Enterobacter spp., S. marcescens and Ps. aeruginosa, and has also been
suggested for A. anitratus (Sanders, 1983; Medeiros, 1984). However,
since the report in 1977 of an inducible cephalosporinase in Acinetobacter
spp., no clear demonstration of such an enzyme can be found in the
literature; the hypotheses of cephalosporinase activities mentioned above
were based mostly upon clinical experience and indirect evidence of
cephalosporinase production rather than in-vitro experiments (Sanders,
1983). J oly-Guillou et al. (1988) also looked for cephalosporinase
production in 100 Acinetobacter clinical strains by using the inhibitory
activity of cloxacillin and isoelectric focusing of crude enzyme extracts
(Matthew and Harris, 1976; Joly-Guillou et al., 1988). The presence of
cephalosporinase activity correlated with variable bands of 13-Iactamase
activity located at pI > 8. The results of this study showed that 32% of the
strains possessed a cephalosporinase plus a penicillinase, while only 9%
exhibited cephalosporinase activity alone. Among the strains with evidence
of cephalosporinase activity, a group of nine strains showed low-level
93
resistance to third generation cephalosporins (geometric mean MICs were
14.7 mg cefotaxime /1 and 4.3 mg ceftazidime /1); a second group of nine
strains exhibited high-level resistance to all cephalosporins, with geometric
mean MICs of 54.g mg cefotaxime /1 and 17.2 mg ceftazidime /1. Analysis of
the results suggested that ceftazidime might be less susceptible to the
cephalosporinase of Acinetobacter than was cefotaxime. Otherwise, B-
lactamase inducibility was not clearly demonstrated in these strains of
Acinetobacter, although there was evidence that the second group of
strains, showing a high level of resistance to third generation cephalosporins,
were producing an increased amount of cephalosporinase constitutively
(Bauernfeind, 1986), resulting in a certain amount of cefotaxime hydrolysis
and a lesser, but detectable, degree of ceftazidime hydrolysis. As described
elsewhere (Piddock and Wise, 1995), these compounds are hydrolysed, but at
a slower rate than in conventional enzymic hydrolysis of first generation
cephalosporins. More recently, a further investigation (Hood and Amyes,
1999) of chromosomally mediated B-lactamases from the genus
Acinetobacter used a novel isoelectric focusing method to identify, in eight
strains, four different B-Iactamases with pIs ranging from 7.3 to g.g.
Additional minor bands at pI 9.8 and 10.1 were also revealed for one enzyme
by this modification of conventional isoelectric focusing techniques.
Further details of these enzymes are given elsewhere (Hood and Amyes, this
volume).
New Enzymes in Acinetobacter spp
The study of Hood and Amyes (1989) provided information on

enzymes previously undescribed in Acinetobacter spp. and illustrated the
apparently limitless power of bacteria to adapt themselves to new antibiotics
and to develop new mechanisms of resistance (Wiedeman et a!., 1989). This
has h2.ppened recently with Klebsiella pneumoniae (Sirot et a!., 1988) and
other Enterobacteriaceae, particularly E. coli. In epidemic strains of
K.pneumoniae, plasmid-mediated broad-spectrum B-Iactamases have been
characterised. Clinical isolates encoding these novel B-Iactamases
appeared in France in 1984-g5 and showed decreased susceptibilities to
aminopenicillins, carboxypenicillins, ureidopenicillins, most cephalosporins,
including third generation compounds (cefotaxime and ceftazidime), and a
monobactam (aztreonam). These strains were still, however, susceptible to
imipenem, cephamycins (cefoxitin, cefotetan) and moxalactam. In late 1989
a strain of A. baumannii isolated in our hospital from a urine sample
exhibited an unusual susceptibility profile (Joly-Guillou and Bergogne-
Ben~zin, 1990), with resistance to ticarcillin and piperacillin (MICs > 256
mg/I), cefoxitin (64 mg/l), cefotaxime (32 mg/l) and ceftazidime (16 mg/I).
In addition, synergy was observed between the inhibition zones obtained in
disc sensitivity tests with amoxicillin/clavulanic acid and cefotaxime or
ceftazidime. This was an unusual pattern of response to cephalosporins -
most strains resistant to cefotaxime and ceftazidime that produce
cephalosporinase (Joly-Guillou et a!., 1998) exhibit antagonism between the
above mentioned zones, due to the induction of cephalosporinase by the
potent cephalosporinase inducer c1avulanic acid (Sanders, 1993). Analysis
94
of crude extracts by the gel iodometric method in the presence of substrate
inhibitors (Labia and Barthelemy, 1979; Joly-Guillou and Bergogne-
Ben~zin, 1988), showed that: (i) the strain was a 13-lactamase producer; (ii)
13-lactamase activity was inhibited by clavulanic acid, which specifically
inhibits penicillinase activity; (iii) 13-lactamase activity was not inhibited by
cloxacillin, which specifically inhibits cephalosporinase activity. Analytical
isoelectric focusing failed to show the presence of either a TEM-type or a
CARB-5 enzyme, and no cephalosporinase activity was detected,although
some enzymic activity was observed with a pI of 7.7. It can be postulated that
this particular Acinetobacter isolate may produce one of the novel broad
spectrum 13-lactamases described recently by Sirot et al. (1988) and
designated "cefotaximases" (e.g. CTX-l of pI 6.3) or "ceftazidimases" (CAZ-1
to CAZ-5), the latter including enzymes of pI 7.7 which otherwise belong to
the SHV 13-lactamase family. A possible role for A. calcoaceticus as a
reservoir or vehicle of transmissible antibiotic resistance in the hospital
environment has been suggested by Chopade et al. (1985); indeed, many
clinical isolates contain a variety of different plasmids (as described in a
subsequent section). Trimethoprim R plasmids have been shown to be
transmissible from E. coli to A. calcoaceticus (Chopade et aI., 1985),
which suggests the potential for interchange of resistance markers with E.
coli and other Enterobacteriaceae in close contact with Acinetobacter
spp. in the clinical environment. If the presence of plasmid-mediated broad-
spectrum B-lactamases is confirmed in Acinetobacter strains, it will be the
first description of this new class of enzyme in aerobic Gram-negative bacilli
and will further confirm the epidemiological role of Acinetobacter strains
in the spread of resistance in hospitals.
B-Lactamase Inhibitors
In order to overcome B-lactamase production and achieve effective

antibiotic therapy of infections due to B-lactamase producers, several B-
lactamase inhibitors have been designed (Wise, 1982) and their in-vitro
activities assayed in combination with various penicillins and
cephalosporins, especially against B-Iactamase-producing aerobic Gram-
negative bacilli (Philippon et aI., 1980; Jacobs et al., 1986). Among the
well-known B-Iactamase inhibitors, clavulanic acid and sulbactam have been
introduced into medical practice; more recently, the newly developed
compound tazobactam (YTR 830) has also been investigated. Although 13-
lactamase inhibitors have poor intrinsic antibacterial activity (Wiedeman et
aI., 1989), sulbactam is surprisingly effective against Acinetobacter spp.
(Kitzis et aI., 1983). Similarly, in our own study (E. Vallee, M.L. Joly-
GuilIou and E. Bergogne-BeH!zin, unpublished data), sulbactam was less
active (Table 4) than tazobactam against ticarcillin-resistant strains (range
MICs: 1-32 mg/l), but both were active by themselves against ticarcillin-
susceptible strains, with geometric mean MICs of sulbactam and tazobactam
of 2.1 and 6.5 mg/l respectively. Whether or not these figures are clinically
relevant is questionable. It was of more interest to examine the inhibitory
activities of these 13-lactamase inhibitors against the 13-lactamases produced
by Acinetobacter strains. MICs of ticarcillin, cefotaxime and ceftazidime
95
Table 4. Susceptibility of 46 clinical strains a of A.calcoaceticus to
ticarcillin in the absence and presence of S-lactamase inhibitors
Drugs MICs (mg/l)

Range 50% 90% Geometric
of strains of strains mean
Ticarcillin 2 - 32 10 25.2 11.5

Tic R 2048 - >2048 >2048 >2048 >2048
Sulbactam - 8 1.5 3.5 2.1

16 - 32 16 22.7 18
Tazobactam 1 - 16 6 13.3 6.5

- 32 5.6 14.8 7.1
Ticarcillin 0.5 - 32 9.7 22.9 10.6

+ clavulanic 32 - >2048 112 >2048
acid (2 mg/l)
Ticarcillin 0.5 - 4 1.12 2.9 1.55

+ sulbactam 8.0 - 64 16.0 30.08 15.50
Ticarcillin 1 - 16 3.75 11.6 5.48

+ tazobactam - 32 23.7 99.55 33.02
a17 strains were ticarcillin-susceptible or moderately resistant;

29 strains were ticarcillin-resistant (MIC >64 mg/l)
96
were therefore determined for these drugs alone and in combination with £-
lactamase inhibitors at a fixed concentration ratio 1: 1. Synergy was defined
as a significant decrease in the geometric mean MICs of the
antibiotic/inhibitor compared with those for antibiotic alone.
Simultaneously, in order to determine the inhibition profiles of £-
lactamases, clavulanic acid, cloxacillin, sulbactam and tazobactam were
added in gel, made from the classical iodine-iodide starch system, at a fixed
concentration of 0.5 mg/I, to crude extracts of bacterial cells in order to
specifically inhibit penicillinase and/or cephalosporinase activities.
Clavulanic acid caused a significant decrease in the MICs for the ticarcillin-
resistant strains, (MIC so =112 mg/l), but not a true reversion to
susceptibility, (Table 4). The MICs of ticarcillin, either singly or combined
with tazobactam, were identical for ticarcillin-susceptible strains, but were
notably diminished for ticarcillin-resistant strains, which returned to
susceptible levels (geometric mean MIC of 33.02 mg/l). Similarly,
sulbactam in combination with ticarcillin significantly reduced the MICs of
ticarcillin (geometric mean MIC of 15.50 mg/l). The geometric mean MICs
of cefotaxime and ceftazidime are shown in Table 5 in relation to the
enzymic profiles of the tested strains. Combinations of cefotaxime and
inhibitors produced a significant reduction in the geometric mean MICs,
when compared with those of cefotaxime alone, for strains that only
produced cephalosporinase (geometric mean MIC of cefotaxime was 41.8
mg/l, compared with 2.2 and 7.2 mg/I when combined with sulbactam or
tazobactam, respectively). This observation was not made for ceftazidime,
with which the geometric mean MICs were identical or only slightly
diminished when tested in combination with inhibitors. Using crude
bacterial extracts, it was found that tazobactam and sulbactam inhibited both
cephalosporinase and penicill inase activities. These data require further
investigation before any clinical application can be proposed.
Imipenem
Imipenem is a new carbapenem antibiotic with the broadest

antibacterial spectrum of any £-lactam antibiotic currently available. It is
active against many multiresistant pathogens, including strains of
Enterobacter, Serratia and Ps. aeruginosa. No cross-resistance has
been observed between imipenem and other £-lactam antibiotics; indeed,
imipenem is theoretically not inactivated by £-lactamases, and "trapping" of
the antibiotic by high concentrations of B-Iactamase in the periplasmic space
(as described above for third generation cephalosporins) does not seem to be
the mechanism by which imipenem is inactivated. With the earliest
preparation of this drug (N-formimidoyl-thienamycin), and before the
antibiotic was clinically available, we started a survey of the imipenem
susceptibility of Acinetobacter strains isolated since 1981 (Bergogne-
BeH!zin and Joly-Guillou, 1986). All strains were initially very susceptible
to imipenem and no changes in susceptibility occurred for six years (Fig. 1).
The MICs of imipenem were in the range 0.016-4 mg/I for A. anitratus and
0.016-1 mg/l for A. lwoffii, with geometric mean MICs of 0.33 and 0.20
mg/l respectively. This survey, which is still in progress, has now shown the
97
to
00
Table 5. In-vitro activity of sulbactam and tazobactam in combination with cefotaxime and ceftazidime
against A.calcoaceticus
Geometric mean HIes (mg!l)

Enzymes produced CTZ CTZ SUL CTX CTX CTX TZB CTZ CTZ
+ SUL + SUL + TZB + TZB
None detected 2.8 1.4 5.6 1.7 2.3 1.4 2.8
Penicillinase 3.2 2.5 16 6.3 6.3 6.3 8.0 4.0 3.2
Cephalosporinase 8.0 1.7 2.3 2.2 41.8 7.2 8.9 S.2 8.0
Penicillinase 6.5 6.8 18.3 11.9 20.9 5.S 6.8 3.6 6.S
+ cephalosporinase
CTZ: ceftazidime; SUL: sulbactam; CTX: cefotaxime; TZB: tazobactam

Geometric mean
mg/l MICs
32
16
8
4
2
1
0.5
0.25
0.125
--,
0.064
0.032
0.016
L---~-----r----'-----~--~~--~-----r----~ Years
82 83 84 85 86 87 88 89
Fig. 1. Imipenem susceptibility of Acinetobacter spp.
emergence of a few resistant strains: two resistant strains of A. baumannii

belonging to two different phage types were isolated in 1987 from a single
patient who had been treated with imipenem (Bergogne-Ben!zin et aI.,
1987), and, within the last two years, 20 strains resistant to imipenem have
been isolated in our hospital. The MICs of imipenem for these isolates
ranged from 4 to 8 mg/l, which actually indicates a diminished susceptibility
rather than high-level resistance. However, the minimal bactericidal
concentrations (MBCs) of imipenem for these strains ranged from 32 to 128
mg/I. Study of their resistance phenotypes has shown that these strains were
resistant to ticarcillin (with no significant effect of clavulanic acid),
piperacillin, moxalactam and cefoxitin; they were variably resistant to
cefotaxime and ceftazidime. Analysis of their enzyme content has shown the
variable presence of a TEM-1 (pI 5.4) or TEM-2 (pI 5.6) enzyme, a
cephalosporinase, and a new enzyme (pI 7.6) found in two strains which were
resistant to all B-Iactams. A few strains, resistant to imipenem only, did not
exhibit any B-Iactamase activity. Although B-Iactamases capable of
inactivating imipenem, produced by Ps. maitophilia, Flavobacterium
odoratum, Bacteroides fragilis and Aeromonas hydrophila, have been
reported occasionally in the literature (Medeiros, 1984; Shannon et aI., 1986;
Quinn et aI., 1988; Wiedeman et aI., 1989), most studies of imipenem
resistance refer to alterations in bacterial outer membrane permeability.
Studies on the penetration of B-Iactams through porin channels have shown
that the rate of penetration appears to be determined by physico-chemical
properties of the drug, such as molecular mass, hydrophobicity and electrical
charge (Hancock, 1984; Piddock and Wise, 1985; Nayler,1987; Quinn et aI.,
1988). Imipenem crosses the outer membrane of E. coli K12 extremely
rapidly, probably as a result of its compact structure, low molecular mass and
its zwitterionic charge (Hancock, 1984; Quinn et ai., 1988). The mechanism
of resistance to imipenem in most resistant bacteria appears to be related to
a selective decrease in permeability across the bacterial outer membrane,
usually associated with discrete alterations in the electrophoretic profiles of
outer membrane proteins, resulting in a decreased rate of antibiotic uptake.
99
This has been analysed primarily in Ps. aeruginosa (Quinn et al., 1988), in
which impaired penetration of imipenem through the outer membrane is
related to the lack of an outer membrane protein that is present in
susceptible parent strains (Quinn et al., 1988). It has also been established
that alteration of the lipopolysaccharide in the outer membrane may
contribute substantially to diminished permeability and thus to B-Iactam
resistance (Piddock and Wise, 1985; Nayler, 1987). So far as
Acinetobacter spp. are concerned, it has been shown that a selective
alteration in outer membrane permeability affects the susceptibility of the
organism to the oxygen-independent antimicrobial activity of rat
polymorphonuclear leucocyte granule contents (Leoffelholz and
Modrzakowski, 1987). This alteration in outer membrane permeability,
related to lipopolysaccharide structure, was shown to be associated with the
presence of plasmid RPI (Loeffelholz and Modrzakowski, 1986). RP1,
which confers resistance to ampicillin, carbenicillin, kanamycin, neomycin
and tetracycline, physically alters the outer membrane structure of the
bacteria. Whether these RP1-mediated outer membrane alterations could
playa role in resistance to imipenem and other antibiotics in A.
calcoaceticus remains to be elucidated. Similarly, the question as to
whether the impaired penetration of imipenem through the outer membrane
observed in Ps. aeruginosa can be extrapolated to Acinetobacter
remains to be determined. Further investigations of possible enzymic
inactivation of imipenem in resistant Acinetobacter strains also remain to
be carried out.
MECHANISMS OF RESISTANCE TO AMINOGLYCOSIDES IN

ACINETOBACTER SPP
Aminoglycosides have been used since the mid-1940s. The emergence

of resistance as a result of the interplay between use of aminoglycosides aJJd
selective pressure (McGowan, 1983) was observed during the early phase of
aminoglycoside therapy when streptomycin and kanamycin were the major
available antibiotics. There are three major mechanisms of resistance: (i)
alteration of the target site on the ribosomes; (ii) reduced aminoglycoside
permeability; (iii) inactivation of the aminoglycoside by enzymic
modification. The latter mechanism is the, most common type of
aminoglycoside resistance found among clinical isolates of staphylococci and
Gram-negative bacilli. Aminoglycosides are modified either by acetylation of
an amino group (acetyltransferases), by phosphorylation
(phosphotransferases). or by adenylation of hydroxyl groups
(adenyl transferases). The distribution of aminoglycoside-modifying
enzymes in various bacterial species is listed in Table 6 (Bauernfeind et aI.,
1986). A number of studies since 1975, have shown the presence of
amino glycoside modifying enzymes in clinical isolates of A c i net 0 b a c ter
(Uwaydah and Taql-Eddin, 1976; Dowding, 1979; Bergogne-Berezin et aI.,
1980; Gomez-Lus et aI., 1980; Murray and Moellering, 1980; Devaud et aI.,
1982; Goldstein et aI., 1983). Resistance of Acinetobacter to tobramycin,
100
Table 6. Substrate profiles of amino glycoside modifying enzymes (based on
Bauernfeind et al., 1986)
Enzyme Modification of Organisms

Genta- Netil- Tobra- Amika- Kana- Strepto
micin micin mycin cin mycin mycin
C A
Phosphorylating enzymes
APH(3')I -/+ + Enterobacter-

APH(3')II iaceae,
Pseudomonas,
Acinetobacter,
Staphylococcus,
S. faecal is
APH(2") + + + Staphylococcus,
S.faecalis
APH(3") + Enterobacter-
iaceae,
Pseudomonas,
ACinetobacter,
Gram-positive
bacteria
Adeny1ating enzymes
AAD(2") + + + + Enterobacter-
iaceae,
Pseudomonas,
Acinetobacter
AAD(4') + + + Staphylococcus
AAD(3") + Enterobacter-
iaceae,
Pseudomonas,
Acinetobacter
AAD(6) + Staphylococcus
continued
101
Table 6. (Continued)
Acet:x:lating enz:x:mes
AAC(6') _/+a + + + + Enterobacter-
iaceae,
Pseudomonas,
Acinetobacter,
Staphylococcus,
S.faecalis
AAC(3)I + ± Enterobacter-
iaceae,
Pseudomonas,
Acinetobacter
AAC(2') + + + Prov.stuartii
P.rettgeri,
Acinetobacter
AAC( 3) III + + + + Pseudomonas
AAC(3) II + + + ± Enterobacter-
iaceae
+, modified; not significantly modified

agentamicin Cl : -ve; CIA and C2 : +ve
dibekacin and sisomycin, plus moderate resistance to kanamycin and

amikacin, were related to the presence of an aminoglycoside 6' -N-
acetyltransferase (AAC(6 » (Bergogne-Berezin et aI., 1980; Gomez-Lus et
aI., 1980; Murray and Moellering, 1980). Other studies have characterised
acetylating enzymes of the AAC(2') (Dowding, 1979) and AAC(3)I types
(Gomez-Lus et aI., 1980), which modify gentamicin and tobramycin,
phosphotransferases of the APH(3')I (Gomez-Lus et aI., 1980) and
APH(3')Il types (Murray and Moellering, 1980), which confer resistance to
kanamycin, and adenyltransferases of the AAD(2") (Murray and Moellering,
1980) and AAD(3") types (Devaud et aI., 1982; Goldstein et aI., 1983), which
modify streptomycin and spectinomycin. Thus the presence of modifying
enzymes belonging to the three main classes is responsible for the resistance
of Acinetobacter strains to a great number of aminoglycosides. In an
extensive study (Bergogne-Berezin et aI., 1980), we investigated the in-vitro
activity of seven aminoglycosides against 215 Acinetobacter strains isolated
from patients between 1971 and 1980. The strains isolated between 1971
and 1975 were initially susceptible to the aminoglycosides gentamicin,
sisomycin and tobramycin, with only 10 to 20% resistant strains isolated in
1971. A similar investigation (Uwaydah and Taql-Eddin, 1976) of the
102
susceptibility of 15 Acinetobacter strains to gentamicin and tobramycin
found MICs ranging from 0.12 to 0.9 mg/I. In our study, a sudden rise in
resistance to gentamicin, kanamycin, sisomycin and tobramycin occurred in
1975, correlating with the appearance of aminoglycoside-modifying enzymes.
An acetylating enzyme of the AAC(3)I type and a phosphorylating enzyme of
the APH(3 / )I type were found to be present in cell-free extracts (Bergogne-
BeH!zin et aI., 1980). More recently, a new phenotype has appeared in
France, with resistance to gentamicin (GEN'), sisomycin (SIS') and amikacin
(AMK'), but not to tobramycin (TOB'). Twelve Acinetobacter strains with
this phenotype (GEN' SIS' AMK' TOB') were analysed in late 1986; the
results indicated the presence of AAC(3)I and APH(3')I enzymes that
inactivate lividomycin, but not amikacin. A possible new
phosphotransferase was also found in four of the 12 Acinetobacter strains.
Enzymic phosphorylation of amikacin, which has been reported previously in
E. coli and Ps. aeruginosa (Joly-Guillou et aI., 1987), was found in A.
calcoaceticus var. anitratus. This new enzyme also inactivated
kanamycin A,B and C, neomycin, neamine, paromomycin, seldomycin,
butirosin, lividomycin, sisomycin and tobramycin. Thus the presence of an
AAC(3)I and the new phosphorylating enzyme, provisionally designated
APH(3 / )III, modifying butirosin, lividomycin and amikacin, may account for
the new Acinetobacter phenotype described above and provides a new
mechanism of amikacin resistance. These observations have been supported
by a further study (Lambert et aI., 1988) which demonstrated that resistance
to amikacin was related to the synthesis of a new type of transferable
APH(3 / ) in a clinical strain of Acinetobacter. The authors of this study
showed that the gene conferring resistance to kanamycin/amikacin was
carried by a 63 kb plasmid, pIP1841, self-transferable to A. baumannii, A.
haemolyticus and A. lwoffii, but not to E. coli. Of the APH(3') types
detected in clinical isolates, only this new type of enzyme, subsequently
designated APH(3' )VI (Lambert et aI., 1990), modifies amikacin in the
manner described above (Joly-Guillou et aI., 1987; Lambert et aI., 1988).
Alternative mechanisms of resistance to aminoglycosides, such as diminished
permeability or alteration of the ribosomal target sites, have not yet been
investigated in Acinetobacter.
RESISTANCE OF ACINETOBACTER TO FLUOROQUINOLONES
During the last few years, the availability of an increasing number of

new fluorinated quinolones has given rise to numerous studies on their
comparative in-vitro activities, mostly against various multiply-resistant
Gram-negative bacilli (King and Philips, 1986; Rolston et aI., 1988).
Additional clinical evaluations of the efficacy of the new quinolones have
included their use in treating severe nosocomial infections involving Gram-
negative bacilli. A. baumannii was not tested extensively in the series of
strains included in the original in-vitro studies (Bergogne-Berezin and Joly-
Guillou, 1985; Godineau-Gauthey et aI., 1988; Rolston et aI., 1988) and thus
only limited data are available on the initial susceptibility of Acinetobacter.
clinical strains to new quinolones. Even so, the extremely interesting data
103
generated by these early in-vitro studies (Bergogne-Berezin and 101y-
Guillou, 1985; 101y-Guillou and Bergogne-Berezin, 1985a; King and
Phillips, 1986) showed enhanced activity against A. baumannii of new
quinolones such as pefloxacin, ciprofloxacin, norfloxacin and ofloxacin,
compared with the parent compound, nalidixic acid. In 1985-86, a study with
a limited number of strains (King and Phillips, 1986) and our own studies of
130 clinical Acinetobacter strains (Bergogne-Berezin and 10Iy-Guillou,
1985; 10ly-Guillou and Bergogne-Berezin, 1985a), compared susceptibility
to the new fluorinated quinolones with susceptibilities to other antibiotics.
Very promising results were obtained, indicating that ciprofloxacin,
ofloxacin, and pefloxacin were substantially more active in vitro than either
nalidixic acid, third generation cephalosporins or aminoglycosides. The
greatest activity was exhibited by ofloxacin, with an MIC so of 0.6 mg/1. The
MICsoof ciprofloxacin was 0.7 mg/I, with mean MICs of 0.98, 1.03 and 1.64
mg/I for ciprofloxacin, ofloxacin and pefloxacin respectively. Similarly,
King and Phillips (1986) reported that the MICs of ciprofloxacin for 35
strains of Acinetobacter ranged from 0.004 to 1 mg/I, with geometric mean
MICs of 0.053, 0.067 and 0.158 mg/I for ciprofloxacin, ofloxacin and
pefloxacin respectively. Since 1985, the new quinolones, especially those
compounds which can be administered parenterally, have been used
increasingly in the treatment of severe nosocomial infections caused by
multiply-resistant bacteria.
It has been suggested that the isolation frequency of quinolone-

resistant variants is higher in non-fermenting organisms than in
enterobacteria; thus it might be expected that emergence of resistant
mutants to the newer quinolones, which occurred rapidly in Ps. aeruginosa
(Neu, 1988), might also occur in Acinetobacter. A few years after addition
of the new quinolones to the antibacterial armamentarium, we carried out a
new in-vitro investigation using Acinetobacter strains collected in 1988-89.
We had noted in our routine work that Acinetobacter strains were
becoming increasingly resistant to pefloxacin. Table 7 compares our latest
data (June 1989) on the susceptibility of 100 recent isolates to new
quinolones with our previous results (Joly-Guillou and Bergogne-Rerezin,
1985a). The results showed that the 1989 MIC values (MIC so ' MIC 90 and
geometric mean MICs) had increased five to ten-fold for pefloxacin and
ciprofloxacin, but not for ofloxacin. In addition, the MIC ranges were
extended, e.g. from 0.016-32 mg/I to 0.03-128 mg/I for ciprofloxacin and
from 0.064-16 mg/l to 0.03-32 mg/I for ofloxacin. The data obtained with
newer compounds (Iomefloxacin, temafloxacin, difloxacin) did not indicate
any significant improvement in activity compared to previous compounds.
As noted in our earlier studies, the most active fluorinated quinolone against
Acinetobacter clinical strains was ofloxacin, while difloxacin exhibited a
very similar activity to that of ciprofloxacin. When the distribution of the
strains as a function of MIC values was examined (Fig. 2), a bimodal
distribution corresponding to two populations (susceptible and resistant)
was apparent. Using the breakpoints recommended in France of 1 and 4
mg/I, the proportions of fully susceptible strains were only 27, 31 and 35%
for pefloxacin, ciprofloxacin and ofloxacin respectively. Such a threatening
104
Table 7. MICs of fluoroquinolones (mg/l) for strains of A. baumannii
isolated during 1985 (130 strains) and 1989 (100 strains)
Compounds Range MIC 50 MIC 90 Geometric

Mean
Pefloxacin
1985 0.25-16 1. 05 6.71 1. 6/~
1989 0.125-128 14.15 74.66 10.26
Ciprofloxacin
1985 0.016-32 0.72 4.64 0.98
1989 <0.03-128 7.3 60.44 5.76
Ofloxacin
1985 0.064-16 0.69 3.73 1. 03
1989 <0.03-32 3.37 14.6 2.03
Lomef1oxacin
1989 0.06-256 10.18 85.33 7.29
Temafloxacin(A.52254)
1989 <0.03-64 2.8 41.14 2.83
Difloxacin(A.56619)
1989 <0.03-128 7.73 88 5.32
situation might have resulted from the early and large usage of pefloxacin in
France; indeed these data can be compared with other studies (e.g. Chow et
aI., 1988) showing a more limited MIC range for ciprofloxacin of 0.5-4 mg/l
and MICsoand MIC 90 values of 2 and 4 mg/l respectively for a low inoculum of
104 cfu. It should, however, be noted that the latter values were notably
increased by a larger inoculum (10 6 cfu), reaching 8 and 16 mg/I, which
suggests, as in our study, a large proportion of resistant strains (MICs > 4
mg/I)
A further in-vitro study with our susceptible strains was carried out in
a micro dilution system which permitted us to evaluate the minimal
bactericidal concentrations (MBCs) for each strain (Table 8). The results of
this study indicated a very close relationship between the MICs and MBCs
(inoculum: 10s cfu) for all new quinolones tested. The ratios of geometric
mean MBCs/MICs ranged from 1.11 to 1.21, which confirmed the excellent
bactericidal activity of this class of antibiotics against susceptible
Acinetobacter strains. Very few studies of antibiotic combinations have
been carried out, although the available in-vitro data (Chow et aI., 1988)
105
30 PEFLOXACIN
25
20
15
10
0
0,03 0,06 0,125 0,25 D,S 2 ~ 8 16 32 64 128
20
CI PROFLOXAC I N
15
10
0+-'---.....
0,03 0,06 0,125 0,25 D,S 2 I 16 32 6~ 121
2S OFLOXACIN
20
15
10
MICs
o
0,03 0.06 0.125 0,25 D,S 4 16 32 128 256
Fig. 2. HIes of pefloxacin, ciprofloxacin and ofloxacin for 100 A.

baumann.U strains
suggest that synergistic interactions with ciprofloxacin rarely occur; most

combinations exhibit at least additive effects at clinically achievahle serum
concentrations. Multiply-resistant Acinetohacter strains involved in
nosocomial infections should respond to comhinations of ciprofloxacin with
amikacin, tohramycin, or ceftazidime.
Even though the frequency of mutational resistance was as low as 10. 9

in early studies, the emergence of resistant variants in clinical
Acinetobacter strains is a serious threat which may result in the failure of
106
Table 8. HICs and MBCs (mg/l) of fluoroquinolones for susceptible strains of A. baumanni j
(number of strains examined in parentheses)

Geometric Mean
MIC 50 MBC 50 HIC 90 MBC 90 MIC MBC MBC/MIC
Pefloxacin (34) 0.31 0.32 1.16 1.8 0.45 0.52 1.15
Ciprofloxacin (29) 0.2 0.24 0.47 0.85 0.31 0.37 1.19
Ofloxacin (35) 0.24 0.29 1. 47 1.8 0.38 0.46 1. 21
Lomefloxacin ( 34) 0.33 0.38 0.9 0.92 0.45 0.51 1.13
Temafloxacin (34) <0.12 <0.12 0.48 0.485 0.18 0.2 1.11
Difloxacin (35) <0.12 <0.12 0.9 0.95 0.197 0.24 1. 21
o
-..J
antibiotic therapy. Cross-resistance may develop by mutations that result in
an altered DNA gyrase, reduced permeability, or both. Precise mechanisms
of resistance in Acinetobacter spp. remain to be analysed, although DNA
gyrase has been shown to exist in all bacterial species examined to date
(Piddock and Wise, 1989). Mutations occur in Ps. aeruginosa, H.
injluenzae and C. jreundii which give rise to a DNA gyrase A sub-unit
that is less susceptible to inhibition by quinolones. The gyr A mutations
result in decreased susceptibility to all quinolones, but not cross-resistance
to chemically unrelated antibiotics. There have also been several reports of
decreased uptake of quinolones causing reduced susceptibility, and
permeability mutants have been described that show cross-resistance to 13-
lactams, e.g. norB mutations in E. coli and Ps. aeruginosa with presumed
nalB mutations (Piddock and Wise, 1989). It is possible to select for cross-
resistance to all fluoroquinolones by in-vitro serial passage of organisms in
the presence of increasing concentrations of the drugs (Neu, 1988).
Repeated transfers may lead to increases in MICs from 0.12 to 16 mg/l for
Ps. aeruginosa and several Enterobacteriaceae, with concomitant
increases in the MICs of 13-Iactams. This suggests that resistance is due to
alteration of the bacterial outer membrane and to permeability factors. In
our experience with Acinetobacter, combined resistance to all 13-Iactams
(83%), all aminoglycosides (42%) and quinolones (30%) is occurring with an
increasing incidence (35% in 1989) (Muller-Serieys et ai., 1989). Although
not yet proven, it is a reasonable assumption that diminished permeability in
Acinetobacter, as described for imipenem resistance, might also be an
acceptable explanation for certain types of multiple resistance, including
resistance to fluoroquinolones, in some Acinetobacter strains. Moreover,
among these strains were isolates with high levels of resistance to penicillins
and cephalosporins, but without any evidence for 13-lactamase activity, which
again suggests the involvement of diminished outer membrane permeability
(Joly-Guillou et aI., 1988). Simultaneous resistance to 13-lactams and
quinolones in these strains also suggests the possible involvement of a gyr B
mutation (Piddock and Wise, 1989). This hypothesis remains to be
investigated.
GENETIC DETERMINANTS OF RESISTANCE IN

ACINETOBACTER Spp
Available information on plasmids and transposons in Acinetobacter

spp. has been reviewed in detail elsewhere (Towner, this volume). So far as
the genetic determinants of resistance are concerned, a limited number of
investigations have provided information on plasmid-mediated resistance.
Although Acinetobacter spp. can easily acquire plasmids from E. coli
(Towner and Vivian, 1976; Chopade et aI., 1985), the opposite transfer
seems to be a rare event (Gomez-Lus et ai., 1980; Goldstein et aI., 1983).
Plasmid patterns of nosocomial Acinetobacter strains were first analysed
by Murray and Moellering (1980) who demonstrated, in a study of
aminoglycoside-modifying enzymes, that loss of resistance in the strains
108
Table 9. Plasmids conferring resistance in Acinetobacter spp.
Plasmid content
of Acinetobacter Molecular size Resistance pattern Ref.
strains
Ap, Km, Sm, Suo (1)
Ap, Tl, Gm, Si, (1)

Cm, Sm, Suo
not named 5 Md Gm, Km, Sm, Ne, (2)

Sp.
not named 16 Md Gm, Sm, Su, Cm, (3)

Tc.
pIP1031 167 kb Ap, Cm, Km, Sm, (4)

Su, Tp.
pIP1841 63 kb Ne, Km, Ak. (5)
Ap, ampicillin; Km, kanamycin; Sm, streptomycin; Su, sulphonamides;

Tl, ticarcillin; Gm, gentamicin; Si, sisomycin; Gm, chloramphenicol;
Ne, neomycin; Sp, spectinomycin; Ak, amikacin; Tc, tetracycline; Tp,
trimethoprim.
amember of the P incompatibility group
References: (1) Gomez-Lus et al., 1980; (2) Murray and Moellering, 1980;
(3) Devaud et al., 1982; (4) Goldstein et al., 1983; (5) Lambert et al.,
1988.
studied was associated with the physical loss of a plasmid; however, no

transfer of resistance, either by filter matings or transformation, was
achieved, and it was not shown conclusively that the plasmid carried the
genes encoding resistance. Nevertheless, despite the inability to
demonstrate transfer of any of these markers, loss of resistance to
kanamycin, gentamicin, sulphadiazine, tetracycline and carbenicillin,
suggested a plasmid location for this genetic information. In another study,
a large self transferable plasmid (pIP1031), belonging to incompatibility
group C( 6) and encoding a TEM-1 B-lactamase and two aminoglycoside
modifying enzymes, (AAD(3")(9» and APH(3 1 )(5")-I), was transferred from
a clinical Acinetobacter strain to E. coli by conjugation at an extremely
low transfer frequency (5 x 10- 1°_ Goldstein et aI., 1983). The multiresistance
of an epidemic strain described by Devaud et al. (1982) could not be
109
transferred to any recipient by various mating procedures, but mobilisation
of resistance to various drugs could be achieved after the conjugative
plasmid RP4 was transferred into the epidemic strain. These authors also
showed that mobilisation resulted from transposition of a 16 Md DNA
sequence from the Acinetobacter chromosome on to plasmid RP4. It was
suggested that a plasmid conferring resistance to multiple antibiotics
(gentamicin, sisomlclll, kanamycin, neomycin, streptomycin,
chloramphenicol, tetracycline, ampicillin and cephalothin) had originally
been transferred from the hospital flora into Acinetobacter spp., and that
the DNA sequence conferring multiple resistance had transposed and stably
integrated into the Acinetobacter chromosome. More recently, in a study
investigating the mechanism of amikacin resistance in Acinetobacter spp.
(Lambert et aI., 1988), the gene conferring resistance to kanamycin and
amikacin in a clinical strain isolated from a urine sample was found to be
encoded by a 63 kb plasmid, p1P1841. This plasmid was self-transferable to
other Acinetobacter strains (A. baumannii, A. haemolyticus and A.
lwoffii), but not to E. coli. The authors suggested that the transfer barrier
resulted from a defect in conjugation, plasmid replication, or both, in the
recipient bacteria (E. coli). The same authors (Lambert et aI., 1990)
showed subsequently that resistance to gentamicin, netilmicin, streptomycin,
sulphonamide, ticarci1lin and tobramycin could also be transferred to strains
of A. haemolyticus and A. lwoffii. Hybridisation experiments using an
aphA-6 probe (the structural gene for the APH(3')VI enzyme - Martin et aI.,
1988) and 14 selected Acinetobacter isolates demonstrated that various
size fragments hybridised with the probe, suggesting that the resistance
determinants were located in different genomic environments. Thus the
authors demonstrated that the aphA-6 gene was involved in the
dissemination of amikacin resistance in Acinetobacter spp. responsible for
nosocomial infections, and that spread of this resistance was due to different
plasmids carrying this epidemic gene. Various other plasmids conferring
resistance or other phenotypic properties have also been studied in
Acinetobacter (Towner and Vivian, 1976; Towner, 1983; Chopade et aI.,
1985); their transfer properties and behaviour in A. calcoaceticus are
described elsewhere (Towner, this volume). Data on naturally-occurring
plasmids conferring resistance in Acinetobacter spp. are gathered together
in Table 9.
CONCLUSIONS
Although a certain amount of work has been done on resistance

mechanisms in Acinetobacter spp., many investigations remain to be
carried out, especially with regard to new antimicrobial agents and the
emergence of resistance to imipenem and the new fluoroquinolones.
Problems of outer membrane permeability are suspected, but not yet
demonstrated. The evolutionary processes which result in the emergence of
new mechanisms of resistance following the therapeutic introduction of new
antibiotics render the task increasingly complicated. This article has tried
to summarise current knowledge on resistance patterns and mechanisms of
110
resistance in Acinetobacter spp. It is clear that epidemic spread of genes
conferring antimicrobial resistance in hospitals will require careful analysis
of plasmid and transposon behaviour in Acinetobacter spp. in order to
determine the possible role of this organism as a reservoir or vehicle of
transmissible drug resistance in the hospital environment.
ACKNOWLEDGEMENTS
We are grateful to Eric Vallee for his contribution to the study of B-

lactamase inhibitors, to K.J. Towner for communication of his publications
and data on plasmids in Acinetobacter spp., to T. Lambert and colleagues
for communication of their most recent data on aminoglycoside resistance,
and to Marie-Jeanne Julliard for careful preparation of the manuscript.
REFERENCES
Bauernfeind, A., 1986, Classification of beta-Iactamases, Rev. I nf ect. Dis.,

8 (Suppl. 5):5470.
Bauernfeind, A., Horl, G., and Monch, V., 1986, Changes in
microbial ecology by therapeutic use of aminoglycosides,
Scand. 1. Infect. Dis., Suppl.49:106.
Bergogne-Berezin, E., and J oly-Guillou, M.L., 1985, An
underestimated nosocomial pathogen, Acinetobacter
calcoaceticus, 1. Antimicrob. Chemother.,16:535.
Bergogne-Beft!zin, E., and Joly-Guillou, M.L., 1986, Comparative
activity of imipenem, ceftazidime and cefotaxime against
Acinetobacter calcoaceticus, 1. Antimicrob. Chemother.,
18(Suppl.E):35.
Bergogne-Berezin, E., and Joly-Guillou, M.L., 1989, Activity of new
quinolones against Acinetobacter from nosocomial
infections, Quinolones Bull., 5:23.
Bergogne-Berezin, E., Zechovsky, N., Piechaud, M., Vieu, J.F., and
Bordini, A., 1971, Sensibilite aux antibiotiques de 240
souches de "M oraxell a" oxydase-negative (Acinetobacter)
isoIees d'infections hospitalieres. Evolution sur 3 ans, Path.
Bioi., 19:981.
Bergogne-Ben~zin E., Joly, M.L., Moreau, N., and Le Goffic, F., 1980,
Aminoglycoside modifying-enzymes in clinical isolates of
Acinetobacter calcoaceticus, Curro Microbiol., 4:361.
Bergogne-Berezin, E., Joly-Guillou, M.L., and Vieu, J.F., 1987,
calcoaceticus, 1. H osp. Infect., 10:105.
Chopade, B.A., Wise, P.J., and Towner, K.J., 1985, Plasmid transfer and
behaviour in Acinetobacter calcoaceticus EBF65/65, J. Gen.
Chow, A.W., Wong, J., and Bartlett, K.H., 1988, Synergistic interactions of
ciprofloxacin and extended spectrum beta-Iactams or
111
aminoglycosides against Acinetobacter calcoaceticus ss.
anitratus, Diagn. Microbiol. Infect. Dis., 9:213.
Dowding, J .E., 1979, A novel aminoglycoside modifying-enzyme from a
clinical isolate of Acinetobacter, 1. Gen. Microbiol., 110:239.
Finland, M., 1979, Emergence of antibiotic resistance in hospital 1935-1975,
Rev. Infect. Dis., 1:4.
1980, A hospital outbreak of antibiotic resistant Acinetobacter
anitratus:epidemiologyandcontrol,l. Hosp. Infect., 1:125.
Freney, J., Bouvet, P.J.M., and Tixier, c., 1989, Identification et
cliniques d'Acinetobacter autres que A. baumannii, Ann.
Bioi. Clin., 47:41.
Garcia, I., Fainstein, V., Leblanc, B., and Bodey, G.P., 1983, 1 n vitro
activities of new beta-Iactam antibiotics against Acinetobacter
spp., Antimicrob. Agents Chemother., 24:297.
Glew, R.H., Moellering, R.C., and Kunz, L.J., 1977, Infections with
Acinetobacter calcoaceticus (H erellea vaginicola).
Clinical and laboratory studies, Medicine, 56:79.
Godineau-Gauthey, N., Lesage, D., Tessier, F., Kollia, D., and Daguet, G.L.,
1988, Acinetobacter calcoaceticus varithe anitratus ou
Acinetobacter baumannii. Etude de la sensibilite aux
antibiotiques de 65 souches hospitali~res, Med. Mal. Infect.,
2bis: 124.
Goldstein, G. W., Labigne-Roussel, A., Gerbaud, G., Carlier, C.,
Collatz, E., and Courvalin, P., 1983, Transferable plasmid
mediated antibiotic resistance in Acinetobacter, PI asmi d,
10: 138.
Gomez-Lus, R., Larrad, L., Rubio-Calvo, M.C., Navarro, M., and Asierra,
M.P., 1980, AAC(3) and AAC(6') enzymes produced by R plasmids
isolated in general hospital, in "Antibiotic Resistance," p.295,
S.Mitsushashi, L. Rosival and V. Kremery, eds, Springer-Verlag,
Berlin.
Hancock, R.E.W., 1984, Alterations in outer membrane permeability, Ann.
Rev. Microbiol., 38:237.
Holton, J., 1982, A report of a further hospital outbreak caused by a
multiresistant Acinetobacter anitratus, 1. Hosp. Infect.,
3:305.
Hood, J., and Amyes, S.G.B., 1989, A novel method for the identification and
distinction of the beta-Iactamases of the genus Acinetobacter, 1.
Appl. Bacteriol., 67:157.
Jacobs, M.R., Aronoff, S.c., Johenning, S., Shlaes, D.M., and Yam abe, S.,_
1986, Comparative activities of the beta-Iactamase inhibitors YTR
830, clavulanate and sulbactam combined with ampicillin and
broad spectrum penicillins against defined beta-Iactamase-
producing aerobic gram-negative bacilli, Antimicrob. Agents
Chemother., 29:980.
112
Joly-Guillou, M.L., and Bergogne-Berezin, E., 1985a, Etude comparative in
a
vitro de l'activite de cinq quinolones vis vis d'Acinetobacter
calcoaceticus, Path. Bioi., 33:416.
Joly-Guillou, M.L., and Bergogne-Berezin, E., 1985b, Evolution
d'Acinetobacter calcoaceticus, en milieu hospitalier, de 1971
a 1984, Press. M ed., 14:2331.
J oly-Guillou, M.L., and Bergogne-Berezin, E., 1990, Presence
d'une beta-Iactamase a spectre elargi chez Acinetobacter
baumanii, Press. M ed., 19:672.
Joly-Guillou, M.L., Bergogne-Berezin, E., and Moreau, N., 1987,
Enzymatic resistance to beta-Iactams and aminoglycosides
in Acinetobacter calcoaceticus, 1. Antimicrob.
Chemother., 20:773.
Joly-Guillou, M.L., Bergogne-Berezin, E., and Philippon, A., 1988,
Distribution of beta-Iactamases and phenotype analysis in clinical
strains of Acinetobacter calcoaceticus, 1. Antimicrob.
Chemother., 22:597.
King, A, and Phillips, 1., 1986, The comparative in vitro activity of eight
newer quinolones and nalidixic acid, J. Antimicrob.
Chemother., 18(Suppl.D):1.
Kitzis, M.D., Goldstein, F.W., Labia, P., and Acar, J.F., 1983, Activite du
sulbactam et de l'acide clavulanique, seuls et associes, sur
Acinetobacter calcoaceticus, Ann. Microbiol. (Inst.Past.),
134A:163.
Knothe, H., and Dette, G.A., 1983, The current state of cephalosporin
antibiotics: microbiological aspects, Infection, 11(Suppl.1):S12.
Labia, R., and Barthelemy, M., 1979, L'enzymogramme des beta-Iactamases:
adaptation en gel de la methode iodometrique, Ann. Microbial.
(lnst.Past.},130B:295.
Chemother., 32:15.
Lambert, T., Gerbaud, G., Bouvet, P., Vieu, J.F., and Courvalin, P., 1990,
Dissemination of amikacin resistance in Acinetobacter spp. due
to synthesis of a 3'-aminoglycoside phosphotransferase type VI,
Antimicrob.Agents Chemother., in press.
Loeffelholz, M.J., and Modrzankowski, M.e., 1986, Plasmid RPI-mediated
susceptibility of Acinetobacter calcoaceticus to rat
polymorphonuclear leukocyte granule contents, Infect. Immun.,
54:705.
Leoffelholz, M.J., and Modrzakowski, M.e., 1987, Outer membrane
permeability of Acinetobacter calcoaceticus mediates
susceptibility to rat polymorphonuclear leukocyte granule
contents, Infect. Immun., 55:2296.
McGowan, J.E., 1983, Antimicrobial resistance in hospital organisms and its
relation to antibiotic use, Rev. I nf ect. Dis., 5: 1033.
Acinetobacter baumannii aphA-6 gene: evolutionary and
113
binding proteins, kinases and other aminoglycoside-modifying
enzymes, Mol. Microbiol., 2:615.
Matthew, M., and Harris, A.M., 1976, Identification of beta-lactamases by
analytical isoelectrofocusing; correlation with bacterial taxonomy,
J. Gen. Microbiol., 94:55.
Mayer, K.H., and Zinner,. S.H., 1985, Bacterial pathogens of increasing sig
nificance in hospital-acquired infections, Rev. I nf ect. Dis.,
7(Supp1.3):S371.
Medeiros, A.A., 1984, Beta-Iactamases, Br. M ed. Bull., 40:18.
Morohoshi, T., and Saito, T., 1977, Beta-lactamase and beta-lactam antibiot-
ic resistance in Acinetobacter anitratus (syn. A. calcoaceti-
cus), 1. Antibiot., 30:969.
Muller-Serieys, c., Lesquoy, J.B., Perez, E., Fichelle, A., Bourjeois, B., Joly-
GuilIou, M.L., and Bergogne-Berezin, E., 1989, Infections noso-
a
comiales Acinetobacter : epictemiologie et difficultes tMrapeu-
tiques, Press. M ed., 18:107.
Murray, B.E., and Moellering, R.C., 1980, Evidence of plasmid-mediated
production of aminoglycoside-modifying enzyme not previously
described inAcinetobacter, Antimicrob. Agents Chemother.,
17:30.
Nayler, J .H.C., 1987, Resistance to beta-lactams in gram negative bacteria:
relative contributions of beta-lactamase and permeability limita-
tions, J. Antimicrob. Chemother., 19:713.
Neu, H.C., 1988, Bacterial resistance to fluoroquinolones, Rev. I nf ect.
Dis., 10 (Suppl.1):S57.
Obana, Y., Nishino, T., and Tanino, T., 1985, In vitro and in vivo activities
1. Antimicrob. Chemother. 15:441.
Paul, G., Joly-Guillou, M.L., Bergogne-Berezin, E., Nevot, P., and Philip-
pon, A., 1989, Novel carbenicillin-hydrolyzing beta-lactamase
(CARB-5) from Acinetobacter calcoaceticus var. anitratus,
Philippon, A.M., Paul, G.c., and Nevot, P.A., 1980, Synergy of clavulanic acid
with penicillins against ampicillin and carbenicillin-resistant
gram-negative organisms, related to their type of beta-lactamase,
Curro Chemother. Immunother., Proc. 11th ICC & 19th
ICAAC, 327.
Phillips, 1., King, A., Shannon, K., and Warren, c., 1981, SQ 26.776: in
vitro antibacterial activity and susceptibility to beta-lactamases,
1. Antimicrob. Chemother., 8(Suppl.E):103.
Piddock, L.J.V., and Wise, R., 1985, Newer mechanisms of resistance to
beta-lactam antibiotics in gram-negative bacteria, J. Antimicrob.
Chemother., 16:279.
Piddock, L.J.V., and Wise, R., 1989, Mechanisms of resistance to quinolones
and clinical perspectives, 1. Antimicrob. Chemother., 23:475.
Pinter, M., and Kantor, M., 1971, Quantitative antibiotic sensitivity pattern
of Acinetobacter strains, Ant. v. Leeuw., 37:197.
114
Quinn, J.P., Studemeister, A.E., Divincenzon c.A., and Lerner, S.A., 1988,
Resistance to imipenem in Pseudomonas aeruginosa: clinical
experience and biochemical mechanisms, Rev. Infect. Dis.,
10:892.
Retailliau, H.F., Hightower, A.W., Dixon, R.E., and Allen, J.R., 1979,
unusual seasonal pattern, 1. Infect. Dis., 139:37l.
Rolston, K.V.I., Ho, D.H., Leblanc, B., Gooch, G., and Bodey, G.P., 1988,
Comparative in vitro activity of the new difluoroquinolone tema-
floxacin (A-62254) against bacterial isolates from cancer patients,
Eur. 1. Clin. Microbiol. Infect. Dis., 7:684.
Rudin, M.L., Michael, J.R., and Huxley, E.J., 1979, Community-acquired
Acinetobacter pneumonia, Amer. 1. M ed., 67:39.
Sanders, e.e., 1983, Novel resistance selected by the new expanded-
spectrum cephalosporins: a concern, 1. Infect. Dis., 147:585.
Sawai, T., Kanno, M. and Tsukamoto, K., 1982, Characterization of eight
beta-Iactamases of gram-negative bacteria, 1. Bacteriol.,
152:567.
Shannon, K.P., King, A., and Phillips, I., 1986, Beta-Iactamases with high
activity against imipenem and SCH 34343 from Aeromonas
hydrophila, 1. Antimicrob. Chemother., 17:45.
Sirot, J., Chanal, c., Petit, A., Sirot, D., Labia, R., and Gerbaud, G., 1988,
Klebsiella pneumoniae and other Enterobacteriaceae produc-
ing novel plasmid-mediated beta-Iactamases markedly active
against third generation cephalosporins: epidemiologic studies,
Rev. Infect. Dis., 10:850.
Towner, K.J., 1983, Transposon-directed mutagenesis and chromosome
mobilization in Acinetobacter cal co acetic us EBF65/65,
Genet. Res., 41:97.
Towner, K.J., and Vivian, A., 1976, RP4-mediated conjugation in Acineto-
bacter calcoaceticus, 1. Gen. Microbiol., 93:355.
Uwaydah, M., and Taql-Eddin, A.R., 1976, Susceptibility of nonfermentative
gram-negative bacilli to tobramycin, 1. Infect. Dis.,
134(Suppl.):S28.
Vila, J., Almela, M., and Jimenez de Anta, M.T., 1989, Laboratory investi-
gation of hospital outbreak caused by two different multiresistant
Acinetobacter calcoaceticus subsp. anitratus strains, 1.
Clin. Microbiol., 27:1086.
Wiedeman, B., Kliebe, E., and Kresken, M., 1989, The epidemiology of beta-
lactamases, 1. Antimicrob. Chemother., 24(Suppl.B):l.
Wise, R., 1982, Beta-Iactamase inhibitors, 1. Antimicrob. Chemother.,
9(Suppl. B):3l.
115
THE CHROMOSOMAL lI-LACTAMASES OF THE GENUS
ACINETOBACTER: ENZYMES WHICH CHALLENGE OUR
IMAGINATION
J. Hood" and S.G.B. Amyes h
-University Department of Bacteriology

Royal Infirmary
Glasgow G4 OSF, UK
bDepartment of Bacteriology, The Medical School

University of Edinburgh
Edinburgh EH8 9AG, UK
INTRODUCTION
At first sight it would appear that the properties of the presumed

chromosomalB-lactamases of the genus Acinetobacter have been fully
characterised. Isoelectric points (pI) in polyacrylamide gel systems were first
described by Matthew and Harris (1976) and reiterated by Sykes and
Matthew (1976). Other workers since then have described the pIs of
Acinetobacter B-lactamases as ranging from 7.5 to > 10 (Medeiros et ai.,
1985); >8 (Joly-Guillou et aI., 1987) and 9.9 (Hikida et aI., 1989). Similarly,
the molecular mass (M) of Acinetobacter B-Iactamase has been described
as 30,000 by Morohoshi and Saito (1977) and as 38,000 by Hikida et al.
(1989).
The kinetics of hydrolysis to various B-Iactam antibiotics have been

most fully described by Morohoshi and Saito (1977) and, subsequently, to
newer agents by Hikida et al. (1989). These authors also examined the effect
of a variety of B-lactamase inhibitors. There is little doubt, however, from
these data that the Acinetobacter B-lactamases are cephalosporinases
(Sykes and Matthew, 1976; Morohoshi and Saito, 1977; Medeiros, 1984;
Bauernfeind, 1986; Neu, 1986; Joly-Guillou et aI., 1987; Hikida et aI., 1989).
117
The inducibility of these cephalosporinases by specific 13-lactam
inducers has been far from clear. They have been described as being
inducible by Morohoshi and Saito (1977) and quoted as inducible by Sykes
and Matthew (1976), Medeiros (1984), Neu (1986) and Hikida et al. (1989).
Bauernfeind (1986) classified them as inducible or constitutive, whereas
J oly-Guillou et al. (1988) s tate d that 13 -lactamase induci bili ty in
Acinetobacter had not yet been clearly demonstrated. On the basis of the
work by Morohoshi and Saito (1977), Bush (1989) classified the
Acinetobacter chromosomal 13-lactamase in Group One, i.e. a
cephalosporinase not inhibited by clavulanic acid (CEP-N). Subsequent
substrate profile and inhibition studies by Hikida et al. (1989) suggested that
this classification was correct.
This article presents evidence that the true character of the

chromosomal 13-lactamase of the genus Acinetobacter is not as simple as
has been proposed. Indeed there appears to be considerable heterogeneity,
with at least four distinct 13-lactamases produced by this genus.
ISOELECTRIC POINTS ON POL YACR YLAMIDE

ELECTROPHORESIS
Matthew and Harris (1976) listed pIs of around 8.6 for two 13-
lactamases produced by strains of Acinetobacter: A. inotti 1786E and
Acinetobacter sp. 1787E. These same two strains appear in Sykes and
Matthew (1976) as A. inotti (1786E) pI 8.6 and A. mallei (1787E) pI 8.7.
The latter strain (A. mallei) 1787E was not an Acinetobacter at all, but a
Pseudomonas mallei. This organism had somewhat controversially been
placed in the genus Acinetobacter by Cowan and Steel (1965) - it became
P. mallei in subsequent editions (Cowan and Steel, 1974). Therefore, this
enzyme can no longer be considered as one produced by the genus
Acinetobacter.
The former strain (A. inotti) 1786E is equally interesting. An

extensive taxonomic literature search has failed to find any previous or
subsequent mention of the specific epithet 'inotti'. This strain is part of the
Glaxo Research Laboratories' collection, and one of the original authors
(Miss A.M. Harris) kindly sent us strain 1786E for further study. The freeze-
dried vial was still marked as A. inotti, but it was identified by the API 20NE
(API System. S.A. France) as A. lwoffii (new specific epithet: A. junii).
On reflection, we suggest that 'lwoffi', when written, could easily have been
wrongly transcribed to 'inotti' and perhaps this error took place somewhere
between its original entry in the strain book and the subsequent publication.
Medeiros et al. (1985) were the first to mention that not all of the
chromosomal 13-lactamases of Acinetobacter focused on conventional
polyacrylamide isoelectric focusing (IEF) systems. Of 14 strains of A.
calcoaceticus, the pI was stated to be 8.8 for one enzyme, 9.0 for one, 9.7
for one, 10.0 for four, > 10 for three and a 'blur' for four. Joly-Guillou et al.
118
(1987) described 30 strains of Acinetobacter with a chromosomal
cephalosporinase of pI > 8. This again suggests poor focusing of these
enzymes on conventional polyacrylamide IEF systems. More recently, Hikida
et al. (1989) described a pI of 9.9 for a purified cephalosporinase from A.
calcoaceticus (ML 4961). However, these authors employed broad range
ampholines of pH 3.5-10.0 in their IEF gel. This suggests that the enzyme had
migrated to, or almost to, the cathode. Unfortunately, there was no
photograph of the IEF gel.
A 10'6
8·3 I
,, _CATHODE
,, •
7·3
pI
.'•
6·45
K',
~,'
5·92
'f
4'75
,
r---,------,-------,-------,- sample
I loaded
5886 66230 H17 H26 H68
B 10'6 -CATHODE
8'3
pI 7·
6'45
5'92
t
if
,
,I-----rl- - - - . ,-------,. ...... sample
H126 H141 H162 SHVI loaded
Fig. 1. Conventional polyacrylamide rEF gels of Acinetobacter chromosomal

B-lactamases and the plasmid B-lactamase SHV-l. Enzyme activity
identified with nitrocephin
119
'"
o
Table 1. Properties of eight aztreonam-resistant strains of Acinetobacter
Acinetobacter strains S-lactamase properties
Mb c
Strain No. sub species new species Specimen type Az~IC ACE type
r
5B86 var. lwoffii A. lwoffii blood 16 >1,000,000 4
6B230 var. anitratus A. junii blood 32 >1,000,000 1
H17 var. anitratus A. baumannii urine 64 640,000 1
H26 var. anitratus A. baumannii sinus swab 32 >1,000,000 1
H68 var. anitratus A. baumannii urine 64 >1,000,000 1
H126 var. lwoffii A. junii wound swab >256 32,500 3
H141 var. anitratus A. baumannii wound swab 64 >1,000,000 1
H162 var. anitratus A. baumannii wound swab 64 60,500 2
aAz aztreonam MIC in mg/1

bM molecular mass
r
cACE = Acinetobacter chromosomal enzyme
We have studied eight partially purified 13-lactamases of
Acinetobacter, obtained from aztreonam-resistant strains collected in the
Royal Infirmary, Edinburgh (Table 1; Hood and Amyes, 1989). These 13-
lactamases were applied to a conventionaliEF system (Fig. 1), but seven of
these enzymes did not focus in this system. Only the 13-lactamase from strain
H126 (A. twoffii, now A. junii) focused, with a pI of about 9.1. We
postulated that the failure of these enzymes to focus resulted from two
factors. Firstly, the basic nature of the enzyme (high pI), and secondly, their
high molecular masses (see Table 1 and subsequent discussion on molecular
masses). An agarose, urea and sorbitol (AUS) gel system was therefore
devised for isoelectric focusing of these enzymes. This combination allowed
the focusing of these basic enzymes (Fig. 2) and their classification into four
groups, termed ACE-l to ACE-4. 13-Lactamases from five of the eight strains
studied were placed in the ACE-l group. Four of these ACE-l enzymes were
derived from A. baumannii - the most prevalent Acinetobacter species
isolated from clinical specimens (Bouvet and Grimont, 1987). A complete
description of the AUS gel technique and further explanation of the IEF
patterns has been published elsewhere (Hood and Amyes, 1989). Further
confirmation of these groupings has been obtained from preliminary studies
employing two additional techniques. Crude enzyme preparations from all
eight strains were applied to either a MONO S or a MONO Q high
performance ion-exchange column connected to a Fast Protein Liquid
Chromatography (FPLC(R» system (Pharmacia, Uppsala, Sweden).
The FPLC method has been described fully elsewhere (Payne et aI.,
1990). Peaks of 13-lactamase activity obtained on the columns were detected
by the nitrocephin spot test (Table 2) and demonstrated the clear differences
between ACE-I, ACE-3 and ACE-4, while suggesting that ACE-2 is similar
to ACE-l. In addition, the results suggest that there may be two subgroups
within the ACE-l group: one with a retention volume of 15-18 ml, the other
with a retention volume of 12-14 mI. However, further work is required to
verify these sub-groupings.
The purified enzymes obtained from the FPLC/ion exchange were also
applied to a gel system on the PhastSystem(R) (Pharmacia, Uppsala, Sweden).
This gel system was devised to resolve basic proteins on a polyacrylamide
(PAGE) minigel. The exact sample preparation and running conditions were
those described by Olsson and Tooke (1988). A Phastgel Homogeneous 20
(i.e. sodium dodecyl sulphate-free) was run with the reversed polarity
electrodes. After electrophoresis, the enzymes were visualised by
nitrocephin solution (500 mg/l). ACE-l (e.g. H141) and ACE-2 (e.g. H162)
preparations migrated similar distances from the cathode, whereas ACE-3
(H126) migrated slightly less far. Insufficient ACE-4 (5B86) was available to
test on this system. Control preparations of the plasmid-mediated 13-
lactamases SHV-l/TEM-l and OXA-2/TEM-2 migrated to appropriate
positions in this gel system, i.e. TEM-l and TEM-2 stayed at the cathode,
while SHY -1 and OXA-2 migrated to a point just behind the ACE enzymes.
Further experimentation employing this technique is required. If it proves
to be reproducible with crude 13-lactamase preparations, it would be an
121
A 10'6 -CATHODE
8·3
7·3
pI
5'85
i i
_sample
loaded
68230 H26 5866 H126 H162
B 10'6 -CATHODE
pI 8'3
7·3
6·45
5'65
4'75
_sample
loaded
H17 H26 H68 H141
Fig. 2. Agarose, urea and sorbitol (AUS) IEF gels. Focused bands of
Acinetobacter chromosomal B-lactamase activity identified with
nitrocephin
A. 6B230 ACE-I B. HI7 = ACE-I

H26 ACE-I H26 = ACE-I
SB86 ACE-4 H68 ACE-l
Hl26 ACE-3 Hl4l ACE-l
Hl62 ACE-2
122
Table 2. FPLC(R} of crude enzyme preparations on high performance ion exchange
Strain no. Retention volume of

Species ACE type
major activity (rnl)
Mono S (cation exchange) in 50 roM phosphate buffer pH7
6B230 ~ junii 1 15-16

H17 A. baumannii 1 17-18
H26 A. baumannii 1 18
Mono Q (anion exchange) in 25 roM Tris/HCl buffer pH8

H126 ~ junii 3 13
5B86 ~~ lwoffii 4 29
eN
'"
extremely useful rapid technique for the study of B-Iactamases obtained from
clinical isolates of Acinetobacter.
MOLECULAR MASS ESTIMATIONS
Morohoshi and Saito (1977) gave the M, of their cephalosporinase

from A. anitratum ,NCrC 7844 as 30,000 when measured by gel filtration on
a Sephadex G-75 column. Hikida et al. (1989) gave the M, of their
cephalosporinase from A. calcoaceticus ML4961 as 38,000. They estimated
this M, by SDS gel electrophoresis after purification of the enzyme with
strong cation exchange and gel filtration.
We estimated the M, of the B-Iactamases in our eight strains of

Acinetobacter by gel filtration, initially on Sephadex G-75 and then, if
required, on Sephacryl S-300 (Hood and Amyes, 1989). The results are listed
in Table 1. One enzyme (ACE-3 from strain H126) had a similar M, (35,000)
to that described by Morohoshi and Saito (1977) and Hikida et al. (1989).
The ACE-2 B-Iactamase from strain H162 had an M, of 60,500, but the other
six enzymes ranged from 640,000 to > 1,000,000. The results suggested that
some of the B-Iactamases of Acinetobacter are either very large, or seem
large because they are linked to other cell envelope components and the
enzyme purification techniques used were incapable of removing these
components. This raises the exciting hypothesis that these enzymes could be
an intermediate step between penicillin binding proteins and the .8-
lactamases. Alternatively, these enzymes may exist in multiple subunit form.
Further experiments are being undertaken to examine these possibilities.
BIOCHEMISTR Y
The specific activities of the eight ACE enzymes were measured for a
selection of cephalosporins, penicillins and a monobactam (Table 3). The
enzyme preparations were partially purified by gel filtration (Hood and
Amyes, 1989). The spectrophotometric assay employed was that of
O'Callaghan et al.(1969) and was carried out on a Pye-Unicam SP1800
UV /VIS spectrophotometer. Cephalosporins and the monobactam were
used at concentrations of either 100 tLM or 1000 tLM (Table 3) and the
penicillins at a concentration of 1000 tLM. The specific activities of the eight
ACE enzymes (Table 3) show clearly that these enzymes are
cephalosporinases, although most of them also have some activity against the
penicillins tested. The apparent detection of weak activity against
cefuroxime in half the strains and weak activity against ampicillin and
carbenicillin in one strain may be dismissed since the spectrophotometer was
working at the limit of its sensitivity and, therefore, activity against these
substrates may also be possible in the other strains. The only clear difference
was the specific. activity of ACE-4 obtained from strain 5B86, which was
considerably lower than the others. Interestingly, ACE-l from strain 6B230
and ACE-3 from H126 seemed similar - they are the only A.junii strains in
the study.
124
Table 3. Specific aGtivities a of ACE enzymes against a selection of cephalosporins and
a monobactam (100 ~M or 1000 ~M) and penicillins (100 ~M)
Strain ACE AZb CAZ b CTX b CXMb CER CED PENG AMP CARB
no. type
5B86 4 25
6B230 1 0.014 330 56 2.8
H17 1 93 69 4.7
H26 1 93 33 6.7
H68 1 0.13 31 19 6.0 0.7 0.46
H126 3 230 46 84
H141 1 0.16 49 17 4.7
H162 2 0.12 93 17 5.8
a~moles of substrate hydrolysed/min/mg protein

bspecific activities also measured with high substrate concentration (1000 ~)
no detectable activity
AZ, aztreonam; CAZ, ceftazidime: CTX, cefotaxime: CXM, cefuroxime; CER, cephaloridine:
CED, cephradine: PENG, penicillin: AMP, ampicillin: CARB, carbenicillin.
N
C11
I\)
m
Table 4. MIC values (mgtl) for ACE-producing strains of a selection of cephalosporins,

penicillins, and a monobactam
Strain no. ACE type AZ CAZ CTX CXM CER CED PENG AMP CARB
5B86 4 16 2 4 8 8 16 16 8 16
6B230 1 32 8 4 16 64 128 >32 128 64
H17 1 64 4 32 64 64 256 >32 32 32
H26 1 32 8 32 64 256 >256 >32 64 32
H68 1 64 8 16 32 64 256 >32 32 32
H126 3 >256 64 64 64 128 256 >32 128 128
H141 1 64 8 16 64 64 256 >32 64 32
H162 2 64 8 32 64 64 256 >32 32 32
Abbreviations as in Table 3.
Table 4 shows the MIC values of the substrates used in Table 3 for
these strains. The results showed that all the Acinetobacter strains
studied had a high degree of resistance to penicillins and first, second and
third generation cephalosporins. There was some correlation between the
resistance levels and the enzyme produced. All the strains encoding the
ACE-1 enzyme had very similar MICs of all the drugs tested. H 162, which
encoded ACE-2, could not be distinguished from the ACE-1 strains on its
resistance profile. However, strain H126, which encoded ACE-3, was
generally more resistant to the drugs tested than the ACE-1 or ACE-2
producers. Specifically, H126 was more resistant to the third generation
cephalosporins and the monobactam, aztreonam. On the other hand, strain
5B86, which encoded ACE-4, was generally less resistant than the other ACE
producers. It was particularly less resistant to the penicillins and the first
generation cephalosporins tested.
Table 5 shows the kinetics of hydrolysis for each of the enzymes, i.e. Km
values with nitrocephin and cephaloridine as substrates. This value was
obtained by measuring the rate of hydrolysis at limiting substrate
concentrations and Lineweaver-Burk plots. The Km values obtained with both
nitrocephin and cephaloridine were broadly similar to each other (Table 5).
The Km values to cephaloridine ranged from 150 ,uM to 710 J.lM and were
similar to the values found by Morohoshi and Saito (1977) and Hikida et al.
(1989) of 250 ,uMand511 ,uMrespectively.
These results show that all four ACE enzymes have moderate affinity
for cephaloridine and nitrocephin. All the Km values were within the same
order of magnitude and could not convincingly be distinguished from one
Table 5. Kinetic data for ACE enzymes
Strain no. ACE type ~a(,um) ~a(J.lM)

Nitrocephin Cephaloridine
5B86 4 220 710

6B230 1 280 150
HI? 1 550 160
H26 1 250 150
H68 1 280 180
H126 3 150 380
H14l 1 220 210
Hl62 2 250 260
aKm , this value was obtained by measuring rates of hydrolysis

at limiting substrate concentrations and determined by
Lineweaver-Burk plots
127
Table 6. 1D50a of ACE enzymes with nitrocephin as substrate (~M)
Strain no. ACE type Aztreonam Cloxacillin Clavulanate
5B86 4 >100 0.18 >100

6B230 1 28 0.012 >100
H17 1 8 0.005 >100
H26 1 5 0.022 >100
H68 1 9 0.1 >100
H126 3 0.08 0.003 >100
Hl41 1 l.8 0.02 >100
H162 2 23 0.022 >100
aID 50 = amount of inhibitor required for 50% inhibition of

nitrocephin hydrolysis
another. Thus none of the ACE enzymes demonstrated greater substrate

affinity than the others and Krn values were a poor discriminator.
Table 6 shows the effect of the B-Iactamase inhibitors

aztreonam, cloxacillin and clavulanic acid, expressed as the concentration
required for 50% inhibition (lDso) of enzyme activity. The enzymes from all
the strains were readily inhibited by cloxacillin. Most were inhibited within
the range of 0.003-0.022 ~M. However, two enzymes required at least 0.1 ~M
cloxacillin for 50% inhibition. One of these enzymes was ACE-I, but this
level of cloxacillin is still considered to be very low. Thus cloxacillin
inhibition also seems to be a poor discriminator of the ACE enzymes,
although a characteristic feature of them all is that they are cloxacillin
sensitive. Similarly, the enzymes from all the strains were extremely resistant
to clavulanic acid inhibition. The inability of any of these enzymes to be
inhibited significantly by 0.1 mM clavulanic acid shows that clavulanic acid
resistance is a characteristic feature of all these enzymes, but is incapable of
being used as a discriminator.
Differences in inhibition were, however, identified when

aztreonam was used as an inhibitor. There was little difference in the IDsofor
the ACE-l and ACE-2 enzymes, which were in the range of 1.8 to 28 ~M
(Table 6). The ACE-3 enzyme from strain H126 was much more sensitive to
aztreonam inhibition than the others (IDso = 0.08 ~M), while the ACE-4
enzyme from strain 5B86 was much more resistant (IDso = > 100 ~M). We
believe that the inhibition profiles with aztreonam provide good
discrimination of the ACE-3 and ACE-4 enzymes.
Table 7 shows the effect of EDTA, HgCl2 and pCMB. All the enzymes
were inhibited to the same extent by a fixed concentration of HgCI 2, but were
128
Table 7. Effect of inhibitors and metal ions on ACE enzymes with
nitrocephin as substrate
Percentage inhibition
Strain no. ACE type EDTAa HgCl 2b pCMB c
5B86 4 <20 89 0
6B230 I <20 98 0
Hl7 I <20 96 0
H26 1 <20 97 0
H68 I <20 95 0
Hl26 3 <20 95 0
Hl41 1 <20 98 0
Hl62 2 <20 98 0
a 1 ruM EDTA (inhibition variable: see text)

b 1 ruM HgC12
c 0.1 ruM p-chloromercuribenzoate
virtually unaffected by moderately high levels of pCMB. Thus no

discrimination of these enzymes could be obtained from these profiles.
The effect of 1 mM EDTA on the enzymes with either nitrocephin or

cephaloridine as the substrate was variable, although generally below 20%.
Hikida et al. (1989) found that EDTA had no effect on their enzyme's
activity. Since many of the enzymes described here are of large M, and,
therefore, may exist in subunit form, it is possible that metal ions playa role
in their function. This may explain the variable effects of EDTA on their
activities. However, further experiments with more purified enzyme will be
required to elucidate these findings.
The above biochemical data (EDTA apart) place all of these 8-

lactamases firmly in Bush Group One (Bush, 1989).
INDUCTION EXPERIMENTS
Morohoshi and Saito (1977) described how the enzyme actIvities of

five strains of A. anitratum "increased by about five to ten-fold" following
treatment with 100 ILg of benzylpenicillin, cephaloridine or 6-
aminopenicillanic acid for 1 h at 37° C. Hikida et al. (1989) quote
Morohoshi and Saito (1977) as having shown that Acinetobacter
cephalosporinases are inducible, but they did not carry out induction
experiments themselves.
129
We have carried out induction experiments with all eight strains,
employing cefoxitin as the inducer at one quarter the MIC value for the
culture. The method employed was that described by Minami et al. (1980).
No discernible £-lactamase induction was found with any of the strains
studied. In discussions with D. Livermore and c.c. Sanders (personal
communications) it was stated that evidence of induction had not yet been
found in any strain of Acinetobacter tested. This includes the use of
cefoxitin, imipenem and penicillin as inducers. Much more convincing
evidence would therefore be required before it can be claimed that these
enzymes are inducible, since current evidence suggests they are not.
ROLE OF PRESUMED CHROMOSOMAL £-LACTAMASES OF

ACINETOBACTER IN £-LACTAM RESISTANCE
Although it seems likely that these enzymes are largely responsible for
the observed resistance to £-lactam drugs, it is also probable that other
factors, e.g. altered permeability or alterations in the penicillin binding
proteins,may be equally important in some strains. The ultimate test must
involve the cloning of the £-lactamase gene from each of these strains into a
suitable recipient, thereby allowing further study of the resistance
mechanism. This would also enable their place on the bacterial chromosome,
rather than on a plasmid, to be confirmed.
CONCLUSIONS
As one would expect from a genus that is very

heterogeneous, with at least 17 genospecies (Grimont and Bouvet, this
volume), the presumed chromosomal H-Iactamases are also heterogenous.
At least four different ACE enzymes have been described in Acinetobacter
on the basis of IEF in AUS gels, molecular mass and biochemical properties.
As has been outlined above, these are very basic proteins with pIs probably
greater than 10. So basic, in fact, that it is difficult to explain how the cell
might secrete them into the periplasmic space; thus they may be bound to
the cell membrane. As a group, they generally do not focus well, or at all, in
conventional polyacrylamide IEF systems. It is possible that further different
chromosomal enzymes exist, since we describe ACE-1 to ACE-4 in only three
of the possible 17 genospecies, i.e. in A. baumannii, A. junii and A.
twoffii.
Further work is needed to explain the large M, of the enzymes found in

six of the eight strains. This may then shed light on the possible evolutionary
place of these interesting enzymes. Future work on the eight enzymes should
include further purification, perhaps by FPLC. This would allow closer
comparison with the biochemical kinetic data obtained by Hikida et al.
(1989) for their enzyme. It would be interesting to know the exact
genospecies of both the Hikida and Morohoshi strains. It is clear, however,
that the enzymes described in this article, together with those described by
130
Morohoshi and Saito (1977) and Hikida et al. (1989), are all
cephalosporinases, not inhibited by clavulanic a~id, and can be placed in
Bush Group One (CEP-N).
ACKNOWLEDGEMENTS
We would like to thank E.R. Squibb and Sons Ltd. for financial
support for some of this work; Mr R. Paton for technical assistance; Miss
A.M. Harris of Glaxo Research Ltd. for kindly providing us with the strain of
A. inotti; Miss Michaela Torrance of Pharmacia for the use of the FPLC
system and the editor of the Journal of Applied Bacteriology for allowing us
to reproduce some of our previous data.
REFERENCES
Bauernfeind, A., 1986, Classification of .B-Iactamases, Rev. 1 nf ect. Dis., 8

(supp1.5):S470.
Bouvet, P.J .M., and Grimont, P.A.D., 1987, Identification and biotyping of
clinical isolates of Acinetob acter, Ann. M icrobiol., 138:569.
Bush, K., 1989, Classification of .B-Iactamases: Groups 1, 2a, 2b and 2b 1 ,
Cowan, S.T., and Steel, K.J., 1965, "Identification of Medical
Bacteria," Cambridge University Press, Cambridge.
Cowan, S.T., and Steel, K.J., 1974, "Manual for the Identification of Medical
Bacteria, 2nd edn.," Cambridge University Press, Cambridge.
Hikida, M., Yoshida, M., Mitsuhashi, S., and Inoue, M., 1989, Purification
and properties of a cephalosporinase from Acinetobacter
calcoaceticus, f. Antibiot., 42:123.
Hood, J., and Amyes, S.G.B., 1989, A novel method for the identification and
distinction of the .B-lactamases of the genus Acinetobacter, f.
Joly-Guillou, M.L., Bergogne-Ben'!zin, E., and Moreau, N., 1987, Enzymatic
resistance to .B-Iactams and aminoglycosides in Acinetobacter
calcoaceticus, 1. Antimicrob. Chemother., 20:773.
Joly-Guillou, M.L., Vallee, E., Bergogne-Ben!zin, E., and Phillipon, A.,
1988, Distribution of .B-lactamases and phenotype analysis in
clinical strains of Acinetobacter calcoaceticus, J.
Matthew, M., and Harris, A.M., 1976, Identification of .B-Iactamases by
analytical isoelectric focusing: correlation with bacterial
taxonomy, 1. Gen. M icrobiol., 94:55.
Medeiros, A.A., 1984, .B-lactamases, Br. M ed. Bull., 40: 18.
Medeiros, A.A., Hare, R., Papa, E., Adam, C, and Miller, G.H., 1985. Gram
negative bacilli resistant to third-generation cephalosporins: .B-
lactamase characterisation and susceptibility to Sch 34343, J.
Antimicrob. Chemother., 15(suppl.C):119.
131
Minami, S., Yotsuji, A., Inoue, M., and Mitsuhashi, S., 1980, Induction of B-
lactamase by various B-Iactam antibiotics in Enterobacter
cloacae, Antimicrob. Agents Chemother., 18:382.
Morohoshi, T., and Saito, T., 1977, B-Lactamase and B-Iactam antibiotics
resistance in Acinetobacter anitratum, J. Antib iot., 30:969.
Neu, H.C., 1986, Antibiotic inactivating enzymes and bacterial resistance, in
"Antibiotics in Laboratory Medicine," p.757, V. Lorian, ed.,
Williams and Wilkins, Baltimore.
O'Caliaghan, C.H., Muggleton, P.W., and Ross, G.W., 1969, Effects of B-
lactamase from Gram negative organisms on cephalosporins and
penicillins, Antimicrob. Agents Chemother., 1968:57.
Olsson, I., and Tooke, N.E., 1988, PAGE of basic proteins, in
"PhastSystemRApplication File No. 300," Pharmacia, Uppsala.
Payne, D.l., Hood, 1., Marriott, M.S., and Amyes. S.G.B., 1990, Separation of
plasmid mediated broad spectrum B-lactamases by Fast Protein
Liquid Chromatography (FPLCR), FEMS Microbiol. Lett.,
69:195.
Sykes, R.B., and Matthew, M., 1976. The B-lactamases of Gram negative
bacteria and their role in resistance to B-lactam antibiotics, J.
132
THE USE OF MOLECULAR TECHNIQUES FOR THE LOCATION
AND CHARACTERISATION OF ANTIBIOTIC RESISTANCE GENES
IN CLINICAL ISOLATES OF ACI NETOBACTER
B.G. Elisha and L.M. Steyn
Department of Medical Microbiology

University of Cape Town & Groote Schuur Hospital
Observatory, 7925
South Africa
INTRODUCTION
Nosocomial isolates of Acinetabacter, particularly A. baumannii,

are often highly resistant to a variety of antibiotics (Bergogne-Berezin et
aI., 1987). Although much is known about the biochemical mechanisms of
antibiotic resistance found in Acinetabacter (Bergogne-Berezin and Joly-
Guillou, this volume), little is known about the structure and regulation of
the expression of these genes in Ac ine ta b act e r at the molecular level.
One of the reasons why there is a paucity of data on the primary

structure of resistance genes in Acinetabacter is the difficulty encountered
in locating these genes by conventional gene transfer experiments (Vivian et
al., 1981; Chopade et al., 1985). For the same reason, data on the precise
genetic location of these genes (i.e. plasmid, chromosome, or both) is often
lacking (Murray and Moellering, 1980; Devaud et al., 1982; Goldstein et al.,
1983; Divers et aI., 1985). This article describes how molecular techniques
can be used to locate and characterise antibiotic resistance genes in clinical
isolates of Acinetabacter. In order to illustrate the use of these
techniques, their application in characterising and determining the genetic
location of chloramphenicol and aminoglycoside (gentamicin and
tobramycin) resistance genes is described in detail. To facilitate the
detection of these genes we first constructed two DNA probes: a 2"-0-
adenylyItransferase (AAD(2"» gene probe and a chloramphenicol
acetyltransferase (CA TI) gene probe. The AAD(2") gene was chosen
because it is the only enzyme with activity against gentamicin and
tobramycin that has been demonstrated in local Acinetabacter isolates
133
(unpublished observations). The construction and use of these probes to
locate the corresponding resistance genes in a multi-resistant clinical isolate
of Acinetobacter (strain SAK) is described. Subsequent cloning
experiments not only confirmed the presence of these genes, but also
revealed another aminoglycoside resistance gene with activity against
gentamicin and tobramycin in strain SAK; this was further analysed using
DNA sequencing techniques. Thus the studies described in this article not
only illustrate the application of molecular techniques to the study of
antibiotic resistance in Acinetobacter, but also serve to demonstrate the
complex nature of the problems which may be encountered in multi-resistant
clinical isolates of this genus.
CONSTRUCTION OF DNA PROBES
CAT Probe
The plasmid pAIOcat2 (Rosenthal et aI., 1983) contains the gene

encoding CATL The structural CATI gene is 650 bp long, of which the first
520 bp was used as a probe. To generate this fragment, pAIOcat2 was
digested with H indIII and N col, the fragments were separated by agarose
gel electrophoresis and the 520 bp fragment was eluted from the gel (Seth,
1984). The eluted DNA was radio-labelled with 32P-dCTP by nick translation.
EB S SB S B E
II I II I I I pGSH10l
E~B sl~ IS f IE
pGSH103
II
HPSm.
Exom_
S B E
I I I pGSH104
I
H
S B
I I
I II pGSH105
H KE
_Exom
S
lkb I
~
1=1 pGSH106
H E
Fig.l. Construction of the AAD(2") gene probe. = , insert DNA;

----, vector DNA. B, BamHI; E, EcaRI; K, KpnI; P, PstI; S, SacI; Sm,
SmaI; ExaIII, exonuclease III. The vector restriction sites are shown
below the bar. The BamHI sites were mapped in pGSHl03 and are therefore
inferred in pGSHlOl.
134
AAD(2/t) Probe
The gentamicin/tobramycin resistance gene, encoding AAD(2"), was

isolated from the multi-resistance plasmid, JR66 (Benveniste and Davies,
1971). JR66 DNA was cut with EeoRI and the restriction fragments cloned
in the EeoRI site of pBR322. Gentamicin-resistant transformants were
selected; all were resistant also to tobramycin and harboured an identical
recombinant plasmid, designated pGSH101, which contained a 5.3 kb insert
(Fig. 1). AS ad fragment of 1.8 kb was deleted from pGSH101 by a partial
S ad digestion followed by re-ligation (forming pGSH102). The 3.5 kb
EeoRI insert from pGSH102 was subcloned in pUC18 to form pGSH103.
This sub cloning was a prerequisite for controlled progressive deletion with
exonuclease III (Henikoff, 1984) to reduce the size of the insert further
while still retaining a functional AAD(2")gene. For the exonuclease III
digestion, pGSH103 was cut with PstI to generate a 3' -cohesive end adjacent
to the vector DNA, while S m aI was used to generate an exonuclease III
susceptible end adjacent to the insert. No gentamicin/tobramycin-resistant
transformants were obtained when more than 1.2 kb of the insert DNA was
deleted. One of the smallest gentamicin-resistant plasmids (pGSH104)
contained an insert of 2.3 kb, from which a B amHI/ H indIII fragment (1.8
kb) was subcloned into pUC18 to form pGSH105. A second digestion with
exonuclease III was performed to delete insert DNA at the opposite end of
the AAD(2") gene. BamHI was used to generate a susceptible site, while
KpnI was used to protect the vector; 0.7 kb of insert DNA was deleted before
loss of AAD(2") activity occurred. The size of the reduced DNA insert
containing the functional AAD(2") gene was 1.1 kb (pGSH106). This insert
was radio-labelled by nick translation and used as a probe.
GENE LOCATION
Preparation of DNAfor Hybridisation Experiments
DNA was isolated as described by Clewell and Helinski (1969).

Plasmid DNA was purified by ultra-centrifugation in a CsCl!EtBr density
gradient. The less-dense DNA band in the gradient was also isolated and
designated chromosomal DNA. Plasmid and chromosomal DNA was
filtered on to Hybond-N membranes in a dot-blot apparatus. When required,
DNA was transferred from agarose gels to Hybond-N by the method of
Southern (1975). Hybridisations were done as described by Johnson et al.
(1984).
Hybridisation with the CAT Probe
The CAT probe was used to detect the CAT structural gene, and to
determine its genetic location in the multi-resistant strain SAK. The probe
gave no hybridisation signal with a sensitive strain of Aeinetobaeter
(YON), but generated a signal with pA10cat2, the source of the probe
fragment. The probe hybridised to both chromosomal and plasmid DNA
preparations from SAK (Fig. 2).
135
-:;'"
U '"
to
"<
0
z
<
u
z
'"
0
A 0
I/)
> a.
B <d:
I/)
0
>
<
a.
S.Oug 5.0u9 1.0ng

1.0ng
1.0u9
0.5ng 1.0ug 0.5ng
o.Sug 0.25ng
0.5ug 0.25ng
Fig.2. Chromosomal and plasmid DNA dot blots hybridised with the CAT
gene probe. (A) Chromosomal DNA, (B) Plasmid DNA. Autoradiography
was for 72 h.
Hybridisation with the AAD( 2") Prob e
The probe for the AAD(2") structural gene hybridised to JR66, the
source of the AAD(2") gene. There was no hybridisation with DNA from a
sensitive strain; however, a signal was obtained from the chromosomal DNA
preparations (Fig. 3), but not from the plasmid preparations of strain SAK.
CLONING AND CHARACTERISATION OF THE ANTIBIOTIC

RESISTANCE GENES
On the basis of the above results it was concluded that strain SAK
contained the resistance genes encoding CAT and AAD(2"). Further support
for this hypothesis was sought from cloning experiments.
Fig.3. Chromosomal DNA dot-blot hybridised with the AAD(2") gene probe.
Autoradiography was for 18 h.
136
Cloning of the CAT Gene
Chromosomal DNA was digested with either BamHI, EcoRI, HindIII

or PstI and cloned into the respective sites in pUC18. E. coli transformants
resistant to chloramphenicol were only obtained from the HindIII
fragments. Of six recombinant plasmids, each of which contained an insert of
11.0 kb and hybridised to the CAT probe, one was selected for further
manipulations. The insert was reduced in size by subcloning a 1.8 kb PstI
fragment which hybridised to the CAT probe into pUC18. This recombinant
plasmid (pGSH201) carried afunctional CAT gene.
A similar procedure was used to isolate the CAT gene from plasmid
DNA. All of the recombinant plasmids contained a 5.0 kb H indill insert and
encoded resistance to chloramphenicol. Hybridisation experiments
demonstrated that this 5.0 kb fragment was present in a 50-60 kb plasmid
harboured by strain SAK (Fig. 4). None of the chloramphenicol-resistant
recombinant plasmids derived from strain SAK encoded amino glycoside
resistance.
Identification of CAT Modifying Enzymes

Assays for CAT activity were performed as described by Gorman et
al. (1982). Crude lysates prepared from strain SAK and E. coli (pGSH201)
acetylated 14C-chloramphenicol to its O-acetoxy derivatives (Fig. 5), thereby
confirming the presence of a CAT gene in the Acinetobacter strain.
Fig.4. Plasmid DNA hybridised to the CAT gene probe. (A) Plasmid profile
of strain SAK following electrophoresis on a 0.8% agarose gel for 16 hat 20
V (1 V/cm). (B) Lane 1: gel-purified 5.0 kb HindIII fragment which
contains the CAT gene cloned from plasmid DNA; lane 2: HindIII digest of
plasmid DNA. A 0.8% agarose gel was electrophoresed·for 2 h at 70 V (7
V/cm). (C) Autoradiograph of DNA shown in (B) hybridised with CAT gene
probe.
137
Fig.S. Assay of CAT activity. Lane 1: extract from strain SAK; lane 2:extract
from E.coli (pGSH20l). (A) l-acetate chloramphenicol; (B) 3-acetate
chloramphenicol; (C) 1,3-diacetate chloramphenicol.
Cloning of Aminoglycoside Resistance Genes
The partial library prepared above was also screened for the presence
of gentamicin resistance genes. Recombinant plasmids encoding gentamicin
resistance were obtained from all the digests. Five recombinants were
screened with the AAD(2") probe: four gave strong hybridisation signals, but
one, which contained a 5.S kb H indIII insert, did not hybridise with the
probe.
The recombinants which hybridised to the probe contained a 2.3 kb

HindIII insert. This was reduced by sub cloning a 1.9 kb BamHI/ HindIII
fragment into pUC18. This recombinant plasmid was designated pGSHlOS
and still conferred gentamicin resistance. The lone recombinant which did
not hybridise with the AAD(2") probe contained an insert of 5.S kb. By
digestion with exonuclease III, the size of the insert was reduced to O.S kb
without altering the expression of gentamicin resistance. This recombinant
plasmid was designated pGSH11O. When the insert from pGSH110 was used
to probe plasmid DNA prepared from strain SAK, a strong hybridisation
signal was obtained (Fig. 6). It was possible to clone this plasmid gene on a
5.S kb H indIII DNA fragment, but it is important to note that the restriction
map of this 5.S kb H indIII fragment was not similar to the 5.S kb H in dIll
fragment cloned from the chromosomal DNA. PstI digested the former
recombinant plasmid into three fragments of 6.5 kb, 1.1 kb and O.S kb
respectively, while the recombinant plasmid containing the plasmid-derived
insert was digested into fragments of 3.S kb, 3.5 kb and 1.1 kb. This insert
was shown to be present in a 50-60 kb plasmid harboured by strain SAK (Fig.
6). Thus the 50-60 kb plasmid carries chloramphenicol and gentamicin
resistance genes which are not closely linked, since the recombinant
plasmids carry either chloramphenicol or gentamicin resistance, but not
both.
138
Fig.6. Plasmid DNA from strain SAK hybridised to the AAC(3) gene probe.
(A) Plasmid DNA dot blot. (B) Lane 1: gel-purified 5.8 kb HindIII fragment
containing the AAC(3) gene cloned from plasmid DNA; lane 2: HindIII
digest of plasmid DNA. A 0.8% agarose gel was electrophoresed for 2 h at 7
V/cm; (C) Autoradiograph of DNA shown in (B) hybridised with the
AAC(3) gene probe.
I dentification of Aminoglycoside Modifying Enzymes
To identify the amino glycoside-modifying enzyme encoded by

pGSHllO, and to confirm that pGSH108 encodes an adenylyItransferase,
aminoglycoside modifying enzymes were assayed by the phosphocellulose
paper binding assay (Shannon and Phillips, 1983). Crude lysates were
prepared from strain SAK and from E. coli cells harbouring the recombinant
plasmids. Sisomicin, a gentamicin derivitive, was used as a substrate in the
screening of cell sonicates for AAC and AAD activity, while 14C-acetyl
coenzyme-A and 14C_ATP were used as co-substrates in the respective
enzymatic reactions. If the amount of radioactivity bound to the paper in the
presence of sisomicin was double the background bound in the absence of
sisomicin, it was assumed that the sisomicin had been modified. No
modifying activity could be detected in lysates from strain SAK (Table 1);
adenylyltransferase activity was detected in lysates prepared from E. coli
(pGSH108) since there was an 83-fold increase in the 14C-ATP counts bound
in the presence of sisomicin. In contrast, the enzyme encoded by pGSH1l0
was shown to be an acetyltransferase as, in this case, there was a 107-fold
increase in the radioactivity bound from I'C-acetyl coenzyme-A. This
enzyme was subsequently identified as an AAC(3) by Bristol Laboratories
(personal communication). Thus strain SAK contains two resistance genes,
AAD(2") and AAC(3), which have activity against gentamicin and
tobramycin. The readily detectable AAD(2") and AAC(3) activity in the
lysates prepared from E. coli (pGSHI08) and E. coli (pGSHI10)
respectively is probably due to the high copy number of the plasmids and the
concomitant high AAD(2") and AAC(3) gene dose.
Nucleotide Sequence of theAAC(3)
There are seven sub-classes of AAC(3): I, la, II, III, IV, V and VII
(Vliegenthart et aI., 1989). The sequences of AAC(3)1 (Tenover et aI.,
139
Table 1. Assay of Aminoglycoside-Modifying Enzymes.
Results are Expressed in Terms of cpm Bound to Phosphocellulose
Paper
i) Assay of O-Adenylyltransferase activity with [14C]_ATP

as co-substrate
Substrate
Bacterial strain None Sisomicin Ratio
SAKa 3291 3505 1.1
E. coli DKI (pGSH106) 339 5432 16.00
E.coli DKI (pGSHl08) 133 11096 83.00
E. coli DKI (pGSH110) 101 128 1.3
ii) Assay of N-Acetyltransferase activity with [14C]-Acety1-

CoA as co-substrate
Substrate
Bacterial strain None Sisomicin Ratio
SAK 612 713 1.2
E.coli DK1 (pGSH106) 2298 2129 0.9
E.coli DKl(pGSHl08) 1823 1866 1.0
E.coli HB101 (pGSH110) 1024 9643 107
a This assay was performed with a radioactive substrate of

higher specific activity than the other assays.
1989), AAC(3)II (Vliegenthart et aI., 1989),AAC(3)III (Allmansberger et

aI., 1985) and AAC(3)IV (Brau et aI., 1984) have been published. It is
important to note that the published sequences for AAC(3)1I and AAC(3)III
are identical. AAC(3)1 has been previously described in Acinetobacter,
but to the best of our knowledge, none of the other sub-classes have been
described in this organism.
To facilitate the classification of the AAC(3) derived from strain SAK,

its corresponding gene was sequenced. Overlapping restriction fragments of
the 0.8 kb DNA insert were subcloned into M13mp18 and M13mp19. The
sequence of the insert was then determined by the dideoxy-chain termination
140
-80 -52 CAP -10 SD-1
GTCCTCCACCTCGGTCAAGCAGCGAGAGGTAG~TATCTCGATAGG!!!!£!GTTTTGCAGTTTAG~Q!Q!TATCGCGA
39 ~!!' 78
ATC CAT Ace CGG AAG eCA ATA Ace GAG GCA ATT eGA A.AA eTC eGA GTC CAA Ace GGT GAG CTG TTG ATC GTG CAT Gee
MET Hi. Thr A:rg Lya Ala lIe Tbr Glu Ala lIe Arg Ly. 'Leu Gly Val GIn Thr Gly Asp Leu Leu Met Val His Ala
117 156
TCA CTT AAA GCG ATT GeT eCG GTC GAA GGA GGA GCG GAG ACe GTC GTT Gee GeG TTA eGe Tce GeG CTT GGG eeG ACT
Ser Leu Lya Ala lIe ely Pro Val Giu Gly Gly Ala Glu Thr Val Val Ala Ala Leu Arg Ser Ala Val ely Pro Thr
195 234
GGC ACT GTG ATC CGA TAC GCG TeG TeG GAC eGA TCA eec TAC GAG GAG ACT CTG AAT GGC GCT eGG TTG GAT GAC AAA
Gly Thr Val Met Cly Tyr Ala Ser Trp Asp Arg Ser Pro Tyr Glu Glu Thr Leu Aao ely Ala Arg Leu Asp Asp Lys
273 312
Gee eGe CGT Ace TeG eeG eCG TTC GAT eec GGA Ace Gce GGG ACT TAC CCT GCG TTC GGC eTG CTG AAT CAA TTT eTC
Ala Ar& Arg Thr Trp Pro Pro Phe Asp Pro Ala Thr Ala Gly Thr Tyr Arg Gly Phe ely Leu Leu Asn GIn Phe Leu
351 390
GTT CAA GCC cee GGe GeG CGG cec AGC GCG CAC eec GAT GCA TCG ATe GTC GCG GTT GGT eCG CTA CCT GAA ACG CTC
Val GIn Ala Pro Gly Ala Arg Arg Ser Ala Hie Pro Asp Ala Ser Met Val Ala Val Gly Pro Leu Ala Glu Thr Leu
G 429 468
Ace GAG ceT CAC GAA CTC GCT eAe Gee TTG ece AAA GGG TCG ecc GTC GAC eGG TTC GTC eGe eTT GGC GGG AAG Gee
Thr Glu Pro His Glu Leu Gly Ria Ala Leu ely iys Gly Ser Pro Val Glu Arg Phe Val Arg Leu Gly Gly Lys Ala
507 546
CTG etG TTG GGT GeG eCG CTA AAC Tce CTT AeC GGA TTG CAC TAC GGe GAG GeG GTT GeG GAT ATC cec AAC AAA eGA
Leu Leu Leu ely Ala Pro Leu A80 Ser Val Thr Ala Leu His Tyr Ala Glu Ala Val Ala Asp Ile Pro Asn Lys Arg
585 624
TeG GTG ACe TAT GAG ATG CCG ATC CTT GGA AGA AAC GGT GAA GTe GCC TeG AAA Ace eCA TCA GAA TAe GAT TCA AAC
Trp Val Thr Tyr Glu Met Pro Met Leu ely Arg Asn Gly Glu Val Ala Trp Ly8 Thr Ala Ser Glu tyr Asp Ser Asn
663 702
GGC ATT eTC CAt TGC TTT GCT ATC GAA GGA AAG CGG GAT GeG GTC GAA ACT ATA GCA AAT GCT TAC GTG AAG CTC GGT
Gly lIe Leu Allp Cy8 Phe Ala lIe Glu Gly LY8 Pro Asp Ala Val Glu Thr lIe Ala Asn Ala Tyr Val Lys Leu Gly
741 780
CGC CAT CGA GAA GGT GTC GTG GGe TTT GCT eAG TGC TAC eTG TTC GAC GeG CAG GAC ATC GTG ACG TTC GGC GTe ACC
Arg Hi. Arg Glu Gly val Val Gly Phe Ala Gin Cys Tyr Leu phe Asp Ala Gin Asp lIe Val Thr Phe Gly Val Thr
819 858
TAT eTT GAG AAG CAC TTC GGA GCC ACT eCG ATe GTG CGA GCA CAC GAA Gce Gec eAG eGC TCT TGC GAG CCT TCC GGT
Tyr Leu Glu Lys His Phe Gly Ala Thr Pro lIe Val Pro Ala His Glu Ala Ala Gin Arg Sel" Cys Glu Pro Ser ely
SaIl
TAG AGGCCGTCGAC
***
Fig. 7. Nucleotide sequence and derived amino acid sequence of the AAC(3) gene. The -10 region, ribosome binding
site (SD) and putative CAP binding site are underlined. The stop codon is indicated by ***. The G to A transition
at position 425 is shown. The Sphl site and the Sall site confining the gene fragment used as the probe are
.,. indicated.
method (Sanger et al., 1977), using 35S-ATP and Sequenase (US
Biochemicals). Internal oligonucleotide primers were used when necessary.
The DNA sequence shown in Fig. 7 was confirmed by sequencing both
strands of DNA. This sequence was compared with the published AAC(3)
sequences; it showed 99% similarity with the DNA sequence of AAC(3 )11,
also designated AAC(3 )111. The only difference in the structural gene
sequence is a transition at position 425 of a G to an A, which results in the
incorporation of lysine rather than glutamic acid (Fig. 7). The sequence
similarity extends into the regulatory region, including the -10 region.
However, unlike the published sequences, the strain SAK AAC(3) sequence
has no sequence corresponding to the E. coli -35 consensus sequence,
TTGACAT (Fig. 8). McClure (1985) suggested that promoters with a poor
homology to the consensus sequence in the -35 region are frequently
controlled by activator proteins. One such protein is CAP, which binds to
cAMP to form an active complex. De Crombrugghe et al. (1984) have
shown that 13 out of 17 regulatory sequences which bind the cAMP ICAP
complex contain a conserved sequence, TGTGA, at variable distances from
the -10 region. A similar sequence, AGTGA, is present in the AAC(3)
sequence at -32 (Fig. 8).
EXPRESSION OF AMINOGLYCOSIDE RESISTANCE GENES IN

STRAIN SAK
Since aminoglycoside modifying activity was not readily detectable in

strain SAK, it was unclear whether the resistance could be attributed to one
or both of the resistance genes. To clarify this, RNA was extracted from
strain SAK and probed by Northern hybridisation with the AAD(2") probe
and the cloned AAC(3) structural gene.
Detection of mRNA Transcripts
Total cellular RNA was extracted from strain SAK with hot phenol as
described by Aiba et al. (1981). RNA from cells grown in the presence and
absence of gentamicin (5 mg/l) was analysed. In addition, RNA from cells
treated with rifampicin (200 mg/l) to block further initiation of transcription
was included. RNA (10 j.tg) was electrophoresed in gels comprising 1 %
agarose and 35% formaldehyde with MOPS buffer (40 mM MOPS, 10 mM
sodium acetate, 1.0 mM EDTA, pH 8.0) and transferred to Hybond-N in 20 x
SSPE (1 x SSPE: 1.8 M NaCl, 10 mM Na 2 HP0 4, 1.0 mM EDTA, pH 7.0). Pre-
hybridisations and hybridisations were carried out as described by Johnson et
al.(1984).
The AAC(3) probe hybridised to a mRNA transcript which remained

detectable for more than 60 mins after treatment of the exponentially
growing cells with rifampicin (Fig. 9). In contrast, the AAD(2") probe did
not hybridise to mRNA from cells grown in the presence or absence of
gentamicin, although it did hybridise to RNA prepared from E. coli
harbouring pGSHI08 (Fig. 9). Based on this, it appears that the AAC(3)
142
-175 -140
AAC(3) - III TGATGCCGTATTTGCAGTACCAGCGTACGGCCCACAGAATGATGTCACGCTGAAAATGCCGGCCTT AAC(3)-III
-70
AAC(3)(SAK) GTCCTCCACCTCG AAC(3)(SAK)
-35
AAC(3)-ll TCACCACCGACTATTTGCAAC AAC( 3) - II
.......................
.. .. .. . .. .. .. .. .. .. . .. . .. .. .. .. .. . . .
AAC(3)-III TGAATGGGTTCATGTGCAGCTCCATCAGCAAAAGGGGATGATAAGTTTATCACCACCGACTATTT~~~~C AAC (3) - I II
-36 -1
AAC(3)(SAK) GTCAAGCAG~~~Q~Q~TAQAGTGATATCTCGA!~QQ!~!~GTQTTTTGCAGTTTAGAGGAGATATCGCG AAC(3)(SAK)
: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :
AAC(3)- II AGTGCCGGCCGAGAGGTAGAGTGATATCTCGATAGGTATAGTGTTTTGCAGTTTAGAGGAGATATCGCG AAC(3)-II
-------------- : : : --------------
: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :
AAC(3) - II I AGTGCC--------------------TCTCGATAGGTATAGTGTTTTGCAGTTTAGAGGAGATATCGCG AAC(3)-III
Fig. 8. Regulatory region of AAC(3) from SAK, AAC(3)II and AAC(3)III. The proposed -35 region of AAC(3)II and
AAC(3)III are underlined, as are the almost perfect direct repeats spanning the -10 region in AAC(3) from SAK and
AAC(3)II.
w
"""
Fig.9. Northern blots hybridised with the AAC(3) and AAD(2") gene probes.
(A) RNA hybridised with the AAC(3) probe; RIF indicates the addition of
rifampicin to the cells; (B) RNA hybridised with the AAD(2") probe; lane
l:RNA from strain SAK; lane 2: RNA from strain SAK grown in the presence
of gentamicin (5 mg/l); lane 3: RNA from E.coli (pGSH108) which contains
the cloned AAD(2") gene from strain SAK.
gene alone is responsible for the gentamlclD and tobramycin resistance

phenotype. In an attempt to try and "switch on" the AAD(2") gene, antibiotic
selective pressure was applied in the form of kanamycin. Kanamycin was
chosen because this antibiotic is included in the AAD(2"), but not the
AAC(3), substrate profile. Strain SAK (kanamycin MIC of 4 mg/l) was
cultured in media containing kanamycin, the concentration of which was
increased incrementally until strain SAK was resistant to 120 mg
kanamycin/l. Total cellular RNA was extracted from cells grown in the
presence of 120 rug kanamycin/l and probed with the AAD(2") and AAC(3)
DNA probes. Transcripts which hybridised only to the AAC(3) probe were
detected (data not shown).
DISCUSSION
The aim of this article is to illustrate the use of techniques designed to

elucidate the molecular nature of antibiotic resistance genes in multi-
resistant clinical isolates of Acinetobacter. DNA probes for the AAD(2")
and CAT genes were prepared and used to probe chromosomal and plasmid
DNA preparations from a resistant strain of Acinetobacter. Both probes
hybridised to chromosomal DNA from the multi-resistant strain SAK, while
the CAT probe hybridised also to plasmid DNA in this strain. Effective
resistance to gentamicin or chloramphenicol could be transformed into
susceptible strains of E. coli after cloning strain SAK chromosomal
fragments into pUC18. Antibiotic modifying enzyme assays confirmed the
presence of AAD(2") and CAT activity in extracts prepared from cells
containing the respective recombinant plasmids.
144
The CAT gene was also shown to be carried on a 50-60 kb plasmid
present in strain SAK. A chromosomal and plasmid locus for the CAT gene
suggests the possibilty of a transposon. A transposon encoding
chloramphenicol, gentamicin, streptomycin and sulphonomide resistance has
been identified in Acinetobacter by Devaud et ai. (1982), and the presence
of inverted repeat sequences flanking a kanamycin resistance gene in the
Acinetobacter plasmid pAV5 (Divers et aI., 1985) indicates that this gene
may also be part of a transposon. The probe specifies a Type I CAT enzyme
that is also encoded by Tn9 and preliminary sequencing data show that the
CAT gene derived from strain SAK is on a transposon which has sequence
and structural similarity to Tn2670 (Womble and Rownd, 1988).
An additional aminoglycoside resistance gene was cloned from

plasmid and chromosomal DNA prepared from strain SAK. This gene was
identified as encoding an AAC(3) which, like AAD(2"), has activity against
gentamicin and tobramycin. Like the CAT gene, the AAC(3) gene is also
contained on a 50-60 kb plasmid harboured by strain SAK. Sequencing data
demonstrated almost perfect sequence homology with the published
sequences for AAC(3)1I (Vliegenthart et aI., 1989) and AAC(3 )111
(Allmansberger et aI., 1985). However, the control region was different from
both of the published sequences. Although the sequence derived from SAK
has a consensus -10 sequence, there is poor similarity with the -35 sequence.
A sequence located 12 bp from the -10 region, is, however, similar to the
consensus sequence described for cAMP ICAP controlled promoters. Little
is known about transcriptional control signals in Acinetobacter, so further
studies are required to prove conclusively that the AAC(3) has a catabolite
sensitive promoter in strain SAK. Catabolite repression of the AAC(3) was
not anticipated since, to our knowledge, only two aminoglycoside modifying
enzymes, AAD(3") (Harwood and Smith, 1971) and APH(3')II (Davies and
Smith, 1978), are known to be subject to catabolite expression. It is well
known that CAT is subject to catabolite repression in Gram-negative bacilli.
Why the AAC(3) in strain SAK, or indeed any antibiotic resistance gene,
should be subject to a form of regulation associated with the availability of
energy sources is puzzling. Perhaps, as Shaw (1983) reasons, special
metabolic demands are placed on microbes which inhabit ecological niches
where antibiotics are found. Acinetobacter is a common soil organism and
it may be that this control aspect in strain SAK is related to the origins of the
organism.
It has been suggested by Prince and Jacoby (1982) that cloning a

resistance gene into a high copy number plasmid is useful in elucidating the
mechanism of aminoglycoside resistance in bacteria with little or no
demonstrable enzyme activity. Although no AAD(2") or AAC(3) activity was
detected in the extracts prepared from strain SAK, it was possible to identify
these enzymes when the corresponding gene was cloned into a high copy
number plasmid in E. coli. This study suggests, however, that cloning a
functional resistance gene is not sufficient in itself to elucidate the
mechanism of resistance. Judged by the cloning experiments alone, strain
SAK contains two mechanisms of resistance to gentamicinltobramycin, but
145
the RNA transcript experiments indicated that only the AAC(3) gene is
transcribed, and is therefore responsible for the resistance phenotype. It
follows that it is insufficient to simply identify a resistance gene with a DNA
probe; the gene must be shown to be expressed in the host strain before a
role in the resistance phenotype can be assigned.
Why the AAD(2") is not expressed in strain SAK is not known, but it is
possible that transcription of the AAD(2") is repressed by a protein which is
absent in E. coli. The AAD(2") gene is currently being sequenced to see if
there are any DNA sequences that could be involved in the control of
transcription of this gene. One other possibility is that the AAC(3) gene is
favoured because it has a stronger promoter.
The presence of an antibiotic provides a strong selective pressure

which favours the survival of organisms that contain a resistance gene
product which inactivates the antibiotic. Even when strain SAK was cultured
in the presence of kanamycin, it was not possible to "switch on" the AAD(2")
gene; only AAC(3) gene transcripts were detected. The mechanism of
resistance to kanamycin in the derivative strains is being investigated. It may
be that strain SAK contains yet another aminoglycoside resistance gene, or
that the AAC(3) is responsible for the kanamycin resistance, in which case it
may have undergone a mutation which broadens its substrate profile to
include kanamycin.
Using a combination of DNA probing and cloning experiments it was

possible to identify genes encoding CAT, AAD(2") and AAC(3) in a clinical
isolate of A. baurnannii. The gene probes will be used to probe current
nosocomial isolates of Acinetobacter. It will be interesting to discover
whether these later isolates also contain an unexpressed AAD(2") gene. The
results described also serve to illustrate the molecular complexity of the
resistance mechanisms which may be encountered in these increasingly
common multi-resistant nosocomial isolates.
ACKNOWLEDGEMENTS
This work was supported by grants from the University of Cape Town
and the Medical Research Council to L.M. Steyn and A.A. Forder. The
computer programmes used in the analysis of the DNA sequences were
obtained from the Molecular Biology Computer Research Resource,
Harvard School of Public Health, Dana-Faber-Cancer Institute - D1154,
Boston, MA, 02115.
REFERENCES
Aiba, H., Adhya, S., and De Crombrugge, B., 1981, Evidence for two
functional gal promotors in intact Escherichia coli, J. Biol. Chern.,
256:11905.
146
Allmansberger, R., Brau, B., and Piepersberg, W., 1985, Genes for
gentamicin-(3)-N-acetyltransferase III & IV, Mol. Gen. Genet.,
198:514.
Benveniste, R., and Davies, J., 1971, R-factor mediated gentamicin
resistance: a new enzyme which modifies aminoglycoside
antibiotics, F EBS Lett., 14:293.
Bergogne-Berezin, E., Joly-Guillou, M.L., and Vieu, J., 1987, Epidemiology
of nosocomial infections due to Acinetobacter cal co acetic us,
Brau, B., Pilz, U., and Piepersberg, W., 1984, Genes for gentamicin-(3)-N-
acetyltransferases III & IV, Mol. Gen. Genet., 193:179.
Chopade B.A., Wise, P.J., and Towner, K.J., 1985, Plasmid transfer and
behaviour in Acinetobacter calcoaceticus EBF65/65, 1.
Gen. Microbiol., 131:2805.
Clewell, D.B., and Helinski, D.R., 1969, Supercoiled circular DNA-protein
complex in Escherichia coli: purification and induced
conversIOn to an open circular DNA form,
Proc. Natl. Acad. Sci. USA, 62:1159.
Davies, J., and Smith, D., 1978, Plasmid determined resistance to
antimicrobial agents, Ann. Rev. Microbiol., 32:469.
De Crombrugghe, B., Busby, S., and Buc, H., 1984, Cyclic AMP receptor
protein: Role in transcription activation, Science, 224:831.
Devaud, M., Kayser, F.H., and Bachi, B., 1982, Transposon mediated
Divers, M., Craven, P., and Vivian, A., 1985, Molecular analysis of an
antibiotic resistance plasmid, pA V5, and its derivative plasmids in
Acinetobacter calcoaceticus, 1. Gen. Microbiol., 131:3367.
Goldstein, F.W., Labigne-Roussel, A., Gerbaud, G., Carlier, C, Collatz E.,
and Courvalin P., 1983, Transferable plasmid-mediated antibiotic
resistance in Acinetobacter calcoaceticus, Plasmid, 10:138.
Gorman, CM., Moffat, L.F., and Harwood, B.H., 1982, Recombinant
genomes which express chloramphenicol acetyltransferases in
mammalian cells, Mol. Cell. Bioi., 2:1044.
Harwood, J., and Smith, D., 1971, Catabolite repression of chloramphenicol
acetyltransferase synthesis in E. coli K12, Biochem. Biophys.
Res. Comm., 42:57.
Henikoff, S., 1984, Unidirectional digestion with exonuclease III creates
targeted breakpoints for DNA sequencing, Gene, 28:351.
Johnson, D.A., Gautsch, J.W., Sportsman, J.R., and Elder, J.H., 1984,
Improved technique utilizing non-fat dry milk powder for analysis
of proteins and nucleic acids transferred to nitrocellulose, Gene
Analyt. Tech., 1:3.
McClure. W.R., 1985. Mechanism and control of transcription in prokaryotes.
Ann. Rev. Biochem .. 54: 171.
Murray, B.E., and Moellering, R.C, 1980, Evidence of plasmid-mediated
production of aminoglycoside modifying enzymes not previously
described in Acinetobacter calcoaceticus., Antimicrob.
Agents Chemother., 17:30.
147
Prince, A.S., and Jacoby, G.A., 1982, Cloning the gentamicin resistance gene
from a Pseudomonas aeruginosa plasmid in Escherichia coli
enhances the detection of aminoglycoside modification,
Rosenthal, N., Kress, M., Gruss, P., and Khoury, G., 1983, B.K. viral
enhancer element and a human cellular homolog, Science,
222:749.
Sanger, F., Nicklen, S., and Coulsen, A.R., 1977, DNA sequencing with chain
terminating inhibitors, Proc. N atl. Acad. Sci. USA, 74:5463.
Seth, A., 1984,Anewmethodforlinkerligation, GeneAnalyt. Tech., 1:99.
Shannon, K.P., and Phillips, I., 1983, Detection of aminoglycoside-modifying
strains of bacteria, in: "Antibiotics: Assessment of Antimicrobial
Activity and Resistance," p.183, A.D.Russell and L.B.Quesnel,
eds., Academic Press, London.
Shaw, W., 1983, Chloramphenicol acetyltransferase: enzymology and
molecular biology, CRC Critical Rev. Biochem., 14:1.
Southern, E.M., 1975, Detection of specific sequences among DNA
fragments separated by gel electrophoresis, i. Mol. Bioi., 98:503.
Tenover, F.e., Phillips, K.L.,Gilbert, T., Lockhart, P., O'Hara, P.J., and
Plorde, J., 1989, Development of a DNA probe for the
deoxyribonucleotide sequence of a 3-N -acetyl transferase
(AAC(3)I) resistance gene, Antimicrob. Agents Chemother.,
33:551.
Vivian, A., Hinchliffe E., and Fewson, e.A., 1981, Acinetobacter
calcoaceticus: some approaches to a problem, i.Hosp.Infect.,
2:199.
Vliegenthart, J.S., Ketelaar-van Gaalen, P.A.G. and Van De Klundert,
J.A.M., 1989, Nucleotide sequence of the aacC2 gene, a
gentamicin resistance determinant involved in a hospital epidemic
of multiply resistant members of the family Enterobacteriaceae,
Womble, D.D., and Rownd, R.H., 1988, Genetic and physical map of plasmid
NRI: comparison with other IncFII antibiotic resistance plasmids,
Microbiol. Rev., 52:433.
148
PLASMID ANDTRANSPOSON BEHAVIOUR IN ACINETOBACTER
K.J. Towner
Department of Microbiology & PHLS Laboratory

University Hospital
Nottingham NG7 2UH
UK
INTRODUCTION
Plasmids and transposons seem to be ubiquitous among bacteria and

it has been increasingly recognised that these genetic elements play an
important role in the overall biology of a wide range of prokaryotic
organisms. Central to this role is the location on plasmids and transposons
of genes encoding a variety of phenotypic properties, ranging from antibiotic
resistance to unusual metabolic activities. In many instances the provision
of these properties enables an organism to survive and prosper in an
otherwise hostile environment. Transfer mechanisms associated with
plasmids, combined with illegitimate recombination events specified by
transposable elements, may also allow spread of such advantageous
properties hetween cells of the same or different species.
In addition to these natural processes, the properties of plasmids and

transposons can also be exploited in the laboratory to enable a detailed
molecular analysis of gene structure and function. Transposons have been
widely used as insertion mutagens and can be employed in this manner to
identify replicons that have no readily scorable phenotype (e.g. cryptic
plasmids). Apart from their involvement in illegitimate recombination
events, they can also be used to provide mobile sites of homology upon which
homologous recombination can act. On a wider basis, conjugation is still
one of the most useful methods for the initial genetic analysis of an organism
and seems to be invariably plasmid-mediated. Plasmids also have an
essential role to play in the utilisation of in-vitro gene manipulation
techniques. Although many such procedures involve the initial cloning of
specific genes in an Escherichia coli K12 host, studies on gene regulation
and expression often require the use of plasmid cloning vectors which are
149
capable of stable replication in the original parent strain. The purpose of
this review is to describe the range of plasmids and transposons which can
exist in Acinetobacter, either naturally or in the laboratory, and the
behaviour and properties associated with them.
BEHA VIOUR OF ENTEROBACTERIAL R PLASMIDS

AND TRANSPOSONS IN ACINETOBACTER
R Plasmids Capable of Transfer to Acinetobacter
Although the presence of extrachromosomal DNA in naturally

occurring strains of Acinetobacter was reported by Christiansen et al.
(1973), most of the early studies on plasmid behaviour in Acinetobacter
were carried out using antibiotic resistance plasmids (R plasmids) which had
originally been isolated from members of the Enterobacteriaceae and
previously characterised in E. coli K12. Olsen and Shipley (1973) reported
that the broad host range P1 incompatibility group plasmid R1822 (later
shown to be indistinguishable from RP1, RP4 and RK2) could be transferred
to Acinetobacter strain ACJl. This observation was followed by the
studies of Towner and Vivian (1976a, 1977) demonstrating that other related
plasmids belonging to the PI incompatibility group could be transferred to
strain EBF65/65.
It was initially thought that members of the PI incompatibility group

were the only enterobacterial plasmids capable of transfer to
Acinetobacter, but work by Hinchliffe et al. (1980) demonstrated that
EBF65/65 contained a cryptic plasmid, pA V2, which was responsible for
variations in the fertility of plasmids and probably determined a restriction-
modification system in its host strain. A subsequent re-evaluation by
Chopade et al. (1985), using a pAV2-minus derivative of EBF65/65 as
recipient, showed that at least one plasmid from each of the incompatibility
groups B, C, F lv ' H/S, la, 10, PI, Wand X could be directly transferred by
conjugation from E. coli K12 to Acinetobacter. Representative
frequencies of transfer obtained using the filter mating method of Towner
and Vivian (1976b) are listed in Table 1, but it is important to note that not
all plasmids of a particular incompatibility group behave identically
(Chopade et aI., 1985). This may be due, at least in part, to the possible
presence of a second restriction system in EBF65/65 and the availability of
specific restriction sites on the plasmids being tested (Towner, 1980;
Chopade et aI., 1985).
In addition to the groups listed in Table 1, Barth et ai. (1981) and

Singer et al. (1986) have demonstrated that non-conjugative plasmids
belonging to the Q incompatibility group can be introduced into
Acinetobacter by either transformation or conjugal mobilisation. Singer
et al. (1986) also showed that strain BD413 was a considerably more efficient
recipient than strain H01-N in mobilisation experiments, thereby
150
Table 1. Transfer frequencies of representative enterobacterial plasmids
from E.coli K12 to Acinetobacter EBF65/65 (pAV2-) (data modified from
Chopade et al., 1985)
Plasmid Inc. Selectiona Frequency of

group transfer b
pUN65 B Tp 5 x 10- 7
R57b C Cm 1 x 10- 3
R124 F IV Tc 1 x 10- 5
R478 H2/S Tc 4 x 10- 4
R64 lex Tc 9 x 10- 6
R82la 10 Ap 3 x 10- 5
RP4 PI Tc 6 x 10- 4
R388 W Tp 3 x 10- 7
R6K X Ap 1 x 10- 4
aAp, ampicillin; Cm, chloramphenicol;

Tc, tetracycline; Tp, trimethoprim
btransfer frequencies are expressed in terms of
tranconjugants / recipient cell following overnight
filter matings at 28°C
emphasising, as with EBF65/65, the importance of choice of recipient strain

in plasmid transfer experiments.
Stability and Subsequent Behaviour of Enterobacterial Plasm ids
All of the enterobacterial plasmids transferred to Acinetobacter by

Chopade et al. (1985) were at least partially unstable in' the EBF65/65
system. In each case it was found that stable maintenance required
continual sub-culture on media containing appropriate selective antibiotics.
Subsequent transfer experiments revealed varied and complex patterns of
plasmid behaviour in EBF65/65. Some of the enterobacterial plasmids were
capable offurther rounds of transfer to other strains of Acinetobacter or of
re-transfer to E. coli K12. In other cases it was observed that further
transfer only occurred, if at all, following the introduction of an additional
mobilising plasmid. Transposition of resistance genes to the mobilising
plasmid was sometimes observed. Plasmids belonging to the same
incompatibility group were found to behave differently, thereby emphasising
the fact that tests with a single "renresentative" plasmid cannot always be
151
regarded as conclusive. A later section of this review will consider the wider
significance of these findings in relation to the possible role of
Acinetobacter as a reservoir of plasmid-mediated antibiotic resistance
genes in hospitals.
In addition to conferring antibiotic resistance, the acquisition of R

plasmids by strains of Acinetobacter may also result in important outer
membrane changes. Transfer of RP1 to the plasmid-free strain HOI-N has
been shown to be associated with changes in the fatty acid composition of the
lipopolysaccharide fraction and increased outer membrane permeability
compared to cells without the plasmid. These changes seem to be linked
with an increase in sensitivity to the antimicrobial activity of extracted
contents from rat polymorphonuclear leukocyte granules (Loeffelholz and
Modrzakowski, 1986; Loeffelholz et a!., 1987). Plasmid acquisition and
associated outer membrane changes may therefore play an important role in
determining the susceptibility of Acinetobacter to host immune cell
responses and phagocytosis.
Ability to Mobilise Chromosomal Genes
Plasmid-mediated mobilisation of Acinetobacter chromosomal

genes was first reported by Towner and Vivian (1976a) using strain
EBF65/65 and the PI incompatibility group plasmid RP4. Transfer of
chromosomal genes only occurred at a detectable frequency (10. 5 -10. 8 /
recipient cell) on solid surfaces, not in liquid matings. Gene mapping and
chromosome organisation in Acinetobacter has been reviewed by Vivian
(this volume), but in summary,two RP4-carrying donor strains which donated
the EBF65/65 chromosome in opposite directions from different origins
were isolated and, as a result, it was possible to demonstrate that the linkage
map of EBF65/65 is circular (Towner and Vivian, 1976b; Towner, 1(78).
Subsequently a number of other plasmids have also been shown to mobilise
chromosomal genes of Acinetobacter (Towner and Vivian, 1977;
Hinchliffe and Vivian, 1980b), but to date this property seems to be
restricted to members of the PI incompatibility group.
The precise mechanism by which chromosome mobilisation occurs in

Acinetobacter has not yet been fully elucidated. RP4 and its relatives are
known to mobilise the chromosomes of a variety of Gram-negative bacteria
at a low frequency (Holloway, 1979). In E. coli K12 it has been proposed
that this mobilisation may occur as a result of an integration event between
two Tnl copies, following the transposition of a Tnl copy from RP4 to the
host cell chromosome (Harayama et aI., 1980). Although Tn 1 can transpose
readily between plasmids, chromosomal insertions are normally only
obtained at a low frequency at a restricted number of sites (Sherratt, 1981),
thereby possibly accounting for the relatively low frequencies of
chromosome mobilisation observed. Other transposons insert into the
E. coli K 12 chromosome on a more random basis at a higher frequency
(Sherratt, 1981) and therefore could possibly mediate higher frequencies of
chromosome mobilisation in Acinetobacter. The behaviour and potential
152
uses of transposons introduced into Acinetobacter from E. coli are
considered in the next section.
Behaviour of Transposons Introduced into Acinetob acter
Transposons conferring antibiotic resistance have been shown to

have considerable utility as mutagenic agents and genetic tools in a variety of
microbial systems; however there have been few studies involving the direct
introduction of these elements into strains of Acinetobacter. Towner
(1980) demonstrated that PI incompatibility group plasmids carrying the
lysogenic bacteriophage Mu have great difficulty in becoming established in
strain EBF6S/6S. This observation formed the basis of a "suicide" vector
system which enabled Towner (1983) to use a RP4:: Mu: :Tn3171 vector
(pUN308) to generate Tn3171 insertions in the chromosome of EBF65/6S at
a frequency of 10-7 / recipient cell. Tn3171 is probably identical to the welI-
characterised trimethoprim/streptomycin resistance transposon Tn7
(Towner 1981, 1983) and seemed to have an insertional preference for one or
more arginine biosynthetic loci in EBF65/65. It is worth noting that Tn7
also has a single insertion site linked to an arginine biosynthetic gene in
Caulobacter ereseentus (Ely, 1982). Subsequent chromosome
mobilisation experiments in EBF6S/6S, using either RP4 or RP4::Tn3171 as
the mobilising vector, showed that the presence of Tn3171 in the
chromosome enhanced the frequency of mobilisation of chromosomal genes
adjacent to the original Tn3171 insertion, but that this process was not
influenced by the presence of Tn3171 on the mobilising plasmid.
A similar suicide plasmid, pJB4JI, containing transposon TnS and

phage Mu was used by Singer and Finnerty (1984) to introduce TnS into
Aeinetobaeter strains BD413 and H01-N. TnS was a particularly
attractive candidate for use as an insertion mutagen in Acinetobaeter since
it has been shown in E. coli K12 to have a low insertional specificity and an
ability to generate non-leaky polar mutations (Kleckner et aI., 1977; Shaw
and Berg, 1979). Although Tn5 insertions were obtained at frequencies of
10-6 -10- 7 / recipient for both Acinetobacter strains, screening of over 10,000
insertions failed to reveal any phenotypic mutations. Further analysis
suggested the existence of a single TnS-specific insertion site in H01-N and
two such sites in BD413. It seems, therefore, that Acinetobacter may
provide an exception to the normal pattern of general insertion of TnS into
numerous sites within prokaryotic genomes. However, in a subsequent
study, Doten et ai. (1987) were readily able to generate TnS insertions in five
or six pea structural genes derived from Acinetobaeter (encoding six
enzymes that convert protocatechuate to citric acid cycle intermediates via B-
ketoadipate) which had been cloned in E.eoli. One possibility is therefore
that, in the absence of the hotspots, there may not be any barriers to
generalised mutagenesis of Aeinetobacter DNA with TnS. The ease with
which certain Aeinetobaeter strains can undergo transformation (Juni,
1972; Ahlquist et aI., 1980) means that it should not be difficult to
reintroduce such cloned genes containing TnS insertions into the
Aeinetobaeter chromosome.
153
An alternative transposon delivery system, based on pRK2013, has
been described by Ely (1985). Since pRK2013 has a broad host range
transfer system linked to a Co1E1 replicon, it can be transferred to, but not
replicated in, a variety of non-enteric Gram-negative bacteria. Delivery
vectors carrying either Tn7 (pBEE7) or Tnl0 (pBEE10 or pBEEI04)
generated transposon insertions at a frequency of 10-6 _10- 8 in Acinetobacter
strain BD413. The specificity of these insertions was not investigated, but in
view of the results outlined above it would not be surprising if the insertional
specificity of these elements differed in Acinetobacter from that previously
observed with E. coli K12. Published data on the transposition frequencies
of transposons introduced into Acinetobacter in the laboratory are
summarised in Table 2.
INDIGENOUS PLASMIDS AND TRANSPOSONS
Plasmid Profiles in Acinetobacter
Following the first demonstration of extrachromosomal DNA in

Acinetobacter by Christiansen et aI. (1973), indigenous plasmids have been
found in the majority of Acinetobacter isolates that have been examined.
Although many of these are cryptic plasmids with as yet unidentified
functions, others (some of which are discussed in more detail below) have
been associated with such properties as antibiotic and heavy metal
resistance, aromatic hydrocarbon degradation, conjugal fertility and
restriction/modification systems. Many strains of Acinetobacter seem to
carry multiple plasmids of variable molecular size and it has also been
observed that unrelated isolates seem to have diverse plasmid profiles
(Hartstein et aI., 1988). Gerner-Smidt (1989) found indigenous plasmids in
75 of 93 strains of Acinetobacter examined, with a range of 0-20 plasmids /
strain. The majority of plasmids (70%) were less than 23 kb in size, with two
thirds of the strains containing two or more plasmids. As discussed by Noble
(this volume), there is a need for suitable typing methods which can be
applied to Acinetobacter. The determination of plasmid profiles may
prove to be a valuable addition to the range of typing methods which can be
used with clinical isolates of Acinetobacter.
Indigenous Antibiotic Resistance Plasm ids
Although clinical isolates of Acinetobacter have shown widespread

and increasing resistance to antibiotics, only slight evidence was obtained in
early studies of resistant hospital strains to suggest the involvement of
indigenous R plasmids (e.g. Murray and Moellering, 1979, 1980; French et
aI., 1980). Failure to detect plasmid-mediated transfer of antibiotic
resistance genes in such studies may, however, simply reflect the absence of a
suitable test system for plasmid transfer. For historical reasons, most
attempts at plasmid transfer from clinical isolates of any Gram-negative
species have initially used E. coli K12 as a recipient strain. The complex
patterns and varied frequencies of plasmid transfer observed between
154
Table 2. Insertion frequencies of transposons introduced into
Acinetobacter from E.coli K12 (data from Towner. 1983; Singer and
Finnerty. 1984; Ely, 1985)
Transposon Vector Reci.pient Frequency of

strain transpositiona
Tn5 pJB4JI BD413 10- 6 _ 10- 7

Tn5 pJB4JI lIOI-N 10- 6 _ 10- 7
Tn? pBEE? BD413
Tn10 pBEE10 BD413

TnlO(HHl04)b pBEE104 BD413
Tn3171 pUN308 EBF65/65
a expressed in terms of insertions per reclplent cell

b Tn10(HH104) is a mutant derivative of Tn10 which over-
produces the Tn10 transposase in E.coli K12
Acinetobacter and E.coli K12 have already been outlined in this review,
but it is pertinant to emphasise at this point that 17 of 31 self-transmissible
trimethoprim R plasmids transferred from E. coli K12 to Acinetobacter
EBF65/65 subsequently required the presence of an additional broad host
range transmissible plasmid for re-transfer to occur (Chopade et al., 1985).
Other studies have confirmed that although Acinetobacter can fairly
readily acquire plasmids from E. coli K12, the opposite transfer seems to be
a relatively rare event (Gomez-Lus et al., 1980; Goldstein et al., 1983). In
addition, the observation that 70% of indigenous Acinetobacter plasmids
are less than 23 kb in size (Gerner-Smidt, 1989) means that antibiotic
resistance genes carried on such indigenous plasmids would, in any event,
probably not be associated with the genetic information necessary for
transfer. It may therefore be necessary to rigorously assess the potential for
mobilisation of resistance genes from clinical isolates of Acinetobacter in
order to clarify the role of plasmids in the increasing spread of such
resistance.
In view of the above findings it is perhaps not surprising that in cases

where indigenous transmissible antibiotic resistance has been identified in
Acinetobacter, it has invariably been subsequently demonstrated to be
associated with a plasmid belonging to a broad host-range incompatibility
group. One of the most well-studied examples is that of the multi-resistant
155
strain JC17, originally isolated from a hospital in Pretoria, South Africa.
This strain carried a 128 kb conjugative plasmid which specified resistance to
sulphonamides, designated pA V1, and a 23 kb non·conjugative plasmid
which specified resistance to kanamycin/neomycin and tetracycline,
designated pA V5 (Hinchliffe et ai., 1980; Divers et ai., 1985). pA V1 is an
unusual example of a P1 incompatibility group plasmid with a restricted host
range (Hinchliffe and Vivian, 1980a); however, although it is unable to
transfer to E. coli K12, it is freely transmissible between several different
strains of Acinetobacter and also serves to mobilise both pAV5 and the
Acinetobacter chromosome (Hinchliffe and Vivian, 1980b; Divers et aI.,
1984).
The non·conjugative plasmid, pA V5, harboured by strain JC17 has

also been the subject of more detailed investigation. Genetic transfer
experiments revealed that the resistance determinants carried by pA V5
occasionally segregate on transfer, resulting in the apparent formation of
independent, compatible replicons: pA V51 (11.8 kb), which specifies
resistance to kanamycin/neomycin by means of an APH(3/)·I inactivating
enzyme, and pAV52 (17.6 kb) which specifies resistance to tetracycline·
(Divers et ai., 1984). Detailed molecular analysis by Divers et al. (1985)
showed that this segregation was associated with deletion of apparently non·
overlapping segments of pA V5, with each of the two deleted segments having
one terminus close to the kanamycin resistance gene (Fig.1). Electron
microscopy of homoduplex structures revealed the presence of 0.81 kb
inverted repeat sequences flanking this gene, while some limited genetic
experiments suggested that the resistance marker was transposable. It
therefore seems probable that the inverted repeats may be at least partially
responsible for generating the observed deletions. Although resistance to
either kanamycin/neomycin or tetracycline could be expressed in E. coli K12
following cloning of appropriate restriction fragments into pBR328, it
isinteresting to note, in the wider context of antibiotic resistance in
Acinetobacter, that pAV5 itself could not be maintained in
E. coli following either transformation or mobilisation experiments.
pAV5 (23 kb)
Km"
DELETED
PORTION ••• pAVS1 (11.8 kb)
,DELETED :r:c.~
- - - · · .... PORTION'··..··...._ - - pAV52 (17.6kb)
Fig. 1. Diagrammatic representation of the probable structures of pAV5

and its deletion derivatives (Divers et al., 1985)
156
A second well-characterised example of transferable plasmid-
mediated antibiotic resistance in a multi-resistant isolate of Acinetobacter
was provided by strain BM2500 isolated from a hospital in Paris, France.
Goldstein et al. (1983) showed that resistance to ampicillin,
chloramphenicol, kanamycin, streptomycin, sulphonamides and
trimethoprim was encoded in this strain by a single 167 kb plasmid,
designated pIP1031, belonging to incompatibility group C. Plasmids
belonging to this group are known to possess a wide host range and have been
shown to be transferable from E. coli to Acinetobacter (Chopade et aI.,
1985). pIPI031 was found to be transferable at an extremely low frequency
(5 x 10- 1°) from the original Acinetobacter isolate BM2500 to E. coli K12;
however this transfer could only be obtained with the initial isolate.
Detailed molecular studies indicated that loss of transferability was not
associated with any major rearrangements to the plasmid genome, nor to
integration of the resistance determinants into the host cell chromosome.
This situation was in sharp contrast to the situation in E. coli K12, where
pIPI031 was stable and could be easily transferred in further rounds of
conjugation. It therefore seemed that pIP1031 was not well-adapted to
carriage by strain BM2500, leading Goldstein et aI. (1983) to suggest that
pIPI031 had only been acquired extremely recently from an exogenous
source and that it was probably not a true indigenous Acinetobacter
plasmid. An additional interesting finding from the associated molecular
studies was that pIP1031 carried an unusual transposable module, designated
IS 15, which can exist in two, apparently not alternative, structural
configurations. IS15 can promote a range of genetic rearrangements and is
found on R plasm ids isolated from a large variety of Gram-negative bacteria,
perhaps thereby indicating its important biological role in plasmid evolution
(Labigne-Roussel and Courvalin, 1983).
A small number of other reports describing transferable antibiotic

resistance in Acinetobacter have also appeared in the literature. Gomez-
Lus et al. (1980) reported the transfer to E.coli K12, again at a low
frequency, of a PI incompatibility group plasmid from a strain of
Acinetobacter isolated in Spain, while Chirnside et al. (1985) described a
single clinical isolate of Acinetobacter, obtained from a London hospital in
1977, which transferred a 112 kb plasmid mediating trimethoprim and
sulphonamide resistance to E. coli K12. In both of these cases it seems
likely that the strain of Acinetobacter had only recently acquired its
plasmid from enteric bacteria occupying the same ecological niche. More
interestingly, Peng (1980) described transfer of several combinations of
resistance genes from 13 strains of Acinetobacter isolated in Taiwan.
Although these were also broad host range plasmids, they appeared to be
well-adapted to existence in Acinetobacter, since intrageneric matings
between strains of Acinetobacter occurred at frequencies 10-103 times
higher than intergeneric matings between Acinetobacter and E. coli KI2.
Of even greater interest was the description by Krcmery et al. (1985) of two
strains of Acinetobacter, isolated from different hospitals in
Czechoslovakia, the first of which transferred amikacin and B-Iactam
resistance to Pseudomonas aeruginosa, but not E.coli K12, while the
157
second strain transferred gentamicin and B-lactam resistance to E. coli KI2,
but not P. aeruginosa. Unfortunately no further details of any of these
plasmids are currently available.
Most recently, Lambert et a1. (1988) described a 63 kb plasmid,

designated pIP1841, which freely transferred resistance to kanamycin and
structurally related antibiotics, including amikacin, between different
taxonomic sub-groups of Acinetobacter, but not to E.coli KI2. Cloning
experiments, in which the resistance gene was transformed into E. coli K12
following in-vitro ligation with the vector plasmid pUC18, demonstrated that
full expression of the resistance gene could be obtained in E. coli K12. It
therefore seems probably that the apparent transfer barrier frequently
observed between Acinetobacter and other organisms does not result from
lack of expression of Acinetobacter genes in heterologous systems, but
rather from defects in conjugation donor or recipient ability, or plasmid
replication in the new host, or both (Lambert et aI., 1988).
C hromosom ally- Lo cated T ransposons
As described by Bergogne-Berezin and Joly-Guillou (this volume),

much of the antibiotic resistance found in clinical isolates of Acinetobacter
seems to result from the presence of enzymic resistance mechanisms that are
essentially identical to those known to he encoded by genes carried on
plasmids in the Enterohacteriaceae and other major genera of bacteria.
This review has already outlined the wide range of enterobacterial plasmids
that are capahle of transfer to Acinetobacter, even if not all are capahle of
suhsequent stable maintenance. It therefore seems likely that many of the
genes conferring antibiotic resistance in Acinetobacter were originally
introduced by plasmid-mediated conjugation from other bacteria sharing the
same environment. The fact that several such genes have been shown to be
transposable in E. coli K 12 means that, potentially, they cou ld become stahly
inserted by transposition into the Acinetobacter chromosome, following
their introduction on a plasmid which was itself unable to undergo stable
replication in the Acinetobacter host. If such genetic events have
occurred, they would provide an additional explanation as to why it is often
difficult in clinical isolates of Acinetobacter to demonstrate an association
between extrachromosomal DNA and resistance mechanisms which are
normally thought of as heing plasmid-encoded. It is therefore important
that appropriate genetic and molecular tests to examine this possibility are
included in studies of antibiotic-resistant clinical isolates.
To date there have been at least two puhlished reports of

chromosomally-located antibiotic resistance transposons in Acinetobacter.
In both cases antibiotic resistance phenotypes were apparently neither self-
transmissible nor associated with plasmid DNA, hut could be mobilised to
other strains by use of the conjugative broad host range PI incompatibility
group plasmid RP4, which underwent a size increase concomitant with
mobilisation. Shimizu et al. (1981) descrihed the mobilisation of a 5 kh
DNA sequence encoding streptomycin/spectinomycin resistance from a
158
strain of Acinetobacter isolated in Japan. In a similar study, Devaud et al.
(1982) mobilised a 24 kb DNA sequence encoding resistance to
chloramphenicol, gentamicin, streptomycin, sulphonamides and possibly
tetracycline from a strain of Acinetobacter found in an intensive care unit
in Switzerland. The occurrence of multi-resistant nosocomial isolates of
Acinetobacter seems to be becoming more common, perhaps in response to
the increasing usage of antibiotics in hospitals, and we have recently shown
that a similar transposable DNA sequence encoding resistance to B-Iactams,
aminoglycosides, tetracycline, chloramphenical and sulphonamides can be
found in endemic hospital strains of Acinetobacter isolated in the United
Kingdom (M. Nikolic and K.J. Towner, unpublished results). Experience
suggests that Acinetobacter is a difficult organism to eradicate from the
hospital environment, raising the worrying possibility that strains of
Acinetobacter could act as reservoirs of antibiotic resistance genes for
more important pathogens if such chromosomally-located transposons
become widely distributed in clinical isolates.
Plasmids Conferring Heavy Metal Resistance
In addition to the role of plasmids in conferring antibiotic resistance,

there have been several reports linking heavy metal resistance in
Acinetobacter to indigenous plasmids. This is particularly true of strains
of Acinetobacter isolated from polluted environments. A typical example
is the study by Olson et al. (1979) describing a strain of Acinetobacter,
designated W45, isolated from an estuarine environment in the USA. Strain
W45 produced mercuric reductase, which conferred resistance to mercury
salts, and. contained three plasmids of molecular size 3.3, 7.1 and 124 kb
respectively. Cured derivatives of W45 were found to have concomitantly
lost the ability to produce mercuric reductase and had consequently regained
their sensitivity to mercuric salts.
A more direct example of the involvement of plasmids in heavy metal

resistance was provided by studies of bacteria contained in polluted soil
surrounding the Khaidarkan mercury mine in the Kirghiz region of the
USSR. (Khesin and Karasyova, 1984; Lomovskaya et aI., 1986). One large
(60 kb) and two small (7.5 and 4.5 kb) plasmids were found in all five strains
of Acinetobacter examined. A mercury resistance determinant was found
to be located on the 7.5 kb plasmid, designated pKLl, which could be
mobilised by the large plasmid present in the same cell. Using conjugative
crosses and transformation experiments, it was established that, although
pKLl was a broad host range plasmid, it could not be transferred to E. coli.
Indeed, a characteristic feature of pKLl was its instability in all foreign host
cells after a few generations in the absence of selective pressure.
Nevertheless, when the mercury resistance determinant from pKLl was
cloned into either pUC19 or pBR322 and transformed into E.coli K12, full
expression of the mercury resistance determinant was obtained (Lomovskaya
et aI., 1986). Preliminary data also suggested that the large 60 kb
conjugative plasmid encoded a mercury resistance determinant but was
capable of replication in Acinetobacter only. As with antibiotic
159
resistance, it therefore seems that the failure to mobilise heavy metal
resistance to E. coli probably results from a transfer or replication barrier
rather than an expression barrier.
An additional interesting finding from the above studies was that a

plasmid apparently identical to pKL1 could be isolated from a strain of
Acinetobacter obtained from a second mercury deposit at Bolshoi Shayan
in the Transcarpathia region of the USSR (Lomovskaya et aI., 1986).
Transcarpathia is thousands of kilometres from Kirghiz, suggesting that
pKL1 is widely distributed amongst strains of Acinetobacter in the USSR.
A plasmid of similar size to pKL 1, although not directly linked to mercury
resistance, was also carried by strain W45 isolated in the USA (Olson et aI.,
1979).
Further evidence that small broad host range plasmids of the pKLl
type may be distributed in strains of Acinetobacter on a world-wide basis
has also been provided by the study of Rochelle et aI. (1988), which
demonstrated that a 7.8 kb broad host range plasmid, designated pOM17,
was responsible for encoding mercury resistance in epilithic strains of
Acinetobacter isolated from the organically polluted River Taff in Wales.
Other mercury-resistant Acinetobacter isolates from this river contained
plasmids varying in size from 2.1 to 220 kb (Rochelle et aI., 1988), but it
seems that small non-conjugative plasm ids of the pKLl/pOM17 type may
play an important role in the maintenance and subsequent spread of mercury
resistance genes in natural bacterial populations.
Metabolic Plasm ids
One of the most important characteristics of bacteria belonging to

the genus Acinetobacter is the ability of different strains to use a wide
range of aromatic hydrocarbon and aliphatic compounds as a source of
carhon and energy. As a consequence of their ability to degrade relatively
recalcitrant compounds of this type, it has been suggested that strains of
Acinetobacter may have an important role to play in the control of
environmental pollution. Although most reports of metabolic plasmids have
involved members of the Pseudomonas genus, there have also been several
studies describing strains of Acinetobacter in which genes encoding
significant degradative steps are plasmid-located.
An important group of environmental pollutants are the

polychlorinated biphenyls (PCBs), normally considered to be non-
degradable due to their unreactive structure and low water solubility.
Furukawa and Chakrabarty (1982) reported the involvement of a 80 kb
plasmid, designated pKFl, in the degradation of PCBs ny Acinetobacter
strain P6. This strain was anle to dissimilate 33 different pure isomers of
PCBs, with the formation of corresponding chlorobenzoic acids, but
spontaneously lost this anility at a high frequency (4 to 8%) after overnight
growth in nutrient broth. Although rigorous proof that pKFl encoded the
degradative pathway was lacking, loss of PCB-utilising activity was directly
160
associated with the generation of small 2.4 kb pKFl deletions. Attempts to
transfer or transform pKFl to other strains of Acinetobacter which could
not utilise PCBs were unsuccessful, but an apparently identical plasmid was
also observed in a strain of Arthrobacter present in the same original
culture.
Plasmid-mediated degradation of 4-chlorobiphenyl (4CB) to 4-

chlorobenzoate in Acinetobacter has also been described by Shields et
al.(198S). Strains of Acinetobacter and Alcaligenes isolated from a
mixed culture were both shown to carry a S3 kb plasmid, designated pSSSO,
which probably encoded a complete pathway for 4CB oxidation. Plasmid
curing resulted in loss of the 4CB-utilising phenotype, and loss of even early
4CB metabolism from the Acinetobacter strain, but the phenotype was
restored following filter-surface matings with the Alcaligenes strain. This
indicated a direct role for pSSSO in 4CB oxidation, but the possibility also
exists that pSSSO may activate other chromosomal genes partly involved in
mineralisation.
A third important degradative plasmid was described by Winstanley

et al. (1987), who investigated a benzene-utilising strain of Acinetobacter
isolated from a selective soil enrichment and found that it carried a large
(approximately 200 kb) transmissible plasmid, designated pWW174,
encoding the enzymes necessary for catabolism of benzene via the B-
ketoadipate pathway. It was possible to demonstrate conjugal transfer of
pWWI74 (and the benzene-utilising phenotype) between different strains of
Acinetohacter at frequencies varying from 10- 3 to 10- 7 , but no transfer
occurred to P.putida (the species in which most reports of catabolic
plasmids have been concentrated). An additional finding was that pWW174
was unstable in strains of Acinetohacter and was lost at high frequency in
the absence of selective pressure. This ease of loss suggested that either
pWW174 had a second alternative host in which it could replicate stably and
from which it had only recently been transferred to Acinetohacter, or that
it was a recently evolved recombinant plasmid in the soil from which it was
isolated and had not been subjected to maintenance selection in the absence
of benzene.
As a final example, there has also been a report of a strain of

Acinetobacter with the ability to degrade crude oil (Rusansky et aI., 1987).
Strain RAS7 contains four plasmids: pSR 1 (S.l kb), pSR2 (S.4 kb), pSR3
(10.S kb) and pSR4 (20 kb). A proportion of colonies which were isolated at
random after growth in the presence of acridine orange were found to have
lost the ability to grow on and disperse crude oil in liquid cultures and to
have concomitantly been cured of pSR4. Cells cured of pSR4 were still,
however, capable of growth on hydrocarbon vapours, thereby indicating that
pSR4 may encode genes that playa role in the physical interaction of the cell
with the hydrocarbon substrate rather than actual catabolism. Initial results
indicated that this role might be concerned with the process of
uptake/internalisation of the hydrocarbon substrate in liquid culture prior to
degradation. It would be of interest to examine the effect of pSR4 on the
161
cell surface properties of other Acinetobacter strains which do not
normally utilise hydrocarbons as the sole course of carbon and energy.
CLONING VECTORS FOR ACINETOBACTER
As outlined at the beginning of this review, most attempts to clone

specific genes derived from Acinetobacter (or any other organism)
invariably use E. coli K12 as an initial host strain in which to recognise
recombinant plasmids. Nevertheless, in order to study gene expression and
regulation, it is often necessary to subsequently reintroduce cloned genes
into their original host strain. As an illustrative example of this sort of
approach, Doten et aI. (1987) cloned a 11 kb EcoRI DNA fragment derived
from Acinetobacter into pUC18 and transformed the recombinant plasmid
into E. coli K12. The 11 kb insert carried the pea structural genes encoding
six enzymes responsible for converting protocatechuate to citric acid cycle
intermediates via B-ketoadipate. The recombinant plasmid was mutagenised
with Tn5 in E. coli K12 and then used to transform appropriate
Acinetobacter mutants in order to study the genetic organisation of the
pea cluster. Similar cloning experiments with a variety of genes derived
from Acinetobacter can be carried out using anyone of the standard
E. coli K12 host/vector systems that are now available. A major drawback,
however, is that in many cases the recombinant plasmids cannot be stably
maintained following their reintroduction into Acinetobacter. Thus, for
example, vector plasmids carrying an origin of replication (ori) gene derived
from pBR322 are not always stably maintained as plasmids in
Acinetobacter, with the consequence that stable reintroduction of cloned
genes can only be obtained following recombination with the chromosome
(Goosen et aI., 1987).
In an attempt to overcome this difficulty, the behaviour in

Acinetobacter strains BD413 and HOI-N of several "broad host range"
cloning vectors was examined by Singer et aI. (1986). Plasmids derived from
RSFlOlO (Bagdasarian et aI., 1979, 1981; Bagdasarian and Timmis, 1982),
such as pKT230 (11.9 kb), pKT231 (13 kb) and pKT248 (12.4 kb),were found
to be stably inherited after 20 generations of growth under non-selective
conditions in both strains. These plasmids could not be reliably transformed
into either BD413 or HOl-N, but could be successfully introduced by means
of a mobilisation system from E. coli K12 (Singer et aI., 1986). Plasmid
pTB107, a 10 kb derivative of R300B carrying a 1.5 kb kanamycin resistance
determinant (Barth et aI., 1981)was transformable into BD413 (not HOI-N),
but was slightly unstable (15% plasmid loss) unless stabilised by low level
antibiotic selection. Plasmid pRK290 is a 20 kb plasmid, derived from RK2,
that can be mobilised into a variety of Gram-negative organisms including
Acinetobacter (Ditta et aI., 1980). pRK290 was, however, found to be
highly unstable in both BD413 (67% plasmid loss) and HOI-N (>99%
plasmid loss). This rather surprising instability of pRK290 limits its
potential use as a cloning vector in Acinetobacter and may result from the
162
deletion of a segment of RK2 DNA that may be required for stable plasmid
maintenance (Ditta et aI., 1980; Lanka and Barth, 1981).
As a consequence of these results, Singer et aI. (1986) suggested that

a vector system based on selected R300B or RSF1010 derivatives would be
useful for routine cloning and complementation studies in Acinetobacter.
Examples of the adoption of this approach include the cloning into pKT230
of genes encoding the ability of AC"inetobacter to grow on benzene
(Winstanley et ai., 1987) and the construction of an Acinetobacter gene
bank by cloning Acinetobacter DNA partially digested with S au3A into a
hybrid vector plasmid, pLV21, derived from RSF1010 and pUB110 (Goosen
et aI., 1987). A number of similar cloning projects have heen successfully
completed and it seems that these vector systems probahly offer the hest
currently availahle approach to the use of in-vitro techniques for studying
genetic organisation and regulation in the AC"inetobacter genus. The
construction and use of shuttle vectors in Acinetobacter for particular
applications is descrihed in more detail hy Gutnick et aI. (this volume).
CONCLUSIONS
Plasmids and transposons undouhtedly play an important role in the

overall biology of Acinetobacter, hut there is still a relative paucity of
information ahout their detailed properties and hehaviour in this genus.
Much of the work described in this review has, either directly or indirectly,
focused on the possibilities for exchange of genetic information between
Acinetobacter and E. coli, largely as a consequence of the way in which the
science of bacterial genetics has developed. It seems clear that, while
certain types of plasmids are confined to the gene pool of E. coli, an
"intermediate group" can cross the generic boundary more freely, although
there may then be further barriers preventing establishment in the new host.
Transposons probably play an important role in ensuring that particular
genes (e.g. those conferring antibiotic resistance) are stably maintained even
in the event of plasmid instahility.
It also seems that there is a pool of plasmid-mediated genetic

information which is largely confined to Acinetobacter and is not readily
mobilisable to E. coli or other genera. Most "indigenous" plasmids so far
characterised in Acinetobacter are those which have heen capable of
transfer/mobilisation to E. coli and which therefore helong to the
"intermediate group". Many strains of Acinetobacter also contain several
cryptic plasmids which presumably encode information which is useful to the
cell. There seem to be few barriers to the expression of Acinetobacter
genes in E. coli, but there are undoubtedly barriers to plasmid transfer and
maintenance. In order to fully understand the role that these accessory
elements play in the lifestyle of the cell, it will therefore he necessary to
utilise in-vitro methods of gene cloning and analysis. It is hoped that the
next few years will see major advances in this direction.
163
REFERENCES
Ahlquist, E.F., Fewson, C.A., Ritchie, D.A., Podmore, J., and Rowell, V.,
1980, Competence for genetic transformation in Acinetobacter
calcoaceticusNCIB 8250, FEMS Microbio!' Lett., 7:107.
Bagdasarian, M., Bagdasarian, M.M., Coleman, S., and Timmis, K.N., 1979,
New vector plasmids for gene cloning in Pseudomonas, in:
"Plasmids of Medical, Environmental and Commercial Importance,"
p.411, K.N. Timmis and A. Puhler, eds., Elsevier/North Holland
Biomedical Press, Amsterdam.
Bagdasarian, M., Lurz, R., Ruckert, B., Franklin, F.C.H., Bagdasarian, M.M.,
Frey, J., and Timmis, K.N., 1981, Specific purpose plasmid cloning
vectors. Broad host range, high copy number, RSFI01O-derived
vectors, and a host-vector system for gene cloning in
Pseudomonas, Gene, 16:237.
Bagdasarian, M., and Timmis, K.N., 1982, Host:vector systems for gene
cloning in Pseudomonas, Curro Top. Micobiol. lmmunol.,
96:47.
Barth, P.T., Tobin, L., and Sharp, G.S., 1981, Development of broad host-
range plasmid vectors, in: "Molecular Biology, Pathogenicity and
Ecology of Bacterial Plasmids," p.439, S.B. Levy, R.C. Clowes and
E.L. Koenig, eds., Plenum Publishing Corporation, New York.
Chopade, B.A., Wise, P.J., and Towner, K.J., 1985, Plasmid transfer and
hehaviour in Acinetobacter calcoaceticus EBF65/6S, .T. Gen.
Christiansen, c., Christiansen, G., Leth Bak, A., and Stenderup, A., 1973,
Extrachromosomal deoxyrihonucleic acid in different
enterobacteria, 1. Bacteriol., 114:367.
Ditta, G., Stanfield, S., Corhin, D., and Helinski, D.R., 1980, Broad host
range DNA cloning system for gram-negative bacteria: construction
of a gene hank of Rhizobium meliloti, Proc. Natl. Acad. Sci.
USA,77:7347.
Divers, M., Craven, P.L., and Vivian, A., 1985, Molecular analysis of an
antihiotic resistance plasmid, pA VS, and its derivative plasm ids in
Divers, M., Hinchliffe, E., and Vivian, A., 1984, Characterisation of an
antihiotic resistance plasmid pA V5 and its constituent replicons in
Acinetobacter caleoaceticus, Genet. Res., 44:l.
Doten, R.c., Ngai, K.L., Mitchell, D.l., and Ornston, L.N., 1987, Cloning and
genetic organisation of the pea gene cluster from Acinetobaeter
calcoaceticus, 1. Bacteriol., 169:3168.
Ely, B., 1982, Transposition of Tn7 occurs at a single si te on the
Caulobaeter crescentus chromosome, 1. Bacterio!., 151: t056.
Ely, B., 1985, Vectors for transposon mutagenesis of non-enteric hacteria,
Mol. Gen. Genet., 200:302.
164
anitratus: epidemiology and control, I. H osp. I nf ect., 1: 125.
Furukawa, K., and Chakrabarty, A.M., 1982, Involvement of plasmids in total
degradation of chlorinated biphenyls, Appl. Environ. M icrob io!.,
44:619.
Gerner-Smidt, P., 1989, Frequency of plasmids in strains of Acinetobacter
calcoaceticus, I. Hosp. Infect., 14:23.
Goldstein, F.W., Labigne-Roussel, A., Gerbaud, G., Carlier, C., Collatz,
E.,and Courvalin, P., 1983, Transferable plasmid-mediated
antibiotic resistance in Acinetobacter, Plasmid, 10:138.
Gomez-Lus, R., Larrad, L., Rubio-Calvo, M.C., Navarro, M. and Assierra,
M.P., 1980, AAC(3) and AAC(6') enzymes produced by R plasmids
isolated in general hospital, in: "Antibiotic Resistance,"
p.295,S.Mitsushashi, L. Rosival and V. Krcmery, eds., Springer-
Verlag, Berlin.
Goosen, N., Vermaas, D.A., and van de Putte P., 1987, Cloning of the genes
involved in synthesis of coenzyme pyrroloquinolone-quinone from
Acinetobacter calcoaceticus, 1. Bacteriol., 169:303.
Harayama, S., Tsuda, M., and Tino, T., 1980, High frequency mobilisation of
the chromosome of Escherichia coli by a mutant of plasmid RP4
temperature-sensitive for maintenance, Mol. Gen. Genet.,
180:47.
Hartstein, A.I., Rashad, A.L., Liebler, J.M., Actis, L.A., Freeman, J.,
Rourke, J.W., Stibolt, T.B., Tomalsky, M.E., Ellis, G.R., and Crosa,
J.H., 1988, Multiple intensive care unit outhreak of Acinetobacter
calcoaceticus subspecies anitratus respiratory infection and
colonisation associated with contaminated, reusable ventilator
circuits and resuscitation bags, Amer . .I. M ed., 85:624.
Hinchliffe, E., and Vivian, A., 1980a, Naturally occurring plasmids in
Acinetobacter calcoaceticus: a P class R factor of restricted
host range,.I. Gen. Microbial., 116:75.
Hinchliffe, E., and Vivian, A., 1980h, Gene transfer in Acinetobacter
calcoaceticus: fertility variants of the sex factor pAVl, 1. Gen.
Hinchliffe, E., Nugent, M.E., and Vivan, A., 1980, Naturally occurring
plasmids in Acinetobacter calcoaceticus: pAV2, a plasmid
which influences the fertility of the sex factor pA VI, .I. Gen.
Holloway, B.W., 1979, Plasmids that mobilise bacterial chromosome,
Plasmid,2:1.
evidence for a ubiquitous genus, 1. Bacterio!., 112:917.
Khesin. R.B., and Karasyova, E.V., 1984, Mercury-resistant plasmids in
bacteria from a mercury and antimony deposit area, Mol. Gen.
Genet., 197:280.
Kleckner, N., Roth, J., and Botstein, D., 1977, Genetic engineering in vivo
using translocatable drug-resistance elements, .I. Mol. BioI.,
116:125.
165
Krcmery, Y., Langsadl, L., Antal, M., and Seckarova, A., 1985, Transferable
amikacin and cefamandole resistance: Pseudomonas
maltophilia and Acinetobacter strains as possible reservoirs of
R plasmids,.I. Hyg. Epidemiol. M icrobiol. Immunol., 28:141.
Labigne-Roussel, A., and Courvalin, P., 1983, ISI5, a new insertion sequence
widely spread in R plasmids of Gram-negative bacteria, Mol. Gen.
Genet., 189:102.
Chemother., 32:15.
Lanka, E., and Barth, P.T., 1981, Plasmid RP4 specifies a deoxyribonucleic
acid primase involved in its conjugal transfer and maintenance, .I.
Bacteriol., 148:769.
Loeffelholz, M.J., and Modrzakowski, M.C., 1986, Plasmid RP1-mediated
susceptibility of Acinetobacter calcoaeeticus to rat
polymorphonuclear leukocyte granule contents, I nf eet. I mmun.,
54:705.
Loeffelholz, M.J., Rana, F., Modrzakowski, M.C., and Blazyk, J., 1987, Effect
of plasmid RP1 on phase changes in inner and outer membranes and
lipopolysaccharide from Acinetobacter ealcoaceticus: a Fourier
transform infrared study, Biochemistry, 26:6644.
Lomovskaya, O.L., Mindlin, S.Z., Gorlenko, Zh.M., and Khesin, R.B., 1986,
A nonconjugative mobilizable broad host range plasmid of
Acinetobacter sp. that determines HgCl 2 resistance, Mol. Gen.
Genet., 202:286.
Murray, B.E., and Moellering, R.C., 1979, Aminoglycoside-modifying
enzymes among clinical isolates of Acinetobacter calcoaceticus
subsp. anitratus (H erellea vaginicola): explanation for high-
level aminoglycoside resistance,Antimicrob. Agents
Chemother., 15:190.
Murray, 8.E., and Moellering, R.C., 1980, Evidence of plasmid-mediated
production of aminoglycoside-modifying enzymes not previously
described in Acinetobaeter, Antimicrob. Agents Chemother.,
17:30.
Olsen, R.H., and Shipley, P., 1973, Host range and properties of the
Pseudomonas aeruginosa R factor R1822, J. Bacterio!.,
113:772.
Olson, T., Barkay, T., Nies, D., Bellama, J.M., and Colwell, R.R., 1979,
Plasmid-mediated mercury volatilization and methylation,
Develop. Indust. Microbiol., 20:275.
Peng, C-F., 1980, Studies on glucose nonfermentative Gram-negative bacilli.
II. Transferable drug resistance in Acinetobacter
calcoaceticus, .I. Formos. M ed. Assoc., 79:583.
Rochelle, P.A., Day, M.J., and Fry, J.C., 1988, Occurrence, transfer and
mobilization in epilithic strains of Acinetobacter of mercury-
resistance plasmids capable of transformation, J. Gen.
166
Rusansky, S., Avigad, R., Michaeli, S., and Gutnick, D.L., 1987, Involvement
of a plasmid in growth on and dispersion of crude oil by
Acinetobacter calcoaceticus RA57, Appl. Environ.
Microbiol.,53:1918.
Shaw, K.J., and Berg, C.M., 1979, Escherichia coli K12 auxotrophs
induced by the insertion of the transposable element Tn5,
Genetics, 92:741.
Sherratt, D., 1981, In vivo genetic manipulation in bacteria, in: "Genetics as
a Tool in Microbiology," p.35, S.W. Glover and D.A. Hopwood, eds.,
Cambridge University Press, Cambridge.
Shields, M.S., Hooper, S.W., and Sayler, G.S., 1985, Plasmid-mediated
mineralization of 4-chlorobiphenyl, 1. Bacteriol., 163:882.
Shimizu, S., Inoue, M., Mitsuhashi, S., Naganawa, H., and Kondo, S., 1981,
Enzymatic adenylylation of spectinomycin by Acinetobacter
calcoaceticus subsp. anitratus, 1. Antibiotics, 34:869.
Singer, J.T., and Finnerty, W.R., 1984, Insertional specificity of transposon
Tn5 in Acinetobacter sp., 1. Bacteriol., 157:607.
Singer, J.T., van Tuijl, J.J., and Finnerty, W.R., 1986, Transformation and
mobilization of cloning vectors in Acinetobacter spp., 1.
calcoaceticus,l. Gen. M icrob iol., 104: 175.
Towner, K.J., 1980, Behaviour of bacteriophage Mu in Acinetobacter
calcoaceticus EBF65/65, 1. Gen. Microbiol., 121:425.
mobilisation in Acinetobacter calcoaceticus EBF65/65, Genet.
Res., 41:97.
Towner, K.J., and Vivian, A., 1976a, RP4-mediated conjugation in
Towner, K.J., and Vivian, A., 1976b, RP4 fertility variants in Acinetobacter
calcoaceticus, Genet. Res., 28:301.
Towner, K.J., and Vivian, A., 1977, Plasmids capable of transfer and
chromosome mobilization in Acinetobacter calcoaceticus, 1.
Winstanley, c., Taylor, S.c., and Williams, P.A., 1987, pWW174: a large
plasmid from Acinetobacter calcoaceticus encoding benzene
catabolism by the B-ketoadipate pathway, Mol. M icrob iol., 1:219.
167
PLASMID STABILITY AND THE EXPRESSION AND REGULATION
OF A MARKER GENE IN ACINETOBACTER AND
OTHER GRAM-NEGATIVE HOSTS
C. Winstanley", J.A.W. Morgan b

J.R. Saunders" and R.W. Pickupb
"Department of Genetics & Microbiology

University of Liverpool
Liverpool L69 3BX
UK
blnstitute of Freshwater Ecology

Windermere Laboratories, Ambleside
Cumbria LA22 OLP
UK
INTRODUCTION
Much controversy has been raised over the possible intentional release
of genetically engineered microorganisms (GEMs) into the natural
environment. The potential for microbial detoxification of pollutants, crop
protection and improved plant productivity make the proposition an
attractive one. However, before any such release can be permitted, there is
a need to understand the mechanisms of survival, expression, transfer and
rearrangement of recombinant DNA in microbial communities.
The construction of safe GEMs has been a priority in order to lessen

the potential risk of adverse consequences due to accidental release from
laboratory or industry into the environment. Safeguards can include the use
of either debilitated host bacteria or cloning vectors incapable of replication
except under defined laboratory conditions. When the deliberate release of
a GEM is considered, an important factor must be that the released strain
should survive long enough to fulfill its intended function. Thus the same
safeguards cannot be applied to the consideration of accidental and
deliberate release. Although reports of actual releases are few in number
(see for example, Drahos et a!., 1986; Gaertner and Kim, 1988; Lindow and
169
Panopoulos,1988; Stewart-Tull, 1988), it is possible to carry out some initial
risk assessments using contained model environments in the laboratory.
Among the questions which need to be asked are: (i) whether the organism or
its recombinant DNA will survive; (ii) whether the organism can multiply
and displace members of the natural population; (iii) whether the organism
will become dispersed away from the intended release site. The potential
consequences of transfer of recombinant DNA into the natural population
must also be considered. Much work has concentrated on the possibility of
gene transfer in the environment, but an equally important consideration is
whether the recombinant DNA would be stable, expressed or regulated in
potential recipients. The possible hazards posed by changes in the activity
or control of genes introduced into natural systems need to be assessed.
To date, work on the study of gene survival and transfer in the

environment has generally relied on antibiotic resistance markers carried on
recombinant plasmids (Hughes and Datta, 1983; Devanas and Stotzky, 1986;
Devanas et al., 1986). Attempts to monitor the transfer of such plasmids,
particularly in aquatic systems, are likely to be hampered by the high
numbers of naturally occuring antibiotic-resistant bacteria. Any study in
freshwater systems is further complicated by the fact that only about 1% of
cells are cultivable by available methods (J ones, 1977).
Table 1. Marker Plasmids used in this Study
Plasmid Inc Group Antibiotic Marker System

Resistance a
pLV1010 Q Sm Ap PL-xylE
pLV1011 Q Sm PL-xylE-eI 8S7
pLV1013 Q Sm Km PR.-xylE-eI 8S7
pLV1016 P Ap Km Tc PR.-xylE-er 8S7
pLV1017 P Ap Km Tc PL-xylE
aSm: streptomycin; Ap: ampicillin; Km: kanamycin;

Tc: tetracycline.
170
In an attempt to overcome these problems we have developed a
versatile marker system designed to operate in a range of common
freshwater and soil microorganisms (Winstanley et aI., 1989). The system
involves detection of the xylE gene which encodes catechol 2,3 dioxygenase
(C230). Colonies expressing xylE can be detected on plates by spraying
them with 1 % catechol solution. A yellow coloration occurs due to the
formation of 2-hydroxymuconic semialdehyde. The marker gene is
expressed from the strong bacteriophage lambda promoters PL or PR which
can be controlled by the presence of the temperature-sensitive lambda
repressor elm. This enables the potentially deleterious metabolic burden
imposed on the cell by high expression of xyLE to be countered. In parallel
with the marker system we have also developed direct detection methods
which preclude the necessity to culture the target organism (Morgan et aI.,
1989). The marker DNA has been introduced into broad host range
plasmids to enable an assessment of xylE expression and regulation in a
range of common freshwater or soil bacteria. This has revealed significant
differences in stability, expression and regulation of the plasmid marker
systems in Acinetobacter compared with Pseudomonas, Klebsiella,
Aeromonas, Serratia and Escherichia coli strains. These differences,
and to a lesser extent differences observed in Pseudomonas, question
whether it is possible to predict the behaviour of recombinant DNA in the
event of gene transfer from one species to another in the natural
environment.
STABILITY OF MARKER PLASMIDS
The marker pJasmids which have been constructed are outlined

in Table 1. The IncQ regulated marker plasmids (pLVIOll and pLVlO13)
and unregulated plasmid (pLVlOlO) are based on the plasmid pKT230
(Bagdasarian et aI., 1981; Fig.I). The IncP regulated (pLVlOI6) and
unregulated (pL VI0l7) marker plasmids are cointegrates based on R68.45
(Haas and Holloway, 1976; unpublished data). Both pKT230 and R68.45
are able to replicate in a wide range of Gram-negative microorganisms.
R68.45 is naturally conjugative, whereas pKT230 may be mobilised by
plasmid pNJ5000 which is unstable in the absence of tetracycline (Grinter,
1983). The marker plasmids were introduced into different host bacteria
(Table 2) to assess the stability of the C230 + phenotype. Cells were
cultured on nutrient agar following growth under non-selective conditions
(nutrient broth) for approximately 20 generations. Resulting colonies were
then screened for C230+ by spraying with catechol.
The IncP cointegrate plasmids pLY 1016 and pL V 10 17 were stable

(>99%) in all hosts except K.pneumoniae (85% stable). The IncQ marker
plasmids showed marked differences in their stability depending on the
plasmid and the host (Table 3). The unregulated system of pLY 1010 was
markedly less stable than the regulated pLY 1013 system. However, there
was a surprising difference in stability in Pseudomonas between the two
regulated systems of pL V 1013 and pL VI 011. xy IE expression from PI rather
171
Fig. 1. IncQ marker plasmids. Abbreviations indicating cleavage sites for
restriction enzymes are: Ba, BamHI, E, EcoRI, H, HindIII, P, PstI, Sm,
SmaI, Xb, XbaI, Xh, XhoI. Ap, Km and Sm indicate resistance to
ampicillin, kanamycin and streptomycin respectively.
than P R apparently caused high instability despite the similar regulation of

the two promoters by cI in bacteriophage lambda (Johnson et ai., 1981).
The largest difference of all was observed in A.calcoaceticus ADPI where
all three plasm ids exhibited high degrees of instability. In particular the
C230+ phenotype of pLY 1010 was retained, at best, by only 4'1c of the
population. Similar high levels of instability were observed in two recent
freshwater isolates, P,putida FBAll and P.fluorescens FHl. Further
observations revealed that the C230 + phenotype was unstable in strains
ADP 1, FBAll and FHl, even when antibiotics selecting for retention of the
plasmid were present in the growth medium. In the presence of
streptomycin (100 mg/l) only 4-5'1c of FBAII colonies and 1-3% of FHl
colonies screened retained the C230 + phenotype after overnight culture of
pLVIOlO-containing strains. The addition of ampicillin to the growth
medium made no difference. Screening of the DNA from four C23o-
SmRApRcolonies from each of FBAll and FHl revealed the presence of
172
Table 2. Host Bacteria used in this Study
Straina Plasmids Assessed
Acinetobacter
calcoaceticus
ADPl All plasmids
Escherichia coli
ED8654 pLV1010, pLV10ll, pLV1013.
AB1l57 pLV1016, pLV1017.
AB2463 pLV1016, pLV1017.
Pseudomonas putida
PRS2000 All plasmids
PaW140 pLV1010, pLV10ll, pLV1013.
PaW8 pLV1016, pLV1017.
FBAll All plasmids.
Pseudomonas aeruginosa
PAOl All plasmids
NCIB8295 pLV1010, pLV10ll, pLV1013.
Pseudomonas fluorescens
FHl All plasmids
Aeromonas hydrophila
NCTC8049 All plasmids.
Klebsiella pneumoniae
NC18418 All plasmids.
K2819 All plasmids.
Serratia rubidiae All plasmids.
amos t strains are described by Winstanley et al. (1989)
deleted pLV1010 derivatives. In each case the deletion included the loss of
one BamHI, one XbaI, one PstI and one EeaRI restriction site. Since one
of the two EeaRI sites on pLVIOIO is located in the still functional Sm R
gene, it would appear that a deletion from an area including the EeaRI site
upstream of xylE to the X b aI site downstream of xylE has occurred.
173
Table 3. Stability of IncQ Plasmid Constructs a
Strain Plasmid
pLV1013 pLV1011 pLV1010
A.calcoaceticus
ADP1 11% 30% 4%
E.coli
ED8654 100% 99% 95%
P.putida
PRS2000 98% 22% 90%
PaW8 100% 26% 98%
P.aeruginosa
PAOl 90% 18% 51%
Aeromonas
NCTC8049 100% 98% 94%
Klebsiella
NCIB418 100% 99% 55%
S.rubidiae 100% 100% 67%
astabili ty is expressed as % retention of C230+

after overnight growth in nutrient broth at 30°C.
The mean of at least two experiments is shown.
Closer analysis of the stability of all three plasmids in

A. calcoaceticus ADP1 has been carried out. Table 4 shows the result of
an experiment to assess the distribution of the three phenotypes
(C230 + SmRAp\ C230-Sm RA pR, C23O-Sm SA p S). Colonies growing on
nutrient agar were screened by patching them on to nutrient agar
supplemented with streptomycin (100 mg/l) as well as nutrient agar alone.
C230+ colonies were then identified by spraying with catechol. In some
cases, particularly when growth was carried out with streptomycin present, a
significant proportion of colonies assumed to be C23o- when grown on
nutrient agar appeared to be C230 + when grown on agar supplemented with
streptomycin. Colonies isolated from the broth cultures containing the
antibiotic selection also appeared to be C230- only on nutrient agar
plates. This indicates that there is a certain amount of plasmid or xylE gene
174
Table 4. Stability of C230+ and S~ Phenotypes in ADPl
Plasmid Phenotype a % Retention of Phenotype b
Growth Medium c
N broth N broth+Sm
pLV1010 C230+SmR 0.3% 6.2%

C230-SmR 7.3% 70.4%
C230-Sm S 92.4% 23.4%
pLV10ll C230+SmR 9.3% 47.4%

C230-SmR 0% 0.6%
C230-Sm S 90.7% 52.0%
pLV1013 C230+SmR 1.0% 74.0%

C230-SmR 0% 0%
C230-Sm S 99.0% 26.0%
aSmR and SmS indicate streptomycin resistance and

sensitivity respectively.
bresults were based on an experiment in which a single
colony, taken from a plate containing streptomycin, was
inoculated into the growth medium. After overnight culture
at 30 C, dilutions were plated on to nutrient agar and
colonies were screened.
cN-broth indicates nutrient broth.
instability during growth on non-selective plates. C230- colonies were

routinely re-screened after patching on to streptomycin-containing media to
take account of this problem.
pLVIOIO gave rise to high numbers of C230-Sm R cells, particularly

under conditions selecting for antibiotic resistance. Fig.2 shows a
restriction digest of DNA isolated from eight such colonies. The gel shows
loss of a PstI site and deletion of approximately 2.5-3.0 kb in each case. This
was also associated with loss of BamHI, XbaI and EeaRI sites. The
deletions appear to be very similar to those observed in strains FHl and
FBAl1. Three derivatives of A. ealeaaeetieus ADPl(pL VlO11) with C23o-
SmR phenotypes were also isolated. Analysis of their DNA revealed loss of
one PstI, one EeaRI and one BamHI site from pLVIOl1. A deletion of
approximately 4 kb including xylE had occurred (data not shown). No
175
1 2 3 4 5 6 7 8 9 10 11 12
Fig. 2. Restriction digests of SmRApRC230- derivatives of pLV1010. Lane

1, HindIII digest of lambda DNA; lane 2, PstI digest of pLV1010; lanes 3-
10, PstI digests of SmRApRC230- derivatives of pLVI010.
separation of the antibiotic resistance and C230+ phenotypes was observed

with strainADP1(pLV1013).
EXPRESSION AND REGULATION OF XY LE
A simple assay for C230 activity (Sala-Trepat and Evans, 1971) has
allowed xylE expression and regulation to be assessed in a range of host
strains containing the different marker plasmids. C230 specific activity is
expressed in Units/mg protein.
Inc P Plasm ids
Unregulated expression of xylE from pLVlO17 gave rise to high levels

of C230 activity in all hosts tested (Table 2). The levels varied from 9.1
U/mg in A. calcoaceticus ADP1 to 25.4 U/mg in E. coli AB1157
(unpublished data). The regulation of xylE expression from pL VlO16 was
assessed by incubation at 37"C after initial growth in nutrient broth at 28°C.
Expression of xylE was increased by incubation at the higher temperature in
all the hosts tried. The Enterobacteriaceae showed a 1000-fold increase
over 12 h, whereas lower induction was observed in Pseudomonas and
Acinetobacter hosts. Uninduced C230 activity in these hosts showed a ten-
fold increase over that observed in E. coli. The specific activity obtained
after 12 h at 37"C varied from 0.5-2 U/mg in Pseudomonas to 4-5 U/mg in
E. coli and Klebsiella. The level of C230 activity in Acinetobacter after
the same period of incubation was 2.5 U/mg (unpublished data).
Inc Q Plasm ids
Detailed analysis of the expression and regulation of xylE from the

IncQ marker plasmids has been carried out (Winstanley et aI., 1989).
176
Unregulated C230 activity from pLV1010 was high in all hosts tested, with
levels of 62 U/mg and 52 U/mg recorded in P.putida PaW140 and PRS2000
respectively. Levels in P. aeruginosa hosts varied from 22 U/mg (PAOl) to
32 U/mg (NCIB8295), whilst good activity was also recorded in E. coli,
Klebsiella and Aeromonas (14-24 U/mg). Levels of 9.5 U/mg were
recorded in S. rubidiae, but the lowest activities were recorded in P.putida
FBAll, P.fluorescens FHl and A. calcoaceticus ADPI (5.5-5.8 U/mg).
It is no coincidence that these are the same strains in which pLVIOI0 is
highly unstable, even in the presence of antibiotics. Where plasmids were
known to be unstable in the host being investigated, antibiotics were
included in the growth medium from which samples were taken for assay.
However, the deletion of xylE from pLV1010 in strains FBAll, FHl and
ADPI inevitably leads to misleading levels of C230 activity being recorded in
these strains.
Analysis of the regulated systems of pLV1011 and pLVI013 revealed

complications (Winstanley et aI., 1989). There was good induction of the PR-
xylE-cl m system of pLVI013 at 3T Cover 25 h in E. coli (680-fold),
Klebsiella (200 to 300-fold), Serratia (260-fold) and Aeromonas (210-
fold). The regulation in Pseudomonas was also good (18 to 67-fold), but
was hampered by a higher uninduced activity (86-290 mU/mg compared with
7.1 mU/mg for E. coli). Maximum activities observed varied from 16 U/mg
in P.aeruginosa NCIB8295 to 2.3 U/mg for S.rubidiae. The worst
induction was found in A. calcoaceticus where the uninduced activity of
560 mU/mg increased to a peak level of 3.4 U/mg over a 25 h period.
A number of differences were observed with the PL -xyIE-cls57 system of

pLV1011, despite the fact that the change in promoter from P R to PLis the
only significant difference from the pL VI013 system. Much lower increases
in C230 activity after incubation at 3T C were observed in all the hosts, but
the most striking difference was found in A. calcoaceticus ADPI where
there appeared to be no regulation of xylE expression. In some hosts
(E. coli, P. aeruginosa and Klebsiella) an improvement in induction of
C230 activity was achieved by raising the incubation temperature to 42° C.
This effect was not observed with the pLY 1013 system. The higher
temperature appeared to have harmful effects on some host strains
(P.putida, P.fluorescens and Aeromonas), but studies of the
ADPl(pLVI013) system indicated that similar levels of induction occurred
at both 3T C and 42° C, although the maximum was achieved earlier at the
higher temperature (data not shown). This indicates that the pL V1011
system is genuinely unregulated in A. calcoaceticus ADP1, where the
uninduced levels of C230 activity (2 U/mg) were the highest recorded.
The generally higher levels of C230 activity which can be achieved with
IncO plasmid systems can be accounted for by a higher copy number
compared with the IncP plasmids. Higher uninduced C230 activity from
pLV1013 in some species may be due to the presence of Pseudomonas
promoters upstream of xylE.
177
DISCUSSION
Considerable differences in the stability, expression and regulation of

the xylE marker systems were observed. Both the unregulated and
regulated IncP plasm ids pL VlO17 and pLVlO16 were apparently stable in the
whole range of hosts tested. Expression and regulation of xylE on these
plasmids was also fairly consistent from species to species. Both IncP
plasmid systems functioned adequately in Acinetobacter. The major
differences were observed with the higher copy number IncQ marker
plasmids. The unregulated plasmid pLVlOlO was highly unstable in ADP1,
FBAll and FHl. A. calcoaceticus is commonly found in both freshwater
and soil, whilst P.putida FBAll and P.fluorescens FH1 were both recent
freshwater isolates (Winstanley et aI., 1989). Closer analysis revealed that
xylE was deleted from pLV1010 in these strains, even in the presence of
antibiotic selection. Similar deleted derivatives of pLVlOlO were obtained
during growth of PRS2000(pL V1010) in sterile lake water model systems
(Morgan et aI., 1989), but the effect has not been observed when this strain is
grown overnight in nutrient broth. Plasmid pL VlOll, containing PL -xylE-
el 857 , was also unstable in a number of hosts and the regulation of this system
was inferior to the pL V1013 system. This was particularly striking in
A. calcoaceticus ADP1 where no regulation was achieved. pLVlOll was
also highly unstable in ADP 1, with some evidence of loss of xylE from the
plasmid being apparent. More surprising was the high instability of
pL V1013, although no evidence for deletion of xylE from the plasmid was
found.
There have been several reports of unstable naturally-occurring

plasmids in Acinetobacter (Divers et aI., 1984; Winstanley et aI., 1987).
However, pKT230, on which the IncQ plasmid systems are based, is known to
replicate and be stably maintained in Acinetobacter (Singer et aI., 1986);
thus the reasons for the instability of the marker plasmids are unclear.
Metabolic burden due to the high expression of xylE could contribute to
instability of pLV1010. However, C230 activity from pLV1011, despite
being unregulated, would not appear to be high enough to account for its
instability, particularly as uninduced levels of C230 activity from pL V1013
are higher in A. calcoaceticus ADP1 than other strains, but no greater than
1 U/mg. Indeed xylE expression from pLVlO17, which is perfectly stable in
strain ADP1, gives rise to levels of 9.1 U/mg. Since both PL and PR are
present on pL V1011 there is a possibility that titration of the cIs57 repressor
could cause reduced regulation because of interaction with both promoters.
However, the introduction of PL in trans on plasmid pPLc245 had no similar
effect on regulation of the pLV1013 system (results not shown). It is
possible that cI repressor is not fully effective in Acinetobacter since high
levels of uninduced C230 activity were apparent with both the pL V1011 and
pL V1013 systems. This would not account for the fact that some control by
cI was apparent with pL VI013, but not pLVIOll. There is also the evidence
provided by the pLV1016 system, which is as well regulated in
Acinetobacter as it is in Pseudomonas.
178
There are a number of examples demonstrating the potential of soil
and aquatic bacteria to receive recombinant DNA in genetic exchanges
(Stotzky and Babich, 1986; Bale et al., 1987; Krasovsky and Stotzky, 1987;
Saye et al., 1987; Genthner et al., 1988; Zeph et al., 1988). There have also
been cases of non-conjugative plasmids being mobilised by indigenous waste-
water bacteria (McPherson and Gealt, 1986; Mancini et al., 1987). In the
event of gene transfer, effects such as differences in gene regulation and
expression or changes in gene stability, depending on the host, suggest that
the behaviour of recombinant DNA following release into the natural
environment would be difficult to predict. These kind of effects would also
make detection of the released DNA more difficult. It may therefore be
considered important to include an assessment of gene expression,
regulation and stability in potential natural recipients in the consideration of
the risk posed by the release of a GEM into the environment.
REFERENCES
Bagdasarian, M., Lurz, R., Ruckert, B., Franklin, F.C.H., Bagdasarian, M.M.,
Frey, J., and Timmis, K.N., 1981, Broad host range, high copy
number, RSF1010-derived vectors and a host-vector system for
gene cloning in Pseudomonas, Gene, 16:237.
Bale,M.J., Fry, J.c., and Day, M.J., 1987, Plasmid transfer between strains of
Pseudomonas aeruginosa on membrane filters attached to river
stones,l. Gen. Microbial., 133:3099.
Devanas, M.A., Rafaeli-Eshkol, D., and Stotzky, G., 1986, Survival of
plasmid-containing strains of Escherichia coli in soil: effect of
plasmid size and nutrients on survival of hosts and maintenance of
plasmids, Curro Microbial., 13:269.
Devanas, M.A., and Stotzky, G., 1986, Fate in soil of a recombinant plasmid
carrying a Drosophila gene, Curro Microbial., 13:279.
Divers, M., Hinchliffe, E., and Vivian, A., 1984, Characterisation of an
antibiotic resistance plasmid pA V5 and its constituent replicons in
Acinetobacter calcoaceticus, Genet. Res., 44:l.
Drahos, D.J., Hemming, B.C., and McPherson, S., 1986, Tracking
recombinant organisms in the environment: il-galactosidase as a
selectable non-antibiotic marker for fluorescent pseudomonads,
Biotechnology, 4:439.
Gaertner, F., and Kim, L., 1988, Current applied recombinant DNA projects,
Trends Biotech., 6:S4.
Genthner, F.J., Chatterjee, P., Barkay, T., and Bourquin, A.W., 1988,
Capacity of aquatic bacteria to act as recipients of plasmid DNA,
Appl. Environ. Microbial., 54:115.
Grinter, N.J., 1983, A broad host-range cloning vector transposable to
various replicons, Gene, 21:133.
Haas, D., and Holloway, B.W., 1976, R factor variants with enhanced sex
factor activity in Pseudomonas aeruginosa, Mol. Gen.
Genet., 144:243.
179
Hughes, V.M., and Datta, N., 1983, Conjugative plasmids in bacteria of the
pre-antibiotic era, Nature, 302:725.
Johnson, A.D., Poteete, A.R., Lauer, G., Sauer, R.T., Ackers, G.K., and
Ptashne, M., 1981, Lambda repressor and cro-components of an
efficient molecular switch, Nature, 294:217.
Jones, J.G., 1977, The effect of environmental factors on estimated viable
and total populations of planktonic bacteria in lakes and
experimental enclosures, Freshwater Bioi., 7:67.
Krasovsky, V.N., and Stotzky, G., 1987, Conjugation and genetic
recombination in Escherichia coli in sterile and non-sterile soil,
Soil Bioi. Biochem., 19:63l.
Lindow, S.E., and Panopoulos, N.J., 1988, Field tests of recombinant Ice-
Pseudomonas syringae for biological frost control in potato, in:
"The Release of Genetically Engineered Microorganisms," p.I2I,
M.Sussman, C.H. Collins, F.A. Skinner and D.E. Stewart-Tull,
eds., Academic Press, London.
McPherson, P., and Gealt, M.A., 1986, Isolation of indigenous wastewater
bacterial strains capable of mobilizing plasmid pBR325, Appl.
Environ. Microbiol., 51:904.
Mancini, P., Fertels, S., Nave, D., and Gealt, M.A., 1987, Mobilization of
plasmid pHSV106 from Escherichia coli HB101 in a laboratory-
scale waste treatment facility, Appl. Environ. Microbiol.,
53:665.
Morgan, J.A.W., Winstanley, e., Pickup, R.W., Jones, J.G., and Saunders,
J.R., 1989, Direct phenotypic and genotypic detection of a
recombinant pseudomonad population released into lake water,
, Sala-Trepat, J.M., and Evans, W.e., 1971, The meta cleavage of catechol by
Azotobacter species: 4-oxalocrotonate pathway, Eur. 1.
Microbiol.,20:400.
Saye, D.J., Ogunseitan, 0., Sayler, G.S., and Miller, R.V., 1987, Potential for
the transduction of plasmids in a natural freshwater environment:
effect of plasmid donor concentration and a natural microbial
community on transduction in Pseudomonas aeruginosa, Appl.
Singer, J.T., Van Tuijl, J.J., and Finnerty, W.R., 1986, Transformation and
mobilisation of cloning vectors in Acinetobacter spp., 1.
Bacteriol.,165:30l.
Stewart-Tull, D.E., 1988, Round table 5: case histories of deliberate release,
in: "The Release of Genetically Engineered Microorganisms,"
p.245, M.Sussman, C.H. Collins, F.A. Skinner and D.E. Stewart-
Tull, eds., Academic press, London.
Stotzky, G., and Babich, H., 1986, Survival of, and genetic transfer by,
genetically engineered bacteria in natural environments, Adv.
Appl. Microbiol., 31:93.
catabolism by the .B-ketoadipate pathway, Mol. Microbiol.,
1:219.
180
Winstanley, c., Morgan, J.A.W., Pickup, R.W., Jones, J.G., and Saunders,
l.R., 1989, Differential regulation of lambda PL and PR promoters
by a cI repressor in a broad-host-range thermoregulated plasmid
marker system. Appl. Environ. Microbial., 55:771.
Zeph, L.R., Onaga, M.A., and Stotzky, G., 1988, Transduction of
Escherichia coli by bacteriophage PI in soil, Appl. Environ.
181
RANDOM INSERTIONAL MUTAGENESIS IN AC1NETOBACTER
B. Vosman, R. Kok and K.J. Hellingwerf
Laboratorium voor Microbiologie

Universiteit van Amsterdam
Nieuwe Achtergracht 127
1018 WS Amsterdam
The Netherlands
INTRODUCTION
Mutant analysis is a powerful tool in the study of biological processes,

the more so if it is combined with the relevant physiological experiments.
Two basically different approaches can be used for the construction of
mutants: (i) mutants can be obtained with chemical mutagens or by
irradiation with UV or X-rays; (ii) mutants can be generated by insertional
mutagenesis. The basic idea behind the latter method is the insertion of an
easily selectable genetic element into the gene of interest. This has the
major advantage that the gene is marked in such a way that it can be isolated
easily.
MODES OF INSERTIONAL MUTAGENESIS
The most frequently used method of insertional mutagenesis involves

the use of transposons (Kleckner et al., 1977; Simon et al., 1983). For
Escherichia coli, bacteriophage lambda is the most commonly used
delivery vehicle, whereas for most other Gram-negative bacteria a
transposon is introduced by conjugation on a plasmid from E. coli. These
plasmid vectors generally contain a ColEl-like origin of replication, which is
unable to initiate replication in strains outside the enteric bacterial group
(Simon et al., 1983).
In Bacillus subtilis, insertional mutagenesis can be accomplished by

use of the transposon Tn917 (Youngman et al., 1983). Since B.subtilis is an
organism that is competent for natural transformation, mutagenesis can also
183
be accomplished by the insertion of a non-replicative plasmid into the gene
of interest (Niaudet et aI., 1982). For the latter method, pBR322 derivatives
are normally used since they are not replicated in B.subtilis. Randomly
generated chromosomal DNA fragments are cloned into these vectors and
the recombinant molecules transformed into competent cells of B.subtilis.
Plasmids containing inserts homologous to regions of the chromosome
integrate efficiently into the chromosome of a wild-type recipient cell
(Niaudet et aI., 1982).
The insertion can occur via two different mechanisms, depending on

the nature of the DNA insert. The first possibility is that the vector
integrates by a Campbell-like mechanism (Campbell, 1962). This type of
integration occurs when one DNA fragment is cloned into the carrier
plasmid. After integration into the chromosome, the plasmid DNA is
flanked by a direct repeat of the cloned fragment. This type of integration
only creates a mutant phenotype if the DNA fragment carried by the vector
does not contain either of the borders of a particular transcriptional unit.
The second possibility is an integration via a double cross-over
recombination event, also referred to as replacement recombination. The
latter mechanism predominates when two non-contiguous DNA fragments
are cloned into a plasmid. It results in deletion of the chromosomal DNA
fragment located between the two cloned fragments.
REPLICATION OF pBR322 DERIVATIVES IN A.CALCOACETICUS
Singer et ai. (1986) showed that plasmid pRK2013, which contains a

ColEl-like origin of replication, can be transferred by conjugation from
E. coli to A. calcoaceticus. Transfer frequencies with this plasmid (1.1 x
10-3 ) were, however, approximately 500-fold lower than transfer frequencies
using RSF1010 (8.2 x 10- 1), or derivatives thereof. After growing cells
containing pRK2013 under non-selective conditions, the plasmid was rapidly
lost. This indicates that plasmids containing a ColE I-like origin of
replication do replicate, although poorly, in A. calcoaceticus.
We have transferred the transposon-containing plasmids pJFF350

(Fellay et aI., 1989), pSUP2021 (Simon et aI., 1983) and pSUP::Tn5-B20
(Simon et aI., 1989), all of which contain a ColEI-like origin of replication, by
conjugation from E. coli S17-1 to A. calcoaceticus AAC1. The latter
strain is a rifampicin-resistant derivative of BD413. Rifampicin and
kanamycin resistant transconjugants were selected. For plasmid
pSUP::Tn5-B20 the transfer frequency was 1.2 x 10-3 , approximately the same
frequency as found for pRK2013 by Singer et ai. (1986).
Tn5-B20 carries a promoterless lacZ gene adjacent to IS50L • With

this construct, transposition should give rise to equal proportions of
transconjugants that do or do not produce B-galactosidase. This prediction
is based on the assumption of random insertion of Tn5-B20 into the
chromosome. However, all transconjugants obtained after transfer of
184
pSUP::Tn5-B20 to strain AACI stained equally blue on Luria-Bertani (LB)
plates containing 5-bromo-4-chloro-3-indolyl-B-D-galactoside (X-gal) and
kanamycin (15 mg/l). In addition to this, all colonies were resistant to
gentamicin (10 mg/l). This resistance is encoded by the vector part of
plasmid pSUP::Tn5-B20. These results indicated that all the
transconjugants still contained the entire plasmid and that very few, if any,
transposition events had taken place. Sub-culturing of several of these
colonies in LB broth (without kanamycin), followed by plating on to agar
containing X-gal and kanamycin, resulted in strains that did show
differences in colour. This suggests that transposition takes place at a low
frequency upon further sub-culturing. However, the latter phenotype was
not accompanied by loss of gentamicin resistance. This shows that these
cells still contained at least part of the vector moiety of pSUP::Tn5-B20.
These observations indicated that straightforward transposon mutagenesis in
A. calcoaceticuswas not possible with the plasmids described above.
MUTAGENESIS OF A.CALCOACETICUS THROUGH RANDOM

INSERTION OF A KANAMYCIN RESISTANCE MARKER
INTO THE CHROMOSOME
The possibility of producing insertion mutations in

A. calcoaceticus BD413 in a manner similar to that described above for
B.subtilis was investigated. Plasmids such as pBR322 do replicate,
although poorly, in A. calcoaceticus. Therefore the entire plasmid was not
used, as for B.subtilis, but only an antibiotic resistance marker, in our case
the nptII gene from pUC4K (Vieira and Messing, 1982) conferring
resistance to kanamycin. This gene was isolated as a 1.3 kb BamHI
fragment from an agarose gel. Chromosomal DNA of A. calcoaceticus was
cleaved simultaneously with BamHI, BgllI and Bell. This digestion yielded
fragments ranging in size from approx. 5 kb to less than 0.5 kb, with an
average of 1.3 kb. Chromosomal DNA was mixed with the DNA fragment
carrying the kanamycin resistance gene in a one to one molar ratio and
ligated (overnight at room temperature) at a final DNA concentration of 50
Itg/mi.
The ligation mixture was transformed into competent cells of

A. calcoaceticus strain BD413-ivllO (Juni, 1972). For competence
development, an overnight culture, grown at 30° C in minimal medium (Juni,
1974), was diluted 50-fold in the same medium and allowed to grow for 1 h.
The culture was then concentrated ten-fold by centrifugation (10 min, 5,000
g) and 1 ml of the cell suspension incubated for 45 min at 30°C with 40 Itl of
the ligation mixture. To allow segregation and expression of the kanamycin
resistance marker, 4 ml LB broth was added and the suspension was
incubated for an additional 4 h. Portions (0.1 ml) of the cells were plated on
to LB plates containing 15 mg kanamycin/I. Approximately 100 kanamycin-
resistant colonies (transformation frequency: 4 x 10- 6 ) were obtained per
plate. The collection of mutants was stored at _80° C after addition of
glycerol (final conc. 15% v/v) untilfurther use.
185
SELECTION OF SPECIFIC MUTANTS
A. calcoaceticus is naturally transformable (Juni, 1972) and

produces extracellular lipase(s) (Breuil and Kushner, 1975). Both
characteristics are major research subjects in our department and mutants
were therefore screened for defects in either of these two properties.
Lipase production was tested on nutrient agar plates containing 1% egg yolk,
1 % N aCI and 15 mg kanamycin/I. Among 9,000 kanamycin-resistant
colonies tested, two (designated AAC300 and AAC302) were unable to form
a precipitate on these plates. This indicated that these strains were mutated
in a gene(s) involved in (phospho)lipase production. To select for
transformation-deficient and auxotrophic mutants, 5,000 colonies were
replica-plated from the nutrient agar plates on to minimal plates containing
15 mg kanamycin/I. In addition, 25 J1.g of wild-type chromosomal DNA was
spread on these plates. Mutants that are unable to take up or integrate this
DNA cannot acquire a wild-type allele of the ivl gene and are therefore
unable to grow on these plates in the absence of isoleucine, leucine and
valine. In addition, auxotrophic mutants which arise through insertion of
the kanamycin resistance gene in a biosynthetic pathway gene are also unable
to grow since the selection for kanamycin resistance was continued. One
mutant (designated AAC211) was found to be completely transformation
deficient. Two others (designated AAC213 and AAC214) were found to be
severely impaired (with a transformation frequency more than 1,000-fold
lower than the wild-type). In addition, 14 mutants were found to be
auxotrophic (a frequency of 3 x 10'3). Thus, the results showed that it was
possible to generate a range of random mutants through insertion of a
kanamycin resistance marker into the chromosome of A. calcoaceticus.
Further Characterisation of the Mutants
The nature of the mutations in the two lipase-deficient strains

(AAC300 and AAC302) and the transformation-deficient strain (AAC211)
were further examined. To ensure that the observed mutant phenotype was
caused by the insertion of the kanamycin resistance marker rather than an
unlinked secondary mutation, A. calcoaceticus strain BD413-ivl10 was
transformed with chromosomal DNA isolated from the mutant strains. The
resulting kanamycin-resistant transformants all showed the mutant
phenotype. This indicated that the mutations were closely linked to the
kanamycin resistance marker and were probably caused by insertion of the
kanamycin resistance marker into the genes of interest.
This insertion was further verified by Southern hybridisation (Maniatis

et aI., 1982). The 1300 bp fragment, encoding kanamycin resistance, was
used as a probe and labelled with digoxigenine (Boehringer-Mannheim,
Germany). Since the fragment to be inserted was obtained by BamHI
digestion, it was still flanked by S all sites which could be used to excise the
marker from the chromosomal DNA of the mutants. Fig. 1 shows that this
fragment was present in mutant strains AAC211, AAC300 and AAC302
(compare lane 1 with lanes 3-5) and had undergone no major
186
rearrangements. Lane 2, which contains wild-type DNA, did not show any
hybridising fragment. The restriction enzyme Bell does not have a
recognition site in the fragment encoding kanamycin resistance. Be II
digestion of the mutant DNA should therefore provide information about the
chromosomal sequences flanking the marker. Fig. 1 shows that the insertion
in the two lipase mutants was not identical (compare lanes 7 and 8). This
Fig. 1. Southern hybridisation of the 1300 bp DNA fragment from plasmid

pUC4K, encoding kanamycin resistance, against chromosomal DNA of the
mutant strains AAC300 (lanes 3 and 7), AAC302 (lanes 4 and 8), AAC2ll
(lanes 5 and 9) and wild-type A.calcoaceticus BD4l3 DNA (lane 2).
Lanes 1 and 6 contain the fragment encoding the kanamycin resistance gene,
isolated as a BamHl fragment. Chromosomal DNA in lanes 2-5 was
digested with Sall; DNA in lanes 7-9 was digested with Bell. The
molecular size of the fragments is indicated in basepairs (bp) in the right
margin.
indicated that the marker in both mutants was situated at different positions
and that the mutants resulted from independent insertions. Whether the
same gene is involved remains to be elucidated. Similar results were
obtained for strain AAC211 (lane 9). Although Fig. 1 shows that this mutant
contained a Bell fragment of approximately the same size as mutant
AAC302, the two strains exhibited different phenotypes.
187
CONCLUDING REMARKS
The results presented above show that it is possible to apply

insertional mutagenesis to the naturally transformable A. calcoaceticus
strain BD413 by use of an isolated antibiotic resistance marker. It is
conceivable that all naturally transformable strains of A. calcoaceticus can
be genetically manipulated in this way. We believe that the availability of
such a reliable system for mutagenesis will simplify and strongly enhance
genetic research in A. calcoaceticus and related organisms.
ACKNOWLEDGEMENTS
We thank Drs. E. Juni, R. Simon and J. Frey for providing us with the
strains and plasmids used in this study.
REFERENCES
psychrophilic and mesophilic Acinetobacter species, Can. 1.
Microbiol., 21:423.
Campbell, A., 1962, Episomes, Adv. Genet., 11:10l.
Fellay, R., Krissch, H.M., Prentki, P., and Frey, J., 1989, Omegon-Km: a
transposable element designed for in vivo insertional mutagenesis
and cloning of genes in Gram-negative bacteria, Gene, 76:215.
Juni, E., 1972, lnterspecies transformation of Acinetobacter: genetic
evidence for a Ubiquitous genus, 1. Bacteriol., 112:917.
Juni, E., 1974, Simple genetic transformation assay for rapid diagnosis of
M oraxella osloensis, Appl. Microbiol., 27:16.
Kleckner, N., Roth, J., and Botstein, D., 1977, Genetic Engineering in vitro
using translocatable drug-resistance elements, 1. Mol. Bi 01.,
116:125.
Maniatis, T., Fritsch, E.F. and Sambrook, J., 1982, "Molecular Cloning. A
Laboratory Manual," Cold Spring Harbor Laboratory, Cold Spring
Harbor, NY.
Niaudet, B., Gose, A., and Ehrlich, S.D., 1982, Insertional mutagenesis in
Bacillus subtilis: mechanism and use in gene cloning, Gene,
19:277.
Simon, R., Priefer, U., and Puhler, A., 1983, A broad host range mobilization
system for in vivo genetic engineering: transposon mutagenesis in
Gram-negative bacteria, Biotechnology, 1:784.
Simon R., Quandt, J., and Klipp, W., 1989, New derivatives of transposon
Tn5 suitable for mobilization of replicons, generation of operon
fusions and induction of genes in Gram-negative bacteria, Gene,
80:161.
Singer, J.T., van Tuijl, J.J., and Finnerty, W.R., 1986, Transformation and
mobilization of cloning vectors in Acinetobacter spp., 1.
188
Vieira, J., and Messing, J., 1982, The pUC plasmids, an M13mp7-derived
system for insertion mutagenesis and sequencing with synthetic
universal primers, Gene, 19:259.
Youngman, P.J., Perkins, J.B., and Losick, R., 1983, Genetic transposition
and insertional mutagenesis in Bacillus subtilis with the
Streptococcus jaecalis transposon Tn917, Proc. N atl. Acad.
Sci. USA,80:2305.
189
GENETIC ORGANISATION OF ACI N ETOBACTER
A. Vivian
Molecular Genetics Group

Science Department
Bristol Polytechnic
Bristol BS16 1QY
UK
INTRODUCTION
Perhaps in the rush to embrace the recent technological advances in

gene manipulation, we have become mesmerised by organisation of the
genome at the molecular level of the gene. Undoubtedly, the power
afforded by the construction of gene libraries, coupled with transposon
mutagenesis and marker exchange, has resulted in a tremendous increase in
knowledge about the nature of genes and their products. Genetic studies of
bacteria in the fifties and sixties concentrated on the identification of genes
in terms of their phenotypes, together with attempts to map the order of
genes on the bacterial chromosome. This phase of activity seemed to reach
its zenith with the publication of the revised map of the Escherichia coli
K12 chromosome (Taylor and Trotter, 1967). This map confirmed that many
functionally-related genes were clustered in operons, although the
disposition of operonic clusters on the map was less clear. Bacteria other
than E. coli were studied on a lesser scale, but included Gram-negative
bacteria such as Salmonella typhimurium (Sanderson, 1967) and
Pseudomonas aeruginosa (Holloway, 1969), as well as the Gram-positive
actinomycete Streptomyces coelicolor (Hopwood, 1967b).
While bacteria such as E. coli seemed to have little apparent order to

the disposition of their genes on the circular map, S. coelicolor exhibited
an intriguing degree of symmetry in the arrangement of biosynthetic genes
(Hopwood, 1967a). Thus, loci for cys mutations were found to map at
(approximate locations on the conventional map) 1 o'clock (cysE), 4 o'clock
(cysC, D), 7 o'clock (cysB) and 10 o'clock (cysA). Another aspect of the
map involved the 'blank' regions at 3 and 9 o'clock, which were apparently
191
devoid of genes. Hopwood (1967a) postulated that the curious features of
this bacterium's genetic map might reflect repeated genome duplication
during evolution, while the so-called blank regions (devoid of both
auxotrophic and temperature-sensitive markers) could be artifacts of the
conjugation process or recombinational hotspots (Hopwood, 1967b).
In E. coli, the reasonably precise understanding of the mechanism of

gene exchange, via the F plasmid and derived Hfr strains, could be utilised to
map genes by an interrupted mating technique that produced time-related
entry of genetic markers into recipient bacteria (Hayes, 1968). In contrast,
gene transfer was less well-defined for many of the other bacteria studied.
In P. aeruginosa mapping was originally based on conjugation mediated by
the FP2 sex factor present in some strains (Stanisich and Holloway, 1969)
and transduction (Holloway, 1969), later superseded by IS-mediated
conjugation involving broad-host range plasmids such as R68.45 (Haas and
Holloway, 1976; Jacob et aI., 1977). In the Gram-positive actinomycete S.
coelicolor, a reliable chromosome mapping procedure, based on mixed
growth of parental strains, was developed well in advance of any clear idea of
the nature of the mechanisms underlying conjugation in this organism
(Hopwood et aI., 1973).
It was against this background that the first efforts were made (Towner
and Vivian, 1976a) to establish a system of conjugal gene transfer which, it
was hoped, would lay the foundations for a complete genetic analysis of
Acinetobacter calcoaceticus and its many-faceted lifestyle. These
efforts have subsequently been largely overtaken by the, possibly mistaken,
belief that gene-mapping in bacteria is obsolete. Cloning and sequencing
have revealed much about the fine detail of particular genes. The debt that
most 'cloners' owe to the pioneers of E. coli genetics, who provided the
basic map on which the detail could be assembled, is often forgotten. In
many other bacteria, induding Acinetobacter, there will be an increasing
need to study the relationships that exist between groups of genes and their
overall genetic organisation; perhaps, as with so many other aspects of
scientific research, chromosome mapping will again become fashionable.
MECHANISMS OF GENE TRANSFER IN ACINETOBACTER
Transformation
The first indication that gene transfer could be effected in

Acinetobacter was a report by Juni and Janik (1969) of transformation in
strain BD4 and its micro-encapsulated derivative, BD413. The efficiency of
transformation was exceptionally high and it was necessary merely to spread
cells freshly produced by growth in broth, together with suitable DNA, on an
agar plate and incubate to obtain transformants. Subsequently, Sawula and
Crawford (1972) used this technique to map tryptophan genes identified in
mutants of BD413. The seven trp genes mapped in three groups in an
arrangement not previously found in other organisms. The nucleotide
192
sequences of these groups have now been determined and potential
regulatory elements identified (Haspel et a!., this volume).
Cruze et al. (1979) improved and defined more precisely the

parameters for efficient transformation of BD413. Modifications to the
procedure for transforming BD413, including a plate-based method for
detecting cloned fragments of A. calcoaceticus DNA in E. coli, were
described by Neidle and Ornston (1986). It is worth noting that BD413 does
not contain any discernible indigenous plasmids. Coupled with its potential
for transformation, BD413 is therefore an extremely useful host for the study
of plasmids and cloned genes in A. calcoaceticus (M. Divers, P.L. Craven
and A. Vivian, unpublished results).
Organisms previously identified as A. calcoaceticus comprise a wide

range of distinct types of bacteria. Perhaps nowhere is this more apparent
than in relation to transformation. In spite of several attempts in our
laboratory, strain EBF65/65 (see below) could not be transformed by either
chromosomal or plasmid DNA. Conditions for the preparation of
competent cells in the closely related strain NCIB 8250 were, however,
described by Ahlquist et a!. (1980). Maximum transformation frequencies
were always lower (10- 4 to 10- 5 ) than those found (0.7%) by Juni and Janik
(1969). In NCIB 8250, competence occurred as a sharp peak at an
early stage of ba tch culture.
Conjugation
The broad host range conjugative plasmid RP4 (Datta et a!., 1971) was
originally isolated from P. aeruginosa. This plasmid, in contrast to its
behaviour in both E. coli and P. aeruginosa, was found to mediate
chromosome transfer between auxotrophically-marked strains of EBF65/65
(Towner and Vivian, 1976a). The origin of transfer, as determined by a
gradient of frequencies of marker inheritance in filter matings, was fixed for
a given strain inheriting RP4 from E. coli. Subsequent transfer of RP4 from
the initial donor strain (designated DO) to other derivatives of EBF65/65
conferred on the recipients the same donor characteristics. By screening
initial RP4 transconjugants (from a cross with E. coli) for those with
enhanced ability to mediate the transfer of an ile-l marker (normally
transferred as a distal marker by the DO donor), a novel donor (designated
D5) was also obtained. This enabled the circularity of the genetic map of
EBF65/65 to be confirmed (Towner and Vivian 1976b; Fig.I).
In a subsequent paper, it was demonstrated that the ability to mobilise

the chromosome of EBF65/65 was a feature of many incP group plasmids,
with the notable exception of R68.45 (Towner and Vivian, 1977). With the
benefit of hindsight and the intervening advances in molecular biology, it
would be interesting to re-examine the phenomenon of chromosome
mobilisation in EBF65/65 by RP4, and its relatives, with a view to
determining whether some special feature of the A. calcoaceticus
193
his-1
trp~hiS-2
mid_1 mdl-1
mdl-2
ile-1 arg-8
pgc-1 arg-10
arg-3
~
-----
trp-4 ala-1 phe-1 trp-1
his-6 cys-3 ~str-2

mdc-1 met-1
met-4 rif-1,-5
thi-1 Ieu _2 mdd-3
Fig. 1 Map of the chromosome of Acinetobacter calcoaceticus strain

EBF65/65 (Vivian, 1987) generated by co-inheritance of markers (Table 1)
from conjugative matings. Markers located on the circle are in the
correct order; those located outside of and overlapping these markers
are linked to them, but their precise order has not been confirmed.
(Reproduced with permission from "Genetic Maps 1987, vol. 4," p.240,
S.J. O'Brien, ed., Cold Spring Harbor Laboratory, New York).
chromosome, such as an insertion sequence, was involved, or whether

mobilisation was solely dependent on the properties of RP4_
Hinchliffe and Vivian (1980a) also used the plasmid pA V1, originally
isolated from a hospital in Pretoria, to mobilise the chromosome. This
plasmid transferred at frequencies up to 10% (about 1000-fold higher than
RP4) between strains of EBF65/65. Transfer of chromosomal genes was
also up to 100 times more efficient, although it was not possible to discern
any clear orientation of transfer (Hinchliffe and Vivian, 1980b).
Nevertheless, the co-inheritance of markers transferred by pAVl has been
successfully used to map a number of catabolic genes on the chromosome of
EBF65/65 (Vakeria et a1., 1984; Fig. 1).
Since this time, regrettably, no further markers have been assigned to

the genetic map of EBF65/65. However, many unanswered questions
remain about the nature of both RP4- and pA Vi-mediated conjugation in
A. calcoaceticus. It is to be hoped that future changes in scientific
priorities may permit these questions to be addressed
194
Table 1. Marker designations (see Fig. 1)
Auxotrophic
a1a-1 alanine
arg-3, -8, -10 arginine
cys-3 cysteine
his-1, -2,-6 histidine
i1e-1 isoleucine
1ys-1 lysine
met-1, -4 methionine
phe-1 phenylalanine
thi-1, -2 thiamine
trp-1, -2,-4 tryptophan
Catabolic
mdc-1 constitutive for mandelate
Rl regulon
mdd-1, -3 unable to grow on D(-)-mandelate as sole
carbon source
md1-1,-2 unable to grow on L(+)-mandelate as sole
carbon source
pgc-l unable to grow on phenylglyoxylate as sole
carbon source
Resistance
rif-1, -5 rifampicin
str-2 streptomycin
Transduction
Bacteriophages, as mediators of gene transfer, have been extremely

useful for fine structure mapping of bacterial genomes (e.g. Hartman et ai.,
1960). Their occurrence in A. calcoaceticus has been documented by
several authors (Twarog and Blouse, 1968; Ackermann et ai., 1973; Vieu et
aI., 1979; Pines and Gutnick, 1981) and sought by others. While phages
attacking strains such as JC17 (Hinchliffe and Vivian, 1980a) can be readily
obtained from fresh activated sludge effluent, some strains seem to be
reluctant hosts, including pAVr strains (lacking a restriction/modification
system) of EBF65/65; indeed, no phages active against EBF65/65, have yet
been obtained from a wide range of potential sources (M. Divers, E.
Hinchliffe and A. Vivian, unpublished results). Hermann and Juni (1974)
noted that, while they had little difficulty in obtaining phages, all ten phages
isolated were restricted in their host range to the strain on which they were
originally isolated.
Twarog and Blouse (1968) described a small short-tailed phage,

morphologically similar to the coliphages T3 and T7, which was thought to
transduce arginine and histidine genes in strain ATCC 15150; however, a
195
personal communication from R. Twarog, cited in Hermann and Juni (1974),
discounts this observation. The latter authors described a generalised
transducing phage (P78) for A. calcoaceticus strain 78. Frequencies of
transduction were low and no evidence for cotransduction of markers was
observed.
Strain NCIB 8250
An apparently novel gene transfer system was found in A.

calcoaceticus NCIB 8250 by Vakeria et al. (1985). This system did not
appear to involve a conjugative plasmid, although the strain is known to
harbour two cryptic plasmids, pAVIO (approx. 100 kb) and pAVll (approx.
15 kb). Mixed culture on membrane filters generated recombinants for
chromosomal genes at frequencies up to 10'. Any strain derived from the
wild type was capable of acting as a donor of markers, but crosses were
frequently observed to be strongly polarised in favour of one or other parent.
This system of gene transfer was quite distinct from the transformation
system in NCIB 8250 and was resistant to treatment with deoxyribonuclease
(Vakeria et ai., 1985).
Transpuson-M ediated Conjugation
Towner (1980) demonstrated that plasmids carrying the lysogenic

bacteriophage Mu have great difficulty establishing in A. calcoaceticus
EBF65/65. Making use of an RP4::Mu vector (pUN308), it was possible to
insert the trimethoprim/ streptomycin transposon, Tn3171, into the
chromosome of EBF65/65 close to the previously mapped ile-1 marker (Fig.
1). The insertion resulted in the appearance of arginine auxotrophy in the
resulting strain (C4508). An approximate ten-fold increase in the frequency
of transfer of the it e -1 marker was observed in crosses using C4508 carrying
RP4, with or without the presence of a second Tn3171 on the mobilising
plasmid. This technique clearly has a useful potential for enhancing the
frequencies of gene transfer over restricted regions of the genome (Towner,
1983).
MUTATION AND MAPPING STUDIES
A variety of treatments has been described for the isolation of mutants

in different strains of A. calcoaceticus. Towner & Vivian (1976a) used UV
and N-methyl-N'-nitro-N-nitrosoguanidine (NTG) to obtain auxotrophs of
strain EBF65/65. In the same strain, ethyl methane sulphonate (EMS) was
used to isolate mutants defective in mandelate catabolism (Vakeria et aI.,
1984). A further procedure for EMS mutagenesis was described by Goosen
et al. (1987) for use with strain LMD79.41. In Ac78, Herman and Juni
(1974) used NTG mutagenesis, together with an enrichment procedure using
cycloserine and ampicillin, to obtain auxotrophs. In a comprehensive
comparison of treatments with UV, NTG, EMS and 8-methoxypsoralen plus
196
near-uv irradiation, Ahlquist et al. (1975) concluded that EMS and NTG
were the most effective methods for the generation of auxotrophs. These
authors used an enrichment procedure involving vancomycin and penicillin V
with strain NCIB 8250.
There appear to have been very few attempts to use transposons as

mutagenic agents in A. calcoaceticus (reviewed by Towner, this volume).
Tn5, which has proven to be so useful in a wide range of Gram-negative
bacteria, was found to show insertional specificity in strains HO-IN and
BD413 (Singer and Finnerty, 1984). Towner (1983) also noted a preferred
site of insertion in an arg gene in EBF65/65 for the Tn7-like transposon,
Tn3171. However, Doten et al. (1987) obtained Tn5 insertions in a number
of pea genes cloned from strain ADP161 (a derivative of BD413), using
mutagenesis mediated by bacteriophage lambda in E. coli. This would
appear to indicate that the specificity of insertion observed in the genome of
BD413 does not operate in E. coli.
Towner (1978) produced the first definitive map of the circular

genome of strain EBF65/65. Since then, further genes have been mapped,
particularly those concerned with the catabolism of mandelate (Vivian,
1987); these are listed in Table 1 and included in the current genetic map
(Fig. 1).
CONCLUSIONS
Two opposing types of model to account for the disposition of genes on

the chromosomes of bacteria have been suggested. One type invokes
chromosome duplication, followed by progressive loss of duplicated
functions, while the other involves the acquisition of gene functions, perhaps
mediated by plasmids and transposons (Riley and Anilionis, 1978). As
already mentioned, duplication has been proposed for S. coelicolor
(Hopwood, 1967a), while Zipkas and Riley (1975) detected a similar
symmetry of functionally related genes in E. coli. Thus, four clusters
located at 90 0 contained virtually all the genes concerned with the
catabolism of glucose in E. coli (Riley and Anilionis, 1978). Knowledge of
the genome organisation of Acinetobacter remains incomplete and,
unfortunately, it is not possible, as yet, to discern any overall features in the
arrangement of the genome. Nevertheless, there are sufficient indications
to conclude that further work aimed at defining the gross topological
structure would be of considerable comparative interest in terms of the
evolution and functional significance of bacterial chromosomes. The tools
for such investigations are available, particularly the efficient conjugation
and transformation systems available for several strains. The prerequisites
for genetic manipulation in A. calcoaceticus are also in place to enable the
fine structure of genes to be determined. Many scientists are perhaps still
inclined to feel that all bacteria are essentially like E. coli; this should not
prevent further work to discover just how different A. calcoaceticus might
be!
197
REFERENCES
Ackermann, H-W., Brochu, G., and Cherchel, G., 1973, Structure de trois
nouveaux phages de Bacterium anitratum (groupe B5W),
I. Microscopie, 16:215.
Ahlquist, E.F., Fewson, e.A., and Ritchie, D.A., 1975, The induction of
mutants of Acinetobacter calcoaceticus NCIB 8250 and their
selection by vancomycin, I. Gen. Microbiol., 91 :338.
Ahlquist, E.F., Fewson, e.A., Ritchie, D.A., Pod more, J., and Rowell, V.,
1980, Competence for genetic transformation in Acinetobacter
calcoaceticus NCIB 8250, FEM S Microbiol. Lett., 7:107.
Cruze, J.A., Singer, J.T., and Finnerty, W.R., 1979, Conditions for
quantitative transformation in Acinetobacter calcoaceticus,
Curro Microbiol., 3:129.
Datta, N., Hedges, R.W., Shaw, E.J., Sykes, R.B., and Richmond, M.H., 1971,
Properties of an R factor from Pseudomonas aeruginosa,
.T. Bacteriol., 108:1244.
Doten, R.e., Ngai, K-L., Mitchell, D.J., and amston, L.N., 1987, Cloning and
genetic organization of the pea gene cluster from Acinetobacter
calcoaceticus, I. Bacteriol., 169:3168.
Goosen, N., Vermaas, D.A.M., and Van de Putte, P., 1987, Cloning of the
genes involved in synthesis of coenzyme pyrrolo-quinoline-
quinolone from Acinetobacter calcoaceticus, I.Bacteriol.,
169:303.
Haas, D., and Holloway, B.W., 1976, R factor variants with enhanced sex
factor activity in Pseudomonas aeruginosa, }Vl 01. Gen. Genet.,
144: 243.
Hartman, P.E., Loper, J.e., and Serman, D., 1960, Fine structure mapping by
complete transduction between histidine-requiring S a I m 0 nell a
mutants,.T. Gen. Microbial., 22:323.
Hayes, W., 1968, "The Genetics of Bacteria and their Viruses," Blackwell
Scientific Publications, Oxford.
Herman, N.J., and Juni, E., 1974, Isolation and characterization of a
generalized transducing bacteriophage for A c i net 0 b act e r,
I. Virol., 13:46.
Hinchliffe, E., and Vivian, A., 1980a, Naturally occurring plasmids in
Acinetobacter calcoaceticus: a P class R factor of restricted
host range, I. Gen. Microbiol., 116:75.
Hinchliffe, E., and Vivian, A., 1980b, Gene transfer in Acinetobacter
calcoaceticus: fertility variants of the sex factor pAV1,
I. Gen. Microbiol., 119:117.
Holloway, B.W., 1969 Genetics of Pseudomonas, Bacteriol. Rev., 33:419.
Hopwood, D.A., 1967a, In discussion to F.W. Stahl's paper: Circular genetic
maps, I. Cell. Physiol., 70(suppl.l):1.
Hopwood, D.A., 1967b, Genetic analysis and genome structure in
Streptomyces coeficolor, Bacterial. Rev., 31:373.
Hopwood, D.A., Chater, K.F., Dowding, J.E., and Vivian, A., 1973, Advances
in Streptomyces coelicalar genetics, Bacteriol. Rev., 37:371.
198
Jacob, A.E., Cresswell, J.M., and Hedges, R.W., 1977, Molecular
characterization of the P group plasmid R68 and variants with
enhanced chromosome mobilizing ability, F EM S
Microbiol. Lett., 1:7l.
Juni, E., and Janik, A., 1969, Transformation of Acinetobacter
calco-aceticus (Bacterium anitratum), I.Bacteriol.,
98:28l.
Neidle, E.L., and Ornston, L.N., 1986, Cloning and expression of
Acinetobacter cal co acetic us catechol 1,2-dioxygenase
structural gene catA in Escherichia coli, I.Bacteriol.,
168:815.
Pines, 0., and Gutnick, D.L., 1981, Relationship between phage resistance
and emulsan production, interaction of phages with the cell-
surface of Acinetobacter calcoaceticus RAG-I,
Arch. Microbio!., 130:129.
Riley, M., and Anilionis, A., 1978, Evolution of the bacterial genome,
Ann. Rev. Microbiol., 32:519.
Sanderson, K.E., 1967, Linkage map of Salmonella typhimurium,
Sawula, R.V., and Crawford, J.P., 1972, Mapping of the tryptophan genes of
Acinetobacter calcoaceticus by transformation, I. Bacterio!.,
112:797.
Singer J.T., and Finnerty, W.R., 1984, Insertional specificity of transposon
Tn5 in Acinetobacter sp., I.Bacteriol., 157:607.
Stanisich, V.A., and Holloway, B.W., 1969, Conjugation in Pseudomonas
aeruginosa, Genetics, 61:327.
Taylor, A.L., and Trotter, C.D., 1967, Revised linkage map of Escherichia
coli, Bacteriol. Rev., 31:332.
calcoaceticus, I. Gen. M icrob io!., 104: 175.
Towner, K.J., 1980, Behaviour of bacteriophage Mu in Acinetobacter
calcoaceticu.s EBF65/65, I. Gen. Microbio!., 121:425.
mobilization in Acinetobacter calcoaceticus EBF65/65,
Genet. Res., 41:97.
Towner, K.J., and Vivian, A., 1976a, RP4-mediated conjugation in
Acinetobacter calcoaceticus, I. Gen. Microbiol., 93:355.
Towner, K.J., and Vivian, A., 1976b, RP4 fertility variants in
Acinetobacter calcoaceticus, Genet. Res., 28:30l.
Towner, K.J., and Vivian, A., 1977, Plasmids capable of transfer and
chromosome mobilization in Acinetobacter calcoaceticus,
I. Gen. Microbiol., 101:167.
Twarog, R., and Blouse, L.E., 1968, Isolation and characterization of
transducing bacteriophage BPI for Bacterium anitratum
(Achromobacter sp.), I. Virol., 2:716.
Vakeria, D., Vivian, A., and Fewson, C.A., 1984, Isolation, characterization
and mapping of mandelate pathway mutants of Acinetobacter
calcoaceticus, 1. Gen. Microbiol., 130:2893.
199
Vakeria, D., Fewson, C.A., and Vivian, A., 1985, Gene transfer in
Acinetabacter calcaaceticus NCIB 8250, FEMS
Microbial. Lett., 26:141.
lysotypie de Acinetobacter, Ann. Microbial., 130A:405.
Vivian, A., 1987, Acinetabacter calcaaceticus strain EBF65/65, in:
"Genetic Maps 1987, vol. 4," p.240, S.J. O'Brien, ed., Cold Spring
Harbor Laboratory, New York.
Zipkas, D., and Riley, M., 1975, Proposal concerning mechanism of evolution
of the genome of Escherichia cali, Prac. Nat!. Acad. Sci.
V SA., 72: 1354.
200
EVOLUTION OF GENES FOR THE fi-KETOADIPATE
PATHWAY IN ACINETOBACTER CALCOACETICUS
L.N. Ornston" * and E.L. Neidle b
"Department of Biology, Yale University

New Haven, CT 06511, USA
bDepartment of Microbiology, Medical School

University of Texas at Houston
Houston, TX 77225, USA
INTRODUCTION
The B-ketoadipate pathway (Fig. 1) is widely distributed in the

microbial world (Stanier et aI., 1966; Ornston and Ornston, 1972; Stanier
and Ornston, 1973; Cain, 1980; Parke and Ornston, 1984) and was one of the
first subjects of physiological investigation of enzyme induction in bacteria
(Stanier, 1951). Enzymes associated with the pathway proved to be inducible
in fluorescent Pseudomonas species, and later studies revealed that
mechanisms of transcriptional control were conserved in this biological
group (Ornston, 1966; Kemp and Hegeman, 1968). Representatives of
Acinetobacter (formerly M oraxella) share induction patterns unlike
those found in Pseudomonas (Omston, 1966; Canovas and Stanier, 1967;
Canovas et aI., 1967; Stanier and Ornston, 1973). Investigation of the
genetic basis for the different forms of transcriptional regulation exercised
in the two genera has given insight into processes underlying their
evolutionary divergence.
This review summarises evidence pointing to five general conclusions.

First, structural genes for isofunctional enzymes from Acinetobacter and
Pseudomonas share common ancestors. Where examined, enzymes from
the two biological groups share amino acid sequence identity that usually
exceeds 40%. The level of amino acid sequence similarity in the enzymes is
remarkable because, with few exceptions, the GC content of the structural
genes for the enzymes differs by about 20%. The difference in GC content is
reflected in highly distinctive patterns of codon usage. Second, extensive
*corresponding author
201
Fig. 1. Designations for structural genes associated with the B-
ketoadipate pathway in Aeinetobaeter. The protocatechuate 3,4-dioxygenase
gene, formerly peaA, now is indicated by peaG and peaH which respectively
encode its nonidentical a and B subunits. Designations eatl and eatJ or
peal and pcaJrepresent the a and B subunits respectively of B-ketoadipate
succinyl GoA transferase. Structural genes for this enzyme were formerly
designated eatE and peaE. Gene designations correspond to enzymes as
follows: pobA, p-hydroxybenzoate hydroxylase (EG 1:14:13:2); peaG and
peaH, protocatechuate 3,4-dioxygenase (EG 1.99.2.3); peaB. B-carboxy-
eis,eis-muconate cycloisomerase (EG 5.5.1.2); peaC, T-
carboxymuconolactone decarboxylase (EG 4.1.1.44); peaD or eatD, B-

ketoadipate enol-lactone hydrolase (EG 3.1.1.24); peal and peaJ or eatl
and eatJ, B-ketoadipate:succiny1-GoA transferase (EG 2.8.3.6); peaF or
eatF. B-ketoadipyl GoA thiolase (EG 2.3.1.16); benABC, reductive
benzoate dioxygenase (EG 1.13.99.2); benD. 1,2-dihydro-l,2-
dihydroxybenzoa te dehydrogenase (EG 1. 3 . 1. 25); eatA. catechol 1,2-
dioxygenase (EG 1.13.11.1); eatB, eis,eis-muconate cycloisomerase (EG
5.5.1.1); eatC, muconolactone isomerase (EG 5.3.3.4).
202
GENES METABOLITES GENES
OCOO-
~I
benzoate
benABC
COO -
OOCO
OOR
~ OR ~ OR
H
P -hydroxybenzoate 1,2-dihydro-l,2-dihydroxybenzoate
pabA
I I benD
O O C O OR I OOR
~ OR "'" I
OR
r0
protocatechuate catechol
peaHG 02l 2
earA
OOC
C"'"
COO -
COO-
C
""
COO -
COO -
B-carboxy-eis, cis -muconate cis, cis -rnuconate
pcaB carS
OOC
LS o
:::----.. =0
COO- ~COO-
~=O
Y-carboxymuconolactone muconolactone
pcaC
/ calC
€~~O
B-ketoadipate enol-lactone
peaD
II ealD
o
C COO -
COO -
B-ketoadipate
peal]
II
B-ketoadipyl-CoA
ealU
pcaF
I J
succinyl-CoA
CalF
+
acetyl-eoA
Fig. 1.
203
gene rearrangements took place as the Il-ketoadipate pathway diverged in
Acinetobacter and Pseudomonas. In some cases, gene rearrangements
resulted in divergent transcriptional controls. In other instances, the amount
of gene rearrangement exceeds the level required for divergent
transcriptional regulation. Third, genes for physiologically interdependent
steps in the B-ketoadipate pathway tend to be linked in supraoperonic
clusters within the chromosomes of Acinetobacter and Pseudomonas.
Selective pressures favouring this clustering have not been defined, hut the
genetic properties of Acinetobacter open this problem to analysis. Fourth,
genes carried in the Acinetobacter chromosome possess the potential
ability to interact. Some of the interactions are physiological and involve the
intermediacy of either regulatory proteins or inducer-metabolites. Other
interactions are genetic, as exemplified by sequence exchange between
nearly identical regions of DNA encoding isofunctional enzymes associated
with the Il-ketoadipate pathway. Fifth, entirely different patterns of DNA
sequence repetition were acq uired as homologous structural genes diverged
in Acinetobacter and Pseudomonas. The patterns suggest that sequence
exchange between slipped DNA strands caused mutations leading to genetic
divergence; repair events involving sequence exchange between slipped DNA
strands may contribute to conservation of the evolved sequences.
The extraordinary competence of one Acinetobacter strain for

natural transformation (Juni, 1972) opens opportunities for analysis of genes
and gene clusters. This review concludes with a summary of some procedures
that may facilitate future examination of the genetic properties of
Acinetohacter.
HOMOLOGOUS GENES WITH ANALOGOUS

FUNCTIONS IN ACINETOBACTER AND PSEUDOMONAS
In so far as has been determined, genes with analogous functions in

Acinetobacter and Pseudomonas share common ancestry (McCorkle et
aI., 1980; Yeh et aI., 1980b; Yeh and Ornston, 1981; Kaplan et aI., 1984;
Neidle et aI., 1988; Hartnett et aI., 1990; Ornston et aI., 1990b).
Homologous proteins from the two genera possess a level of amino acid
sequence identity that is particularly striking in the light of divergence at the
level of DNA that has produced a difference of about 20% in GC content
(Neidle et aI., 1988; Hartnett et aI., 1990). These observations are
illustrated by comparison of the respective catM BC and catRBC genes
from Acinetobacter and Pseudomonas (Table 1). The catB and catC
genes encode enzymes catalysing consecutive metabolic transformations of
muconate (Fig. 1). Common ancestry of the catB and catC genes from the
two genera is demonstrated by conservation of amino acid sequence
identities of ahout 50% in the respective gene products (Aldrich et aI., 1987;
Aldrich and Chakrabarty, 1988; J. Houghton and J. Hughes, unpublished
observations). The crystal structures of the protein products of the P.
putida catB and catC gene products have been determined (Goldman et
aI., 1985; Katz et aI., 1985; Katti et aI., 1988; AlleweIl, 1(89); many of the
204
Table 1. Properties of homologous genes from Acinetobacter and Pseudomonas
Acinetobacter Pseudomonas
catHa catB catC catR a catB catC
GC in DNA 44.6 1+4.7 41. 6 66.2 63.6 61. 8
ATT codons for Ile b 72 .0 63.6 57.1 22.2 26.7 33.3
TTA codons for Leu b 6.0 29.6 37.5 2.4 0.0 0.0
AGA codons for Argb 40.0 0.0 0.0 0.0 0.0 0.0
athe catH and catR regulatory gene products recognise muconate as an inducer
and share an amino acid sequence identity of 40%. The genes differ in that
catH encodes a repressor whereas catR encodes an activator of transcription.
bExpressed as a percentage of all the codons
amino acid residues conserved with those of the corresponding

Acinetobacter gene products make essential contributions to protein
structure (Katti et al., 1988).
The crystal structures of Acinetobacter CatM and Pseudomonas

CatR are unknown, but the 40% amino acid sequence identity shared by
these proteins indicates that they have experienced selective pressure for
extensive structural conservation (N eidle et al., 1989; Rothmel et al., 1990;
J. Houghton, unpublished observations). Some regions of conservation,
notably those associated with a potentially DNA-binding helix-turn-helix, are
conserved with other members of the wide ranging LysR family (Henikoff et
al., 1988) of regulatory proteins (Fig.2). Other regions of sequence
conservation are unique to CatM and CatR; some of these regions may have
been selected on the basis of their common function, binding of the inducer
muconate so as to govern expression of the cat structural genes.
In general, CatM and CatR resemble each other more closely than
they resemble a consensus sequence for other members of the LysR family
(Henikoff et al., 1988). As illustrated in Fig. 2, the sequence similarities are
widely distributed through the primary structures of CatM and CatR, and it
thus seems probable that the muconate-binding regulatory proteins shared
their evolutionary history after divergence from other transcriptional
regulators. The inferred common evolutionary history for the muconate-
binding proteins is remarkable because they acquired opposite mechanisms
205
A
m ~ ~
Pseudomonas CatR ArgAl aAl aGl uLeuLeuHi sI 1eAl aGl nProProLeuSerArgGl nI 1eSerGl nLeuGl u
* * ~ * * * * * * * * * * * * *
Aci netobacter CatM LysAl aAl aGl uLysLeuCysI 1eAl aGl nProProLeuSerArgGI n1 I eGl nLysLeuGl u
** * ** **** **
Consensus Al aAl aAl aArgAl aLeuHi sLeuSerGl nProAl all eSerArgGl nIl eAl aArgLeuGl u
130 140
Pseudomonas CatR Gl nVal Gl uAI aLeuLysSerGl yArgI I eAspI leAl aPheGl yArgIl eArgIl eGI uAsp
* * * * * * * * * * * *
Aci netobacter CatH Gl n1 1eAsnAI aLeuLysGI nGl yLys1 1eAspLeuGl yPheGI yArgLeuLysIl eThrAsp
* * * * *
Consensus ProGI uGl uAI aLeuArgAl aGl yGl uLeuAspLeuAl all eSerXXXAspProLeuHi sSer
230 240
Pseudomonas CatR GI uLeuGI nThrAI all eGl yLeuVal Al aVal Gl yVal Gl yVal ThrLeuVal ProAl aSer
* * * * * * * * * * * *
Aci netobacter CatH GI uIl eGl nLeuAI aLeuGl yLeuVal Al aAl aGI yGI uGl yVal CysIl eVal ProAl aSer
* * * * *
Consensus Gl ySerVal XXXMetVal Val HetLeuAI aAl aGI yVal Gl yII eAl aAl aLeuProLeuVal
Fig. 2. Conserved regions in the deduced amino acid sequences of

Acinetobacter CatM and Pseudomonas CatR. Numerals indicate positions of
amino acid. residues in the CatR sequence; asterisks indicate amino acid
sequence identity of CatM with either CatR or a consensus sequence
(Henikoff et al., 1988). The consensus sequence, a rough representation
of an ancestral sequence, contains the most frequently shared amino acid
residues in the LysR family of transcriptional regulators. (A) All three
sequences share substantial similarity in the helix-turn-helix region
that is likely to be associated with DNA binding. In this region, and in
internal regions (exemplified by (B) and (C)) the CatR and CatM proteins
resemble each other more closely than they resemble the putative
ancestral sequence.
of transcriptional control: CatM represses gene expression in

Acinetobacter (Neidle et al., 1989) whereas CatR activates transcription
in Pseudomonas (Wheelis and Ornston, 1972; J. Hughes et al., unpublished
observations). A possibility raised by these observations is that
activation is a preferred mode of regulation for the GC-rich genes of
Pseudomonas, whereas repression may be a favoured mechanism for control
of the AT-rich genes in Acinetobacter.
Selection for different mechanisms of transcriptional control was

one force favouring divergence of catM and catR. Another selective
force, acting generally upon genes in bacteria, appears to have been a
demand for codon usage consonant with the distribution of transfer RNA
molecules in the host cell (Grantham et al., 1981; Ikemura, 1981a,b).
An illustration of differences in codon usage, shown in Table 1, is the
206
predominance of ATT codons for isoleucine in the relatively AT -rich genes
from Acinetob acter.
Consistent with the notion that the translational process exerts a

powerful selective force on gene divergence, patterns of codon usage for
different genes within a species tend to be remarkably constant (Kaplan et
aI., 1984; West and Iglewski, 1988). Constancy of GC content also tends to
be observed (Chargaff, 1950), and it is difficult to escape the conclusion that
selection for codon usage exerts a profound effect on the GC content of
genes (Grantham et al., 1981; Ikemura, 1981a,b). However, it should be
noted that selective demand for codon usage may not be the sole factor
influencing the GC content of genes. As described below, gene conversion
mutations may shift DNA sequences within the chromosome and therefore
may serve as a genetic source providing to a gene a GC content consistent
with that of surrounding genes.
There are significant exceptions to general patterns of codon usage.

Known exceptions include regulatory genes (Konigsberg and Godson, 1983;
Sharp and Li, 1986), and an additional example is the Acinetobacter catM
gene (Neidle et aI., 1989). As shown in Table 1, this gene has resisted the
tendency to relatively high usage of the TTA codon illustrated by the
Acinetobacter catB and catC genes. Even more striking is the presence of
AGA as four of the ten arginine codons in the Acinetobacter catM gene.
The AGA codon is found in none of the seven cat structural genes from
Acinetobacter, yet occupies two of the first seven codons translated during
synthesis of the catM encoded repressor molecule (Fig. 3). One
interpretation of the unusual codon usage patterns in regulatory genes is that
their low levels of expression have not challenged pools of activated transfer
RNA molecules to the extent demanded by highly expressed structural genes
(Konigsberg and Godson, 1983; Sharp and Li, 1986). The unusual
distribution of AGA codons in catM prompts the additional speculation that
the exceptional codon usage pattern may have been selected, perhaps in part
as a mechanism for controlling expression of the repressor gene. It is
Pseudomonas catR ATGGAGCTGCGCCACCTGCGTTACTTCAAG

MetGluLeUArgHisLeUArgTyrPheThr
* * * * * * * * *
MetGluLeuArgHisLeUArgTyrPheval
Acinetobacter catM ATGGAACTAAGACACCTCAGATATTTTGTG
Fig. 3. The translational origin of the Pseudomonas catR and

Acinetobacter catM genes. The amino acid sequences of the translated
gene products are nearly identical in this region. Highlighted in bold
and underlined are two of the four AGA arginyl codons in catM. The
presence of these codons, absent in the Pseudomonas catRBC region (Table
1) and unrepresented in the seven Acinetobacter cat structural genes,
suggests that the normally infrequent AGA codons may have been selected
during divergence of the catM gene.
207
noteworthy that a mutation in the cognate tRNA for AGA selectively lowers
expression of AGA-rich genes in Escherichia coli (Chen et al., 1990).
GENE REARRANGEMENTS DURING DIVERGENCE

OF THE B-KETOADIPATEPATHWAY
Without known exception, genes for isofunctional enzymes from

Acinetobacter and Pseudomonas can be traced to a common ancestor.
This knowledge opens opportunities to explore how genes were rearranged
during evolution and, as described later, to analyse how DNA sequences
within genes changed during their divergence.
Extensive gene rearrangements took place during evolution of the B-

ketoadipate pathway. As shown in Fig. 4A, the catI] FD region,
unrepresented in fluorescent Pseudomonas species (Ornston, 1966; Kemp
and Hegeman, 1968; Hughes et a1., 1988), is fused to the catBC genes in
Acinetobacter (Shanley et al., 1986; A. Harrison et al., unpublished
observations). The cat genes share an element of organisation common to
many operons governed by members of the LysR family of transcriptional
regulators (Henikoff et al., 1988): the regulatory genes lie upstream and are
divergently transcribed from the structural genes that they govern (Fig. 4A).
Organisation of the Pseudomonas pea genes is somewhat

fragmentary as compared with the tight linkage of these genes in
Acinetobacter. Depicted in Fig. 4B, the pcaBDC operon represents the
most extensive linkage of pea genes in P. putida (Hughes et al., 1988). In
contrast, all of the Aeinetobacter pea structural genes are linked in a
single operon (Doten et al., 1987a). Comparison of the two operons reveals
that the order of the neighbouring peaD and peaB genes was reversed
during their divergence in Aeinetobacter and Pseudomonas (Fig. 4B).
Rearrangement of genes during evolution of the B-ketoadipate

pathway within Acinetobaeter appears to have occurred because the
catI] F D region, lying at the end of the cat operon (Shanley et al., 1986), is
represented by its homologue pcaIJ F D placed at the beginning of the pea
operon (Doten et al., 1987a; Fig. 4C). DNA sequences of the cat IJ F and
peaIJ F regions are nearly identical and, as described below, the similarity
appears to afford opportunity for DNA sequence exchange between the two
regions.
SUPRAOPERONIC CLUSTERING
Transductional analysis of fluorescent Pseudomonas strains

(Rosenberg and Hegeman, 1969; Wheelis and Stanier, 1970) gave the first
indications that independently regulated genes for aromatic catabolism were
clustered in the bacterial chromosome. The clustered genes frequently
participate in the same overall physiological function, such as the utilisation
208
A
<--- ----------------------------->
Ad netobacter leatM leatB eatC eatI eatJ eatF caW I
Pseudomonas leatR leatB eatC I
<--- ---------->
--------------------------------------->
Acinetobacter Ipear peaJ peaF peaD peaB peaC peaH peaG I
Pseudomonas IpeaB pcaD peaC I
-------------->
----------------------------->
Ad netobacter leatB eatC cat! eatJ eatF eatD I
IIII11111 t i l l
I I I I I I II I I I I I
Acinetobacter Ipear peaJ peaF pc aD peaB peaC peaH peaG I
--------------------------------------->
Fig. 4. Organisation of eat and pea genes. Genes within operons are
enclosed in boxes, and arrows indicate the direction of transcription.
(A) In Aeinetobaeter, the eatBC sequence forms an operon with the
downstream eatlJFD region. A direct counterpart of the catIJFD region is
not present in fluorescent Pseudomonas strains such as P. putida. In
these organisms, the physiological function of the region is fulfilled by
the peal, peaJ, peaF and pc aD genes. (B) All of the structural genes
specifically required for metabolism of protocateehuate (Fig. 1) are
linked within the pea operon of Aeinetobaeter. Organisation of pea genes
in P. putida is relatively fragmentary. In the longest P. putida pea
operon, the transcriptional order of the linked peaB and peaD genes is
the reverse of that found in Aeinetobaeter. (C) In Aeinetobaeter, the
eatlJFD region appears at the end of the eat operon whereas the
homologous region lies at the beginning of the pea operon. Vertical
lines indicate near identity of the eatlJF and the pealJF genes in DNA
sequence. The eatD and peaD genes have diverged substantially in
Aeinetobaeter (Yeh et al., 1980a).
209
of benzoate. The most fully characterised supraoperonic cluster, associated
with benzoate metabolism in Acinetobacter (Neidle et al., 1987) is
depicted in Fig. SA. The complete DNA sequence of this 16 kb cluster has
been determined. It contains three open reading frames (ORF1, ORF2 and
ORF3; Fig. SA). The functions of these open reading frames are yet to be
clearly established.
Positioned to be expressed in a transcript downstream from catA, the

structural gene for catechol oxygenase, are ORF2 and ORF3 (Fig. SA).
Genetic disruption of these open reading frames does not prevent growth
with either anthranilate or benzoate, compounds that are metabolised via
catechol (Neidle et al., 1989). As described below, some evidence suggests
that these open reading frames may contribute to transcriptional regulation.
The catM gene, the only regulatory gene to be characterised in the 16

kb segment, represses expression of both c a tA and the cat B C] J F D operon
(Neidle et al., 1989). As is often the case with members of the lysR family of
regulatory genes (Henikoff et al., 1988), the catM gene is transcribed
divergently from the genes it controls. The orientation of the genes may
contribute to regulation of their expression, and this possibility can be
explored by examining the physiological consequences of reversing the
orientation of the catA gene or the catBC]] FD operon.
Distance (Kbp) 5 10 15
1 I I I ,----, rl- - - - - - - - - - - - - , 1
IbenA
I
benB benC benD ORF1 H catA ORF2 ORF3 H catM HcatB catC cat! catJ catF catD I
1 I 1 ~IL._ _ _ _ _ _ _ _ _ _ ---'1
-------------------------> --------------> <----- ----------------------------->
Distance (Kbp) 5 10 15
I ,----,
lpea! pcaJ pcaF peaD pcaB peaC peaH pcaG I-----------i~
----------------------------------------->
Fig. 5. Supraoperonic clustering. The distances are in kb from the
translational origin of the first operon in the cluster. Arrows indicate
the direction of transcription. (A) The ben and cat genes are clustered
within 16 kb of DNA and are transcribed as four separate transcriptional
units. The catM gene encodes a repressor of both catA and the catBCIJFD
operon. Functions are yet to be assigned to the open reading frames
designated ORF1, ORF2 and ORF3. (B) The direction of transcription of
pobA is yet to be determined; this gene lies about 5 kb downstream from
the end of the pcaIJFDBCHG operon.
210
As described below, transcriptional control of the ben-cat genes is
complex, and it is evident that additional regulatory genes lie outside the 16
kb segment. The most persuasive evidence in support of this conclusion is
the observation that the cloned catA gene, carried on a broad host-range
plasmid in an Acinetabacter strain from which the 16 kb ben-cat segment
has been deleted, remains subject to induction in response to either
benzoate or muconate (Neidle et aI., 1989).
All of the eight structural genes specifically associated with

protocatechuate metabolism are contained within the pcal J P DBCR G
operon (Doten et aI., 1987a; Hartnett et aI., 1990). As shown in Fig. 5B, the
independently transcribed pabA gene lies about 5 kb downstream from the
pcaG gene (B. Averhoff and G. Hartnett, unpublished observations). Thus
an additional supraoperonic cluster in the Acinetobacter chromosome has
been identified. Other genes that may prove to be associated with this
cluster are those concerned with conversion of hydroaromatic growth
substrates, quinate and shikimate, to protocatechuate (Tresguerres et aI.,
1970). These genes, like the pcaIlPDBCRG operon, are expressed in
response to the inducer protocatechuate, and all of the genes governed by
protocatechuate appear to share a common regulatory gene (Canovas et aI.,
1968b).
Thus far we have discussed chromosomal genes, and genes for the B-
ketoadipate pathway generally are associated with the chromosome.
Exceptions exist (Hewetson et aI., 1978). Of particular interest is the full set
of cat genes carried on an Acinetabacter plasmid associated with benzene
metabolism (Winstanley et aI., 1987). The distance between catA and catB
is greater in the plasmid than in the chromosome, and it is possible that there
are other differences in gene organisation. It is not known if the plasmid
carries a counterpart of the regulatory gene catM.
Clustering of plasmid-borne genes can be readily interpreted as the

consequence of a demand for their cotransfer during conjugation (Wheelis,
1975). This argument is less effective when applied to chromosomal genes
that, as described above, have demonstrated marked potential for gene
rearrangement during their divergence. In the absence of selection,
rearrangement mutations would be expected to lead to randomisation of the
chromosomal location of independently transcribed genes. Absence of such
randomisation, exemplified by closer clustering of catA and catB on the
chromosome than on a plasmid, indicates that selective demands can
maintain close linkage of genes that contribute to an overall nutritional
function such as metabolism of benzoate. One set of demands may be
physiological: examples might be either global transcriptional controls
exercised over linked operons or the formation of DNA loops that may
influence gene expression (Schleif, 1988). Other selective demands may be
genetic. DNA sequence exchange may buffer genes against change by
repairing mutations (Edelman and Gally, 1970; Nagylaki and Petes, 1982;
Xiong et aI., 1988; Meagher et aI., 1989). The possibility exists that such
genetic interactions contribute to the stabilisation of genes within linkage
211
groups (Orns ton et a!., 1990a, b). Physiological and genetic interactions
between chromosomal Acinetobacter genes for the B-ketoadipate pathway
are discussed below.
PHYSIOLOGICAL AND GENETIC INTERACTIONS

BETWEEN DIFFERENT CHROMOSOMAL DNA SEGMENTS
Many of the physiological interactions between different chromosomal

DNA segments are indirect because they are mediated in part by inducer-
metabolites and regulatory proteins. Such interactions can be interpreted as
the consequence of the movement of regulatory molecules from one
chromosomal region to another. Other controls exhibit complexity, raising
the possibility that one region of the chromosome may influence regulation
exercised in another. Direct physical interaction between different regions
of the chromosome is indicated by observed repair of a mutation in the peal
gene, a process that depends upon the presence of the unlinked catl
sequence (Doten et ai., 1987b).
Physiological Interactions
Mutant strains that constitutively express B-ketoadipate-

:succinyl eoA transferase. Among the first regulatory mutant strains
isolated from Acinetobacter were organisms that had lost the ability to
form protocatechuate 3,4-dioxygenase (Canovas and Stanier, 1967). These
strains were selected with B-ketoadipate, a growth substrate that demands
constitutive synthesis of B-ketoadipate:succinyl CoA transferase and B-
ketoadipyl CoA thiolase. Further analysis revealed that loss of the
oxygenase was accompanied by constitutive expression of the other pea
genes, including those that encode the transferase and thiolase activities
(Canovas et ai., 1968a). The apparent cause of constitutivity was metabolic
formation of the inducer-metabolite protocatechuate from endogenously
synthesised shikimate (Fig. 6).
3-Hydroxy-4-methylbenzoate, a chemical analogue of

protocatechuate, acts as an anti-inducer and prevents constitutive
expression of the remaining functional pea genes in mutant organisms
lacking protocatechuate-oxygenase (Canovas et aI., 1968b). When these
strains are exposed to B-ketoadipate in the presence of the protocatechuate
analogue, secondary mutant strains with altered regulation can be selected.
Mutations in these strains change the specificity of induction so that the
analogue, rather than the natural substrate protocatechuate, acts as inducer
of the pea enzymes. The gene associated with altered inducer-specificity is
yet to be mapped.
Selection for growth of Acinetobacter with B-ketoadipate yields

additional strains in which cat genes are expressed constitutively. Some of
these organisms carry mutations in the repressor gene catM (Neidle et aI.,
1989). Others among the mutants constitutively express the catCIIFD
212
BIOSYNTHESIS DISSIMILATION
AROMATIC AMINO ACIDS SHIKIMATE
Induced by
PROTOCATECIlUATE
SHIKIMATE I
----------------------> PROTOCATECIlUATE
Induced by
I
PROTOCATECHUATE ==+== Blocked by Mutation
! in ~ or I2£2!i
!
Induced by
PROTOCATECHUATE
COMMON METABOLITES COMMON METABOLITES

Fig. 6. The dual role of shikimate as a biosynthetic intermediate and as a
growth substrate. Shikimate, formed as an intermediate in the biosynthesis
of aromatic amino acids, can serve as a growth substrate for Acinetobacter.
Protocatechuate is an intermediate in the dissimilation of shikimate and
induces synthesis of all of the enzymes associated with this metabolic
pathway. Mutations in pcaG or pcaH block synthesis of protocatechuate
oxygenase and result in the metabolic accumulation of protocatechuate.
Such mutants constitutively form other enzymes associated with shikimate
catabolism. Apparently, catabolism of shikimate formed biosynthetically in
the mutant strains causes accumulation of endogenous inducer in quantities
sufficient to trigger synthesis of the full set of catabolic enzymes.
region, but retain inducible control over catA and eatB (Canovas and
Johnson, 1968; Patel et aI., 1975). An interpretation of this finding is that
these mutant strains have acquired a promoter within or near eatB. This
proposal has not been explored genetically.
Demand for growth of Aeinetobaeter strains with 13-ketoadipate

yields one set of mutant organisms that are constitutive for neither pea nor
cat genes. Biochemical investigation revealed that such strains
constitutively form a 13-ketoadipate:succinyl CoA transferase that exhibits
relatively low affinity for 13-ketoadipate and high affinity for adipate as a
competitive inhibitor (Canovas and Johnson, 1968). Evidently this
transferase is associated with adipate metabolism, yet contains substrate
specificity sufficiently broad to allow growth with 13-ketoadipate. The
properties of these strains suggest that they constitutively form a 13-
ketoadipyl CoA thiolase that is not encoded by pcaF or by catF. The
213
existence of such a thiolase has been confirmed by demonstration that
growth with adipate is unimpeded in Acinetobacter strains from which the
pcaF and catF genes have been deleted (L. Gregg, unpublished
observations).
Regulation of the expression of the ben genes. The benABC

region contains genes for the three proteins that catalyse the reductive
oxygenation of benzoate to the corresponding 1,2 dihydrodiol (Fig. 1). This
metabolite is converted to catechol by a dehydrogenase encoded by benD
(Fig. 1). DNA sequence evidence (E.Neidle and C.Hartnett, unpublished
observations) suggests that the three ben genes are expressed as a single
transcript that is likely to include ORFI (Fig. 5). The terminus of benC is
separated by only 18 bp from the translational origin of benD, and there is
no discernible sequence evidence for separation of transcriptional control of
benD from benABC. The activity of the NAD-dependent dehydrogenase
encoded by b enD is relatively easy to determine, and measurement of this
activity reveals complexity of transcriptional control.
A mutation within benABC prevents metabolism of benzoate, but

does not prevent this metabolite from inducing full levels of benD
expression in Acinetobacter (Neidle et al., 1987). Thus benzoate is
identified as an inducer of the putative benABCD operon. Benzoate
dihydrodiol, the product of the benABC-encoded enzyme, does not support
growth of the wild-type cells. This may be because the diol does not induce
benD expression, but the possibility that the compound does not permeate
the cell membrane at a rate sufficient to support growth has not been
excluded.
Physiological complexity arises from the fact that growth with

muconate, a compound three metabolic steps removed from benzoate (Fig.
1), elicits expression of benD but not of benC (Neidle et al., 1987).
Furthermore, muconate-grown cells do not oxidise benzoate (Canovas and
Stanier, 1967). Thus induction with muconate separates the physiological
expression of benABC from expression of benD. The possibility that
benzoate and muconate induce separate enzymes with the benD-encoded
activity is eliminated by the fact that mutations within benD prevent its
expression in response to either benzoate or muconate (Neidle et al., 1987).
Physiological separation of the benABC and benD encoded activities

does not require that the genes be transcribed separately. It is possible that
control is exercised by inactivation of either the benABC transcript or the
protein gene products. This possibility is heightened by the fact that
mutations within benD lead to minimal expression of the oxygenative
function associated with the upstream benABC genes (Neidle et aI., 1987).
These genes express a reductive benzoate oxygenase at high levels when
placed under a lac promoter in E. coli. Thus elements within the genetic
background of Acinetobacter shut down benABC expression in the
absence of metabolism of the benzoate diol, the end-product of their
encoded enzymes and the substrate of the enzyme encoded by the possibly
214
cotranscribed benD gene. Similar observations have been reported with
genes associated with benzoate oxidation in Pseudomonas (Harayama et
aI., 1984).
The fact that benD is expressed in response to muconate raises the

possibility that CatM, a repressor protein that responds to muconate,
contributes to control of this structural gene. This possibility is rendered
unlikely by the fact that benD remains under inducible control in mutants
carrying a dysfunctional catM gene. Such mutants constitutively express
enzymes encoded by catA and by the catBCIJFD operon (Neidle et aI.,
1989).
Regulation of the expression of catA. Several genes contribute to

regulation of the expression of catA. This structural gene has a regulatory
role in its own right because the oxygenase it encodes gives rise to muconate,
the inducer of all of the cat genes in Acinetobacter (Canovas and Stanier,
1967). These genes are expressed when interaction of muconate with CatM
in wild-type cells relieves repression. The wild-type catM gene appears to
be dominant over a mutated variant because repression can be exerted from
a plasmid-borne catM gene in a strain in which the chromosomal catM gene
is dysfunctional. Selection for constitutive expression of the cat structural
genes in such a strain selects for loss of the plasmid (Neidle et at, 1989).
Inferences concerning the regulation of catA can be drawn from

examination of the influence of flanking regions of Acinetobacter DNA
upon expression of the gene in E. coli (Neidle and Ornston, 1986). As
shown in Table 2, catA can be expressed inducibly in E. coli from the lac
promoter in pUC19. In the absence of other Acinetobacter genes, levels of
catA expression are relatively low (Table 2 - pIB1345). Presence of the
Acinetobacter open reading frame ORF2 downstream from catA increases
its uninduced expression ten-fold and its IPTG-induced expression 20-fold
(Table 2 - pIB1343). Expression of catA was lowered somewhat by inclusion
of ORF3 within the insert (Table 2 - pIB1341) and almost eliminated by the
presence of upstream DNA including ORF1 (Table 2 - pIB1362).
One interpretation is that ORF2, coexpressed with catA when the

neighbouring genes undergo in-vitro transcription-translation (N eidle et
aI., 1989), contributes to the expression of catA. This contribution
cannot be essential for gene expression because a deletion introduced
into ORF2 does not discernibly impede the growth of Acinetobacter with
benzoate, anthranilate or muconate (Neidle et aI., 1989). Similar
observations were obtained with Acinetobacter strains carrying insertions
within ORF3. The possibility that ORF2 and ORF3 may assist in expression
of catA was strengthened by the observation that a mutation in ORF2
resulted in catechol accumulation during growth with benzoate, and
similar accumulation was observed when a strain carrying a mutation in
ORF3 was grown with anthranilate (Neidle et aI., 1989). It is possible
that unexplored physiological conditions, such as nitrogen limitation,
might reveal more dramatically regulation exerted by ORF2 and ORF3.
215
Table 2. Expression in E.coli of catA carried in Acinetobacter DNA
fragments inserted downstream from a pUC19 lac promoter
Recombinant IPTG IPTG Presence (+) or Absence ( - ) of

Plasmid Absent a Presenta ORFI catA ORF2 ORF3
pIB1345 0.05 0.11 +
pIB1343 0.48 2.0 + +
pIB1341 0.16 0.8 + + +
pIB1362 0.004 0.02 + + + +
athe specific activity of the catA gene product was measured as units of
catechol 1,2-dioxygenase activity I mg protein and is presented as the
ratio of observed activity to the activity found in extracts of benzoate-
grown Acinetobacter cultures. All of the recombinant plasmids carried
DNA inserts containing catA in the multiple cloning site of pUC19 and
orientated so that its direction of transcription lies downstream from
the lac promoter. The Acinetobacter DNA inserts in pIB1345, pIB1343 and
pIB1341 begin at the same site upstream from catA. The inserts in
pIB1341 and pIB1362 end at the same site downstream from ORF3 (Neidle and
Ornston, 1986; Neidle et al., 1989).
The lowering of catA expression in the presence of upstream DNA

extending as far as ORFI (Table 2 - pIB1362) could be attributed to the
presence of a possible transcriptional terminator (ATTCGCTTTA TTCT AA-
GGCGAATTTTTCTTTTTT) 40 bp downstream from this open reading
frame. Transcriptional termination upstream from catA would demand the
presence of promoters upstream from the gene, and the presence of such
promoters can be inferred from the response of DNA fragments, including
catA, to inducers when carried on a broad host range plasmid in
Acinetobacter strains from which the entire ben-cat region has been
deleted (Neidle et ai., 1989). When placed under control of the lac
promoter of pRK415, an Acinetobacter DNA segment containing catA and
50 bp upstream from the gene expressed the gene at levels corresponding to
about 10% of the specific activity observed in fully induced wild-type cells;
expression of the gene from this plasmid, probably in response to the lac
promoter of pRK415, was not influenced by the presence of benzoate or
muconate in the growth medium. Derivatives of pRK415 containing catA
216
and 1 kb or more of Acinetobacter DNA upstream from this gene expressed
its product at relatively high levels corresponding to about 50% of the
specific activity in fully induced cells. This level of gene expression,
independent of the orientation of the insert with respect to the lac
promoter, was elevated six-fold when either benzoate or muconate was
included in the growth medium. Thus the 1 kb of DNA upstream from catA,
a region that does not contain a complete open reading frame, appears to
contain a promoter that recognises one or more regulatory molecules that
respond to either benzoate or muconate as an inducer (Neidle and Ornston,
1987). The regulatory molecules must be encoded by DNA lying outside the
16 kb of DNA which, containing known ben and cat genes, is lacking in the
Acinetobacter deletion mutant that acted as host for the recombinant
plasmids (Neidle et al., 1989).
Induction of plasmid-borne catA in the deletion mutant can take

place in the absence of catM, the known repressor of the structural gene,
and in the absence of ORF2, a gene that was implicated in control of the
catA gene when it was cloned in E. coli (Neidle et aI., 1989). The general
pattern that emerges from these observations is that catA is subject to
overlays of transcriptional control to which a number of regulatory genes
contribute~ It is possible that genetic interactions associated with these
control mechanisms contribute to conservation of the supraoperonic
clustering of ben-cat genes. This possibility can be explored by examining
the physiological consequences of transposition of catA and neighbouring
DNA fragments from their present locus to other loci in the Acinetobacter
chromosome.
Expression of Acinetobacter genes in heterologous hosts.

Differences in codon usage pattern (Table 1) present no barrier to
expression of Acinetobacter genes in Pseudomonas. The
Acinetobacter catBCll FD operon, contained as a 5.0 kb EcoRl insert in
the broad host range plasmid pKT230, was selected initially on the basis of
its ability to complement a pcall (pcaE) mutation in P. putida (Shanley et
aI., 1986). The recombinant plasmid also allowed growth of complemented
P. putida strains carrying mutations in catB, catC or pcaD. In every case,
the complementing Acinetob acter gene was expressed at levels comparable
to those found in fully induced wild-type cells.
When placed under control of a lac promoter in pUC13 and induced

by growth of host E. coli cells in the presence of IPTG, the products of the
catBCIJFD genes represent more than 10% of the cell protein (Shanley et
aI., 1986). Even higher levels of expression, with Acinetobacter enzymes
accounting for about half of the soluble protein in E. coli, are produced
when the Acinetobacter pcaJ.lFDBCHG operon under the control of a
lac promoter is expressed in response to IPTG (Doten et aI., 1987a). In
these cells, levels of most of the protocatechuate-metabolising enzymes are
ten to 30-fold higher than those found in fully-induced Acinetobacter
cultures. The sole exception to this pattern is protocatechuate oxygenase
which, at a three-fold higher level than in induced Acinetobacter cultures,
217
represents about 10% of the E. coli soluble protein. The oxygenase purified
from E. coli contains only 20% of the specific activity of the same gene
product purified from Acinetobacter. It is possible that E. coli lacks
factors that are required for effective assembly of the oxygenase when it is
expressed at extremely high levels. Perhaps overproduction of the oxygenase
demands that iron be supplied at a rate that exceeds the limit which can be
supplied physiologically.
The pea genes encode all of the enzymes required for metabolism of
protocatechuate to citric acid cycle intermediates, and E. coli cultures in
which the enzymes are expressed might be expected to grow at the expense of
protocatechuate (Doten et a!., 1987a). Growth of such cultures is not
observed. Rather, the cultures rapidly and quantitatively convert
protocatechuate to B-ketoadipate. Evidently B-ketoadipate succinyl CoA
transferase, expressed at high levels in the E. coli cultures, is physiologically
ineffective in these cells. A possible reason for inactivity of the expressed
enzyme is the relatively low level of succinyl CoA, the required thioester
donor for B-ketoadipate metabolism, in E. coli (Jackowski and Rock, 1986).
As described above, the benABC genes can be expressed effectively in

E. coli cells in which the encoded enzymes efficiently convert benzoate to
the corresponding dihydrodiol. Such conversion is not observed when the
benABC genes are expressed in Acinetobacter cells carrying a
dysfunctional benD gene (Neidle et aI., 1987). This evidence suggests the
presence in Acinetobacter of regulatory mechanisms that govern the rate
of conversion of benzoate to the dihydrodiol. Thus examination of
Acinetobacter gene regulation in E. coli can give an insight that might not
be gained by examination of Acinetobacter alone. A further example is the
demonstration that ORF2 contributes to the expression of catA (Table 2),
an observation that might have been obscured by knowledge that mutations
within ORF2 do not prevent growth of Acinetobacter with substrates that
demand catA expression for their metabolism.
Genetic Interaction
Participation of the catIl region in repair of a mutation in

the pcall region. In Acinetobacter strains containing the wild type
catIJ genes, the peaJ3125 mutation is repaired at a frequency of 10- 4 •
Deletion of a region containing the catI} region from the chromosome
causes this frequency to drop 300-fold to 3 x 10 7 • Thus it appears that the
cati J region contributes DNA sequences that repair the pcal3125 lesion
(Doten et aI., 1987b). The repair process depends upon a functional reeA
gene and is not inhibited by deoxyribonuclease at concentrations that
prevent natural transformation of the mutation to wild type by exogenous
DNA (L. Gregg and D. Elsemore, unpublished observations). Therefore it
seems likely that the repair process is mediated by direct intrachromosomal
DNA sequence transfer from the eatll region into the peal} region. The
linear chromosomal distance between these two regions is not known, but it
appears to exceed 15 kb. Therefore folding of DNA must bring the regions
218
into close physical proximity so that the sequence exchange may take place.
DNA sequence exchange as a mechanismfor conservation of

exceptional DNA sequences in catlIF and pcalIF. Repair of
mutations in the pcaI.T region by cat 1J sequences may serve as a mechanism
for gene stabilisation: genetic variation can be counterbalanced by
substitution of the altered DNA with a sequence copy carried elsewhere on
the chromosome. The effectiveness of such a genetic process may be
assessed by the degree to which it has overcome selective pressures for genes
to take on the GC content and codon usage patterns that are characteristic of
the host chromosome (Grantham et al., 1981; Ikemura, 1982; Muto et al.,
1984; Ohama et a1., 1987). The existence of such pressures is evident from
the closeness with which these acquired traits are shared by most genes
associated with catabolism of aromatic compounds in Acinetobacter. As
summarised in Table 3, most genes for aromatic catabolism, derived from
separate ancestors, vary only slightly in GC content and in codon usage
pattern. These genes, the reference genes in Table 3, have no close
homologues in the Acinetobacter chromosome, but share similar GC
contents of 44±2% and, with rare exception, exhibit codon usage patterns
that are typical for Acinetobacter genes.
Genes in the cat 1.T F and pc a I 1 F regions, unlike other

Acinetobacter genes, contain DNA sequences that are nearly identical:
comparison of 1.2 kb encoding catIJ and pcaI.T revealed only six base pair
differences in DNA sequence (A. Benson and J. Houghton, unpublished
observations). It seems likely that this close sequence similarity underlies
the capacity of the genes to exchange sequence information, as revealed by
catl-mediated repair of the pcal3125 mutation. Most remarkable is the
set of genetic properties shared by catI, catl and catF. As shown in Table
3, these genes share closely similar GC contents ranging from 54% to 57%,
values more than 10% higher than the average of those shared by all other
known Acinetobacter genes. Furthermore, the catI, catl and catP genes
share similar patterns of codon usage that distinguish them from the other
Acinetobacter genes Cfable 3). For example, the codon CTG accounts for
roughly one sixth of the leucyl codons in the reference genes and represents
half or more of the leucyl codons in catl, catl and catF (Table 3).
The presence in catI, cat.T and catP of highly distinctive properties,

traits shared with the nearly identical pca1, pc al and pc a P genes, is open to
three possible interpretations. One proposal is that both sets of genes are
recent evolutionary acquisitions that have not undergone the selective
pressures that conferred characteristic GC contents and codon usage
patterns upon the other Acinetobacter genes. This interpretation seems
somewhat unlikely because it is difficult to account for evolution of genes
catalysing most reactions in the £-ketoadipate pathway (the reference genes
in Table 3) in the absence of genes for the enzymes that most directly yield
the selective benefit of the pathway. The suggestion that the nearly identical
catI.TP and pcaIJP regions replaced Acinetobacter genes relatively
recently in evolutionary history requires the assumption that highly
219
Table 3. GC content and codon usage frequency in A. calcoaceticus genes
reference catI catJ catF
GC content 44+2 57 55 54
TTA (Leu) 24+8 5 10 7

TTG (Leu) 23+7 o o 7
CTG (Leu) 17+10 55 65 50
ATT (He) 68+10 22 29 34
ATC (He) 29+12 67 64 62
TCC (Ser) 03+04 33 29 6
,CCT (Pro) 31+17 8 8 6
CCG (Pro) 20+16 38 50 61
ACC (Thr) 35+14 69 69 52
GCT (Ala) 20+7 4 5 5
GCC (Ala) 21+10 56 59 24
TAC (Tyr) 18+14 71 33 56
CAG (GIn) 39+15 50 57 80
AAC (Asn) 31+20 73 60 53
GAC (Asp) 28+8 82 53 45
CGT (Arg) 69+14 43 20 38
CGC (Arg) 18+9 43 60 48
AGC (Ser) 15+10 33 29 31
GGC (G1y) 31+17 48 46 49
All values are given as percentages. The reference values are the
average observed with the structural genes pcaC, pcaG, pcaH, benA, benB,
benC, benD, catA, catB, catC and catD. These values do not differ
significantly from those found with biosynthetic structural genes from A.
calcoaceticus (Kaplan et a1., 1984).
improbable events happened twice. A second interpretation is that some

intrinsic property of CoA transferase and thiolase genes demands
conservation of properties at the level of DNA sequence. This interpretation
is rendered unlikely by the fact that widely divergent thiolase genes from
other organisms have been sequenced (Peoples and Sinskey, 1989).
Furthermore, the CoA transferase and thiolase genes that are associated
with adipate metabolism in Acinetobaeter have diverged significantly, as
evidenced by their failure to hybridise with DNA containing the eatIJ F or
the peaIJ F genes (Shanley et ai., 1986; Doten et ai., 1987a). The third
interpretation, and in our view the most likely, is that the documented
capability for sequence exchange between the cat! J F and the peal J F
regions has allowed repair to buffer them against events that conferred
characteristics typical of the Acinetobacter chromosome upon other genes.
220
The proposal that sequence exchange between cat /.T F and pc a I.T F
contributes to their conservation can be explored by examination of the
corresponding regions from divergent Acinetobacter strains. A major
contribution of sequence exchange to conservation would be reflected in
concerted divergence so that the two genetic regions retained their sequence
similarity as they diverged in different organisms. The natural
transformation system of Acinetobacter should allow replacement of either
catI.TF or pcaI.TF with a divergent region from another strain; genetic
interplay between the regions could be monitored in the recombinant
organism.
A major question to be addressed is the basis for the genetic stability

of the nearly identical catIJ F and pca/.T F regions in the Acinetobacter
chromosome. Recombinational events lead to rapid excision of most
repeated DNA sequences from the prokaryotic chromosome (Horuichi et al.,
1962; Jackson and Yanofsky, 1973; Rigby et al., 1974; Anderson and Roth,
1977). Notable exceptions are ribosomal genes (Nomura et al., 1984;
Widom et al., 1988) and now, catIJ F and pca/.T F. It is possible that the
chromosomal location of these two regions allows their conservation in the
Acinetobacter chromosome. This possibility can be explored by selection
for recombinants in which the genes have been rearranged. For example, the
catBCIJ FD operon, completely deleted from the chromosome of strain
ADP161, is carried on a 5.0 kb EcoRI fragment (Shanley et al., 1986).
Insertion of this fragment in the correct orientation into pcaH G, the genes
encoding protocatechuate oxygenase, would inactivate the oxygenase and
result in constitutive expression of the inserted genes (Hartnett et al., 1990).
A recombinant that had acquired this construct could be selected by
demanding growth of strain ADP161 with muconate after transformation
with engineered DNA containing eatBC IJ F D within peaH G. The
recombinant would carry the nearly identical cat IJ F and peaIJ F DNA
segments within 3 kb of each other, and the genetic stability of this
arrangement could be examined by monitoring the ability of the
recombinant's progeny to grow with muconate.
DNA SEQUENCE EXCHANGE AS A GENETIC

MECHANISM FOR EVOLUTIONARY DIVERGENCE
In contrast to the remarkable conservation of cat 1J F and peal J F

sequences, substantial sequence variation accompanied divergence of the
isofunctional catD and pcaD genes in Aeinetobaeter (Yeh et al., 1980a).
The DNA sequence of the pc aD gene has not yet been determined, but
divergence of the NH 2-terminal amino acid sequences of the enol-lactone
hydrolases encoded by Acinetobaeter catD and pc aD is so extreme that
inclusion of the amino acid sequence encoded by the Pseudomonas peaD
gene was required to provide evidence for common ancestry of all of the
hydrolases (McCorkle et al., 1(80). Thus, somewhere before the terminal
gene in cat 1 J F D and peal J F D, the evolutionary pattern shifts from
extreme conservation to extensive divergence.
221
The NH 2 -terminal amino acid sequences of the enol-lactone
hydrolases show patterns of internal repetition (McCorkle et aI., 1980) and
also give evidence of sequence exchange between their structural genes and
catC during their evolutionary divergence (Yeh et al., 1978; Yeh and
Ornston, 1980). This evidence fostered the hypothesis that gene conversion,
the substitution of a DNA sequence from one chromosomal region into
another, may have produced extensive genetic variation during evolutionary
divergence (Ornston and Yeh, 1979, 1982). DNA sequences for some of the
relevant genes are now available, and they support the conclusion that such
sequence exchange played a significant role in the acquisition of DNA
sequences that distinguish homologous genes for the B-ketoadipate pathway
(Neidle et al., 1988; Hartnett et aI., 1990). The evolved genes seem to
possess the potential ability to form sets of slippage structures created by
hybridisation between misaligned DNA strands. An inference from this
finding is that the slippage structures, formed during evolutionary
divergence, may make a contribution to maintenance of the evolved DNA
sequence through mismatch repair between slipped DNA strands. In such
events, complementary components of repeated DNA sequences arc
envisaged as hybrid ising in slippage structures and exchanging sequence
information so that the wild type repetitions are maintained (Ornston et al.,
~
Q--£
1- -a
Q- -£
Q--£
£--Q
Q--£
110 C·-G 140 150 160 170 180
3' GCGGACGTTGCACTTCCCTGAAC- -TAACGGCCCGTGATGCGTTTGATGTCGCACAAGCTACACGGGTCGCAG-AAA--,
11111111 111I11 11111111 I
11111111 111111 11111111
10 11111111 20
11111111
30 111111 40
11111 I 50 60
11111111
11111111
5' ATGCTGTTCCACGTGAAGATGACCGTGAAACTGCCGGTGGACATGGACCCGGCCAAGGCT-AAA-TGGCCCAGCGCCTGC--
I I
::: II: I
III I
:: : ::::: II: I:
11111111 III 11111111
:::::::: 100 II
111111
1111I1
3' GTTGACGGCAACAAGGGC==no==1
111111111111
111111111111
11111111
11111111
=gap=====ATGTACCTGTAGCTCCAGCTGCCGGACACGGCCGT-AAA
I
1I
I
210 220 I!!!!:!!!II !
111111111111
230 240 250 260 I
111111111111
111111111111 I
3' TTATACTCCATGCGGGCTACTGGCACTTTTAAAGGGTGTGACGGCTGTAAGTTCGGAAGAAGTTATTTCACCG
1320 330 340 350 360 370 380 390 I
Fig. 7. An apparent slippage structure formed by the P.putida catC

sequence. The beginning of the 5' strand of the structural gene shows
the potential ability to enter into hybridisation in three different
alignments with the slipped 3' strand. The 5' strand is shown in bold,
and the 3' strand is underlined. Vertical lines indicate >,here six or
more contiguous residues in each strand exhibit perfect complementarity
and may hybridise.
222
1990a,b). According to this view, mutations causing nucleotide substitutions
within one component of a repetition could be eliminated by mismatch
repair. Thus, as suggested for catIJF and pea/IF, events taking place at
the level of DNA metabolism could account for the maintenance of sequence
information.
The most clearly elucidated slippage structures are those associated

with the P. putida catC gene (Aldrich et al., 1987; Aldrich and
Chakrabarty, 1988; J. Houghton, unpublished observations). As illustrated
in Fig. 7, the beginning of the 5' strand of this gene appears to have the
capacity to hybridise to three different regions of the 3' strand. Thus the
slippage structure can be envisaged as a three-dimensional structure in
which the 3' strand coils around so that different portions of its sequence can
achieve hybridisation with the 5' strand. The central role of the 5' strand in
the structure allows its other components to be discerned relatively directly.
Other slippage structures are less apparent, and are reflected largely in the
presence of different patterns of sequence repetition that were acquired as
genes diverged.
Some sets of sequence repetitions lie close to each other in positions

that might facilitate ratcheting of DNA out of its conventional duplex
structure so that single strands may hybridise in slippage structures
elsewhere. Examples, shown in Fig. 8, are segments of catC encoding the
middle of the primary sequence of muconolactone isomerase from
Aeinetobacter and Pseudomonas. As shown in Fig. 8A, the amino acid
sequences of the isomerases are quite similar, especially in light of
substantial divergence in the GC content of the structural genes (Table 1).
Clusters of similar DNA sequence remain in the divergent genes (Fig. 8B),
but they have acquired quite different patterns of sequence repetition (Fig.
8C).
Additional examples of genetic events accompanying evolutionary

divergence emerge from comparison of genes encoding intradiol
dioxygenases. The crystal structure of Pseudomonas protocatechuate 3,4-
dioxygenase has been solved (Ohlendorf et al., 1988). The protein is formed
by association of equal quantities of a and 13 subunits, the respective products
of the peaG and peaR genes. The 13 subunit binds ferric ion and catalyses
the oxygenative reaction. The a subunit is homologous with the 13 subunit
and the two subunits bind to each other to form a complementary array in the
enzyme. An arginyl side chain in the a subunit contributes to binding of the
negatively-charged carboxylate of protocatechuate in the active site of the
enzyme. The Acinetobacter pcaRG region has been sequenced (Hartnett
et al., 1990), and the encoded proteins share amino acid sequence identity of
about 50% with their homologues from Pseudomonas. Residues indicated
by crystallography to be essential have been conserved, and comparison
of the peaR gene with Acinetobacter catA (the catechol 1,2-
oxygenasestructural gene) and with the Pseudomonas cleA (the
chlorocatechol 1,2-dioxygenase structural gene) gives some insight into
events accompanying their evolutionary divergence. For example, the
223
A. )
Acinetobacter Isomerase
40 50 60
LeuGlnArgGlnGlyLysTrpArgHislleTrpArgIleThrGlyGlnTyrSerAsnIleSerIlePheAspValGluSer
*** * *** *** *' * * * *** *'
LeuGlnArgGluGlyThrTrpArgHisLeuTrpArglleAlaGlyHisTyrAlaAsnTyrSerValPheAspValProSer
~ ~ ~
Pseudomonas Isomerase
B. )
Acinetobacter catC
lW 120 130 140 150 160 170 180
TTGCAACGCCAAGGCAAATGGCGTCATATCTGGCGTATTACCGGACAATATTCCAATATCAGTATTTTTGATGTCGAAAGT
IIIIII IIII IIIII IIIII IIII IIIII
IIIIII IIII IIIII IIIII IIII IIIII
CTGCAACGTGAAGGGACTTGGCGCCACCTGTGGCGCATTGCCGGGCACTACGCAAACTACAGCGTGTTCGATGTGCCCAGC
110 120 130 140 150 160 170 180
Pseudomonas catC
C. )
Acinetobacter catC
lW 120 ---li0 140 150 160
5' TTGCAACGCCAAGGCAAATGGCGTCATATCTGGCGTATTACCGGA~TC~CAGTATTTTTGATGTCGAAAGT
r·---> <---., <_ .. _, r·-·> <._-,
1 l' l' 2 2
3 3 4 4
L•• _._> < ___ ._J L._ •• > <_. __ J
5' CTGCAACGTGAAGGGACTTGGCGCCACCTGTGGCGCATTGCCGGGCACTACGCAAACTACAGCGTGTTCGATGTGCCCAGC
110 'i'2O 13-0-- 140 """"i'5o' ---;60 170 180
Pseudomonas ~
Fig. 8. Sequence conservation and divergence within catC, the structural

gene for muconolactone isomerase from Acinetobacter and Pseudomonas. (A)
Identical amino acids in the aligned sequences are marked with an
asterisk. (B) Conserved regions of nucleotide sequence are marked with
vertical lines where four or more contiguous bases in the 5' strand are
identical. (C) Different patterns of repetition were acquired as the
catC gene diverged in Acinetobacter and Pseudomonas. Direct repetitions
of five or more contiguous base pairs are marked by single or double
underlining or overlining. Direct slippage of 5' and 3' strands followed
by hybridisation between the complementary components of sequence
repetitions could help to melt DNA into single strandedness so that
slippage structures such as that depicted in Fig. 7 could be formed.
Inverted repetitions, marked with numbered arrows, could facilitate the
localised melting of DNA by forming loops within strands.
catechol oxygenases are formed by self-association of a single subunit, the

evolutionary counterpart of the B subunit of protocatechuate oxygenase
(Frantz and Chakrabarty, 1987; Neidle et al., 1988; Ngai and Ornston, 1988).
Thus divergence of protocatechuate and catechol oxygenase was
224
Substrate m: His !!i!
Protocatechuate ProGlyProlEProTrpArgAsnArglleAsnGlu---------------------------------TrpArgProAlaHislleJ!i!PheSerLeulleAlaAspGly
* * * * * * * * * * *
Catechol ProAlaGlylEGlyCysProProGluGlyProThrGlnGlnLeuLeuAsnGlnLeuGlyArgHisGlyASnArgProAla!!i!lle!!i!TyrPheValSerAlaAspGly
* * * * * * * * * * * * * * * * * * *
Chlorocatechol ProValPr0!l-rGlnlleProTyrGlyGlyProThrGlyArgLeuLeuGlyHisLeuGlySerHisThrTrpArgProAlaHisValHisPheLysValArgLysAspGly
Fig.9. Conserved amino acid residues in the iron-binding region of intradiol dioxygenases. Highlighted are two
histidyl residues that bind ferric ion within the enzymes. The amino acid sequence of the Bsubunit of Acineto-
bact:er protocatechuate oxygenase is aligned with the corresponding sequences of Acinetobacter catechol oxygenase
and Pseudomonas chlorocatechol oxygenase. Asterisks mark identical residues shared by the catechol oxygenase
sequence and either of the other sequences.
"-l
"-l
(1l
'"'"OJ
A
210 220 His 4His 26 230

Catechol oxygenase GlyProThrGlnGlnLeuLeuAsnGlnLeuGlyArgHisGlyAsnArgProAlaHis¥~eHis~yrphevalSerAlaAspGlY
* * * * * * * * * * * * * * * *
Chlorocatechol Oxygenase GlyProThrGlyArgLeuLeuGlyHisLeuGlySerHisThrTrpArgProAlaHisValHisPheLysValArgLysAspGly
170 180 His 188His 190
620 630 640 650 660 670 680 690

Acinetobacter catA 5' GGTCCAACGCAACAGTTGCTGAATCAGTTGGGCCGTCATGGTAACCGCCCTGCGCACATTCACTATTTTGTTTCTGCCGATGGA 3'
IIIII IIIII IIIIII IIII
IIIII IIIII IIIIII IIII
Pseudomonas clcA 5' GGGCCGACTGGGCGTCTGCTGGGCCACCTGGGCAGCCATACCTGGCGTCCGGCGCACGTGCACTTCAAGGTGCGCAAGGACGGT 3'
510 520 530 540 550 560 570 580 590
c
680 690
620 630 640 650 660 670 CACTATTTTGTTTCTGCCGATG
Acinetobacter catA 5' GGTCCAACGCAACAGTTGCTGAATCAGTTGGGCCGTCATGGT------AACCGCCCTGCGCACATT G
-- -- IIIIII IIII IIIIII
IIIIII IIII IIIIII A
Acinetobacter catA:same strand 3' CCCAACGTATTCGTTTGCAGTATCCACATGCCTAGCGG--TCGGTGTAA C
760 750 740 730 TTAAACGCATCAATCAAACGCCA
7~ 710 --700
D
< _____ 380
510 520 530 540 550 560 570

Pseudomonas clcA 5' GGGCCGACTGGGCGTCTGCTGGGCCAC--CTGGGCAGCCATACCTGGCGTCCGGCGCACGTGCAC 3'
IIII IIIIII IIIIII IIIIII
IIII IIIIII IIIIII IIIIII
Pseudomonas clcA:opposite strand 3' CCCGGCT--GACCCGCAGACGACCCG-----GTGGACCCGTCGGTATGGACCGCAGGCCGCGTGCACGTG 5'
510 520 530 540 550 560 570
<------124
Fig. 10. Different DNA slippage structures in the divergent Acilletobacter catA and Pseudomonas clcA genes.
Numerals above or below residues indicate their position in pr1mary sequence. (A) A portion of the iron-
binding region in the two proteins. Asterisks mark identical residues occupying the same position in the
aligned sequences. (B) DNA sequences corresponding to the aligned amino acid sequences. Vertical bars
indicate positions in which four or more contiguous DNA bases are identical in the aligned sequences.
(C) A potential loop formed by the 5' Acinetobacter catA DNA strand corresponding to the compared region.
Direct sequence repetitions that could serve as a basis for alignment with the slipped 3' strand are
under lined. CD) A potential slippage structure formed between misaligned 5' and 3' Pseudomonas c1 cA DNA
strands corresponding to the compared region. Arrows indicate DNA that is pinched into single strandedness
in this slippage structure and which has the potential to hybridise with DNA in the same strand at positions
indicated by the numerals.
""
"""
accompanied by gain or loss of the c> subunit that specifically binds the
protocatechuate carboxylate, a functional group that is absent in the
catechols.
Amino acid sequences associated with three of the four iron-binding

ligands in the oxygenases are aligned in Fig. 9. It is evident that the two
catechol oxygenases resemble each other more closely than they resemble
protocatechuate oxygenase. Particularly noticeable is the gap required to
align the latter sequence with the former sequences. The gap, lying near a
loop in the protocatechuate oxygenase structure (Ohlendorf et aI., 1988), is
filled with an 11 amino residue amino acid sequence that, judging by its
conservation, appears to serve a significant function in the catechol
oxygenases. Conservation of the amino acid sequences masks considerable
divergence at the level of DNA. As shown in Fig. 10, Acinetobaeter eatA
and Pseudomonas cleA acquired entirely different slippage structures
during their evolution (Neidle et aI., 1988; Ornston et aI., 1990b). A portion
of the aligned amino acid sequences of the two genes is depicted in Fig. lOA.
Fig lOB indicates the remaining DNA sequence homology between the 5'
strands encoding the amino acid sequences. As shown in Fig. 10C, the
Acinetobaeter eatA sequence has diverged to form a slippage structure in
which an intrastrand loop is the predominant component, whereas
hybridisation between the 5' and 3' strands forms the major portion of the
slippage structure in the Pseudomonas cleA sequence.
It is difficult to account for how the distinctive slippage structures of

divergent genes could have been formed without hybridisation between their
complementary strands. It is likely that mismatch repair (Kunkel and Soni,
1988; Meselson, 1988) helped to create the repeated sequences, and a similar
process may contribute to maintaining the repetitions in the evolved
sequences. One avenue to exploring the process would be to disrupt slippage
structures by site-directed mutagenesis (Kunkel, 1985) and to examine the
effects of sequence repetitions on repair. A more general approach,
probably more warranted at our present stage of enlightenment about
slippage structures, would be to examine DNA sequences that undergo
spontaneous mutation and repair. This approach would take into account
the fact that, as exemplified by repair events mediated between cat.! and
pea.!, genetic interactions may involve DNA sequences that are separated by
a substantial linear distance on the chromosome.
It is possible that sequence interactions between separate genes may

be a selective force accounting for supraoperonic clustering of genes for the
B-ketoadipate pathway. If so, gene rearrangements, readily created in the
naturally transformable strain of Acinetobaeter, may influence patterns of
spontaneous mutagenesis and repair. An example, described above, would
be design of a rearrangement that brings the nearly identical eatl.!F and
peal.! F regions into linear proximity. In light of the general tendency of
duplicated regions to be unstable in prokaryotic chromosomes,it seems
possible that such a rearrangement would elicit genetic instability.
228
TECHNIQUES FOR ANALYSIS AND MANIPULATION OF GENES
Analysis of the B-ketoadipate pathway has raised questions about the

genetic and physiological consequences of gene arrangement. Natural
transformation (Juni, 1972) should facilitate investigation of these and other
questions that are raised by the fascinating biology of Acinetobacter. This
article concludes with a general summary of insights that may be gained by
application of relatively simple genetic procedures to the naturally
transformable strain of Acinetobacter.
Selection of Spontaneous Mutant Strains
As noted in a preceding section, B-ketoadipate selects for strains

carrying dysfunctions in pcaG and pcaH, structural genes for
protocatechuate oxygenase(Canovas and Stanier, 1967). Such strains
accumulate the inducer protocatechuate from endogenously-formed
shikimate and consequently express the pca genes constitutively. An
alternative and more specific procedure for mutant selection emerged from
the observation that a mutation in pcaD prevents growth with succinate in
the presence of compounds that are metabolised via protocatechuate (G.
Hartnett and B. Averhoff, unpublished observations). The physiological
basis for the growth inhibition is unknown; accumulation of the chemically
reactive B-ketoadipate enol-lactone (amston and Stanier, 1966), the
substrate of the enzyme encoded by pcaD, may be toxic to cells. Secondary
mutants that resist the toxic effect can be isolated readily, and most of these
are blocked in early steps of the protocatechuate pathway (G. Hartnett,
unpublished observations). Thus exposure of pcaD mutants to p-
hydroxybenzoate yields mutants blocked in pobA, the structural gene for p-
hydroxybenzoate hydroxylase (Fig. 1). Such mutants remain unable to grow
in the presence of quinate or shikimate, compounds that are metabolised to
protocatechuate independently of p-hydroxybenzoate. The initial selection
also yields some strains that grow in the presence of quinate or shikimate;
these strains usually carry mutations in the structural genes for
protocatechuate oxygenase.
Selection for secondary mutants derived from apc aD deficient strain

opens the genetics of the early steps of the protocatechuate pathway to
analysis. It also allows selection of recombinant strains carrying inserts
within the structural genes for protocatechuate oxygenase. Because such
strains endogenously accumulate the inducer protocatechuate, they
constitutively express the inserted DNA.
Analysis of the genetic properties of mutants lacking protocatechuate

oxygenase is facilitated if they carry a wild-type pcaD gene, and this gene can
be introduced from a cloned DNA fragment by transformation. For
successful selection of the wild-type recombinant, the initial recipient carries
mutations in both catD and pcaD (G. Hartnett, unpublished observations).
The wild-type pcaD gene, constitutively expressed in the recombinant
because of the dysfunction in protocatechuate oxygenase, is selected by
229
demanding its complementation of the catD mutation during growth with
benzoate. Thus it is possible to select and maintain mutants lacking
protocatechuate oxygenase in a genetic background containing a mutant
pcaD gene. This gene can be replaced with the wild-type gene when genetic
analysis necessitates separation of the protocatechuate oxygenase mutations
from other pca mutations. This procedure should facilitate investigation of
unstable protocatechuate oxygenase mutations and may give an insight into
the processes underlying repair of such mutations.
Location of Mutations and Identification of Cloned DNA
As noted above for pcaD, wild-type Acinetobacter DNA carried in

recombinant E. coli plasmids can be used to replace mutations in the
Acinetobacter chromosome (Neidle and Ornston, 1986; Neidle et aI.,
1987). This procedure permits mutations to be located within known DNA
fragments carried in plasmids. Restriction inactivates donor DNA if the site
of cleavage lies within 100 bp of the mutant allele (Hartnett et aI., 1990).
Thus sets of donor DNA fragments can be used to determine the
approximate position of a mutant allele, and further restriction of donor
DNA can locate mutations within distances that can be resolved on a single
DNA sequencing gel.
Strains carrying known mutations in the Acinetobacter chromosome

can be used as recipients in natural transformation to identify the
corresponding wild-type DNA either in gels or in E. coli colonies carrying
recombinant plasmids. DNA from gel slices containing an Acinetobacter
gene can be placed directly upon a lawn of recipient cells in which the gene is
dysfunctional and prevents growth on the selective growth medium (Neidle
and Ornston, 1986; Neidle et al., 1987). Recombinant colonies appear where
DNA containing the wild-type gene comes into contact with the plate. E.
coli colonies carrying the Acinetobacter pabA gene in pUC18 were
identified by replica-plating of colonies containing a library of cloned
Acinetobacter DNA within the plasmid directly upon a lawn of recipient
Acinetobacter strains, deficient in pabA, and spread upon p-hydroxy-
benzoate plates (B.Averhoff, unpublished observations). Growth of
transformants, produced from donor DNA released from lysed cells within
the E. coli colonies, revealed the colonies that contained the wild type
pabA gene.
Design of Strains Carrying Deletions and Gene Rearrangements
DNA that has undergone genetic engineering can be introduced into

Acinetobacter by natural transformation (Doten et al., 1987b). This
procedure is so efficient that recombinants which have acquired the
engineered DNA arise with a frequency of several percent of the population
after transformation has taken place in the absence of selection. Thus a
successful screening procedure is all that is required to isolate recombinants
containing designed mutations. Such transformations were used to examine
the effect of catIJ on repair of pcaIJ mutations (Doten et al., 1987b) and to
230
create Acinetobacter strains carrying deletions in pcaD (G. Hartnett,
unpublished observations). Successful selection procedures can simplify
genetics, and it should be noted that the pcaD mutants provide a genetic
background in which constitutively expressed DNA inserts within pcaG or
pcaH can be selected.
Recovery of DNA from the Acinetobacter Chromosome
Genetic analysis of Acinetobacter is facilitated by procedures that

allow the direct recovery of specified DNA fragments from the chromosome,
and such a technique is presented by gap repair (Orr-Weaver et aI., 1983).
This technique rests upon the necessity for circularisation of a plasmid in
order to achieve its replication. A donor plasmid contains a copy of
Acinetobacter DNA extending beyond the chromosomal target for
recovery. The plasmid is linearised by introduction into its copy of
Acinetobacter DNA of a gap encompassing the chromosomal target. The
linearised DNA is introduced into a recipient Acinetobacter strain by
transformation, and strains containing a circularised plasmid are selected by
demanding a plasmid-encoded function for growth. Recombination of target
DNA from the chromosome into the plasmid allows its circularisation. After
their selection in Acinetobacter, plasmids carrying the target DNA can be
introduced into E. coli, and thus specified chromosomal DNA segments can
be isolated for analysis (L. Gregg, unpublished observations).
Gap repair has been applied successfully to recover mutations from

the Acinetobacter chromosome (L. Gregg, unpublished observations) and
also to obtain DNA lying between po bA and the pca region in the
supraoperonic cluster containing these genes (B. Averhoff, unpublished
observations). There is every reason to believe that gap repair and other
techniques made possible by the natural transformation of Acinetobacter
will open much of the biology of this extraordinary genus to genetic
investigation.
ACKNOWLEDGEMENTS
Research from our laboratory has been supported by grants from the
Army Research Office, the National Science Foundation and the National
Institutes of Health. Support from the Celgene Corporation has been
particularly valuable because it has allowed development of Acinetobacter
genetics. Kind permission to report unpublished research from our
laboratory was given by B. Averhoff, A.Benson, D. Elsemore, L. Gregg, G.
Hartnett, A. Harrison, J. Houghton, J. Hughes, D. Mitchell, E. Neidle, R.
Parales and M. Shanley. DNA sequences from the benABCD region,
determined by E. Neidle and C. Hartnett, were analysed in collaboration
with A. Bairoch, M. Rekik and S. Harayama from the University of Geneva,
Switzerland. This manuscript was organised and prepared under the expert
guidance of Susan Voigt.
231
REFERENCES
Aldrich, T. L., and Chakrabarty, A. M., 1988, Transcriptional regulation,

nucleotide sequence, and localization of the promoter of the
catBC operon in Pseudomonas putida, 1. Bacteriol.,
170:1297.
Aldrich, T. L., Frantz B., Gill B., Kilbane, J. F., and Chakrabarty, A. M.,
1987, Cloning and complete nucleotide sequence determination of
the catB gene encoding cis, cis-muconate lactonizing enzyme,
Gene 52:185.
Allewell, N., 1989, Evolving to dissimilate hydrocarbons, Trends in
Biochem. Sci., 168:473.
Anderson, R. P., and Roth, J. R., 1977, Tandem genetic duplications in phage
and bacteria, Ann. Rev. Microbiol., 31:473.
Cain, R. B., 1980, The uptake and catabolism of lignin-related aromatic
compounds and their regulation in microorganisms, in "Lignin
Biodegradation: Microbiology, Chemistry and Potential
Applications, vol. 1," p.21, T. Kent Kirk, T. Higuchi, and H.
Chang, eds., CRC Press, Boca Raton FL.
Canovas, J. L., and Johnson, B. F., 1968, Regulation of the enzymes of the B-
ketoadipate pathway in M oraxella calcoacetica 4. Constitutive
synthesis of B-ketoadipate succinyl-CoA transferases II and III,
Eur. 1. Biochem., 3:312.
Canovas, J. L., and Stanier, R. Y., 1967, Regulation of the enzymes of the B-
ketoadipate pathway in M oraxella calcoaceticus, Eur. 1.
Biochem.,1:289.
Canovas, J. L., Ornston, L. N., and Stanier, R. Y., 1967, Evolutionary
significance of metabolic control systems, Science 156:1695.
Canovas, J. L., Wheelis, M. L., and Stanier, R. Y., 1968a, Regulation of the
enzymes of the B-ketoadipate pathway in M oraxella
calcoacetica 2. The role of protocatechuate as inducer, Eur. 1.
Biochem. 3:293.
Canovas, J. L., Johnson, B. F., and Wheelis, M. L. 1968b, Regulation of the
enzymes of the B-ketoadipate pathway in M oraxell a
calcoacetica 3. Effects of 3-hydroxy-4-methylbenzoate on the
synthesis of enzymes of the protocatechuate branch, Eur. 1.
Biochem., 3:305.
Chargaff, E., 1950, Chemical specificity of nucleic acids and mechanism of
their enzymatic degradation, Experimentia, 6:201.
Chen, K.-S., Peters, T. c., and Walker, J. R., 1990, A minor arginine tRNA
mutant limits translation preferentially of a protein dependent on
the cognate codon, 1. Bacteriol., 172:2504.
Doten, R. c., Ngai, K.-L., Mitchell, D. J., and Omston, L. N., 1987a, Cloning
and genetic organization of the pca gene cluster from
Acinetobacter calcoaceticus, 1. Bacteriol., 169:3168.
Doten, R. c., Gregg, L. A., and Omston, L. N., 1987b, Influence of the
catBCE sequence on the phenotypic reversion of apcaE mutation
in Acinetobacter calcoaceticus, 1. Bacteriol., 169:3175.
232
Edelman, G. M., and GaIly, J. A., 1970, Arrangement and evolution
of eukaryotic genes, in "The Neurosciences: Second Study
Program," p.962, F. O. Schmitt, ed., Rockefeller University Press,
New York.
Frantz, B., and Chakrabarty, A. M., 1987, Organization and nucleotide
sequence determination of a gene cluster involved in 3-
chlorocatechol degradation, Proc. Natl. Acad. Sci. USA,
84:4460.
Goldman, A., Ollis, D. L., Ngai, K.-L., and Steitz, T. A., 1985, Crystal
structure of muconate lactonizing enzyme at 6.5 A resolution, I.
Mol. Biol., 182:353.
Grantham, R., Gautier, c., Gouy, M., Jacobzone, M., and Mercier, R., 1981,
Codon catalog usage is a genome strategy modulated for gene
expressivity, Nucl. Acids Res., 9:r43.
Harayama, S., Lehrbach, P. R., and Timmis, K. N., 1984, Transposon
mutagenesis analysis of meta-cleavage pathway operon genes of
the TOL plasmid of Pseudomonas putida mt-2, I.Bacteriol.,
160:251.
Hartnett, C.S., Neidle, E.L., Ngai, K.-L., and Ornston, L.N., 1990, DNA
sequences of genes encoding Acinetobac:ter calcoaceticus
protocatechuate 3,4-dioxygenase: Evidence indicating shuffling of
genes and of DNA sequences within genes during their
evolutionary divergence,.I. Bacteriol., 172:956.
Henikoff, S., Haughn, G. W., Calvo, J. M., and Wallace, J. c., 1988, A large
family of activator proteins, Proc. Natl. Acad. Sci. USA,
85:6602.
Hewetson, L., Dunn, H. M., and Dunn, N. W., 1978, Evidence for a
transmissible catabolic plasmid in Pseudomonas putida
encoding the degradation of p-cresol via the protocatechuate
ortho cleavage pathway, Genet. Res., 32:249.
Horuichi, T., Tomizawa, J., and Novick, A., 1962, Isolation and properties of
bacteria capable of high rates of B-galactosidase, Biochim.
Biophys.Acta, 55:152.
Hughes, J., Shapiro, M., Houghton, J., and Ornston, L.N., 1988, Cloning and
expression of pca genes from Pseudomonas putida in
Escherichia coli,.I. Gen. Microbial., 134:2877.
Ikemura, T., 1981a, Correlation between the abundance of E. coli
transfer RNAs and the occurrence of the repetitive codon in
protein genes, .I. Mol. Biol., 146:1.
Ikemura, T., 1981b, Correlation between the abundance of E. coli transfer
RNAs and the occurrence of the repetitive codon in protein genes:
a proposal for a synonymous codon choice that is optimal for E.
coli translational system, 1. Mol. Bioi., 151:389.
Ikemura, T., 1982, Correlation between the abundance of yeast transfer
RNAs and the occurrence of the repetitive codon in protein genes:
differences in synonymous codon choice patterns of yeast and E.
coli with references to the abundance of isoacceptor transfer
RNAs, 1. Mol. Biol., 158:573.
233
Jackowski, S., and Rock, e.O., 1986, Consequences of reduced intracellular
coenzyme A content in Escherichia coli., I. Bacterial.,
166:866.
Jackson, E. N., and Yanofsky, e., 1973, Duplication-translocations of
tryptophan operon gene in Escherichia coli, J. Bacterial.,
116:33.
Juni, E., 1972, Interspecies transformation of Acinetobacter: Genetic
evidence for a ubiquitous genus, J. Bacterial., 112:917.
Kaplan, J. B., Goncharoff, P., Seibold, A. M., and Nichols, B. P., 1984,
Nucleotide sequence of the Acinetobacter calcoaceticus
trpGDC gene cluster, Mol. Biol. Evol., 1:456.
Katti, S.K., Katz, B., and Wyckoff, H.W., 1988, Crystal structure of
muconolactone isomerase at 3.3 A resolution, I. Mol. BioI.,
205:557.
Katz, B., Ollis, D. L., and Wyckoff, H. W., 1985, Low resolution crystal
structure of muconolactone isomerase, I. Mol. Biol., 184:311.
Kemp, M. B., and Hegeman, G. D., 1968, Genetic control of the JJ-
ketoadipate pathway in Pseudomonas aeruginosa, .I.
Bacteriol.,96: 1488.
Konigsberg, W., and Godson, G. N., 1983, Evidence for use of rare codons in
the dnaG gene and other regulatory genes of Escherichia coli,
Kunkel, T., 1985, Rapid and efficient site specific mutagenesis without
phenotypic selection, Proc.Natl. Acad. Sci. USA, 82:488.
Kunkel, T. A., and Soni, A., 1988, Mutagenesis by transient misalignment,
J.Biol. Chem., 263:14784.
McCorkle, G. M., Yeh, W. K., Fletcher, P., and Ornston, L. N., 1980,
Repetitions in the NH 2 -terminal amino acid sequence of Jl-
ketoadipate enol-lactone hydrolase from Pseudomonas putida,
I. BioI. Chem., 255:6335.
Meagher, R., Berry-Lowe, S., and Rice, K., 1989, Molecular evolution of the
small subunit of ribulose bisphosphate carboxylase: Nucleotide
substitution and gene conversion, Genetics 123:845.
Meselson, M., 1988, Methyl directed repair of DNA mismatches, in
"Recombination of Genetic Material," p.91, K.B. Low, ed.,
Academic Press, New York.
Muto, A., Kawauchi, Y., Yamao, F., and Osawa, S., 1984, Preferential use of
A- and U-rich codons for Mycoplasma capricolom ribosomal
proteins S8 and L6, Nucl. Acids Res., 12:8209.
Nagylaki, '1'., and Petes, '1'.D., 1982, Intrachromosomal gene conversion and
the maintenance of sequence homogeneity among repeated genes,
Genetics, 100:315.
N eidle, E.L., and Ornston, L.N., 1986, Cloning and expression of the
Acinetobacter calcoaceticus catechol 1,2-dioxygenase
structural gene, catA, in Escherichia coli, .I. Bacterio!.,
168:815.
Neidle, E.L., and Ornston, L.N., 1987, Benzoate and muconate, structurally
dissimilar metabolites, induce expression of catA in
Acinetobacter calcoaceticus, I. Bacteriol., 169:414.
234
Neidle, E.L., Shapiro, M., and Omston, L. N., 1987, Cloning and expression
in Escherichia coli of Acinetobacter calcoaceticus genes for
benzoate degradation, l. Bacteriol., 169:5496.
Neidle, E. L., Hartnett, c., Bonitz, S., and Omston, L. N., 1988, DNA
sequence of the Acinetobacter calcoaceticus catechol 1,2-
dioxygenase I structural gene catA: evidence for evolutionary
divergence of intradiol dioxygenases by acquisition of DNA
repetitions, l. Bacteriol., 170:4874.
Neidle, E. L., Hartnett, C. S., and Omston, L. N., 1989, Characterization of
Acinetobacter calcoaceticus catM, a repressor gene
homologous in sequence to transcriptional activator genes,
l.Bacteriol., 171:5410.
Ngai, K.-L., and Ornston, L. N., 1988, Abundant expression of
Pseudomonas genes for chlorocatechol metabolism, l.
Nomura, M., Gourse, R., and Baughman, G., 1984, Regulation of the
synthesis of ribosomes and ribosomal components, Ann. Rev.
Biochem., 53:75.
Ohama, T., Yamao, F., Muto, A., and Osawa, S., 1987, Organization and
codon usage of the streptomycin operon in Micrococcus luteus,
a bacterium with a high genomic G+C content, l. Bacteriol.,
169:4770.
Ohlendorf, D.H., Lipscomb, J. D., and Weber, P. c., 1988, Structure and
assembly of protocatechuate 3,4-dioxygenase, Nature, 336:403.
Omston, L.N., 1966, The conversion of catechol and protocatechuate to B-
ketoadipate by Pseudomonas putida, IV. Regulation, 1. Bioi.
Chem., 241:3800.
Oms ton, M. K., and Ornston, L. N., 1972, The regulation of the B-
ketoadipate pathway in Pseudomonas acidovorans and
Pseudomonas testosteroni, l. Gen. Microbiol., 73:455.
Omston, L.N., and Stanier, R.Y., 1966, The conversion of catechol and
protocatechuate to B-ketoadipate by Pseudomonas putida. I.
Biochemistry, l. Bioi. Chem., 241:3776.
Omston, L. N., and Yeh, W.K., 1979, Origins of metabolic
diversity: Evolutionary divergence by sequence repetition,
Ornston, L.N., and Yeh, W. K., 1982, Recurring themes and repeated
sequences in metabolic evolution, in "Biodegradation and
Detoxification of Environmental Pollutants," p.105, A.M.
Chakrabarty, ed., CRC Press, Miami.
Omston, L. N., Neidle, E. L., and Houghton, J.E., 1990a, Gene
rearrangements, a force for evolutionary change; DNA sequence
arrangements, a source of genetic constancy, in "The Bacterial
Chromosome," p.325, M. Riley and K. Drlica, eds., American
Society for Microbiology, Washington, D.C.
Oms ton, L. N., Houghton, J. E., Neidle, E. L., and Gregg, L. A., 1990b,
Subtle selection and novel mutation during evolutionary
divergence of the B-ketoadipate pathway, in "Pseudomonas:
Biotransformation, Pathogenesis and Evolving Biotechnology,"
235
p.207, S.Silver et aI., eds., American Society for Microbiology.
Washington, D.C.
Orr-Weaver, T. L., Szostak, J. W., and Rothstein, R. J., 1983,
Genetic applications of yeast transformation with linear
and gapped plasmid, M eth. Enzymol., 101:228.
Parke, D., and Ornston, L. N., 1984, Nutritional diversity of Rhizobiaceae
revealed by auxanography, 1. Gen. M icrobiol., 130:1743.
Patel, R. N., Mazumdar, S., and Ornston, L. N., 1975, B-Ketoadipate enol-
lactone hydrolases I and II from Acinetobacter calcoaceticus,
1. Bioi. Chem., 250:6567.
Peoples, O. P., and Sinskey, A. J., 1989, Poly-B-hydroxybutyrate biosynthesis
in Alcaligenes eutrophus H16: Characterization of the genes
encoding B-ketothiolase and acetoactyl-CoA reductase, 1. BioI.
Chem., 264:15293.
Rigby, P. W. J., Burleigh, B. D., and Hartley, B. S., 1974, Gene duplication in
experimental enzyme evolution, Nature, 251:200.
Rosenberg, S. L., and Hegeman, G. D., 1969, Clustering of functionally
related genes in Pseudomonas aeruginosa, 1. Bacteriol.,
108:1270.
Rothmel, R. K., Aldrich, T. L., Houghton, J. E., Coco, W. M., Ornston, L.
N., and Chakrabarty, A.M., 1990, Nucleotide sequencing and
characterization of Pseudomonas putida catR: A positive
regulator of the catBC operon is a member of the LysR family, 1.
BacterioL., 172:922.
Schleif, R., 1988, DNA looping, Science, 240:127.
Shanley, M. S., Neidle, E. L., Parales, R. E., and Ornston, L. N., 1986,
Cloning and expression of Acinetobacter calcoaceticus
catBCDE genes in Pseudomonas putida and Escherichia
coli, 1. Bacteriol., 165:557.
Sharp, P.M., and Li, W.-H., 1996, Codon usage in regulatory genes in
Escherichia coli does not reflect selection for "rare" codons,
Nue!. Acids Res., 14:7737.
Stanier, R. Y., 1951, Enzymatic adaptation in bacteria, Ann. Rev.
Microbiol.,5:35.
Stanier, R. Y., and Ornston, L. N., 1973, The B-ketoadipate pathway, in
"Advances in Microbial Physiology, vo1.9," p.g9, A. H. Rose and D.
W. Tempest, eds., Academic Press, London.
Stanier, R. Y., Palleroni, N. J., and Doudoroff, M., 1966, The aerobic
Pseudomonads: A taxonomic study,.I. Gen. M icrobiol., 43: 159.
Tresguerres, M. E. F., de Torrontequi, G., lngledew, W. M., and Canovas, J.
L.,1970, Regulation of enzymes of the B-ketoadipate pathway in
M oraxella: Control of quinate oxidation by protocatechuate,
Eur. 1. Biochem., 14:445.
West, S.E.H., and 19lewski, B., 1988, Codon usage in Pseudo monas
aeruginosa, N ue!. Acids Res., 16:9323.
Wheelis, M.L., 1975, The genetics of dissimilarity pathways III
Pseudomonas aeruginosa, Ann.Rev.Microbiol., 29:505.
236
Wheelis, M. L., and Ornston, L. N., 1972, Genetic control of enzyme
induction in the 13-ketoadipate pathway of Pseudomonas putida:
Deletion mapping of cat mutations, I.Bacteriol., 108:790.
Wheelis, M. L., and Stanier, R. Y., 1970, The genetic control of dissimilatory
pathways in Pseudomonas putida, Genetics, 66:245.
Widom, R. L., Jarvis, E. D., LaFauci, G., and Rudner, R., 1988, Instability of
rRNA operons in Bacillus subtilis, I. Bacterial., 170:605.
Winstanley, c., Taylor, S. c., and Williams, P. A., 1987, pWW174: A large
catabolism by the 13-ketoadipate pathway, Mol. M icrob iol., 1:219.
Xiong, Y., Sakaguchi, B., and Eickbush, T. H., 1988, Gene conversions can
generate sequence variants in the late chorion multigene families
of Bombyx mori, Genetics, 120:221.
Yeh, W. K., and Ornston, L. N., 1980, Origins of metabolic diversity:
Substitution of homologous sequences into genes for enzymes with
different catalytic activities, Proc. Natl. Acad. Sci. USA,
77:5365.
Yeh, W. K., and Ornston, L. N., 1981, Evolutionarily homologous 0<2132
oligomeric structures in 13-ketoadipate succinyl CoA transferases
from Acinetobacter calcoaceticus and Pseudomonas putida,
I. Bioi. Chem., 256:1565.
Yeh, W.K., Davis, G., Fletcher, P., and Ornston, L.N., 1978, Homologous
amino acid sequences in enzymes mediating sequential metabolic
reactions, I. Bioi. Chem., 253:4920.
Yeh, W. K., Fletcher, P., and Ornston, L. N., 1980a, Evolutionary divergence
of coselected 13-ketoadipate enol-lactone hydrolases in
Acinetobacter calcoaceticus,.I. BioI. Chem., 255:6342.
Yeh, W. K., Fletcher, P., and Ornston, L. N., 1980b, Homologies in the NH 2-
terminal amino acid sequence of ){ -carboxymuconolactone
decarboxylases and muconolactone isomerases, .I. Bi 0 l. C hem.
255:6347.
237
ORGANISATION, POTENTIAL REGULATORY ELEMENTS
AND EVOLUTION OF T RP GENES IN
ACINETOBACTER
G. Haspel, V. Kishan and W. Hillen *
Lehrstuhl fur Mikrobiologie

Friedrich Alexander Universitat ErIangen-Nurnberg
Staudtstrasse 5
D-8520 Erlangen
Germany
INTRODUCTION
This article discusses the evolutionary and regulatory implications

derived from the nucleotide sequence of trp genes from Acinetobacter
calcoaceticus. The genetics of tryptophan biosynthesis have been studied
in a wide variety of organisms (Crawford, 1989). As depicted in Fig. 1,
organisation of trp genes differs considerably among the prokaryotes studied
so far. For instance, Escherichia coli has five genes in a single operon for
tryptophan biosynthesis (Yanofsky et aI., 1981), Bacillus subtilis has six
genes in a single operon (Henner et aI., 1984) and A. calcoaceticus has
seven genes in three unlinked clusters (Sawula and Crawford, 1972, 1973).
Unlinked clusters of trp genes have also been reported in Pseudomonas
aeruginosa, P. putida (Holloway et aI., 1979; Shinomiya et aI., 1983) and
Rhizobium meliloti (Johnston et aI., 1978). Sequence information is
available for trp genes from E. coli (Yanofsky et aI., 1981), B. subtilis
(Henner et aI., 1984), R. meliloti (Bae et aI., 1989), P. putida (Crawford
and Eberly, 1989; Essar et aI., 1990a) and P. aeruginosa (Crawford et aI.,
1986; Essar et aI., 1990b). From A. calcoaceticus the trpGDC operon
(Kaplan et aI., 1984), the trpE gene (Haspel et aI., 1990) and the trpFB
operon (Ross et aI., 1990; Kishan and Hillen, in preparation) have been
sequenced. Using a newly developed shuttle vector (Hunger et aI., 1990) for
A. calcoaceticus, the promoter structures of the trpE and trpFB genes
have been determined. This article compares the organisation of trp genes
239
E. coli
E (G) D c (F) B A
S. marcescens
E G D c (F) B A
E G D C F B A
A. ca/coaceticus II II
E G D C F B A
P. aeruginosa H..........J'
B. iactofermentum
E "G D c (F) B A
GP E D C F B A
B. subtilis L.........J!
E D C F B A
S. aureus
R. meli/oti
E (G) D c F B A
II II
Fig. 1. Genetic organisation of trp genes in different prokaryotes.

Universal designations for the trp genes are used. A solid unbroken
line indicates a unit of transcription. Gene fusions are indicated by
brackets. The organisms are specified on the left side of the figure.
C ~ indoleglycerol phosphate synthase; D ~ phosphoribosyltransferase;
E, G ~ subunits of anthranilate synthase; F ~ N-phosphoribo-
sylanthranilate isomerase. The gene arrangements are taken from: E.
coli: Yanofsky et a1. (1981); S. marcescens: Yanofsky (1984); A.
calcoaceticus: this work, Kaplan et al. (1984); P. aeruginosa: Essar
et a!. (1990b); B.lactofermentum: Matsui et a1. (1986), Sano and
Matsui (1987); B.subtilis: Henner et a1. (1984); S. aureus: Crawford
(1975); R.meliloti: Bae et al. (1989).
in A. calcoaceticus with that in other organisms and discusses potential

common regulatory sequences for their expression. Striking similarities of
trp genes from P. putida and A. calcoaceticus with respect to these
features are presented and discussed.
THE ORGANISATION OF TRP GENES IN ACINETOBACTER
It has been reported previously that the A. calcoaceticus trp genes

are dispersed in three unlinked clusters, namely trpE, trpGDC and trpFBA
(Sawula and Crawford, 1972, 1973). We have recently isolated and
sequenced the trpFB operon from A. calcoaceticus (Kishan and Hillen, in
preparation) and found no indication for the presence of the trpA gene in
that operon. Thus, we propose that trpA is distinct from the trpF B operon.
The comparison with the other trp gene organisations in Fig. 1 reveals that
A. calcoaceticus and P. aeruginosa both contain four unlinked clusters;
however, the distribution of genes in the clusters is different. While P.
aeruginosa contains a single trpF gene and a trpBA operon, A.
calcoaceticus contains a trpF B operon and an unrelated trpA gene. The
organisation of trp genes in P. putida seems to be the same as in P.
aeruginosa (Hadero and Crawford, 1986). Three clusters are found in
R. meliloti, two in B. subtilis and the other organisms show single
240
Table l. Codon usage in A.calcoaceticus trp genes
Codon AA trpE trpG trpD trpe trpF trpB
GGG G1y 4 0
GGA G1y 2 0 3
GGT G1y 16 10 25
GGG G1y 10 0 10
GAG G1u B 2 5 4
GAA Glu 27 20 19 4 19
GAT Asp 25 16 16 14 15
GAe Asp 6 1 4 0 11
GTG Val 14 9 6 10
GTA Val 2 5 6
GTT Val 11 3
GTG Val 14 6
GGG Ala 11 7 9
GGA Ala 14 15 12 17
GCT Ala 5 B 5 6
GGG Ala 7 5
AGG Arg 0
AGA Arg 0
AGT Ser
AGe Ser
AAG Lys 6 6
AAA Lys 17 17 11 11
AAT Asn 10 11 10
AAG Asn 7 7
ATG Met 11 15 4 2 15
ATA Ile 2 0 0 1 1 0
ATT Ue 22 11 24 12 10 20
ATC Ue S 6 2 2 7
AGG Thr 5
ACA Thr 10
ACT Thr 5
ACC Thr 10
TGG Trp
TGA End
TGT Gys
TGC Cys
TAG End
TM End 1
TAT Tyr 17 14
TAG Tyr 1
TTG Leu 13 9
TTA Leu 9 12 5
TTT Phe 17 6
TTC Phe 1 4
TCG Ser
TCA Set 5
TCT Ser 4
TGG Ser
CGG Arg 0
CGA Arg 1
GGT Arg 17 16
CGG Arg 5
CAG Gln 11 7 5 11
GM Gln 12 4 14 13 14 6
eAT His 9 8 6 5 12
GAG His 2
eTG Leu 10
CTA Leu 2
eTT Leu 10
CTC Leu 3
CGG Pro
CCA Pro 12 3
GGT Pro 10 6
ceG Pro 1 0
241
operons. In E. coli the GD and CF functions are encoded in single
polypeptides, while in Serratia marcescens and B. lactofermentum only
the CF functions are encoded in a single polypeptide. In R. meliloti the
EG functions are located on a single polypeptide. The other examples
shown in Fig. 1 encode a separate polypeptide for every function. Thus, the
organisation of trp genes in A. calcoaceticus shows most similarity to
those in P. aeruginosa and P. putida. This result is also supported by the
amino acid homologies (see below).
NUCLEOTIDE SEQUENCE OF THE TRP GENES ENCODED

BY ACINETOBACTER
The trpGDC (Kaplan et aL, 1984), trpE (Haspel et aL, 1990) and
t rp F B (Ross et aI., 1990; Kishan and Hillen, in preparation) determinants
have been isolated and sequenced. The results constitute a data hase for the
analysis of codon usage in A. calcoaceticus in general, and in the trp genes
in particular. The codon usage frequencies are presented in Table 1. The
codons with maximal AT content are generally preferred over the ones with
GC nucleotides in their wohhle positions. This preference is particularly
pronounced for the His, Pro, Phe, Tyr, Cys, lie, Asn, Lys, Asp and Glu codons
in all t rp genes presented in Table 1. Among the lie codons A TA seems to
be avoided and, as a result, the AT preference in this case is almost
exclusively achieved hy use of the ATT codon. No clear preference for AT
rich codons is found for Gly, Val, Ser, Thr, Leu and GIn. Interestingly the
codon preference of trpE differs for some amino acids from that of the other
trp genes. In particular, in the cases of Leu and Val, a preference for GC-
rich codons is found. Since trpE is unlinked to any other trp gene, this could
reflect a different evolutionary source of trpE (see below). The amino acid
Arg can be encoded by six codons. Among these, AGG is not used at all, and
AGA only at two positions. CGG is used only four times while CGC is used
23 times. CGA and CGT are used 11 and 53 times, respectively. This shows
a clear tendency to avoid certain codons. The reason for this is not clear;
however, they are under-represented in many other prokaryotic genes as well
(Grantham et aL, 1980).
We have recently reported the trpE nucleotide sequence from A.

calcoaceticus (Haspel et aL, 1990). Here we extend this sequence by
about 850 nucleotides upstream of trpE. The sequence is displayed in Fig.
2. It contains a reading frame with a maximal coding capacity for 234 amino
acids constituting a potential protein of 30,817 DaL We do not have any
data regarding the expression of this reading frame. In addition, other start
codons appear in the N -terminal region which could lead to smaller proteins.
The 3'-end of the reading frame overlaps with the trpE promoter as
determined by primer extension analysis (Haspel and Hillen, in preparation)
and indicated in Fig. 2. This argues strongly against expression of this
reading frame because no primer extension product of the corresponding
length has been found. The nucleotide sequencing result from the trpE
gene of P. putida revealed the same organisation with a reading frame
242
10 30 50
GTTGTTTTTTAAATAATCTTGCCTGAGCTAGACAAACGTCTGTAAAGGCATGCGAAAAGC
M R K A
70 90 110
ATGCCTTTACTTTTTTATAAGGTGCAATATGTCTTTAGTACAATTAGAACGTCGTCAACT
C L Y F FIR C N M S L V Q L ERR Q L
130 150 170
GGTCTTGTTTGATCTAGATGGAACATTGGTCGATACAGCATCAGATATGTATCGTGCGAT
V L F 0 LOG T L V 0 T A S 0 M Y RAM
190 210 230
GAATCTGACTTTGGATCATCTGGGGTGGTCTCGTGTTACAGAAGCTCAAATTCGCCAATG
N L T L 0 H L G W S R V TEA Q IRQ W
250 270 290
GGTTGGACAGGGTACTGGCAAGTTATGCGATGCAGTTTTAAAACATTTGTTTGAAGAAGT
V G Q G T G K LCD A V L K H L FEE V
310 330 350
AGAACCTGCAAAGCATCAAATGCTGTTAACCACTTATCTCGAAATTTATGCACAAGAGTT
EPA K H Q M L L TTY LEI Y A Q E L
370 390 410
GTGTGTGACCAGTCGCTTATTTGAAGGTGTGCAAGCATTTCTTGATGAGTGTAAGGCGCG
C V T S R L F E G V Q A FLO E C K A R
430 450 470
TAAAATCGAGATGGCTTGTGTAACCAACAAGCCTGAGCAATTGGCGAGAAATTTATTGGA
K I E MAC V T N K P E Q L A R N L L E
490 510 530
AACACTTAAGATTGGCGATTACTTTGACTTGGTCGTGGGTGGGGATACTTTGCCTGTACG
T L K I GOY F D L V V G G D T L P V R
550 570 590
AAAACCAGACCCGCTTCCATTGCTACATAGTGTGCAGGTGATGAAGACCACTATTGAGAA
K PDP L P L L H S v Q v M K TTl E N
610 630 650
TACCTTGATGATCGGTGATTCAAAAAATGATGTTGAAGCAGCCAGACGTGCGGGTATCGA
T L M I G 0 S K N D V E A A R RAG I D
-35 670 -10 710
TTGTATTGTAGTCAGCTATGGCTATAATCATGGAGAAAATATTTATGATAGTCATCCGCA
C I V V S Y G Y N H G E N I Y D S H P Q
730 750 770
AGAAGTTGTGGATCGTCTGGATCAATTGATTTAATCATTGAGGAGTGGGGAATGTATCAC
E V V 0 R L 0 Q L I *
790 810 830
CTTATGGTTTTTCGATGGCGATAATCGCGTCTTAAATCACCTTTATCGTACTTAACTTAA
850 870 890
AACCATCGAGAATACTCATGACATCACTAACTCAATTTGAACAGCTTAAAACAGCAGGCT
M T S L T Q F E Q L K TAG
Fig. 2. Nucleotide and deduced amino acid sequence of 850 bp upstream of
the A. calcoaceticus trpE gene. The amino acid sequence is given in the
one letter abbreviation. The -10 and -35 regions of the trpE promoter
are indicated by the lines above the sequence. The trpE gene starts at
position 858.
upstream of trpE (Essar et aI., 1990a). A comparison of the encoded amino

acid sequence is shown in Fig. 3. The P. putida N-terminus shows
homology beginning at amino acid number 31 of the A. calcoaceticus
sequence. At the C-terminal end the P. putida reading frame extends for a
further 36 codons. Both encoded amino acid sequences have 39% identical
amino acids, which are mainly clustered in the C-terminal half. This is far
above background and indicates clearly that both trpE genes originate from
the same source, which may be different from that of the other trpE genes.
This conclusion is also supported by the homology in the primary amino acid
structures of both trpE genes. They contain about 53% identical amino
acids and only three gaps need to be introduced to reach this strong
homology (Deveraux et aI., 1984). The average homology among trpE genes
243
.
31 GTLVDTASDMYRAMNLTLDHLGWSRVTEAQIRQWVGQGTGKLCDAVLKHL
.... .. ........ .. . 80
1 GTLIDSVPDLAAAVDRMLLELGRPPADLEAVRHWVGNGAQVLVRRALAGG 50
...... .. .
81 FEE •• VEPAKHQMLLTTYLEIYAQELCVTSRLFEGVQAFLDECKARKIEM 128
51 IEHDAVDDVLAEQGLALFMEAYAQSHELTV.VYPGVKDTLRWLQKQGVEM 99
129 ACVTNKPEQLARNLLETLKIGDYFDLVVGGDTLPVRKPDPLPLLHSVQVM 178

. ..............
100 ALITNKPERFVAPLiDQMKiGRYFRWMIGGDTLPQKKPDPAALLFVMQMA 149
179 KTTIENTLMIGDSKNDVEAARRAGIDCIVVSYGYNHGENIYDSHPQEVVD 228

. ........ . .
150 GVTPQQSLFVGDSRSDVLAAKAAGVQCVGLTYGYNHGRPiHDETPSLVID 199
229 RLDQLI 234

200 DLRALL 205
Fig. 3. Homology of the open reading frames upstream of trpE in A.
calcoaceticus and P. putida. The deduced amino acid sequences are
compared according to Deveraux et a1. (1984). Only identical amino
acids are marked by dots between the sequences. The upper sequence
originates from A. calcoaceticus.
Strain Gene Promoter Sequence
E.c. trp TGAGCTG TTGACA ATTAATCATCGAACTAG TTAACT AGTACGCA*

E.c. trpR GTCGTTA CTGATC CGCACGTTTATGATATGC TATCGT ACTCTTTA*
S.t. trp ATAGGTG TTGACA TTATTCCATCGAACTAG TTAACT AGTACGA*
A.c. trpE GTATCGA TTGTAT TGTAGTCAGCTATGGC TATAAT CATGGA*
A.c. trpFB TTCAACT TTGACG CAAAGCACAAAAATTGCAT TACAAT ACTTA*
S.m. trp AAGAGGG TTGACT TTGCCTTCGCGAACCAG TTAACT AGTACACA*
B.1. trp CGGAAAC TACACA AGAACCCAAAAATGATT AATAAT TGAGACAA
B.s. trp TTTATCA TTGACA AAAAATACTGAATTGTAA TACGAT AAGAACA*
Consensus TTGACA TATAAT

Fig. 4. Promoter structures of different trp genes. The nucleotide
sequences of different trp promoters are shown. The nuc1eotides for
ini tiation of transcription are indicated by asterisks. The consensus
boxes are indicated in bold print. Sequences are aligned with respect
to their consensus boxes. The abbreviations used and the respective
references are: A.c.: A. calcoaceticus; B.s.: B. subtilis, Henner et a1.
(1984); B.1.: Brevibacterium lactofermentum, Matsui et a1. (1986);
E.c.: E. coli, Yanofsky et al. (1981), Gunsa1us and Yanofsky (1980);
S.m.: Serratia marcescens, Miozzari et al. (1978); S.t.: S.typhimurium,
Bennett et a1. (1978).
244
ranges between 30% and 40% (Haspel et aI., 1990). Only the trpE gene
encoded by Clostridium thermocellum shows a similar homology (49%
identical amino acids). However, it does not contain an upstream reading
frame (Sato et al., 1989).
COMPARISON OF TRP PROMOTERS FROM A. CALCOACETICUS

WITH THOSE OF OTHER ORGANISMS
In Fig. 4 trp promoters from different organisms are compared with

those from A. calcoaceticus as determined by primer extension analyses
(Haspel and Hillen, in preparation; Kishan and Hillen, in preparation).
Generally, the A. calcoaceticus promoters seem to use the same consensus
sequence as the other organisms (Hawley and McClure, 1983). Among the
trp promoters shown in Fig. 4 they have the best conservation of the -10
regions, while in the -35 regions only the TTG sequence is strictly conserved
and seems to be most important. It is noteworthy, that the A.
calcoaceticus promoters show the most extreme spacings between the -35
and -10 regions, with only 16 bp for the trpE and as many as 19 bp for the
trpF B promoter. Both spacings would not be optimal for E. coli (Hawley
and McClure, 1983). Furthermore, the distances of the -10 regions to the
start nucleotides of transcription are very short. It remains to be shown in
the future whether these features are peculiarities of these two promoters or
whether they are characteristic for A. calc 0 a ce ti c us promoters in general.
POTENTIAL REGULATORY FEATURES OF THE

A. CALCOACETICUS TRP GENES
Fig. 5 shows a comparison of the upstream sequences from the A.

calcoaceticus trpGDC, trpE and trpFB determinants. A conserved
sequence is found at different distances with respect to the start codons. A
similar element is also found in the trpE and trpGDC sequences of P.
pu tid a (Essar et al., 1990a). It is not clear whether this seq uence is involved
in regulation; however, in trpE and trpFB of A. calcoaceticus it is located
just downstream of the -10 promoter regions. Thus, it may well be a cis
active regulatory site for these promoters. Promoter studies for the
trpGDC operon of A. calcoaceticus and the trpE and trpGDC
determinants of P. putida have not been published so far. However, it is
most unlikely that the conservation of this sequence has occurred by chance
in the upstream region of five functionally linked genes. Blumenberg and
Yanofsky (1982) have shown consensus regions in the trp leader sequences
of some Enterobacteriaceae, which are most probably involved in
attenuation of transcription. This mechanism involves eight blocks of
homologous sequences, six of which are involved in the formation of
alternative secondary structures of the mRNA, depending on the extent of
translation of a leader peptide. Most interestingly, the consensus sequence
found here for the trp genes of A. calcoaceticus and P. putida is nearly
identical to two of the homology blocks (designated 6 and 3) described for
245
N
(1)
""
!rpFB A.c. - 91 TTTG'CGCAA'GC'CAAAAATTG~TTACAATACTTAGCCC~GATGGATAG'TCGGCTGTCTGTCAGGCA - 20

trpE A.c. -180 ~±G~CTATAA~C~TG~AG~~~A; ;ii~~~if~~f;~~GC: ~AA~TTGT~GATC~TC;~~AT~AAT;~~ -109
trpE P.p. -142 ACCGTGGCGGTTACTGGCAAATT
. . TGGATGAAAGTCATCAA
. . .. -... ... .. TCTGGCCCGTTGGCGTTGGCGCGCCTGA
... - 71
trpGDC A.c. -184 CCTCATAAGTGGGC ATCATGATAGTTAGCCCG CTCATAGTCTAATGTATAAAAAATAGTG -122
trpGDC P.p. -370 TCAGGGAACATC~T±TTC±TT~T:TtTi~iGGTAC~tt~: ~GACCTTAGcAAATCGCCGTGcA±CCCT -299
tTtATGAtAgTCAgCCcg
Fig. 5. Homology of upstream sequences of Lrp genes from A,calcoaceticus and P. putida,Upstream sequences,
numbered from the respective start codon and specified on the left side using the same abbreviations as in
Fig.4, are shown for various trp genes. Identical nucleotides of neighbouring sequences are indicated by
dots. The conserved sequence is boxed and printed in bold letters. A concensus sequence is given at the
bottom. A capital letter indicates conservation in four or more genes, while a small letter indicates
conservation in three genes.
the Enterobacteriaceae. However, the other elements necessary to cause
attenuation of transcription are not present. This indicates that the
evolutionary source of the trp genes may have been regulated by attenuation.
Why most of the necessary sequence elements except the two shown in Fig. 5
were removed by evolutionary drift is not clear. Their conservation would
imply, however, an important function for these elements. If they exhibited
a regulatory function one would have to assume that the trp genes are co-
regulated; indeed Cohn and Crawford (1976) showed that the expression of
A. calcoaceticus-encoded trp genes is subject to repression. However,
according to these results, trpGDC and trpE are co-regulated, while trpFB
and trpA show a different type of regulation. A previous report has focused
on the feedback inhibition of the first two enzymes of the pathway and on
repression of anthranilate synthase (Twarog and Liggins, 1970). Thus, the
question of regulation of trp genes in A. calcoaceticus has not yet been
fully resolved.
REFERENCES
Bae, Y.M., Holmgren, E., and Crawford J.P., 1989, Rhizobium meliloti
anthranilate synthase gene: Cloning, sequence and expression in
Escherichia coli, 1. Bacteriol., 171:3471.
Bennett, G.N., Brown, K.D., and Yanofsky, c., 1978, Nucleotide sequence of
promoter-operator region of the tryptophan operon of
Salmonella typhimurium, J. Mol. Bioi., 121:139.
Blumenberg, M., and Yanofsky, c., 1982, Regulatory region of the K I e b-
siella aerogenes tryptophan operon, J. Bacteriol., 152:49.
Cohn, W., and Crawford, J.P., 1976, Regulation of enzyme synthesis in the
tryptophan pathway of Acinetabacter calcaaceticus,
J. Bacteriol., 127:367.
Crawford, J.P., 1975, Gene rearrangements in the evolution of the
tryptophan synthetic pathway, Bacteriol. Rev., 39:87.
Crawford, J.P., 1989, Evolution of a biosynthetic pathway: the tryptophan
paradigm,Ann. Rev. Microbial., 43:567.
Crawford, J.P., and Eberly, L., 1989, DNA sequence of the tryptophan
synthase genes of Pseudomonas putida, Biachimie, 71:521.
Crawford, J.P., Wilde, A., Yelverton, E.M., Figurski, D., and Hedges, R.W.,
1986, Structure and regulation of the anthranilate synthase genes
in Pseudomonas aeruginosa. II. Cloning and expression in
Escherichia coli, Mol. BioI. Eva!., 3:449.
Deveraux, J., Haeberli, P., and Smithies, 0., 1984, A compressive set of
sequence analysis programs for the VAX, Nucl. Acids Res.,
12:387.
Essar, D.W., Eberly, L., and Crawford, J.P., 1990a, Evolutionary differences
in chromosomal locations of four early genes of the tryptophan
pathway in fluorescent pseudomonads: DNA sequences and
characterization of Pseudomonas putida trpE and trpGDC, 1.
247
Essar, D.W., Eberly, L., Han, c.J., and Crawford, J.P., 1990b, DNA
sequences and characterization of four early genes of the
tryptophan pathway in Pseudomonas aeruginosa,
.f. Bacteriol., 172:853.
Grantham, R., Gautier, c., and Gouy, M., 1980, Codon frequencies in 119
individual genes confirm consistent choices of degenerate bases
according to genome type, Nucl. Acids Res., 8:1893.
Gunsalus, R.P., and Yanofsky, c., 1980, Nucleotide sequence and expression
of Escherichia coli trpR, the structural gene for the trp
aporepressor, Proc. Natl. Acad. Sci. USA., 77:717.
Hadero, A., and Crawford, I.P., 1986, Nucleotide sequence of the genes for
tryptophan synthase in Pseudomonas aeruginosa, Mol. Bio!.
Evo!., 3:191.
Haspel, G., Hunger, M., Schmucker, R., and Hillen, W., 1990, Identification
and nucleotide sequence of the Acinetobacter calcoaceticus
encoded trpE gene, Mol. Gen. Genet., 220:475.
Hawley, O.K., and McClure, W.R., 1983, Compilation and analysis of
Escherichia coli promoters, Nue!. Acids Res., 11:2237.
Henner, D.J., Band, L., and Shimotsu, H., 1984, Nucleotide sequence of the
Bacillus subtilis tryptophan operon, Gene, 34: 169.
Holloway, B.W., Krishnapillai, Y., and Morgan, A.F., 1979, Chromosomal
genetics of Pseudomonas, Microbial. Rev., 43:73.
Hunger, M., Schmucker, R., Kishan, Y., and Hillen, W., 1990, Analysis and
nucleotide sequence of an origin of DNA replication for
Acinetobacter calcoaceticus and its use for shuttle plasmids
with Escherichia coli, Gene, 87:45.
Johnston, A.W.B., Bibb, M.J., and Beringer, J.E., 1978, Tryptophan genes in
Rhizobium - their organization and their transfer to other
bacterial genera, Mol. Gen. Genet., 165:323.
Kaplan, J.B., Goncharoff, P., Seibold, A.M., and Nichols, B.P., 1984,
trpGDC gene cluster, M of. BioI. Evol., 1:456.
Matsui, K., Sano, K., and Ohtsubo, E., 1986, Complete nucleotide and
deduced amino acid sequences of the Brevibacterium
lactofermentum tryptophan operon, Nucl. Acids Res.,
14:10113.
Miozzari, G.F., and Yanofsky, c., 1978, The regulatory region of
the trp operon of Serratia marcescens, Nature, 276:684.
Ross, C.M., Kaplan, J.B., Winkler, M.E., and Nichols, B.P., 1990,
An evolutionary comparison of the Acinetobacter
calcoaceticus trpF gene with trpF genes of several organisms,
Mol. BioI. Evol., 7:74.
Sano, K., and Matsui, K., 1987, Structure and function of the trp operon
control regions of Brevibacterium lactofermentum, a glutamic
acid producing bacterium, Gene, 53:191.
Sato, S., Nakada, Y., Hon-Nami, K., Yasui, K., and Shiratsuchi,
A., 1989, Molecular cloning and nucleotide sequence of the
Clostridium thermocellum trpE gene,.I. Biochem., 105:362.
248
Sawula, R.V., and Crawford, J.P., 1972, Mapping of the tryptophan genes of
Acinetobacter calcoaceticus by transformation, 1.
Sawula, R.V., and Crawford, J.P., 1973, Anthranilate synthetase of
Acinetobacter calcoaceticus: separation and partial
characterization of subunits, 1. Bioi. Chern., 248:3573.
Shinomiya, T., Shiga, S., and Kageyama, M., 1983, Genetic determinant of
pyocin R2 in Pseudomonas aeruginosa PAOl. Localization of
the pyocin R2 gene cluster between the trpCD and trpE genes,
Mol. Gen. Genet., 189:375.
Twarog, R., and Liggins, L., 1970, Enzymes of the tryptophan pathway in
Acinetobacter calcoaceticus, .I. Bacterial., 104:254.
Yanofsky, c., 1984, Comparison of regulatory and structural regions of genes
of tryptophan metabolism, Mol. BioI. Evol., 1:143.
Yanofsky, c., Platt, T., Crawford, J.P., Nichols, B., Christie, G., Horowitz,
H., van Clemput, M., and Wu, A., 1981, The complete nucleotide
sequence of the tryptophan operon of Escherichia coli,
Nucl. Acids Res., 9:6647.
249
CODON USAGE IN ACINETOBACTER STRUCTURAL GENES
P.l. White", I.S. Hunter"and C.A. Fewson\>*
Departments of 'Genetics and bBiochemistry

Glasgow G12 8QQ
UK
INTRODUCTION
The first gene sequence from an Acinetobacter was described by

Kaplan et al. (1984). An increasing number of such sequences have been
reported subsequently, reflecting the growing interest in understanding the
molecular biology of this organism. It is now possible to analyse codon
usage from a substantial number of Acinetobacter genes and this is
important for several reasons (Andersson and Kurland, 1990). In particular:
(a) it forms a valuable basis for evolutionary speculations; (b) it is an
important consideration when attempting the expression of heterologous
genes, either of Acinetobacter genes in Escherichia coli and other hosts,
or of foreign genes in Acinetobacter; (c) DNA may be analysed and
potential protein coding regions identified from a knowledge of codon usage;
(d) knowledge of codon usage can aid in the design of effective synthetic
oligonucleotide probes used to clone genes when partial amino acid
sequences are known.
This article presents an analysis of the sequences of various structural

genes from a number of acinetobacters, including the codon usage and
overall base composition at each of the three codon positions.
CODON USAGE AND OVERALL BASE COMPOSITION
The codon usage patterns for all Acinetobacter structural gene

sequences in the EMBL and NBRF (nucleic) data bases and in the published
251
Table 1. Acinetobacter structural genes used in codon analysis
Gene Phenotype or gene product Bacterial strain Ref.
apbA-6 kanamycin resistance A. baumannii 1

BM 2580
gdbA glucose dehydrogenase A. calcoaceticus 2
LMD 79.41
gdhB glucose dehydrogenase A. calcoaceticus 3
LMD 79.41
cyclohexanone mono oxygenase Acinetobacter sp. 4
NCIB 9871
est esterase A. calcoaceticus 5
RAG-l
catA catechol 1,2-dioxygenase A. calcoaceticus 6
BD 413
mro mutarotase A. calcoaceticus 7
DSM 30007
pqq PQQ biosynthesis A. calcoaceticus 8
(I,II,III) LMD 79.41
trpG,D,C tryptophan biosynthesis A. calcoaceticus 9
BD 413
trpF tryptophan biosynthesis A. calcoaceticus 10
BD 413
gltA citrate synthase A. anitratum 11
NCTC 7844
pcaC carboxymuconolactone A. calcoaceticus 12
decarboxylase BD 413
pcaG,H protocatechuate A. calcoaceticus 12
3,4-dioxygenase BD 413
trpB tryptophan biosynthesis A. calcoaceticus 13
BD 413
trpE anthranilate synthase A. calcoaceticus 14
BD 413
References: (1) Martin et al., 1988; (2) Cleton-Jansen et al., 1988;

(3) Cleton-Jansen et al., 1989; (4) Chen et al., 1988; (5) Reddy et
a1., 1989; (6) Neidle et a1., 1988; (7) Gatz and Hillen, 1986; (8)
Goosen et al., 1989; (9) Kaplan et al., 1984; (10) Ross et al., 1990;
(11) Donald and Duckworth, 1987; (12) Hartnett et al., 1990; (13)
Haspel et al., this volume; (14) Haspel et al., 1990.
literature were determined. Twenty gene sequences were available for

analysis in May 1990 (Table 1). The University of Wisconsin Genetics
Computer Group (UWGCG) program CODONFREQ (Devereux et aI.,
1984) was used to calculate codon frequencies for each of the genes and the
results for all 20 genes were summed Cfable 2). These data were then used
252
Table 2. Codon usage in Acinetobacter structural genes
Amino acid Codon Total % Amino acid Codon Total %
G1y GGG 39 8 Met ATG 154 100

GGA 70 14
GGT 286 55 Ile ATA 14 3
GGC 120 23 ATT 286 69
ATC 116 28
G1u GAG 78 22
GM 284 78 Thr ACG 64 17
ACA 94 25
Asp GAT 307 71, ACT 130 35
GAC 107 26 ACC 88 23
Val GTG 118 25 Trp TGG 91 100

GTA 101 22
GTT 164 35 Cys TGT 43 72
GTC 83 18 TGC 17 28
Ala GCG 92 18 Tyr TAT 178 74

GCA 195 38 TAC 63 26
GCT 142 27
GCC 90 17 Leu TTG 105 18
TTA 200 35
Arg AGG 3 1 CTG 65 11
AGA 12 4 CTA 46 8
CGG 8 3 CTT 108 19
CGA 33 11 CTC 54 9
CGT 185 62
CGC 56 19 Phe TTT 182 71
TTC 75 29
Ser AGT 65 18
AGC 52 15 GIn CAG 117 35
TCG 28 8 CM 216 65
TCA 99 28
TCT 96 27 His CAT 109 70
TCC 17 5 CAC 46 30
Lys MG 88 24 Pro CCG 54 16

AM 274 76 CCA 156 47
CCT 105 32
Asn MT 218 63 CCC 16 5
MC 129 37
Term TGA 6 30
TAG 2 10
TM 12 60
253
Table 3. Percentage base composition of Acinetobacter structural genes
Codon position
Base First Second Third
A 26.9 33.4 27.0
G 34.3 16.2 16.6
C 20.7 22 .1 17.0
T 18.0 28.2 39.3
A+T 44.9 61. 6 66.3
to calculate the overall base composition at each of the three codon positions
in the Acinetobacter genes analysed (Table 3).
The results of these analyses show that A and T predominate in the

third codon position (over 66% of third position bases). Similarly, A and T
are again favoured over G and C in the second codon position, comprising
nearly 62% of the bases used. In contrast, there is no bias toward A and T in
the first codon position. The overall A + T content of the 20 genes, as
determined by the average composition of the three codon positions, is
57.6%. These conclusions are in agreement with those of authors who have
observed a predominance of A and T bases in the third position of codons in
individual genes (e.g. Chen et al., 1988; Haspel et al., this volume). The
overall A + T content of these 20 genes is also in agreement with the
predicted A+T content of the Acinetobacter genome (55-62%; Henriksen,
1976). There is a preference for G in the first codon position (34.3%) and a
paucity of G in the second codon position (16.2%). This pattern of
dinucleotide usage has been observed in other protein coding sequences and
may be responsible for monitoring the correct reading frame during
translation (Trifonov, 1987).
Several codons seem to be used very rarely by acinetobacters. The

co dons AGG, AGA and CGG for arginine and the ATA codon for isoleucine
represent less than 5% of the co dons used from specific synonymous codon
groups.
Only one of the gene sequences listed in Table 1 showed any gross
differences in codon usage when compared with the calculated mean codon
usage. This was the gltA gene (citrate synthase) from A. anitratum NCTC
7844, which showed differences, especially in the preferences for codons for
phenylalanine, leucine, histidine, asparagine, aspartate and tyrosine. This
254
difference in codon usage may be a reflection of the level of expression of
this gene, or of its evolution, or it may reflect some other feature of the
genome of this particular strain.
The abundance of A and T bases in the degenerate position of amino

acid codons in Acinetobacter genes should help in the construction of
oligonucleotide probes based upon amino acid sequences. For example, the
codon for glutamic acid is found to be GAA in 78% of all cases. Similarly
the codon for aspartate is predominantly GAT.
OTHER GENES AND SEQUENCES
The catM gene, which encodes a transcriptional repressor, has been

sequenced (Neidle et aI., 1989). Its codon usage shows some differences
from that of the structural genes listed in Table 1, for instance four of its ten
arginine codons are AGA. The overall codon usages of the structural genes
summarised in Table 2 seem to be very similar to those of 11 of the structural
genes of the 13-ketoadipate pathway of Acinetobacter (Ornston and Neidle,
this volume). However, all of them are different from the codon usages of
the structural genes cati, catl and catF. The special evolutionary factors
and selective forces that may have been involved in the emergence of the 13-
ketoadipate pathway of A. calcoaceticus are discussed by Ornston and
Neidle (this volume).
The sequence of the mutarotase gene indicated the existence of a 20

amino acid leader sequence not found in the mature enzyme (Gatz and
Hillen, 1986), but no sequence corresponding to a signal peptide for protein
export was found within the sequence for the est gene, even though the
enzyme may be located external to the cytoplasmic mem.brane (Reddy et aI.,
1989).
Several Acinetobacter genes have been expressed in E. coli and

Pseudomonas putida (e.g. Neidle et aI., 1987; Chen et ai., 1988; Reddy et
ai., 1989; Ornston and Neidle, this volume). Although it is too early to
generalise about promoter sequences, ribosome binding sites and other
regulatory elements in Acinetobacter, the sequences involved in the
regulation of expression of a few of the genes are now being clarified (e.g.
Haspel et ai., this volume; Ornston and Neidle, this volume).
CONCLUSIONS
This article provides an analysis of the codon usage and base

composition of 20 Acinetobacter structural genes. The distribution of
genes amongst different strains and metabolic pathways should be
sufficiently random to ensure that the results are representative of
Acinetobacter structural genes in general. The analysis therefore provides
a reference for comparison of individual genes and a basis for the
construction of oligonucleotide probes.
255
REFERENCES
Andersson, S.G.E., and Kurland, e.G., 1990, Codon preferences in free-

living microorganisms, Microbio!' Rev., 54:198.
Chen, y.-e.J., Peoples, O.P., and Walsh, e.T., 1988, Acinetobacter
cyclohexanone monooxygenase: gene cloning and sequence
determination, J. Bacteriol., 170:781.
Cleton-Jansen, A.-M., Goosen, N., and van de Putte, P., 1988, Nucleotide
sequence of the gene coding for quinoprotein glucose
dehydrogenase from Acinetobacter calcoaceticus, Nucl.
Acids Res., 16:6228.
Cleton-Jansen, A.-M., Goosen, N., Vink, K., and van de Putte, P., 1989,
Cloning, characterization and DNA sequencing of the gene
encoding the M,50,000 quinoprotein glucose dehydrogenase from
Acinetobactercalcoaceticus, Mol. Gen. Genet., 217:430.
Devereux, J., Haeberli, P., and Smithies, 0., 1984, A comprehensive set of
sequence analysis programmes for the VAX, Nucl. Acids Res.,
12:387.
Donald, L.J., and Duckworth, H.W., 1987, Expression and base sequence of
the citrate synthase gene of Acinetobacter anitratum,
Biochem. Cell. Bioi., 65:930.
Gatz, e., and Hillen, W., 1986, Acinetobacter calcoaceticus
encoded mutarotase: nucleotide sequence analysis of the gene. and
characterisation of its secretion in Escherichia coli, Nucl.
Acids Res., 14:4309.
Goosen, N., Horsman, H.P.A., Huinen, R.G.M., and van de Putte, P., 1989,
Acinetobacter calcoaceticus genes involved in biosynthesis of
the coenzyme pyrrolo-quinoline-quinone: nucleotide sequence and
expression in Escherichia coli K-12, J. Bacterio!., 171:447.
Hartnett, c., Neidle, E.L., Ngai, K., and Omston, L.N., 1990, DNA
sequences of genes encoding Acinetobacter calcoaceticus
protocatechuate 3,4-dioxygenase: Evidence indicating shifting of
genes and of DNA sequences within genes during their
evolutionary divergence, J. Bacteriol., 172:956.
Haspel, G., Hunger, M., Schmucher, R., and Hillen, W., 1990, Identification
and nucleotide sequence of the Acinetobacter calcoaceticus
encoded trpE gene, M o!. Gen. Genet., 220:475.
Henriksen, S.D., 1976, M oraxella, Neisseria, Branhamella and
Acinetobacter, Ann. Rev. Microbiol., 30:63.
Kaplan, J.B., Goncharoff, P., Siebold, A.M., and Nichols, B.P., 1984,
trpGDC gene cluster, Mol. Biol. Evol., 1:456.
Acinetobacter baumannii aphA-6 gene: evolutionary and
binding proteins, kinases and other aminoglycoside-modifying
enzymes, Mol. Microbia!., 2:615.
Neidle, E.L., Shapiro, M.K., and Omston, L.N., 1987, Cloning and expression
in Escherichia coli of Acinetabacter calcoaceticus genes for
benzoate degradation, J. Bacteria!., 169:5496.
256
Neidle, E.L., Hartnett, c., Bonitz, S., and Omston, L.N., 1988, DNA
sequence of the Acinetobacter calcoaceticus catechol 1,2-
dioxygenase 1 structural gene catA: evidence for evolutionary
sequence repetitions, J. Bacteriol., 170:4874.
Neidle, E.L., Hartnett, c., and Omston, L.N., 1989, Characterization of
Acinetobacter calcoaceticus catM, a repressor gene
homologous in sequence to transcriptional activator genes,
1. Bacteriol., 171:5410.
Reddy, P.G., Allon, R., Mevarech, M., Mendelovitz, S., Sato, Y., and
Gutnick, D.L., 1989, Cloning and expression in Escherichia coli
of an esterase-coding gene from the oil-degrading bacterium
Acinetobacter calcoaceticus RAG-I, Gene, 76:145.
Ross, C.M., Kaplan, J.B., Winkler, M.E., and Nichols, B.P., 1990, An
evolutionary comparison of Acinetobacter calcoaceticus trpF
with trpF genes of several organisms, Mol. Bioi. Evol., 7:74.
Trifonov, E.N., 1987, Translation framing code and frame-monitoring
mechanism as suggested by the analysis of mRNA and 16 S rRNA
nucleotide sequences, 1. Mol. Bioi., 194:643.
257
THE OUTER MEMBRANE OF ACINETOBACTER:
STRUCTURE-FUNCTION RELATIONSHIPS
P. Borneleit and H.-P.Kleber
Sektion Biowissenschaften
Bereich Biochemie
Karl-Marx-Universitat Leipzig
Talstrasse 33
7010 Leipzig
Germany
INTRODUCTION
The fine structure of the cell envelope of Acinetobacter has been

determined from electron micrographs of thin sections, negatively-stained
and freeze-etched prcparations of whole cells, and of envelopes treated to
remove the various layers (Thornley and Glauert, 1968; Thorne et aI., 1973;
Sleytr et aI., 1974; Thornley and Sleytr, 1974; Aurich et aI., 1977). These
early studies revealed the presence of an outer mcmhrane typical of Gram-
negative bacteria. Furthermore, there are numerous reports of superficial
material covering the outer surface of several Acinetobacter strains; such
material includes crystalline surface layers of protein (S-laycrs) (Thornley et
aI., 1973, 1974), and polysaccharide capsules (Taylor and Juni, 1961; Pines et
aI., 1983). These additional surface layers are of tremendous importance in
determining the surface properties of the cells, and the polysaccharide
capsular material has proven to be a potent extracellular emulsifying agent
(reviewed by Rosenherg and Kaplan, 1987).
This article concentrates on the outer memhrane of Acinetobacter in

the strict sense. Since the early seventies, much knowledge has accumulated
on the structure and function of outer membranes, hut interest has focused
mainly on the Enterobacteriaceae (for reviews see Lugtenberg and van
Alphen, 1983; Hancock, 1984; Nikaido and Vaara, 1985). Much less is known
about non-enteric bacteria, and it has been speculated that their ollter
membrane organisation may he completely different. In this review we
attempt to summarise the known facts about the outer membrane of
Acinetobacter which, in comparison to the Enterobacteriaceae, is
259
characterised by some special permeability properties. These are, as we will
show, most probably related to unusual features of its lipopolysaccharide
(LPS) constituent.
PERMEABILITY PROPERTIES
One of the main functions of the outer membrane is to serve as a

selective barrier towards the external medium. Small solutes may pass
through the outer membrane via water-filled proteinaceous diffusion
channels termed porins (reviewed by Benz, 1985). On the other hand, in
enteric bacteria the outer membrane has developed into a very effective
barrier giving protection against the penetration of hydrophobic molecules
and the lytic action of detergents (Nikaido, 1979).
There are several lines of evidence that suggest the existence of a

functioning so-called hydrophohic pathway in Acinetobacter spp. First,
their usual habitats, soil, water and the skin and mucous membranes of
humans (Henriksen, 1973; Juni, 1978) do not require the protection
mechanisms effective in enteric hacteria living in the intestinal tract of
animals. Second, many Acinetobacter strains are capahle of using
hydrophobic growth suhstrates such as long-chain hydrocarhons (Baumann et
al., 1968), which, prior to degradation, have to pass the outer memhrane.
Strains especially well-characterised in this respect are Acinetobacter sp.
HOI-N, growing on n-alkanes (C 10 to C 20 ) (Stewart et al., 1959), primary
aliphatic alcohols (C 12 to C 16 ) and aliphatic aldehydes (C 12 to C 16 ) (Singer et
aL, 1985), A.calcoaceticus 69-V (Kleber et al., 1973) and A.
calcoaceticus RAG-l (Reisfeld et ai., 1972). According to a screening test
performed by Kleber et al. (1983), 12 of 13 Acinetobacter strains
investigated were capahle of degrading hydrocarhons with chain lengths
longer than C 10 ' Finally, the inclusion of Acinetobacter in the family
Neisseriaceae (Juni, 1984) also renders the existence of a hydrophobic
pathway highly likely. For this hacterial family several cases of
hypersensitivity towards hydrophobic antibacterial agents such as
erythromycin, fusidic acid, Triton X-lOO, acridine orange (Sarubbi et al.,
1975), crystal violet (Guymon and Sparling, 1975) and hydrophobic steroid
hormones (Lysko and Morse, 1981) have heen reported.
However, apart from these general lines of investigation, a systematic

analysis of the permeability properties of the outer memhrane of
Acinetobacter has not been performed. We therefore now present some
additional data concerning the functioning of the hydrophohic pathway (in
part recently published in Borneleit et al., 1990). In contrast to
Esc her i chi a col ian d Sal m 0 n ell a t y phi m uri u m, ten of 11
Acinetobacter strains tested were found to be sensitive to the hydrophobic
antibiotic erythromycin, with minimal inhibitory concentrations of between
1.2 and 50 mg/l (Table 1). Furthermore, for two of the strains, sensitivity
towards novobiocin and gentian violet was observed. Interestingly, in
A. calcoaceticus strain CCM 5593, sensitivity to hydrophobic agents was
260
Table 1. Minimal inhibitory concentrations (mg/l) of antibacterial agents
Strain Erythromycin Novobiocin Gentian violet
Escherichia coli K-12 >150.0 >150.0 >2.0
Salmonella typhimurium LT 2 >150.0 nd >2.0
CCM 2355 b 1.2 nd nd
Acinetobacter sp. EB 10d c 5.0 nd nd

EB 114 c 5.0 nd nd
EB 6 c 5.0 nd nd
Acinetobacter lwoffii
CCM 5572 b 7.5 nd nd
Acinetobacter sp. EB 104 c 7.5 nd nd

EB 113 c 10.0 nd nd
69-Vd 10.0 50.0 0.25
Acinetobacter sp. EB 1l0c 15.0 nd nd
CCM 5595 b 50.0 nd nd
CCM 5593 b 150.0 100.0 1.5
aminimal inhibitory concentrations were determined in nutrient broth

containing varying concentrations of the agent tested and were defined as
the lowest concentration which inhibited visible growth
bCzechoslovak Collection of Microorganisms, Brno, Czechoslovakia
cStrain Collection of the Institute of Biotechnology, Leipzig, Germany
disolated from soil specimens and identified as described by Kleber et al.
(1973)
nd: not determined.
not as pronounced as in strain 69-V. According to Kleber et al. (1983), these

strains differ greatly in their ability to grow on long-chain hydrocarbons. In
agreement with these data, we were able to demonstrate uptake of the
hydrophobic dye gentian violet by the hydrocarbon-degrading strain 69- V,
but not by strain CCM 5593 or the enteric bacteria E. coli and S.
typhimurium (Fig. 1). Finally, we found A. calcoaceticus strains to be
261
A. calcoaceticus CCM 5593
0.30 E. coli K-12
~
S. typhimurium LT 2
o
a>
It)
<
A. calcoaceticus 69V
1
4 6 8 10
Time (min)
Fig. 1. Gentian violet uptake by intact cells of A. calcoaceticus 69-V

and CCM 5593, E. coli K-12 and S. typhimurium LT-2. Cells were grown in
nutrient broth to the early logarithmic phase and uptake rates determined
at 30°C as described by Gustafsson et al. (1973).
much more susceptible to the lytic action of the strong ionic detergent
sodium dodecyl sulphate (SDS) than Enterobacteriaceae. While E. coli has
been reported to tolerate as much as 50 gil in the growth medium
(Lugtenberg and van Alphen, 1983), we could observe visible growth of
Acinetobacter spp. only up to concentrations of about 0.05 g SDS/I.
PROTErN EXPORT
Another important function of the outer membrane of Gram-negative

bacteria is to confine the periplasmic enzymes and other proteins to the
periplasmic space. The enzymes localised in the periplasm are mainly
hydrolytic, with functions that include enabling the cell to digest
macromolecular molecules or low molecular mass organic compounds as
nutrients and conferring protection against certain antibacterial agents.
Several hydrolytic enzymes have been described in Acinetobacter spp.
(Table 2), while the outer membrane itself has been shown to contain a
phospholipase Aj (Torregrossa et a!., 1978) and a palmitoyl-CoA esterase
(Claus et a!., 1985).
Most interestingly, however, some Acinetobacter spp. have been

found to produce extracellular hydrolytic enzymes_ An extracellular lipase
has been described for A.lwoffii 016 (Breuil and Kushner, 1975a,b; Breuil
et a!., 1978) and A_ calcoaceticus 69-V (Haferburg and Kleber, 1982, 1983;
Fischer and Kleber, 1987). An extracellular phospholipase C (haemolysin)
262
Table 2. Hydrolytic enzymes in Acinetobacter spp.
Enzyme Reference
Aminopeptidase I Ludewig, 1983

Endopeptidase Fricke et al., 1983,1986
Exopeptidase Jahreis and Aurich, 1989
Collagenase Monboisse et a1., 1979a,b
a-amylases I and II Onishi and Hidaka, 1978; Onishi et al., 1979
B-G1ucanase Katohda et al., 1979
Lipases Breuil and Kushner, 1975a,b;
Haferburg and Kleber, 1982;
Ka1ogridov-Vassiliadov, 1984
Phospholipase Al Torregrossa et a1., 1977
Phospholipase A2 Thorne et a1., 1976
Phospholipas,? C Lehmann, 1971a,b; Lehmann, 1972, 1973a;
Zaikina and Robakidze, 1976
Esterase Claus et al., 1985; Fischer, 1986
Penicillin amidase Baker, 1984
B-Lactamase Jo1y-Gui11ou et a1., 1988; Paul et a1., 1989
was found in six A. calcoaceticus and two A. lwoffii strains (Lehmann

1973a,b), while an extracellular B-Iactamase is produced by A.
calcoaceticus strains 69-V and CCM 5593 (Blechschmidt et aI., 1989). The
regularly arranged surface protein A of Acinetobacter sp. 199A has also
been shown to have phospholipase A2 activity, and about half of this protein
is released into the growth medium (Thorne et aI., 1976).
Protein export across the outer membrane of Gram-negative bacteria

has, in the past, been considered a rare phenomenon, and has attracted
attention only recently (Pugsley and Schwartz, 1985). In view of the
numerous reports of extracellular enzymes (see above), the outer membrane
of Acinetobacter seems to be especially well-equipped for this process. As
to the underlying mechanism, very few investigations have been performed,
but a genuine selective protein export system is obviously involved.
Non-specific periplasmic leakage could be excluded for the export of

B-Iactamase by A. calcoaceticus 69-V because other periplasmic enzymes,
such as glucose dehydrogenase or alkaline phosphatase, were shown to be
retained within the cells (Blechschmidt et aI., 1989). When the time course
of lipase location was followed during the growth of A. calcoaceticus 69-V,
an accumulation of the enzyme on the outer membrane was observed prior to
export (Fischer et aI., 1987). According to a model based on these results,
the lipase is released from the "overloaded" outer membrane into the
263
surrounding medium without cell lysis or membrane blebbing. The latter
statement is in contrast to our observations concerning the export of B-
lactamase by A. calcoaceticus CCM 5593. In this strain the appearance of
an extracellular B-Iactamase is clearly accompanied by a release of the outer
membrane constituent LPS (Fig. 2). However, as postulated by Pugsley and
Schwartz (1985), cells may easily have developed different strategies to
overcome the outer membrane permeability barrier.
COMPOSITION AND STRUCTURE
The composition of the outer membrane of Acinetobacter is

basically the same as that of the Enterobacteriaceae, with proteins,
phospholipids and LPS as the main constitutents.
With regard to the protein component, a major protein species of

approximate M,40,OOO was demonstrated in Acinetobacter sp. H01-N
(Scott et ai., 1976) and A. calcoaceticus 69-V (Borneleit et ai., 1988), and
was additionally shown to be heat-modifiable (A. calcoaceticus 69-V:
15
•• •
...
•
5
•
.
• o•
0
• •
200 300 400 500

Extracellular B-Iactamase activity (U/g bacteria)
Fig. 2. LPS release as a function of extracellular B-lactamase production

by A _ calcoaceticus CCM 5593. Cells were grown on complex growth media
or minimal media (with acetate as carbon source) supplemented with yeast
extract. B-Lactamase production was varied by adding different
concentrations (0-750 mg/l) of B-lactam inducers (ampicillin or
cefotaxime) _
264
Fischer et ai., 1984; Acinetobacter sp. ATCC 23055 and Its radiation-
resistant mutant FO-l: Nishimura et aI., 1986), thus being comparable to the
structural OmpA protein of E. coli (Benz, 1985). Furthermore, a pore-
forming peptidoglycan-associated protein (porin) of M, 53,000 was identified
(Fischer et al., 1984), analogous to the OmpC/F porins in E. coli (Benz,
1985). A distinguishing feature in the protein composition of the outer
membrane of the Neisseriaceae might be the absence of a small (M, 10,(00)
lipoprotein which is present in the Enterobacteriaceae in large amounts
(Hebeler et aI., 1978). The presence of a similar protein in A.
calcoaceticus 69-V was, however, claimed by Fischer et al. (1984) and
Borneleit et al. (1988).
As to the phospholipid composition, in all Acinetobacter outer

membranes isolated so far, the typical phospholipids of bacterial
membranes, i.e. phosphatidylethanolamine (main component),
phosphatidylglycerol and cardiolipin, were found. In addition, the
lysophosphoJipids lysocardiolipin and lysophosphatidylethanolamine were
detected (Scott et aI., 1976; Borneleit et aI., 1988), as would be expected
from the presence of an outer membrane phospholipase A especially active
against cardiolipin (Torregrossa et aI., 1977).
The LPS component in early studies was a matter of debate because it

was assumed to be practically devoid of the LPS-specific sugar constituent 3-
deoxy-D-manno-2-octulosonic acid (KDO) (Adams et aI., 1970; Scott et aI.,
1976). Later, the presence of KDO was demonstrated by the usual
thiobarbituric acid assay in whole cells of several A. calcoaceticus strains
(McDougall et aI., 1983), in isolated outer membranes (Fischer et aI., 1984;
Borneleit et aI., 1988), and in isolated LPS (Thorne et aI., 1973; Brade and
Galanos, 1982; Brade et aI., 1987; Borneleit et aI., 1989a).
Close examination of the chemical composition and structure of the

Acinetobacter LPS has shown that it is unusual. First, in contrast to the
enteric bacteria, the LPS isolated so far from Acinetobacter spp. has been
of the R-type, i.e. it lacks an O-specific sugar chain (Rietschel et aI., 1987).
There is, however, some evidence that a small fraction of S-LPS containing
an O-antigenic chain is also present in A. calcoaceticus 69-V cells
(Borneleit et aI., 1990). Second, in A. calcoaceticus NCTC 10305, the core
oligosaccharide and lipid A component have been shown to be interlinked by
a 2-octulosonic acid instead of the KDO found in Enterobacteriaceae
(Kawahara et aI., 1987). As a consequence, the respective ketosidic linkage
is highly resistant to acid hydrolysis, and this might be advantageous in view
of the ability of many Acinetobacter strains to form organic acids from
several sugars (Juni, 1978). Furthermore, in A. calcoaceticus NCTC 10305
LPS, a previously undescribed 3-deoxyaldulosonic acid (2-keto-3-deoxy-1,7-
dicarboxyheptonic acid) was detected (Brade and Rietschel, 1985). As
another unusual feature, C 12 fatty acids were found to be attached to the lipid
A region (Brade and Galanos, 1982; Borneleit et aI., 1989a) instead of the C14
fatty acids predominant in lipid A from many Enterobacteriaceae, and this is
probably a characteristic of the Neisseriaceae (Horisberger and Dentan,
1980).
265
The special structure of the Acinetobacter LPS might be expected to
exert some effect on the physico-chemical properties of the molecule and its
behaviour as an integral constituent of the outer membrane. Indeed, the LPS
of different Acinetobacter strains has aroused interest because, in contrast
to the Enterobacteriaceae, it has been found to be liberated from the cells
into the surrounding medium in considerable amounts (Brade and Galanos,
1982). This LPS release is especially pronounced when cells are grown on
hydrophobic substrates such as long-chain hydrocarbons (Kappeli and
Finnerty, 1979; Claus et aI., 1984; Borneleit et aI., 1988). In such
circumstances, LPS has been shown to be released in the form of
sedimentahle membrane vesicles over the whole surface of the cell. This
process is, however, more complex than the simple blebbing of the complete
outer membrane structure that is sometimes observed in Enterobacteriaceae
(Nikaido and Vaara, 1985), if only because the composition of the outer
membranes and the membrane vesicles is completely different. While the
outer membranes contain about 24% LPS (weight % of the three main
constituents: protein, phospholipid, LPS), the vesicles are composed of
about 90% LPS (Borneleit et aI., 1988), and, on the basis of the KDO content
/mg protein, are about 360-fold enriched in LPS over their respective outer
membranes (Kappeli and Finnerty, 1979).
With regard to the mechanism of LPS release, a disturbance of the

interaction between LPS and outer membrane proteins is most probably not
the crucial factor. The use of an affinity electrophoresis system showed that
there was a lower dissociation constant (indicative of a stronger interaction)
for an LPS-protein complex from the LPS vesicle-forming A. calcoaceticus
69-V than for a similar complex from A. calcoaceticus CCM 5593 where no
vesicle formation has been observed (Borneleit et aI., 1989b). Instead,
fluorescence investigations indicated a clustering of the LPS molecules,
leading to the formation of large aggregates which are probably no longer
stabilised within the membrane structure (Borneleit et al.. 1990).
In Enterobacteriaceae. the structural basis for the barrier properties

of the outer membrane against hydrophobic molecules, and the lytic action
of detergents or digestive enzymes, resides in the tightly sealed outer LPS
leaflet which covers the cells (Nikaido, 1979). Therefore, as a consequence
of a disturbance of the intact lateral organisation of this leaflet by the
shedding of LPS, severe changes in the permeability properties are easily
conceivable, and susceptibility to phospholipases has been reported for
vesicle-forming cells of A. calcoaceticus 69-V (Borneleit et ai., 1988).
The LPS release, mentioned above. which is observed during B-

lactamase export across the outer membrane of A. calcoaceticus CCM 5593
is, in our opinion, in agreement with the general view that Acinetobacter
LPS is easily removed from the membrane structure, leading to a disruption
of the permeability barrier. Both disruptive processes may occur together.
This may explain the increase in production of extracellular enzymes
observed when cells are grown on hydrophobic carbon sources (Breuil et ai.,
1978; Haferburg and Kleber, 1983).
266
REFERENCES
Adams, G.A., Oadling, C, Yaguchi, M., and Tornabene, T.G., 1970, The
chemical composition of cell wall lipopolysaccharides from
M oraxella duplex and Micrococcus calcoaceticus, Can. 1.
Microbiol.,16:1.
Aurich, H., Sorger, H., and Muller, H., 1977, Isolierung und Charak-
terisierung der Zellgrenzschichten von Acinetobacter
calcoaceticus, Zeits. Allg. Mikrobiol., 17:333.
Baker, W.L., 1984, A sensitive procedure for screening microorganisms for
the presence of penicillin amidase, Aust. 1. Bioi. Sci., 37:257.
M oraxella group. II. Oxidase-negative species (genus
Acinetobacter), l. Bacterio!., 95:1520.
Benz, R., 1985, Porin from bacterial and mitochondrial outer membranes,
CRC Crit. Rev. Biochem., 19:145.
Blechschmidt, B., Borneleit, P., Emanouilidis, I., Lehmann, W., and Kleber,
H.-P., 1989, Extracellular location of a .B-Iactamase produced by
Acinetobacter calcoaceticus, Appl. Microbiol. Biotech.,
32:85.
Borneleit, P., Hermsdorf, T., Claus, R., Walther, P., and Kleber, H.-P., 1988,
Effect of hexadecane-induced vesiculation on the outer membrane
of A ci net 0 b act e rca I c 0 ace tic us, 1. G en. M i c rob i 0 I. ,
134:1983.
Borneleit, P., Blechschmidt, B., Blasig, R., Franke, P., Gunther, P., and
Kleber, H.-P., 1989a, Preparation of R-type lipopolysaccharide of
Acinetobacter calcoaceticus by EDTA-salt extraction, Curro
Microbiol., 19:77.
Borneleit, P., Blechschmidt, B., and Kleber, H.-P., 1989b, Interaction
between lipopolysaccharide and outer membrane proteins of
Acinetobacter calcoaceticus studied by an affinity
electrophoresis system, Electrophoresis, 10:234.
Borneleit, P., Binder, H., and Kleber, H.-P., 1990, Outer membrane
vesiculation of Acinetobacter calcoaceticus, Acta Biotech.,
10:127.
Brade, H., and Galanos, C, 1982, Isolation, purification and chemical
analysis of the lipopolysaccharide and lipid A of Acinetobacter
calcoaceticus NCTC 10305, Eur. 1. Biochem., 122:233.
Brade, H., and Rietschel, E.T., 1985, Identification of 2-keto-3-deoxy-1,7-
dicarboxyheptonic acid as a constituent of the lipopolysaccharide
of Acinetobacter calcoaceticus NCTC 10305, Eur. 1.
Biochem., 153:249.
Brade, H., Tacken, A., and Christian, R., 1987, Isolation and identification of
a rhamnosyl-rhamnosyl-3-deoxy-D-manno-octulosonic acid
trisaccharide from the lipopolysaccharide of Acinetobacter
calcoaceticus, Carbohyd. Res., 167:295.
Breuil, C, and Kushner, D.l., 1975a, Lipase and esterase formation by
Microbiol., 21:423.
267
Breuil, C., and Kushner, 0.1., 1975b, Partial purification and
characterization of the lipase of a facultatively psychrophilic
bacterium (Acinetobacter 016), Can. J. Microbiol., 21:434.
Breuil, c., Shindler, D.B., Sijher, 1.S., and Kushner, D.J., 1978, Stimulation
of lipase production during bacterial growth on alkanes, J.
Bacteriol.,133:60l.
Claus, R., Kappeli, 0., and Fiechter, A., 1984, Possible role of extracellular
membrane particles in hydrocarbon utilization by Acinetobacter
calcoaceticus 69-V, J. Gen. Microbiol., 130: 1035.
Claus, R., Fischer, B.E., and Kleber, H.-P., 1985, An esterase as marker
enzyme of the outer membrane of Acinetobacter
calcoaceticus, 1. Basic Microbiol., 25:299.
Fischer, B.E., 1986, Localization of hydrolytic enzymes in Acinetobacter
Fischer, B.E., and Kleber, H.-P., 1987, Isolation and characterization of the
extracellular lipase of Acinetobacter calcoaceticus 69-V, 1.
Basic Microbiol., 27:427.
Fischer, B.E., Claus, R., and Kleber, H.-P., 1984, Isolation and
characterization of the outer membrane of Acinetobacter
calcoaceticus, 1. Biotech., 1:111.
Fischer, B.E., Koslowski, R., Eichler, W., and Kleber, H.-P., 1987,
Biochemical and immunological characterization of lipase during
its secretion through cytoplasmic and outer membrane of
Acinetobacter calcoaceticus 69 V, 1. Biotech., 6:271.
Fricke, B., Jahreis, G., and Sorger, H., 1983, Characterization of membrane-
bound endoprotease activities in Acinetobacter, Wiss. Beitr.
Martin -Luther- U niv. Hall e- W ittenb erg, 52:50.
Fricke, B., Jahreis, G., Sorger, H., and Aurich, H., 1986, Cell envelope-
bound proteinase activities in Acinetobacter calcoaceticus,
Biomed. Biochim. Acta, 45:257.
Gustafsson, P., Nordstrom, K., and Normark, S., 1973, Outer penetration
barrier of Escherichia coli K-12: kinetics of the uptake of
gentian violet by wild type and envelope mutants, J. Bacteriol.,
116:893.
Guymon, L.F., and Sparling, P.F., 1975, Altered crystal violet permeability
and lytic behaviour in antibiotic-resistant and -sensitive mutants of
Neisseria gonorrhoeae, 1. Bacteriol., 124:757.
Haferburg, D., and Kleber, H.-P., 1982, Extracellular lipase from
Acinetobacter calcoaceticus, Acta Biotechnol., 2:337.
Haferburg, D., and Kleber, H.-P., 1983, Regulation of extracellular lipase of
an alkane-utilizing Acinetobacter strain, Acta Biotechnol.,
3:185.
Hancock, R.E.W., 1984, Alterations in outer membrane permeability, Ann.
Hebeler, B.H., Morse, S.A., Wong, W., and Young, F.E., 1978, Evidence for
peptidoglycan-associated protein(s) in Neisseria gonorrhoeae,
Biochem. Biophys. Res. Comm., 81:1011
Henriksen, S.D., 1973, M oraxella, Acinetob acter and M imeae,
268
Horisberger, M., and Dentan, E., 1980, Chemical composition and
ultrastructure of cellular and extracellular M oraxell a
glucidolytica lipopolysaccharides, Arch. Microbiol., 128:12.
Jahreis, G., and Aurich, H., 1989, Identification and localization of
exopeptidase activities bound to membranes of Acinetobacter
calcoaceticus, Biomed. Biochim. Acta, 48:517.
Joly-Guillou, M.L., Vallee, E., Bergogne-Berezin, E., and Philippon, A.,
1988, Distribution of B-lactamase and phenotype analysis in
clinical strains of Acinetobacter calcoaceticus, 1.
Microbiol., 32:349.
"Bergey's Manual of Systematic Bacteriology, vo!.l," p.303,
N.R.Greig and J.G. Holt, eds., Williams and Wilkins, Baltimore.
Kalogridov-Vassiliadov, D., 1984, Lipolytic activity and heat-resistance of
extracellular lipases of some gram-negative bacteria,
Milchwissenschaft, 39:601.
Kappeli, 0., and Finnerty, W.R., 1979, Partition of alkane by an extracellular
vesicle derived from hexadecane-grown Acinetobacter, 1.
Bacteriol.,140:707.
Katohda, S., Suzuki, P., Katsuki, S., and Sato, T., 1979, Studies on microbial
enzyme active in hydrolyzing yeast polysaccharides. Part II.
Purification and some properties of endo-beta-l,6-glucanase from
Acinetobacter sp., Agric. Bioi. Chem., 43:2029.
Kawahara, K., Brade, H., Rietschel, E.T., and Zahringer, U., 1987, Studies
on the chemical structure of the core-lipid A region of the
lipopolysaccharide of Acinetobacter calcoaceticus NCTC
10305. Detection of a new 2-octulosonic acid interlinking the core
oligosaccharide and lipid A component, Eur. 1. Biochem.,
163:489.
Kleber, H.-P., Schopp, W., and Aurich, H., 1973, Verwertung von n-Alkanen
durch einen Stamm von Acinetobacter calcoaceticus, Zeits.
Allg. Mikrobiol., 13:445.
Kleber, H.-P., Claus, R., and Asperger, 0., 1983, Enzymologie der n-
Alkanoxidation bei Acinetobacter, Acta Biotech., 3:251.
Lehmann, V., 1971a, Production of phospholipase C in Acinetobacter
calcoaceticus, Acta Path. Microbiol. Scand., B79:789.
Lehmann, V., 1971b, Phospholipase activity of Acinetobacter
calcoaceticus, Acta Path. Microbiol. Scand., B79:372.
Lehmann, V., 1972, Properties of purified phospholipase C from
Acinetobacter calcoaceticus, Acta Path. Microbiol.
S cand., B80:827.
Lehmann, V., 1973a, Nature of phospholipase C from Acinetobacter
calcoaceticus. Effects of whole red cells and red cell
membranes, Acta Path. Microbiol. Scand., B81:419.
Lehmann, V., 1973b, Hemolytic activity of various strains of Acinetobacter,
Acta Path. Microbiol. Scand., B81:427.
269
Ludewig, M., 1983, An aminopeptidase bound to cell envelopes of
Acinetobacter calcoaceticus, Wiss. Beitr. M artin-Luther-
Univ. Halle-Wittenberg, 52:47.
Lugtenberg, B., and van Alphen, L., 1983, Molecular architecture and
functioning of the outer membrane of Escherichia coli and other
gram-negative bacteria, Biochim. Biophys. Acta, 737:51.
Lysko, P.G., and Morse, S.A., 1981, Neisseria gonorrhoeae cell envelope:
permeability to hydrophobic molecules, I. Bacteriol., 145:946.
McDougall, G.1., Fixter, L.M., and Fewson, C.A., 1983, The occurrence of 2-
keto-3-deoxy-octonic acid in Acinetobacter calcoaceticus,
Monboisse, 1.-C., Labadie, 1., and Gouet, P., 1979a, Attachment of a
collagenolytic bacteria to its substrate, Ann. Microbiol.,
130A:435.
Monboisse, 1.-C., Labadie, 1., and Gouet, P., 1979b, Purification and
physiochemical properties of collagenase synthesized by
Acinetobacter sp. bacteria, Biochimie, 61:1169.
Nikaido, H., 1979, Permeability of the outer membrane of bacteria, Ang.
Chemie, 18:337.
Nikaido, H., and Vaara, M., 1985, Molecular basis of bacterial outer
membrane permeability, M icrobiol. Rev., 49:1.
Nishimura, Y., Ino, T., and Iizuka, H., 1986, Isolation and characterization of
the outer membrane of radiation resistant Acinetobacter sp. FO-
1, I. Gen. Appl. Microbiol., 32:177.
Onishi, H., and Hidaka, 0., 1978, Purification and properties of amylase
produced by a moderately halophilic Acinetobacter species,
Can. I. Microbiol., 24:1017.
Onishi, H., Hidaka, 0., and Nishimura, A., 1979, Halophilic alpha-amylase I
and II, lap. Kokai Tokkyo Koho, 9.
Paul, G., 101y-Guillou M.L., Bergogne-Berezin, E., Nevot, P., and
Philippon, A., 1989, Novel carbenicillin-hydrolizing B-lactamase
(CARB-5) from Acinetobacter calcoaceticus var. anitratus,
FEM S Microbiol. Lett., 59:45.
Pines, 0., Bayer, E., and Gutnick, D., 1983, Localization of emulsan-like
polymers associated with the cell surface of Acinetobacter
calcoaceticus, I. Bacteriol., 154:893.
Pugsley, A.P., and Schwartz, M., 1985, Export and secretion of proteins by
bacteria, FEMS Microbio!' Rev., 32:3.
Reisfeld, A., Rosenberg, E., and Gutnick, D., 1972, Microbial degradation
of crude oil: factors affecting the dispersion in sea water by mixed
and pure cultures, I. Appl. Microbiol., 24:363,
Rietschel, E.T., Brade, L., Schade, U., Seydel, U., Zahringer, U., Kusumoto,
S., and Brade, H., 1987, Bacterial endotoxins: Properties and
structure of biologically active domains, in "Surface Structures of
Microorganisms and their Interaction with the Mammalian Host,"
p.l, E. Schirmer, M.H. Richmond, G. Seibert and U. Schwarz, eds.,
Hoechst, Ringberg.
270
Rosenberg, E., and Kaplan, N., 1987, Surface-active properties of
Acinetobacter exopolysaccharides, in "Bacterial Outer
Membranes as Model Systems," p. 311, M. Inouye, ed., Wiley, New
York.
Sarubbi, F., Sparling, P.F., Blackman, E., and Lewis, E., 1975, Loss of low-
level antibiotic resistance in Neisseria gonorrhoeae due to env
mutations, 1. Bacteriol., 124:750.
Scott, C., Makula, R.A., and Finnerty, W.R., 1976, Isolation and
characterization of membranes from a hydrocarbon-oxidizing
Acinetobacter sp., 1. Bacteriol., 127:469.
Singer, M.E., Tyler, S.M., and Finnerty, W.R., 1985, Growth of
Acinetobacter sp. strain H01-N on hexadecanol: physiological
and ultrastructural characteristics, 1. Bacteriol., 162:162.
Sleytr, V.B., Thornley, M.J., and Glauert, A.M., 1974, Location of the
fracture faces within the cell envelope of Acinetobacter species
strain MJT/F5/5, 1. Bacteriol., 118:693.
Stewart, J.E., Kallio, R.E., Stevenson, D.P., Jones, A.C., and Schissler, D.O.,
1959, Bacterial hydrocarbon oxidation. I. Oxidation of n-
hexadecane by a gram-negative coccus, 1. Bacteriol., 78:441
Taylor, W.H., and Juni, E., 1961, Pathways for biosynthesis of bacterial
capsular polysaccharide. L Characterization of the organism and
polysaccharide, 1. Bacteriol., 81:688.
Thorne, K.J.L, Thornley, M.L, and Glauert, A.M., 1973, Chemical analysis
of the outer membrane and other layers of the cell envelope of
Acinetobacter sp., 1. Bacteriol., 116:410.
Thorne, K.J.I., Oliver, R.C., and Heath, M.F., 1976, Phospholipase A2
activity of the regularly arranged surface protein of
Acinetobacter sp. 199A, Biochim. Biophys. Acta, 450:335.
Thornley, M.J., and Glauert, A.M., 1968, Fine structure and radiation
resistance in Ac i ne tob acter: studies on a resistant strain, 1. Cell
Sci., 3:273.
Thornley, M.J., and Sleytr, V.B., 1974, Freeze-etching of the outer
membranes of Pseudomonas and Acinetobacter, Arch.
Microbiol.,100:409.
Thornley, M.J., Glauert, A.M., and Sleytr, V.B., 1973, Isolation of outer
membranes with an ordered array of subunits from
Acinetobacter, 1. Bacteriol., 115:1294.
Thornley, M.J., Thorne, K.J.I., and Glauert, A.M., 1974, Detachment and
chemical characterization of the regularly arranged subunits from
the surface of Acinetobacter, 1. Bacteriol., 118:654.
Torregrossa, R.E., Makula, R.A., and Finnerty, W.R., 1977, Outer membrane
phospholipase A from Acinetobacter sp. H01-N, 1. Bacteriol.,
131:493.
Torregrossa, R.E., Makula, R. A., and Finnerty, W.R., 1978, Hydrolysis of
phosphatidylethanolamine and phosphatidylglycerol by outer
membrane phospholipase A from Acinetobacter sp. H01-N, 1.
Zaikina, N.A., and Robakidze, T.N., 1976, Phospholipase C of fungi and
staphylococci, Mikrobiologiya, 45:466.
271
ENERGY RESERVES IN ACINETOBACTER
L.M. Fixter and M.K. Sherwani
Glasgow, G12 8QQ
UK
INTRODUCTION
The first systematic investigation of the occurrence of energy reserves

in Acinetobacter formed part of the delineation of the genus by Baumann
et al. (1968). Their study indicated that poly-B-hydroxybutyrate was not
present in any of the 106 strains examined. They examined two strains in
detail and concluded that they did not accumulate special non-nitrogenous
reserve materials. However, since this work, polyphosphates, poly-B-
hydroxybutyrate and simple wax esters, a novel energy reserve, have been
identified in a large number of acinetobacters. In several of these organisms,
the compounds show many, but not all, of the characteristics of energy
reserves. As the study of these compounds arose largely from investigations
of other aspects of acinetobacter microbiology, this review will touch on
these areas as well. The biochemistry and molecular biology of energy
reserves in Acinetobacter has not been extensively studied and so this
review will attempt to relate this area to what is known of the general
biochemistry of bacterial energy reserves.
GENERALITIES: DEFINITION AND FUNCTIONS OF

ENERGY RESERVES
Wilkinson (1959) proposed three criteria which compounds must fulfil

to be considered as energy reserves. First, they should accumulate when
growth is limited by a nutrient in the presence of excess carbon and energy
source. Second, they should be utilised when there is insufficient exogenous
carbon and energy source for growth or the maintenance of viability. Third,
their catabolism must yield energy. Certain overflow metabolites meet at
least some of these criteria and therefore only those compounds which
273
accumulate· intracellularly and meet these criteria are generally regarded as
true energy reserves. Such materials have an ancillary characteristic,
insolubility in water. This is an advantage as they do not disturb the osmotic
balance of the bacterium in which they occur and, because of this, they are
found as inclusions within bacteria (reviewed by Shively, 1974). Glycogen,
glycogen-like polymers, poly-fl-hydroxybutyrate and other poly-B-
hydroxyalkanoates meet these criteria in many bacteria (reviewed by Dawes
and Senior, 1973; Dawes, 1986; Preiss, 1989). A general role for
polyphosphate as an energy reserve is more problematical (reviewed by
Harold, 1966; Kulaev and Vagabov, 1983) and it serves simply as a phosphate
store in many bacterial species.
Energy reserves may have secondary metabolic functions depending on

their chemical structure. In Azotobacter vinelandii, poly-B-
hydroxybutyrate acts as a sink for reducing equivalents when the oxygen
supply is limited. In the presence of oxygen, its breakdown provides
substrates for the high rate of respiration which helps to protect the oxygen-
sensitive nitrogenase of this organism (Dawes and Senior, 1973). Euglena
gracilis utilises the difference in oxidation state between its two energy
reserves, paramylon (a glucose polymer) and wax esters, to allow it to carry
out wax fermentation (Inui et aI., 1982). Upon transition from aerobic to
anaerobic conditions, this organism converts its paramylon to the more
highly reduced wax esters, obtaining the use of a sink for reducing
equivalents. In the reverse transition, waxes are reconverted to paramylon
and the reducing equivalents released are used to generate energy.
Polyphosphate is a polyanion at intracellular pH and is normally found
complexed with polyamines or metal ions. Thus, it may function as a store of
ions such as magnesium (Kulaev and Vagabov, 1983).
Recent work on poly-fl-hydroxybutyrate in Escherichia coli has

shown that a compound which is generally regarded as an energy reserve may
serve a quite different role which is critically dependent on its chemical
structure. In this bacterium, only small amounts of poly-fl-hydroxybutyrate
are found and it is associated with the cytoplasmic membrane (Reusch et aI.,
1986, 1987). Its content is increased in conditions which increase the
potential of this organism for genetic transformation. Poly-fl-
hydroxy butyrate can adopt a helical configuration in hydrophobic conditions
(Dawes and Senior, 1973). Two mechanisms have been suggested to explain
how the competent state is induced. Poly-B-hydroxybutyrate helices in the
membrane may cause changes in membrane structure (Reusch et al. 1987),
allowing the entry of DNA (Reusch et ai. 1987), or transmembrane helices,
possibly in association with polyphosphate, may form channels for DNA
import (Reusch and Sadoff, 1988).
The accumulation of energy reserves when bacterial growth is

nutrient-limited in the presence of excess carbon and energy source has been
studied in batch and continuous culture. In such batch cultures, the content
of glycogen, poly-fl-hydroxybutyrate, or polyphosphate increases at about the
time of cessation of exponential growth and during the stationary phase. In
continuous cultures, the concentration of energy reserves increases at low
274
specific growth rates when the culture is limited by a nutrient other than the
carbon or energy source (Dawes and Senior, 1973). Similar results are
usually obtained irrespective of whether the cultures are limited by the
supply of nitrogen, sulphur or phosphorus. In such experiments it is also
often found that at low growth rates the specific activities of the enzymes
involved in energy reserve metabolism are increased. There are, however,
exceptions to these generalisations. Glycogen accumulation can occur
during exponential growth in some bacteria (reviewed by Dawes and Senior,
1973; Preiss, 1984). Polyphosphate accumulation probably shows the
greatest number of exceptions because of the "phosphate overplus"
phenomenon which is obtained when bacteria are transferred from
phosphate-limited conditions to phosphate-excess conditions (Harold, 1966;
Dawes and Senior, 1973). This phenomenon consists of a rapid uptake of
inorganic phosphate by the phosphate-starved bacteria on transfer to a
phosphate-rich medium containing a carbon and energy source. Some of the
inorganic phosphate taken up is converted to polyphosphate which can be
seen as volutin granules within the cell. After a short period of growth this
polyphosphate is degraded and utilised to supply phosphate for synthesis of
nucleic acids by the growing bacteria. In cases where two carbon-containing
energy reserves are present in an organism, their accumulation usually
occurs under the same type of conditions. Thus in Bacillus megaterium,
which contains both glycogen and poly-B-hydroxybutyrate, the choice of
energy reserve accumulated depends on the carbon and energy source, with
acetate favouring poly-B-hydroxybutyrate synthesis and glucose favouring
glycogen synthesis, but the reserve accumulates only when growth is limited
in the presence of the carbon and energy source (Dawes and Senior, 1973).
In Alcaligenes eutrophus, poly-B-hydroxybutyrate and polyphosphate,
which in this bacterium does not serve as an energy reserve, accumulate
together under oxygen-limiting conditions (Doi et al. 1989).
There are examples of bacteria in which accumulated carbon-

containing energy reserves maintain viability or growth during carbon and
energy source starvation. In E. coli, glycogen increases the survival rate of
the bacterium (Dawes and Senior, 1973). A high content of poly-B-
hydroxybutyrate in the nitrogen-fixing bacterium Azospirilfum brasilense
even allows the organism to grow when starved in a phosphate buffer (Tal
and Okon, 1985). There are, however, observations showing that the
presence of energy reserves in certain starvation conditions does not help to
maintain viability. In Sarcina lutea, the presence of the polyglucose
energy reserve reduces viability when the organism is starved in phosphate
buffer (Dawes and Senior, 1973). Bacteria with high levels of energy
reserves have a lower rate of degradation of proteins and RNA than bacteria
with low levels of energy reserves (Dawes and Senior, 1973). This suggests
that one of the mechanisms by which energy reserves maintain a cell's
viability during carbon and energy source starvation is by providing an
alternative energy source, thereby reducing the need to catabolise essential
cellular components. The processes that yield energy by degrading proteins
and RNA also result in the production of ammonia, which itself may have
deleterious effects on the survival of the cell (Dawes and Senior, 1973).
275
ACINETOBACTER ENERGY RESERVES
Simple Wax Esters
The first investigations of wax ester metabolism in Acinetobacter

were those of Stewart and Kallio (1959), who described the production of
wax esters during alkane oxidation by "Micrococcus cerificans" H01-N,
which was later reclassified as an acinetobacter (Baumann et aI., 1968).
During alkane oxidation, cultures of this strain can produce up to 4.3 g wax
esters/l and they are largely extracellular. The difficulty of separating
biomass from extracellular wax esters prevented the identification of
intracellular wax esters. Later work showed that the bacteria in such cultures
contained about 9 mg wax esters/g dry wt (Scott et al., 1976). The
distribution of wax esters between the intracellular and extracellular
compartments in such cultures is difficult to derive from the results
presented (Stewart and Kallio, 1959; Makula et ai. 1975). Using a molar
growth yield of 20.1 g dry wt/mol acetate (Fewson, 1985) to calculate the
biomass present in cultures, taken together with the results of Makula et ai.
(1975) for diauxic growth of this strain with acetate and hexadecane as
substrates, 97% of the wax esters produced can be estimated to be
extracellular. Thus wax esters produced during alkane oxidation are
overflow metabolites rather than energy reserves because they do not
accumulate to a significant extent within the organism. This strain contained
about 9 mg intracellular wax esters/g dry wt when grown in the absence of
alkanes (Finnerty, 1977), supporting the idea that a large proportion of the
wax esters present in alkane-grown cultures is extracellular. Attempts to
increase wax ester yield from alkanes have produced mutants of this strain
which accumulate larger amounts of wax esters than the parent strain after
growth on both alkane and non-alkane feedstocks (Geigert et aI., 1984;
Neidelman, 1987).
Gallagher (1971) provided the first report of endogenous wax esters as

part of a taxonomic study of acinetobacters which had been isolated from
lamb carcasses. The composjtion of the intact waxes that were isolated was
not reported, but the waxes contained saturated and unsaturated alkan-1-0Is
which had three to five carbon atoms or eight to 20 carbon atoms. The
proportion of short and long chain alkanols, and their degree of
un saturation, was dependent on the strain and the growth temperature.
Subsequently, Acinetobacter NCIB 8250 was shown to contain wax esters
(about 1 mg/ g dry wt) when grown with succinate or other carbon sources in
carbon-limited batch cultures (Fixter and Fewson, 1974). The waxes of this
strain have chain lengths of between 30 and 36 carbon atoms and are largely
saturated. The fatty acid and alkan-1-01 composition of the waxes, in terms
of chain length and degree of unsaturation, is very similar to the fatty acid
composition of the phospholipids of this strain, with 16 and 18 carbon
saturated species predominant. It was shown that the wax content increased
to 40-60 mg/ g dry wt in the stationary phase of nitrogen-limited batch
cultures. There was no large change in the composition of the wax esters,
and the predominant waxes contained 32 carbon atoms. The increase in the
276
wax content of nitrogen-limited stationary phase cultures, compared with
similar carbon-limited cultures, is typical of energy reserve compounds and,
solely on this basis, it was proposed that wax esters in this strain may be
energy reserves. In taxonomic investigations, Bryn and co-workers extended
the number of acinetobacters known to contain wax esters (Jantzen et aI.,
1975; Bryn et aI., 1977). The quantitative data presented in the latter paper
showed that only four of nine strains analysed contained wax esters. In these
strains the wax esters had chain lengths of between 32 and 36 carbon atoms
and were largely mono-unsaturated species. The total wax esters of strains
grown on complex solid media ranged from 1-5 mg/g dry wt. This value for
wax content is very similar to that found in other studies for either
exponentially growing or carbon-limited cultures (Fixter and McCormack,
1976; Fixter et ai. 1986). These more extensive studies showed that waxes
with similar general composition occurred in a further 17 different strains of
Acinetobacter (Fixter et aI., 1986). However, there was one significant
difference in that this work showed that strain NCTC 10306, which is
identical to strain ATCC 17907 included in the study of Bryn et ai. (1977),
contained significant amounts of wax in carbon-limited conditions (2.3 mg/ g
dry wt), while the previous authors found that this and a related strain
contained no waxes. The taxonomic potential of waxes or their alkan-l-ols is
still being investigated (Moss et aI., 1988), but it appears that clinical
isolates of Acinetobacter lack hexadecan-l-01 and octadecan-l-ol, and this
apparently indicates an absence of wax ester. This observation should be
treated with caution as it is based on the analysis of cultures grown on a
complex solid medium. Bacterial colonies isolated from this medium would
be expected to be a mixture of growing bacteria and bacteria in carbon-
limited stationary phase. The very low levels of wax esters found in such
bacteria may not yield detectable amounts of alkan-l-ols.
The composition of endogenous wax esters from Acinetobacter has

been examined in detail for only a few strains (Bryn et aI., 1977; Fixter et aI.,
1986; Neidelman, 1987). The composition of wax esters isolated from two
strains grown into nitrogen-limited stationary phase is shown in Table 1; the
wax esters are predominantly saturated 32 and 34 carbon atom species.
Determination of the isomeric composition of each wax ester species by gas
liquid chromatography / mass spectrometry for strain NCIB 8250 (Fixter et
aI., 1986) showed that, in terms of specific alkanol-fatty acid combinations,
the composition found was that expected from the random combination of
alkanols and acids present in the wax esters. The random composition
combination of wax esters has been reported in another species of wax ester-
containing bacteria (Russell and Volkman, 1980). A comparison of the
composition of waxes from strains studied by Bryn et ai. (1977) indicates
that, in these strains, mono-unsaturated and di-unsaturated species
constitute about 80% of the wax esters. This may simply reflect the different
growth temperatures, 30°C and 22°C, used in the two studies. It has been
shown that, in strain HOI-N, the degree of unsaturation of the wax esters is
dependent on growth temperature (Neidelman, 1987). In the case of cultures
grown with ethanol as substrate, the percentage of mono-and di-unsaturated
species increases from 24% when the growth temperature is 30°C to 90%
277
Table 1. The composition of wax esters isolated from Acinetobacter NCIB
8250 and Acinetobacter NCIB 10487 grown into stationary phase in
nitrogen-limited batch culture. Wax ester analysis was done using capillary
column gas liquid chromatography and the identity of the waxes was
confirmed by mass spectrometry. Tc, Trace «1% of the total). Data from
Fixter et al., (1986); by permission of the Editors of the Journal of General
Microbiology.
Wax Composition (% by weight)

ester of total wax esters
Strain Strain
NCIB 8250 NCIB 10487
30:0 2.9 2.2

30:1 0.4 0.6
31:0 Tc Tc
32:0 44.9 33.3
32:1 4.8 9.8
32:2 0.4 0.8
33:0 Tc Tc
34:0 26.9 23.1
34:1 8.5 13.6
34:2 l.9 l.9
35:0 Tc Tc
36:0 4.0 5.7
36:1 3.1 6.2
36:2 2.1 2.8
when the growth temperature is 17°C. In terms of chain length, as opposed to

degree of unsaturation, there is little difference caused by changes in growth
temperature. In the case of the wax esters produced during growth on
alkanes, the major determinant of their composition is the chain length of
the alkane substrate (Stewart and Kallio, 1959; Makula et aI., 1975; DeWitt et
al. 1982). The wax esters produced during alkane oxidation are composed
largely of alkan-1-0Is of the same chain length as the alkane, esterified to
fatty acids of the same chain length or two or four carbon atoms shorter
(Neidelman, 1987). This again indicates that such wax esters are best
regarded as overflow metabolites of alkane oxidation.
Fixter and McCormack (1976) and Fixter et ai. (1986) showed that 16
of the 17 strains which they studied had increased (5-20 fold) wax ester
contents in nitrogen-limited, compared with carbon-limited, stationary phase
cultures. This indicated the possibility that simple wax esters may be widely
used as energy reserves within this genus. Interestingly, the two strains
examined in detail by Baumann et ai. (1968) for poly-B-hydroxybutyrate,
278
strains ATCC 17902 and NCIB 8250, are among these wax-accumulating
strains, and this is possibly the reason for the failure to detect poly-.B-
hydroxybutyrate. The earliest report showing an energy reserve, possibly a
carbohydrate, in an acinetobacter (Clifton, 1937) was on the neotype strain
ATCC 23055, which is also a wax ester-accumulating strain (Fixter et al.,
1986). In strain NCIB 8250, more detailed studies (Fixter et al., 1986) showed
that bacterial wax content increased dramatically as growth ceased in
nitrogen-limited batch culture, and then increased more slowly in stationary
phase (Fig. 1). In continuous nitrogen-limited cultures, wax content
increased as specific growth rate decreased, while in carbon-limited cultures,
wax levels were lower and increased with specific growth rate. Limitation by
oxygen in continuous culture did not greatly change the wax content of this
strain (Dawes, 1986), which means that this strain is unable to use this highly
reduced reserve as a sink for reducing equivalents. Wax esters in these 16
strains meet the first criterion of energy reserves. Further, as wax esters are
degraded to carbon dioxide and water-soluble products during carbon and
energy source starvation in strain NCIB 8250 (Fixter et al. 1986), there is
evidence that wax esters meet the second criterion, since their catabolism
yields energy by the same final pathway as alkane catabolism (Asperger and
Kleber, this volume). Based on the measured rates of wax ester degradation
and the assumption that the major water-soluble product is acetate, the rapid
50
40
6'
o
U')
o
Q.
.r:::
.1 ~
Cl
10
o -t-.........or-,-rT""r-r""'..,...,.....,-.-T""!"....--r--r..,..,;-r-or-,....--f-. 01
o 5 10 15 20 25
Time (h)
Fig. 1. The accumulation of simple wax esters in a nitrogen-limited cuI ture of
Acinetobacter strain NCIB 8250 . • , Wax content; 0, growth. Reproduced
from Fixter et al., (1986) by permission of the Editors of the Journal of
General Microbiology.
279
initial phase of wax breakdown would give about 40% of the maintenance
energy of this strain, calculated on the basis of the maintenance oxygen
requirement (Hardy and Dawes, 1985). The final slower phase of wax ester
degradation, when the rate is only 10% of the initial rate, makes an
insignificant contribution to maintenance energy needs.
Wax esters at concentrations greater than 20 mg/g dry wt in

acinetobacters cause the appearance of characteristic inclusions (Fig. 2;
Fixter, 1976; Singer et al., 1984). The inclusions are plate-like and have
varying thicknesses. This means that they appear either as discs, when their
major plane is parallel to the plane of the section, or as sharp rectilinear
inclusions, when their major plane is perpendicular to the plane of the
section (Fig. 2a). Freeze-fracture techniques show the inclusions to be
multi-layered (Fig. 2b). The thicknesses of the inclusions are multiples of 4
nm, which corresponds to the thickness of a bilayer composed of wax esters
with 32-36 carbon atoms (Aleby et al. 1971). Analysis of the inclusions has
shown that wax esters are their major component in strain HOI-N grown on
hexadecanol (Singer et al., 1984). The presence of phospholipids in these
inclusions may mean that they are bounded by membranes. Thus wax esters,
like other energy reserves, occur as characteristic inclusions within the
bacterium.
1.um
Fig.2. Inclusions in Acinetobacter strain NCIB 8250. The strain was grown
into stationary phase in nitrogen-limited batch culture where it had a wax
ester content of 56.8 mg /g dry wt. Samples were prepared for transmission
electron micrography and freeze fracture electron micrography.
Transmission electron micrography (Fig. 2a) showed numerous inclusions as
discs (D) or rectilinear inclusions (R). Freeze fracture micrography (Fig.
2b) showed the plate-like nature of the inclusions and that they were layered
structures. Bar mark ~ 1 ~m.
280
I
R.COO.CHZ·R •
,
Wax ester ,'"
, '
""""
R-CHZOH ..
Alkan·l-ol ~
, '
t ;
"
R-CHO "
Alkanal Alkanal '
\
R·CO-SCoA__ - R.COOH--- R-CH3
-,,/
Fatty acyl eoA Fatty acid Hydrocarbon
\ I
CH 3 -CO-SCoA
Acetyl CoA
Fig.3. Proposed pathway for the metabolism of simple wax esters in

Acinetobacter. Synthetic reactions are indicated by solid lines and
degradative reactions by dashed lines.
The biochemistry of endogenous wax metabolism in Acinetobacter is

largely unknown, but there is evidence for the pathway in Fig. 3 which is
based on the metabolism of alkanes in acinetobacters (Aperger and Kleber,
this volume), wax metabolism in other bacteria (Albro, 1976; Lloyd and
Russell, 1983), and the pattern of labelling of the intermediates with
radioactive fatty acids and alkan-1-ols in strain NCIB 8250 (Nagi, 1981)_ The
source of fatty acids is synthesis de novo which, in most bacteria, yields fatty
acyl-ACP derivatives rather than fatty acyl-CoAs, and there is evidence that,
in E. coli, lipid biosynthesis uses fatty acyl-ACPs rather than CoA
derivatives (Rock and Cronan, 1985). Therefore, it is possible that ACP
derivatives are used in wax biosynthesis rather than the CoA esters.
Although there are well-studied enzyme systems for alkane catabolism in
acinetobacters, there are very few published observations which are directly
relevant to endogenous wax synthesis. A constitutive NAD-dependent long
chain alkanal (fatty aldehyde) dehydrogenase and an inducible NADP-
dependent alkanal dehydrogenase have been shown to occur in the
membranes of strain HOl-N (Singer and Finnerty, 1985a). In some strains
there is a constitutive NADP-linked alcohol dehydrogenase which may
reduce long chain aldehydes· to alkanols for wax synthesis (Fixter and Nagi,
1984). and similar activities have been found in other wax ester-producing
bacteria (De Bruyn et aL, 1981). The inducible NAD-dependent
dehydrogenase specific for long chain alkan-1-ols in strain H01-N (Singer
and Finnerty, 1985b) appears to be involved in exogenous alkane and
alkan-1-01 degradation because, although there is a low activity of whole-cell
hexadecanol oxidation in uninduced cultures, this enzyme could not be
detected. The breakdown of wax esters is presumably initiated by hydrolysis
of the ester link. and esterolytic activities have been identified in several
acinetobacters (Breuil and Kushner, 1975; Shabtai and Gutnick, 1985;
Sherwani and Fixter, 1989). In most of the organisms studied, the activities
appear to be secreted forms involved in the breakdown of extracellular lipids
(Breuil and Kushner, 1975; Shabtai and Gutnick, 1985), but there are
281
intracellular forms in several strains (Sherwani and Fixter, 1989). There is
evidence that the regulation of wax ester content is by changes in the rate of
wax degradation (Nagi and Fixter, 1981) rather than by changes in the rate of
synthesis, and this would be compatible with the proposal that mutants of
strain HOI-N which accumulate larger amounts of wax esters are unable to
degrade wax esters (Geigert et aI., 1984).
Polyphosphates
Biological removal of phosphate from waste-water by conventional

sewage treatments removes somewhat more than half the phosphate input
(Nesbitt, 1969). Modification of sewage treatment to include an anaerobic
treatment stage increases the efficiency of phosphate removal (Yall et aI.,
1970). In these more efficient phosphate-removing plants there have been
numerous investigations of the sewage sludge ecology to determine which
microorganisms are responsible for phosphate removal (Wentzel et aI.,
1988). Acinetobacters have been isolated from this sludge which are capable
of polyphosphate synthesis (reviewed by Stephenson, 1987) and, on this
basis, models of the phosphate-removing process in sewage sludge have been
proposed (Fuhs and Chen, 1975; Wentzel et aI., 1986). The data upon which
models and research programmes have been built pose a number of
problems. Although there is a consensus opinion that acinetobacters are
important in phosphate removal (Lotter et aI., 1986; Lotter, 1987a; Streichan
et aI., 1990), and the general view of their application in treatment of waste-
water has been summarised by Gutnick et ai. (this volume), there are
opposing views about the role of acinetobacters in this process (Brodisch and
Joyner, 1983; Suresh et a!., 1985; Roberts, 1987). Even though novel
approaches, such as estimation of the type and quantity of quinones have
been used (Hiraishi et aI., 1989), enumeration of the bacterial types within
the activated sludge remains a problem (Streichan et aI., 1990). Even the
identification of acinetobacters must be treated with caution when this is
based on a limited number of phenotypic tests (Duncan et al., 1988; Beacham
et aI., 1990). Initially, the mechanism by which anaerobic pre-treatment
increases phosphate removal in a subsequent aerobic phase centred upon the
role of an anaerobic stress in inducing extra phosphate uptake (Nicolls and
Osborn, 1979). However, more recent studies have emphasised the
significance of the production of volatile short chain fatty acids during the
anaerobic phase, and the subsequent utilisation of these fatty acids by
aerobic organisms during the aerobic phosphate-removing phase (Fuhs and
Chen, 1975; Brodisch, 1985; Ye et aI., 1988).
The acinetobacters responsible for phosphate remov(l.l accumulate

polyphosphate as typical volutin granules (Deinema et aI., 1980). The
general view is that the phosphate removed is converted to polyphosphate
within the bacterium. Whether this is the major mechanism of phosphate
removal is open to doubt as quantitative studies on activated sludge have
shown that only a third of the phosphate removed could be accounted for by
polyphosphate granules in acinetobacters (Cloete and Steyn, 1988). The
removal of phosphate and accumulation of polyphosphate in such studies
282
shows a typical feature of energy reserves in that the presence of a carbon
and energy source, whether added as acetate or already present as high
biological oxygen demand nutrients in the sewage, can improve the uptake of
phosphate and production of polyphosphate granules (Fuhs and Chen, 1975;
Brodisch, 1985; Ye et aI., 1988).
Pure cultures of acinetobacters isolated from activated sludge have

been used to investigate the culture conditions in which polyphosphate
accumulates (Deinema et aI., 1980, 1985; Murphy and Lotter, 1986).
Experiments in which the phosphate content of the bacteria is derived by
phosphate uptake from the medium, and where no evidence is presented that
the accumulated phosphate is in the polyphosphate form (Hao and Chang,
1987), must be regarded critically as it is well known that changes in growth
rates change the amount of RNA, largely ribosomal RNA, within bacteria
and hence increase their phosphate content (Tempest, 1969). In nitrogen-
limited batch cultures, the polyphosphate content of acinetobacters
increases on cessation of growth (Halvorson et aI., 1987). There is some
evidence that acetate as carbon source, rather than other carbon and energy
sources, gives the greatest accumulation of polyphosphate (Murphy and
Lotter, 1986). Rather confusing results have been obtained in continuous
culture experiments. The phosphate content of one strain (210A) does not
increase at low dilution rates in nitrogen-limited cultures, but does increase
in sulphur-limited cultures (Van Groenestijn et aI., 1989b). The effects of
limitation of potassium and magnesium ions showed that potassium ions
played a role in promoting phosphate uptake, and similar limitation in
continuous culture reduced the phosphate content of strain 210A. Although
there is good evidence that magnesium ions are the predominant counterions
for polyphosphate in acinetobacters (Van Groenestijn et aI., 1988),
magnesium limitation in continuous culture did not reduce phosphate
content dramatically, and an explanation for this is that other ions,
particularly calcium ions, may substitute for magnesium ions (Buchan, 1983;
Beacham et aI., 1990). Overall, accumulation of polyphosphate in
acinetobacters occurs as expected for an energy reserve compound, but as
explained below, it is probably simply a phosphate store in some strains.
The nature and distribution of acinetobacter polyphosphate has been

studied only in strain JWll, a phosphate-removing strain isolated from
activated sludge (Suresh et aI., 1985). The highly polymeric polyphosphate
isolated from this strain contains more than 200 phosphate residues /
molecule (Halvorson et al. 1987), which is similar to the polyphosphate
isolated from E. coli (Rao et aI., 1987). Using the effect of EDTA on 3 1p
NMR signals and the metachromatic shift of a non-penetrating dye, toluidine
blue (Suresh et aI., 1985; Halvorson et aI., 1987), it was shown that there
were two forms of polyphosphate in this strain. The bulk of the
polyphosphate was cytoplasmic, with a small fraction (about 2%) associated
with the outer membranes or periplasmic space. This bi-phasic distribution
is also found in other bacteria (Rao et aI., 1987). It is interesting to note that
some of the strains examined by Streichan et ai. (1990) show electron-dense
material associated with the bacterial wall, and this may be the periplasmic
283
polyphosphate. Suresh et al. (1986) showed that the small pool of
polyphosphate was utilised for cadmium ion transport into strain JWll, but
that the large cytoplasmic pool was not used for ion transport.
Important activities in bacterial polyphosphate metabolism include

polyphosphatases, ATP:polyphosphate phosphotransferases (polyphosphate
kinases), polyphosphate glucokinases and polyphosphate:AMP
phosphotransferases (reviewed by Kulaev and Vagabov, 1983; Wood and
Clark, 1988). The role and the existence of 1,3-bisphosphoglycerate
polyphosphate phosphotransferase is doubtful (Wood and Clark, 1988),
while the polyphosphate:NAD kinase is not considered to be central to
polyphosphate metabolism because of its restricted distribution among
bacteria and the fact that it uses cyclic rather than linear polyphosphates
(Wood and Clark, 1(88). There is evidence for polyphosphatase activity,
polyphosphate:AMP phosphotransferase, 1,3-bisphosphoglycerate poly-
phosphate phospho-transferase and polyphosphate kinase in a small number
of acinetobacters (T'Seyen et aI., 1985; Van Groenestijn et aI., 1989a;
Vasiliadis et aI., 1(90). Attempts to detect polyphosphate glucokinase and
polyphosphate :NAD kinase in several of the strains in these studies proved
fruitless. The activities of all the enzymes of polyphosphate metabolism
detected are low, in the O-l()O nmol/min/mg protein range, and in many
cases this has prevented studies of the properties of the enzymes in crude
extracts. Strain 210A, an extensively studied strain (Van Groenestijn et aI.,
1(87), contains polyphosphate:AMP phosphotransferase and poly-
phosphatase activity. However, strain 210A lacks the enzyme polyphosphate
kinase, which is known to be responsible for polyphosphate biosynthesis in
other genera of bacteria (Wood and Clark, 1988), and so
polyphosphate:AMP phosphotransferase (Eqn.1) may playa synthetic role as
well as a degradative role in this strain. The polyphosphatase activity is
enhanced by ammonium ions and, in crude extracts, 300 mM ammonium
chloride increases the activity 20-fold. The stimulated activity is about half
that of the phosphotransferase reaction in the direction of polyphosphate
degradation. Thus, in certain circumstances, polyphosphatase may be
relatively important in polyphosphate breakdown. This strain, together with
other acinetobacters, contains significant adenylate kinase activity (Eqn.2).
The two enzymes allow this organism to generate ATP from polyphosphate:
(P),+l + AMP -----> (P)n + ADP ...... (Eqn.1)
2ADP -----> ATP + AMP ............ (Eqn.2)
If the large pool of polyphosphate in the cytoplasmic compartment of the

bacterium is to serve as an energy reserve in this strain, these two reactions
are the only ones available. Calculations based on the activity of this system
in vitro show that it could supply about 25% of the maintenance energy
requirement (Van Groenestijn et aI., 1987). Adenylate kinase activity would
also be essential for the synthesis of polyphosphate in this organism as only
ADP functions as a phosphate donor, and in fact there is increased
phosphate removal by sludges with high adenylate kinase activities. Other
284
strains of Acinetobacter, such as NCIB 8250, contain a polyphosphatc
kinase and a polyphosphatase which are responsible for their polyphosphate
metabolism (Van Groenstijn et al. 1989a), and in these strains
polyphosphate is simply a phosphate reserve. The most recent work
(Vasiliadis et al., 1990) has shown a high activity of 1,3-bisphosphoglycerate
polyphosphate phosphotransferase in certain acinetobacters, and indeed the
activities reported (55-92 nmoles/min/mg protein are approximately 100-
fold higher than those found in other species of bacteria (Wood and Clark,
1988). It would seem, therefore, that these types of Acinetobacter contain
a very unusual polyphosphate-metabolising system. Polyphosphate kinase
appears to be the major enzyme responsible for the synthesis of
polyphosphate in these strains because its activity is much higher than the
polyphosphate:AMP phosphotransferase activity present. Vasiliadis et al.
(1990) propose that the 1,3-bisphosphoglycerate polyphosphate
phosphotransferase, acting in the direction 3-phosphoglycerate to 1,3-
b isphosphoglycerate, plays a role in the utilisation of polyphosphate. This
proposal is based upon the fact that the strains were grown on acetate and
would be carrying out gluconeogenesis which requires the conversion of 3-
phosphoglycerate to 1,3-bisphosphoglycerate. In view of the difficulties
which other workers have experienced in assaying 1,3-b isphosphoglycerate
polyphosphate phosphotransferase (Wood and Clark, 1988), the purification
and characterisation of this enzyme would be essential to confirm the ideas
put forward by Vasiliadis et al. (1990). The factors which regulate
polyphosphate metabolism in acinetobacters have not been identified.
Poly-B- Hydroxyb u tyra te
PolY-13-hydroxybutyrate accumulation is associated with the

polyphosphate-storing acinetobacters isolated from sewage (Fuhs and Chen,
1975; Deinema et al., 1980; Lotter et al., 1986; Lotter, 1987b; Heyman et al.
1989), but not all these strains contain polY-13-hydroxybutyrate. Although13-
hydroxybutyrate, 13-hydroxyvalerate and other 13-hydroxyalkanoates are found
in the "poly-13-hydroxybutyrate" of sewage sludge (Odham et aI., 1986), only 13-
hydroxybutyrate was detected in the one acinetobacter analysed using gas
liquid chromatography (Deinema et aI., 1980). PolY-13-hydroxybutyrate has
been postulated to playa special role in phosphate accumulation by
acinetobacters (Fuhs and Chen, 1975; Lotter et aI., 1986). The exact nature
of this role is not clear as some authors feel that poly-13-hydroxybutyrate is
important in the anaerobic zone of activated sludge systems, while others
feel that its importance is in the aerobic phase. In the first case, the
synthesis of poly-13-hydroxybutyrate during the anaerobic phase allows its use
in the subsequent aerobic phase to provide oxidisable substrate for
phosphate uptake and polyphosphate synthesis (Lotter, 1987b). This process
depends upon the ability of the acinetobacter to use nitrate as an electron
acceptor in anaerobic conditions. In the second case, the production of
volatile fatty acids during the anaerobic phase provides the conditions of
excess carbon and energy source required in the aerobic phase for the
accumulation of energy reserves, with the result that both poly-13-
hydroxy butyrate and polyphosphate accumulate.
285
[
rt ~"':O,'""'_~
HS-CoA
CHrCH-CHrCO-I
[ CHr CfH-CH2-CO-1 n CHrCHOH -CHrCOOH

0- D (-) ~-Hydro,ybutyrate
CH 3-CHOH-CH2-CO-SCoA I
D (-) ~-Hydro,ybutyryl-CoA t
~
CHrCO-CHz-COOH
Acetoacetate
CH3-CO-CHrCO-SCoA
Acetoacetyl-eoA
•
~
J 2CHr CO-SCoA
Acetyl-eoA
CH3-CHOH-CHr CO-SCoA
L (-) ~-Hydro,ybutyryl-CoA
• ,,
CHrCH=CH-CO-SCoA
Crotonyl-eoA
•
,,
CH3-CHrCHrCO-SCoA
•
Butyryl eoA
CHJ-CH2 -CH z-COOH

Butyrate
Fig. 4. ~acterial poly-B-hydroxybutyrate metabolism. Dashed lines indicate

the conversion of butyrate to poly-B-hydroxybutyrate without its metabolism
to acetate.
The pathway of poly-B-hydroxybutyrate metabolism (Fig. 4; Dawes and

Senior, 1973) has been modified by the finding that butyrate can be
incorporated directly without being converted to acetyl eoA (Doi et al.,
1989). Only in two species, Azotobacter beijerinckii and Alcaligenes
(Hydrogenomonas) eutrophus have there been detailed studies of control
of PolY-B-hydroxybutyrate metabolism at the enzyme level. The important
regulated steps are probably the conversion of acetyl eoA to acetoacetyl
eoA by 3-ketothiolase, which is inhibited by eoA (Oeding and Schlegel,
1973; Haywood et aI., 1988), and the conversion of D(-)B-hydroxybutyrate to
acetoacetate by an NAD-dependent dehydrogenase which is inhibited by
NADH, pyruvate and either oxaloacetate or 3-oxoglutarate (Senior and
Dawes 1973). The first reaction plays a key role in regulating synthesis and
the second in regulating breakdown. A B-hydroxybutyrate dehydrogenase has
been found in several acinetobacters (Lotter and Dubery, 1989), but this
activity is not well-characterised as all assays were done with D,L-B-
hydroxybutyrate. The relationship of this activity to the stereospecific
dehydrogenase of poly-B-hydroxybutyrate breakdown should be determined.
A partially purified preparation from the strain with the highest activity was
286
used to investigate the effects of acetyl CoA, NADH and oxaloacetate on the
oxidation of B-hydroxybutyrate, and these compounds did indeed inhibit the
activity, but at high concentrations (Lotter and Dubery, 1989). The fact that
this enzyme activity has similar properties to the D( -)B-hydroxybutyrate
dehydrogenase from other PolY-B-hydroxybutyrate-storing bacteria may
indicate that this enzyme is involved in poly-B-hydroxybutyrate breakdown,
but this is not certain.
SUMMARY
It would appear that there is good evidence that simple wax esters,
poly-B-hydroxybutyrate and polyphosphate occur in acinetobacters, or
bacterial strains at least partially characterised as acinetobacters. Wax
esters meet the three criteria which are used to define energy reserves: they
accumulate in conditions of carbon and energy source excess when growth is
limited by another nutrient; they are degraded during carbon and energy
source starvation; and their metabolism by the pathways known to exist in
acinetobacters would yield energy. On the basis of the known role of poly-B-
hydroxybutyrate as an energy reserve in other bacteria, this polymer can be
presumed to fulfil a similar role in acinetobacters. There are, however, few
studies on the physiology and biochemistry of this compound in well-
characterised acinetobacter strains. Polyphosphate can act as an energy
reserve in certain strains, as either the enzymes required for ATP synthesis
from polyphosphate are present, or there is evidence that polyphosphate is
utilised for ion transport in these acinetobacters. Other acinetobacters
would appear to use it simply as a phosphate store because they lack any
means of energy generation from polyphosphate. The fact that there are two
polyphosphate pools in some organisms, one of which can serve as an energy
reserve and the other as a simple phosphate store, means that
generalisations about the functions of this compound are difficult without
detailed investigations of the strains in question.
The occurrence of more than one carbon-containing energy reserve

within a single bacterial species is not unusual. Glycogen and poly-B-
hydroxybutyrate are found in pseudomonads and bacilli, and in some cases in
the same strain. This may, in the case of the bacilli, be a reflection of the
metabolic capacity of these organisms since it is known that, with acetate as
substrate, poly-B-hydroxybutyrate synthesis is favoured, while with glucose as
substrate, glycogen synthesis predominates. It is possible that, in other
cases, the choice of energy reserve depends on environmental factors which
may favour the selection of poly-B-hydroxybutyrate because it can also serve
as a sink for reducing equivalents. Are considerations such as these
influencing the choice of energy reserves among acinetobacters? It is
certainly possible because, in very simple terms, the wax ester-containing
strains were isolated from soil, water and clinical materials, while the poly-B-
hydroxybutyrate-containing strains were isolated from a single ecological
niche, activated sewage sludge. There is no evidence in the case of wax esters
that their synthesis is increased in oxygen-limiting conditions, and so they
287
would provide no extra advantage to acinetobacters from sewage plants
which cycle bacteria between aerobic and anaerobic compartments. Perhaps
poly-B-hydroxybutyrate synthesis in these strains can serve as a transient sink
for reducing equivalents during the aerobic to anaerobic transition, and this
has led to the selection o'f POlY-B-hydroxybutyrate-containing acinetobacters
in this particular niche. It would certainly be of interest to learn more about
the physiological factors regulating PolY-B-hydroxybutyrate synthesis in this
group of acinetobacters. More subtle selection pressures may also be playing
a role. The density of poly-B-hydroxybutyrate is greater than that of water,
and the density of bacteria increases significantly as their poly-B-
hydroxybutyrate content increases (Pedros-Alio et al., 1985). In contrast,
wax esters have densities of about 0.8 g/ cm3 , much less than that of water
(Sargent, 1978), and thus an increase in the wax content of organisms would
give them a greater buoyancy. The acinetobacters from sewage sludge exist
in a bacterial community associated with particulate material. It is possible
that the greater buoyancy of wax-containing strains could make it more
difficult for them to become or remain attached to this material.
The genus Acinetobacter contains a wide range of strains, and it

could be that the type of carbon-containing energy reserve is simply a
reflection of a strain's position within the genus, and thus energy reserves
could be determined by the taxonomic position of the strain within the genus.
Again, there are not really enough data to reach any firm conclusions about
this, and some of the existing results are contradictory. The survey of
Acinetobacter for the use of wax esters as reserves (Fixter et al., 1986) was
based on the classification scheme of Baumann et al. (1968), and wax-
accumulating organisms were found in both of the original major phenotypic
groups and their sub-divisions. More recent and extensive taxonomic work
has changed the internal sub-divisions -considerably (Bouvet and Grimont,
1986; Grimont and Bouvet, this volume). Only a small number of strains
from the survey of wax ester-containing acinetobacters are included in the
later classification scheme and this reduces the value of the survey. The wax
ester-containing strains are found in genospecies 1,2,4,5, 7, 9 and 11 of the
scheme proposed by Bouvet and Grimont (1986). The taxonomic position of
the polyphosphate and poly-B-hydroxybutyrate-accumulating strains is more
problematic. The acinetobacters from sewage sludge belong almost
exclusively to genospecies 5, 7 and 8/9 (Duncan et al., 1988; Beacham et al.,
1990). As polyphosphate and PolY-B-hydroxybutyrate-accumulating strains
are isolated from this environment, they are presumably members of these
genospecies, but there is no direct evidence for this. Thus it is possible that
the taxonomic distribution of these two energy reserves is more restricted.
The lack of a study focusing on the distribution of all three energy reserves in
well-characterised strains represents a major gap in our knowledge of the
genus and its physiology and biochemistry.
REFERENCES
Albro. P.W .. 1976. Bacterial waxes. in "Chemistry and Biochemistry of
288
Natural Waxes," p.419, P.E. Kolattukudy. ed .. Elsevier.
Amsterdam.
Aleby, S., Fischmeister, I., and Iyengar, B.T.R., 1971, The infrared spectra
and polymorphism of long chain esters, Lipids, 6:421.
Baumann, P., Doudoroff, M., and Stanier, R. Y., 1968, A study of the
Moraxella group. II Oxidative-negative species (genus
Acinetobacter), J. Bacterial., 95:1520.
Beacham, A.M., Seviour, R.J., Lindrea, K.C., and Livingston, 1., 1990,
Genospecies diversity of Acinetobacter isolates from a biological
nutrient removal pilot plant of a modified UCT configuration,
Water Res., 24:23.
caLcoaceticus and Acinetobacter lwoffii, into J. Syst.
Bacterial., 36:228.
Microbiol.,21:423.
Brodisch, K.E. U., 1985, Interaction of different groups of micro-organisms
in biological phosphate removal, Water Sci. Tech., 17:89.
Brodisch, K.E. U., and Joyner, S.J., 1983, The role of microorganisms other
than Acinetobacter in biological phosphate removal in activated
sludge processes, Water Sci. Tech.,15:117.
Bryn, K., Jantzen, E., and Bovre, K., 1977, Occurrence and patterns of waxes
in Neisseriaceae, J. Gen. jVlicrobiol., 102:33.
Buchan, L., 1983, Possible biological mechanism of phosphorus removal,
Water Sci. Tech., 15:87.
Clifton, C.E., 1937, On the possibility of preventing assimilation in respiring
cells, Enzymologia, 4:246.
Cloete, T.E., and Steyn, P.L., 1988, The role of Acinetobacter as a
phosphorus removing agent in activated sludge, Water Res.,
22:971
Dawes, E.A., 1985, The effect of environmental oxygen concentration on the
carbon metabolism of some aerobic bacteria, in "Environmental
Regulation of Microbial Metabolism," p.121, 1.S. Kulaev, E.A.
Dawes and D. W. Tempest, eds., Academic Press, London.
Dawes, E.A., 1986, "Microbial Energetics," p.145, Blackie, London.
Dawes, E.A., and Senior, P.J., 1973, The role and regulation of energy
reserve polymers in microorganisms, Adv. Microb. Physiol.,
10: 136.
De Bruyn, J., Johannes, A., Weckx, M., and Breumer-Jochmans, M-P., 1981,
Partial purification and characterization of an alcohol
dehydrogenase of Mycobacterium tuberculosis var. bovis
(BCG),l. Gen. Microbial., 124:359.
289
Deinema, M.H., Habets. L.H.A., Schalten. J., Turkstra, E., and Webers,
H.A.A.M., 1980. The accumulation of polyphosphate in
Acinetobacter spp. F EM S M icrobiol. Lett., 9:275.
Deinema, M.H., Van Loosdrecht, M., and Schalten, A., 1985, Some
physiological characteristics of Acinetobacter spp accumulating
large amounts of phosphate, Water Sci. Tech., 17:119.
DeWitt, S., Ervin, J.L., Howes-Orchison, D., Daletas, D., Neidelman, S.L.,
and Geigert, J., 1982, Saturated and unsaturated wax esters
produced by Acinetobacter sp. HOI-N grown on C 16 -C 20 n-
alkanes, f. Arner. Oil Chern. Soc., 59:69.
Doi, Y., Kawaguchi, Y., Nakamura, Y., and Kunioka, M., 1989, Nuclear
magnetic resonance studies of poly(3-hydroxybutyrate) and
polyphosphate metabolism in Alcaligenes eutrophus, Appl.
Environ. M icrobiol., 55:2932.
Duncan, A., Vasiliadis, G., Bayly, R.C., May, J.W., and Rapor, W.G.c.,
1988, Genospecies of Acinetobacter isolated from activated
sludge showing enhanced removal of phosphate during pilot scale
treatment of sewage, Biotech. Lett., 10:831.
Acinetobacter calcoaceticus, f. Gen. Microbiol., 131:865.
Finnerty, W.R., 1977, The biochemistry of microbial alkane oxidation, new
insights and perspectives, Trends Biochern. Sci., 2:73.
Fixter, L.M., 1976, Ultrastructural studies of Acinetobacter strains grown
in carbon and nitrogen limiting batch cultures, Proc. Soc. Gen.
Microbiol.,4:177.
Fixter, L.M., and Fewson, C.A., 1974, The accumulation of waxes by
Acinetobacter calcoaceticus NCIB 8250, Biochem. Soc.
Trans., 2:944.
Fixter, L.M., and McCormack, J.G., 1976, The effect of growth
conditions on the wax content of various strains of
Acinetobacter, Biochern. Soc. Trans., 4:504.
Fixter, L.M., and Nagi, M.N., 1984, The presence of an NADP-
dependent alcohol dehydrogenase in Acinetobacter
calcoaceticus, FEM S Microbiol. Lett., 22:297.
Fixter, L.M., Nagi, M.N., McCormack, J.G., and Fewson, C.A., 1986,
Structure, distribution and function of wax esters in
Acinetobacter calcoaceticus, f. Gen. M icrobiol., 132:3147.
Fuhs, G.W., and Chen, M., 1975, Microbiological basis of phosphate removal
in the activated sludge process for treatment of wastewater,
Microb. Ecol., 2:119.
Gallagher, LH.C., 1971, Occurrence of waxes in Acinetobacter, J. Gen.
Microbiol.,68:245.
Geigert, J., Neidleman, S.L., and DeWitt, S.K., 1984, Further aspects of wax
ester biosynthesis by Acinetobacter sp. HOI-N, f. Amer. Oil
Chern. Soc., 61:1747
Halvorson, H.O., Suresh, N., Roberts, M.F., Cocca, M., and Chikarmane,
M.M., 1987, Metabolic surface polyphosphate pool in
Acinetobacter Iwoffi, in "Phosphate Metabolism and Cellular
Regulation in Microorganisms," p.220, A Torriani-Gorini, F.G.
290
Rothman, S. Silver, A. Wright and E. Yagi!, eds., American Society
for Microbiology, Washington D.e.
Hao, 0.1., and Chang, e.H., 1987, Kinetics of growth and phosphate uptake
in pure culture studies of Acinetobacter species, Biotech.
Bioeng., 29:819.
Hardy, G.A., and Dawes, E.A., 1985, Effect of oxygen concentration on the
growth and respiratory efficiency of Acinetobacter
calcoaceticus,.!. Gen. Microbiol., 131:855.
Harold, F .M., 1966, Inorganic polyphosphates in biology: structure,
metabolism and functions, Bacteriol. Rev., 30:772.
Haywood, G.W., Anderson, A.J., Chu, L., and Dawes, B.A., 198/),
Characterization of two 3-ketotbiolases possessing differing
substrate specificities in the polyhydroxyalkanoate synthesizing
organism Alcaligenes eutrophus, FEM S Iv1 icrobiol. Lett.,
52:9l.
Heymann, 1.B., Eagle, L.M., Greben, H.A., and Potgieter, D.J.J., 1989, The
isolation and characterisation of volutin granules as subcellular
components involved in biological phosphorus removal, Water
Sci. Tech., 21:397.
Hiraishi, A., Masumune, K., and Kitamura, M., 1989, Characterisation of the
bacterial population in aerobic-anaerobic activated sludge system
on the basis of respiratory quinone profiles, App l. En vir 0 n.
M icrobiol., 55:897.
Inui, H., Miyatake, K., Nakano, Y., and Kitaoka, S .. 1982, Wax ester
fermentation in Euglena gracilis, FEBS Lett., 150:89.
Jantzen, E., Bryn, K., Bergan, T., and Bovre, K., 1975, Gas chromatography
of bacterial whole cell methanolysates VII. Fatty acid composition
of Acinetobacter in relation to the taxonomy of Neisseriaceae,
Acta Path. Microbiol. Scand., 83B:569.
Kulaev, I.S., and Vagabov, V.M., 1983, Polyphosphate metabolism in
microorganisms, Adv. M icrob. Physiol., 4:83.
Lloyd, G.M., and Russell, N.J., 1983, Biosynthesis of wax esters in the
psychrophilic bacterium Micrococclls cryophilus, .I. Gen.
M icrobiol., 129:264l.
Lotter, L.H., 1987a, Metabolic behaviour of Acinetobacter spp. in
enhanced biological phosphorus removal - a biochemical model.
Reply to comments, Water SA, 13:251.
Lotter, L.H., 1987b, Preliminary observations on polyhydroxybutyrate
metabolism in the activated sludge process, Water SA, 13: 189.
Lotter, L.H., and Dubery, LA., 1989, Metabolic regulation of beta-hydroxy-
butyrate dehydrogenase in Acinetobacter calcoaceticus var
lwoffi, Water SA, 15:65.
Lotter, L.H., Wentzel, M.e., Loewenthal, L.E., Bkama, G.A., and Marais,
GvR., 1986, A study of selected characteristics of Ac i n e to b a c te r
spp. isolated from activated sludge in aerobic/ anoxic/ aerobic and
aerobic systems, Water SA, 12:203.
Makula, R.A., Lockwood, P.]., and Finnerty, W.R., 1975, Comparative
analysis of lipids of Acinetobacter grown on hexadecane, .I.
Bacteriol.,121:250.
291
Moss. CW .. Wallace. P.L.. Hollis. D.G .. and Weaver. R.E .. 1988. Cultural
and chemical characterization of CDC groups EO-2. M-5. and M-6.
Moraxella species. Olifiella urethralis. Acinetobacter
species. and Psychrobacter immobilis, J. Clin. Microbiol .•
26:484.
Murphy, M., and Lotter, L.H., 1986, The effect of acetate and succinate on
polyphosphate formation and degradation in activated sludge with
particular reference to Acinetobacter calcoaceticus, Appl.
Microbiol. Biotech., 24:512.
Nagi, M.N., 1981, Some studies on wax esters of Acinetobacter
calcoaceticus, PhD thesis, University of Glasgow, UK.
Nagi, M.N., and Fixter. L.M., 1981, Regulation of the wax content of
Acinetobacter calcoaceticus. NCIB 8250, Abstr. Ann. Meet.
Amer. Soc. Microbio!., Kll
Neidleman, S.L., 1987, Effects of temperature on lipid unsaturation,
Biotech. Genet. Eng. Rev., 5:245.
Nesbitt, J.B., 1969, Phosphorus removal, the state of the art, J. Water
Pollut. Cont. Fed., 41:701.
Nicolls, H.A., and Osborn, D.W., 1979, Bacterial stress, a prerequisite for
biological removal of phosphorus, J. Water Pollut. Cont. Fed.,
51:557.
Odham, G., Tunlid, A., Westerdahl, G., and Marden, P., 1986, Combined
determination of poly-J3-hydroxyalkanoic acid and cellular fatty
acids in starved marine bacteria and sewage sludge by gas
chromatography with flame ionization or mass spectrometry
detection, Appl. Environ. jV1 icrobiol., 52:905.
Oeding, V., and Schlegel, H.G., 1973, J3-Ketothiolase from
Hydrogenomonas eutropha H16 and its significance in the
regulation of poly-J3-hydroxybutyrate metabolism, Bi 0 c hem. .I.,
134:239.
Pedros-Alio, C, Mas, J., and Guerrero, R., 1985. The influence of poly-J3-
hydroxybutyrate accumulation on cell volume and buoyant density
in Alcaligenes eutrophus, Arch. Microbiol., 143:178.
Preiss, J., 1984, Bacterial glycogen synthesis and its regulation,
Ann. Rev. M icrobiol., 38:419.
Preiss, J., 1989, Chemistry and metabolism of intracellular reserves, in
"Bacteria in Nature, vol. 3," p.189, 1.S. Poindexter and E.R.
Leadbetter, eds. Plenum Press, New York.
Rao, N.N., Roberts, M.F., and Torriani, A., 1987 Polyphosphate
accumulation and metabolism in Escherichia coli, in "Phosphate
Metabolism and Cellular Regulation in Microorganisms," p.213, A.
Torriani-Gorini, F.G. Rothman, S. Silver, A. Wright and E. Yagil,
eds., American Society for Microbiology, Washington D.C
Reusch, R.N., and Sadoff, H.L., 1988, Putative structure and functions of a
poly-J3-hydroxybutyrate-calcium polyphosphate channel in
bacterial plasma membranes, Proc. N atl. Acad. Sci. USA.,
85:4176.
Reusch, R.N., Hiske, T.W., and Sadoff, H.L., 1986, Poly-J3-hydroxybutyrate
membrane structure and its relationship to genetic transformation
in Escherichia coli, J. Bacterio!., 168:553.
292
Reusch, R., Hiske, T., Sadoff, H., Harris, R., and Beveridge, T., 1987,
Cellular incorporation of polY-B-hydroxybutyrate into plasma
membranes of Escherichia coli and Azotobacter vinelandii
alters native membrane structure, Can. 1. Microbiol., 33:435.
Roberts, M.R., 1987, Metabolic behaviour of Acinetob acter spp. in
enhanced biological phosphorus removal - a biochemical model.
Comments. Water SA., 13:251.
Rock, C.O., and Cronan, J.E., 1985, Lipid metabolism in prokaryotes, in
"Biochemistry of Lipids and Membranes," p.73, D.E. Vance and
J.E. Vance, eds., Benjamin/Cummings, Menlo Park, California.
Russell, N.J., and Volkman, J.K., 1980, The effect of growth temperature on
wax ester composition in psychrophilic bacterium Micrococcus
cryophilusATCC 15174,1. Gen. Microbiol., 118:131.
Sargent, J.R., 1978, Marine wax esters, Sci. Prog., 65:437.
Scott, c.c.L., Makula, R.A., and Finnerty, W.R., 1976, Isolation and
characterization of membranes from a hydrocarbon-oxidizing
Acinetobacter sp., J. Bacteriol., 127:469.
Senior, P.J., and Dawes, E.A., 1973, The regulation of poly-B-
hydroxybutyrate metabolism in Azotobacter beijerinckii,
Biochem. J., 134:225.
Shabtai, Y., and Gutnick, D.L., 1985, Exocellular esterase and emulsan
release from the cell surface of Acinetobacter calcoaceticus,
Sherwani, M.K., and Fixter, L.M., 1989, Multiple forms of carboxylesterase
activity in Acinetobacter calcoaceti cus, F EM S M icrob iol.
Lett., 58:75.
Shively, 1.M., 1974, Inclusion bodies of prokaryotes, Ann. Rev. Microbiol.,
28:167.
Singer, M.E., and Finnerty, W.R., 1985a, Fatty aldehyde dehydrogenases in
Acinetobacter sp. strain HO-1N: role in hexadecane and
hexadecanol metabolism, J. Bacterial., 164: 1011.
Singer, M.E., and Finnerty, W.R., 1985b, Alcohol dehydrogenases in
Acinetobacter sp. strain HO-1N: role in hexadecane and
hexadecanol metabolism, J. Bacteriol., 164:1017.
Acinetobacter sp. strain H01-N on hexadecanol: physiological
and ultrastructural characteristics, J. Bacteriol., 162:162.
Stephenson, T., 1987, Acinetobacter: its role in biological phosphate
removal, in "Biological Phosphate Removal from Wastewaters," p.
313, R. Ramadori, ed., Pergamon Press, Oxford.
Stewart, J.E., and Kallio, R.E., 1959, Bacterial hydrocarbon oxidation. II
Ester formation from alkanes, 1. Bacteriol., 78:726.
Streichan, M., Golecki, J.R., and Schon, G., 1990, Polyphosphate removal
from sewage plants with different processes for biological
phosphorus removal, FEM S Microbiol. Ecol., 73:113.
Suresh, N., Warburg, R., ,Timmerman, M., Wells, J., Coccia,M., Roberts,
M.F., and Halvorson, H.O., 1985, New strategies for the isolation
of microorganisms for phosphate accumulation, Water Sci.
Tech., 17:99.
293
Suresh, N., Roberts, M.F., Coccia, M. Chikarmane, H.M., and Halvorson,
H.O., 1986, Cadmium-induced loss of surface polyphosphate in
Adnetobacter lwoffi" FEMS Microbiol. Lett., 36:91.
Tal, S., and Okon, Y., 1985, Production of the reserve material, poly-B-
hydroxybutyrate and its function in Azospirillum brasiliense,
Can. f. Microbiol., 31:608.
Tempest, D.W., 1969, Quantitative relationship between inorganic cations
and anionic polymers in growing bacteria, in "Microbial Growth,
19th Symp. Soc. Gen. Microbiol.," p.87, P.M. Meadow and S.J.
Pirtm, eds., Cambridge University Press, Cambridge.
T'Seyen, J., Malnou, D., Block, J.e., and Faup, G., 1985, Polyphosphate
kinase activity during phosphate uptake by bacteria, Water S d.
Tech 17:43.
Van Groenestijn, J.W., Deinema, M.M., and Zehnder, A.J.B., 1987, ATP
production from polyphosphate in Acinetobacter strain 210A,
Arch. Microbiol., 148:14.
Van Groenestijn, J.W., Vlekke, G.F.M., Anink, D.M.E. Deinema, M.M., and
Zehnder, A.J.B., 1988, Role of cations in the accumulation and
release of phosphate by Acinetobacter strain 210A, Appl.
Van Groenestijn, J.W., Bentvelsen, M.M.A., Deinema, M.M., and Zehnder,
A.J.B., 1989a, Polyphosphate-degrading enzymes in
Acinetobacter spp. and activated sludge, Appl. Environ.
Microbiol., 55:219.
Van Groenestijn, J.W., Zuidena, M., Van de Worp, J.J.M., Deinema, M.M.,
and Zehnder, A.J.B., 1989b, Influence of environmental
parameters on polyphosphate accumulation in Acinetobacter sp.,
Ant. v. Leeuw. f. Microbiol., 55:67.
Vasiliadis, G., Duncan, A., Bayly, R.e., and May, J.W., 1990, Polyphosphate
production by strains of Acinetobacter, FEMS Microbiol.
Lett., 70:37.
Wentzel, M.e., Lotter, L.H., Loewenthal, R.E., and Marais. Gv.R., 1986,
Metabolic behaviour of Acinetobacter spp. in enhanced
biological phosphorus removal. A biochemical model, Water SA,
12:209.
Wentzel, M.e., Loewenthal, R.E., Ekama, G.A., and Marais, G.V.R., 1988,
Enhanced polyphosphate organism cultures in activated sludge
systems - Part I Enhanced culture development, Water SA,
14:14l.
Wilkinson, J.F., 1959, The problem of energy-storage compounds in bacteria,
Exp. Cell Res., Suppl. 7:11l.
Wood, H.G., and Clark, J.E., 1988, Aspects of inorganic polyphosphates,
Ann. Rev. Biochem., 57:235.
YaH, I., Ba!lgwan, W.H., Knudsen, e. and Sinclair, N.A., 1970, Biological
uptake of phosphate by activated sludge, Appl. Microbiol.,
20:145.
Ye, Q., Ohtake, H., and Toda, K., 1988, Phosphorus removal by pure and
mixed cultures of microorganisms, 1. Ferment. Tech., 66:207.
294
ENERGY GENERATION AND THE GLUCOSE DEHYDROGENASE
PATHWAY IN ACINETOBACTER
l.A. Duine
Vakgroep Microbiologie & Enzymologie

Technische Universiteit Delft
lulianalaan 67
2628 BC Delft
The Netherlands
INTRODUCTION
Direct, non-phosphorylative, bacterial glucose oxidation was

discovered nearly half a century ago (Barron and Friedemann, 1941;
Lockwood et aI., 1941). The possession of such a property was deduced from
the fact that gluconic and ketogluconic acids were formed from glucose. The
first step in this pathway is the oxidation of glucose at the C , position to form
1,5-gluconolactone, which is hydrolysed by enzyme action or spontaneously
to form gluconic acid; direct glucose oxidation can also occur at other
positions, but these conversions are not relevant here. The reaction is
catalysed by a glucose dehydrogenase, whereas glucose oxidase (EC 1.1.3.4)
occurs only in yeasts and fungi. The enzyme belongs either to the category of
NAD(P)-dependent dehydrogenases or to the dye-linked dehydrogenases.
In the former case, NAD(P) is reduced in the reaction and then re-converted
into NAD by the NADH dehydrogenase of the respiratory chain; in the latter
case, the reduction equivalents on the cofactor are directly transferred in
vivo to a component of the respiratory chain, and in vitro to an artificial
dye. During the 1950's, the latter type of glucose dehydrogenase was
discovered in many Pseudomonas and Acetobacter strains, and also in
Bacterium anitratum. It should be noted that the bacterial
dehydrogenase discussed here is not the flavoprotein glucose dehydrogenase
(EC 1.1.99.10) which has only been reported to occur in Aspergillus oryzae
(Bak,1967).
295
Historical Overview
Bacterium anitratum, nowadays known as Acinetobacter

calcoaceticus, was used by Hauge and colleagues to obtain the dye-linked
glucose dehydrogenase and study its properties. In a series of papers it was
demonstrated that the enzyme contained a co-factor whose identity was not
known at that time (Hauge, 1960a,b, 1961, 1964, 1966a,b; Hauge and
Hallberg, 1964; Hauge and Murer, 1964). The presence of the co-factor
could easily be detected by the ability to reconstitute the glucose
dehydrogenase activity in cell-free extracts prepared from anaerobically-
grown Rhodopseudomonas (now Rhodobacter) sphaeroides
(Niederpruem and Doudoroff, 1965). Based on this assay, it was shown that
glucose dehydrogenase containing this co-factor occurs in other aerobically-
grown Gram-negative bacteria. In addition, the assay demonstrated the
natural occurrence of the glucose dehydrogenase apo-enzyme.
From what is currently known, the choice of the particular B.

anitratum strain (provided to us by Prof. J.G. Hauge and deposited in our
cultwe collection, first as A. calcoaceticus LMD 79.39 and later as LMD
79.41) may not have been fortuitous as it contains a type of glucose
dehydrogenase that is most easily purified, as described below. On the other
hand, only low levels of this enzyme are produced, which could be the reason
why identification of its co-factor was achieved only many years later. Thus,
15 years after the peculiar properties of the co-factor of methanol
dehydrogenase had been described (Anthony and Zatman, 1964), the
similarity to the co-factor of glucose dehydrogenase was noticed. both by us
and by Imanaga et al. (1979). After the elucidation of the co-factor
structure of methanol dehydrogenase (Salisbury et aI., 1979; Duine and
Frank. 1980), that of glucose dehydrogenase was also shown to be 2,7,9-
tricarboxyl-l H -pyrrolo [2,3-fl-quinoline-4,5 -dione, or pyrroloquino line
quinone (POO) (Duine et aI., 1979). In the past ten years, the quinoprotein
nature (a quinoprotein is defined as an enzyme with PQO as a co-factor) of
glucose dehydrogenase from many other Gram-negative bacteria has been
established. Although it has been reported that the glucose dehydrogenase
from the Gram-positive organism Zymomonas mobilis is also a
quinoprotein (Strohdeicher et aI., 1989), the enzyme is probably quite
different because the co-factor is covalently bound.
Glucose Dehydrogenase: Different Forms or Different Enzymes?
It was noted quite early (Hauge, 1966a) that glucose dehydrogenase

activity present in a cell-free extract could be separated into two fractions by
ion exchange chromatography; one fraction being adsorbed, the other not.
Another distinction could be made on "solubility", as centrifugation studies
suggested that one part of the activity was soluble (i.e. remaining in the
supernatant), while the other part remained attached to membranes (in the
pellet). Although these fractions showed different substrate and electron
acceptor activities, they were considered to derive from one and the same
enzyme since the properties of the particle-bound form could be changed
296
into those of the soluble form by detergent treatment (Hauge and Haliberg,
1964; Hauge, 1966a). On the basis of what is currently known, namely that
two quite different glucose dehydrogenases occur in this strain (soluble and
membrane-bound glucose dehydrogenases, represented by "s-GDH" and "m-
GDH" respectively), this observation is difficult to understand. The
"transformation" could be explained by assuming that both s-GDH and m-
GDH were present in the pellet and that m-GDH activity was selectively
destroyed by the treatment; membrane preparations do indeed show activity
with disaccharides and artificial electron acceptors (see e.g. Matsushita et
aI., 1988), and this points to the presence of s-GDH. In studies carried out
with several Acinetobacter strains, it was concluded that different forms of
glucose dehydrogenase exist (Duine et aI., 1982). However, the data on the
induction of these forms in the original strain (at that time deposited as
LMD 79.39) may have been biased by the fact that, as was later discovered,
the culture slant was contaminated with a Pseudomonas putida strain,
leading to the erroneous conclusion that A. calcoaceticus contains a
quinoprotein alcohol dehydrogenase (Duine and Frank, 1981; Beardmore-
Gray and Anthony, 1983; B.W. Groen, R.A. van der Meer and J.A. Duine,
unpublished results). On the other hand, it is clear that the glucose
dehydrogenases from the Acinetobacter strains are dissimilar. Different
glucose dehydrogenases have also been reported to occur simultaneously in
certain Pseudomonas species (Hauge, 1966a; Hayaishi, 1966; Vicente and
Canovas, 1973). It remains to be investigated whether these phenomena can
be explained by the presence of both an m-GDH and an s-GDH, or by a
single glucose dehydrogenase manifesting itself in different forms because it
is present in different environments. The curious result that the substrate
specificity for aldose sugar oxidation by whole cells is quite different from
that observed in cell-free extracts (Dokter et aI., 1987) can now be explained
from the fact that s-GDH shows no activity in vivo, as discussed below.
Scope and Limitations
When reading the literature on the glucose dehydrogenase of the

A. calcoaceticus strain originally used by Hauge, the conclusions in the
early papers should be reinterpreted in the light of what it now known about
s-GDH and m-GDH. Since glucose dehydrogenases from other strains have
scarcely been studied, it is not known whether this caution should apply in
general.
Detection of the glucose oxidation pathway in Acinetobacter strains

is achieved by measuring growth on glucose or by measuring acid formation
under aerobic conditions in special media supplemented with glucose.
However, identification of the enzyme responsible for the observed
phenomenon has generally been neglected. Moreover, the non-acid forming
strains (frequently classed as A. lwoffii) should be reinvestigated as they
may produce only the protein part of glucose dehydrogenase, which can be
detected by screening with and without POO (van Schie et aI., 1984).
Although this approach has recently been followed by Bouvet and Bouvet
(1989), it will be already clear that data on the distribution of the enzyme in
297
this genus cannot be fully trusted. Similarly, since most of the work has been
carried out with A. calcoaceticus strain LMD 79.41, the conclusions drawn
in this review apply only to this strain, unless explicitly indicated otherwise,
and care should be taken in making generalisations. In this connection, it
should be mentioned that this strain does not grow on glucose or gluconic
acid. Furthermore, compared with the sub-group which does have these
characteristics (Baumann et aI., 1968), it shows a somewhat atypical
behaviour as it does not grow on pentoses and does not produce lactobionic
acid from lactose. A recent identification by the criteria of Bouvet and
Grimont (1986) in two separate laboratories has now revealed that strain
LMD 79.41 belongs to either Acinetobacter genospecies 2 or 3 (Dijkshoorn
et aI., 1990).
THE SOLUBLE TYPE OF GLUCOSE DEHYDROGENASE (s-GDH)
Although the distinction between s-GDH and m-GDH has only

recently been made, from the properties reported it can be deduced that the
enzyme investigated by Hauge was s-GDH. Since classification of this
glucose dehydrogenase was carried out some years ago, it seems logical to
reserve the EC number 1.1.99.17 for s-GDH and to give m-GDH a novel EC
number.
General Properties
s-GDH has been purified (Dokter et aI., 1986; Geiger and Gorisch,
1986) and most of its properties are shown in Table 1. It is a dimeric,
strongly basic protein with a broad substrate specificity. Thus, not only
pentose and hexose sugars, but also disaccharides are oxidised; in this
respect, aldose dehydrogenase would be a better name. There is a
preference for the fl-anomer of glucose and the product formed is the 1,5-
gluconolactone (Hauge, 1960b). A substantial isotopic effect on the
reaction rate was observed with glucose labelled at the C 1 position with 2H
(Hauge et aI., 1968). Several artificial electron acceptors exist for the
enzyme (Hauge, 1966b). All are cationic or neutral species, as deduced
from the fact that 2,6-dichlorophenolindophenol (DCPIP) is active only
below pH 6.5, but the pH optimum observed with Wurster's Blue is 9.0
(Dokter et aI., 1986). In a search for the natural electron acceptor, only a
cytochrome appeared to be active, not ubiquinones. Purification revealed
that it is a periplasmic, small, basic haemoprotein, called cytochrome b S62 '
and similar to a cytochrome from Escherichia coli with an unknown
function (Dokter et aI., 1988). A. calcoaceticus does not contain
cytochromes c. A start has been made on unravelling the cytochrome b
complexes (Geerlof et aI., 1989; Dokter et aI., 1990).
The mechanism of action seems straightforward, namely reaction of

the oxidised form with the substrate to yield the 2-electron reduced form.
Stopped flow experiments revealed a fluorescing intermediate (1. Frank and
J.A.Duine, unpublished results) which could be a POO-substrate complex,
298
Table 1. Comparison of the properties of s-GDH and m-GDH
s-GDH m-GDH
Mr (enzyme) 100,462 86,956

Mr (subunit) 50,231 86.956
Cofactors (per enzyme 2 PQQ x PQQ
molecule) 4 Ca 2+ x Ca 2+jMg2+?
Metal ions in reconstitution Ca 2+(Mn2+,Cd 2+) Mg2+, Ca 2+
Removal of PQQ by chelators no yes
Midpoint potential (pH 7) +50 mV ?
Isoelectric point 9.5 (low)
Substrate specificity
glucose + +
2-deoxyglucose +
lactose +
maltose +
pH optimum 9.0 (8.0,4.6)
Substrate inhibition yes not observed
Natural electron acceptor cytochrome b 562 ubiquinone (Q9)
Location periplasm membrane-integrated
(outside)
just as has been observed for methanol dehydrogenase (Frank et aI., 1989).
Ca2+ ions are required for the reconstitution of the apo-form of s-GDH to
active holo-enzyme (Geiger and Gorisch, 1989; B.W. Groen and I.A. Duine,
unpublished results). Absorption spectra of the redox forms of the enzyme
are depicted in Fig.I. The semiquinone form, which should theoretically
occur in the oxidative half reaction in which reduced glucose dehydrogenase
reacts with one-electron acceptors of the respiratory chain, has not been
studied to date.
Structural Aspects
In reconstitution studies, Ca'+ can be replaced by Mn'+ and Cd 2+,

but not by Mg2+. Four Ca 2+ ions occur per enzyme molecule, and these
cannot be removed from the holo-enzyme with chelators. The cofactor and
the metal ions also playa structural role since they are required in the
refolding process after heat denaturation, and Ca 2+ stabilises the enzyme
against this denaturation (Geiger and Gorisch, 1989). Detachment of PQQ
requires rather drastic conditions (Hauge and Miirer, 1964; Duine et aI.,
1979; Kilty et aI., 1982) and the effects are not always reproducible (l.A.
Duine, unpublished results). As it has only recently become apparent that
both the metal ion and PQQ are required for reconstitution, past
experiments in which apo-enzyme was prepared to assay PQQ, or to establish
299
0.12
:::c
<1l 0.08
-E0
'"
-"
<J:
004
000
300 400 500 600
00
300 400 500 600 700

Wavelength (nrn)
Fig.l. Absorption spectra of s-GDH: , reduced form (maximum 338 nm);

----, oxidised form (maximum 350 nm).
activity of POO-analogues, may have suffered from Jack of knowledge in this

respect. Thus, although it was claimed that synthetic phenanthroline
quinones are active in reconstitution (Duine et aI., 1980; Conlin et aI.,
1985), this could not be confirmed in later experiments (J. Frank and J.A.
Duine, unpublished results). A plausible explanation could be that the
samples tested contained the tiny amount of Ca'+ required for
reconstitution and that POO was not completely removed from the apo-
enzyme preparation.
The gene of s-GDH has been cloned, sequenced and expressed in an

E. coli strain (Cleton-Jansen et aI., 1989). It contains a leader sequence of
24 amino acids which has the characteristics of a signal peptide. Growth in
the presence of the detergent Triton X-IOO had already shown that s-GDH
occurs in the periplasm, as deduced from its presence in the culture fluid
under these conditions (Dokter et aI., 1985). Similar observations have been
made recently for strain 69- V where selective detachment of the activity (m-
GDH?) depended on the detergent used (Hommel et aI., 1989).
Applications
The gene of s-GDH has been combined with a strong promoter and
expressed in E. coli so that substantial amounts of apo-enzyme could be
obtained; there was no holo-enzyme since E. coli does not produce free
POO (Cleton-Jansen et aI., 1989; van der Meer et aI., 1990). Since the holo-
enzyme has an extremely high turnover number of around 400,000 /min
(Hauge, 1966b; Dokter et aI., 1986), a sensitive assay for POQ could be
developed with these preparations (van der Meer et aI., 1990). Production
of the enzyme may also be important for other applications as the natural
holo-enzyme has been applied in an amperometric biosensor; the high
catalytic activity and oxygen-insensitivity of s-GDH are attractive aspects
compared with glucose oxidase (D'Costa et aI., 1986). Studies on the
binding and catalysis of POO analogues have been carried out with glucose
300
dehydrogenase from E. coli (Shinagawa et aI., 1986) and are in progress with
apo-s-GDH from A. calcoaceticus.
A Physiological Role?
The typical substrate specificity of s-GDH is reflected in that of cell-

free extracts, not in the oxidation pattern of whole cells (Dokter et aI., 1987).
A mutant unable to produce s-GDH showed the same substrate specificity as
wild-type cells and no unusual growth behaviour (Cleton-Jansen et aI.,
1988b). This leads to the conclusions that s-GDH is not involved in cellular
sugar oxidation and has no role at all in the organism or is operating in a
dispensable function. Since both s-GDH and cytochrome b S62 are localised
in the periplasm, the non-functionality is not related to transport problems
for the substrate. Thus, an essential component for electron transfer to the
respiratory chain may be lacking which could have been lost in the
evolutionary history of this organism; this assumed incompleteness is in
accordance with the low reaction rate of reduced s-GDH with cytochrome
b 562 in vitro (Dokter et aI., 1988). However, this assumption leads to the
curious implication that in the past the organism would have possessed the
ability to oxidise glucose with two dehydrogenases having the same cofactor,
but no structural similarity of the protein components. However, the real
conclusion seems to be that, despite all the research effort that has been
concentrated on the characterisation of s-GDH from this strain, the
physiological role of the enzyme, if any, is completely unknown.
MEMBRANE-BOUND GLUCOSE DEHYDROGENASE (m-GDH)
Since m-GDH was previously thought to be identical with s-GDH, no

real efforts were made in the past to characterise it. Another reason for its
elusiveness is the hydrophobic nature of the protein, requiring detergents to
remove it from the membranes and making purification difficult (Matsushita
et aI., 1989). Now, 30 years after its discovery (Hauge, 1960), it is realised
that this enzyme should have been preferentially studied as it is the one
which is responsible for cellular glucose oxidation.
General Properties
From the primary structure deduced from the DNA sequence (Cleton-
Jansen et aI., 1988a), it appears that the N-terminal part of the enzyme (140
amino acids) is very hydrophobic. Five hydrophobic regions, separated from
each other by hydrophilic amino acids, are putative membrane spanning
segments. There is no signal sequence and it has been suggested that the N-
terminal part anchors the enzyme to the cytoplasmic membrane (Cleton-
Jansen et aI., 1988b). Attachment should occur in such a way that the active
site is still available for sugars coming from the outside because no transport
mechanisms have been found. A domain motif has been found which also
occurs in mitochondrial NADH dehydrogenase, coding for interaction with
301
ubiquinone (Friedrich et aI., 1990). Although the gene has been cloned and
expressed in an E. coli strain, production of adequate amounts of the
enzyme is still problematic since the cells were killed in attempts to achieve
high expression (A-M. Cleton-Jansen and B.W. Groen, unpublished results).
Comparison with data from the literature suggests that m-GDH is

widespread among Gram-negative bacteria (van Schie et aI., 1987a). This
has been further substantiated for a range of bacteria by showing that an
immunological cross-reaction occurs between the antibody directed against
m-GDH from A. calcoaceticus (Matsushita et aI., 1986) and the glucose
dehydrogenases from E. coli and Pseudomonas aeruginosa, where the
genes for the enzyme showed sequence similarity with that of A.
calcoaceticus (Cleton-Jansen et aI., 1988b). In studies on the partially
purified enzyme from Ps. fluorescens, selective protein modification
indicated several amino acids in the active site which are essential for
binding of PQQ or for activity (Imanaga, 1989).
The enzyme has been purified and partly characterised (Table

1; Matsushita et aI., 1989). It is a monomeric protein with a
narrower substrate specificity and lower catalytic activity than s-
GDH. The substrate specificity of m-GDH from other bacteria is
variable, ranging from strict specificity for glucose to very broad
specificity, including disaccharides. Evidence has been provided
that the natural electron acceptor is ubiquinone Q 9 (Matsushita et aI.,
1989). If it appears that the redox potential of m-GDH is as high as that of s-
GDH (+ 50 mY; Dokter et aI., 1988), problems may arise if the oxidation
state of the respiratory chain is inadequate. Perhaps the decrease of the
cellular glucose oxidation rate, but not of the ethanol oxidation rate, at low
oxygen tensions (Dokter et aI., 1987) is an indication for that. Binding of
PQQ also requires a bivalent metal ion (Ca 2+ or Mg2+). In contrast to s-
GDH, the metal ion, and concomitantly PQQ, could be easily removed by a
chelator from m-GDH of Ps. aeruginosa, enabling easy preparation of an
apo-enzyme suited for PQQ determinations (Duine et aI., 1983). m-GDH
apo-enzyme is present in A.lwoffii (van Schie et aI., 1984, 1987a) and, of
course, in PQQ- mutants (van Kleef et aI., 1987). Thus whole cells of these
strains can be used for the detection of PQQ. Upon transfer of the genes for
PQQ biosynthesis (obtained from A. calcoaceticus) to A. lwoffii, holo-m-
GDH was produced (Goosen et aI., 1989).
Function in the Entner-Doudoroff Route
Glycolysis has not been detected in A. calcoaceticus, probably

because of the absence of hexokinase. A few percent of the strains grow on
glucose via the Entner-Doudoroff pathway (Juni, 1978). m-GDH is required
in these cases to provide the gluconic acid. Most strains do not grow on
glucose, although half of them possess m-GDH, as judged by acid formation
from aldose sugars (Baumann et aI., 1968). Mutation may occur to produce
gr'owth on glucose (Juni, 1972), as has also been reported for strain LMD
79.41 (van Schie et aI., 1989). All observations suggest that cryptic genes are
302
present (especially for gluconokinase) which become derepressed upon
mutation. The ability to grow on glucose can also be acquired by genetic
transformation (Juni, 1972). In principle, acquisition of m-GDH could
enable growth on those sugars which can be converted to assimilatable
lactones; indeed, all strains which can grow on pentoses contain glucose
dehydrogenase (Baumann et aI., 1968). However, in this connection it
should be mentioned that a lactonase, observed in several Pseudomonas
strains, has so far not been reported for this bacterium, and spontaneous
hydrolysis, especially of an aldonolactone from a pentose, scarcely occurs at
pH 7 (van Schie et aI., 1989). This could imply that, although an assimilation
route for a certain aldonic acid is available, in practice the rapid
acidification during the lag phase may prohibit growth if buffering of the
medium is not adequate.
Natural and man-made mutants occur which lack the ability to

synthesise m-GDH, but show normal growth behaviour. This indicates that
the absence of the protein is not crucial for growth on non-sugar substrates.
The viability of PQQ- mutants supports this view. The apo-enzyme functions
as a scavenger for PQQ because nearly stoichiometric amounts added to the
growth medium are sufficient to achieve the holo-enzyme status. However,
other quinoprotein dehydrogenases perform even better. Thus, by applying
low levels of PQQ, it was found that the in-vivo affinity of quinate
dehydrogenase for PQQ is higher than that of m-GDH (van Kleef et aI.,
1988).
ENERGY CONSERVATION
Since oxidation of glucose to gluconolactone is achieved by a

dehydrogenase, the question can be posed as to whether respiratory activity
via this enzyme leads to any useful energy being conserved for the organism.
In principle, this process could be most easily studied with a strain such as
LMD 79.41 as the enzymes for the following interfering processes are absent:
glycolysis; NAD(P)-dependent glucose dehydrogenation; ketogluconate
formation; gluconolactone or gluconate conversion.
Growth Experiments
To create optimal conditions for ameasurable effect of m-GDH,

growth yield experiments should be performed using a carbon source with a
low energy content (i.e. highly oxidised) so that the carbon source is mainly
used for assimilation and energy provision derives to a large extent from the
m-GDH pathway. These conditions are not easily met in batch growth
experiments as it is generally energy spillage, rather than energy limitation,
that applies. This may be why negative results were reported in the past for
growth yield improvement, although there were some reports with a positive
outcome (Bell and Marus, 1966; Kitagawa et aI., 1986a). Several research
groups have performed growth experiments of this type with continuous
cultures. Studies with strain 69-V on an acetate medium supplemented with
303
glucose, added in a gradient form, gave a high yield of biomass (Muller and
Babel, 1986). The presence of an unknown pathway for glucose or gluconic
acid assimilation in this strain can be excluded since only low amounts of
label were incorporated by the cells in the presence of labelled glucose;
some intracellular labelled gluconate and 6-phosphogluconate was detected
(Kleber et aI., 1984), and no glucose uptake, as previously reported by Cook
and Fewson (1973). Very high growth yields were obtained for strain LMD
79.41 (De Bont et aI., 1984; van Schie et aI., 1987c). These increases were
found at pH 8.3, not at pH 7 or 5. Yield improvement was correlated with
lactone hydrolysis, as was especially clear for the experiments with xylose at
pH 7 where no hydrolysis and no yield increase was observed; however, it is a
puzzling phenomenon that energisation of the transport system via m-GDH
occurred under this condition (van Schie et aI., 1985). Checks with labelled
sugars also revealed that assimilation cannot account for the high yields;
with glucose, incorporation of 2.9% label in biomass and 9.2% in CO 2 ; with
xylose, 0.2% and 2.6%, respectively. Since the increase could not even be
explained from coupling of the dehydrogenase at the level of NADH, it was
assumed that the glucose oxidation step was able to improve the sub-optimal
metabolism of acetate. Unfortunately, the literature reports on the
efficiency of acetate metabolism and ratios are not unanimous. Thus,
acetate conversion has been reported to occur with variable yields (Du Preez
et aI., 1981; Fewson, 1985; Mi.iller and Babel, 1986; Gommers et aI., 1988);
these may be related to strain differences andlor dilution rates, and P 10
ratios of one (Fewson, 1985) and three (Meyer and Jones, 1973) have been
mentioned. Nevertheless, the conclusions with respect to strain LMD 79.41
may be premature because the experiments have subsequently been repeated
with a slightly different experimental arrangement and the very high yield
improvements could not be confirmed. This discrepancy could be related to
the easy flotation of this organism under certain conditions; cells are in the
gas liquid boundary layer and removed with increased efficiency in the
effluent (H. Noorman, personal communication). From recent experiments,
a more realistic increase was observed, as reflected by the number of moles
of ATP generated per mole of 02' which is twice as large on acetate plus
glucose than on acetate (H. Noorman, personal communication).
Summarising, it is clear that the pathway via m-GDH can lead to increases in
yield under appropriate conditions of growth, but that the quantitative
aspects still need attention.
Transport Studies
Transport processes can be driven by glucose oxidation via m-GDH

(van Schie et aI., 1985, 1987b; Kitagawa et aI., 1986b). The proton
translocation achieved can be used to energise transport and ATP synthesis.
Since several dehydrogenases deliver their electrons to the respiratory

chain at the level of ubiquinone Q9' one may wonder whether competition
occurs. However, studies on the availability of the respiratory chain for
different dehydrogenases have not been carried out for this organism. Since
the same rates were observed for the oxidation of sugars with strain 69-V
304
(Asperger et aI., 1981), it was concluded that the dehydrogenase step is not
rate-limiting and the bottleneck should occur somewhere else in the
respiratory chain. However, the conclusion does not apply to strain LMD
79.41, where differences in the affinity constant as well as in the maximal
rate were observed with different sugars (Dokter et aI., 1987). Thus, the
rate-limiting step may vary according to the organism and the growth
condition. Kinetically distinct pathways have been suggested for NADH and
PQQH 2 0xidation (Beardmore-Gray and Anthony, 1986). The dissimilarity
may be related to the location of m-GDH at the periplasmic site and of
NADH dehydrogenase at the cytosolic site of the cytoplasmic membrane.
Induction Experiments
Studies on overflow metabolism (Neijssel et aI., 1989) proceeding via

m-GDH in Klebsiella aerogenes have shown that energy stress leads to
induction of the enzyme. This correlation was found either during growth in
the presence of uncoupler (collapse of the proton motive force) or during
growth limitation with certain nutrients, leading to futile cycling.
Continuous culture experiments with A. calcoaceticus strain LMD 79.41
showed that the enzyme is produced constitutively, independently of the
growth substrate or the presence of an aldose sugar (van Schie et aI., 1988).
Since the highest levels were attained at low growth rates, the energy status
of the cells may also have at least a partial influence on the induction in this
bacterium. The conclusions drawn here may, however, be premature, since
on repeating the experiments with acetate plus xylose (instead of glucose), it
appeared that both constitutive as well as inducible synthesis occurs (H.
Noorman, personal communication).
The synthesis of apo-m-GDH and PQQ can occur independently (van

Kleef and Duine, 1989). This independence is very pronounced in the case
of the PQQ-Iess strain A. lwoffii LMD 73.1, but is also indicated by the
occurrence of non-coordinated synthesis in the case of A. calcoaceticus
LMD 82.3 (van Schie et aI., 1989). A curious phenomenon is the absence of
m-GDH activity in cells grown under oxygen limitation (van Schie et aI.,
1988). This negative result is not due to absence of the cofactor because
PQQ addition had no effect and cells resumed their activity upon a short
oxygen exposure. Since respiration of other substrates was not impaired, the
route for glucose conversion may involve a special factor which has to be
kept in its oxidised form.
CONCLUSIONS AND PROSPECTS
Until recently, s-GDH had been found only in A. calcoaceticus

strain LMD 79.41, but it is now known to exist in other Acinetobacter
strains (K. Matsushita, personal communication). Although the enzyme
oxidises a broad range of aldose sugars in vitro, all the data suggest that it
does not participate in sugar oxidation or in any other cellular process. Thus
one is left with the conclusion that either the enzyme is a relic from the
305
ancient past, for which natural selection had no tools to remove the genetic
information, or that it has a still unknown function. The same difficulty is
met when one tries to explain the widespread occurrence of m-GDH apo-
enzyme in bacteria which do not synthesise free PQQ and which will most
probably never see PQQ in their natural environment (Neijssel, 1987).
Studies on glucose dehydrogenase remain relevant in order to answer these
questions.
s-GDH is an attractive model enzyme to study the mechanism of

quinoprotein dehydrogenases: the spectra of its redox forms are similar to
those of PQQ, suggesting a rather simple binding to the protein chain; the
oxidised form is stable and no activator is required for the substrate
conversion step; the availability of stable apo-enzyme gives possibilities for
studying the binding and catalytic mechanism of PQQ, labelled PQQ, and
P.QQ-analogues. Since the enzyme has a high turnover number and is
specific for PQQ (only slight modifications of the molecule are allowed), a
very sensitive and reliable assay for PQQ has been developed, based on apo-
enzyme which is free of holo-enzyme, prepared by heterologous expression in
E. coli.
m-GDH is responsible for the oxidation of sugars by whole cells. The

hydrophobic enzyme has been only partially characterised. Its protein
structure is quite different from that of s-GDH, but the global information
available suggests mechanistic similarity between the two enzymes.
Although m-GDH probably has a high redox potential, ubiquinone is the
natural electron acceptor. Perhaps this discrepancy will be explained when
more is known about the interplay between the Q-cycle of this organism and
the mechanism of one-electron transfer steps between m-GDH and the
respiratory chain.
Growth experiments and studies of transport and ATP formation

convincingly demonstrate that the glucose oxidation step can provide useful
energy to the organism. For some organisms containing m-GDH, but which
are unable to metabolise gluconic acid, glucose oxidation via m-GDH
provides an auxiliary energy source. Although these organisms form an ideal
object for energetic studies, the results of growth experiments with mixed
substrates (acetate plus glucose) are still confusing. Efforts to improve
experimental arrangements might be rewarding as in principle, related to
their simplicity, opportunities are provided to manipulate carbon flow,
energy dissipation and redox balances. This could also have practical
significance since Gluconobacter oxydans, the industrial organism used
for gluconic acid production, is much more complicated as it has multiple
catabolic routes for glucose degradation.
Finally, what is the significance of m-GDH for A. calcoaceticus in

its natural environment? The affinity constant for the aldose sugars is
rather high: 2 mM for glucose, 4 mM for D-xylose, and much higher for other
sugars (Dokter et aI., 1987). Cellular oxidation of glucose has a pH optimum
of about 6, but as no lactonase is present, the benefit of the oxidation is
306
questionable at this pH. Moreover, the oxidation rates rapidly drop at lower
oxygen tensions. Taking these restrictions together, one may wonder
whether A. calcoaceticus in its natural environment will benefit from this
enzyme for energetic purposes. The objections apply even more strongly in
the case of A. lwoffii, because it not only requires the high concentrations
of sugars, but also PQQ to functionalise its enzyme. Although one could be
optimistic about the chance that these conditions are met, is m-GDH really
designed for energy provision? For instance, the production of
gluconolactone or gluconic acid per se may be beneficial as it creates an
attractive micro-environment for A. calcoaceticus, but not for other
organisms. Support for this idea can be derived from the situation which
exists for those yeasts and fungi that contain glucose oxidase; this enzyme
catalyses the same reaction (although additional HzOzformation occurs, this
becomes rapidly decomposed by the catalase), but without yielding useful
energy. Perhaps the constitutive synthesis of the enzyme serves that
purpose, while the gene(s) to make a complete route for glucose catabolism
(gluconate kinase) remain silent until conditions are favourable for
derepression by mutation. The final conclusion is that, although it is clear
that m-GDH is able to function in an energy conserving process in A.
calcoaceticus, its ecological role and real function are still far from
understood.
REFERENCES
Anthony, c., and Zatman, L.J., 1967, The microbial oxidation of methanol.
The prosthetic group of the alcohol dehydrogenase of
Pseudomonas sp. M27, Biochem. J., 104:960.
Asperger, 0., Borneleit, P., and Kleber, H.-P., 1981, Untersuchungen zum
Elektronentransport in Acinetobacter calcoaceticus,
Ab. Akad. Wissen. DDR., 3:259.
Bak, T.G., 1967, Studies on glucose dehydrogenase from Aspergillus
oryzae. II Purification and physical and chemical properties,
Biochim. Biophys. Acta, 139:277.
Barron, E.S.G., and Friedemann, T.E., 1941, Studies on biological
oxidations. XIV. Oxidations by microorganisms which do not
ferment glucose, J. Bioi. Chem., 137:593.
Baumann, P., Doudoroff, M., and Stanier, R. Y., 1968, A study of the
M oraxell a group. II Oxidative-negative species (genus
Acinetobacter), J. Bacteriol., 95:1520.
Beardmore-Gray, M., and Anthony, c., 1983, The absence of quinoprotein
alcohol dehydrogenase in Acinetobacter calcoaceticus,
Beardmore-Gray, M., and Anthony, c., 1986, The oxidation of glucose by
Acinetobacter calcoaceticus: interaction of the quinoprotein
glucose dehydrogenase with the electron transport chain,
307
Bell, E.J., and Marus, A., 1966, Carbohydrate metabolism of Mima
polymorpha I. Supplemental energy from glucose added to a
growth medium, 1. Bacteriol., 91:2223.
Bouvet, P.J.M., and Bouvet,O.M.M., 1989, Glucose dehydrogenase activity in
Acinetobacter species, Res. Microbiol., 140:531.
nov., and emended escriptions of Acinetobacter calcoaceticus
and Acinetobacter lwoffii, Int. i. Syst. Bacteriol., 36:228.
Cleton-Jansen, A.M., Goosen, N., OdIe, G., and van de Putte, P.,
1988a, Nucleotide sequence of the gene coding for
quinoprotein glucose dehydrogenase from Acinetobacter
calcoaceticus, Nucl. Acid Res., 16:6228.
Cleton-Jansen, A.M., Goosen, N., Wenzel, T.J., and van de Putte, P., 1988b,
Cloning of the gene encoding quinoprotein glucose dehydrogenase
from Acinetobacter calcoaceticus: evidence for the presence
of a second enzyme, i. Bacteriol., 170:2121.
Cleton-Jansen, A.M., Goosen, N., Vink, K., and van de Putte, P., 1989,
Cloning, characterization and DNA sequencing of the gene
encoding the M,50,000 quinoprotein glucose dehydrogenase from
Acinetobacter calcoaceticus, Mol. Gen. Genet., 217:430.
Conlin, M., Forrest, H.S., and Bruice, T.C., 1985, Replacement of
methoxatin by 4,7-phenanthroline-5,6-dione and the inability of
other phenanthroline quinones, as well as 7,9-
didecarboxymethoxatin, to serve as cofactors for the methoxatin-
reqUIflng glucose dehydrogenase of Acinetobacter
calcoaceticus, Biochem. Biophys. Res. Comm., 131:564.
Cook, A.M., and Fewson, C.A., 1973, Role of carbohydrates in the
metabolism of Acinetobacter calcoaceticus, Biochim.
D'Costa, E.J., Higgins, I.J., and Turner, A.P.F., 1986, Quinoprotein glucose
dehydrogenase and its application in an amperometric glucose
sensor, Biosensors 2:71.
De Bont, J.A.M., Dokter, P., van Schie, B.J., van Dijken, J.P., Frank, J.,
Duine, J.A., and Kuenen, J.G., 1984, Role of quinoprotein glucose
dehydrogenase in gluconic acid production by Acinetobacter
calcoaceticus, Ant. v. Leeuw., 50:76.
Dijkshoorn, L., Tjernberg, I., Pot, B., Ursing, M.J.F., and Kersters, K., 1990,
Numerical analysis of cell envelope profiles of Acinetobacter
strains classified by DNA-DNA hybridization, Syst. Appl.
Microbiol., in press.
Dokter, P., van Kleef, M.A.G., Frank, J., and Duine, J.A., 1985, Production
of quinoprotein D-glucose dehydrogenase in the culture medium
of Acinetobacter calcoaceticus, Enz. Microb. Tech., 7:613.
Dokter, P., Frank, J., and Duine, J.A., 1986, Purification and
characterization of quinoprotein glucose dehydrogenase from
Acinetobacter calcoaceticus LMD 79.41, Biochem. i.,
239:163.
308
Dokter, P., Pronk, 1.T., van Schie, B.l., van Dijken, 1.P., and
Duine, 1.A., 1987, The in vivo and in vitro substrate
specificity of Acinetobacter calcoaceticus LMD 79.41, F EM S
Dokter, P., van Wielink, 1.E., van Kleef, M.A.G., and Duine, 1.A., 1988,
Cytochrome-562 from Acinetobacter cal co acetic us LMD
79.41, Biochem. J., 254:13l.
Dokter, P., van Wielink, 1.E., Geerlof, A., Oltman, F., Stouthamer, A.H., and
Duine,l.A., 1990, Purification and partial characterization of the
membrane-bound haem-containing proteins from Acinetobacter
cal coaceticus LMD.41, J. Gen. M icrob iol., in press.
Du Preez, J.e., Toerien, D.F., and Lategan, P.M., 1981, Growth parameters
of Acinetobacter calcoaceticus on acetate and ethanol,
Appl. Microbiol. Biotech., 13:45.
Duine, J.A., and Frank, J., 1981, Quinoprotein alcohol
dehydrogenase from a non-methylotroph, Acinetobacter
calcoaceticus, J. Gen. Microbiol., 122:20l.
Duine, 1.A., Frank, J., and van Zeeland, 1.K., 1979, Glucose
dehydrogenase from Acinetobacter calcoaceticus: a
quinoprotein, FEBS Lett., 108:443.
Duine, 1.A., Frank, 1., and Verwiel, P.E.J., 1980, Structure and
aCtIvity of the prosthetic group of methanol
dehydrogenase, Eur. J. Biochem., 108:187.
Duine, 1.A., Frank, 1., and van der Meer, R., 1982, Different forms
of quinoprotein aldose-(glucose-) dehydrogenase in
Acinetobacter calcoaceticus, Arch. Microbiol., 131:27.
Duine, J.A., Frank, J., and Jongejan, J.A., 1983, Detection and
determination of pyrroloquinoline quinone, the coenzyme of
quinoproteins, Anal. Biochem., 133:239.
Acinetobacter calcoaceticus, J. Gen. Microbio!., 131:865.
Frank, J., van Krimpen, S.H., Verwiel, P.E.J., Jongejan, J.A., Mulder, A.e.,
and Duine, J.A., 1989, On the mechanism of inhibition of methanol
dehydrogenase by cyclopropane-derived inhibitors,
Eur. J. Biochem., 184:187.
Friedrich, T., Strohdeicher, M., Hofhaus, G., Preis, D., Sahm, H., and Weiss,
H., 1990, The same domain motif for ubiquinone reduction in
mitochondrial or chloroplast NADH dehydrogenase and bacterial
glucose dehydrogenase, FEBS Lett., 265:37.
Geerlof, A., Dokter, P., van Wielink, J.E., and Duine,J.A., 1989, Haem-
containing protein complexes of Acinetobacter calcoaceticus,
in "PQQ and Quinoproteins," p.106, J.A. Jongejan, and J.A. Duine,
eds., Kluwer Academic Publishers, Dordrecht.
Geiger, 0., and Gorisch, H., 1986, Crystalline quinoprotein glucose
dehydrogenase from Acinetobacter calcoaceticus,
Geiger, 0., and Gbrisch, H., 1989, Reversible thermal inactivation of the
quinoprotein glucose dehydrogenase from Acinetobacter
calcoaceticus, Biochem. J., 261:415.
309
Gommers, P.J.F., van Schie, B.J., van Dijken, J.P., and Kuenen, J.G., 1988,
Biochemical limits to microbial growth yields: an analysis of mixed
substrate utilization, Biotech. Bioeng., 32:86.
Goosen, N., Horsman, H.P.A., Huinen, R.G.M., de Groot, A., and van de
Putte, P., 1989, Genes involved in the biosynthesis of PQQ from
Acinetobacter calcoaceticus, in "PQQ and Quinoproteins,"
p.169, J.A.Jongejan and J.A.Duine, eds., Kluwer Academic
Publishers, Dordrecht.
Hauge, J.G., 1960a, Purification and properties of glucose dehydrogenase
and cytochrome b from Bacterium anitratum, Biochim.
Hauge, J.G., 1960b, Kinetics and specificity of glucose dehydrogenase from
Bacterium anitratum, Biochim. Biophys. Acta, 45:263.
Hauge, J.G., 1961, Glucose dehydrogenase in bacteria: a comparative study,
1. Bacteriol., 82:609.
Hauge, J.G., 1964, Glucose dehydrogenase of Bacterium anitratum: an
enzyme with a novel prosthetic group, 1. Biol. Chem., 239:3630.
Hauge, J.G., 1966a, Glucose dehydrogenase-particulate. I Pseudomonas
species and Bacterium anitratum, M eth. Enzymol., 9:92.
Hauge, J.G., 1966b, Glucose dehydrogenase soluble. II Bacterium
anitratum, Meth. Enzymol., 9:107.
Hauge, J .G., and Hallberg, P.A., 1964, Solubilization and properties of the
structurally-bound glucose dehydrogenase of Bacterium
anitratum, Biochim. Biophys. Acta, 81:251.
Hauge, J.G., and Murer, E.H., 1964, Studies on the nature of the prosthetic
group of glucose dehydrogenase of Bacterium anitratum,
Hauge, J.G., Foulds, G., and Bentley, R., 1968, Kinetic isotope effects in
enzymic oxidations of D-glucose-l-H and D-glucose-1- 2 H,
Hayaishi, 0., 1966, Lactose dehydrogenase, M eth. Enzymol., 9:73.
Hommel, R., G6tzrath, M., and Kleber, H-P., 1989, Enzyme production by
growing cells of Acinetobacter in presence of sophoroselipid and
Triton X-lOO, Acta Biotech., 5:461.
Imanaga, Y., 1989, Investigations on the active site of glucose dehydrogenase
from Pseudomonas fluorescens, in "PQQ and Quinoproteins,"
p.87, J.A. Jongejan and J.A. Duine, eds., Kluwer Academic
Publishers, Dordrecht.
Imanaga, Y., Hirano-Sawatake, Y., Arita-Hashimoto, Y., Iton-Shibouta, Y.,
and Katon-Semba, R., 1979, On the cofactor of glucose
dehydrogenase of Pseudomonas fluorescens,
Proc. J ap. Acad., 55B:264.
Microbiol., 32:349.
Kilty, e.G., Maruyama, K., and Forrest, H.S., 1982, Reconstitution
of glucose dehydrogenase using synthetic methoxatin,
Arch. Biochem. Biophys., 218:623.
310
Kitagawa, K., Tateishi, A., Nakano, F., Matsumoto, T., Morohoshi,
T., Tanino, T., and Usui, T., 1986a, Generation of energy
coupled with membrane-bound glucose dehydrogenase in
Acinetobacter caicoaceticus, Agric. Bioi. Chem., 50:1453.
Kitagawa, K., Tateishi, A., Nakano, F., Matsumoto, T., Morohoshi,
T., Tanino, T., and Usui, T., 1986b, Sources of energy and
energy coupling reactions of the active transport systems in
Acinetobacter caicoaceticus, Agric. Bioi. Chem., 50:2939.
Kleber, H.P., Haferburg, D., Asperger, 0., Schmidt, M., and Aurich, H.,
1984, Aufnahme und Oxidation von monosacchariden bei
Acinetobacter calcoaceticus, ZeUs. Allg. Mikrobiol.,
24:691.
Lockwood, L.B., Tabenkin, B., and Ward, G.E., 1941, The production of
gluconic acid and 2-ketogluconic acid from glucose by species of
Pseudomonas and Phytomonas, J. Bacteriol., 42:51.
Matsushita, K., Shinagawa, E., Inone, T., Adachi, 0., and Ameyama, M.,
1986, Immunological evidence for two types of POO-dependent D-
glucose dehydrogenases in bacterial membranes and the location
of the enzyme in Escherichia coli, FEMS Microbiol. Lett.,
37:141.
Matsushita, K., Shinagawa, E., Adachi, 0., and Ameyama, M., 1988,
Quinoprotein D-glucose dehydrogenase in Acinetobacter
calcoaceticus LMD 79.41: the membrane-bound enzyme is
distinct from the soluble enzyme, FEMS MicrobioI.Lett., 55:53.
Matsushita, K., Shinagawa, E., Adachi, 0., and Ameyama, M., 1989,
Ouinoprotein D-glucose dehydrogenase of the Acinetobacter
calcoaceticus respiratory chain: membrane-bound and soluble
forms are different molecular species, Biochemistry, 28:6276.
Meyer, D.J., and Jones, C.W., 1973, Oxidative phosphorylation in bacteria
which contain different cytochrome oxidases, Eur. f. Biochem.,
36: 144.
Muller, R.H., and Babel, W., 1986, Glucose as an energy donor in acetate
growing Acinetobacter calcoaceticus, Arch. Microbiol.,
144:62.
Neijssel, O.M., 1987, POO-linked enzymes in enteric bacteria,
Microbiol. Sci., 4:87.
Neijssel, O.M., Hommes, R.W.J., Postma, P.W., and Tempest, D.W., 1989, in
"PQQ and Ouinoproteins," p.57, J.A. Jongejan, and J.A. Duine,
eds., Kluwer Academic Publishers, Dordrecht.
Niederpruem, D.J., and Doudoroff, M., 1965, Cofactor-dependent aldose
dehydrogenase of Rhodopseudomonas spheroides,
f. Bacteriol., 89:697.
Salisbury, S.A., Forrest, H.S., Cruse, W.B.T., and Kennard, 0., 1979, A novel
coenzyme from bacterial primary alcohol dehydrogenase, Nature,
280:843.
Shinagawa, E., Matsushita, K., Nonobe, M., Adachi, 0., Ameyama, M.,
Ohshiro, Y., Hoh, S., and Kitamura, Y., 1986, The 9-carboxylic acid
group of pyrroloquinoline quinone, a novel prosthetic group, is
essential in the formation of holo-enzyme of D-glucose
dehydrogenase, Biochem. Biophys. Res. Comm., 139:1279.
311
Strohdeicher, M., Bringer-Meyer, S., Neusz. B., van der Meer, R., Duine,
J .A., and Sahm, H., 1989, Glucose dehydrogenase from
Zymomonas mobilis: evidence for a quinoprotein, in "POO and
Ouinoproteins," p.103, J.A. Jongejan and J.A. Duine, eds., Kluwer
Academic Publishers, Dordrecht.
van Kleef, M.A.G., and Duine, J.A., 1988, Bacterial NAD(P)-
independent quinate dehydrogenase is a quinoprotein,
van Kleef, M.A.G., and Duine, J.A., 1989, Factors relevant in bacterial
pyrroloquinoline quinone production, Appl. Environ.
van Kleef, M.A.G., Dokter, P., Mulder, A.C., and Duine, J.A., 1987,
Detection of the cofactor pyrroloquinoline quinone,
Anal. Biochem., 162:143.
van der Meer, R.A., Groen, B.W., van Kleef, M.A.G., Frank, J., Jongejan,
J.A., and Duine, J.A., 1990, Isolation, preparation and assay of
pyrroloquinoline quinone (POO)' M eth. Enzymol. in press.
van Schie, B.J., van Dijken, J.P., and Kuenen, J.G., 1984, Non-
coordinated synthesis of glucose dehydrogenase and its prosthetic
group POO in Acinetobacter and Pseudomonas species, FEM S
van Schie, B.J., Hellingwerf, K.J., van Dijken, J.P., Elferink, M.G.L., van
Dijl, J.M., Kuenen, J .G., and Konings, W.N., 1985, Energy
transduction by electron transfer via a pyrroloquinoline quinone
Pseudomonas aeruginosa, and Acinetobacter calcoaceticus
(var. lwoffi), .I. Bacteriol., 163:493.
van Schie, B.J., de Mooy, O.H., Linton, J.D., van Dijken, J.P., and Kuenen,
J.G., 1987a, POO-dependent production of gluconic acid by
Acinetobacter, Agrobacterium and Rhizobium species,
.T. Gen. Microbiol., 133:867.
van Schie, B.J., Pronk, J.T., Hellingwerf, K.J., van Dijken, J.P., and Kuenen,
J.G., 1987b, Glucose-dehydrogenase-mediated solute transport
and ATP synthesis in Acinetobacter calcoaceticus,
.T. Gen. Microbiol., 133:3427.
van Schie, B.J., Rouwenhorst, R.J., de Bont, J.A.M., van Dijken, J.P., and
Kuenen, J.G., 1987c, An in vivo analysis of the energetics of
aldose oxidation by Acinetobacter calcoaceticus,
van Schie, B.J., van Dijken, J.P., and Kuenen, J.G., 1988, Effects of growth
rate and oxygen tension on glucose dehydrogenase activity in
Acinetobacter calcoaceticus, Ant. v. Leeuw., 55:53.
van Schie, B.J., Rouwenhorst, R.J., van Dijken, J.P., and Kuenen, J.G., 1989,
Selection of glucose assimilating variants of Acinetobacter
calcoaceticus LMD 79.41 in chemostat culture, Ant. v. Leeuw.,
55:39.
Vicente, M., and Canovas, J.L., 1973, Glucolysis in Pseudomonas putida:
physiological role of alternative routes from the analysis of
defective mutants,.T. Bacterio!., 116:908.
312
ACINETOBACTER - CITRIC ACID CYCLIST WITH
A DIFFERENCE
P.D.l. Weitzman
Faculty of Life Sciences

Cardiff Institute of Higher Education
Cardiff CFS 2YB
UK
INTRODUCTION
My own introduction to the genus Acinetobacter was the result of a

purely chance encounter - indeed an experimental accident - for which I have
reason to be most grateful. I had decided to examine enzyme regulatory
mechanisms in the Krebs' citric acid cycle of bacteria. The ubiquitous nature
of this pathway, its dual metabolic function (catabolic and anabolic) in the
generation of both energy and biosynthetic intermediates, and its cyclic
character combine to present an interesting challenge to the investigator of
metabolic regulation. The present article summarises work carried out in my
laboratory over a number of years. Fig. 1 shows the reaction scheme of the
citric acid cycle.
CITRATE SYNTHASE
I started investigations on citrate synthase - the enzyme that condenses

acetyl-CoA and oxaloacetate to produce citrate. This enzyme may be
considered the "first" or "initiating" enzyme of the citric acid cycle since it
catalyses the only step in which carbon atoms enter the cycle by combining
with a cycle intermediate. In extracts of Escherichia coli I found that the
enzyme was strongly inhibited specifically by NADH (Weitzman, 1966). As
NADH may be thought of as the product of the cycle, this represented a case
of product feedback inhibition in a cyclic metabolic pathway and clearly
merited further investigation. The fact that NADH was without effect on
citrate synthases from eukaryotic organisms suggested the possibility that the
effect observed with the E. coli enzyme might be characteristic of other
bacteria, perhaps even a general property of prokaryotes.
313
acelyl-CoA
~
m. ,?K"", d""\
r isocitrate
rum\te )
SUCci~. 71utarate
succlnyl-CoA
Pi g. 1, The citric acid cycl e
To examine the E. coli citrate synthase in more detail, I embarked on

its large-scale purification and it was here that a lack of skill resulted in a
fortunate accident. It was suggested that if I used a heavy inoculum of
E. coli, sterilisation of the bottles containing 15 I of growth medium could
probably be dispensed with, particularly as these were made up with freshly-
distilled water. I innocently followed this advice, harvested the cells
produced and devised an effective procedure to purify the citrate synthase.
The pure enzyme retained its sensitivity to inhibition by NADH, though the
kinetic characteristics of this inhibition differed from those observed with
crude extracts of E. coli. Whereas the latter had shown a hyperbolic
dependence of inhibition on NADH concentration, the pure enzyme
displayed a sigmoidal dependence. Possible explanations in terms of
restricted conformational flexibility in the presence of (association with?)
other proteins in the crude extract offered themselves as speculations.
However, another marked difference between the behaviour of the crude and
pure enzymes was observed. With the pure enzyme, the NADH inhibition
could be relieved by AMP (the "reactivation" again displaying a sigmoidal
dependence on AMP concentration), whereas no such effect could be seen
with the crude enzyme.
The reader will already have jumped to the correct conclusion, though
for us, blinkered by our expectations, it took some time to realise and
convince ourselves that the large-scale growth of E. coli had, in reality,
resulted in the production of another bacterium which had dramatically
outgrown the E. coli inoculum. One organism dominated the large-scale
cultures and this was shown to be Acinetobacter calcoaceticus. Crude
extracts of this bacterium showed the same citrate synthase properties as did
the purified enzyme. It was the difference between these properties and
those of E. coli citrate synthase that suggested it would be worthwhile to
examine citrate synthases from a wide range of bacterial species to ascertain
both the generality of NADH inhibition and any other regulatory diversity.
314
Our survey (Weitzman and Jones, 1968) revealed a striking pattern of
enzyme behaviour with two clear-cut divisions. First, citrate synthases from
Gram-negative bacteria are all inhibited by NADH, whereas those from
Gram-positive bacteria are unaffected. Second, within the Gram-negative
group, it is only in the case of the citrate synthases from strictly aerobic
organisms that AMP reactivation of NADH inhibition occurs; in the
facultative anaerobes (such as E. coli) AMP exerts no such effect. The
subsequent examination of yet more bacterial citrate synthases, both by
ourselves and by other investigators, has essentially confirmed these
regulatory patterns.
The striking difference between citrate synthases from Gram-negative

and Gram-positive bacteria with respect to NADH inhibition is paralleled by
the difference between the molecular complexity and sizes of the two classes
of enzyme (Weitzman and Dunmore, 1969; Weitzman, 1981). Gram-negative
bacterial citrate synthase appears to be a hexameric molecule of identical or
similar subunits (M,250,000-300,000), whereas the Gram-positive bacterial
enzyme is a dimer (M,approx. 95,000).
Why the phenomenon of NADH inhibition should occur only in the

Gram-negative organisms remains a mystery. On the other hand, the
incidence of AMP reactivation exclusively in strictly aerobic organisms may
be rationalised in terms of the absolute dependence of such organisms on the
citric acid cycle (and hence on citrate synthase) for energy production. It
thus makes some sense that AMP, a signal of depleted metabolic energy,
should serve as a positive effector (stimulator) of a key enzyme in the cycle,
thereby increasing the activity of the cycle and overcoming the low-energy
condition. In contrast, organisms such as E. coli are metabolically capable
of producing energy without recourse to the citric acid cycle, i.e. by
fermentation, and would thus not require AMP stimulation of citrate
synthase.
Although the properties of the Acinetobacter citrate synthase are

not so different from those of the enzyme produced by other Gram-negative
aerobes, e.g. the pseudomonads, it was the difference between the properties
of the enzyme from Acinetobacter and E. coli which initially prompted the
whole survey of citrate synthases. Acinetobacter may therefore claim to
have played a crucial part in the discovery of the remarkable patterns of
regulatory diversity displayed by this enzyme. We shall later see again that
the distinctive properties of Acinetobacter enzymes may prompt broader
investigations and lead to other interesting and novel findings.
The Acinetobacter citrate synthase possesses several features that

facilitate its further study: (i) it may be purified to homogeneity by
conventional procedures and is a particularly stable enzyme; (ii) it may be
stored cold for several months with little loss of activity or regulatory
sensitivity; (iii) unlike theE. coli enzyme, the citrate synthase produced by
Acinetobacter does not appear to have sensitive thiol groups whose
chemical modification or oxidation would lead to inactivation or
315
desensitisation and, moreover, its hexameric molecule shows no ready
tendency to disaggregate into smaller units. These features enabled us to
examine the enzyme by electron microscopy and to look for gross
conformational changes accompanying interactions with the regulatory
effectors (Rowe and Weitzman, 1969). Measurement of individual enzyme
molecules in the electron micrographs clearly demonstrated that the enzyme
molecule undergoes an asymmetrical "swelling" on interaction with NADH
(i.e. on transition to the inactive state), and that this is completely reversed
in the additional presence of AMP (return to the active state). This
"visualisation" of an allosteric transition was a gratifying confirmation of
other, indirect, probes of conformational alterations.
We have also examined the effects of "cross-linking" the

Acinetobacter citrate synthase molecule with bifunctional reagents,
thereby "freezing" the conformation and restricting its alteration (Mitchell
and Weitzman, 1983; Lloyd and Weitzman, 1987). Interestingly, the cross-
linked enzyme continues to exhibit sensitivity to both NADH and AMP, but
the cooperative interactions between effector sites, which give rise to the
sigmoidal responses to both NADH and AMP, are abolished. Cleavage of
the cross-links restores these sigmoidal responses, presumably by releasing
the conformational constraints. Such cross-linking experiments are a useful
aid to further exploration of the molecular mechanisms of the regulatory
inhibition and reactivation. The finer details of the molecular structure of
Acinetobacter citrate synthase have been successfully investigated by
Duckworth and co-workers (e.g. Morse and Duckworth, 1980; Donald and
Duckworth, 1987) and should lead, in the future, to a fuller molecular
understanding of the catalytic and regulatory functions of this enzyme.
ISOCITRATE DEHYDROGENASE
Isocitrate dehydrogenase catalyses the first oxidative decarboxylation

reaction of the citric acid cycle, converting isocitrate into oxoglutarate.
Although some bacteria contain NAD-Iinked isocitrate dehydrogenase, the
majority appear to contain only the NADP-linked enzyme.
In the course of examining Acinetobacter isocitrate dehydrogenase

we observed that, with extracts of acetate-grown cells, the enzyme activity
appeared to increase during the course of the assay; in contrast, extracts of
cells grown in nutrient broth gave normal linear rates in the assay. It was
found that the progressive activation of isocitrate dehydrogenase resulted
from the accumulation in the assay reaction mixture of glyoxylate, produced
by the action of isocitrate lyase on the isocitrate substrate. The absence of
isocitrate lyase in cells grown in nutrient broth accounted for the absence of
the "activation" effect in extracts of such cells. In the course of further
investigation of this glyoxylate effect it was discovered that Acinetobacter
produces two distinct isoenzymes of NADP-linked isocitrate dehydrogenase -
IDH-I, a "small" isoenzyme (M, approx. 100,000) and IDH-II, a "large"
isoenzyme (M,approx. 300,000) (Self and Weitzman, 1970, 1972). It is the
"large" isoenzyme which shows several unusual and interesting properties.
316
The activity of IDH-ll is markedly stimulated by AMP and, to a lesser
extent, ADP (Parker and Weitzman, 1970). As in the case of citrate synthase,
stimulation of IDH-II by AMP may be rationalised in terms of "energy
control" of the citric acid cycle. The low-energy signal AMP (or ADP)
stimulates metabolite flux through the cycle, thereby overcoming the low-
energy condition and constituting a homeostatic mechanism. Confirmatory
studies on the nucleotide control of IDH-II have been reported by Kleber
and Aurich (1976).
Some activation of A c in e t 0 b a c te r isocitra te dehydrogenase by

glyoxylate was referred to above. Separation of the two isoenzymes revealed
that it is only IDH-IJ that is stimulated by glyoxylate (Self et aI., 1973). A
similar stimulatory effect is also exerted by pyruvate, which bears a strong
structural similarity to glyoxylate. It is tempting to suggest that the
activation of IDH-U by glyoxylate constitutes a mechanism for controlling
metabolic flux at the isocitrate branchpoint. Isocitrate may be metabolised
either via isocitrate dehydrogenase and the citric acid cycle, or via isocitrate
lyase and the glyoxylate cycle. The latter route results in the formation of
glyoxylate which, by stimulating IDH-II, promotes metabolism via the
dehydrogenase route and hence exerts a balance between the two options. It
is also significant that we found the adaptation of Acinetobacter to growth
on acetate to be accompanied by an increase in the level of the IDH-II
isoenzyme (Reeves et aI., 1983, 1986).
It is also noteworthy that the two types of activator - AMP and

glyoxylate - exert their actions in an apparently independent manner; their
joint presence results in a summation of their individual activating effects
(Self et al., 1973).
As if these activations do not reflect sufficient functional complexity

of IDH-II, this enzyme also displays the phenomenon of "hysteresis" (O'Neil
and Weitzman, 1988). Partial purification of IDH-II in low-salt buffer by
dye-ligand chromatography results in a preparation which takes around 20
min to display a constant steady rate in the normal spectrophotometric
enzyme assay. Such a hysteretic effect suggests the slow transition of the
enzyme from an inactive to an active form. Both AMP and glyoxylate abolish
this hysteresis. It remains to be determined whether the hysteresis plays any
part in the normal physiological functioning of Acinetobacter isocitrate
dehydrogenase, but it certainly adds yet another facet to the regulatory
complexity observed in vitro. It should also be emphasised that the unusual
properties of isocitrate dehydrogenase described here confer a
distinctiveness on Ac i n e to b act e r and are not general features of the
bacterial enzyme.
OXOGLUTARATE DEHYDROGENASE
Following isocitrate dehydrogenase, the next step in the citric acid

cycle is the oxidative decarboxylation of oxoglutarate to succinyl-CoA. This
step is catalysed by the multi-enzyme complex oxoglutarate dehydrogenase.
317
Once again, the Acinetobacter enzyme displays distinctive features
(Weitzman, 1972; Parker and Weitzman, 1973). The Acinetobacter
oxoglutarate dehydrogenase is inhibited by NADH, and this inhibition is
relieved by AMP (or ADP). Moreover, AMP and ADP stimulate enzyme
activity even in the absence of NADH. In contrast, ATP produces no
stimulation but, instead, a small degree of inhibition. The AMP and ADP
stimulatory effects are associated with a marked reduction in the value of K
for oxoglutarate. Again, this suggests the operation of a physiological
regulatory mechanism whereby the low-energy signals AMP or ADP
stimulate metabolite flux round the cycle.
PYRUVATE DEHYDROGENASE
Observation of the unusual regulatory properties of the oxoglutarate

dehydrogenase from Acinetobacter prompted an examination of the
analogous multi-enzyme pyruvate dehydrogenase complex, which catalyses
the conversion of pyruvate to acetyl-CoA. Broadly similar observations were
made (Jaskowska-Hodges, 1975). Acinetobacter pyruvate dehydrogenase
exhibits sensitivity to inhibition by NADH and to activation by AMP.
Although this enzyme is not, strictly speaking, a component of the citric acid
cycle, it functions to produce acetyl-CoA, one of the substrates for citrate
synthase. Stimulation by AMP may therefore be considered as enhancing
flux through the cycle.
SUCCINATE THIOKINASE
Succinate thiokinase catalyses the breakdown of succinyl-CoA to

succinate and coenzyme A, with the concomitant phosphorylation of a
nucleoside diphosphate to the corresponding triphosphate. The nucleoside
diphosphate may be either ADP or GOP. The generally held view (e.g. in
textbooks of biochemistry) is that succinate thiokinase of animal origin
functions with GDP, whereas bacterial and plant succinate thiokinases
utilise ADP.
We found a surprisingly high Krn value for AOP (about 1 mM) with
Acinetobacter succinate thiokinase compared with the value of the E. coli
enzyme (about 0.01 mM). This prompted us to test the Acinetobacter
enzyme with GDP. The Krn value for this nucleotide was around 0.02 mM, but
the same V rna< was displayed with GDP as with ADP. The apparent
"preference" of Acinetobacter succinate thiokinase for GDP over ADP
prompted a survey of a range of bacteria for the nucleotide utilisation of
their succinate thiokinase (Weitzman and Jaskowska-Hodges, 1982). The
bacteria could be classified into four groups. One group contained
Acinetobacter and a few other species, all of which had a succinate
thiokinase with a very high Krn value for ADP and a low Km for GDP. The
enzyme of a second group had a low Km value for both ADP and GDP, while
that of a third group showed a low Km value for ADP and a higher Km value for
318
GDP. The fourth group of bacteria had a succinate thiokinase which
appeared to function only with ADP. It is noteworthy that the grouping
together of bacteria on the basis of this nucleotide utilisation of their
succinate thiokinases shows some taxonomic rationale.
These findings stem from the initial unexpected observations on

Acinetobacter succinate thiokinase, in much the same way as the extensive
studies on the diversity of citrate synthases have their origin in the initial
observations of the distinctive properties of the Acinetobacter enzyme. In
the case of succinate thiokinase it is interesting to note that the initial
prompt by Acinetobacter has led to new findings about this enzyme in the
animal kingdom. The diversity of bacterial succinate thiokinases tempted
us to question the accepted view of the nucleotide utilisation of animal
succinate thiokinase. Although, as mentioned above, succinate thiokinase
from animal sources is usually claimed to be specific for GDP, we found a
range of animal tissues to show succinate thiokinase activity with both GDP
and ADP (Weitzman et aI., 1986). Interestingly, the ratio of activity with
GDP to that with ADP varies across tissues and this unexpected observation
was resolved by the demonstration of the presence of two distinct succinate
thiokinases in these tissues - one enzyme specific for GDP and the other for
ADP. Further studies have shed light on the distinct physiological roles of
the two succinate thiokinases and have suggested that it is the ADP-linked
enzyme which functions in the energy-generating citric acid cycle, whereas
the GDP-linked enzyme is associated with ketone body utilisation and haem
biosynthesis (Jenkins and Weitzman, 1986, 1988; Jenkins et aI., 1988).
Fig. 2. Biosynthetic branch-points in the citric acid cycle. Heavy

arrows indicate biosynthetic reactions
319
MUL TlPOINT CONTROL
The regulatory properties of the citric acid cycle enzymes of

Acinetobacter, described above, constitute an unusual system of AMP
control of the cycle in this organism, which I have termed "multipoint
control" (Weitzman, 1981). Reference to the scheme of the citric acid cycle
shown in Fig. 2 emphasises the multi-branchpoint nature of the cycle.
Pyruvate (C 3), acetyl-CoA (C 2 ) oxaloacetate (C 4), isocitrate (C 6 ) and
oxoglutarate (C) are branchpoint metabolites which may be metabolised
either via the cycle or into other (biosynthetic) pathways. As a prime
function of the cycle is in the provision of energy (via the formation of
NADH and its oxidation coupled to phosphorylation and ATP production), it
is logical that metabolites signalling a low-energy condition, e.g. AMP,
should act to channel flux round the cycle rather than allow diversion into
competing pathways. Acinetobacter displays a coordinated enzymic
sensitivity to AMP which may permit multipoint control of its citric acid
cycle. By acting as a positive effector of pyruvate dehydrogenase (C3~,)'
citrate synthase (C 2 + C4~6)' isocitrate dehydrogenase (C6~) and
oxoglutarate dehydrogenase (C S--;;.C 4 ), AMP can enhance metabolite flux
through the cycle and thereby exert a sensitive control over its energy-
generating role.
ENVOI
I hope that the brief description given here of the novel features of
some citric acid cycle enzymes from Acinetobacter has justified the choice
of title for this contribution. The study of the Acinetobacter enzymes has
revealed some very unusual and interesting properties, served as a stimulus
to the examination of other organisms, and led to the discovery of extensive
enzyme diversity.
At the outset, I confessed to the accidental nature of my introduction

to Acinetobacter. In conclusion, I must acknowledge that this accidental
encounter has proved a most productive and valuable association.
REFERENCES
Donald. L.J., and Duckworth, H.W., 1987, Expression and base sequence of
the citrate synthase gene of Acinetobacter anitratum,
Biochem. Cell. Bio!., 65:930.
J askowska-Hodges, H., 1975, The control of citric acid cycle enzymes in
Acinetobacter, PhD thesis, Univ. Leicester.
Jenkins, T.M., and Weitzman, P.D.l., 1986, Distinct physiological roles of
animal succinate thiokinases: association of guanine nucleotide-
linked succinate thiokinase with ketone body utilization, F EBS
Lett., 205:215.
320
Jenkins, T.M., and Weitzman, P.D.J., 1988, Physiological roles of animal
succinate thiokinases: specific association of the guanine
nucleotide-linked enzyme with haem biosynthesis, F EBS Lett.,
230:6.
Jenkins, T.M., Eisenthal, R., and Weitzman, P.D.J., 1988, Two distinct
succinate thiokinases in both bloodstream and procyclic forms of
Trypanosoma brucei, Biochem. Biophys. Res. Comm.,
151:257.
Kleber, H.-P., and Aurich, H., 1976, Control of NADP-specific isocitrate
dehydrogenase from Acinetobacter by nucleotides, FEBS Lett.,
61:282.
Lloyd, A.l., and Weitzman, P.D.J., 1987, Modification of the
regulatory properties of Acinetobacter citrate synthase by cross-
linking, Biochem. Soc. Trans., 15:840.
Mitchell, e.G., and Weitzman, P.D.J., 1983, Reversible effects of
cross-linking on the regulatory cooperativity of Acinetobacter
citrate synthase, FEBS Lett., 151:260.
Morse, D., and Duckworth, H. W., 1980, A comparison of the citrate
synthases of Escherichia coli and Acinetobacter anitratum,
Can. J. Biochem., 58:696.
O'Neil. S., and Weitzman, P.D.J., 1988, Acinetobacter catcoaceticus
isocitrate dehydrogenase - fractionation, AMP stimulation and
hysteretic behaviour, Biochem. Soc. Trans., 16: 870.
Parker, M.G., and Weitzman, P.D.l., 1970, Regulation of NADP-linked
isocitrate dehydrogenase activity in Acinetobacter, FEBS Lett.,
7:324.
Parker, M.G., and Weitzman, P.D.l., 1973, The purification and regulatory
properties of a-oxoglutarate dehydrogenase from Acinetobacter
twoffi, Biochem. J., 135:215.
Reeves, H.C., O'Neil, S., and Weitzman, P.D.l., 1983, Modulation of
isocitrate dehydrogenase activity in Acinetobacter
calcoaceticus by acetate, FEBS Lett, 163:265.
Reeves, H.C., O'Neil, S., and Weitzman, P.D.l., 1986, Changes in NADP-
isocitrate dehydrogenase isoenzyme levels in Acinetobacter
calcoaceticus in response to acetate, FEMS Micobiol. Lett.,
35:209.
Rowe, A.J., and Weitzman, P.D.J., 1969, Allosteric changes in citrate
synthase observed by electron microscopy, J. Mol. BioI., 43:345.
Self, C.H., and Weitzman, P.D.l., 1970, Separation of isoenzymes by zonal
centrifugation, Nature, 225:644.
Self, C.H., and Weitzman, P.D.J., 1972, The isocitrate dehydrogenases of
Acinetobacter twofIi: separation and properties of two
nicotinamide-adenine dinucleotide phosphate-linked isoenzymes,
Biochem. J., 130:211.
Self, e.H., Parker, M.G., and Weitzman, P.D.l., 1973, The isocitrate
dehydrogenases of Acinetobacter Iwoffi: studies on the
regulation of a nicotinamide-adenine dinucleotide phosphate-
linked isoenzyme, Biochem. J., 132:215.
321
Weitzman, P.D.l., 1966, Regulation of citrate synthase actIvity In
Escherichia coli, Biochim. Biophys. Acta, 128:213.
Weitzman, P .0.1., 1972, Regulation of a-ketoglutarate dehydrogenase
activity in Acinetobacter, FEBS Lett., 22:323.
Weitzman, P.D.l., 1981, Unity and diversity in some bacterial citric acid
cycle enzymes, Adv. M icrob. Physiol., 22: 185.
Weitzman, P.D.l., and Dunmore, P., 1969, Citrate synthases: allosteric
regulation and molecular size, Biochim. Biophys. Acta, 171:198.
Weitzman, P.DJ., and laskowska-Hodges, H., 1982, Patterns of nucleotide
utilisation in bacterial succinate thiokinases, FEBS Lett.,
143:237.
Weitzman, P.DJ., and lones, D., 1968, Regulation of citrate synthase and
microbial taxonomy, Nature, 219:270.
Weitzman, P.D.l., Jenkins, 1'., Else, A.l., and Holt, R.A., 1986, Occurrence
of two distinct succinate thiokinases in animal tissues, F EBS
Lett., 199:57.
322
METABOLISM OF ALKANES BY ACINETOBACTER
O. Asperger and H.-P. Kleber
Bereich Biochemie
Karl-Marx-UniversWit Leipzig
Talstrasse 33
7010 Leipzig
Germany
INTRODUCTION
The ability to oxidise alkanes is widely distributed among prokaryotic

and eukaryotic microorganisms. A c i net 0 b act e r is among the bacterial
genera most often found in petroleum-contaminated habitats and has been
extensively used in studies of n-alkane oxidation (Einsele, 1983). Research
with this bacterium has contributed to the exploration of many aspects of
microbial n-alkane metabolism.
Interest in microbial alkane oxidation is multifarious. It encompasses

various practical aspects such as ecology, pollution and waste disposal,
problems of deterioration in the petroleum industry, geological prospecting,
manufacture of single cell protein or diverse biosynthetic products (amino
acids, organic acids, vitamins, lipids, emulsifiers etc.) on the base of
petroleum as sole source of carbon and energy, as well as the
biotransformation of hydrophobic chemicals to defined products by regio-
and stereoselective oxidation. Besides these practical applications, there is
a fundamental interest of chemists and biochemists in the biological
mechanisms underlying the oxidative functionalisation of such extremely
hydrophobic and poorly reactive compounds as n-alkanes.
Ever since the description of wax ester formation (Stewart and Kallio,
1959) by a Gram-negative, hexadecane-oxidising coccus (Stewart et aI.,
1959), originally called Micrococcus cerificans HOI-N (Finnerty et aI.,
1962), research with strains of Acinetobacter has contributed substantially
to most reviews of microbial n-alkane metabolism (e.g. McKenna and Kallio,
1965; van der Linden and Thijsse, 1965; Davis, 1967; Klug and Markovetz,
323
1971; Gutnick and Rosenberg, 1977; Ratledge, 1978: Aurich, 1979;
Einsele, 1983; Buhler and Schindler, 1984; Singer and Finnerty, 1984a,b).
There are some review articles devoted mainly to aspects of n-alkane
metabolism by Acinetobacter (Finnerty, 1977; Kleber et aI., 1983, 1985;
Finnerty and Singer, 1985; Rosenberg and Kaplan, 1987), but most consider
only special topics or are restricted to individual strains. It is now more than
ten years since this subject was covered in a general review on
Acinetobacter (Juni, 1978). It is therefore the aim of this article, by
summarising most of the available results related to n-alkane metabolism of
Acinetobacter, to outline the importance of Acinetobacter research for
the understanding of microbial hydrocarbon oxidation, and contribute to a
comprehensive view of the biochemistry and physiology of this genus.
Generally, the growth of bacteria on oil is thought to involve three

specific processes (Erickson et aI., 1973): (i) interaction of the cells with, and
uptake of, the insoluble substrate via adherence at the oil-water interface,
pseudo-solubilisation of the carbon source (Goma et aI., 1973), or uptake of
hydrocarbon dissolved in the bulk aqueous phase; (ii) the introduction of
molecular oxygen into the hydrocarbon and its subsequent biotransformation
to a substrate which can then enter the normal metabolic pathway of the
bacterial cell (Singer and Finnerty, 1984a): and (iii) the expression of
specific regulatory signals governing the primary pathways (Grund et aI.,
1975) and metabolism as a whole. These aspects will be considered by
presenting results on n-alkane assimilation and product formation, on
alkane uptake and alkane-induced morphological changes, as well as on the
enzymology of primary oxidation pathways and other metabolic sequences.
GENERAL FEATURES OF n-ALKANE OXlDA TION
Aliphatic hydrocarbon assimilation seems to be a common property of

most Acinetobacter strains. Non-utilisation appears to be the exception.
Of 106 strains tested by Baumann et ai. (1968) in their fundamental study on
the taxonomy of Acinetobacter, only 19 strains failed to assimilate and
utilise n-dodecane, n-tridecane and n-hexadecane. The alkane utilisers
were distributed among all seven phenotypic groups postulated by Baumann
et ai. (1968) and, of ten representative strains, only one did not assimilate
hydrocarbons. Similarly, the alkane-assimilating strains tested by Baumann
et ai. (1968) were also distributed among all six of the Acinetobacter
species proposed by Bouvet and Grimont (1986) on the basis of 12
genospecies identified by DNA hybridisation experiments. This means that
acinetobacters in general, regardless of whether they originate from soil and
water, as it is the case for most strains of the species "A. calcoaceticus
sensu stricto (Bouvet and Grimont, 1986), or from animal or human
1/
habitats, possess the ability to assimilate aliphatic hydrocarbons.
This conclusion is in good agreement with the assumption that,

generally, the genes encoding n-alkane assimilation are located on the
chromosome. The assumption is supported by the fact that plasmids have
324
not been detected in strains that have been well-studied with respect to n-
alkane assimilation (Acinetobacter sp. HOI-N, Singer & Finnerty, 1984b;
A.calcoaceticus 69-V, O. Asperger, unpublished results.) Data on the
genes encoding alkane assimilation by Acinetobacter are so far only
preliminary. Attempts by Singer and Finnerty (1984c) to obtain mutants by
means of transpositional methods failed, but, as a result of chemically
induced point mutations, 120 alkane-negative mutants (C 12 -C16 ) that were
palmitate-positive and also grew well with alcohols and aldehydes (C 12 -C 16 )
have been isolated from strains H01-N and BD413 (Singer and Finnerty,
1984b). The mutants clustered into ten groups on the basis of reciprocal
transformation crosses. From intergroup crosses, it was concluded that at
least two distinct unlinked loci (alk x and alk y) were necessary for alkane
oxidation by Acinetobacter.
In addition, certain strains of Acinetobacter may have special

capabilities with respect to n-alkane metabolism that are encoded by
plasmid DNA. Thus, the plasmid-carrying strain A. calcoaceticus RA57,
which can grow on, and disperse, crude oil, loses this ability concomitantly
with the curing of one of its four plasmids (pSR4), although its abilities to
assimilate long-chain n-alkanes when applied in the vapour phase, to adhere
to hydrocarbons, and to produce bioemulsifiers, were retained (Rusansky et
aI., 1987). Perhaps pSR4 encodes a factor which is tightly associated with
the cell surface.
High specific growth rates during cultivation on liquid minimal media

with n-hexadecane as sole carbon source, up to 1.3 /h in the case of
A. calcoaceticus EB104, indicate that Acinetobacter is highly adapted to
the oxidation of aliphatic hydrocarbons. However, n-alkanes are oxidised
only by cells grown on them or cultivated in the presence of appropriate
inducers (Aurich and Eitner, 1973; Asperger and Aurich, 1977; Asperger,
1990), thereby indicating the inducibility of the n-alkane assimilation
system. Since several non-hydrocarbon substrates did not repress the
induction of n-alkane oxidation capability (Haferburg et aI., 1983; Asperger,
1990), it is clear that catabolite repression does not operate. Thus,
induction must occur in direct response to the alkane molecule.
SUBSTRATE SPECIFICITY, PRODUCT PATTERNS

AND DEGRADATION PATHWAY
Most Acinetobacter strains tested for the ability to utilise n-alkanes

assimilate only compounds with more than 10 to 12 carbon atoms (Table 1).
Hence, Acinetobacter strains can be used as model organisms for the
assimilation of long-chain n-alkanes. Only a few strains, all of them capable
of synthesising cytochrome P-450 (cf. Table 7), additionally utilise middle-
chain n-alkanes (C 6 -C 9 ) as a sole source of carbon, provided that the
hydrocarbons are supplied in the vapour phase (Kleber et aI., 1983; Eremina
et aI., 1987).
325
Table 1. Utilisation of n-alkanes as sole carbon source by Acinetobacter
strains as a function of the chain-length
Strain Alkane chain length

Ref.
5 6 7 8 9 10 11 12-18 >lB
Acinetobacter sp. H01-N + + + + (1)

A.calcoaceticus 69-V + + + (2,3)
A.calcoaceticus EB106 + + (3)
A.calcoaceticus EBl14 + + (3)
A.calcoaceticus CCM2355 + + (3)
A.lwoffii CCM2376 + + (3)
A.lwoffii CCM5572 + + (3)
A.lwoffii NCIBB250 + + + + + (4)
A.lwoffii NCIB10533 + (5)
A.lwoffii 0-16 + (6)
A.calcoaceticus EB102 + + + + (3)
A.calcoaceticus EB104 + + + + + + + + (3)
A.calcoaceticus EBl13 + + + + (3)
References: (1) Finnerty et al., 1962; (2) Kleber et al., 1973; (3)
Kleber et al., 1983; (4) Fewson, 1967; (5) Grant, 1973; (6) Breuil et
al., 1978.
Chain-length specificity during assimilation of n-alkanes as sole

source of carbon does not reside in the substrate specificity of the primary n-
alkane-oxidising enzymes. Thus, respiratory studies revealed that all n-
alkanes (from n-hexane upwards) were oxidised not only by the middle-chain
n-alkane-assimilating strain EB104, but also by the strain A. calcoaceticus
69-V (Fig.l; Asperger and Aurich, 1977). Inhibition or retardation of
growth after addition of middle-chain n-alkanes to cultures of both strains
growing on n-hexadecane, or on a complex medium, indicated that sensitivity
of Acinetobacter to the toxic effect of middle-chain alkanes is probably the
reason for the restricted chain-length growth specificity. This fact has also
been reported for other n-alkane-assimilating microorganisms (Ratledge,
1978). Why certain strains of Acinetobacter (Table 1) are able to utilise
middle-chain n-alkanes (Eremina et al., 1987) has still to be resolved, but
there is a striking correlation between this property and the occurrence of an
alkane-inducible cytochrome P-450 in these strains (cf. Table 7).
It is clear from the respiratory studies (Fig. 1) that the substrate

spectrum of acinetobacters comprises all liquid and even solid n-alkanes.
Furthermore, a broad variety of n-alkane derivatives are oxidised, as can be
326
A
100
Z
to 75
0::
>-
l-
e
E 50
'0.
'"'"
0::
"" 25
10 15 20
Chain length
B
100
~'" 75
>-
l-
e
~ 50
'0.
'"
Qj
0::
.. 25
0
5 15
Chain length
Fig. 1. Dependence of respiratory rates on the chain length of n-alkanes

for resting cells of A. calcoaceticus EB104 CA) and A. calcoaceticus 69-V
(B) after growth on n - hexadecane ( _ ) or nonane (0- - - -0) . The
relative values are given as percentages of the respiration rate with n-
hexadecane as substrate, for which the rate was 700 nmol 02/min/mg
protein (data from Asperger and Aurich, 1977; Asperger, 1990).
deduced from growth and respiratory experiments (Table 2) or from studies

of product formation (Tables 3 and 4). According to Amund and Higgins
(1985), even-numbered phenylalkanes are totally degraded, via phenylacetic
acid, by an oil-degrading strain of A.fwoffii, whereas odd-numbered
phenylalkanes are transformed into dead-end products such as trans-
cinnamic and 3-phenylpropionic acid.
Stoichiometric oxygen/alkane coefficients determined by respiratory

measurements with the Clark electrode show that alkyl chains are rapidly
oxidised beyond the carboxylic acid stage to intermediary metabolites and
CO 2 (Asperger and Aurich, 1977). Although growth on, or respiration with
327
Table 2. Substrates of n-alkane monooxygenases of Acinetobacter as detected
by growth experiments or by respiration studies with n-alkane-induced cells
Class of Compounds positively tested with the strains

compound HOI-N a 69-Vb EB 104 c
n-Alkanes
Secondary Hexadecan-2-01 Hexadecan-2-01

alkanols Octan-2-01
Ketones Heptadecan-9-
one
Hexadecan-3-one Hexadecan-3-
one
Tridecan-7-one Tridecan-2-one
Alkenes l-Octadecene I-Hexadecene I-Hexadecene

I-Hexadecene I-Tetradecene
l-Octene
1so- Methylpentadecanes
alkanes (2-; 3-; 4-; 5-;
6 -; 7 - and 8-
isomers)
Phenyl- I-Phenyldodecane I-Phenyltridecane I-Phenyltri-

alkanes decane
2-Phenyldodecane l-Phenyldo-
de cane
I-Phenylundecane
I-Phenyldecane
Halo- l-Chlorohexa- I-Bromotetradecane I-Bromotetra-

alkanes decane decane
I-Bromodecane
Ethers Didecyl ether Dioctyl ether

Dinonyl ether Dihexyl ether
Dioctyl ether
Diheptyl ether
Thioether Dioctylsulphide
References: aFinnerty et al., 1962; Hankin and Kollatukudy, 1968; Stewart

et al., 1960; Kollatukudy and Hankin, 1968; Modrzakowski et al., 1977;
110drzakowski and Finnerty, 1980; bAsperger and Aurich, 1977; Aurich and
Eitner, 1973; Kleber et al., 1973; CAsperger, 1990.
328
Table 3. Lipid accumulation by Acinetobacter sp. HOl-N in (A) cells and
(B) culture broth during late exponential phase growth on n-hexadecane or n-
hexadecanol. Cell accumulation is expressed in terms of ~mol/mg dry
weight, and broth accumulation in terms of ~mol/l
Lipid Nutrient broth Hexadecane a Hexadecanol b

A B A B A B
Phospholipids 46 129 100

Glycerides 2.2 2.4 9.3 436 29
Wax esters 11 18 280 306 320
Free fatty acids 7.5 4 8.2 60 0.3
References: aMakula et al., 1975; bSinger et al., 1985.
di-terminally branched alkanes was not observed, the oxidation of the

terminal methyl group in such compounds may be possible since the
formation of pristanol and pristanediol from pristane by Acinetobacter has
been reported in the patent literature (Hiratsuka et aI., 1980).
A rather unusual method of degradation of the carbon skeleton has

been proposed with respect to the oxidation of symmetrical dialkyl ethers by
Acinetobacter sp. H01-N (Modrzakowski et aI., 1977; Modrzakowski and
Finnerty, 1980). In cells grown on these substrates as sole source of carbon,
alkoxycarboxylic acids having the same carbon number as the substrates were
detected in the cellular phospholipid fraction, and alkoxyacetic acids were
also found in the neutral lipid fraction as well as in the spent culture broth.
Only acetate derivatives, but not propionate derivatives (even in the case of
odd-numbered alkyl residues of the ethers) were described. On the basis
that dicarboxylic acids with a carbon number of n-2 (n =carbon number of
ether alkyl residue) were detected in the cells, the authors postulated a
scission of the initial alkoxycarboxylic acid product in the position II to the
ether oxygen. More recently, Finnerty (1984) reported the formation of
short-chain n-alkanes, together with a dicarboxylic acid, from long-chain n-
alkanes or long-chain fatty acids, but did not give experimental details and
yields of products.
In addition to the different types of bioemulsifiers produced in

enlarged amounts during growth on n-alkanes (cL Table 6), various cellular
constitutents and products are accumulated in which metabolites of primary
alkane oxidation are incorporated. These include wax esters, glycerides,
phospholipids, and small amounts of free fatty acids (cf. Table 3; Vachon et
aI., 1982; Muller and Voigt, 1982). Phospholipids accumulate mainly inside
the cell, reflecting the generation of intracytoplasmic membranes during
growth on hydrophobic substrates (Kennedy and Finnerty, 1975; Muller et
329
Table 4. Components of wax esters produced by Acinetobacter sp. HOl-N
after growth on different n-alkanes and alkane derivatives
Growth substrate Alcohol Fatty acid Ref.
n-Tetradecane Tetradecyl Hexadecanoic (1)

n-Hexadecane Hexadecyl Hexadecanoic (1, 2)
n-Heptadecane Heptadecyl Heptadecanoic (3)
Pentadecanoic
Hexadecanoic
n-Octadecane Octadecyl Octadecanoic (1)
Hexadecanoic
l-Hexadecene l5-Hexadecenyl Hexadecanoic (4)
l-Octadecene l7-0ctadecenyl Hexadecanoic (4)
Heptadecanoic
Octadecanoic
l-Chlorohexadecane l6-Chlorohexa- l6-Chlorohexa- (5)
decyl decanoic
Hexadecanoic
References: (1) Stewart and Kallio, 1959; (2) Stewart et al., 1959; (3)
Stevenson et al., 1962; (4) Stewart et al., 1960; (5) Kollatukudy and
Hankin, 1968.
aI., 1983). Wax ester accumulation by Acinetobacter is not restricted to n-

alkane assimilation. It is found in cells growing on n-hexadecanol (Table 3),
and has also been reported for non-hydrocarbon-grown acinetobacters
(Gallagher, 1971; Fixter and Fewson, 1974; Bryn et aI., 1977; Neidleman
and Geigert, 1983; Fixter et aI., 1986). All 19 strains of Acinetobacter
tested by Fixter et al. (1986) contained wax esters and 16 of these strains had
increased wax contents when harvested from stationary phase of N-limited
batch cultures. It appears that wax esters are widespread energy storage
components in the genus Acinetobacter and their mobilisation under
conditions of C-starvation has been shown (Fixter et al. 1986).
The chemical structures of the fatty alcohol moiety and, to a certain

degree, those of the fatty acids in the waxes of n-alkane-grown cells
correspond to those of the hydrocarbon substrates (Table 4). Other cellular
lipids are characterised by a significant increase in the percentage of fatty
acids having the same carbon number as the alkane that has been used as
carbon source (Table 5). This shows that fatty acids originating from the
primary oxidation of the alkanes are directly incorporated into the cellular
lipids. Furthermore, a terminal oxidation pathway of the n-alkanes can be
deduced from these results. Additional support for this suggestion is given
by the failure of diterminal substituted alkane derivatives, such as
330
Table 5. Percentage fatty acid composition of Acinetobacter sp. H01-N after
growth on different n-a1kanes (data from Maku1a and Finnerty, 1968)
Fatty Nutrient Acetate n-A1kanes

acid broth C13 C14 C15 C16 C17
C10 0.3 0.1 l.6 l.7 l.1 0.8

Cll l.7 l.2 2.0
C12 3.2 5.0 7.2 l.3 7.8 l.4
c13 7.7 l.1 l.8
C14 0.8 0.5 2.2 28.4 0.2 2.0 0.4
C15 a l.2 68.B 10.2
C16 a 36.6 41.1 28.B 27.B 4.5 72.9 4.4
Cll a 15.4 7.0 0.9
C1B a 48.B 53.3 34.0 34.B 14.9 9.1 9.1
a The sum of saturated and unsaturated fatty acids.
hexadecane-1,I5-dion or 1,12-dibromododecane (Asperger and Aurich,

1977; Asperger, 1990), to support respiration by alkane-grown cells, whilst
the corresponding mono terminal substituted derivatives are oxidised (cf.
Table 2). Still more stringent evidence was provided by the identification of
a fatty acid of the same substrate chain-length as the main product of alkane
oxidation catalysed by cell-free extracts of A. calcoaceticus strain 69-V
(Asperger et al., 1978a).
ALKANE UPTAKE AND MORPHOLOGICAL IMPLICATIONS OF

ALKANE ASSIMILATION
A serious problem in the biological oxidation of hydrocarbon

compounds arises from their extremely low solubility in water. Therefore, in
addition to enzymes for the initial oxidative attack on the hydrocarbon
skeleton, alkane-oxidising microorganisms also require appropriate
mechanisms to facilitate translocation of the alkane molecule from the bulk
hydrocarbon phase to the active site of the enzyme. With respect to the
intact cell, the interaction with alkanes seems to be facilitated by
Acinetobacter in two ways: emulsification and adherence.
Surface-active compounds of low molecular mass have, to date, not

been described for Acinetobacter. The bioemulsifiers produced by
Acinetobacter are either high molecular weight soluble compounds such as
"Emulsan" (Table 6; Rosenberg et al., 1979; Zuckerberg et al., 1979;
Rosenberg, 1986) or they are particulate and form vesicles constituted out of
lipopolysaccharides, phospholipids and proteins Cfable 6; Kappeli and
Finnerty, 1979; Claus et al., 1984; Borneleit et al., 1988). While the former
331
Table 6. Formation of surface-active compounds by Acinetobacter
Strain Product designation Chemical composition Ref.
RAG-l/ Emulsan Polysaccharide-protein (1)

BD413 complex (covalently bound (2)
fatty acids) (3)
(4)
2CA2 Emulsifier 2CA2 Protein-lipid-carbohydrate (5)
complex
H01-N Extracellular Lipopolysaccharide-phospho- (6)
vesicles lipid-protein vesicles
Soluble Emulsifier Lipoprotein (7)
69-V Extracellular Lipopolysaccharide-phospho- (8)
vesicles lipid-protein vesicles
References: (1) Rosenberg et al., 1979; (2) Zuckerberg et al., 1979; (3)
Kaplan and Rosenberg, 1982; (4) Rosenberg and Kaplan, 1987; (5) Neufeld
and Zajic, 1984; (6) Kappeli and Finnerty, 1979; (7) Kappeli and Finnerty,
1980; (8) Claus et al., 1984.
seem to represent partially d~graded capsular constituents (Rosenberg and

Kaplan, 1987), the latter appear to derive from the outer membrane
(Kappeli and Finnerty, 1979; Borneleit et ai., 1988). They all affect n-
alkane accommodation, mainly by the stabilisation of emulsions (Rosenberg
and Kaplan, 1987).
However, since the enhancement of alkane uptake rates in the

presence of extracellular vesicles (Claus et al., 1984) did not exceed the rates
of alkane respiration measured with ultrasonically dispersed alkanes
(Asperger and Aurich, 1977; Haferburg et al., 1983), it appears unlikely that
the extracellular surface-active agents are directly involved either in the
mechanism of alkane transport across the cellular membranes or in the
interaction of the alkanes with the alkane-oxidising enzymes. In agreement
with this, vesicle phospholipids were found not to be imported into the cell of
Acinetobacter sp. HOI-N (Finnerty and Singer, 1985).
The other mechanism - adherence to alkane droplets, as demonstrated

by ultramicrographs for various strains of Acinetobacter (Kennedy et aI.,
1975; Rosenberg and Kaplan, 1987) - seems to be achieved by thin fimbriae.
These have been observed for A. calcoaceticus RAG-l (Rosenberg et aI.,
1982) and reported for another alkane-grown strain formerly designated as
M oraxella glucidolytica (Horisberger, 1977). Furthermore, Neufeld and
Zajic (1984) demonstrated the reduction of surface tension by hexadecane-
grown cells themselves.
332
One of the exciting discoveries arising from studies of n-alkane
assimilation by Aci net 0 b a c te r has been the detection of intracellular
membrane-surrounded alkane inclusions and intracytoplasmic membranes in
alkane-grown Acinetobacter sp. H01-N (Kennedy et aI., 1975; Kennedy
and Finnerty, 1975; Scott and Finnerty, 1976a,b) and A. calcoaceticus 69-
V (Muller et aI., 1983). Inclusion membranes of strain H01-N have been
described as monolayers (Scott and Finnerty, 1976b), whereas bilayers have
been claimed for strain 69-V (Miiller et aI., 1983). These intracellular
membranes have been discussed as possible sites for the location of the
initial enzymes of alkane oxidation, but firm evidence in support of this view
has still to be provided.
ENZYMOLOGY RELATED TO ALKANE ASSIMILATION AND ITS

REGULATION
In most organisms, the initial metabolism of n-alkanes takes the form

of monoterminal oxidation:
R-CH3 ----;;. R-CHpH ----;;. R-CHO ----;;. R-COOH
A similar pathway of n-alkane oxidation by Acinetobacter can be deduced

from studies with whole cells and cell-free extracts as described above.
Enzymes required for such a pathway are: alkane monooxygenase, alcohol
dehydrogenase or oxidase, and aldehyde dehydrogenase or oxidase.
Furthermore, the regulation of enzymes controlling intermediary
metabolism and lipid metabolism may be substantially affected in n-alkane-
assimilating cells.
Initial Step ofn-Alkane Oxidation
Evidence in support of the oxygenative nature of the initial enzyme in

alkane oxidation by Acinetobacter sp. H01-N was first provided by Stewart
et ai. (1959), who demonstrated the incorporation of 1802into the alkane-
derived constituents of wax esters. Based on the fact that n-alkyl
hydroperoxides are rapidly oxidised by Acinetobacter sp. HOI-N (Stewart
et aI., 1959), a pathway has been suggested that involves n-alkyl
hydroperoxide as the initial product of alkane oxidation and, consequently,
requires a dioxygenase as the first enzyme, followed by an alkyl
hydroperoxide reductase (Finnerty, 1977; Singer and Finnerty, 1984a).
However, this hypothesis lacks enzymological proof, since alkanecoxidising
activity has not, to date, been demonstrated in cell-free extracts of this
strain.
Cell-free n-alkane oxidation - generally until now demonstrated only

for a few alkane-assimilating bacteria that had mostly been grown on middle-
chain n-alkanes and tested with such substrates - was achieved in our
laboratory with strain A. calcoaceticus 69-V (Aurich et aI., 1977; Asperger
et aI., 1978a) and strain EBI04 (Asperger et aI., 1985a). In both cases,
333
Table 7. Cytochrome composition of various strains of Acinetobacter
grovm on n-hexadecane
Strain Cytochromes Ref.
abc d 0 P-450
69-V + + (1)
CCM2355 + + (2)
CCM5591+ + + (2 )
CCM5595 + + (2)
CCM5572 + + (2)
CCM2376 + + (2)
EB1l4 + + (2)
EB102 + + + (2)
EB103 + + + (2)
EB104 + + + (2)
EB10C + + + (2)
EB10D + + + (2)
EB1l3 + + + (2)
A6E + (3)
ATCC33305 (3)
AH60 (3)
DON2 (3)
H01-N + + + (4)
References: (1) Asperger et al., 1978b; (2) Asperger et al.,

1981b; (3) Wyndham, 1987; (4) Ensley and Finnerty, 1980.
alcohols and fatty acids were identified as the main products of alkane
oxidation, indicating the functioning of alkane monooxygenases, alcohol
dehydrogenases and aldehyde dehydrogenases in cell-free extracts. Alkane
oxidation by cell-free extracts of strain 69-V and EB104 was oxygen- and
NADH-dependent, providing evidence for a monooxygenase as the initial
enzyme of alkane oxidation.
Troublesome assay procedures and enormous loss of activity after cell

disruption render the purification of the oxygenases based on activity
measurements highly difficult. This is reflected by the fact that to date,
apart from methane monooxygenases (Dalton, 1980), the rubredoxin-
dependent system of the middle-chain n-alkane-utilising bacterium
Pseudomonas oleovorans (Coon et a!., 1973) represents the only alkane
monooxygenase isolated from bacteria other than Acinetobacter. For
Ac in e to b a c te r, a partially successful approach to the characterisation of
the alkane monooxygenases was to search for, and isolate, electron transport
334
proteins that were potential components of monooxygenases. Cytochrome
P-450 - the alkane monooxygenase of yeasts - is absent from most of the
strains of Acinetobacter that have been traditionally investigated, but
occurs in certain strains (Table 7; Asperger et aI., 1978b, 1981b). It is
induced in strain EB104 by a broad variety of aliphatic and aromatic
hydrocarbons (Asperger et aI., 1984, 1985b) and induction can be enhanced
by oxygen limitation (Asperger et al., 1986).
Cytochrome P-450 was isolated from A. calcoaceti cus EB104 - either

grown on n-hexadecane (Miiller et aI., 1989) or in nutrient broth with the
addition of n-hexane as inducer (Asperger, 1990) - as an electrophoretically
homogeneous protein with a M, of 52,000 and spectral properties typical of
other cytochromes P-450 (Muller et aI., 1989). Purified cytochrome P-450 is
enzymically reduced by NADH in the presence of a ferredoxin and a
ferredoxin reductase, both isolated from the same bacterium .(Asperger et
aI., 1983, 1985a). The ferredoxin has a M, of about 9,000 and absorption
maxima of the oxidised form at 415 and 460 nm, characteristicfor a (2Fe-2S)-
ferredoxin. Identification of the cytochrome P-450 system (Fig.2) as an
alkane monooxygenase was based on the following evidence: (a) induction of
cytochrome P-450 by the unoxidised n-alkane, but not by an intermediate
oxidation product, as shown by regulatory studies with a broad variety of
different inducers (Kleber et aI., 1985); (b) inhibition of cell-free alkane-
oxidising activity by CO, typical of cytochrome P-450-dependent reactions
(Asperger et aI., 1985a); (c) recovery of cytochrome P-450, ferredoxin and
ferredoxin reductase as well as alkane-oxidising activity from the cytosolic
protein fraction (Asperger et aI., 1985a); (d) concomitant induction of
cytochrome P-450 and alkane monooxygenase activity by the non-aliphatic
hydrocarbon biphenyl (Asperger et al., 1985a); (e) reconstitution of alkane
monooxygenase activity by the isolated proteins cytochrome P-450,
ferredoxin and ferredoxin reductase (Asperger et al., 1985a).
Hitherto, the detection of alkane-inducible cytochrome P-450 in

partially purified alkane monooxygenase preparations of Rhodococcus
rhodochrous ATCC 19067 (formerly designated as Corynebacterium sp.
7EIC) (Cardini and Jurtshuk, 1970) had been the only indication for the
existence of cytochrome P-450-dependent alkane monooxygenases in
bacteria. The isolation of such a system from Acinetobacter provides
more convincing evidence and it may be supposed that still more cytochrome
P-450-dependent alkane monooxygenases will be found in bacteria.
There are, however, other n-alkane-assimilating bacteria in which

cytochromes P-450 have definitely not been found, including several strains
of Acinetobacter (Table 7). The terminal oxidase of the alkane
monooxygenase of these cytochrome P-450-negative Acinetobacter strains
remains unknown, but a rubredoxin (Aurich et al., 1976) and NADH-
dependent rubredoxin reductase (Claus et al., 1978, 1979) have been isolated
from alkane-grown cells of A. calcoaceticus 69-V. The rubredoxin
exhibited an optical absorption spectrum typical of this class of iron-sulphur
proteins with maxima at 380 and 490 nm in the oxidised state, a M, of 6,000
335
and an iron content of 1 mol/mol (Aurich et aI., 1976). Since the rubredoxin
was induced by growth on n-alkanes, as determined by means of a
radioimmunoassay (Claus et aI., 1980a,b), it can be supposed that the alkane
monooxygenase of A. calcoaceticus 69-V is a rubredoxin-dependent
enzyme comparable to that of P.oleovorans (Fig.2; Coon et aI., 1973).
While alkanes are oxidised only by alkane-grown cells or by induced

cells, fatty alcohols, fatty aldehydes and fatty acids are also respired by cells
grown on non-hydrocarbons (Aurich and Eitner, 1973; Asperger and Aurich,
1977). Hence, the alkane monooxygenases of Acinetobacter are inducible
enzymes, which is confirmed by the inducibility of cytochrome P-450
(Asperger et a!., 1984) and rubredoxin (Claus et a!., 1980b), whereas
constitutive alcohol and aldehyde dehydrogenases, as well as a constitutive £-
oxidation pathway, probably also exist. So far as the cellular location of the
alkane monooxygenases of Acinetobacter is concerned, there is no
convincing evidence for them being integral proteins of the cytoplasmic
membranes. Activity of the enzymes from strains 69-V (Asperger et aI.,
1978a) and EB104 (Asperger, 1990) is stimulated by particulate fractions,
and also cytochrome P-450 can, in part, be recovered from particulate
fractions (Miiller et aI., 1989). However, it could be unequivocally
demonstrated that cytochrome P-450 is not an integral protein of the
cytoplasmic membrane (Muller et aI., 1989) and that, under conditions of low
intracellular alkane concentration, it is localised in the cytosolic protein
69-V
NADH, Alkane
Rubredoxin
0,
H,o
Rubredoxin Terminal Alkanol
Reductase Oxidase
NAD'
EB-104
Alkane
NADH,
0,
H,o
Ferredoxin Cytochrome
Reductase Alkanol
NAD" P-450
Fig. 2. Schematic representation of the proposed alkane monoxygenases in

Acinetobacter strains 69-V and EB104.
336
fraction (Asperger et aI., 1983). It may therefore be proposed that the
enhancement of activity by particulate cell fractions occurs because alkanes
can be accommodated by the help of membranous phospholipid-rich
structures. Reconstitution experiments with cell-free components of the
alkane monooxygenase system should help to test the validity of this
hypothesis. With the cytochrome P-450 system from Acinetobacter, a cell-
free alkane monooxygenase from long-chain n-alkane-assimilating bacteria
is available for the first time. Because of its soluble nature, it should be a
suitable model system for exploring the mechanisms that achieve high
intracellular conversion rates of such extremely hydrophobic substrates as n-
alkanes.
The substrate specificity of the alkane monooxygenases of

A.calcoaceticus 69-V and EB104 is reflected by the substrate pattern
listed in Table 2, considering that these substrates were oxidised only by n-
alkane-induced cells and that alcohol and aldehyde dehydrogenases are
present constitutively.
Alcohol and Aldehyde Dehydrogenases
Some data on alcohol and aldehyde dehydrogenases of strains 69- V,

EB104 and HOI-N are summarised in Tables 8 and 9. Obviously there are
multiple enzymes, some of which are induced by growth on n-alkanes. Of
the alcohol dehydrogenases, only the constitutive, cytosolic NADP+-
dependent enzyme from A. calcoaceticus 69-V has been purified (Tauchert
et aI., 1976). Firm conclusions about the importance of inducible alcohol
dehydrogenases for the degradation of n-alkanes by Acinetobacter
therefore cannot yet be drawn. Induction of a membrane-bound, NADP+-
dependent aldehyde dehydrogenase is well established (Aurich and Eitner,
1977; Aurich et aI., 1983; Singer and Finnerty, 1985a). This enzyme has
been purified (Sorger and Aurich, 1978), reconstituted into liposomes
(Aurich et aI., 1985) and kinetically characterised (Aurich et aI., 1987).
There is controversy with respect to the location of the aldehyde

dehydrogenase of A.calcoaceticus 69-V. Sorger et ai. (1986) interpreted
the ultracytochemically-demonstrated aldehyde-dependent accumulation of
NADPH around the inclusions of hexadecane-grown cells as the location in
an inclusion-surrounding membrane. However, Fischer et al. (1984)
postulated cytoplasmic membrane-bound enzyme in the same strain.
Enzymes of Intermediary Metabolism
Fatty acids produced by the initial steps of alkane oxidation are either
directly incorporated into cellular lipids or waxes, or they are rapidly further
degraded by B-oxidation. Nothing is known about the specificity or
regulation of the enzymes involved in these processes. There has been
intensive study of the tricarboxylic acid cycle enzymes and the anaplerotic
sequences involved in alkane assimilation by A. calcoaceticus 69-V (Kleber
and Aurich, 1973, 1974; Kleber, 1978). lsocitrate lyase and malate
337
(.oJ
(.oJ
OJ
Table S. Alcohol dehydrogenases of alkane-cultivated strains of Acinetobacter
Strain Enzyme Cosubstrate Localisation Induction Maximum Ref

Sp.Act. a
69 - V I NADP+ Cytosol Constitutive 7(C S )b (1)

II DCPIP Membranes ? 4(C S ) (2)
EB104 1 NADP+ Cytosol Constitutive 24(C S ) (3)
II DCPIP Membranes Constitutive 3(C S ) (3)
HOI-N ADH-Ac NAD+ Cytosol Ethanol 92.6(C 2 ) (4)
ADH-B NADP+ Cytosol Constitutive 11 (C2) (4)
HDH NAD+ Cytosol Hexadecanol 5.9(C l6 ) (4)
HDH Membranes Hexadecane (4)
anmol/min/mg protein; bChain-length of the substrate; cADH: Alcohol

dehydrogenase; HDH: Hexadecanol dehydrogenase.
References: (1) Tauchert et al., 1976; (2) Tauchert et al., 1975; (3)
Jirausch et al., 19S6; (4) Singer and Finnerty, 19S5b.
Table 9. Aldehyde dehydrogenases of alkane-cultivated strains of Acinetobacter
Strain Enzyme Co- Localisation Induction Maximum Ref.

substrate Sp.Aet. a
69-V NADP+ Membranes Alkane 600(C S )b (l)

Alkanol (2 )
EBlO4 NADP+ Membranes ? ? (3 )
HOl-N FALDH-a c NADP+ Membranes Hexadeeane 14/(C lO ) (II)
Hexadecanol (4)
FALDH-b c NADP+ Membranes Constitutive ? (4)
ALDH c NAD+ Cytosol Ethanol 16.5(C 2 ) (4)
anmol/min/mg protein; bChain-length of the substrate; cFALDH: Fatty aldehyde

dehydrogenase; ALDH: aldehyde dehydrogenase.
References: (1) Aurich and Eitner, 1977; (2) Aurich et a1., 1983; (3)
Klossek et a1., 198.5; (4) Singer and Finnerty, 198.5a.
w
w
co
dehydrogenase, key enzymes for the metabolism of excess C2 -substrates, are
repressed in the presence of carbohydrates, phosphoenolpyruvate, or during
growth on C4-dicarboxylic acids. However, they are induced by n-alkanes,
fatty acids, and acetate, whereas malic enzyme' is repressed under these
conditions (Kleber, 1973, 1978), avoiding the loss of C4 -units of the citric
acid cycle. In addition, fatty acids and acetyl-CoA inhibit the latter enzyme
(Kleber, 1975). Thus, this Acinetobacter is well-equipped for the
metabolism of n-alkanes via oxidation to fatty acids followed by B-oxidation.
Induced levels of isocitrate lyase and malate dehydrogenase, as well as the
repression of malic enzyme in n-alkane-grown cells, contribute to such a
catabolic pathway.
Enzymes of Lipid Metabolism
Alkane metabolism leads to significant changes in cellular lipid

composition and to the formation of several lipid products such as waxes,
glycerides or glycolipids. However, little is known about how enzymes
involved in lipid biosynthesis or catabolism are regulated during n-alkane
assimilation. Scott et al. (1976) found that cytoplasmic membranes of
Acinetobacter sp.H01-N contain an unusually high specific activity of
phosphatidic acid cytidyl transferase (McCaman and Finnerty, 1968) which is
increased threefold in n-alkane-grown cells (Finnerty, 1980). Elevated
activities of c..-glycerol-CMP phosphatidyl transferase, L-serine-CMP
phosphatidyl transferase and phosphatidyl serine decarboxylase in the
cytoplasmic membrane have also been measured, whereas phosphatidic acid
phosphatase was decreased during growth on n-hexadecane compared with
nutrient broth-grown cells (Finnerty and Singer, 1985). Since neither
phospholipid content nor the composition of cytoplasmic membranes
depended on the carbon source (Makula and Finnerty, 1970,1971; Makula et
al., 1975), it is possible that the elevated enzyme activities may reflect the
formation of intracellular membranes during growth on n-alkanes.
A phospholipase A, detected in the outer membrane of

Acinetobacter sp. H01-N (Torregrossa et al., 1977b) by studying the
occurrence of lysocardiolipin (Torregrossa et al., 1977a), was two to three
times more active in hexadecane-grown than in nutrient broth-grown cells
(Torregrossa et al., 1977b). In addition to cardiolipin, it also hydrolyses
phosphatidylglycerol and phosphatidylethanolamine (Torregrossa et al.,
1978). An esterase with high activity against palmitoyl-CoA has been
described as a marker enzyme of the outer membrane of A. calcoaceticus
69-V (Claus et al., 1985; Fischer,1986); however, the ability of this enzyme
to serve as a phospholipase and its regulation by the carbon source have not
been studied. It would be of interest to discover whether these outer
membrane hydrolase activities are functionally linked with the formation of
vesicles and/or capsular materials and therefore with the capacity of
acinetobacters to deal with n-alkanes. A. calcoaceticus RAG-l produces
an extracellular esterase for which the bioemulsifier emulsan is a substrate
(Shabtai and Gutnick, 1985), and it has been proposed that it has a role in
emulsan release from the cell surface.
340
It remains an open question whether or not formation of lipase by
Acinetobacter plays a functional role in n-alkane metabolism. Production
of lipase during growth on n-alkanes has been reported for strain A.lwoffii
016 (Breuil et aI., 1978) and for strain A. calcoaceticus 69-V (Haferburg
and Kleber, 1982, 1983). Little or no activity is observed during growth on
defined media with non-hydrocarbons as carbon sources. Lipase production
occurs in complex media and is largely enhanced by fatty acid esters and
certain detergents (Breuil et aI., 1978; Haferburg and Kleber, 1982). Thus,
lipase production may be more a consequence of, than a prerequisite for,
alkane metabolism. There are no other extracellular enzymes known so far
whose production is linked to alkane utilisation.
Enzymes of fatty acid biosynthesis have not yet been studied in detail,
although Sampson and Finnerty (1974) found evidence for their repression in
Acinetobacter sp. H01-N growing at the expense of n-hexadecane.
Respiratory Chain and Electron Transport Systems
Alkane assimilation does not seem to influence energy transducing

systems. Thus, studies with A. calcoaceticus 69-V (Asperger et aI., 1978b,
1981a) and Acinetobacter sp. H01-N (Ensley and Finnerty, 1980) did not
reveal significant changes of respiratory enzyme activities in response to n-
alkanes as carbon sources. In hexadecane-grown A. calcoaceticus 69-V,
membrane-bound oxidase systems for L-malate, D-glucose and L-glutamate
(as well as the corresponding substrate:acceptor oxidoreductases) occur in
addition to the classical oxidase systems for NADH and succinate (Asperger
et aI., 1981a). The same enzymes with similar specific activities are also
found in cells grown at the expense of other carbon sources. Both strains
contain cytochromes band 0, but no cytochrome c; under conditions of low
aeration cytochrome d is also found. Compared with non-hydrocarbon
carbon sources, growth on n-hexadecane did not significantly change the
specific cytochrome content of A. calcoaceticus cells (Asperger et aI.,
1978b). In the case of Acinetobacter sp. H01-N, only moderate changes of
the cytochrome 0 content of cytoplasmic membranes have been observed in
response to the carbon source (Ensley and Finnerty, 1980).
OUTLOOK
This review demonstrates that numerous insights into bacterial n-

alkane metabolism have been gained by investigating several strains of the
genus Acinetobacter. However, various questions remain to be resolved
before the pathway taken by an alkane molecule from the bulk hydrocarbon
phase to the active site of enzymes inside the cell and its subsequent
conversion to a fatty acid can be depicted in an even more convincing
fashion. One of the main intellectual challenges concerns the penetration of
alkanes across the cell envelope to the initial oxygenative enzyme system,
wherever it is located. All the hypotheses about possible mechanisms of this
process still lack unequivocal evidence. Location of the alkane
341
monooxygenases in cellular membranes is not very probable. Nevertheless,
the necessity of superstructural assemblies for n-alkane oxidation becomes
evident when rates of n-alkane conversion by intact cells (up to 100
nmol/min/mg protein), as calculated from the growth rates, are compared
with maximum activities of cell-free extracts or reconstituted alkane
monooxygenase systems (about 1 nmol/min/mg protein). It is to be hoped
that the isolated cytochrome P-450 system from A. calcoaceticus EBI04
will provide an appropriate instrument for studying these questions and for
gaining more insights into the cellular organisation of bacterial systems that
can oxidise strongly hydrophobic compounds.
Furthermore, several practical aspects of the capability of

Acinetabacter for alkane oxidation should be considered. Principally, the
high growth rate on long-chain n-alkanes makes it appear that
Acinetabacter might be an appropriate organism for the manufacture of
biosynthetic products, the production of single cell protein and the
bioremediation of oil-contaminated soil or water. The latter field of
application is favoured by the generation of surface-active compounds for
which a broad spectrum of other applications exists (Haferburg et aI., 1986;
Rosenberg, 1986). In addition, the surface-active compounds, wax esters or
glycerides produced during growth on n-alkanes or derivatives of them may
be exploited as sources of special carbohydrates, fatty acids or fatty alcohols.
Another aspect to be considered is the use of alkane-induced

Acineto b ac ter cells or' enzymes for the biotransformation of selected
substrates to defined products. Practical use of the alkane monooxygenases
of Acinetobacter may be favoured by the following features: (a) high
regioselectivity for the hydroxylation of the alkyl terminus; (b) high specific
activities for the substrate conversion, which can be deduced from the rapid
growth rates on n-hexadecane or from respiratory measurements with resting
cells; and (c) broad specificity against chain-length or different substituted
derivatives, which allows the conversion of a broad variety of alkyl chain-
containing compounds. Biotransformation processes may be achieved with
whole cells or with isolated enzymes. A serious disadvantage of whole cells
is the rapid consecutive degradation of the primary oxidation products. This
could be circumvented by product recovery in wax esters under appropriate
fermentation conditions, by the selection of mutants prepared by classical
methods, or by transfer of the alkane monooxygenase genes into other hosts
via appropriate vectors. The isolation of cytochrome P-450 from
A.calcoaceticus EB104 and the start that has been made on its cloning
provides the basis for such strategies.
REFERENCES
Amund, 0.0., and Higgins, I.J., 1985, The degradation of I-phenylalkanes

by an oil-degrading strain of Acinetobacter lwaffi, Ant. van
Leeuw. J. Microbiol. Serol., 51:45.
342
Asperger, 0., 1990, Alkanoxidation und Elektronentransportsysteme bei
Acinetobacter calcoaceticus unter besonderer Bertick-
sichtigung des Vorkommens von Eisenschwefelproteinen und
Cytochrom P-450, Dissertation B. Leipzig: Karl-Marx-
Universitiit.
Asperger, 0., and Aurich, H., 1977, Anwendung polarographischer O 2,-
Messungen auf die Veratmung von n-Alkanen und deren
Derivaten durch Acinetobacter calcoaceticus, Zeits. Allg.
Mikrobiol., 17:419.
Asperger, 0., Futtig, A., Behrends, B., Kleber, H.-P., and Aurich, H., 1978a,
Oxidation langkettiger n-Alkane durch zellfreie Extrakte von
Acinetobacter calcoaceticus. Identifizierung und
Bestimmung von n- Tetradekansaure nach Inkubation mit n-
Tetradekan, Wiss. Zeits. Karl-Marx-Univ. Leipzig, Math.
N aturwiss. Reihe, 27:3.
Asperger, 0., Kleber, H.-P., and Aurich, H., 1978b, Zytochrom-
Zusammensetzung von Acinetobacter calcoaceticus, Acta
Bioi. M ed. German., 37:191.
Asperger, 0., Borneleit, P., and Kleber, H.-P., 1981a, Untersuchungen zum
Elektronentransportsystem in Acinetobacter calcoaceticus,
Abh. Akad. Wiss. DDR, 3:259.
Asperger, 0., Naumann, A., and Kleber, H.-P. 1981b, Occurrence of
cytochrome P-450 in Acinetobacter strains after growth on n-
hexadecane, FEMS Microbiol. Lett., 11:309.
Asperger, 0., Miiller, R., and Kleber, H.-P., 1983, Isolierung von Cytochrom
P-450 und des entsprechenden Reductasesystems aus
Acinetobacter calcoacticus, Acta Biotech., 3:319.
Asperger, 0., Naumann, A., and Kleber, H.-P., 1984, Inducibility of
cytochrome P-450 in Acinetobacter calcoaceticus by n-
alkanes, Appl. Microbiol. Biotech., 19:398.
Asperger, 0., Miiller, R., and Kleber, H.-P., 1985a, Reconstitution of n-
alkane-oxidising activity of a purified cytochrome P-450 from
Acinetobacter calcoaceticus strain EB104, in: "Cytochrome P-
450, Biochemistry, Biophysics and Induction," p.447, L. Vereczkey,
and K. Magyar, eds., Elsevier Science Publishers, Amsterdam.
Asperger, 0., Stiiwer, B., and Kleber, H.-P., 1985b, Aryl hydrocarbons as
inducers of cytochrome P-450 in Acinetobacter calcoaceticus,
Asperger, 0., Sharychev, A.A., Matyashova, R.N., Losinov, A.B., and Kleber,
H.-P., 1986, Effect of oxygen limitation on the content of n-
hexadecane-inducible cytochrome P-450 in Acinetobacter
calcoaceticus strain EB104, J. Basic Microbial., 26:571.
Aurich, H., 1979, Die Oxidation aliphatischer Kohlenwasserstoffe durch
Bakterien, Sitz. Akad. Wiss. DDR, Math. Naturwiss. Tech.,
16(N):3.
Aurich, H., and Eitner, G., 1973, Oxidation von n-Hexadecan durch
Acinetobacter calcoaceticus. Bedingungen und Induktion
beteiligter Enzyme, Zeits. Allg. M ikrobiol., 13:539.
343
Aurich, H., and Eitner, G., 1977, Induktion der NADP+ -abhangigen
Aldehyddehydrogenase durch Kohlenwasserstoffe bei
Acinetobacter calcoaceticus, Zeits. Allg. Mikrobiol.,
17:263.
Aurich, H., Sorger, D., and Asperger, 0., 1976, Isolierung und
Charakterisierung eines Rubredoxins aus Acinetobacter
calcoaceticus, Acta Biol.M ed. German., 35:443.
Aurich, H., Bruckner, A., Asperger, 0., Behrends, B., and Futtig, A., 1977,
Oxidation von n- Tetradecan-1- 14 C durch zellfreie Extrakte aus
Acinetobacter calcoaceticus, Zeits. AUg. Mikrobiol.,
17:249.
Aurich, H., Sorger, H., Bergmann, R., Lasch, J., and Koelsch, R., 1983,
Purification and characterization of membrane-bound aldehyde
dehydrogenase from Acinetobacter calcoaceticus grown on
long-chain alkanes, in "Enzyme Technology," p.37, R.M. Lafferty,
ed., Springer Verlag, Berlin.
Aurich, H., Bergmann, R., Lasch, l., Koelsch, R. and Sorger, H., 1985,
Wechselwirkung der Aldehyddehydrogenase aus Acinetobacter
calcoaceticus mit Membranlipiden. (11),1. Basic Microbiol.,
25:631.
Aurich, H., Sorger, H., Bergmann, R., and Lasch, l., 1987, Zur Kinetik der
membrangebundenen Aldehyd-Dehydrogenase aus
Acinetobacter calcoaceticus, Bioi. Chem. Hoppe-Seyler,
368:10l.
Baumann, P., Doudoroff, M., and Stanier, R.Y. 1968, A study of the
Moraxella group. II. Oxidase-negative species (genus
Acinetobacter), 1. Bacteriol., 95:1520.
Borneleit, P., Hermsdorf, T., Claus, R., Walther, P., and Kleber, H.-P., 1988,
Effect of hexadecane-induced vesiculation on the outer membrane
of A c i net 0 b act e rca I c 0 ace tic us, 1. G en. M i c rob i 0 I. ,
134:1983.
Aci n e to b ac ter j 0 hnsonii sp. nov., Acine to b ac ter ju nii sp.
nov. and emended description of Acinetobacter calcoaceticus
andAcinetobacter lwoffii, Int. 1. Syst. Bacteriol., 3:228.
Breuil, c., Shindler, B.D., Sijher, l.S., and Kushner, D.l., 1978, Stimulation
of lipase production during bacterial growth on alkanes, 1.
Bryn, K., Jantzen, E., and Bovre, K., 1977, Occurrence and patterns of waxes
in Neisseriaceae, 1. Gen. Microbiol., 102:33.
Buhler, M., and Schindler, l., 1984, Aliphatic hydrocarbons, in
"Biotechnology, vo1.6a," p.331, H.-J. Rehm, and G. Reed, eds.,
Verlag Chemie, Weinheim.
Cardini, G., and Jurtschuk P., 1970, The enzymatic hydroxylation of n-
octane by Corynebacterium sp. strain 7EIC, 1. Bioi. Chem.,
245:2789.
344
Claus, R., Asperger. 0., and Kleber, H.-P., 1978, Nachweis und partielle
Anreicherung von Rubredoxin-Reductase aus Acinetobacter
calcoaceticus, Wiss. Zeits. Karl-Marx-Univ. Leipzig,
Math. Naturwiss. Reihe, 27:17.
Claus, R., Asperger, 0., and Kleber, H.-P., 1979, Eigenschaften der
Rubredoxin-Reductase aus dem alkanassimilierenden
Bakterienstamm Acinetobacter calcoaceticus, Zeits. Allg.
Mikrobiol., 19:695.
Claus, R., Hadge, D., Asperger, 0., Fiebig, H., and Kleber, H.-P., 1980a,
Quantitative, immunologische Bestimmungsmethode fur
Rubredoxin in Bakterienrohextrakten von Acinetobacter
Claus, R, Asperger, 0., and Kleber H.-P., 1980b, Influence of growth phase
and carbon source on the content of rubredoxin in Acinetobacter
calcoaceticus, Arch. Microbiol., 128:263.
Claus, R., Kappeli, 0., and Fiechter, A., 1984, Possible role of extracellular
membrane particles in hydrocarbon utilization by Acinetobacter
calcoaceticus 69-V,.r. Gen. Microbiol., 130:1035.
Claus, R., Fischer, B.E., and Kleber, H.-P., 1985, An esterase as marker
enzyme of the outer membrane of Acinetobacter
calcoaceticus, J. Basic Microbiol., 25:299.
Coon, M.J., Autor, A.P., Boyer, R.F., Lode, E.T., and Strobel, H.W., 1973,
On the mechanism of fatty acid, hydrocarbon, and drug
hydroxylation in liver microsomal and bacterial enzymes systems,
in "Oxidase and Related Redox Systems," p.529, T.E. King, H.S.
Mason, and M. Morrison, eds., University Park Press, Baltimore.
Dalton, H., 1980, Oxidation of hydrocarbons by methane monooxygenase
from a variety of microbes, Adv. Appl. Micro bioi., 262:71.
Davis, J. B., 1967, "Petroleum Microbiology," Elsevier Publishing Company,
Amsterdam.
Einsele, A., 1983, Biomass from higher n-alkanes, in "Biotechnology, vo!.3,"
p.44, H.J. Rehm, and G. Reed, eds., Verlag Chemie, Weinheim.
Ensley, B.D., and Finnerty, W.R., 1980, Influence of growth substrates and
oxygen on the electron transport system in Acinetobacter
species H01-N, J. Bacteriol., 142:859.
Eremina, S.S., Asperger, 0., and Kleber, H.-P., 1987, Cytochrome P-450 and
the respiration activity of Acinetobacter calcoaceticus growing
on n-nonane, Mikrobiologija, 5:764.
Erickson, L.E., Nakahara, T., and Prokop, A., 1973, Growth in cultures with
two liquid phases: hydrocarbon uptake and transport, Process
Biochem., 10:9.
Fewson, C.A., 1967, The growth and metabolic versatility of the Gram-
negative bacterium NCIB 8250 (,Vibrio 01'), J. Gen.
Microbio!.,46:255.
Finnerty, W.R., 1977, The biochemistry of microbial alkane oxidation: new
insights and perspectives, Trends Biochem. Sci., 2:73.
Finnerty, W.R., 1980, Physiology and biochemistry of bacterial phospholipid
metabolism, Adv. Bacterial Physiol., 18:177.
345
Finnerty, W.R., 1984, The application of hydrocarbon utilizing
microorganisms for lipid production, in "Biotechnology for the oils
and fats industry," AGCS Monograph, 11:199.
Finnerty, W.R., and Kallio, R.E., 1964, Origin of palmitic acid carbon in
palmitate formed from hexadecane-1- 14 C and tetradecane-1- 14 C by
Micrococcus cerificans,.T. Bacteriol., 87:1261.
Finnerty, W.R., and Singer, M.E., 1985, Membranes of hydrocarbon-
utilizing microorganisms, in "Organization of Procaryotic Cell
Membranes," p.l, B.K. Ghosh, ed., CRC-Press, Boca Raton.
Finnerty, W.R., Hawtrey, E., and Kallio, R.E., 1962, Alkane-oxidizing
Micrococci, Zeits. Allg. Mikrobiol., 2:169.
Fischer, B.E., 1986, Localization of hydrolytic enzymes in Acinetobacter
calcoaceticus,.T. Basic Microbiol., 26:9.
Fischer, B., Claus, R., and Kleber, H.-P., 1984, Isolation and
characterization of the outer membrane of Acinetobacter
calcoaceticus, .T. Biotech., 1:111.
Fixter, L.M., and Fewson, C.A., 1974, The accumulation of waxes by
Acinetobacter calcoaceticus NCIB 8250, Biochem. Soc.
Trans., 2:944.
Fixter, L.M., Nagi, M.N., McCormach, J.G., and Fewson, C.A., 1986,
Structure, distribution and function of wax esters in
Acinetobacter calcoaceticus,.T. Gen. Microbiol., 132:3147.
Gallagher, LH.C. 1971, Occurrence of waxes in Acinetobacter, .T. Gen.
Microbiol.,68:245.
Goma, G., Pareilleux, A., and Durand, G., 1973, Cinetique de degradation
des hydrocarbures par Candida /ipolytica, Arch. Microbiol.,
88:97.
Grant, 0.1. W., 1973, The degradative versatility, aryl-esterase activity and
hydroxylation reactions of Acinetobacter lwoffii NCIB 10553,
J. Appl. Microbiol., 36:47.
Grund, A., Shapiro, 1., Fennewald, M., Bacha, P., Leahy, J., Markbreiter, K.,
Nieder, M., and Toepfer, M., 1975, Regulation of alkane oxidation
in Pseudomonas putida, J. Bacteriol., 123:546.
Haferburg, D., and Kleber, H.-P., 1982, Extrazellulare Lipase aus
Acinetobacter calcoaceticus, Acta Biotech., 2:337.
Haferburg, D., and Kleber, H.-P., 1983, Regulation der extrazelluHiren
Lipase aus einem alkanverwertenden Stamm von Acinetobacter,
Acta Biotech., 3:185.
Haferburg, D., Asperger, 0., Lohs, V., and Kleber, H.-P., 1983, Regulation
der Alkanverwertung bei Acinetobacter calcoaceticus, Acta
Biotech., 3:371.
Haferburg, D., Hommel, R., Claus, R., and Kleber, H.-P., 1986, Extracellular
microbial lipids as biosurfactants, Adv. Biochem. Eng.
Biotech., 33:53.
Hankin, c., and Kollatukudy, P.E., 1968, Metabolism of a plant wax paraffin
(n-nonacosane) by a soil bacterium (Micrococcus cerificans),
Hiratsuka, 1., Furukawa, T., and Noda, M., 1980, Microbial production of
pristanol and pristanediol, Japan K okai K oho, 8000:065.
346
Horisberger, M., 1977, Structure of the peptidoglycans of M oraxella
glucidolytica and M oraxella twoffi grown on hydrocarbons,
Arch. Mikrobiol., 112:297.
Jirausch, M., Asperger, 0., and Kleber, H.-P., 1986, Alcohol oxidation by
Acinetobacter calcoaceticus EB 104 - a n-alkane-utilizing and
cytochrome P-450-producing strain, 1. Basic Microbiol., 26:351.
Microbio!.,32:349.
Kaplan, N., and Rosenberg, E., 1982, Exopolysaccharide distribution of and
bioemulsifier production by Acinetobacter calcoaceticus BD4
and BD413, Appl. Environ. Microbiol., 44:1335.
Kappeli, 0., and Finnerty, W.R., 1979, Partition of alkane by an
extracellular vesicle derived from hexadecane-grown
Acinetobacter, J. Bacterio!., 140:707.
Kappeli, 0., and Finnerty, W.R., 1980, Characteristics of hexadecane
partition by the growth medium of Acinetobacter sp., Biotech.
Bioeng., 22:495.
Kennedy, R.S., and Finnerty, W.R., 1975, Microbial assimilation of
hydrocarbons. II. Intracytoplasmic membrane induction in
Acinetobacter sp., Arch. Microbio!., 102:IB.
Kennedy, R.S., Finnerty, W.R., Sudarsanan, K., and Young, R.A., 1975,
Microbial assimilation of hydrocarbons. I. The fine structure of a
hydrocarbon oxidizing Acinetobacter sp., Arch. Microbiol.,
102:75.
Kleber, H.-P., 1973, Repression des Malatenzyms durch n-Alkane in
Acinetobacter calcoaceticus, Zeits. Allg. Mikrobiol.,
13:467.
Kleber, H.-P., 1975, Hemmung des Malatenzyms aus Acinetobacter
calcoaceticus durch Acetyl-CoA, Zeits. Allg. Mikrobiol.,
15:19.
Kleber, H.-P., 1978, Regulation der Synthese und Aktivitat von Enzymen des
Citrat- und Glyoxylatzyklus in Acinetobacter calcoaceticus
unter dem Einfluss der n-Alkanassimilation, Wiss. Zeits. Karl-
Marx-Univ. Leipzig, Math. Naturwiss. Reihe, 27:55.
Kleber, H.-P., and Aurich, H., 1973, Einfluss von n-Alkanen auf die Synthese
der Enzyme des Glyoxylatzyklus in Acinetobacter
Kleber, H.-P., and Aurich, H., 1974, Verhalten der Enzyme des Citratzyklus
wah rend der n-Alkan-Assimilation bei Acinetobacter
Kleber, H.-P., Schopp, W., and Aurich, H., 1973, Verwertung von n-Alkanen
durch einen Stamm von Acinetobacter calcoaceticus, Zeits.
Allg. Mikrobio!., 13:445.
Kleber, H.-P., Claus, R., and Asperger, 0., 1983, Enzymologie der n-
Alkanoxidation bei Acinetobacter, Acta Biotech., 3:251.
Kleber, H.-P., Muller, R., and Asperger, 0., 1985, Cytochrome P-450 in
Acinetobacter: Occurrence, isolation and regulation, in
"Environmental Regulation of Microbial Metabolism," p.89, I.S.
Kulaev, E.A. Dawes, and D.W. Tempest, eds., Academic Press,
London.
347
Klossek, P., Kirchner, M., and Kurth, J., 1985, Immobilisierung partikel-
gebundener Aldehyddehydrogenase aus Acinetobacter
calcoaceticus EB 104,1. Basic Microbiol., 25:429.
Klug, M.J., and Markovetz, A.J., 1971, Utilization of aliphatic
hydrocarbons by microorganisms, Adv. Microbial Physiol., 5:1.
Kollatukudy, P.E., and Hankin, e., 1968, Production of -haloesters from
alkylhalides by Micrococcus cerificans, 1. Bacteriol., 54:145 ..
McCaman, R.E., and Finnerty, W.R., 1968, Biosynthesis of cytidine
diphosphate-diglyceride by a particulate fraction from
Micrococcus cerificans, .I. Bioi. Chem., 243:5074.
McKenna, E.l., and Kallio, R.E., 1965, The biology of hydrocarbons, Ann.
Makula, R., and Finnerty, W.R., 1968, Microbial assimilation of
hydrocarbons. 1. Fatty acid derived from normal alkanes, 1.
Makula, R.A., and Finnerty, W.R., 1970, Microbial assimilation of
hydrocarbons: Identification of phospholipids, 1. Bacteriol.,
103:348.
Makula, R.A., and Finnerty, W.R., 1971, Microbial assimilation of
hydrocarbons: Phospholipid metabolism, 1. Bacteriol., 107:806.
Makula, R.A., Lockwood, P.J., and Finnerty, W.R., 1975, Comparative
analysis of the lipids of Acinetobacter species grown on
hexadecane,.I. Bacterial., 121:250.
Modrzakowski, M.e., and Finnerty, W.R., 1980, Metabolism of symmetrical
dialkyl ethers by Acinetobacter species HOI-N, Arch.
Microbiol.,126:285.
Modrzakowski, M.e., Makula, R.A., and Finnerty, W.R., 1977, Metabolism
of the alkane analogue n-dioctyl ether by Acinetobacter
species, 1. Bacterial., 131:92.
Muller, H., and Voigt, B., 1982, Untersuchungen zur chemischen
Zusammensetzung der Lipidfraktion von Acinetobacter
calcoaceticus, Acta Biotech., 2:155.
Muller, H., Naumann, A., Claus, R., and Kleber, H.-P., 1983, Intracyto-
plasmic membrane induction by hexadecane in Acinetobacter
calcoaceticus, ZeUs. Allg. Mikrobiol., 23:645.
Muller, R.,.Asperger, 0., and Kleber, H.-P., 1989, Purification of
cytochrome P-450 from n-hexadecane-grown Acinetobacter
calcoaceticus, Biomed. Biochim. Acta, 48:243.
Neidleman, S.L., and Geigert, J., 1983, Long-chain wax esters, US
Patent 4, 404, 283.
Neufeld, R.J., and Zajic, J.E., 1984, The surface activity of Acinetobacter
cal co acetic us sp. 2CA2, Biotech.Bioeng., 26:1108.
Ratledge, c., 1978, Degradation of aliphatic hydrocarbons, in
"Developments in Biodegradation of Hydrocarbons," p.1, R.J.
Watkinson, ed., Applied Science Publishers, London.
Rosenberg, E., 1986, Microbial surfactants, CRC Crit. Rev.
M icrob i 01., 3: 109.
Rosenberg, E., and Kaplan, N., 1987, Surface-active properties of Acineto-
bacter exopolysaccharides, in "Bacterial Outer Membranes as
Model Systems," p.311, M.lnouye, ed., Wiley, New York.
348
Rosenberg, E., Zuckerberg, A., Rubinovitz, c., and Gutnick, D.L., ] 979,
Emulsifier of Arthrobacter RAG-I: isolation and emulsifying
properties, Appl. Environ. Microbiol., 37:409.
Rosenberg, M., Bayer, E.A., DeLarea, J., and Rosenberg, E., 1982, Role of
calcoaceticus RAG-Ion hexadecane, Appl. Environ.
Microbiol., 44:929.
Rusansky, S., Avigad, R., Micheali, S., and Gutnick, D.L., 1987, Involvement
Acinetobacter calcoaceticus RA57, Appl. Environ.
Sampson, K.L., and Finnerty, W.R., 1974, Regulation of fatty acid
biosynthesis in the hydrocarbon-oxidizing microorganism,
Acinetobacter sp., Arch. Microbiol., 99:203.
Scott, c.c.L., and Finnerty, W.R., 1976a, A comparative analysis of the ultra-
structure of hydrocarbon-oxidizing microorganisms, J. Gen.
Microbiol., 94:342.
Scott, c.c.L., and Finnerty, W.R., 1976b, Characterization of intracyto-
plasmic hydrocarbon inclusions from the hydrocarbon-oxidizing
Acinetobacter species HOI-N, J. Bacteriol., 127:481.
Scott, c.c.L., Makula, R.A., and Finnerty, W.R., 1976, Isolation and charac-
terization of membranes from a hydrocarbon-oxidizing
Acinetobacter sp., J. Bacteriol., 127:469.
Shabtai, Y., and Gutnick, D.L., 1985, Exocellular esterase and emulsan
release from the cell surface of Acinetobacter calcoaceticus,
Singer, M.E., and Finnerty, W.R., 1984a, Microbial metabolism of straight-
chain and branched alkanes, in "Petroleum microbiology," p.l,
R.M. Atlas, ed., Macmillan, New York.
Singer, J.T., and Finnerty, W.R., 1984b, The genetics of hydrocarbon-
utilizing microorganisms, in: "Petroleum Microbiology,"
p.299,R.M. Atlas, ed., Macmillan, New York.
Singer, J.T., and Finnerty, W.R., 1984c, Insertional specificity of transposon
Tn5 in Acinetob acter sp., J. Bacteri 01., 157:607.
Singer, M.E., and Finnerty, W.R., 1985a, Fatty aldehyde dehydrogenases in
Acinetobacter species strain H01-N: Role in hexadecane and
hexadecanol metabolism, J. Bacteriol., 164:1011.
Singer, M.E., and Finnerty, W.R., 1985b, Alcohol dehydrogenases in
Acinetabacter species strain H01-N: Role in hexadecane and
hexadecanol metabolism, J. Bacterial., 164:1017.
Acinetabacter species H01-N on n-hexadecanol: Physiological
and ultrastructural characteristics, 1. Bacteriol., 162:162.
Sorger, H., and Aurich, R., 1978, Mikrobielle Aldehyddehydrogenasen und
ihre Bedeutung fur die Assimilation aliphatischer
Kohlenwasserstoffe. Wiss. Zeits. Karl-Marx-Univ. Leipzig,
Math. N aturwiss. Reihe, 27:35.
Sorger, H., Aurich, R., Fricke, B., and Vorisek, J., 1986, Ultracytochemical
localization of aldehyde dehydrogenase in Ac i net 0 b a c ter
349
Stevenson, D.P., Finnerty, W.R., and Kallio, R.E., 1962, Esters produced
from n-heptadecane by Micrococcus cerificans, Biochem.
Biophys. Res. Comm., 9:426.
Stewart, J.E., and Kallio, R.E., 1959, Bacterial hydrocarbon oxidation. II.
Esterformationfrom alkanes, J. Bacteriol., 78:726.
Stewart, J.E., Kallio, R.E., Stevenson, D.P., Jones, A.c., and Schissler, D.O.,
1959, Bacterial hydrocarbon oxidation. I. Oxidation of n-
hexadecane by a gram-negative coccus, J. Bacteriol., 78:441.
Stewart, J.E., Finnerty, W.R., Kallio, R.E., and Stevenson, D.P., 1960, Esters
from bacterial oxidation of olefins, Science, 132:1254.
Tauchert, H., Roy, M., Schopp, W., and Aurich, H., 1975,
Pyridinnudeotidunabhangige Oxidation von langkettigen
aliphatischen Alkoholen durch ein Enzym aus Acinetobacter
calcoaceticus, Zeits. Allg. M ikrobiol., 15:457.
Tauchert, H., Grunow, M., Harnisch, R., and Aurich, R., 1976, Reinigung
und elTIlge Eigenschaften der NADP + -abhangigen
Alkoholdehydrogenase aus Acinetobacter calcoaceticus, Acta
Bioi. Med. German., 35:1267.
Torregrossa, R.E., Makula, R.A., and Finnerty, W.R., 1977a,
Characterization of lysocaroliolipin from Acinetobacter species
H01-N, J. Bacteriol., 131:486.
Torregrossa, R.E., Makula, R.E., and Finnerty, W.R., 1977b, Outer
membrane phospholipase A from Acinetobacter species R01-
N, I. Bacteriol., 131:493.
Torregrossa, R.E., Makula, R.E., and Finnerty, W.R., 1978, Hydrolysis of
phosphatidylglycerol by outer membrane phospholipase A from
Acinetobacter species HOl-N, J. Bacteriol., 136:803.
Vachon, V., McGarrity, J.T., Breuil, c., Armstrong, J.B., and Kushner, D.J.,
1982, Cellular and extracellular lipids of Acinetobacter lwoffi
during growth on hexadecane, Can. J. Microbiol., 28:660.
van der Linden, A.C., and Thijsse, G.J .E., 1965, The mechanism of
microbial oxidation of hydrocarbons, Adv. Enzymol., 27:469.
Wyndham, R.C., 1987, A screening method for cytochrome P-450 organic
peroxidase activity and application to hydrocarbon-degrading
bacterial popUlations, Can. J. M icrob iol., 33: 1.
Zuckerberg, A., Diver, A., Peeri, Z., Gutnick, D.L., and Rosenberg, E., 1979,
Emulsifier of Arthrobacter RAG-I: chemical and physical
properties, Appl. Environ. M icrob iol., 37:414.
350
METABOLISM OF AROMATIC COMPOUNDS BY ACINETOBACTER
C.A. Fewson
;Department of Biochemistry
iGlasgow G12 800
UK
INTRODUCTION
One of the most notable features of strains of Acinetobacter is their

ability to grow on a wide range of aromatic compounds. In addition,
acinetobacters isolated from natural environments can grow in simple
defined mineral media with no requirement for growth factors, so they are
clearly synthesising all their own aromatic amino acids and related
compounds. This review contains a description of the mechanisms used for
the dissimilation of aromatic compounds by Acinetobacter strains; it
concentrates on the best-characterised of these pathways, the mandelate
pathway, but includes at least outline descriptions of the ways in which many
other aromatic compounds are catabolised. In addition, there is a brief
summary of the pathways for the biosynthesis of aromatic amino acids and an
indication of the potential competition between these pathways and those for
the degradation of aromatic compounds.
UTILISATION OF AROMA TIC COMPOUNDS TO SUPPORT GROWTH
It is easy to isolate acinetobacters which can use aromatic or

hydroaromatic compounds as sole sources of carbon and energy for growth,
or which can partially metabolise them. Such strains are found in many
natural environments and in industrial effluents and treatment plants. One
of the first strains to be isolated was the so-called 'Vibrio strain 01', which
was obtained by Happold and Key (1932) in the course of their work on the
destruction of phenols when spent gasworks' liquor was mixed with sewage
and then passed through filter beds. This organism, which is probably A.
calcoaceticus NCIB8250 (Fewson, 1967b; Baumann et aI., 1968), can
351
degrade a great range of aromatic compounds (Fewson, 1967a) and was used
in some of the earliest studies on the elucidation of the metabolic pathways
by which dihydric phenols are subjected to intra-dial ring-cleavage and then
converted into B-ketoadipate (Evans and Happold, 1939; Evans, 1947; Kilby
1948, 1951; reviewed by Stanier and Oms ton, 1973).
In their monumental study of the nutritional and biochemical

properties of Acinetobacter, Baumann et al. (19611) found that many of
their strains could grow on aromatic or hydro aromatic compounds. Of 106
strains tested, 116 grew on benzoate. 80 on quinate, 79 on L-tyrosine, 76 on 4-
hydroxybenzoate, 66 on phenylacetate, 63 on kynurenate, 60 on anthranilate,
58 on L-phenylalanine, 53 on L-tryptophan, 44 on L-kynurenine, 28 on
phenol, 17 on 2-hydroxybenzoate (salicylate), 14 on 3-hydroxybenzoate, five
on phenylglyoxylate (benzoylformate), four on L-mandelate and three on D-
mandelate. Even these proportions may be underestimates because of
possible false-negative results caused by toxicity of compounds such as
phenol and salicylate. Many of these strains were obtained from clinical
sources or from enrichments on non-aromatic substrates, so clearly the
genetic potential to metabolise aromatic compounds is very widespread
within the genus. Individual strains can sometimes use more than 30
different aromatic compounds as sale sources of carbon (e.g. Fewson, 1967a;
Hasegawa and Suzuki, 1983). All these substrates are used aerobically, but
there is a report in the literature of a 'Moraxella species' metabolising
benzoate anaerobically, supported by nitrate respiration (Williams and
Evans,1975). Although there seems no intrinsic reason why acinetobacters
should not carry out this type of reaction, this particular organism seems to
. have been wrongly identified and was actually a Paracoccus sp. (Evans,
1988).
EFFICIENCY OF UTILISATION OF AROMATIC COMPOUNDS
Many observations on the utilisation of aromatic compounds by

Acinetobacter are rather qualitative, but a few attempts have been made to
quantify the efficiency of metabolism and growth. These experiments have
been done by microcalorimetry (e.g. Lovrien et aI., 1980) and by
determination of molar growth yields (e.g. Fewson, 1985). Growth yields on
both aromatic and non-aromatic substrates are low. There are various
possible explanations for this. The most likely is that the effective P /0 ratio
is only about one under most conditions, even though there may be two sites
of oxidative phosphorylation. The values for oxygen uptake used in these
calculations were those for respiratory oxygen and excluded the large
amounts needed for oxygenative reactions (Fewson, 1985). It is perhaps
disappointing that an organism which is so metabolically versatile is so
apparently inefficient in this respect. However, this may simply be further
evidence that the principal selective pressures operating on bacteria have
not been those which have tended to maximise yield (Tempest, 1978).
Acinetobacters may be better advertisements for the success of metabolic
versatility and rapid growth rate than for high molar growth yields.
352
GENERAL PATHWAYS FOR THE CATABOLISM OF AROMATIC
COMPOUNDS
The soluble and insoluble aromatic compounds which can be

metabolised by Acinetobacter and other microorganisms may consist of
one or more rings that often carry various types of substituent groups. The
complex and apparently bewildering variety of catabolic pathways can best
be thought of as parts of a systematic attack (Fig. 1): (a) chemotaxis or
directed growth towards the substrates; (b) adsorption to the surface of
insoluble substrates; (c) excretion of extracellular enzymes or emulsifying
agents in order to obtain diffusible products of small molecular mass; (d)
entry into the cell; (e) metabolism by peripheral pathways to manipulate the
substituent groups and form substrates for ring-cleavage; (f) ring-cleavage;
(g) conversion of the products of ring-cleavage into amphibolic
intermediates; (h) utilisation of the amphibolic intermediates for growth
and energy production.
So far as the initial stages are concerned, Acine to b ac ter is not motile
and has to make do with substrates that are present in its immediate
environment. The possibility of it using emulsifying agents to solubilise
potential substrates in natural environments is only hinted at, despite the
~~{T"'" ?"'>~y
peripheral pathways
\ 1
SUBSTRATES FOR RING CLEAVAGE
\ /
p-ketoadipate and mela pathways
\1
n
AMPHIBOLIC INTERMEDIATES
amphibolic, biosynthetic and energy·yielding pathways
V
ENERGY PRODUCTION AND GROWTH
Fig. 1. Microbial utilisation of aromatic compounds
353
OOH
~
O(H
~ OH HOOC I OH
a7
catechol
protocateclllJ Ic
aCid
\2 b d
C 01
,\2
cC a
CHO CHO
(COOH
QOH COOH
~ COOH ~OH HOO ~ COOH HOOC ~
.,1'0
COOH
OH HOOC OH I
.oo~ fH 2 COOH
Q.
I
((
HOO~OCH3
~I ::-.,.1
::-.,. OH
OH OH OH ~I ~ OH
H
gentisic homogentisic
et2
3-0'methYlgalllc
fr
aCid acid OH
acid
9r if
3-(2. 3-dlhydroxyphenyl-) homoprotocatechu I c
propionic acid aCid
hr
02 02
y
HOOC 02 O2
I
HOOC~OH
... I H2~H HOOC O
COOH
::-.,. ~I ~ c&<ir3
~OOH I
~~
OH 1H2
OH OH
,H2
~ COOH
C=O
(J: OH
Fig. 2. Intra-dial (ortho) (a and c) and extra-dial (meta) (b. d-i)

oxygenative ring-cleavage reactions found in various strains of
Acinetobacter. Based on Baumann et al. (1968). Stanier and Ornston
(1973), Dagley (1978), Keil et al. (1983), Sze and Dagley (1987).
great deal of work that has been done on the emulsifiers which it produces
(Gutnick et aI., this volume). Entry into the cell is often assumed to be by
free diffusion, but there is evidence that A. calcaaceticus possesses
specific transport mechanisms for even such simple aromatic compounds as
benzoate and mandelate (Cook and Fewson, 1972; Fewson, 1988),
The great majority of work that has been done on the utilisation of
aromatic compounds by Acinetabacter and other microorganisms has been
concerned with elucidating the metabolic pathways which are involved in the
breakdown of these compounds and how these pathways are regulated. All
the dozens, probably hundreds, of aromatic compounds that can be degraded
by microorganisms are converted by converging, peripheral metabolic
pathways into a small handful of dihydric phenols. These key intermediates
are then subjected to intra-diol (artha) or extra-diol (meta) ring-cleavage
(e.g. Dagley, 1978; Fewson, 1981). In Acinetabacter spp., just seven such
compounds have been identified (Fig. 2). Of these, catechol and
protocatechuate serve as the points of convergence of a particularly wide
range of peripheral degradative pathways (Figs. 3 and 4). In the great
354
ethyl methyl
m,"~ 7''''''
L-tryptophan
D-ma~ /ndelate 2-hYdrOXYfandelate
~
L-lormylkynurenine phenylglyoxylate 2-hydroxyphenylglyoxylate 2-hydroxybenzyl alcohol
benzyl _____ (saligenin)
alcohol '-..... ~ I ____
~ benzaldehyde 2-hydroxybenzaldehyde
"""""''"
acetylsalicylate
L-kynurenlne
~\ . 1 I
4-hydroxyqurnazolrne
. . \ 2-h ydroxy benzoate ________
------.. aniline benzoate phenol (salicylate)
anthranilate ~
guaiacol
glycerol ether
j~
--... gUaiaCOI~ benzene
~catechOI~
Fig. 3.
~
Conversion of aromatic compounds into catechol by various strains of
Acinetobacter. Based on Kennedy and Fewson (1968), Grant (1971), Stanier and
Ornston (1973), Grant et al. (1975), Wyndham (1986), Winstanley et al. (1987),
Vasudevan and Mahadevan (1989).
w
(11
(11
3-methoxy-4-hydroxymandelate
Ivanillylmandelate)
4·hydroxymandelate +I 3·methoxy·4·hydroxybenzyl
alcohol
~ 3.methOXY.4.hYdrOXYPhenYlgIYOXYlate/
4·hydroxyphenylglyoxylate ~ 3,4·dihyroxymandelate
.hYdrOXYbenZYI 3·methoxy-4·hydroxybenzaldehyde /
3'PhenrroPionate alCr OI Ivatllin) 3,4·dihydroxyphenylglyoxylate
\
. j
cinnamate .. .. / /qUinate
j
4-hydroxybenzaldehyde 3-methoxy-4-hydroxybenzoate
\
(vanillate) 5-dehydroquinate
4'hYdrOXy~ 3,4·dihydroxybenzaldehyde ~ shlklmate
4·hydroxybenzoale / 5·dehydroshikimale ~
~protocatechuale ~
Fig. 4. Conversion of aromatic compounds into protocatechuate by various

strains of Acinetobacter. Based on Kennedy and Fewson (1968). Stanier
and Ornston (1973). Keil et al., (1983).
majority of cases, art h a fission of catechol by catechol l,2-dioxygenase (Fig.

2a) and of protocatechuate (3,4-dihydroxybenzoate) by protocatechuate 3,4-
dioxygenase (Fig. 2c) have been demonstrated or assumed to take place
(Baumann et aI., 1968; Stanier and arnston, 1973). However, meta-
cleavage of protocatechuate by protocatechuate 4,5-dioxygenase (Fig. 2d)
occurs in a strain of A. twaffii after growth on 4-hydroxymandelate (Sze
and Dagley, 1987), while strain 20 of Baumann et al. (1968) was reported to
catalyse both artha and meta fission of catechol after growth on phenol.
The B-ketoadipate pathway (Fig. 1 in arnston and Neidle, this volume)

converts products of artha-cleavage into succinate and acetyl-eoA, and the
various meta-cleavage pathways produce a range of compounds such as
pyruvate, acetaldehyde and formate (Stanier and amston, 1973; Dagley,
1978). The small molecular mass compounds then enter the amphibolic and
anabolic pathways and support growth (Fig. 1). Rather little energy and few,
if any, biosynthetically useful intermediates are produced in the early stages
of these aerobic pathways for the degradation of aromatic compounds.
Indeed, in many cases it is only the very last products of degradation that can
be used for either energy production or synthesis. Some of the early
catabolic steps may actually be energy-dependent or lower the potential
amount of energy that is available for growth; e.g. monooxygenases divert
reducing power from the electron transfer system, and therefore from
oxidative phosphorylation, into the non-energy producing reduction of one of
the oxygen atoms to water. In a few cases, however, there is evidence that
early oxidative steps may give rise to useful energy by oxidative
356
phosphorylation; e.g. mandelate supports a slightly higher molar growth
yield of A. calcoaceticus NCIB8250 than does phenylglyoxylate,
presumably because the flavin-linked mandelate dehydrogenase leads to
additional production of ATP (Fewson, 1985).
One of the problems that dogs much of the research in this area is the
lack of commercial availability and/or instability of many of the potential
metabolic intermediates. In many cases they have to be synthesised
enzymically as required and acinetobacters can be used for this purpose (e.g.
Darrah and Cain, 1969).
INVOLVEMENT OF PLASMIDS IN THE DEGRADATION

OF AROMA TICS
The widespread presence of plasmids in strains of Acinetobacter

isolated from natural environments (Towner. this volume) has invited the
speculation that some of them might be involved in encoding catabolic
pathways, as is quite often the case in pseudomonads. However, so far as is
known, the vast majority of aromatic pathways in acinetobacters are
chromosomally-encoded. Nevertheless, there are a few well-documented
cases of aromatics being degraded by acinetobacters using plasmid-encoded
enzymes; examples include the utilisation of chloro-substituted compounds
(Furukawa and Chakrabarty, 1982; Shields et al., 1985; see below), resorcin
(Barkovsky and Shub, 1986) and benzene (Winstanley et aI., 1987).
METABOLISM AFTER RING-CLEA VAGE
The present chapter concentrates on the peripheral pathways used by

acinetobacters for the conversion of aromatic compounds into one or other
of the substrates for ring-cleavage. The pathways found in Acinetobacter
spp., Pseudomonas spp. and other organisms for the ortha-cleavage of
catechol and protocatechuate, and the subsequent formation of B-
ketoadipate, have been extensively reviewed (e.g. Kilby, 1951; Cain, 1961;
Omston and Stanier, 1964; Canovas et aI., 1967; Stanier and Ornston, 1973;
Ornston and Parke, 1977; Kozarich, 1988; Ornston and Neidle, this
volume). The regulation of expression of the enzymes of the B-ketoadipate
pathway has been thoroughly studied, especially in strain 73 of A.
calcoaceticus CM oraxella calcoacetica') (Canovas and Stanier, 1967;
Canovas et ai., 1967, 1968; Canovas and Johnson, 1968; Neidle and Omston,
1987; Doten et ai., 1987b). Many of the enzymes have been purified, mostly
from the transformable strain BD413 (ATCC33305 = ADPI; Juni and Janik,
1969), but in some cases from strains such as 80-1. The enzymes have been
characterised and compared with respect to mechanisms of action, N-
terminal amino acid sequences, amino acid compositions and immunological
cross-reactivity (Katagiri and Wheelis, 1971; Patel et ai., 1974, 1975, 1976;
Zaborsky and Schwartz, 1974; Hou et al., 1976, 1977, 1978; Patel and
Omston,1976; Yeh et aI., 1978, 1980a,b, 1981, 1982; Durham et aI., 1980;
McCorkle et a!., 1980; Yeh and Ornston, 1980, 1981, 1984; Chari et aI.,
357
1987a,b). Many of the relevant genes have been cloned and sequenced
(Neidle and Ornston, 1986; Shanley et al., 1986, 1989; Doten et al., 1987a;
Neidle et al., 1987, 1988; Hughes et al., 1988; Ornston and Neidle, this
volume). The combined results from this sustained attack have led to some
of the most original and intriguing speculations concerning the evolution of
any metabolic pathway (e.g. Canovas et al., 1967; Stanier and Ornston, 1973;
Ornston and Parke, 1977; Ornston and Yeh, 1981). Current thinking is
summarised by Ornston and N eidle (this volume).
Meta-cleavage pathways in Acinetobacter are much less well

understood. A few of the enzymes have been purified (Seltzer, 1973; Leung
et al., 1974; Burlingame and Chapman, 1983) and the regulation of the 4-
hydroxyphenylacetate meta-cleavage pathway in Acinetobacter strain 3B-1
has been reviewed by Bayly and Barbour (1984).
THE MANDELATE AND BENZYL ALCOHOL PATHWAYS
Of all the aromatic compounds that are degraded by Acinetobacterspp.,

the converging pathway for the catabolism of mandelate and benzyl alcohol is by
far the best characterised (Fig. 5). The information that has been obtained
parallels that availableforvarious Pseudomonas spp., which in turn builds on
the pioneering work of R.Y. Stanier and his colleagues (references in Fewson,
1988). The term 'mandelate pathway' generally refers to the pathway down to
the point of ring cleavage.
Growth on Mandelate and Benzyl Alcohol
Of 141 strains of A. calcoaceticus that have been tested, 11 can grow

on both L- and D-mandelate. One strain (NCIB8250) can grow on only L-
mandelate and a further one (EBF65/65) on only D-mandelate (Fewson,
1988). So far as is known, none of these strains was isolated by procedures
involving enrichment on mandelate media; indeed, the majority of the
acinetobacters which can grow on mandelate were originally isolated from
frozen chicken carcasses by E.M. Barnes (Fewson et al., 1988). There seems
to have been no systematic survey of the ability of acinetobacters to grow on
benzyl alcohol, but it appears to be quite a common feature.
The Mandelate Pathway and its Enzymes
It is not known how mandelate and benzyl alcohol enter the cell. In
very preliminary experiments it appeared that there was a barrier against the
entry of mandelate into A. calcoaceticus NCIB8250 unless the bacteria
had been grown under conditions where the mandelate-utilising enzymes
were induced (Cook and Fewson, 1972). The implication was that there is a
specific and inducible uptake system for mandelate. If so, it may be required
only at low extracellular concentrations of mandelate when the lipid-
solubility of the substrate will not be sufficient to maintain adequate
intracellular concentrations. Transport systems are usually stereospecific,
358
eOOH
"'(5"
I
D-mandel ic acid L-mandellc acid
phenylglvoxvliC acid
'~"7
Y-
2H
6 00
benzyl alcohol
6
CHO
benza I dehyde
r
e H20""j
2H~~ ~""""2H
H20 f
benzoic acid
Fig. 5. The mandelate pathway in Acinetobacter. Based on Kennedy and

Fewson (1968a,b), Fewson (1988).
so presumably there must be a separate system for each of the mandelate

enantiomers.
The pathway for the conversion of mandelate and benzyl alcohol into
benzoate by A. calcoaceticus (Fig. 5) is the same as that found in some
pseudomonads, and was elucidated by: (a) experiments on simultaneous
adaption; (b) measurement of the relevant enzyme activities; and (c) the
detection and utilisation of intermediates (Kennedy and Fewson, 1968a,b;.
359
6
COOH
benzoic acid
HOOaOH OH
I
rr
H
h benzene
3, 5-cyclohexadi ene- O2
6
OH 1,2-diol-l-carboxylic acid f 2H
b ~C02 H OH
t- 2H
~~
D
phenol
OH ~v-
OH ~{ benzene cis -glycol
1 2H
CO 2 h
~cateCrhOI
OI"
COOH
......::
~TH
OH02
e H20
c
0
2
2-hydroxybenzolc acid QOOH

(salicylic acid) 1 h COOH
cis,cis -mucon I c ac I d
succinate
+
acetyl-CoA
Fig. 6. Oxygenati ve formation and metabolism of catechol by various strains
of Acinetobacter. Based on Kennedy and Fewson (1968a,b), Reiner
(1971), Winstanley et a1. (1987), Fewson (1988).
Fewson, 1988). Benzoate is oxygenatively converted into catechol (Fig. 6;

Reiner, 1971), which is then artha-cleaved and metabolised to £-
ketoadipate (Fig. 1 in Ornston and Neidle, this volume).
Those strains of A. calcaaceticus (e.g. EBF65/174) which can grow

on both enantiomers of mandelate have two stereospecific mandelate
360
dehydrogenases (Fig. 5a,b). Strain NCIB8250 can grow on only the L-
enantiomer and has an L-mandelate dehydrogenase (Fig. 5b), but no D-
mandelate dehydrogenase, whereas strain EBF65/65 can grow on only the D-
enantiomer and has just the D-mandelate dehydrogenase (Fig. 5a). Both of
these strains give rise to mutants that can grow on the second enantiomer.
In every case this is because of the appearance of an extra dehydrogenase,
specific for the other enantiomer (Hills and Fewson, 1983a). All of the
enzymes that convert mandelate and benzyl alcohol into benzoate have now
been purified from A. calcoaceticus NCIB8250, EBF65/65 or their
mutants.
The mandelate dehydrogenases (Fig. 5a,b) are membrane-bound

flavoproteins and can be solubilised in an active form by careful treatment
with appropriate detergents (Allison et aI., 1985a). L-Mandelate
dehydrogenase contains FMN and has a sub-unit Mr of 44,000 (Hoey et aI.,
1987), whereas D-mandelate dehydrogenase contains FAD and has a sub-
unit Mrof 60,000 (Allison et aI., 1985b). A. calcoaceticus also contains a
pair of membrane-bound flavin-linked lactate dehydrogenases which have
been purified (Allison et aI., 1985b; Allison and Fewson, 1986).
Comparison of the properties of the various enzymes has led to the following
conclusions:- (a) The two enzymes specific for the L-enantiomers of
mandelate and lactate strongly resemble each other, as do the two enzymes
specific for the D-enantiomers, with respect to ease of extraction from the
membrane by detergents, sub-unit Mr values, pH optima, pI values, and
susceptibility to thiol reagents. Each of these pairs of enzymes may
therefore have had a common evolutionary origin. If so, the ability of D-
mandelate dehydrogenase to oxidise D-lactate (Allison et aI., 1985b) may be
a relic of the recruitment of a lactate dehydrogenase to oxidise mandelate, or
it may simply be a consequence of the similarity in chemical structures. (b)
The two pairs of enzymes, although similar to each other in a general way,
are sufficiently different (e.g. in flavin specificity and sub-unit M) that a
common origin for all four enzymes is either unlikely or was much more
remote. (c) The 'evolved' dehydrogenases in mutants of strains NCIB8250
and EBF65/65 that were selected on the basis of ability to grow on the
second enantiomer of mandelate are almost identical with the 'original'
dehydrogenases in the other strains with respect to immunological cross-
reactivity, solubility in detergents, net charge at pH 7.5, pI values, salting-out
properties, sub-unit Mr values, apparent Km values, heat-stability and
sensitivity to 4-chloromercuribenzoate (Fewson et aI., 1988). This suggests
that the most likely explanation for the appearance of the 'evolved'
dehydrogenases is some event(s) that allowed the expression of previously
silent or cryptic genes. This idea is strengthened by the observation that, in
each of the mutants tested, the 'evolved' dehydrogenases were found to be
regulated co-ordinately with the pre-existing dehydrogenases, whether tested
by the isolation of constitutive mutants or by measuring the effects of various
inducers, anti-inducers and compounds that cause catabolite repression
(Hills and Fewson, 1983b).
361
Phenylglyoxylate decarboxylase (Fig. 5c) is a tetrameric, TPP-
dependent enzyme with a sub-unit M, of 58,000 (Barrowman and Fewson,
1985) which is at least superficially similar to the familiar yeast pyruvate
decarboxylase (Barrowman et al., 1986).
There are two benzaldehyde dehydrogenases in A. calcoaceticus

(Livingstone et al., 1972). Benzaldehyde dehydrogenase I (Fig. 5e) is a heat-
stable, K + -activated enzyme and is associated with mandelate oxidation.
Benzaldehyde dehydrogenase II (Fig. 5f) is a heat-labile, K + -independent
enzyme and is associated with benzyl alcohol metabolism. Both enzymes
have been purified and shown to be NAD-dependent, soluble and tetrameric
(MacKintosh and Fewson, 1987, 1988a,b; Chalmers and Fewson, 1989b).
Benzyl alcohol dehydrogenase (Fig. 5d) is also a soluble, NAD-dependent
tetrameric enzyme (MacKintosh and Fewson, 1987, 1988a,b).
A study of the two benzaldehyde dehydrogenases and benzyl alcohol

dehydrogenase affords a good opportunity to test the hypothesis of
retrograde evolution of enzymes as well as ideas about gene duplication. All
three enzymes share two common substrates (NAD and benzaldehyde) and
participate in a peripheral pathway that may be more likely to undergo rapid
evolution than central pathways which are constrained by the necessity of
preserving features concerned with regulation. The evidence obtained so far
from N -terminal amino acid sequencing, amino acid compositions,
immunological cross-reactivity and kinetic studies is that benzyl alcohol
dehydrogenase is probably not homologous with the other two enzymes, and
that there is therefore no evidence for retrograde evolution in this pathway.
The two benzaldehyde dehydrogenases may, however, be homologous with
each other and, if so, could have arisen as a result of a gene doubling event
(Chalmers and Fewson, 1989a, unpublished observations).
The two enzymes that convert benzoate into catechol (Fig. 6a,b) have
not been purified from A. calcoaceticus, but they have been well-
characterised in other organisms (Reiner, 1971; references in Fewson,
1988).
Regulation of the Mandelate Pathway
The mandelate pathway enzymes are expressed co-ordinately and are

induced by phenylglyoxylate, but not by mandelate. The evidence for this
rests on a variety of approaches, including: (a) estimation of the degree of
correlation amongst differential rates of synthesis of the enzymes under
various conditions of induction and repression; (b) characterisation of
blocked mutants; (c) gratuitous induction by thiophenoxyacetate and
phenoxyacetate; (d) co-ordinate anti-induction by 2-phenylpropionate; and
(e) characterisation of constitutive mutants (Livingstone and Fewson, 1972;
Fewson et al., 1978). The enzymes converting benzoate into catechol are
regulated independently of the other mandelate enzymes, as are those of the
B-ketoadipate pathway (Stanier and Ornston, 1973).
362
The mandelate enzymes of A. caLcoaceticus are subject to some sort
of catabolite repression (e.g. by succinate) and to feedback repression (e.g.
by benzoate), but the mechanisms of this are unknown (Cook et aI., 1975;
Fewson, 1988). In addition, mandelate metabolism dominates benzyl
alcohol metabolism even though benzyl alcohol supports a faster growth rate
and gives a higher molar growth yield than does mandelate (Beggs et aI.,
1976; Beggs and Fewson, 1977; Fewson, 1985). Batch cultures of A.
calcoaceticus growing on mandelate or phenylglyoxylate show an unusual
non-exponential pattern of growth. There are transient accumulations of
benzyl alcohol and benzaldehyde in the medium which are caused by the
limitation of mandelate oxidation by low activities of benzaldehyde
dehydrogenase and by the diversion of reducing power to the formation of
benzyl alcohol (Cook et aI., 1975). The basis for the apparent inability to
induce the two enzymes specific for benzyl alcohol oxidation whenever
phenylglyoxylate decarboxylase is being induced is not clear (Beggs et aI.,
1976; Beggs and Fewson, 1977).
There is no evidence for feedback inhibition of any of the mandelate

enzymes by subsequent metabolites, and there is no substantial evidence that
they exist in any sort of ordered complex within the cell (Hoey and Fewson,
1990). The general presumption seems to be that peripheral pathways of
this sort are regulated only at the level of gene expression, and not at all by
regulation of enzyme activity by allosteric or other mechanisms (except, of
course, by the availability of substrates and cofactors). If this is correct, it
may explain why intermediates often accumulate while aromatic and similar
substrates are being metabolised, and why growth rates are less than
maximum.
Genetic studies of the mandelate pathway in A. calcoaceticus are at

an elementary stage. The mandelate genes in strain EBF65/65 appear to be
clustered near the auxotrophic marker phe-l, but are not all contiguous with
each other. A gene responsible for the expression of the cryptic L-
mandelate dehydrogenase appears to be close to a gene required for the
activity of D-mandelate dehydrogenase (Vakeria et aI., 1984).
Metabolism of Ring-Substituted M andelates and Benzyl Alcohols
A. calcoaceticus can grow on several ring-substituted mandelates

and benzyl alcohols, and can partially metabolise others. This is possible
because the enzymes involved in the conversion of mandelate into benzoate
can tolerate extensive substitution on the benzene ring, especially at the 3-,
4- and 5-positions, and the enzymes are also induced by the substituted
compounds (Kennedy and Fewson, 1968a,b; Fewson, 1988). Strain
NCIB8250 can convert L-4-hydroxymandelate and 4-hydroxybenzyl alcohol
into 4-hydroxybenzoate (Fig. 7b-d), and there is an inducible 3-hydroxylase
(Fig. 7e) to form protocatechuate prior to ortho- cleavage (Fig. 7). This
strain also forms protocatechuate from 3,4-dihydroxymandelate by means of
the mandelate enzymes, and from 4-hydroxy-3-methoxymandelate
(vanillylmandelate) and 4-hydroxy-3-methoxybenzyl alcohol (vanillyl
363
"6
COOH CooH CooH
I I
¢
I
H~H
VOCH 3
IlH OH OH
0-4-hvdroxyrnandeIIC acid l-4-hvdroxymandellc acid D-4-hydrOxY-3- L-4-hydroxy-3-
metnoxymancJellc acid methoxvmandellc acid
o
COOH COOH
¢
I I
YOCH 3
OH OH
4-hydroxyohenyig\voxvllc acid 4-hvdroxy-3-methoxvphenylg 1yoxyll c ae Id
cH c'~~
¢ OH
~
V O CH 3
OH
Lj-hydroxybenZa 1dehyde Lj -hyd roxy- 3-methoxybenZa Ide hyde
dH d'~ ,:
COOH •
¢ OH
4-hydroxvbenzolc acid 4-hydroxy-3-methoxvbenzoIC acid
'"
eU
.
" J
"'~"
~
COOH
6~OH
OH
HO~OCH3
OH
~OH
OH
3-0-methylgaIiIC acid protocatechuic acid
protocatechu Ie ae I d
,, ik
, , t
(]t
S
COOH
COOH OH
" COOH
QOOH
lA~~OOH
COOH
COOH
2-hvd rOXV-4 -C3rt)O)(V- 3-carhOxv-cis.cis-muconlc acid 3-carboxy- cis,cis -muconlc acid
muconlc semi ildf'hVd~
methanol
formate suec I nate succinate
pyruvate
acetyl-eoA acetYl-eM
oxaioacetate
Fig. 7. Utilisation of 4-hydroxymandelate and 4-hydroxy-3-methoxy-

mandelate (vanillylmandelate) by A.calcoaceticus NCIB8250 (---»
(Kennedy and Fewson, 1968a,b) and A.lwoffii (---7) (Sze and Dagley,
1987).
364
alcohol) by the additional involvement of an inducible O-demethylase (Fig.
7j). 2-Hydroxymandelate and 2-hydroxybenzyl alcohol (saligenin) support
growth by oxidation to salicylate, and then conversion to catechol, by means
of salicylate hydroxylase (Figs. 3,6e). The much more restricted availability
and substrate specificities of the ring-cleavage enzymes means that
compounds can be completely oxidised only if they can be converted into an
appropriate substrate for one of the ring-cleavage enzymes. The
consequence is that many of the substrates of the mandelate enzymes can be
only partially oxidised. For example, 3-hydroxymandelate and 3-
hydroxybenzyl alcohol are converted into 3-hydroxybenzoate, and this
accumulates because there are no hydroxylases which might convert it into
protocatechuate or gentisate (2,5-dihydroxybenzoate). Monofluoro-
mandelates are oxidised beyond the level of the fluorobenzoates because the
small fluorine substitutions can he tolerated to some extent by the later
enzymes of the pathway (Clarke et ai., 1975).
Sze and Dagley (1987) isolated a strain of Acinetobacter which could

grow on D,L-4-hydroxy-3-methoxymandelate or D,L-4-hydroxymandelate,
but not on either D- or L-mandelate. They isolated the strain from a cattle
yard that had presumably been exposed to large amounts of urinary
4-hydroxy-3-methoxymandelate derived from adrenaline. As in strain
NCIB8250, the initial enzymes of the mandelate pathway are fairly non-
specific in this strain and they convert 4-hydroxy-3-methoxymandelate and 4-
hydroxymandelate into 4-hydroxy-3-methoxybenzoate and 4-
hydroxybenzoate, respectively, and these are then hydroxylated to form 3- 0-
methylgallate and protocatechuate, which in this strain undergo meta
cleavage (Fig. 7). Mandelate can be converted into benzoate, but this strain
has no benzoate oxygenase, which explains why unsubstituted mandelate
cannot support its growth. It would clearly be worthwhile trying to isolate
other acinetobacters which can be grown on a variety of substituted
mandelates.
METABOLISM OF A WIDE RANGE OF SIMPLE

AROMATIC COMPOUNDS
This section summarises some of the work that has been done on the
degradation of a great range of aromatic compounds by acinetobacters.
Three provisos must be borne in mind: (i) a few of the organisms identified
as Acinetobacter may in fact have belonged to other genera; however,
whenever possible papers have been cited only if the evidence for
identification appeared to be adequate; (ii) the literature almost certainly
contains descriptions of aromatic degradations carried out by
acinetobacters, but where the organisms were not identified or were wrongly
assigned to other genera; (iii) there is an unfortunate tendency to postulate
catabolic pathways on the basis of inadequate evidence. Whatever the
shortcomings of some of the work, the accelerating rate of appearance of
publications dealing with aromatic degradation by acinetobacters means that
this selection can represent only a very small part of the total repertoire of
this versatile organism.
365
Benzene: a Plasmid-Encoded Pathway in Some Strains
Hogn and Jaenicke (1972) isolated a benzene-utilising bacterium from

Rhine mud and named it M oraxell a B; it was, however, oxidase-negative
and probably an Acinetobacter. Extracts required catalytic amounts of
NADH for the conversion of benzene into catechol. This is consistent with
the dioxygenation of benzene to benzene cis-glycol (cis-l,2-
dihydroxycyclohexa-3,5-diene) which is then dehydrogenated to catechol, as
occurs in other bacteria (Fig. 6f,g; Fewson, 1988). A. calcoaceticus RJE
7 4, which was isolated from soil, contains the large ( about 200 k b),
transmissible catabolic plasmid pWW174, and this encodes the enzymes for
the catabolism of benzene via the B-ketoadipate pathway (Winstanley et al.,
1987). This was the first report of a plasmid-encoded catechol pathway, and
the catechol 1,2-dioxygenase (Fig. 2a) appeared to be significantly more
thermostable than the chromosomally-encoded enzyme.
Phenol and Salicylate: Biodeterioration and Bioremediation
Phenol and salicylate are presumably converted into catechol by

specific mono-oxygenases, although the evidence rests more on results from
pseudomonads and other organisms than from acinetobacters (Figs. 3 and
6d,e). Both compounds are rather toxic to microorganisms, but can support
growth of various acinetobacters provided that the concentration is kept low
enough (Baumann et al., 1968; Fewson et al., 1970; Jones and Carrington,
1972).
Acinetobacters may be responsible for the biodeterioration of certain

pharmaceutical preparations based on salicylate. Acetylsalicylate can be
subjected to esterase attack by some acinetobacters, with the salicylate being
converted into catechol prior to ortho-cleavage (Fig. 3; Grant et al., 1970;
Grant, 1971, 1973). Salicylamide degradation by an acinetobacter has also
been reported, but in this case the salicylate formed by hydrolysis appeared
to be hydroxylated to form gentisate (2,5-dihydroxybenzoate) before ring-
cleavage (Grant et al., 1975b).
Kim et al. (1986) reported that Acinetobacter strain DY-1 could

utilise phenanthrene as sole source of carbon and energy by degrading it to
salicylate. There was some evidence that the pathway might be plasmid-
encoded.
Some of the classical experiments on the bacterial degradation of

aromatics concerned the oxidation of phenol in gasworks' liquor by 'Vibrio
01' (A. calcoaceticus NCIB8250) (Evans and Happold, 1939; Evans, 1947;
Kilby, 1948). More recently, a great deal of work has been done to clarify
the role of acinetobacters in the degradation of phenol in various sorts of
industrial and agricultural waste waters. Studies have included examination
of the effects of acclimation in activated sludge and various modelling
procedures especially aimed at bioremediation of high phenol
concentrations, often in the presence of cyanide, thiocyanate or other
366
noxious substances (Carbidenc and Zakharchenko, 1969; Jones and
Carrington, 1972; Suzuki and Fujii, 1980, 1985; Borighem and Vereecken,
1981; Anselmo and Novais, 1984; Bourque et aI., 1987; Liang et aI., 1987;
D'Aquino et aI., 1988; Korol et aI., 1989; Tibbles and Baecker, 1989).
Aniline: Evolution in Action
Various strains of A. calcoaceticus can use aniline as sole source of

carbon, nitrogen and energy for growth (Wyndham, 1986; Cho et aI., 1988).
Aniline is converted into catechol (Fig. 3), presumably by a dioxygenase
which releases ammonia, and this is followed by ortho-cleavage. Most
aniline-utilising strains of A. calcoaceticus isolated from river water were
Anio, with a low half-saturation constant and a high specific affinity for
aniline. A much smaller proportion of the population exhibited an Ani +
phenotype and could grow on relatively high concentrations of aniline (up to
16 mM). Adaptation of mixed river water communities to aniline appeared
to depend on the dynamics of parent and mutant populations. In this study
there was no evidence that plasm ids were involved in aniline degradation,
but the frequencies of acquisition and loss of the Ani + phenotype were
thought to be commensurate with repair and subsequent excision of genes by
homologous recombination (Wyndham, 1986); however, there is some
evidence that aniline degradation may be plasmid-encoded in some
acinetobacters (Cho et aI., 1988).
3-Hydroxybenzoate and the Gentisate Pathway; 4-Hydroxy-

benzoate and the Protocatechuate ortho-Cleavage Pathway
All the Acinetobacter strains found by Baumann et aI. (1968) to

utilise 3-hydroxybenzoate appeared to carry out a monohydroxylation to
form gentisate, and this was then ring-cleaved (Fig. 2e). 4-Hydroxybenzoate
is usually hydroxylated at the 3-position and the protocatechuate is generally
ortho-cleaved (Fig. 2c), giving rise to LI-ketoadipate (Cain, 1961), although
in some strains it is meta-cleaved (Fig. 2d; Sze and Dagley, 1987).
Phenylalanine, Tyrosine, Phenylacetate, Phenylalkanes and

4-Hydroxyphenylacetate: Inter-Related Pathways
Phenylacetate metabolism is common amongst acinetobacters and

appears to involve the homogentisate (2,5-dihydroxyphenylacetate) pathway
(Figs. 2f and 8d,e; Dagley et aI., 1952; Chapman and Dagley, 1962;
Baumann et aI., 1968; Keil et aI., 1983). 4-Hydroxyphenylacetate is
hydroxylated to give homoprotocatechuate (3,4-dihydroxyphenylacetate;
Fig. 8j), and this is then subjected to meta-cleavage by
homoprotocatechuate 2,3-dioxygenase (Fig. 2i; Sparnins et aI., 1974;
Barbour and Bayly, 1977). The regulation of this meta pathway has been
studied in some detail (Bayly and Barbour, 1984).
367
[DOH
I
fH.NH Z
(]
L-phenylalanlne
a
""l
(transamInation)
~ 0'2H H:O
o
eaOH COOH
I
I
[=0 f H,NH 2
(]
I
OIlenylpvruv\c acid OH
L-tvrosjn~
.~:::"-,,,,,
o
CHO COOH
6
I I
e=o
I
phenylacetaldehyrle
OH
4-hvdroxyphenV!PyrUvlC acid
h~C02
Q
CHO
phenylacetic acid OH
4-hydrOxYPllenylacetaldehyde
AOH
COOH [DOH
Q
I I
V
2-nydroxypnenvlacetlc acid OH
~
02 2H x4_hYdrOXYDhenylacetiC acid 0,
e
H20
COOH
j 2H02 ~H
k
HO
2
h
MOH
tH
HO
2 ~
eDOH
I
HO::::-""
homogentisIc aCid
VOHOH
nomoprotocatechulc acId
Fig. 8. Degradation via homogentisate and homoprotocatechuate in various

strains of Acinetobacter. Based on Dagley et al. (1952, 1953), Chapman
and Dagley (1962), Kei1 et a1. (1983), Sparnins et a1. (1984), Amund and
Higgins (1985).
368
3-phenylproplonlc acid
COOH COOH
I I
fH 2 HC
II
CH 2
6
CH
6~H
~ OH
H
3-( cis -2.3-dlhvdro-2.3- /IanS-clnnamlc acid
dlhydroxvphenvl-) propionic aCid
COOH COOH
I
fH2 H!
H
&:
3-(2. 3-dlhydroxvphenvl-)
propionic acid
6 OH
4-hydroxyclnnamlc acid
I
I
COOH I
J
~H2 I
fH2 I
c=o
r(v~OOH ~
protocatechuic acid
~OH
l-hydroxy-5-oxo-
L3-heptadlene-L 7-
carboxylic acid
Fig. 9. Metabolism of 3-phenylpropionate and trans-cinnamate in various

strains of Acinetobacter. Based on Keil et al. (1983).
369
Experiments with 'Vibrio 0]' (A. calcoaceticus NCIB8250; Dagley
et aI., 1953), A. calcoaceticus EBF65/61 (Barrowman and Fewson, 1985),
various strains of A. catcoaceticus isolated from soil (Keil et aI., 1983) and
Achromobacter eurydice (Asakawa et aI., 1968; - Brisou and Prevot, 1954,
have suggested that this organism is probably an acinetobacter) all indicated
that L-phenylalanine is degraded via phenylacetate and homogentisate (Fig.
8a-e). Phenylpyruvate decarboxylase (Fig. 8b) has been purified and is a
TPP-dependent tetramer with a subunit M, value of 56,800 (Asakawa et aI.,
1968; Barrowman and Fewson, 1985).
Amund and Higgins (1985) have suggested, although with little

supporting evidence, that A. twoffii AS1 (isolated by elective culture on
North Sea Forties crude oil from an activated sludge sample) converts L-
phenylalanine to homogentisate via L-tyrosine (Fig. 8f-j). Dagley et al.
(1953) had earlier obtained a few results suggesting a similar pathway in
'Vibrio 01'. If this is correct, then it must involve the migration of the acetyl
substituent around the aromatic ring in the final hydroxylation step (Fig. 8j *;
see Hareland et aI., 1975). It remains to be seen if this is the pathway
followed by the many acinetobacters which can grow on L-tyrosine (Baumann
et aI., 1968).
A. twoffii AS1 can completely degrade 1-phenyldodecane. Amund

and Higgins (1985) suggested that this was achieved by initial (J-oxidation,
followed by B-oxidation of coenzyme A derivatives, and then production of
phenylacetate which was in turn metabolised via homogentisate (Fig. 8d,e).
This organism converted 1-phenyltridecane into 3-phenylpropionate and
trans-cinnamate, again presumably viaw and B-oxidation. However, in this
case these two compounds accumulated, apparently because of the absence
of any enzymes for their further metabolism.
Cinnamate and 3-Phenylpropionute: Alternative Pathways
Keil et aI. (1983) isolated several strains of A. calcoaceticus which

could grow on 3-phenylpropionate or trans-cinnamate. Most of the 3-
phenylpropionate was subjected to attack by a dioxygenase which added two
hydroxyl groups to the ring (Fig. 9a) and ring-cleavage was catalysed by 3-
(2,3-dihydroxyphenyl- )propionate 1,2-dioxygenase (Figs. 2h and 9c).
However, some of the 3-phenylpropionate was oxidised to cinnamate, and
this appeared to be converted into protocatechuate (Fig. 9d-f).
DEGRADATION OF LIGNIN-RELATED COMPOUNDS
Many acinetobacters can grow on lignin monomers such as vanillin,

vanillate and anisate (e.g. Fewson, 1967a; Golovleva et aI., 1983; Hasegawa
and Suzuki, 1983). However, this is a common property, shared with many
soil microorganisms. The ability of bacteria to grow on larger molecules
appears to be a more specialised feature, and it may well be that the primary
attack on lignin is generally made by fungi and that bacteria then utilise the
370
6 CH 2- O - - O - O H
~
4-benzyloxyphenol
~
6 OH
4-hydroxybenzaldehyde
AOH
~
~
bj
Q
COOH
OH
4-hydroxybenzoic acid protocatechuic acid

Fig. 10. Metabolism of 4-benzyloxyphenol by an acinetobacter isolated from
forest soil. Tentative scheme based on Vasudevan and Mahadevan (1988).
products of fungal metabolism. Nevertheless, it is clear that some bacteria

can use substrates more complex than the monomeric products of lignin
degradation. For instance, a strain of A. calcoaceticus isolated from
forest soil could use 4-benzyloxyphenol as sole source of carbon and energy
for growth. This compound contains an 0<-0-4 linkage of the type found in
lignin. The evidence for the degradative pathway involved after cleavage of
the ether bond (Fig. lOa) is extremely scanty, but is unusual because two
ring-cleavage substrates, catechol and protocatechuate, may be formed (Fig.
10; Vasudevan and Mahadevan, 1988). The lignin model compound guaiacol
glyceryl ether [3-(2-methoxyphenoxy) 1,2-propanediol] appeared to be
degraded by another strain of Acinetobacter in which the side chain was
removed to form guaiacol (o-methoxyphenol) and this was converted into
catechol prior to ortho-cleavage (Vasudevan and Mahadevan, 1989).
Another lignin model compound, veratrylglycerol-( o-methoxyphenyl) ether
was degraded by the mutualistic action of two bacteria in which an
acinetobacter first formed guaiacol and this was then used by Nocardia
corallina (Crawford, 1975).
371
DEGRADA TION OF MULTI-RING COMPOUNDS CONTAINING
NITROGEN OR SULPHUR
4-Hydroxyquinazoline was degraded by an acinetobacter isolated from

garden soil. There was an initial attack on the heterocyclic ring (Fig. I1a-c)
to give anthranilate, which was then deaminated to catechol (Fig. lId),
followed by meta-cleavage (Fig. 11; Grant and AI-Najjar, 1975; Grant et
al.,1975a).
(JQ
~ ~N
a
.. ~'yOH
~ I ~N
OH OH
4-hydroxYQulnazol ine 2, 4-d i hydroxyqu i nazo line
b~
H
(]CNH2
I ..
c O:-N'C~O
~ COOH ::--.
I COOH
'NH
2
anthranilic acid 2-carboxvpheny I urea
'j
~OH
~ meta-cleavage
~OH
catechol
Fig. 11. Metabolism of 4-hydroxyquinazoline by an acinetobacter isolated

from garden soil. Based on Grant et al. (1975).
Organisms identified as Acinetobacter, or related to

Acinetobacter, have been isolated which can degrade quinoline (Grant and
AI-Najjar, 1976) and isoquinoline (Aislabie et aI., 1989).
There is a preliminary report that an Acinetobacter strain can use

dibenzothiophene as sole source of carbon and energy for growth, but the
metabolic pathway involved is not clear (reviewed by Ensley, 1984).
372
HALOGENATED COMPOUNDS, INCLUDING PCBs AND VARIOUS
PESTICIDES
The general topic of biodegradation of halogenated compounds has

been well-reviewed (Ghosal et al., 1985; Muller and Lingens, 1986;
Rochkind et al., 1986; Reineke and Knackmuss, 1988).
Monohalogenated phenols and benzoates can be mineralised, partially

metabolised, or co-metabolised by various strains of Acinetobacter. There
have been detailed studies of the ability of halogenated compounds to act as
inducers and substrates for the relevant enzymes (Beveridge and Tall, 1969;
Clark et al., 1975; Reber and Thierbach, 1980; Reber, 1982). Strains of A.
calcoaceticus able to use 3- or 4-chlorobenzoate as sole sources of carbon
and energy for growth have been isolated (Zaitsev and Baskunov, 1985;
Adriaens et al., 1989) and 3-chlorobenzoate can be completely oxidised by
mixed cultures containing A. calcoaceticus and Alcaligenes faecalis
(Zaitsev, 1988).
Acinetobacters seem to be common members of soil microflora

following the application of pentachlorophenol, although this seems to be
more a question of tolerance rather than of ability to degrade it (Kato et al.,
1980,1981; Sato et al., 1987).
The herbicide 2,4-dichlorophenoxyacetate can be degraded by various

bacteria, including a strain of Acinetobacter, and plasmids encoding the
enzymes involved in degrading 2,4-dichlorophenoxyacetate can be
transferred to Acinetobacter strains (Don and Pemberton, 1981). The
pathway involves the intermediate formation of 2,4-dichlorophenol, and the
enzyme 2,4-dichlorophenol hydroxylase has been purified and partially
characterised (Beadle and Smith, 1982; Beadle et al., 1984).
4-Chlorobiphenyl can be completely mineralised by bacteria, including

Acinetobacter spp., that contain the catabolic plasmid pSS50 (Shields et
al., 1985; Hooper et al., 1989). 2,4,4'-Trichlorobiphenyl can be partially
metabolised by Acinetobacter strain P6, and the pathway of degradation is
a useful model for the biodegradation of polychlorinated biphenyls (PCBs)
(Fig. 12; Furukawa et al., 1979b). Various acinetobacters are capable of
partial metabolism of a range of PCBs, and this has led to systematic studies
of the action of acinetobacters on pure PCBs and on commercial mixtures of
PCBs such as Kaneclors and Araclor. By and large, the greater the degree of
chlorination, the less readily are the PCBs degraded. In addition, isomers
with chlorine atoms on only one ring are more readily metabolised than those
with substituents on both rings. Isomers with two ortho-substitutions (e.g.
2,2' or 2,6) are very recalcitrant to degradation (Furukawa et al., 1978a,b,
1979a, 1982, 1983, 1989; Focht and Brunner, 1985; Kohler et al., 1988).
Several bacteria, including Acinetobacter, can carry out the

complete dechlorination, demethylation and cleavage of one of the aromatic
rings of Methoxychlor [1,1, 1-trichloro-2,2-bis(p-methoxyphenyl)ethane]
373
(Golovleva et aI., 1984). However, it is likely that the synergistic action of
microbial communities is important in the degradation of many pesticides
and other residues in Nature. Acinetobacters may well be important
constituents of such communities. Thus, co-cultures of A. calcoaceticus
4CB1 and P6 were thought to have potential applications for de halogenation
and mineralisation of specific polychlorobiphenyl cogeners because of their
complementary metabolic activities (Adriaens et aI., 1989). Biodegradative
Cl Cl Cl
Cl
OH
OH
Cl Cl Cl
2,4,4' -trlchl oro- l-chloro-2,3- 2,4,4' -tr I chloro-
biphenyl d ihydroxy-4- 2' 3' -dl hydroxyblpheny 1
(2,4-dichlorophenyl )-
hexa-4, 6-d I ene
Cl
COOH
2,4-dlchlorobenzoic acid
Cl
3-chloro-2-hYdroxy-
6-oxo-6-
(2,4-dichlorophenyl )-
hexa-2,4-dienic acid
Fig. 12. Primary metabolism of 2,4,4'-trichlorobiphenyl by Acinetobacter

P6. Based on Furukawa et al. (1979).
studies on Aroclor 1242 indicated that analogue enrichment with biphenyl is

the most important factor affecting PCB degradation in soil, but additional
enhancement can be achieved by inoculation with Acinetobacter (Brunner
et aI., 1988). A microbial community isolated from wheat root systems was
capable of growth on Mecoprop [2-(2-methyl-4-chlorophenoxy )propionate 1
and contained two Pseudomonas species, an Alcaligenes species, a
Flavobacterium species and A. calcoaceticus (Lappin et aI., 1985).
374
METHYLATION OF HALOGENATED AROMATICS: A DANGEROUS
AL TERNATIVE TO BIODEGRADATION
Microbial O-methylation of halogenated phenols can give neutral

products that have a high potential for bioconcentration and may be more
toxic than the starting compounds. There have been reports of
acinetobacters that can O-methylate halogenated phenols and thiophenols
such as 2,6-dibromophenol, 4,5,6-trichloroguaiacol, pentachlorothiophenol
and high molecular mass chlorinated lignin (Neilson et aI., 1984, 1988;
Allard et aI., 1987).
POSSIBLE APPLICATIONS OF ACINETOBACTERS IN AROMATIC

WASTE TREATMENT AND THEIR ROLE IN BIODETERIORATION
There are several potential or actual applications of acinetobacters in

the degradation of hazardous and unpleasant waste and residue aromatic
compounds in addition to those described in previous sections. Examples
include the degradation of cresols (Suzuki and Fujii, 1982), phthalic acid
esters (Taylor et aI., 1981), diphenylethylene (Focht and Joseph, 1984) and
anisate (Ventosa et aI., 1984; Ferrer et aI., 1985); the utilisation of
acinetobacters with the potential to degrade aromatic fractions of petroleum
(Walker et aI., 1975, 1976; Yall, 1979; Ling et aI., 1986) and crude oil
(Fedorek and Westlake, 1983); fluoranthene-utilisation by bacterial
communities growing on the polycyclic aromatic hydrocarbon components of
creosote (Mueller et aI., 1989); degradation of aromatic hydrocarbons in
nuclear reactor coolants (Catelani et aI., 1970); degradation of linear
alkylbenzene sulphonates (Hrsak et aI., 1982).
The possible role of acinetobacters as agents of biodeterioration is

exemplified by strain NA V2 which excretes an emulsifier that is capable of
emulsifying the saturate and naphthene aromatic fractions of asphalt
(Pendrys, 1989).
BIOSYNTHESIS OF AROMATIC AMINO ACIDS AND POSSIBLE

INTERACTIONS WITH DEGRADATIVE PATHWAYS
Competition for 5-Dehydroshikimate Between the Route for the

Biosynthesis of Aromatic Amino Acids and that for the Catabolism
of Hydroaromatic Compounds
Many strains of A. calcoaceticus can grow on quinate or shikimate

and they all seem to use the shikimate pathway for the biosynthesis of
aromatic amino acids and related compounds. Chorismic acid, the key
intermediate in the synthesis of aromatic compounds, is formed from
erythrose 4-phosphate and phosphoenolpyruvate by seven enzymes (Fig.
13a,b,f-j). Shikimate and quinate are dehydrogenated by a single, NAD(P)-
independent, enzyme (Fig. 13c; Tresguerres et aI., 1970a,b). There are two
375
quinic acid
I
'e
5-dehYdrojuinic acid ~ 5-dehydroshikimic acid ~ shikimic acid
t b
ile f t9
lh
I
3-deoxy-D-
t
protocatechuic acid
arabino heptulosonic acid
7-phosphate
~ i
ta
chorismic acid
erythrose 4-phosphate ortho

d~~
+ cleavage aromatic amino
phosphoenolpyruvate acids and
related compounds
Fig. 13. Competition in A. calcoaceticus between the shikimate pathway

for the biosynthesis of aromatic amino acids and related compounds (---»
and the pathway for the degradation of quinate and shikimate (- - -» . Based
on Tresguerres et al. (1972), Berlyn and Giles (1973).
dehydroquinase isoenzymes, one of which is constitutive and is involved in

biosynthesis (Fig. 13j), and one which is inducible and is involved in the
catabolic pathway (Fig. 13d; Ingledew et al., 1971). The biosynthetic and
catabolic pathways therefore share common intermediates. It was thought
at one time that this problem was overcome by channeling of intermediates
within multi-enzyme complexes (Tresguerres et al., 1972), but this does not
seem to be the case. Rather, end-product induction of the degradative
enzymes by protocatechuate (Canovas et al., 1968; Ingledew et al.. 1971)
ensures that the biosynthetic intermediates do not induce enzymes for their
own degradation, since only at high levels of exogenously supplied quinate or
shikimate is enough protocatechuate presumed to be present to induce the
catabolic enzymes (Berlyn and Giles, 1973; Ornston and N eidle, this
volume). This inclusion of the shikimate/ quinate enzymes within the
degradative group induced by protocatechuate contrasts with the fact that 4-
hydroxybenzoate 3-hydroxylase (Fig. 7e), which also feeds into the
protocatechuate pathway, is substrate-induced (Canovas and Stanier, 1967),
but this probably causes no difficulty since 4-hydroxybenzoate is not an
important anabolic intermediate.
Metabolism of Anthranilate and Tryptophan
Strains of A. calcoaceticus that can grow on tryptophan appear to

use the kynurenine pathway for catechol formation (Fig. 3). Tryptophan
induces the first enzyme, tryptophan oxygenase. The next two enzymes,
formamidase and kynureninase, are induced by kynurenine. The last enzyme,
anthranilate oxidase, is substrate-induced (Wheelis, 1972), and exogenous
anthranilate is a good growth substrate for these strains. Tryptophan is
synthesised from anthranilate, which is in turn formed from chorismate
produced by the shikimate pathway (Fig. 13; Cohn and Crawford, 1976).
Again, the patterns of induction of the catabolic enzymes appear to minimise
problems that might arise if the pool of anthranilate required for anabolic
purposes was depleted by unregulated catabolism.
376
EvolutionaryAspects of the Biosynthesis of Tyrosine, Phenylalanine
and Tryptophan
Phenylalanine and tyrosine are synthesised from products of the

shikimate pathway (see Fig. 13), but are degraded by pathways, described
elsewhere in this chapter (see Fig. 8), which do not overlap significantly with
the synthetic pathways, and which therefore introduce no particular
problems in their regulation. Several enzymes involved in the biosynthesis
of aromatic amino acids in A. cal co acetic us have been purified and
characterised: these include anthranilate synthetase (Sawula and Crawford,
1973), the bifunctional chorismate mutase-prephenate dehydratase (Berry et
aI., 1985; Ahmad et aI., 1988) and two 3-deoxy-D-arabino-heptulosonate-7-
phosphate synthases (Byng et aI., 1985). In addition, some of the genes have
been mapped, cloned and sequenced (Sawula and Crawford, 1972; Kaplan et
aI., 1984; Ross et aI., 1990; Haspel et aI., this volume). The results obtained
from examination of enzyme properties and DNA sequences have led to
some intriguing speculations about the evolution of these biosynthetic
pathways and about the Acinetobacter line of bacterial descent;
unfortunately these are beyond the scope of the present chapter, but they
have been comprehensively discussed elsewhere (Byng et aI., 1985; Ahmad
et aI., 1987; Ahmad and Jensen, 1988; Jensen and Ahmad, 1988; Crawford,
1989; Ross et aI., 1990; Haspel et aI., this volume).
ENVOI
The very plethora of aromatic compounds that can be utilised by

acinetobacters, together with the scattered and incomplete nature of many of
the observations, makes it difficult to draw any significant generalisations,
beyond those of the sort illustrated in Fig. 1. Nevertheless, it is clear that
detailed studies of all aspects of aromatic metabolism will shed light on the
ways in which peripheral metabolic pathways have evolved and are regulated,
and will underpin attempts to exploit acinetobacters and other organisms for
the biodegradation and bioremediation of hazardous and unpleasant
aromatic wastes, spillages and residues.
REFERENCES
Adriaens, P., Kohler, H.P.E., Kohler-Staub, D., and Focht, D.D., 1989,
Bacterial dehalogenation of chlorobenzoates and coculture
biodegradation of 4,4'-dichlorobiphenyl, Appl. Environ.
Microbiol.,55:887.
Ahmad, S., and Jensen, R.A., 1988, The phylogenetic origin of the
bifunctional tyrosine-pathway protein in the enteric lineage of
bacteria, Mol. BioI. Evol., 5:282.
Ahmad, S., Johnson, J.L., and Jensen, R.A., 1987, The recent evolutionary
origin of the phenylalanine-sensitive isozyme of 3-deoxy-D-
ara bin o-heptulosonate-7 -phosphate synthase in the enteric
lineage of bacteria, 1. Mol. Evol., 25:159.
377
Ahmad, S., Wilson, A.T., and Jensen, R.A., 1988, Chorismate mutase:
prephenate dehydratase from Acinetobacter calcoaceticus.
Purification, properties and immunological cross-reactivity, Eur.
I. Biochem., 176:69.
Aislabie, J., Rothenburger, S., and Atlas, R.M., 1989, Isolation of
microorganisms capable of degrading isoquinoline under aerobic
conditions, Appl. Environ. Microbiol., 55:3247.
Allard, A.S., Remberger, M., and Neilson, A.H., 1987, Bacterial 0-
methylation of halogen-substituted phenols, Appl. Environ.
Microbiol.,53:839.
Allison, N., and Fewson, C.A., 1986, Purification and properties of L-lactate
dehydrogenase from Acinetobacter calcoaceticus, FEMS
Allison, N., O'Donnell, M.J., Hoey, M.E., and Fewson, C.A., 1985a,
Membrane-bound lactate dehydrogenases and mandelate
dehydrogenases of Acinetobacter calcoaceticus. Location and
regulation of expression, Biochem. I., 227:753.
Allison, N., O'Donnell, M.J., and Fewson, C.A., 1985b, Membrane-bound
lactate and mandelate dehydrogenases of Acinetobacter
calcoaceticus: purification and properties, Biochem. I.,
231:407.
Amund, 0.0., and Higgins, 1.1.,1985, The degradation of 1-phenylalkanes by
an oil-degrading strain of Acinetobacter lwoffi,
Ant. v.Leeuw., 51:45.
Anselmo, A.M., and N ovais, J .M., 1984, Isolation and selection of phenol-
degrading microorganisms from an industrial effluent, Biotech.
Lett., 6:601.
Asakawa, T., Wada, H., and Yamano, T., 1968, Enzymatic conversion of
phenylpyruvate to phenylacetate, Biochim. Biophys. Acta,
170:375.
Barbour, M.G., and Bayly, R.C., 1977, Regulation of the 4-hydroxy-
phenylacetic and meta-cleavage pathway in an Acinetobacter sp.,
Biochem. Biophys. Res. Comm., 79:663.
Barkovsky, A.L., and Shub, G.M., 1986, An Acinetobacter calcoaceticus
strain utilising various aromatic compounds and carrying a plasmid
for resorcin degradation, M ikrobi ologiya, 55:237.
Barrowman, M.M., and Fewson, C.A., 1985, Phenylglyoxylate decarboxylase
and phenylpyruvate decarboxylase from Acinetobacter
calcoaceticus, Curro Microbiol., 12:235.
Barrowman, M.M., Harnett, W., Scott, A.J., Fewson, C.A., and Kusel, J.R.,
1986, Immunological comparison of microbial TPP-dependent
non-oxidative a-keto acid decarboxylases, FEMS Microbiol.
Lett., 34:57.
Moraxella group. II. Oxidase-negative species (genus
Acinetobacter), I. Bacteriol., 95:1520.
Bayly, R.C., and Barbour, N.G., 1984, The degradation of aromatic
compounds by the meta and gentisate pathways, in "Microbial
Degradation of Aromatic Compounds," p.253, D.T. Gibson, ed.,
Marcel Dekker, New York.
378
Beadle, C.A, and Smith, A.R. W., 1982, The purification and properties of
2,4-dichlorophenol hydroxylase from a strain of Acinetobacter
species, Eur. f. Biochem., 123:323.
Beadle, C.A., Kyprianou, P., Smith, A.R.W., Weight, M.L., and Yon, R.J.,
1984, Rapid purification of 2,4-dichlorophenol hydroxylase by
biospecific desorption from 10-carboxydecylamino Sepharose,
Biochem. Internat., 9:587.
Beggs, J.D., and Fewson, C.A., 1977, Regulation of synthesis of benzyl
alcohol dehydrogenase in Acinetobacter calcoaceticus
NCIB8250, f. Gen. Microbiol., 103:127.
Beggs, J.D., Cook, AM., and Fewson, C.A, 1976, Regulation of growth of
Acinetobacter calcoaceticus NCIB8250 on benzyl alcohol in
batch culture, f. Gen. Microbiol., 96:365.
Berlyn, M.B., and Giles, N.H., 1973, Organisation of interrelated aromatic
metabolic pathway enzymes in Acinetobacter calco-aceticus,
f. Gen. Microbio!., 74:337.
Berry, A., Byng, G.S., and Jensen, R.A., 1985, Interconvertible molecular-
weight forms of the bifunctional chorismate mutase-prephenate
dehydratase from Acinetobacter calcoaceticus, Arch.
Biochem. Biophys., 243:470.
Beveridge, E.G., and Tall, D., 1969, The metabolic availability of phenol
analogues to bacterium NCIB8250, f. Appl. Bacteriol., 32:304.
Borighem, G., and Vereecken, J., 1981, Model of a chemostat utilising phenol
as inhibitory substrate, Ecolog. Modelling, 12:231.
Bourque, D., Bisaillon, J.G., Beaudet, R., Sylvestre, M., Ishaque, M., and
Morin, A., 1987, Microbiological degradation of malodorous
substances of swine waste under aerobic conditions, Appl.
Environ. Microbio!., 53:137.
Brisou, J., and Prevot, A-R., 1954, Etudes de systematique bacterienne. X.
Revision des especes reunies dans Ie genre Achromobacter,
Brunner, W., Sutherland, F.H., and Focht, D.D., 1988, Enhanced
biodegradation of polychlorinated biphenyls by analog enrichment
and bacterial inoculation, 1. Environ. Qual., 14:324.
Burlingame, R., and Chapman, P.J., 1983, Stereospecificity in meta-fission
catabolic pathways, f. Bacteriol. 155:424.
Byng, G.S., Berry, A, and Jensen, R.A, 1985, Evolutionary implications of
features of aromatic amino acid biosynthesis in the genus
Acinetobacter, Arch. Microbiol., 143:122.
Cain, R.B., 1961, The metabolism of protocatechuic acid by a Vibrio,
Biochem. f., 79:298.
Canovas, J.L., and Johnson, B.F., 1968, Regulation of the enzymes of the B-
ketoadipate pathway in Moraxella calcoacetica. 4.
Constitutive synthesis of B-ketoadipate succinyl-CoA transferases
II and III, Eur. f. Biochem., 3:312.
Canovas, J.L., and Stanier, R.Y., 1967, Regulation of the enzymes of the B-
ketoadipate pathway in M oraxella calcoacetica. 1. General
aspects, Eur. f. Biochem., 1:289.
379
Canovas, J.L., Ornston, L.N., and Stanier, R.Y., 1967, Evolutionary
significance of metabolic control systems, Science, 156:1695.
Canovas, J.L., Johnson, B.F., and Wheelis, M.L., 1968, Regulation of the
enzymes of the B-ketoadipate pathway in M oraxella
calcoacetica. 3. Effects of 3-hydroxy-4-methylbenzoate on the
synthesis of enzymes of the protocatechuate branch, Eur. 1.
Biochem., 3:305.
Canovas, J.L., Wheelis, M.L., and Stanier, R.Y., 1968, Regulation of the
enzymes of the B-ketoadipate pathway in M oraxell a
calcoacetica. 2. The role of protocatechuate as inducer, Eur. 1.
Biochem., 3:293.
Carbidenc, R., and Zakharchenko, V., 1969, Possibilities of biological
decomposition of an effluent incorporating formol and phenol,
Terres Eaux, 61:8.
Catelani, D., Mosselmans, G., Nienhaus, J., Sorlini, c., and Treccani, V.,
1970, Microbial degradation of aromatic hydrocarbons used in
reactor coolants, Experientia, 26:922.
Chalmers, R.M., and Fewson, C.A., 1989a, The evolution of metabolic
pathways: an immunological approach to the evolution of
aromatic and aldehyde dehydrogenases, in "Enzymology and
Molecular Biology of Carbonyl Metabolism", p.193, H.Weiner and
T.G. Flynn, eds., Arthur Liss, New York.
Chalmers, R.M., and Fewson, C.A., 1989b, Purification and characterization
of benzaldehyde dehydrogenase I from Acinetobacter
calcoaceticus, Biochem. I., 263:913.
Chapman, P.J., and Dagley, S., 1962, Oxidation of homogentisic acid by cell-
free extracts of a vibrio, 1. Gen. Microbiol., 28:251.
Chari, R.V.J., Whitman, c.P., Kozarich, J.W., Ngai, K.L., and Ornston,
L.N., 1987a, Absolute stereochemical course of the 3-
carboxymuconate cycloisomerase from Pseudomonas putida and
Acinetobacter calcaceticus: analysis and implications, 1.
Amer. Chem. Soc., 109:5514.
Chari, R.V.J., Whitman, c.P., Kozarich, J.W., Ngai, K.L., and Omston, L.N.,
1987b, Absolute stereochemical course of muconolactone 6-
lactoneisomerase and of 4-carboxymuconolactone decarboxylase:
proton NMR 'ricochet' analysis, 1. Amer. Chem. Soc., 109:5520.
Cho, K. Y., Ha, I.H., Baf, K.S., and Kho, Y.H., 1988, Isolation
and characterization of aniline degrading bacteria, San 0 p
Misaengmul H akhoechi, 16:421.
Clarke, K.F., Callely, A.G., Livingstone, A., and Fewson, C.A., 1975,
Metabolism of monofluorobenzoates by Acinetobacter
calcoaceticus NCIB8250, Biochim. Biophys. Acta, 404:169.
Cohn, W., and Crawford, I.P., 1976, Regulation of enzyme synthesis in the
tryptophan pathway of Acinetobacter calcoaceticus, 1.
Cook, A.M., and Fewson, C.A., 1972, Evidence for specific transport
mechanisms for aromatic compounds in bacterium NCIB8250,
380
Cook, A.M., Beggs, J.D., and Fewson, c.A., 1975, Regulation of growth of
Acinetobacter calcoaceticus NCIB8250 on L-mandelate in
batch culture, 1. Gen. Microbiol., 91:325.
Crawford, J.P., 1989, Evolution of a biosynthetic pathway: the tryptophan
paradigm, Ann. Rev. Microbiol., 43:567.
Crawford, R.L., 1975, Mutualistic degradation of the lignin model compound
veratrylglycerol (o-methoxyphenyl)ether by bacteria, Can. 1.
Dagley, S., 1978, Pathways for the utilization of organic growth substrates, in
"The Bacteria, vol.VI, Bacterial Diversity," p.305, L.N. Ornston
and J.R. Sokatch, eds., Academic Press, New York.
Dagley, S., Fewster, M.E., and Happold, F.C., 1952, The bacterial oxidation
of phenylacetic acid, 1. Bacteriol., 63:327.
Dagley, S., Fewster, M.E., and Happold, F.C., 1953, The bacterial oxidation
of aromatic compounds, 1. Gen. Microbiol., 8:1.
D'Aquino, M., Korol, S., Santini, P., and Moretton, J., 1988, Biodegradation
of phenolic compounds. I. Improved degradation of phenol and
benzoate by indigenous strains of Acinetobacter and
Pseudomonas, Revista Latinoamer.Microbioiogia, 30:283.
Darrah, J .A., and Cain, R.B., 1969, A convenient biological method for
preparing B-ketoadipic acid, Lab. Pract., 16:989.
Don, R.H., and Pemberton, J.M., 1981, Properties of six pesticide
degradation plasmids isolated from Alcaligenes paradoxus and
Alcaligenes eutrophus, J. Bacterio!., 145:681.
Doten, R.C., Ngai, K.L., Mitchell, D.J., and Ornston, L.N., 1987a, Cloning
and genetic organization of the pca cluster from Acinetobacter
calcoaceticus, 1. Bacterio!., 169:3168.
Doten, R.C., Gregg, L.A., and Ornston, L.N., 1987b, Influence of the
catBCE sequence on the phenotypic reversion of a pcaE mutation
in Acinetobacter calcoaceticus, J. Bacteriol., 169:3175.
Durham, D.R., Stirling, L.A., Ornston, L.N., and Perry, J.J., 1980,
Intergeneric evolutionary homology revealed by the study of
protocatechuate 3,4-dioxygenase from Azotobacter vinelandii,
Ensley, B.D., 1984, Microbial metabolism of condensed thiophenes, in
"Microbial Degradation of Aromatic Compounds," p.309, D.T.
Gibson, ed., Marcel Dekker, New York.
Evans, W.c., 1947, Oxidation of phenol and benzoic acid by some soil
bacteria, Biochem. J., 41:373.
Evans, W.c., 1988, Anaerobic degradation of aromatic compounds, Ann.
Rev. Microbio!., 42:289.
Evans, W.c., and Happold, F.C., 1939, The utilization of phenol by bacteria,
1. Chem. Soc., 58:55.
Fedorak, P.M., and Westlake, D.W.S., 1983, Selective degradation of
biphenyl and methylbiphenyls in crude oil by two strains of marine
bacteria, Can. 1. Microbio!., 29:497.
Ferrer, M.R., Del Moral, A., Ruiz-Berraquero, F., and Ramos-Cormenzano,
A., 1985, Ability of o-anisate degrading microorganisms to
cometabolize DICAMBA and other related pesticides,
Chemosphere, 14: 1645.
381
Fewson, C.A., 1967a, The growth and metabolic versatility of the gram-
negative bacterium NCIB8250 ('Vibrio 01'), 1. Gen. Microbiol.,
46:255.
Fewson, C.A., 1967b, The identity of the Gram-negative bacterium
NCIB8250 ('Vibrio 01'), 1. Gen. Microbiol., 48:107.
Fewson, C.A., 1981, Biodegradation of aromatics with industrial relevance,
in "Microbial Degradation of Xenobiotics and Recalcitrant
Compounds," p.143, T.Leisinger, R. Hutter, A.M. Cook and J.
Nuesch, eds., Academic Press, London.
Fewson, C.A., 1988, Microbial metabolism of mandelate: a microcosm of
diversity, FEMS Microbiol. Rev., 54:85.
Fewson, C.A., Barclay, A., and Eason, R., 1970, Batch culture and computer
simulation of growth of bacterium NCIB8250 using inhibitory
substrates, 1. Gen. Microbiol., 61:i.
Fewson, C.A., Livingstone, A., and Moyes, H.M., 1978, Mutants of
Acinetobacter calcoaceticus NCIB8250 constitutive for the
mandelate enzymes, J. Gen. Microbiol., 106:233.
Fewson, c.A., Allison, N., Hamilton, I.D., Jardine, J., and Scott, A.J., 1988,
Comparison of mandelate dehydrogenase from various strains of
Acinetobacter calcoaceticus: similarity of natural and evolved
forms, J. Gen. Microbiol., 134:967.
Focht, D.D., and Brunner, W., 1985, Kinetics of biphenyl and
polychlorinated biphenyl metabolism in soil, Appl. Environ.
Focht, D.D., and Joseph, H., 1984, Degradation of 1,1-diphenylethylene by
mixed cultures, Can. J. Microbiol., 20:631.
Furukawa, K., and Chakrabarty, A.M., 1982, Involvement of plasmids in total
degradation of chlorinated biphenyls, Appl. Environ.
Microbiol., 44:619.
Furukawa, K., Matsumura, F., and Tonomura, K., 1978a, Alcaligenes and
Acinetobacter strains capable of degrading polychlorinated
biphenyls, Agric. Bioi. Chern., 42:543.
Furukawa, K., Tonomura, K., and Kamibayashi, A., 1978b, Effect of chlorine
substitution on the biodegradability of polychlorinated biphenyls,
Furukawa, K., Tomizuka, N., and Kamibayashi, A., 1979a, Effect of chlorine
substitution on the bacterial metabolism of various
polychlorinated biphenyls, Appl. Environ. Micrbiol., 38:301.
Furukawa, K., Tonomura, K., and Kamibayashi, A., 1979b, Metabolism of
2,4,4'-trichlorobiphenyl by Acinetobacter sp. P6, Agric. BioI.
Chern., 43:1577.
Furukawa, K., Tomizuka, N., and Kamibayashi, A., 1982, Bacterial
degradation of polychlorinated biophenyls (PCB) and their
metabolites, Adv. Exp. M ed. BioI., 136A:407.
Furukawa, K., Tomizuka, N., and Kamibayashi, A., 1983, Metabolic
breakdown of Kenoclors (polychlorobiphenyls) and their products
by Acinetobacter sp., Appl. Environ. Microbiol., 46:140.
382
Furukawa, K., Hayase, N., Taira, K., and Tomizuka, N., 1989, Molecular
relationship of chromosomal genes encoding
biphenyl/polychlorinated biphenyl catabolism: some soil bacteria
possess a highly conserved bph operon, 1. Bacteriol., 171:5467.
Ghosal, D., You, I.-S., Chatterjee, D.K., and Chakrabarty, A.M., 1985,
Microbial degradation of halogenated compounds, Science,
228:135.
Golovleva, L.A., Pertsova, R.N., Fedechkina, I.E., Ganbarov, K.G., and
Baskunov, B.P., 1983, Microbial decomposition of lignin and a
possible role of dioxygenases in its destruction, Prikladnaya
Biokhimiya i Mikrobiologiya, 19:709.
Golovleva, L.A., Polyakova, A.B., Pertsova, R.N., and Finkel'shtein, Z.I.,
1984, The fate of methoxychlor in soils and transformation by soil
microorganisms, 1. Environ. Sci. Health, B19:523.
Grant, D.l.W., 1971, Degradation of acetylsalicylic acid by a strain
ofAcinetobacter lwoffii, 1. Appl. Bacteriol., 34:689.
Grant, D.l. W., 1973, The degradative versatility, arylesterase activity and
hydroxylation reactions of Acinetobacter lwoffii NCIBI0553, 1.
Grant, D.l.W., and AI-Najjar, T.R., 1975, Degradation of 4-hydroxy-
quinazoline by an Acinetobacter and a Pseudomonas isolated
from soil, Microbios, 14:55.
Grant, D.l.W., and AI-Najjar, T.R., 1976, Degradation of quinoline by a soil
bacterium, Microbios, 15:177.
Grant, D.l.W., de Szecs, l., and Wilson, l.V., 1970. Utilization of
acetylsalicylic acid as sole carbon source and the induction of its
enzymatic hydrolysis by an isolated strain of Acinetobacter
lwoffii, 1. Pharmaceut. Pharmacol., 22:461.
Grant, D.l.W., AI-Najjar, T.R., Barber, H.L., and Harbour, M.l., 1975a,
Tangential pathways of catechol cleavage in an Acinetobacter
which degrades 4-hydroxyquinazoline and in a Micrococcus
which degrades 2,4-dihydroxyquinazoline, Micro bios, 14: 195.
Grant, D.l.W., Burgess, S.M., Rambow, E.J., and Yelling, P.W., 1975b,
Bacterial degradation of salicylamide, J. Pharmaceut.
Pharmacol.,27:49P.
Happold, F.e., and Key, A., 1932, The bacterial purification of gasworks'
liquors. The action of the liquors on the bacterial flora of sewage,
1. Hyg., 32:573.
Hareland, W.A., Crawford, R.L., Chapman, P.l., and Dagley, S., 1975,
Metabolic function and properties of 4-hydroxyphenylacetic acid
I-hydroxylase from Pseudomonas acidovorans, 1. Bacteriol.,
121:272.
Hasegawa, S., and Suzuki, M., 1983, Microbiological decomposition of
aromatic compounds by Acinetobacter calcoaceticus,
Hokkaidoritsu Eisei Kenkynshoho, 33:134.
Hills, e.A., and Fewson, e.A., 1983a, Mutant strains of Acinetobacter
calcoaceticus possessing additional mandelate dehydrogenases.
Identification and preliminary characterization of the enzymes,
Biochem. I., 209:379.
383
Hills, C.A., and Fewson, C.A., 1983b, Regulation of expression of novel
mandelate dehydrogenases in mutants of Acinetobacter
calcoaceticus,.T. Gen. Microbiol., 129:2009.
Hoey, M.E., Allison, N., Scott, A.J., and Fewson, C.A., 1987, Purification and
properties of L-mandelate dehydrogenase and comparison with
other membrane-bound dehydrogenases from Acinetobacter
calcoaceticus, Biochem . .I., 248:871.
Hoey, M.E., and Fewson, C.A., 1990, Is there a mandelate complex in
Acinetobacter calcoaceticus or Pseudomonas putida? .T.
Hogn, T., and Jaenicke, L., 1972, Benzene metabolism of M oraxella
species, Eur . .T. Biochem., 30:369.
Hooper, S.W., Dockendorff, T.C., and Sayler, G.S., 1989, Characteristics
and restriction analysis of the 4-chlorobiphenyl catabolic plasmid
pSS5(), Appl. Environ. Microbiol., 55:1286.
Hou, c.T., Lillard, M.O., and Schwartz, R.D., 1976, Protocatechuate 3,4-
dioxygenase from Acinetobacter calcoaceticus,
Biochemistry, 15: 582.
Hou, C.T., Patel, R., and Lillard, M.O., 1977, Extradiol cleavage of 3-
methylcatechol by catechol 1,2-dioxygenase from various
microorganisms, Appl. Environ. Microbiol., 33:725.
Hou, c.T., Patel, R.N., Lillard, M.O., Felix, A., and Florance, J., 1978,
Circular dichroism of catechol 1,2-dioxygenase from
Acinetobacter calcoaceticus, Bioorgan. Chem., 7:115.
Hrsak, D., Busnjak, M., and Johanides, V., 1982, Enrichment of linear
alkyl benzene sulphonate (LAS) degrading bacteria in continuous
culture,.l. Appl. Bacteriol., 53:413.
Hughes, E.J., Shapiro, M.K., Houghton, J.E., and Ornston, L.N., 1988,
Cloning and expression of pca genes from Pseudomonas putida
in Escherichia coli, .I. Gen. Microbiol., 134:2877.
Ingledew, W.M., Tresguerres, M.E.F., and Canovas, J.L., 1971, Regulation of
the enzymes of the hydro aromatic pathway in Acinetob acter
calco-aceticus, .I. Gen. Microbiol., 68:273.
Jensen, R.A., and Ahmad, S., 1988, Evolution and phylogenetic distribution
of the specialized isozymes of 3-deoxy-D-arabino-heptulosonate 7-
phosphate synthase in Superfamily-B prokaryotes, Microbiol.
Sci., 5:316.
Jones, G.L., and Carrington, E.G., 1972, Growth of pure and mixed cultures
of microorganisms concerned in the treatment of carbonization
waste liquors,.l. Appl. Bacteriol., 35:395.
Juni, E., and Janik, A., 1969, Transformation of Acinetobacter
calco-aceticus (Bacterium anitratum),.l. Bacteriol., 98:281.
Kaplan, J.B., Goncharoff, P., Seibold, A.M., and Nichols, B.P., 1984,
trpGDC gene cluster, M of. BioI. Evol., 1:456.
Katagiri, M., and Wheelis, M.L., 1971, Comparison of the two isofunctional
enol-lactone hydrolases from Acinetobacter calcoaceticus, .T.
384
Kato, H., Sato, K., and Furusaka, C., 1980, Effect of pentachlorophenol
(PCP) on soil bacterial flora in soil suspension, Tohoka Daigaku
N agaku K enkyuosho H okoku, 32: l.
Kato, H., Sato, K., and Furusaka, c., 1981, Effects of PCP on bacterial flora
in water-logged soil, Nippon N oyaku Gakkaishi, 6:163.
Keil, H., Tittmann, U., and Lingens, F., 1983, Degradation of aromatic
carboxylic acids by Acinetobacter, Syst. Appl. Microbiol.,
4:313.
Kennedy, S.l. T., and Fewson, C.A., 1968a, Metabolism of mandelate and
related compounds by bacterium NCIB8250, J. Gen. Microbio!.,
53:259.
Kennedy, S.I.T., and Fewson, C.A., 1968b, Enzymes of the mandelate
pathway in bacterium NCIB8250, Biochem. J., 107:497.
Kilby, B.A., 1948, The bacterial oxidation of phenol to B-ketoadipic acid,
Biochem. 1., 43:v.
Kilby, B.A., 1951, The formation of B-ketoadipic acid by bacterial fission of
aromatic rings, Biochem. J., 49:67l.
Kim, C.K., Kim, l.W., Kim, W.C., and Mheen, T.I., 1986, Isolation of
aromatic hydrocarbon-degrading bacteria and genetic
characterization of their plasmid genes, Misaengmal
H akhoechi, 24:67.
Kohler, H.-P. E., Kohler-Staub, D., and Focht, D.D., 1988, Cometabolism of
polychlorinated biphenyls: enhanced transformation of Aroclor
1254 by growing bacterial cells, Appl. Environ. Microbiol.,
54:1940.
Korol, S., Orsingher, M., Santani, P., Moretton, l., and D'Aquino, M., 1989,
Biodegradation of phenolic compounds. II. Effects of inoculum,
xenobiotic concentration and adaptation on Acinetobacter and
Pseudomonas phenol degradation, Revista Latinoamer.
Microbiol., 31:117.
Kozarich, l.W., 1988, Enzyme chemistry and evolution in the B-ketoadipate
pathway, in "Microbial Metabolism and the Carbon Cycle," p.283,
S.R. Hagedorn, R.S. Hanson and D.A. Kunz, eds., Harwood,
London.
Lappin, H.M., Greaves, M.P., and Slater, l.H., 1985, Degradation of the
herbicide Mecoprop [2-(2-methyl-4-chlorophenoxy)propionic
acid] by a synergistic microbial community, Appl. Environ.
Microbiol.,49:429.
Leung, P.-T., Chapman, P.l. and Dagley, S., 1974, Purification and properties
of 4-hydroxy-2-ketopimelate aldolase from Acinetobacter, 1.
Bacteriol.,120:168.
Liang, c., Li, L., Quin, c., Fan, L., and Yuan, X., 1987, Phenol degradation
by Acinetobacter in wastewater from Xiangan Chemical
Industrial Plant, Beijing Daxne Xuebao, Ziran K exueban, 77.
Ling, Z., Xiong, X., and Qian, c., 1986, Application of microbial degradation
in the treatment of petrochemical waste, Huanjing Huaxne, 5:7.
Livingstone, A., and Fewson, C.A., 1972, Regulation of the enzymes
converting mandelate into benzoate in bacterium NCIB8250,
Biochem. J., 130:937.
385
Livingstone, A., Fewson, c.A., Kennedy, S.I.T., and Zatman. L.J., 1972, Two
benzaldehyde dehydrogenases in bacterium NCIB8250.
Distinguishing properties and regulation, Bi 0 c hem. i., 130:927.
Lovrien, R., Jorgenson, G., Ma, M.K., and Sund, W.E., 1980,
Microcalorimetry of microorganism metabolism of
monosaccharides and simple aromatic compounds, Biotech.
Bioeng., 22:1249.
MacKintosh, R.W., and Fewson, C.A., 1987, Microbial aromatic alcohol and
aldehyde dehydrogenases, in "Enzymology and Molecular Biology
of Carbonyl Metabolism," p.259, H. Weiner, and T.G. Flynn, eds.,
Arthur Liss, New York.
MacKintosh, R.W., and Fewson, C.A., 1988a, Benzyl alcohol dehydrogenase
and benzaldehyde dehydrogenase II from Acinetobacter
calcoaceticus. Purification and preliminary characterization,
Biochem. i., 250:743.
MacKintosh, R.W., and Fewson, c.A., 1988b, Benzyl alcohol dehydrogenase
and benzaldehyde dehydrogenase II from Acinetobacter
calcoaceticus. Substrate specificities and inhibition studies,
Biochem. i .., 255:653.
McCorkle, G.M., Yeh, W.-K., Fletcher, P., and Omston, L.N., 1980,
Repetitions in the NH 2 -terminal amino acid sequence of B-
ketoadipate enol-lactone hydrolase from Pseudomonas putida,
i. Bioi. Chem., 255:6335.
Mueller, J.G., Chapman, P.J., and Pritchard, P.H., 1989, Action of a
fluoranthene-utilizing bacterial community on polycyclic aromatic
hydrocarbon components of creosote, Appl. Environ.
Microbiol.,55:3085.
Miiller, R., and Lingens, F., 1986, Microbiological degradation of
halogenated hydrocarbons: a biological solution to pollution
problems?, Angewandte Chemie, Internat. Edn., 25:779.
Neidle, E.L., and Ornston, L.N., 1986, Cloning and expression of
Acinetobacter calcoaceticus catechol 1,2-dioxygenase
structural gene catA in Escherichia coli, 1. Bacterial.,
168:815.
Neidle, E.L., and Ornston, L.N., 1987, Benzoate and muconate, structurally
dissimilar metabolites, induce expression of catA in
Acinetabacter calcaaceticus, i. Bacterial., 169:414.
Neidle, E.L., Shapiro, M.K., and Ornston, L.N., 1987, Cloning and expression
in Escherichia cali of Acinetabacter calcaaceticus genes for
benzoate degradation, 1. Bacterial., 169:5496.
Neidle, E.L., Hartnett, C., Bonitz, S., and Ornston, L.N., 1988, DNA
sequence of the Acinetabacter calcaaceticus catechol 1,2-
dioxygenase I structural gene catA: evidence for evolutionary
sequence repetition, i. Bacterial., 170:4874.
Neilson, A.H., Allard, A.S., Reilands, S., Remberger, M., Taernom, A.,
Victor, T., and Landner, L., 1984, Tri- and tetrachloroveratrole,
metabolites produced by bacterial O-methylation of tri- and
tetrachloroguaiacol: an assessment of their bioconcentration
386
potential and their effects on fish reproduction, Can. 1. Fisher.
Aquat. Sci., 41:1502.
Neilson, A.H., Lindgren, c., Hynning, P.A., and Remberger, M., 1988,
Methylation of halogenated phenols and thiophenols by cell
extracts of gram-positive and gram-negative bacteria, Appl.
Environ. Microbio!., 54:524.
Oms ton, L.N., and Parke, D., 1977, The evolution of induction mechanisms
in bacteria: insights derived from the study of the B-ketoadipate
pathway, Curro Top. Cell. Reg., 12:210.
Omston, L.N., and Stanier, R. Y., 1964, Mechanism of B-ketoadipate
formation by bacteria, Nature, 204: 1279.
Omston, L.N., and Yeh, W.K., 1979, Origin of metabolic diversity:
evolutionary divergence by sequence repetition, Proc. N atl.
Acad. Sci. USA" 76:3996.
Ornston, L.N., and Yeh, W.K., 1981, Toward molecular natural
history, Microbiology, 140-143.
Patel, R.N., and Omston, L.N., 1976, Immunological comparison of enzymes
of the B-ketoadipate pathway, Arch. M i crob i 01., 110:27.
Patel, R.N., Meagher, R.B., and Omston, L.N., 1974, Relationships among
enzymes of the B-ketoadipate pathway. IV. Muconolactone
isomerase from Acinetobacter calcoaceticus and
Pseudomonas putida, 1. Bioi. Chern., 249:7410.
Patel, R.N., Mazumdar, S., and Omston, L.N., 1975, B-Ketoadipate enol-
lactone hydrolases I and II from Acinetobacter calcoaceticus,
1. BioI. Chern., 250:6567.
Patel, R.N., Hou, C.T., Felix, A., and Lillard, M.O., 1976, Catechol 1,2-
dioxygenase from Acinetobacter calcoaceticus: purification
and properties, 1. Bacteriol., 127:536.
Pendrys, J.P., 1989, Biodegradation of asphalt cement-20 by aerobic
bacteria, Appl. Environ. Microbiol., 55:1357.
Reber, H.H., 1982, Inducibility of benzoate oxidizing cell activities in
Acinetobacter calcoaceticus strain Bs5 by chlorobenzoates as
influenced by the position of chlorine atoms and the inducer
composition, Eur. 1. Appl. Microbiol. Biotech., 15:138.
Reber, H.H., and Thierbach, G., 1980, Physiological studies on the oxidation
of 3-chlorobenzoate by Acinetobacter calcoaceticus strain
Bs5, Eur. 1. Appl. Microbiol., 10:223.
Reineke, W., and Knackmuss, H.-J., 1988, Microbial degradation of
haloaromatics, Ann. Rev. Microbiol., 42:263.
Reiner, A.M., 1971, Metabolism of benzoic acid by bacteria: 3,5-
cyclohexadierie-l,2-diol-l-carboxylic acid is an intermediate in the
formation of catechol, 1. Bacterio!., 108:89.
Rochkind, M.L., Blackbum, J.W., and Sayler, G.S., 1986, "Microbial
Decomposition of Aromatic Compounds," United States
Environmental Protection Agency, Cincinnati.
Ross, C.M., Kaplan, J.B., and Winkler, M.E., 1990, An evolutionary
comparison of Acinetobacter calcoaceti'cus trpF with trpF
genes of several organisms, Mol. BioI. Evol., 7:74.
387
Sato, K., Kato, H., and Furusak, c., 1987, Comparative study of soil bacteriai
flora as influenced by the application of a pesticide,
pentachlorophenol (PCP), Plant Soil, 100:333.
Sawula, R.V., and Crawford, l.P., 1972, Mapping of the tryptophan genes of
Acinetobacter calcoaceticus by transformation, 1.
Sawula, R.V. and Crawford, I.P., 1973, Anthranilate synthetase of
Acinetobacter calcoaceticus. Separation and partial
characterisation of subunits,.I. BioI. Chem., 248:3573.
Seltzer, S., 1973, Purification and properties of maleylacetone cis-trans
isomerase from Vibrio 01,.1. BioI. Chem., 248:215.
Shanley, M,S., Neidle, E.L., Parales, R.E., and Ornston, L.N., 1986, Cloning
and expression of Acinetobacter calcoaceticus cat BCDE
genes in Pseudomonas putida and Escherichia coli, .I.
Shanley, M.S., Neidle, E.L., Parales, R.E., Harrison, A., and Ornston, L.N.,
1989, Organization of catabolic pathway genes in Acinetobacter
cal coacet i cus, SAAS Bull. Bio c he m. Bi 0 tec h .., 2: 1.
Shields, M.S., Hooper, S.W., and Sayler, G.S., 1985, Plasmid-mediated
mineralization of 4-chlorobiphenyl,.I. Bacteriol., 163:882.
Sparnins, V.L., Chapman, P.l., and Dagley, S., 1974, Bacterial degradation of
4-hydroxyphenylacetic acid and homoprotocatechuic acid, 1.
Stanier, R.Y., and Omston, L.N., 1973, The B-ketoadipate pathway, Adv.
M icrob. Physio/., 9:89.
Suzuki, M., and Fujii, T., 1980, Biological treatment of phenolic waste water,
Seisan Kenkyu, 32:110.
Suzuki, M., and Fujii, T., 1982, Biodegradation rate of aromatic compounds
in wastewater treatment, S eisan K enkyu, 34: 156.
Suzuki, M., and Fujii, '1'., 191\5, Microbiological treatment of phenol
wastewater. Effect of coexisting cyanide and thiocyanate ion on
decomposition rate, S eisan K enkyu, 37:312.
Sze, I.S.-Y., and Dagley, S., 1987, Degradation of substituted mandelic acids
by meta fission reactions,.I. Bacteriol., 169:3833.
Taylor, B.F., Curry, R.W., and Corcoran, E.F., 1981, Potential for
biodegradation of phthalic acid esters in marine regions, Appl.
Tempest, D.W., 1978, The biochemical significance of microbial growth
yields: a reassessment, Trends Biochem. Sci., 3: 180.
Tibbles, B.J., and Baecker, A.A.W., 1989, Effect of pH and inoculum size on
phenol degradation by bacteria isolated from landfill waste,
Environ. Poll., 59:227.
Tresguerres, M.E.F., de Torrontegui. G., and Canovas, J .L., 1970a,
Metabolism of quinate by Acinetobacter calcoaceticus, Arch.
Mikrobiol.,70:110.
Tresguerres, M.E.F., de Torrontegui, G., Ingledew, W.M., and Canovas, l.L.,
1970b, Regulation of the enzymes of the B-ketoadipate pathway in
M oraxella. Control of quinate oxidation by protocatechuate,
Eur . .I. Biochem., 14:445.
388
Tresguerres, M.E.F., Ingledew, W.M., and Canovas, J.L., 1972, Potential
competition for 5-dehydroshikimate between the aromatic
biosynthetic route and the catabolic hydroaromatic pathway,
Arch. Mikrobiol., 82:11l.
Vakeria, D., Vivian, A., and Fewson, C.A., 1984, Isolation, characterization
and mapping of mandelate pathway mutants of Acinetobacter
calcoaceticus, 1. Gen. Microbiol., 130:2893.
Vasudevan, N., and Mahadevan, A., 1988, Degradation of p-benzyloxyphenol
by Acinetobacter sp., FEM S Microbiol. Lett., 56:349.
Vasudevan, N., and Mahadevan, A., 1989, Degradation of 3(0-
methoxyphenoxy)1,2-propanediol (guaiacol glyceryl ester) by
Acinetobacter sp., 1. Biotech., 9:107.
Vento sa, A., Quesada, E., Ferrer, M., Ruiz-Berraquero, F.L., and Ramos-
Cormenzana, A., 1984, Degradation of o-anisate by soil
microorganisms, Pol. J. Soil Sci., 15:37.
Walker, J.D., Colwell, R.R., and Petrakis, L., 1975, Evaluation of petroleum-
degrading potential of bacteria from water and sediment, Appl.
Walker, J.D., Colwell, R.R., and Petrakis, L., 1976, Biodegradation of
petroleum by Chesapeake Bay sediment bacteria, Can. J.
Microbiol., 22:423.
Wheelis, M.L., 1972, Regulation of the synthesis of enzymes of tryptophan
dissimilation in Acinetobacter calcoaceticus, Arch.
Mikrobiol., 87:1.
Williams, R.J., and Evans, W.c., 1975, The metabolism of benzoate by
M oraxell a species through anaerobic nitrate respiration.
Evidence for a reductive pathway, Biochem . .T., 148:1.
metabolism by the B-ketoadipate pathway, Molec. Microbiol.,
1:219.
Wyndham, R.C., 1986, Evolved aniline catabolism in Acinetobacter
calcoaceticus during continuous culture of river water, Appl.
Environ. M icrobiol., 51:781.
Yall, I., 1979, A petroleum-eating bacterium, Nat. H ist., 88:41.
Yeh, W.-K., and Ornston, L.N., 1980, Origins of metabolic diversity:
substitution of homologous sequences into genes for enzymes with
different catalytic activities, Proc. N atl. Acad. Sci. USA,
77:5365.
Yeh, W.-K., and Ornston, L.N., 1981, Evolutionary homologues Cl<2B2
oligomeric structures in B-ketoadipate succinyl-CoA transferases
from Acinetobacter calcoaceticus and Pseudomonas putida,
.T. BioI. Chem., 256:1565.
Yeh, W.-K., and Omston, L.N., 1984, p-Chloromercuribenzoate specifically
modifies thiols associated with the active sites of B-ketoadipate
enol-lactone hydrolase and succinyl-CoA:B-ketoadipate-CoA
transferase, Arch. Microbiol., 138:102.
Yeh, W.-K., Davis, G., Fletcher, P., and Omston, L.N., 1978, Homologous
amino acid sequences in enzymes mediating sequential metabolic
reactions, I. Bioi. Chem., 253:4920.
389
Yeh, W.-K., Fletcher, P., and Ornston, L.N., 1980a, Evolutionary divergence
of co-selected B-ketoadipate enol-lactone hydrolases in
Acinetobacter calcoaceticus, f. BioI. Chern., 255:6342.
Yeh, W.-K., Fletcher, P., and Ornston, L.N., 1980b Homologies in the NH2 -
terminal amino acid sequences of ~ -carboxymuconolactone
decarboxylases and muconolactone isomerases, f. BioI. Chern.,
255:6347.
Yeh, W.-K., Durham, D.R., Fletcher, P., and Ornston, L.N., 1981,
Evolutionary relationships among 'if -carboxymuconolactone
decarboxylases, f. Bacteriol., 146:233.
Yeh, W.-K., Shih, c., and Ornston, L.N., 1982, Overlapping evolutionary
affinities revealed by comparison of amino acid compositions,
Proc. Natl. Acad. Sci. USA,79:3794.
Zaborsky, O.R., and Schwartz, R.D., 1974, The effect of imidoesters on the
protocatechuate 3,4-dioxygenase activity of A c in e to b a c te r
calcoaceticus, FEBS Lett., 46:236.
Zaitsev, G.M., 1988, Utilization of 3-chlorobenzoic acid by a mixed culture
of microorganisms, Mikrobiologiya, 57:550.
Zaitsev, a.M., and Baskunov, B.P., 1985, Utilization of 3-chlorobenzoic
acid by Acinetobacter calcoaceticus, Mikrobiologiya,
54:203.
390
METABOLISM OF CYCLOHEXANE AND RELATED
ALICYCLIC COMPOUNDS BY ACINETOBACTER
P.W. Trudgill
University College of Wales
Aberystwyth
Dyfed SY23 3DD
UK
INTRODUCTION
Acinetobacter strains deposited in culture collections have been

variously described as species of Vibrio, Alcaligenes, Achromobacter
and Bacterium. One of these (NCIB 8250) deposited in the National
Collection of Industrial Bacteria, Aberdeen, Scotland, as Vi b ri 0 01 in 1950
by Professor W.e. Evans, has provided those of my colleagues who have been
interested in the degradation of aromatic compounds with a topic of interest
for decades (Fewson, this volume). The involvement of Acinetobacter spp.
in the degradation of alicyclic compounds was preceded by a period of
gradual development of our understanding of the metabolism of these
compounds. Early attempts to isolate organisms capable of growth with
simple alicyclic hydrocarbons such as cyclohexane were unsuccessful (Pelz
and Rehm, 1971; Beam and Perry, 1973, 1974; De Klerk and van der Linden,
1974). Although reports of organisms capable of growth with cyclohexane
have appeared in the literature from time to time (Tausz and Peter, 1919;
Johnson et aI., 1942; Skarzynski and Czekalowski, 1946; Fredericks, 1966),
none has been followed by significant metabolic study.
Commensal and co-oxidative systems have been reported in which the

degradation of cycloalkanes has been initiated. Ooyama and Foster (1965)
made use of an alkane-degrading strain of Mycobacterium vaccae (JOB 5)
to convert cyclohexane into cyclohexanol and cyclohexanone. Beam and
Perry (1974) and de Klerk and van der Linden (1974) demonstrated that
inclusion of cyclohexanol-utilising strains in mixed cultures with organisms
capable of the gratuitous hydroxylation of cyclohexane resulted in complete
degradation of the cyclic hydrocarbon.
391
More recently, two groups of workers (Stirling et aI., 1977; Magor et
aI., 1986) have clearly established growth of Nocardia, Pseudomonas and
Xanthobacter spp. on cyclohexane as sole source of carbon. Hydroxylation
of the compound to form cyclohexanol is the initial step in oxidation
(Anderson et aI., 1980; Trower et aI., 1985). In addition, Kamata et al.
(1986) have recently isolated a strain, which they named Acinetobacter
cyclohexanophilus, that grows rapidly with cyclohexane as sole source of
carbon, although no pathway studies have been reported.
Our knowledge of the only clearly established pathway for

cyclohexanol oxidation by bacteria was first obtained with Nocardia and
Acinetobacter spp. isolated by enrichment with cyclohexanol and capable
of growth on the compound as sole carbon source.
CYCLOHEXANOL METABOLISM BY
ACI N ETOBACTER SP. NCIMB 9871
Acinetobacter sp. NCIMB 9871, used in our studies of cyclohexanol

metabolism, was a generous gift from Dr. P.J. Chapman who isolated it by
elective culture with the alcohol.
It is self-evident that energy and biosynthetic precursor provision from

cyclohexanol must involve a ring cleavage step. Cyclohexanol-grown cells
oxidised the growth substrate and cyclohexanone rapidly, and extracts of
cyclohexanol-grown Acinetobacter sp. NCIMB 9871 contained an inducible
cyclohexanol dehydrogenase, providing evidence that cyclohexanone is an
intermediate in cyclohexanol degradation (Donoghue and Trudgill, 1976).
The importance of cyclohexanol dehydrogenase in cyclohexanol degradation
has recently been reinforced by the work of Park et al. (1986) in their study of
induction of the enzyme in A. calcoaceticus strain C10 during growth of the
organism on cyclohexanol.
Fig. 1. Mechanism of the chemical Baeyer-Villiger reaction: (a) protonated

cyclohexanone; (b) peroxyacid reagent; (c) protonated € -caprolactone. 1,
nucleophilic attack by the peroxyacid; 2, deprotonation; 3, loss of the
leaving group; 4, product formation.
392
The Baeyer- Villiger Reaction
A chemical approach to the problem of ring cleavage of cyclic ketones

was first described by Baeyer and Villiger (1899). These authors showed that
peroxyacid reagents were able to insert oxygen into cyclic ketones with
consequent ring expansion and lactone formation. The accepted mechanism
for this reaction is shown in Fig. 1. The reaction is initiated by protonation
of the carbonyl group, followed by nucleophilic attack by the peroxy acid.
The intermediate resulting from addition of the peroxyacid to the carbonyl
has a good leaving group, a carboxylate anion on an oxygen atom, since the
oxygen-oxygen bond is weak (approx. 197 kllmo]) and not difficult to break.
As the leaving group departs, a partial positive character develops at the
oxygen and a 1-2 alkyl shift from a carbon atom to an adjacent oxygen atom
takes place, driven in part by the energy recovery of carbonyl bond
reformation.
The Biological Bayer-Villiger Reaction
Ample evidence now exists to show that Acinetobacter sp. NelMB

9871 makes use of a bio] ogical parallel to the chemical Baeyer- Villiger
reaction in cyclohexanol oxidation. Whole cell oxidation studies, subcellular
enzymology and isolation of catabolic intermediates all served to establish
the partial catabolic pathway shown in Fig. 2. Manipulation of soluble
protein fractions in the presence of the esterase inhibitor paraoxon (diethyl-
p-nitrophenylphosphate) confirmed that the €-caprolactone arose directly
from cyclohexanone and not by ring-closure of 6-hydroxyhexanoate
(Donoghue and Trudgill, 1976). All assayed enzymes were induced by growth
with cyclohexanol. Subsequent isolation of the enzyme responsible for
- 02- o
OH
o
6
~-OXIDATION -
Fig. 2. Oxidation of cyclohexanone by bacteria. Compounds: (a)

cyclohexane; (b) cyclohexanol; (c) cyclohexanone; (d) E-caprolactone; (e)
6-hydroxyhexanoate; (f) adipate. Enzymes: 1, cyclohexane monooxygenase;
2, cyclohexanol dehydrogenase; 3, cyclohexanone 1,2-monooxygenase; 4, E-
caprolactone hydrolase; 5, 6-hydroxyhexanoate and 6-oxohexanoate
dehydrogenases; 6, CoA ester synthetase leading to B-oxidation.
393
formation of thef-caprolactone from cyclohexanone showed it to be a simple
flavoprotein (M,59,OOO; 1 FAD) that constituted about 3% of the soluble
protein of the induced cell, required NADPH as an electron donor, and
molecular oxygen. It had a reaction stoichiometry that was consistent with it
being a monooxygenase (Donoghue et al., 1976). The correct systematic
name for the enzyme is cyclohexanone NADPH:oxygen oxidoreductase (1,2-
lactonising) (EC 1.14.13.-), hereafter called cyclohexanone 1,2-
monooxygenase.
Biological Baeyer-Villiger monooxygenases are widely distributed and

are also involved in the cleavage of cyclopentanone (Griffin and Trudgill,
1972), both rings of camphor (Conrad et' al., 1965a,b; Ougham et al., 1983;
Taylor and Trudgill, 1986) and the D-ring of androstene-3,17-dione (Prairie
and Talaley, 1963). Removal of the side chains of progesterone (Rahim and
Sih, 1966) and acetophenone (Cripps, 1975), and conversion of 2-
tridecanone to undecyl acetate by Pseudomonas cepacia (Britton and
Markovetz, 1977) are also catalysed by this class of enzyme. In all cases
where significant characterisation has been undertaken, the enzymes are
flavoproteins. Within this group of enzymes the cyclohexanone 1,2-
monooxygenase from Acinetobacter sp.NCIMB 9871 has a dominant
position. Its stability and ease of purification has made it the logical choice
for thorough and extensive studies of the reaction mechanism.
Reaction Mechanism of Cyclohexanone 1, 2-M onooxygenase
Using an experimental approach that included anaerobic rapid

reaction stopped-flow spectrophotometry, Ryerson et al. (1982) were able to
establish a ter-ter mechanism for the reaction, as shown in Fig. 3, which is
compatible with the experimental observation that NADPH and
cyclohexanone bind to different forms of the enzyme. This is in accord with
the failure of cyclohe:tanone to influence the rate of enzyme-bound FAD
reduction by NADPH. An unusual feature of the enzyme was the very low Krn
for oxygen (approx. 10-15 JLM) which made it impossible to determine the
order of addition of oxygen with any certainty. No stable flavin-oxygen
intermediate could be detected in the absence of cyclohexanone and this
o .
E_ f E. NADPH. • [EH2.NADP~~.0
O2 ] _ _ _ ] _ E. NADP+ _ E
t EH2 .NADP; O2 E. S~O.NADP. Hp
t ~
Q
NADPH
NADP'
O2 ?
o o
Fig. 3. Reaction sequence for the oxygenation of cyclohexanone by the 1,2-

monooxygenase from Acinetobacter sp. NCIMB 9871 (after Ryerson et al.,
1982) .
394
s6:~ H/ CO
H
0
cyclohexanone
b
Fig. 4. Proposed mechanism for the biological Baeyer-Villier oxygenation of
by the 1,2 -monooxygenase from Acinetobacter sp. NCIMB
9871. The reactive oxygen donor is 4a-hydroperoxy FAD (after Ryerson et
al., 1982).
observation, plus the fact that the flavin-dependent phenolic

monooxygenases all add organic substrate before oxygen, suggests that
oxygen adds to the reduced enzyme-NADP complex. Stopped-flow studies, in
which the reduced flavoprotein was reacted with oxygen in the presence of
NADP, resulted in the formation of an intermediate, proposed from its
electronic spectrum to be 4a-hydroperoxyflavin (Entsch et aI., 1976), within
the dead time (2-3 ms) of the instrument. In the presence of cyclohexanone
an additional intermediate, 4a-hydroxyflavin, was also observed.
Reaction of 4a-hydroperoxyflavin with cyclohexanone could

theoretically involve cyclohexanone acting as an electrophile at the carbonyl
carbon or as a nucleophile through the intermediacy of an enolate tautomer.
Studies with deuterated cyclohexanone strongly support an electrophilic role
for the cyclohexanone, as shown in Fig. 4, that provides a clear parallel to the
peroxy acid mediated Baeyer-Villiger reaction. Schwab (1981), using
deuterium NMR, established that the configuration of the migrating C
centre was completely retained, further sustaining the analogy between the
microbial and chemical versions of the reaction.
The cyclohexanone 1,2-monooxygenase of Acinetobacter sp. NCIMB

9871 displays a characteristic unique among flavoprotein monooxygenases in
having the ability to deliver an oxygen atom, derived from 02' to both
nucleophilic and electrophilic substrates. The cyclic sulphide thiane yields a
stable sulphoxide (thiane-1-oxide), demonstrating that the nature of the
peroxide oxygen-oxygen bond cleavage is controlled by electron availability
in the compound to be oxygenated (Fig. 5).
The catabolic flexibility of the enzyme is further illustrated by its

ability to oxygenate aldehydes (phenylacetone), selenides, boronic acids, a
phosphite ester and an iodide ion (Branchaud and Walsh, 1985). Abril et al.
(1989) have established a broad ketone specificity for the enzyme using
395
R
o +.O .... H-4· S
·v
0
.
'c~··
I _
:0:
.. V
H-S+
V
2 3 4
Fig. 5. Electrophilic and nucleophilic oxygenation reactions catalysed by

cyclohexanone 1,2-monooxygenase from Acinetobacter sp. NCIMB 9871:
A. Cyclohexanone acting as an electrophile (carbonyl carbon); B. Thiane
acting as a nucleophile with the formation of thiane l-oxide.
immobilised enzyme and regenerating the NADPH in situ with glucose 6-

phosphate and glucose-6-phosphate dehydrogenase from Leuconostoc
mesenteroides.
Gene Sequencing and Structure of

eycl ohexanone-l, 2-M ono oxygenase
The gene coding for cyclohexanone 1,2-monooxygenase has been

located, isolated and its nucleotide sequence determined (Chen et aI., 1988).
The complete sequence of 542 amino acids has been derived by translation of
the nucleotide sequence. Analysis of the amino acid sequence showed seven
cysteine residues, of which four may be involved in disulphide bridge
formation, leaving three available for titration as experimentally determined
(Donoghue et al., 1976; Latham and Walsh, 1987). A binding site for FAD
has been identified as a .Ba.B-unit characterised by a pair and a triplet of
glycine residues. Typically these regions are near the NH2 terminus of the
polypeptide chain (Wierenga et aI., 1983), and cyclohexanone 1,2-
monooxygenase is no exception as the region is encompassed by residues 6 to
18. In addition, a potential ADP (NADPH) binding site, based upon a
proposed fingerprint of 11 amino acids (Wierenga et aI., 1986), has been
identified in the region extending from residue 176 to residue 208.
An over-production system for the gene was constructed by using the

tac promoter vector pKK223-3 in Escherichia coli. Although the amount
of pure enzyme available from this strain is not significantly greater than that
obtained from the same mass of Acinetobacter sp. NCIB 9871, the system
provides a means for performing genetic manipulations on the plasmid
(pKKCO) and creating mutant forms of the enzyme (Chen et aI., 1988).
Cyclohexanone 1,2-monooxygenase from Acinetobacter sp. NCIMB

9871 is the only enzyme of this class for which the nucleotide and, by
396
inference, the amino acid sequence has been determined. The potential now
exists for an interplay of approaches, including X-ray crystallographic
methods, that will provide a complete understanding of biological Baeyer-
Villiger oxygenation.
A novel exploitation of the biological Baeyer-Vi1liger reaction has

been reported by Alphand et ai. (1990). These authors made use of whole
cell cultures of two strains of Acinetobacter in the conversion of 2-
undecylcyclopentanone in the synthesis of (S)-( -)-5-hexadecanolide (a wasp
pheromone ).
LACTONE HYDROLYSIS
The interest in enzymes that incorporate molecular oxygen into

organic molecules has spanned several decades, a variety of biological source
material, and a wide range of cyclic and acyclic compounds. In the case of
alicyclic ring cleavage, ring oxygen insertion with the consequent formation
of a lactone is not, of itself, a ring cleavage mechanism. The E-caprolactone
formed from cyclohexanone, for example, is stable at near neutral pH, and
ring cleavage is catalysed by an inducible lactone hydrolase (EC 3.1.1.-). The
lactone hydrolase from cyclohexanol-grown Acinetobacter sp. NCIMB 9871
has been purified and characterised (Bennett et aI., 1988). It constitutes
about 1 % of the soluble protein of induced cells, has an M, of about 60,000,
and consists of two electrophoretically-identical subunits. Substrate
specificity is narrow (5% activity with 5-valerolactone; no activity with
}(-lactones) and, of a wide range of putative inhibitors, only paraoxon (diethyl
p-nitrophenylphosphate) was effective, giving 50% inhibition at 8 JLM. In this
respect it most closely resembles mammalian acetylcholine esterase (Froede
and Wilson, 1971) and carboxyesterase (Heymann et aI., 1970), both of which
have a functional catalytic centre serine residue, and it contrasts sharply with
other bacterial 5-lactone hydrolases that are inhibited by Co 2 +, Cu 2 +, Mn 2 +or
Zn 2 + ions and thiol-reactive reagents, and are thus considered to have
catalytically active thiol groups (Kersten et aI., 1982; Maruyama, 1983), and
with the Ca 2 +-dependent 'l{-lactone hydrolases of human blood and rat liver
microsomes (Fishbein and Bessman, 1966a,b).
SPONTANEOUS CLEAVAGE OF LACTONES
The formation of stable lactones as a consequence of oxygenation,

leading to ring expansion of simple cyclic ketones such as cyclohexanone and
cyclopentanone, dictates the requirement for an esterase if ring cleavage is
to proceed. Some ketonic substrates, however, yield unstable lactones upon
ring oxygenation. Acinetobacter sp. NCIMB 9871 and N. globerula CLI
are also capable of growth with trans-cyclohexan-l,2-dioi. This is initially
dehydrogenated to 2-hydroxycyclohexanone, with subsequent insertion of a
ring oxygen between the two substituent groups forming an unstable lactone
that spontaneously ring-opens to form 6-oxohexanoate. Although the first
397
two steps are catalysed by enzymes that are common to cyclohexanol
degradation, the induced lactone hydrolase is redundant (Donoghue and
Trudgill, 1973). In contrast, Acinetobacter sp. strain 6.3, which was
isolated by elective culture on trans-cyc1ohexan-1,2-diol, is incapable of
growth with cyc1ohexanol, cyclohexanone or c-caprolactone, although the
induced secondary alcohol dehydrogenase and Baeyer- Villiger oxygenase are
of broad specificity and active towards trans-cyclohexan-1,2-
diol! cyc1ohexanol and 2-hydroxycyc1ohexanonel cyc1ohexanone respectively.
One possible explanation of this very restricted growth spectrum is that the
organism does not synthesise a lactone hydrolase. Certainly no hydrolase
activity towards E-caprolactone can be detected in extracts of trans-
cyclohexan-1,2-diol-grown cells (Davey and Trudgill, 1977).
CYCLOHEXANECARBOXYLIC ACID METABOLISM
The presence of a carboxyl group on the cyclohexane ring has a

profound effect upon catabolic metabolism and, to date, no organism has
been isolated that utilises a biological Baeyer-Villiger reaction in the
oxidation of the compound to central metabolites. Organisms that grow with
the compound as sole carbon source use one of two catabolic routes.
Hydroxylation at the 4-position, followed by oxidation to 4-
oxocyclohexanecarboxylic acid and aromatisation to p-hydroxybenzoate,
leads to protocatechuic acid. Both ortho and meta fission has been
variously demonstrated for strains of Arthrobacter, Corynebacterium
and Pseudomonas (Blakley, 1974; Kaneda, 1974; Smith and Callely, 1975;
Taylor and Trudgill, 1978). Alternatively, because the carboxyl group is
attached to a ring carbon, the molecule is also a potential substrate for ring
cleavage by B-oxidation.
Experimental evidence for this simple direct catabolic route was first
obtained by Rho and Evans (1975) who, using a strain of Acinetobacter
anitratum isolated by enrichment culture on cyclohexane carboxylic acid,
demonstrated that pimelate was a metabolite and that ATP, CoA, Mg2+ and
FAD were required for its formation by cell extracts. Blakley (1978),
working with a strain of Alcaligenes, has confirmed the aerobic B-oxidation
of cyclohexane carboxylic acid to pimelate and demonstrated inducible
enzymes, including a cyclohexane carboxyl-CoA synthetase, in extracts of
induced cells.
CONCLUSIONS
This brief survey illustrates the key role played by Acinetobacter

spp. in the development of our understanding of the pathways and
enzymology of the degradation of simple alicyclic compounds. Although
important discoveries have also been made with strains of Arthrobacter,
Corynebacterium, Nocardia and Pseudomonas spp., the unique
contribution of Acinetobacter spp. to our understanding of biological
398
Baeyer- Villiger oxygenation is well documented. In particular,
Acinetobacter sp. NCIMB 9871 was chosen for detailed study because of its
rapid growth with cyclohexanol and production of large amounts of a stable
oxygenase that is easy to purify.
REFERENCES
Abril, 0., Ryerson, CC, Walsh CT., and Whitesides, G.M., 1989, Enzymic
Baeyer- Villiger type oxidations of ketones catalysed by
cyclohexanone oxygenase, Biorgan. Chern., 17:41.
Alphand,V., Archelas, A., and Furstoss, R., 1990, Microbiological
transformations. 13. Direct synthesis of both Sand R enantiomers
of 5-hexadecanolide via an enantioselective microbiological
Baeyer-Villiger reaction, 1. Organ. Chern., 55:347.
Anderson, M.S., Hall, R.A., and Griffin, M., 1980, Microbial metabolism of
alicyclic hydrocarbons: cyclohexane catabolism by a pure strain of
Pseudomonas sp., 1. Gen. Microbiol., 120:89.
Baeyer, A., and Villiger, V., 1899, Einwirkung des caro'schen Reagens auf
Ketone, Berich. Deuts. Chern. Gesellschaft, 32:3625.
Beam, H.W., and Perry, J.J., 1973, Co-metabolism as a factor in microbial
degradation of cycloparaffinic hydrocarbons, Arch. Microbiol.,
91:87.
Beam, H.W., and Perry, J.J., 1974, Microbial degradation and assimilation of
n-alkyl-substituted cycloparaffins, 1. Bacteriol., 118:394.
Bennet, A.P., Strang, E.J., Trudgill, P.W., and Wong, V.T.K., 1988,
Purification and properties of E-caprolactone hydrolase from
Acinetobacter NCIB 9871 and Nocardia globerula CL1, 1.
Gen. Microbial., 134:161.
Blakley, E.R., 1974, The microbial degradation of cyclohexanecarboxylic
acid: a pathway involving aromatization to form p-hydroxybenzoic
acid, Can. 1. Microbiol., 20:1297.
Blakley, E.R., 1978, The microbial degradation of cyclohexanecarboxylic
acid by a B-oxidation pathway with simultaneous induction to the
utilization of benzoate, Can. 1. Microbiol., 24:847
Branchaud, B.P., and Walsh, CT., 1985, Functional group diversity in
enzymatic oxygenation reactions catalysed by bacterial flavin-
containing cyclohexanone oxygenase, 1. Arner. Chern. Soc.,
107:2153.
Britton, L.N., and Markovetz, A.J., 1977, A novel ketone monooxygenase
from Pseudomonas cepacia. Purification and properties, 1.
Bioi. Chern., 252:8561.
Chen, J. Y-C, Peoples, O.P., and Walsh, CT., 1988, Acinetobacter
cyclohexanone monooxygenase; gene cloning and sequence
determination, 1. Bacteriol., 170:781.
Conrad, H.E., Dubus, R., Namtvedt, MJ., and Gunsalus, I.C, 1965a, Mixed
function oxygenation II. Separation and properties of enzymes
catalysing camphor lactonization, 1. Bioi. Chern., 240:495.
399
Conrad, H.E .. Lieb, K., and Gunsalus, I.e. 1965b, Mixed function oxidation
III. An electron transport complex in camphor ketolactonization,
J. Biol. Chem., 240:4029.
Cripps, R.E., 1975, The microbial metabolism of acetophenone. Metabolism
of acetophenone and some chloroacetophenones by an
Arthrobacter species, Biochem. 1., 152:233.
Davey, J.F., and Trudgill, P.W., 1977, The metabolism of trans-cyclohexan-
1,2-diol by an Acinetobacer species, Eur. 1. Biochem., 74:115.
De Klerk, H., and Van Der Linden, A.e., 1974, Bacterial degradation of
cyclohexane. Participation of co-oxidation reaction, Ant. v.
Leeuw. 1. Microbiol. Serol., 40:7.
Donoghue, N.A., and Trudgill, P.W., 1973, The metabolism of cyclohexane-
I,2-diols by Nocardia g/oberula CLl, Trans. Biochem. Soc.,
1:1287.
Donoghue, N .A., and Trudgill, P.W., 1976, The metabolism of cyclohexanol
by Acinetobacter NClB 9871, Eur . .1. Biochem., 60:1.
Donoghue, N.A., Norris, D.B., and Trudgill, P.W., 1976, The purification and
properties of cyc!ohexanone oxygenase from Nocardia
globerula CLI and Acinetobacter NCIB 9871, Eur. 1.
Biochem., 63:175.
Entsch, B., Ballou, D.P., and Massey, V., 1976, Flavin-oxygen derivatives
involved in hydroxylation by p-hydroxybenzoate hydroxylase, 1.
Bioi. Chem., 251:2550.
Fishbein, W.N., and Bessman, S.P., 1966a, Purification and properties of an
enzyme in human blood and rat liver microsomes catalysing the
formation and hydrolysis of )s"-lactones. 1. Tissue localization,
stoichiometry, specificity, distinct from esterase, .I. BioI. Chem.,
241:4835.
Fishbein, W.N., and Bessman, S.P., 19Mb, Purification and properties of an
enzyme from human blood and rat liver microsomes catalysing the
formation and hydrolysis of ls'-lactones. II. Metal ion effects,
kinetics and equilibria, .I. Bi 0 t. C hem., 241 :4842.
Fredericks, K.M., 1966, Adaptation of bacteria from one type of hydrocarbon
to another, Nature, 209: 1047.
Froede, H.C., and Wilson, I.B., 1971, Acetylcholine esterase, in "The
Enzymes, vol. 5," p.87, P.D. Boyer, ed., Academic Press, New York.
Griffin, M., and Trudgill, P.W., 1972, The metabolism of cyclopentanol by
Pseudomonas NCIB 9872, Biochem . .I., 129:595.
Heymann, E., Krisch, K., and Pahlich, E., 1970, Structure of the active centre
of microsomal carboxyesterase from pig kidney and liver, H oppe-
Seyler'S Zeits. Physiol. Chem., 351:931.
Johnson, F.H., Goodale, W.T., and Turkevich, J., 1942, The bacterial
oxidation of hydrocarbons, 1. Cell. Compo Physiol., 19:163.
Kamata, A., Mikawa, T., and Imada, Y., 1986, Bacterial degradation of
cyclohexane, Nippon Kokai Tokkyo Koho, 61128890.
Kaneda, T., 1974, Enzymatic aromatization of 4-ketocycohexanecarboxylic
acid to p-hydroxybenzoic acid, Biochem. Biophys. Res.
Comm.,58:140.
400
Kersten, P.J., Dagley, S., Whittaker, J.W., Arciero, D.M., and Lipscomb,
J.D., 1982, 2-Pyrone-4,6-dicarboxylic acid, a catabolite of gallic
acid in Pseudomonas species, 1. Bacteriol., 152:1154.
Latham, J.A., and Walsh, e.T., 1987, Mechanism-based inactivation of the
flavoenzyme cyclohexanone oxygenase during oxygenation of cyclic
thiol ester substrates, J. Amer. Chem. Soc., 109:3421.
Magor, A.M., Warburton, J., Trower, M.K., and Griffin, M., 1986,
Comparative study of the ability of three Xanthobacter species
to metabolize cycloalkanes, Appl. Environ. Microbiol. 52:665.
Maruyama, K.C., 1983, Purification and properties of 2-pyrone-4,6-
dicarboxylate hydrolase, I. Biochem., 93:557.
Ooyama, J., and Foster, J.W., 1965, Bacterial oxidation of cycloparaffinic
hydrocarbons, Ant. v. Leeu w. I. M i crobiol. S erol., 31:45.
Ougham, H.J., Taylor, D.G., and Trudgill, P.W., 1983, Camphor revisited:
involvement of a unique monooxygenase in the metabolism of 2-
oxo- 3-4,5,5-trimethylcyclopentenylacetic acid by Pseudomonas
putida, .I. Bacteriol., 153:140.
Park, H.D., Choi, S., and Rhee, l.K., 1986, Induction of cyclohexanol
dehydrogenase in Acinetobacter calcoaceticus ClO, Han'guk
N onghwa H akhoechi, 29:304.
Pelz, B.F., and. Rehm, H.J., 1971, lsolierung, Substrassimilation und einige
produkte alkenabbauender Schimmelpilze, Arch. M ikrobiol.,
84:20.
Prairie, R.L., and Talaley, P., 1963, Enzymatic formation of testololactone,
Rahim, M.A., and Sih, e.l., 1966, Mechanisms of steroid oxidation by
microorganisms. XI Enzymatic cleavage of the pregnone side
chain, I. BioI. Chem., 241:3615.
Rho, E.M., and Evans, W.C., 1975, The aerobic metabolism of
cyclohexanecarboxylic acid by Acinetobacter anitratum,
Biochem. I., 148:11.
Ryerson, e.e., Ballou, D.P., and Walsh, e.T., 1982, Mechanistic studies on
cyclohexanone oxygenase, Biochemistry, 21:2644.
Schwab, J.M., 1981, Stereochemistry of an enzymic Baeyer-Villiger reaction.
Application of deuterium NMR, 1. Amer. Chem. Soc., 103:1876.
Skarzynski, B., and Czekalowski, l. W., 1946, Utilization of phenols and
related compounds by Achromobacter, Nature, 158:304.
Smith, D.l., and Callely, A.G., 1975, The microbial degradation of
cyclohexanecarboxylic acid, I. Gen. Microbial., 91:210.
Stirling, L.A., Watkinson, R.J., and Higgins, I.J., 1977, Microbial
metabolism of alicyclic hydrocarbons: isolation and properties of a
cyclohexane degrading bacterium, I. Gen. Microbial., 99:119.
Tausz, J., and Peter, M., 1919, Neue Methode der Kohlenwasserstoffanalyse
mit Hilfe von Bakterien, Zent. Bact. Parasit. Infekt. Hyg.,
49:497.
Taylor, D.G., and Trudgill, P.W., 1978, Metabolism of cyclohexanecarboxylic
acid by Alcaligenes strain WI, J. Bacterial., 134:401.
Taylor, D.G., and Trudgill, P.W., 1986, Camphor revisited: studies of the 2,5-
diketocamphane 1,2-monooxygenase from Pseudomonas putida
ATCC 17453, I. Bacterial., 165:489.
401
Trower, M.K., Buckland, R.M., Higgins, R., and Griffin, M., 1985, Isolation
and characterization of a cyclohexane-metabolizing X antho-
bacter sp., Appl. Environ. Microbiol., 49:1282.
Wierenga, R.K., Drenth, J., and Schulz, G.E., 1983, Comparison of the three-
dimensional protein and nucleotide structure of the FAD-binding
domain of p-hydroxybenzoate hydroxylase with FAD as well as
NADPH-binding domains of glutathione reductase, 1. Mol. Bioi.,
167:725.
Wierenga, R.K., Terpstra, P., and HoI, W.G., 1986, Prediction of the
occurence of the ADP-binding Bod~-fold in proteins using an
amino acid sequence fingerprint, 1. Mol. Bioi., 187: 101.
402
METABOLISM OF TRIMETHYLAMMONIUM COMPOUNDS BY
ACINETOBACTER
H.-P. Kleber
Bereich Biochemie
Karl-Marx-Universitat Leipzig
Talstrasse 33
7010 Leipzig
Germany
INTRODUCTION
L( - )-Carnitine (R( - )-3-hydroxy-4-trimethylaminobutyrate) is a

ubiquitously-occurring substance which is essential for the .B-oxidation of
long-chain fatty acids in mitochondria. Bacteria are able to metabolise this
trimethylammonium compound in different ways. Species of the genus
Pseudomonas can utilise L-carnitine as sole source of carbon and nitrogen
under aerobic conditions. After growth on L( -)-carnitine, the highly specific
L(-)-carnitine dehydrogenase (EC 1.1.1.108) is induced in P. aeruginosa
(Aurich et aI., 1967) and P. putida (Kleber et aI., 1978). A second group of
carnitine-metabolising microorganisms compnsing different
Enterobacteriaceae, e.g., Escherichia coli, Salmonella typhimurium
and Proteus vulgaris, can almost quantitatively convert L(-)-carnitine into
'If-butyrobetaine under anaerobic conditions (Seim et aI., 1980, 1982a,c,d).
They do not assimilate the carbon and nitrogen skeleton. The metabolism of
L(-)-carnitine by E. coli includes at least a two-step reduction of L(-)-
carnitine to '){-butyrobetaine with crotonobetaine as intermediate (Seim et
aI., 1982d). The function of this reaction sequence is still unknown. Seim et
ai. (1982a,c,d) postulated that crotonobetaine serves as an external electron
acceptor similar to (e.g.) nitrate (Haddock and Jones, 1977). One of the two
enzymes involved, the carnitine dehydratase, was recently purified and
characterised (Jung et aI., 1989). Conversion of crotonobetaine to L(-)-
carnitine also seems to occur in different strains of Acinetobacter under
aerobic (Yokozeki and Kubota, 1984) and anaerobic (Kawamura et aI.,
403
Table 1. Utilisation of carnitine and structurally-related trimethyl-
ammonium compounds [N+(CH3)3-Rj as sole carbon source by A. calcoaceticus
69-V
Quarternary N-compound R Growth Product

(0.3%) formed by
cleavage
of C-N
bond
Glycine betaine - CH 2 -COO

B-Homobetaine - CHZ-CH 2 -COO-
'6-Butyrobetaine - CH 2 -CH 2 -CH 2 -COO- + trimethyl-
amine
Crotonobetaine - CH2-CH~CH-COO
D(+)-Carnitine - CH 2 -CH(OH)-CH 2 -COO-
L(-)-Carnitine - CH 2 -CH(OH)-CH 2 -COO- + trimethyl-
amine
Trimethylamine N-oxide - 0
Trimethylammonium - H
Choline - CH 2 -CH 20H
'6-Homocholine - CH 2 -CH 2 -CH 20H
D,L-B-Methylcholine - CH 2 -CH(CH 3 )OH
D,L-B-Ethylcholine - CH 2 -CH(CH 3 -CH 2 )OH
Acetonyltrimethylammonium - CH 2 -CO-CH 3
1986b) conditions. So far as the catabolism of quaternary nitrogen

compounds is concerned, under aerobic conditions Acinetobacter
calcoaceticus 69-V has a different pathway from that which occurs in the
other L( -)-carnitine-metabolising bacteria mentioned above.
SPLITTING OF THE C-N BOND
The utilisation of carnitine and structurally-related trimethyl-

ammonium compounds (betaines and nitrogen-bases) by A. calcoaceticus
69-Y is summarised in Table 1 (Kleber et aI., 1977). Only L(-)-carnitine and
){-butyrobetaine are used as the sole carbon sources. Growth on i-
butyrobetaine proceeded with a lag phase considerably longer than that
observed with L- or D,L-carnitine as sole sources of carbon (Kleber et aI.,
1977). L- O-Acylcarnitines were also assimilated by this strain of
Acinetobacter. The utilisation of these compounds and the growth of the
organism correlated with the stoichiometric formation of trimethylamine
and, presumably, with the cleavage of the C-N bond and the degradation of
the carbon backbone (Fig. 1). The bacteria oxidised choline to glycine
betaine in the presence of additional carbon sources, but glycine betaine
404
/ 1.5
I
60-l9---+--~h...
I
/
/ ..c" ······0
A /.:'
~ // LO
1 ·f J
.s;;
OJ
:§ 30
I! 1
~
0.5
/ j
; ....
15 /0:
--/
/
....
/
........
.. ~.~.~"., .....
4 8 12
h
Fig. 1. Degradation of L(-)-carnitine (1 gil) during growth (.--.) of A.
calcoaceticus 69-V. Disappearance of L(-)-carnitine (~, thin-
layer chromatography; ~,optical activity) and formation of
trimethylamine (TMA, 0 ..... --0) in the culture medium.
itself was not assimilated. D-Carnitine was metabolised by A.

calcoaceticus 69-V if an additional carbon source, such as L-carnitine, was
present in the incubation mixture or if the bacteria were preincubated with
L- or D,L-carnitine, but no growth was observed with D-carnitine as the sole
carbon source (Kleber et aI., 1977).
Miura-Fraboni et al. (1982) investigated the metabolic patterns of

utilisation of [1,2,3,4-14C, methyPH]'GI'-butyrobetaine and D- and L-[1-14C,
methyPH]carnitine with resting cell suspensions of A. calcoaceticus that
had been grown under various conditions. They found that all three betaines
were utilised at almost the same rate. Disappearance of each of these
quaternary ammonium compounds was accompanied by stoichiometric
formation of trimethylamine. The utilisation of the betaines and the
corresponding formation of trimethylamine by resting cell suspensions of A.
calcoaceticus was essentially abolished under conditions of anaerobiosis
and was severely impaired in the presence of sodium cyanide, sodium azide,
2,4-dinitrophenol or 2,2 1 -bipyridine (Miura-Fraboni et aI., 1982). The
results of these investigations with resting cell suspensions of A.
calcoaceticus do not support an earlier suggestion (Kleber et aI., 1977)
that lI'-butyrobetaine degradation in this organism proceeds by its prior
hydroxylation to L-carnitine. Indeed, cell-free preparations of A.
calcoaceticus grown on either D,L-carnitine or 'If-butyrobetaine show no
detectable a-ketoglutarate-linked )( -butyrobetaine hydroxylase activity
(Mirua-Fraboni and Englard, 1982; Miura-Fraboni et aI., 1982).
Furthermore, failure of A. calcoaceticus to grow on D-carnitine (Kleber et
aI., 1977) does not accord with the observations that resting cell suspensions
of organisms, grown on either D,L-carnitine or 'If-butyrobetaine, vigorously
405
Table 2. Specifity of the cleavage of the C-N bond of quaternary
ammonium compounds by the membrane fraction prepared from A.
calcoaceticus cells grown on different carbon sources (10 gjl). The
protein concentrations in membrane fractions of D,L-carnitine-, acetate-
and succinate-grown bacteria were 11.8, 13.6 and 11.2 mgjml
respectively; incubation time 2 h.
Quaternary ammonium Trimethylamine formation after growth on

compound
D,L-Carnitine Acetate Succinate
D,L-Carnitine +++ + +
L(-)-Carnitine +++ + +
D(+)-Carnitine +++ + +
Acetyl-L(-)-carnitine ++
ls'-Butyrobetaine +++
Glycine betaine
Acetonyltrimethyl-
ammonium
13-Methylcholine
Choline
Acetylcholine
+++, >10.0 nmoljmg protein; ++, 1.0-10.0 nmoljmg protein;

+, 01-1.0 nmoljmg protein; -, non-detectable (>0.1 nmoljmg protein).
metabolised D-carnitine with stoichiometric formation of trimethylamine

and complete degradation of the carbon skeleton (Miura-Fraboni et ai.,
1982). In addition, Miura-Fraboni and Englard (1983) showed that D-
carnitine effectively supported the growth of A. calcoaceticus, and that
utilisation of carnitine resulted in stoichiometric formation of
trimethylamine and equivalent loss of the [I4C]carboxyl-labelled carbon
backbone from the growth medium. An explanation for the discrepancy
between these results and those reported previously (Kleber et ai., 1977) is
not obvious.
Cell-free extracts of D,L-carnitine-grown A. calcoaceticus are able

to split the C-N bond of labelled carnitine and form trimethylamine (Seim et
ai., 1982b). This activity was found exclusively in the membrane fraction.
All quaternary ammonium compounds with four carbon atoms in the chain
were split by the membrane fraction of A. calcoaceticus cells which had
been grown on D,L-carnitine, whereas glycine betaine and the closely related
trimethylammonium bases were not (Table 2). Thus, the four carbon atom
chain and the negative charge of the C,-atom, but not the hydroxyl group of
the C 3-atom, are necessary for the enzyme action to occur. Cleavage of the
C-N bond of 4-N-trimethylaminocrotonic acid by A. calcoaceticus, and
lack of stereochemical specifity with respect to the configuration of C-3 for
406
other betaines undergoing a similar degradation with formation of
trimethylamine, are inconsistent with a mechanism involving a simple
Hofmann-type of elimination reaction (Englard et aI., 1983).
The carnitine-splitting activity of acetate- and succinate-grown

bacteria was very low, but was increased by more than two orders of
magnitude when the enzyme was induced by growth of the bacteria on D,L-
carnitine. The formation of trimethylamine from ¥-butyrobetaine could not
be detected after growth on succinate or acetate. However, in these
experiments the specific radioactivity of 'X-butyrobetaine was less than one-
tenth that of carnitine, so a degradation comparable to that of carnitine
cannot be ruled out (Seim et aI., 1982b).
Although the remaining unlabelled carbon skeleton of the carnitine

has not yet been identified, the results show that A. calcoaceticus 69-V
possesses a carnitine degradation pathway which is different from that in
Pseudomonas spp. (Aurich et aI., 1967; Kleber et aI., 1978),
Enterobacteriaceae (Seim et aI., 1980, 1982a,c,d), and other Acinetobacter
spp. (see below).
SYNTHESIS OF L(- )-CARNITINE
At present there is an increasing demand for L( -)-carnitine in

medicine since it is a registered drug for the treatment of carnitine
deficiency syndromes. It is used in replacement therapy for haemodialysis
patients, for therapy and prevention of heart diseases linked to disturbed fat
metabolism and related indications, e.g. hyperlipoproteinaemia (Gitzelmann
et aI., 1987).
Cheap racemic carnitine can no longer be applied in clinics because

the D-enantiomer is not as harmless as was assumed in the past (Gitzelmann
et aI., 1987; Seim and Kleber, 1988). Therefore, many chemical, enzymic
and microbiological procedures have been developed in the last few years for
the production of enantiomerically pure L(- )-carnitine (Seim and Kleber,
1988).
Different strains of Acinetobacter have been used for the resolution

of racemic D,L-carnitine or its derivatives. Besides Achromobacter
parafinoclastus, Aeromonas liquefaciens, E. coli, P. aeruginosa,
Rhodotorula rubra, etc., the stereoselective hydrolysis of O-acyl-D,L-
carnitines was also effected by A. calcoaceticus (Nakamura et aI., 1985).
Another process for resolving a racemic mixture of D,L-carnitine, described
by Sih (1985, 1988), involves the growth of organisms of the genus
Acinetobacter (A. calcoaceticus ATCC 39648, A. lwoffii ATCC
39770), or mutants of these strains (A. calcoaceticus ATCC 39647, A.
lwoffii 39769), which are characterised by the unique ability to
preferentially degrade D( + )-carnitine in the racemic mixture under aerobic
conditions, followed by recovery of the L( - )-carnitine. It should be
407
mentioned that the enzymes for the stereoselective hydrolysis of O-acyl-D,L-
carnitines and the degradation of D( + )-carnitine have not yet been
described.
In contrast to our results (Jung and lung, 1988; Seim and Kleber,
1988), various strains of bacteria and yeasts, among them A. calcoaceticus
IFO 13006 (Kawamura et al., 1986a) and A. lwoffii ATCC 9036 (Yokozeki
and Kubota, 1984), have been described as producing L( -)-carnitine from
crotonobetaine under aerobic growth conditions. Culture media used for
growth of these microorganisms contained the usual carbon and nitrogen
sources and inorganic ions plus crotonobetaine. In addition, A.
calcoaceticus IFO 13006 also forms L(-)-carnitine from crotonobetaine
under anaerobic conditions (Kawamura et al., 1986b). In our experiments,
however, a variety of Acinetobacter strains (A. calcoaceticus 69-V, A.
lwoffii CCM 2376, A. lwoffii ATCC 9036) were unable to synthesise L(-)-
carnitine from crotonobetaine under both aerobic and anaerobic growth
conditions (lung and lung, 1988). We have always found that aerobic growth
conditions, as well as the addition of nitrate to anaerobically growing
cultures, diminished or inhibited completely the formation of L( -)-carnitine
(Seim and Kleber, 1988). The ability to synthesise L( -)-carnitine from
crotonobetaine, as well as the activity of carnitine dehydratase, which
catalyses the reversible dehydration of L( -)-carnitine:
L( -)-carnitine crotonobetaine
was detected only in cells of E. coli which had been grown under anaerobic
conditions in complex media supplemented with L( - )-carnitine or
crotonobetaine as inducers (lung and lung, 1988; lung et al., 1989). Only if
the carnitine dehydratase and the transport system had been induced
previously could aerobic carnitine synthesis be achieved (Seim and Kleber,
1988).
On the basis of the facts known so far, and even taking into account the
different cultivation conditions that have been used, it appears that different
acinetobacters have different pathways for the degradation/metabolism of
carnitine and its derivatives.
408
REFERENCES
Aurich,H., Kleber, H.-P., and Schopp, W.-D., 1967, An inducible

carnitine dehydrogenase from Pseudomonas aeruginosa,
Englard, S., Blanchard, J.S., and Miura-Fraboni, J., 1983, Production of
trimethylamine from structurally related trimethylammonium
compounds by resting cell suspensions of )(-butyrobetaine- and
D,L-carnitine- grown Acinetobacter calcoaceticus and
Pseudomonas putida, Arch. Microbiol., 135:305.
Gitzelmann, G., Baerlocher, K., and Steinmann, B., eds., 1987, "Carnitin in
der Medizin," Schattauer, Stuttgart.
Haddock, B.A., and Jones, C.W., 1977, Bacterial respiration,
Jung, K., and Jung, H., 1988. Zur Charakterisierung und Regulation der L( -)-
Carnitinmetabolisierung in Escherichia coli, Dissertation A,
Karl-Marx Universitiit, Leipzig.
Jung, H., Jung, K., and Kleber, H.-P., 1989, Purification and properties of
carnitine dehydratase from Escherichia coli - a new enzyme of
carnitine metabolization, Biochim. Biophys. Acta, 1003:270.
Kawamura, M., Akutsu, S .• Fukuda, H., Hata, H., Morishita, T., Kano, K.,
and Nishimori, H., 1986a, Microbial production of L-carnitine
from crotonobetaine, j P, 61,271,995.
Kawamura, M., Akutsu, S., Fukuda, H., Hata, H., Morishita, T., Kano, K.,
and Nishimori, H., 1986b, Production of L-carnitine, j P,
61,234,794.
Kleber, H.-P., Seim, H., Aurich, H., and Strack, E., 1977, Verwertung von
Trimethylammoniumverbindungen durch Acinetobacter
calcoaceticus, Arch. Microbiol., 112:201.
Kleber, H.-P., Seim, H., Aurich, H., and Strack, E., 1978,
Beziehungen zwischen Carnitinstoffwechsel und
Fettsaureassimilation bei Pseudomonas putida,
Miura-Fraboni, J., and Englard, S., 1982, Utilization of iI'-butyrobetaine and
D- and L-carnitine by Acinetobacter calcoaceticus and
Pseudomonas putida, Fed. Proc., 41:1166.
Miura-Fraboni, J., and Englard, S., 1983, Quantitative aspects of )(-
butyrobetaine and D- and L-carnitine utilisation by growing cell
cultures of Acinetobacter calcoaceticus and Pseudomonas
putida, FEMS Microbiol. Lett., 18:113.
Miura-Fraboni, J., Kleber, H.-P., and Englard, S., 1982, Assimilation of 'If-
butyrobetaine, and D- and L-carnitine by resting cell suspensions
of Acinetobacter calcoaceticus and Pseudomonas putida,
Nakamura, T., Ito, M., and Ueno, M., 1985, Optically active carnitines, J P,
60,214,898.
Seim, H., and Kleber, H.-P., 1988, Synthesis of L( -)-carnitine by hydration of
crotonobetaine by enterobacteria, Appl. Microbiol. Biotech.,
27:538.
409
Seim, H., Ezold, R., Kleber, H.-P., and Strack, E., 19HO, Stoffwechsel des L-
Carnitins bei Enterobakterien, Zeits. Allg. M ikrobiol., 20:591.
Seim, H., U:ister, H., Claus, R., Kleber, H.-P., and Strack, E., 1982a,
Formation of 'l( -butyrobetaine and trimethylamine from
quaternary ammonium compounds structure-related to L-carnitine
and choline by Proteus vulgaris, FEMS Microbiol. Lett.,
13:201.
Seim, H., Ulster, H., Claus, R., Kleber, H.-P., and Strack, E., 1982b, Splitting
of the C-N bond in carnitine by an enzyme (trimethylamine
forming) from membranes of Acinetobacter calcoaceticus,
Seim, H., Loster, H., Claus, R., Kleber, H.-P., and Strack, E., 1982c,
Stimulation of the anaerobic growth of Salmonella
typhimurium by reduction of L-carnitine, carnitine derivatives
and structure-related trimethylammonium compounds, Arch.
M icrobiol., 132:91.
Seim, H., Loster, H., and Kleber, H.-P., 1982d, Reduktiver Stoffwechsel des
L-Carnitins und strukturverwandter Trimethylammonium-
verbindungen in Escherichia coli, Acta Bioi. M ed. German.,
41: 1009.
Sih, c.J., 1985, L-Carnitine, WD, 85/04,900.
Sih, c.J., 1988, L-Carnitine and its preparation from D,L-carnitine with D-
carnitine-metabolizing microorganisms, US, 475,1182.
Yokozeki, K., and Kubota, K., 1984, Method for producing L-carnitine, EP,
0,122,794.
410
APPLICATIONS OF ACINETOBACTER AS AN INDUSTRIAL
MICROORGANISM
D.L. Gutnick, R. AllDn, C. Levy, R. Petter and W. Minas
Department Df MicrDbiDIDgy
GeDrge S. Wise Faculty Df Life Sciences
Tel Aviv University
Ramat Aviv, 69978
Israel
INTRODUCTION
Bacteria have a unique rDle to. play in the develDpment Df the

biDtechnDIDgy industry, nDt Dnly because Df their impDrtance in mDdern
recDmbinant DNA technDIDgy and mDlecular biDIDgy, but mDre impDrtantly
because Df the incredible diversity Df micrDbial processes and prDducts which
are Df pDtential cDmmercial and industrial interest. In many cases, these
systems are readily Dbtained frDm Drganisms whDse iSDlatiDn depends simply
Dn suitable enrichment prDcedures, sDmetimes cDupled with specific
screening techniques.
In the case Df Acinetobacter, several strains Df industrial impDrtance

have been iSDlated by either specific enrichment Dr simple screening
(Gutnick, 1987). Two. majDr activities Df applied interest assDciated with
variDus Acinetobacter iSDlates which have received a great deal Df
attentiDn in recent years are (i) the productiDn Df surface pDlysaccharide-
cDntaining biDpDlymers Df pDtential industrial impDrtance (Gutnick and
Minas, 1987; Gutnick and Shabtai, 1987), and (ii) the ability Df
acinetDbacters to. sequester large quantities Df inDrganic phDsphate in the
fDrm Df intracellular pDlyphDsphate, suggesting an applicatiDn in phDsphate
remDval frDm waste-water envirDnments (Fuhs and Chen, 1975).
HDspitals are also. a "natural envirDnment" fDr Acinetobacter in its

rDle as an DPPDrtunistic pathDgen (Ramphal and Kluge, 1979; Hinchliffe and
Vivian, 1980; NDble, this vDlume). Large cDllectiDns Df hDspital iSDlates
represent pDtentially useful banks Df Acinetobacter strains. A study Df
hospital iSDlates shDwed that a significant percentage Df strains grew Dn
411
hexadecane and produced extracellular bioemulsifier activity (Sar and
Rosenberg, 1983; Gutnick et aI., US Patent No. 4,883,757; see below).
Other activities of applied interest include the potential for

hydrocarbon and xenobiotic degradation (Gutnick and Rosenberg, 1977;
Gutnick, 1984; Cain, 1987; Asperger and Kleber, this volume; Fewson, this
volume), coal desulphurisation (Stevens and Daniels, European Patent No.
EP 126443; Isbister, US Patent No. 411(11535) and manganese leaching
(Karavaiko et aI., 1986a,b; Yurchenko et aI., 1(87).
This article examines Acinetobacter as an industrial microorganism.

The identification, chemical and physical properties, and mode of action of
the products themselves are discussed wherever possible. We have
attempted to include relevant information on the physiology of production
which is essential for the design of an efficient fermentation process. In
some cases it has been possible to describe simple approaches to select
improved mutant strains which either (a) exhibit enhanced productivity, or
(b) produce modified products with a broader range of activities. In this
regard we present some of our own results dealing with the cloning and
expression of relevant Acinetobacter genes in either Escherichia coli or
heterologous strains of A. calcoaceticus.
Some of the work from our own laboratory has been partially reviewed
elsewhere (Gutnick and Rosenberg, 1977; Gutnick, 1987; Gutnick and
Minas, 19117; Gutnick and Shabtai, 1987; Gutnick et aI., 1(87).
BIOPOLYMER PRODUCTION BY ACINETOBACTER
Emulsan - Acinetobacter strain RAG-l
Oil degradation and bioemulsification. Over 200 different

species of eukaryotic and prokaryotic hydrocarbon-utilising microorganisms
have been described (Beerstecher, 1954; ZoBel!, 1973; Gutnick and
Rosenberg, 1977; Rosenberg and Gutnick, 1981; Atlas, 1984).
Acinetobacter spp. are generally among the most prevalent natural
isolates. This is not a surprising observation since, in addition to possessing
the catalytic and regulatory genetic potential for hydrocarbon oxidation,
Acinetobacter spp. also exhibit a major requirement for efficient
utilisation of this potential in the natural environment: i.e. the ability to
interact physically with the hydrocarbon substrate (Gutnick and Rosenberg,
1(77). Two major forms of physical interaction are (i) direct contact
between the cells and the hydrophobic growth substrate, which is apparently
mediated by cell surface fimbrae that are similar in function to those of other
organisms which interact with hydrophobic surfaces (Rosenberg and
Kjelleberg, 1(86), and which are responsible for the avid adherence of the
cells to the oil; (ii) the production of biopolymers which bring about the
emulsification of the hydrocarbon growth substrate in water.
412
The possibility of exploiting the in-situ biodegradation of oil
pollutants by microbes led to the isolation and characterisation of A.
calcoaceticus RAG-1 (ATCC 31012) from a polluted beach in Tel Aviv
(Reisfeld et aI., 1972). Four basic observations characterised the
degradation of crude oil by RAG-I: (i) microscopic - the organism initially
adhered to and appeared to grow at the oil/water interface; (ii) nutritional -
. there was an absolute requirement for nitrogen and phosphorus; (iii)
physiological - only the alkane components of the oil were degraded
(Horowitz et aI., 1975); (iv) physical- the residual oil was emulsified. These
factors were incorporated into a large-scale field experiment aboard an oil
tanker (Gutnick and Rosenberg, 1977). The experiment had three basic
objectives: (a) the in-situ growth of RAG-Ion crude oil residues in an oil
storage tank; (b) the dispersion of these polluting residues as a stable
oil/water emulsion capable of discharge without the concomitant oil slick
formation; (c) the cleaning of the dirty oil storage tank as a result of the
efficient emulsification of the oil waste pollutants. All these objectives were
achieved (Gutnick and Rosenberg, 1977; US Patent No. 3,941,692) and some
of the potential advantages and disadvantages of this approach have been
discussed previously (Gutnick, 1987; Gutnick and Minas, 1987). One of the
major consequences of this type of large-scale experiment (Gutnick and
Shabtai, 1987) was the development of a programme to isolate, characterise
and examine the potential applications of the major factor(s) from
Acinetobacter and other oil-degrading isolates responsible for the
formation of stable emulsions. In the case of RAG-I, emulsification was
found to be due to the action of an extracellular non-dialysable polymer
which was termed emulsan.
Characteristics and interaction with hydrocarbons. The

emulsan produced by strain RAG-l is an extracellular anionic
heteropolysaccharide bioemulsifier (Rosenberg et aI., 1979a,b; Zuckerberg
et aI., 1979; Gutnick, 1987). Unlike many surfactants produced by oil-
degrading microorganisms, emulsan is also produced on a water-soluble
growth substrate, ethanol. Purified emulsan (M, 10 6 ) consists of D-
galactosamine, D-galactosamine uronic acid (pKa 3.05) and an unidentified
hexosamine. The amphipathic properties of emulsan are due in part to the
presence of fatty acids linked to the polysaccharide backbone in both ester
and amide linkages (Belsky et aI., 1979). In addition, when recovered from
the growth medium, emulsan is complexed with about 15-25% protein.
Removal of the emulsan-associated protein (ranging in M, from 10-90,000)
yields a product, termed apoemulsan, which retains most of the original
properties of emulsan itself (Zuckerberg et aI., 1979; Zosim et aI., 1982).
Emulsan exhibits several activities which are of applied and industrial

interest: (i) stabilisation of preformed oil-in-water emulsions; (ii)
formation of oil-in-water emulsions under conditions of low energy input;
(iii) removal of adherent microorganisms from hydrophobic surfaces. The
last activity, while related to the high affinity of emulsan for hydrophobic
surfaces, will be discussed separately.
413
Under conditions of low energy input, emulsan requires a mixture of
both aliphatic and aromatic hydrocarbons for optimum emulsification
(Rosenberg et aI., 1979b). However, we have recently found that this
specificity can be modified, either by the addition of small amounts (JLg) of
aliphatic alcohols, or by the addition of RAG-l proteins normally associated
with crude emulsan which are inactive by themselves, but which enhance the
activity of apoemulsan towards pure aliphatics (Zosim et aI., 1986, 1987,
1989).
Under conditions of high energy input, such as rapid mixing or sonic

oscillation, both emulsan and apoemulsan bind tightly at the interface
between the droplet and the bulk aqueous phase. Both activities (at either
high or low energy input) require relatively low ratios of emulsan to oil
(between 1:500 and 1:1000) because of the very high affinity of the polymer
for the oil/water interface (Zosim et aI., 1982). Emulsan forms a film of
about 20 nm on the surface of the oil droplet with the hydrophilic groups
orientated towards the aqueous phase, thus preventing coalescence of the
droplets (stabilisation). The affinity is sufficiently high that, even under
conditions of strong centrifugation, the emulsion is not broken, but separates
into an upper cream layer and a lower aqueous phase. The cream, termed an
emulsanosol, is itself an oil/water emulsion consisting of about 70% by
weight oil (Zosim et aI., 1982). Because of the orientation of the polymer on
the surface of the droplet, both organic and inorganic cations can be
partitioned into the oil layer as a result of electrostatic interactions with the
negatively charged uronic acid residues in the polymer (Zosim et aI., 1983).
In several instances the binding of the cations to the emulsanosol w'as even
greater than equimolar, probably because of a conformational change in the
polymer associated with its binding to the oil/water interface (Zosim et aI.,
1983).
The stability of the emulsanosols arises from the ability of the polymer
to form a stable film around the surface of the oil droplet (Zosim et aI.,
1983), which is probably stabilised by hundreds of points of contact between
the large amphipathic polymer and the oil droplet. One prediction of this
model for the rigid interaction of the biopolymer with the interface is that
cleavage of the polymer to lower M, subunits should severely impair the
stabilising ability of emulsan. An emulsan depolymerase, isolated from the
organism YUV-l (Shoham and Rosenberg, 1983; Shoham et aI., 1983), has
been shown to degrade emulsan via a trans elimination reaction and to
produce oligosaccharides with an average M, of about 3,000 after exhaustive
digestion. Depolymerisation of the emulsan polymer required the presence
of ester linkages. Breakage of only 5% of the glycosidic linkages (as assayed
by release of reducing groups) was sufficient to reduce the emulsifying
activity of emulsan by over 75%, thereby indicating a strong dependence for
activity on polymer size. Furthermore, in addition to its ability to degrade
emulsan in solution, the enzyme was able to degrade the biopolymer at the
oil/water interface (Shoham and Rosenberg, 1983) or on the surface of the
producing cell (Pines et aI., 1988; see below).
414
Physiology of emulsan production. The successful applications
and technologies for emulsan utilisation depend to a great extent on the
availability of the bioemulsifier and the economics of its manufacture
(Gutnick, 1984). A cell-associated form of emulsan constitutes a
minicapsular layer on the surface of RAG-1 cells growing exponentially on
minimal medium, and is released into the medium as the cells approach
stationary phase (Goldman et aI., 1982; Pines et aI., 1983). This capsule is
orientated on the cell surface such that (a) emulsan is a receptor for a
specific bacteriophage, and (b) the cells are partially protected from the
effects of toxic cationic surfactants such as cetyltrimethylammonium
bromide (Shabtai and Gutnick, 1985b). Evidence for a natural role for the
emulsan capsule comes from the finding that mutants of strain RAG-!
defective in emulsan do not grow on crude oil (Pines and Gutnick, 1986).
This deficiency was not corrected by either the addition of emulsan to the
culture or by growth in mixed culture with the wild-type RAG-! which itself
produces extracellular emulsan. The results indicate that the form of
emulsan required for growth on crude oil is the cell-bound biopolymer (Pines
and Gutnick, 1986). Emulsan-producing revertants of the emulsan-defective
mutants could be isolated on the basis of their ability to grow on crude oil.
The emulsan minicapsule was easily detected using a crossed

immunoelectrophoretic technique and has been shown to be identical to
purified apoemulsan (Bayer et aI., 1983). The cell-associated biopolymer,
which can comprise as much as 30% of the dry cell weight, also serves as a
receptor for a specific RAG-1 phage, ap3 (Pines and Gutnick, 1981, 1984a).
Thus, mutants of RAG-1 defective in emulsan production are readily isolated
on the basis of ap3 resistance (Pines and Gutnick, 1981). Emulsan-deficient
strains exhibit a characteristic pale translucent colony morphology which is
also observed when the cells are treated with emulsan depolymerase
(Shoham et aI., 1983; Pines et aI., 1988). These enzyme-treated RAG-1
cells become resistant to phage ap3 and simultaneously become sensitive to a
second Acinetobacter phage, ncp (Pines and Gutnick, 1984b; Pines et aI.,
1988). Evidence that emulsan undergoes a conformational change once
released from the cell surface comes from the finding that phage receptor
activity is not normally observed with cell-free emulsan, but can be
reconstituted in an emulsanosol. The results are consistent with the cation-
binding experiments described above (Zosim et aI., 1983) and suggest that
the conformation of emulsan at the oil/water interface resembles its
conformation at the cell surface (Pines and Gutnick, 1984a).
Emulsan release from the microbial cell surface is mediated by the

action of an exocellular esterase which appears in the growth medium just
prior to the appearance of cell-free emulsifying activity (Shabtai and
Gutnick, 1985a). A role for esterase in capsule release was indicated by
studies involving emulsan production in the presence of the protein synthesis
inhibitor chloramphenicol (Rubinovitz et aI., 1982; Shabtai and Gutnick,
1985a; Shabtai et aI., 1985). Both emulsan and esterase release were
stimulated in the presence of the antibiotic provided that the medium
contained both a carbon and a nitrogen source. In the absence of a carbon
415
source only the esterase was released. Subsequent addition of a carbon
source in the presence of chloramphenicol did not bring about release,
presumably because the esterase had already been released from the cell
surface and could not be made when protein synthesis was inhibited.
Mutants defective in esterase were also found to be defective in emulsan
production (Shabtai, 1982). In addition to weakening the association of the
emulsan minicapsule with the cell surface during fermentation, the esterase
allows the cells to utilise simple triglycerides as carbon sources.
Studies using a metabolic control strategy, combined with a computer-

aided system for regulating the rate of oxygen uptake by the bacteria during
their growth on triglycerides, revealed a complex and highly organised
network of cells interconnected through cell surface appendages (Gutnick
and Shabtai, 1987; Shabtai and Wang, in preparation). Under conditions of
oxygen limitation, these aggregates become de stabilised as the emulsan
concentration increases significantly. The results of the fermentation study
support the hypothesis that restricting the bacterial oxidative metabolism
enhances the de stabilisation of the minicapsule, resulting in a faster
production rate and easier separation of biopolymer from the cells. In this
regard it should be noted that emulsan is not the only molecule released
from the cell surface during the fermentation. About 25-30% of the
released extracellular material consists of a number of proteins ranging in M,
from 20,000 to 90,000 (Shabtai, 1982).
Genes involved in emu/san production. An esterase-coding

gene from RAG-1 (est) was cloned into E. coli (Gutnick et al., 1987; Reddy
et al., 1989; Fig. 1). A simple screen for esterase-positive clones exploited
the fact that strains of E. coli normally do not utilise triglycerides such as
triacetin as carbon sources, even though the same strains grow on the
0 0.5 1.0 1.5 2.0 2.5 3.0
pRA176 pRA172
H
cV
~
11
P
(( , i" 1
E X SpE SpS
C v
I I Ie: II II
Fig. 1. Physical map of pRA17. The open box indicates the insert of
RAG-l DNA. The arrow at the bottom shows the direction and extent of
the open reading frame of the est gene. pRA176 (esterase-negative) and
pRA172 (esterase-positive) carry TnS insertions at the positions
indicated by the long downward pointing arrows. TnphoA insertions into
est are identified by the dark arrow heads. Restriction enzyme sites
are as follows: C,ClaI; E,EcoRI; V,EcoRV; H,HindIII; P,PvuII; S,SalI;
Sp,SphI; X,SbaI.
416
5' ATG AAA TTT GGT ACT GTT TGG AAA TAT TAT TTT ACT GAA TCG TTA
M K F G T V W K Y Y F T E S L
CTA AAG GCG 3'

L K A
Fig. 2 . Putative leader sequence of the est gene from A. lwoffii RAG-l.
The translation product of the first 54 bp of the N-terminal sequence of
the esterase (est) gene from RAG-l (from Reddy, et al., 1989).
hydrolysis products, glycerol and acetate. Esterase-negative mutants of

RAG-l grow on acetate, but not on triacetin (RAG-l itself cannot grow on
glycerol). The same insert was obtained in four independent cloning
experiments. Esterase-positive clones exhibited high levels of esterase
activity, even when assayed with intact cells. The original esterase-positive
plasmid pRA17 carried a 2.2 kb insert which gave a single band following
Southern hybridisation with RAG-l DNA. It was, however, interesting to
note that other Acinetabacter strains which exhibit cell surface esterase
activity and grow on triglycerides, such as BD413 or RA57 (Rusansky et al.,
1987), show no homology with the pRA17 estprobe. After subcloning, the
est gene was sequenced. An open reading frame of 909 bp suggested that est
encoded a peptide of M,32,700, which correlated well with an in-vitro
translation product of 32,500. This polypeptide was not found in clones in
which Tn5 was inserted in the gene. In contrast to RAG-I, no esterase
activity was found in the cell-free broth. In addition, TnphaA insertions
were obtained within the first 300 bp of the insert. Localisation of these
fusions by sequencing and restriction analysis indicate that the enzyme is
exported out of the cytoplasm in E. coli. A putative signal peptide for est is
shown in Fig. 2.
As described above, expression of the RAG-l est gene in E. cali

conferred a new physiological function on the E. coli host, i.e. the ability to
grow on triacetin. Similar results were obtained with esterase-positive
clones of BD413 (Gutnick et al., 1987). A second Acinetabacter system
involves the ability to utilise ethanol as a sole source of carbon and energy.
This system is important for emulsan fermentation since polysaccharide
yields with RAG-1 are highest when the cells are grown on ethanol (Shabtai
et al., 1985). There have been a number of reports describing multiple
alcohol and aldehyde dehydrogenases from various strains of
Acinetabacter (Duine and Frank, 1981; Beardmore-Gray and Anthony,
1983; Fixter and Nagi, 1984; Singer and Finnerty, 1984, 1985).
Several clones were isolated when the RAG-l genomic library was
transformed into E. cali and plated on minimal plates containing ethanol as
a sole source of carbon and energy. These strains all possessed a plasmid,
pRETl, which carried a 5.3 kb insert (Fig. 3). When tested for alcohol
dehydrogenase activity (ADH), crude extracts of pRETl-containing cells
417
.,.
00 activity
NAO I NAOP
C
H' Be Se Be H
+ + pRET-l
H lie
rlC~~==~==~r-----~!H
C C
H Be 8e Be H
NO NO pRP3a
II
P.
h •
NO NO
~ pRP4
•
H P.
C
U
7 Se H
+ +
II) ~6 Ji
~tth------~----- pRP6a
8cp.
H SciSm
+ NO pRP7a
• ) I w/07Ai.
• 8 .. H P.
Be C LL...kJu
Fig. 3. Subclones of the alcohol dehydrogenase (ADH) complex from RAG-I. Plasmid pRETI was digested with various
restriction endonucleases and cloned into suitably digested pUCIB to generate plasmids pRP3a, 4, 6a and 7a respectively.
Cells carrying each of the plasmids were analysed in crude extracts (as shown on the left) or by zymogram analysis for
either NAD- or NADP-dependent ADH activities (ND: no detectable ADH activity). The site of TnS insertion in pRETI leading
to the elimination of either growth on or oxidation of ethanol is shown by the dark arrow head. Restriction enzyme sites
are as follows: C, ClaI; H, HindlII; Ps, Pstl ; S, SalI ; Sc, SacI ; Sm, Smal.
showed both NAD-dependent as well as NADP-dependent activities. In
addition, using in-situ specific staining for zymogram analysis on native gels,
three distinct NADP-dependent ADH-positive bands were detected, one of
which was also NAD-dependent. Expression of the insert in mini-cells
showed three unique polypeptides of apparent Mr 60, 49, and 48 kDa
respectively. Insertion of Tn5 into one end of the insert in pRET1
eliminated growth on ethanol, all ADH activities, and appearance of the
three peptides in a minicell expression system. The results suggested that
the genes encoding these activities might be organised or associated within
an "ethanol regulon". When pRETl was subcloned into pUC18, two
plasmids, pRP6 and pRP7 were obtained. Clones containing pRP6 exhibited
both NAD- and NADP-dependent activities, while the ADH from pRP7 was
dependent on NADP alone. Experiments are currently in progress in our
laboratory to identify which of the cloned ADH genes are involved in capsule
production and to study their regulation.
Mutants over-producing emulsan can be isolated on the basis of

resistance to a cationic surfactant such as cetyltrimethylammonium bromide
(CTAB) which is highly toxic to RAG-1 (Shabtai and Gutnick, 1986).
Emulsan-overproducers generate elevated levels of the polyanionic
bioemulsifier which neutralises the toxic effect of cetyltrimethylammonium
bromide (Shabtai and Gutnick, 1985b, 1986). This protection is mediated by
either cell-associated or cell-free emulsan (Shabtai and Gutnick, 1985b).
Over-producers isolated in this fashion have yielded between two and three
times more emulsan / g cells than the wild-type strain. Our most recent
experiments indicate that this approach may be applicable to the isolation of
over-producers which produce a variety of different polyanionic
bioemulsifiers.
A second approach to strain improvement involves the use of flow

cytometry, in conjunction with immunocytochemical labelling and cell
sorting, to isolate cells which exhibit larger amounts of a specific antigen on
the cell surface. In addition, this approach may be used to screen genomic
libraries for the expression of new specific cell-surface antigens in E. coli.
The feasibility of this approach has been demonstrated with a population of
E. coli clones containing a genomic library from RAG-1 (Minas et a!.,
1988).
Adhesion to hydrophobic surfaces. Strains of Acinetobacter

grow at an oil/water interface as a result of their ability to adhere to the
substrate (Rosenberg and Rosenberg, 1985). The role of hydrophobic
interactions in micro bial adherence has been reviewed extensively by
Rosenberg and Kjelleberg (1986). A simple and semi-quantitative assay for
adherence was initially developed to study the interaction of strain RAG-1
with oil (Rosenberg et a!., 1980). In this method washed bacterial
suspensions are mixed with liquid hydrocarbons, an emulsion is formed and,
upon settling, the adherent organisms rise to the surface with the
hydrocarbon droplets, resulting in a concomitant measurable decrease in
optical density of the aqueous phase. This decrease was taken as an
419
approximation of the cell surface hydrophobicity of the bacterial cells.
Adherent organisms could be removed from the oil droplets with alcohols or
with emulsan (Pines and Gu tnick, 1986). Using this approach it was possible
to describe a number of common features related to the cell surface
hydrophobicity of RAG-1 and other acinetobacters: (i) The adherence to the
hydrocarbon (whether liquid or solid) was not dependent on the ability of the
organism to degrade it. (ii) Cells grown under conditions which were optimal
for emulsan production were actually less hydrophobic than when grown on
rich medium in which emulsan is not produced. (iii) Mutants defective in
emulsan production were more hydrophobic than the wild-type (Pines and
Gutnick, 1984b). Similar results were obtained when emulsan was removed
from the RAG-1 cell surface with emulsan depolymerase, or from the cell
surface of A. calcoaceticus BD4 with a similar enzyme active against the
BD4 capsule (Rosenberg et ai., 19~Bc; Pines et ai., 1988). (iv) RAG-l
mutants lacking thin fimbrae on the cell surface were found to be defective in
adherence to hydrocarbons (Rosenberg et al., 1982). (v) Mutants defective in
adherence to hydrocarbon could still produce emulsan (Pines and Gutnick,
1984b). Selection for emulsan-defective derivatives of these non-adherent
mutants restored the hydrophobicity even though the revertant cells still
lacked the thin fimbrae. The results suggest that alternative sites on the
surface of the RAG-1 cell can mediate adhesion to hydrocarbons (Pines and
Gutnick, 1984b). (vi) Under conditions of poor mixing and with a small
inoculum, mutants defective in adherence lost the ability to grow on
hydrocarbons, suggesting a role for adherence in the early stages of growth
on these substrates (Rosenberg and Rosenberg, 1985). Growth was restored
by the addition of emulsan to the medium to disperse the growth substrate.
It thus appears that, in the case of RAG-I, early growth on hydrocarbons
occurs primarily via direct contact at the oil/water interface. During
subsequent growth, emulsan which accumulates in the culture broth: (a)
brings about a detachment of the cells from the hydrocarbon droplets, and
(b) enhances emulsion formation which further increases the interfacial
area.
According to the adherence assay described above, many other

bacteria such as pathogenic streptococci (Rosenberg et aI., 1983b),
Yersinia spp. (Lachica and Zink, 1984), Serratia (Rosenberg and
Kjellenberg, 1986), cyanobacteria (Fattom and Shiloh, 1985), myxobacteria
(Kupfer and Zusman, 1984), group A streptococci (Ofek et aI., 1983) and
gramicidin-producing bacilli (Rosenberg, 1986) present hydrophobic cell
surfaces. Moreover, the adherence of many of these strains to hydrocarbons
was found to be inhibited by emulsan. It was of interest that RAG-l itself
could bind to other biological (buccal epithelial cells) or non-biological
(polystyrene) hydrophobic surfaces (Rosenberg et aI., 1981). Finally, in
addition to removing the bound cells, emulsan prevented adherence of
washed suspensions to the surface (Rosenberg et aI., 1983a,d). The
potential industrial applications of these "anti-adherence" properties of
Acinetobacter biopolymers will be discussed in a separate section.
420
Other Acinetobacter Biopolymers
Strains BD4 and BD413. Taylor and Juni (1961a,b,c) originally

characterised the capsular polysaccharide of the heavily encapsulated A.
calcoaceticus strain BD4. Kaplan and Rosenberg (1982) studied the
protein-dependent emulsifying factors from BD4 and a minicapsulated
derivative mutant, BD413, originally characterised by Juni and Janick
(1969). Both these strains are highly competent in specific transformation
assays with DNA from virtually all known isolates of Acinetobacter. It was
found that these strains grow on hydrocarbons and produce extracellular
emulsifiers. Both strains produce an extracellular polysaccharide, with a 3: 1
molar ratio of rhamnose to glucose, which was similar in composition to the
cell-bound capsular material. The structure of this material was shown to
consist of a repeating heptasaccharide branched subunit (Kaplan et aI.,
1985). While strain BD4 produced about four times more total
polysaccharide (cell bound and extracellular) than the mutant BD413, the
latter strain released up to half of the polysaccharide into the medium. The
parent strain BD4 released only about 10% of the rhamnose-containing
polymer into the medium. Unlike RAG-I, the BD strains grow and produce
emulsifier on glucose. This may be of importance in any scale-up process for
emulsifier production. The emulsifying factor from BD413 was precipitated
in the presence of 50% ammonium sulphate and shown to resemble emulsan
in its requirement for mixtures of aliphatic and aromatic hydrocarbons as
substrates for emulsification. It is of interest that, in contrast to emulsan,
the BD413 product is relatively inactive with crude oil. In this regard it
should be noted that, unlike RAG-I, A. calcoaceticus BD413 does not
grow on crude oil (Rusansky, 1985). When the capsular polysaccharide of
BD4 is released as an extracellular polymer during growth, the product is
active as an emulsifier. However, when the capsule is mechanically removed
by shearing from the cell surface, the polysaccharide is not active as an
emulsifier. Extracellular protein produced by a mutant which does not make
the polysaccharide has been shown to reconstitute the emulsifying activity of
the rhamnose polymer. The amphipilic properties of the BD413 polymer are
thus provided through the interaction of protein with the polysaccharide. It
remains to be determined whether the hydrophobic portion of the emulsifier
complex is due to the presence of hydrophobic proteins, or whether the
amphilicity is due to conformational changes in the complex. Recently the
lipoprotein from E. coli has been shown to be active in restoring emulsifying
activity to the BD413 polysaccharide (Z. Zosim, unpublished results).
Biodispersan - A. calcoaceticus strain A2. The Acineto-

bacter emulsifiers, emulsan or the polysaccharide-protein complex of BD4,
exert their effects by adhering to a hydrophobic organic surface. Rosenberg
et aI., (1988a,b) described the production by A. calcoaceticus strain A2 of
an extracellular polysaccharide, termed Biodispersan, which binds to
inorganic materials such as calcium carbonate. The organism was first
isolated from soil by an enrichment procedure for Acinetobacter
(Baumann, 1968), followed by a screening of the extracellular culture broth
for its ability to disperse limestone powders in water. Strain A2 was
421
identified as a member of the genus Acinetobacter on the basis of its ability
to transform BD413. However, since the organism was unable to grow on
glucose (Rosenberg et al., 1988a), it may be more appropriate to assign it to
the species A. lwoffii. The biodispersan polymer accumulated in the
culture broth of an ethanol-grown culture during late exponential and
stationary growth (Rosenberg et al., 1988a). The material was precipitated
from the culture broth with ammonium sulphate and subsequently purified
by hot phenol treatment to produce a polysaccharide of M,51,400; yields of
about 30% by weight and 84% by activity were obtained. Biodispersan
appeared to consist of four different amino sugars, including glucosamine,
rhamnosamine or fucosamine, galactosamine uronic acid and an unidentified
amino sugar. Potentiometric titration and 13C NMR analyses suggested that
at least 90% of the amino groups were acetylated. The polymer showed no
emulsifying activity, but was effective in dispersing the inorganic limestone
powders (Rosenberg et al., 1988b). In addition to its dispersing activity,
Biodispersan was shown to dramatically reduce the time necessary to grind
calcium carbonate to fine particles, apparently as a result of binding of the
polymer (Rosenberg et al., 1989).
A. calcoaceticus strain H OJ-N - Particulate surfactant.

Kappeli and Finnerty (1979, 1980) observed the accumulation of
extracellular membrane vesicles in cell-free broth from a culture of A.
calcoaceticus strain HOI-N growing on hexadecane. These vesicles were
composed of proteins, phospholipids and lipopolysaccharides, suggesting
that they might be derived from the outer membrane of the cells. However,
the phospholipid to protein ratio was five times greater than in the outer
membrane, while the lipopolysaccharide content (determined by assaying for
keto-deoxyoctulosonic acid) appeared to be enriched some 360-fold. The
vesicles were found only in the culture broths of cells growing on alkanes.
The authors suggest that these extracellular vesicles play an important role
in alkane uptake since: (i) they are only found in cultures growing on
hydrocarbons; (ii) radioactive hexadecane bound to the vesicles was taken
up by intact cells; (iii) the vesicle-solubilised hexadecane was transferred to
hydrocarbon inclusion bodies normally observed in alkane-growing cells of
this strain (Singer and Finnerty, 1984).
Molecular Analysis of Bioemulsifier Production by Acinetobacter
Apart from RAG-I, BD4 and BD413, a number of other strains of

Acinetobacter have been isolated from natural environments and shown to
grow on hydrocarbons and to produce extracellular emulsifying activity (Sar
and Rosenberg, 1983). Studies with specific phages (Pines and Gutnick,
1981) indicated that these emulsifier complexes exhibit widely different
structures. In most cases the emulsifiers required both a capsular
polysaccharide and a mixture of extracellular proteins for activity.
Both BD4 and BD413 are naturally competent for transformation with
genomic DNA from all known Acinetobacter strains (Juni, 1972),
suggesting a high degree of homology in both sequence and function.
422
Moreover, it has been found (W.Minas and D.Gutnick, in preparation) that
while plasmids derived from ColE 1 are not stably maintained in BD413, they
do integrate into the chromosome provided a fragment of homologous
Acinetobacter DNA is present in the plasmid. It was postulated that if
RAG-1 DNA contains homologous sequences to emulsifier-related genes in
BD4 or BD413, it might be possible to detect such sequences by virtue of
their ability to integrate into the BD4 chromosome, thereby giving rise to
emulsifier-defective mutants (Levy, 1990; W.Minas and D.Gutnick, in
preparation). Genomic libraries of RAG-1 DNA cloned into a derivative of
pBR322 were used to transform either BD413 or BD4 to kanamycin
resistance. Among the BD4 transformants, six mutants defective in capsular
polysaccharide production were isolated, initially on the basis of translucent
(tlu) colony morphology. Further evidence that these mutants were
defective in exopolysaccharide (EPS) production included: (i) India ink
staining; (ii) resistance to exopolysaccharide-specific phages; (iii) lack of
extracellular emulsifying activity; (iv) enhanced adherence to hydrocarbon.
Chromosomal DNA isolated from the kanamycin-resistant transformants
could be used to transform BD4 to tlu. Southern blot analysis demonstrated
EVHBHB H
~I ::t
8.411
I I 1 K8
pHLl7/lU II I I
I I
EIHEV 5 P H P
5 HB EI PS H EI
11.25
pHLIZ5 I II I II I I
k\I
BHEV
II
5
I
p
I
H
I
P
I
H P PSH B
pHLZZ7 ~~)~(=~I~'(=~Ih-_____-rI~r-l8.7
Bk;}IEV I~ ~ ~ ~
5 B
I I 8.7
pHLza.
ski II
5 P
I I
H
I
P
P
H P 5 H SSEVH
I I 13.011
I I I )1 I
pHLU.
~
BHEV
II
5
I
P H
I i
P
Fig. 4. Physical maps of rescued plasmids from exopolysaccharide mutants

of strain BD4. Followinginsertional mutagenesis by homologous
recombination, the integrated plasmids were rescued from the chromosome
by digestion with Sac1 followed by self-ligation and transformation into
E. coli MM294 as described in the text. The inserts are represented by
the blank boxes. The total size (kb) of each plasmid is shown at the
right hand end; the vector itself, pWM3, is 6.5 kb. Restriction enzyme
sites are as follows: E1, EcoR1; EV,EcoRV; H,HindIII; P,PstI; S,Sa11.
423
that the entire plasmid was integrated into the chromosome. The flanking
sequences adjacent to the integrated vectors were rescued from the
chromosomal DNA of the six different mutants by cleavage with S acJ (which
does not cut the vector), and transformed following self-ligation into E. coli.
As evidenced by restriction analysis (Fig. 4), five unique sequences were
isolated from six mutants. These were used as probes in subsequent Southern
blot analysis of DNA from either RAG-lor BD4.
The functions encoded by these sequences are not known. However, it

was of interest to note that three of the six BD4 mutants were unable to grow
on minimal medium with glucose as carbon source (Levy, 1990). This was at
first surprising in view of the finding that RAG-1 cannot grow on glucose
(Reisfeld et al., 1972). One possible explanation emerged from the finding
that the mutants unable to grow on glucose are also defective in glucose
oxidation. In certain strains of A. calcoaceticus, such as BD4, glucose
oxidation is generally catalysed by glucose dehydrogenase (GDH), whose
activity is mediated by the co-factor pyrroloquinoline-quinone (POO)
(Duine et al., 1979; van Schie et al., 19R4, 1985). In addition,GDH activity
can be restored by the addition of POO to many of the strains of A. lwoffii
(e.g. RAG-I) which cannot grow on or oxidise glucose. Such strains possess
an apo-GDH protein, but lack the genes necessary for the biosynthesis of
POO (Duin e, this vol ume). In some cases it is, therefore , likely, that the
mutants generated by homologous recombination between RAG-l and BD4
are the result of insertion of the gene for apo-GDH. This cannot, however,
explain the behaviour of all of the mutants since the restriction maps are
completely different (Fig. 4). The genes from BD4 that encode the synthesis
of POO have recently been cloned and sequenced (Goosen et al., 1987,
1989). Four gene products, encoded on a 5.1 kb fragment, are required for
production of POO in A. calcoaceticus. When a plasmid carrying all four
genes was introduced into a strain of A. lwoffii which produced apo-GDH,
the resulting transconjugants could produce acid from a variety of sugars.
All four genes were required to restore holo-GDH activity in the A. lwoffii
system. The ability to synthesise POO could also be introduced into a
phosphotransferase-defective mutant of E. coli K12 that was unable to grow
on glucose, produced apo-GDH, but lacked the POO co-factor.
Transformants were able to grow slowly on glucose provided the POO genes
were placed under the control of the lac promoter. Restoration of the POO
synthesising system was, however, less than complete since exogenous
addition of the cofactor enhanced the growth of E. coli transformants even
further. The results with the POO system suggest the possibility of
manipulating RAG-1 genetically so that growth and/or production of
emulsan or other biopolymers could be achieved using glucose or other
sugars as carbon substrates.
Cryptic plasmids in BD413 and the construction of shuttle

vectors. In one of the BD413 mutants generated by the homologous
recombination method mentioned above, extrachromosomal DNA was
detected which was not observed before the transformation. However, using
Southern blot analysis, no homology was detected between this DNA and the
424
original plasmid vector used for transformation. Further analysis of the
extrachromosomal DNA revealed that BD413 itself (even before
transformation) contains three cryptic plasmids of 2.3, 2.6 and 7.4 kb
respectively. The three plasmids were observed only in preparations of total
genomic DNA. Two plasmids were detected with a similar extraction
procedure in BD4. Southern blot analysis revealed homology between all
three plasmids of BD413 and the two plasmids of BD4. Two of the BD413
plasmids were sub-cloned into pUCI9, and restriction maps established (Fig.
5). Identical restriction sites were found in a 2.1 kb fragment. The
recombinant plasmids were stably maintained as independent replicons in E.
coli MM294 and A. calcoaceticus BD413 or BD4. No apparent
modification of the plasmids was detected, even after several reciprocal
transfers between E. coli and A. calcoaceticus. When sub-clones
carrying partial deletions of the cloned 2.3 kb Acinetobacter plasmid were
examined for transformation efficiency (Fig. 6), it was found that a minimum
size fragment of about 190 bp (Figs. 6 and 7) was sufficient to support
Ikb
I
PS XC SM H P5
km·R !!!!
HII HII
BI sc c ~
...........1.....
.............I' ~{:1 IlcHe £51 I
81 All SC EO
SA SA BII SC C P5 xc SM H PS 8.5
pWM5 ~c=~1==~====~==~='~rr--------.XP-5~;~I~!~I~'____~'~"~
81 PVEvr--.----------------E-v§~V 81 All SC_ SC~ EO
SM_ SM~A HII- Hlle
EV SA SA 811 SC C BN EI 8.5
pWMB ~=j'E<:tIi:::E:i'E:,:,:'1rr"'.Z':':::E'/:Ei:E:"E",-,,:;td:z,,',Z"':':E):]:k:E: :~~==~I~Ir./.v-------~l~A~,f""'~-L-1E~
sHe He
SM
SA
PS
B
Fig. 5. Partial restriction maps of the cloned cryptic plasmids from

strain BD4l3. pWM4 carries the cloned 2.3 kb plasmid from BD4l3 (shown as
a box), linearised with EcoRV and cloned into the SmaI site of pUC19.
Plasmid pWM6 carries a 3.9 kb HindIII fragment of the 7.4 kb cryptic
plasmid from BD4l3. pWM5 is derived from pWM4 and carries the kanamycin
resistance gene of Tn903 (heavy line) inserted into the bla gene (striped
box) of the pUC19 part (line) of the shuttle vector. Shaded areas of
the inserts indicate regions with identical restriction sites in pWM4 and
pWM6. Restriction enzyme sites are as follows: AII,AvaII; B,BamHI;
BI,BgII; BII,BglII; BS,BstEII; C,ClaI; EO,EcoOl09; EI,EcoRI;
EV,EcoRV; H,HindIII; HII,HincII; PS,PstI; PV,PvuII; SA,SalI; SM,SmaI;
X,XhoI.
425
Transformants per
)Jg of DNA in
,.
.. ~
'" .. plasmid BD413 MM294
.
0 0
0 r:t r:t
\
0
0 ~~
:: ~ ~~
~
~ 0
Ii
0 flI fII_ ~ 0
r:t ~ Jf 9~ J: ,Jf r:t
I I 5 5
I I I pWM4 6x10 1xlO
6.5
4 5
6500 9x10 2xlO
6.0
~~ 3 5
61700 3x10 2x10
4.8
1
62200 4xlO 3x10 6
4.3
Fig. 6. Efficiencies of transformation in A. calcoaceticus BD413 and E.

coli MM294 using shuttle plasmids pWM4 and deletions. Deletions in pWM4
were generated by partial digestion with SaIl followed by self-ligation.
Sites of the deletions are indicated as gaps in the restriction maps
shown on the left side of the figure. The size of the deletion is
expressed in bp. The remaining portion of the insert is indicated by
the filled boxes. Transformation efficiency is expressed as the number
of ampicillin-resistant transformants of either BD413 or E. coli MM294
per ~g plasmid DNA.
10 20 30 40 50 60
* * * *
CCCATCCGCT TCAATTGGAA AATCAGACT GTTTGAACTG GCGAAGCAAA GCTATATTTG
70 80 90 100 110 120
* * * *
ATGAATCCTG CCAATTGAAA GTCTCACAA CCATTGCCGA TTGCATTACC TTGAGATAAA
130 140 150 160 170 180
* * * *
TCCATTGGGA TTTCACAATA CAAAGCAAC ATGGTCATAG'TGACGGTTTG AACCGTCAGA
190
CGGTTTAAAG
Fig. 7. Sequence of the 190 bp fragment which still permits independent
replication of pUC19 in A. calcoaceticus BD413.
426
replication in the Acinetobacter strains. However, the efficiency of
transformation with the various sub-clones was reduced by up to four orders
of magnitude, depending on the size of the deletion (Fig. 6). No effects
resulting from the deletions were observed with respect to replication in E.
coli. Although we were not able to introduce DNA into RAG-l via
conventional methods of transduction, transformation or conjugation, the
shuttle vector could be transferred into the emulsan-producing strain using
electroporation. Although the copy number of the vector in RAG-l was too
low for it to be observed using normal extraction procedures, it could be
readily isolated as an unmodified shuttle vector following extraction of total
DNA from the electroporants and transformation into either E. coli or
BD413. Both restriction and Southern blot analysis demonstrated that the
plasmid was not integrated into the RAG-l chromosome. Finally, the
shuttle vector was used to clone alcohol dehydrogenase genes from RAG-I,
which were shown subsequently to positively complement mutants unable to
grow on ethanol (selected on the basis of their resistance to allyl alcohol).
In further experiments the 1.2 kb gene for the extracellular esterase from
RAG-l has been inserted. Expression of this gene in E. coli allows growth
on simple triglycerides such as triacetin. Subsequent insertion of foreign
sequences within the esterase gene abolishes growth on these substrates (R.
Alon, R.Petter and D.Gutnick, unpublished results).
Hunger et al. (1990) described the construction of a shuttle vector by

fusing a cryptic plasmid from A. lwoffii with pBR322. This vector was
found to be functional in both E. coli and A. calcoaceticus. The plasmid
was sub-cloned and shown to contain a 1350 bp fragment which apparently
contained the origin of replication. This ori consisted of an AT -rich region,
an 18 bp element with palindromic symmetry, and 11 repeats of a consensus
sequence, AAAAATAT, eight of which were clustered within about 360 bp.
Interestingly, this sequence was not present in the 190 bp segment minimally
required for stability, maintenance and replication of pWM4 in
Acinetobacter (Figs. 6 and 7). The 9 bp repeat was, however, found in
several locations throughout the 2.1 kb insert.
Kaplan and Silverman (1989) have begun a study of the genetic and
regulatory aspects of exopolysaccharide biosynthesis (EPS) in BD4. A
recombinant cosmid genomic library bank of BD4 constructed in E. coli was
used to complement spontaneous EPS- mutants. All of the EPS- mutants
mapped to one of two separate loci, epsA or epsB. Most of the EPS-
mutants carried deletions in epsA, and transposon mutagenesis
demonstrated that a region of about 10 kb in epsA was essential for EPS
production. A strange result was obtained when an EPS- mutant, termed
4011, carrying a deletion of over 30 kb, of which 20 kb were missing from
epsA and 10 kb were missing from epsB, was transformed with purified epsB
DNA. The transformants were found to contain both "missing" fragments,
the 20 kb of epsA and the 10 kb of epsB. Moreover, chromosomal DNA
from mutant 4011 could transform other epsA mutants to wild-type despite
the fact that the deletion in 4011 was larger than the deletion in the recipient
epsA mutants. Southern blot analysis indicated that the transformants
427
contained a wild-type epsA locus. Kaplan and Silverman (1989) suggested
that the eps DNA in deletion 4011 is maintained in some sort of "archived" or
"invisible" form which escapes detection by classical means of gene
expression or hybridisation. In addition to elucidating the mechanisms
associated with DNA "archiving", it will be of interest to determine whether
the phenomenon is general, or specific to the EPS system in Acinetobacter,
and whether the process plays some role in the regulation of capsule
production.
APPLICATIONS OF ACl N ETOBACT ER SURFACE POLYMERS
The chemical industry has produced very large numbers of synthetic

materials which carry out many of the activities described above. This vast
myriad of products, which ostensibly do the same thing, stems largely from
industrial experience which has shown that there is no universally accepted
surfactant which fits the needs of every application. Moreover, the ability to
formulate such materials in combination with others to generate surfactant
packages for specific applications has added to the demand for surfactants in
a number of industries. Over 2.5 million tons of surfactants were sold in the
USA alone in 1983 (Gutnick, 1984). If chemically-enhanced oil recovery
technology becomes applicable, surfactant consumption is likely to double by
1995. These optimistic projections for growth of the surfactant industry
have stimulated a great deal of interest in the potential application of
biotechnology as a new dimension for commercial development in this area.
The advantages of the approach include the ability to: (i) exploit the vast
amount of knowledge and experience from the fermentation industry; (ii)
design simple yet powerful selection techniques for the isolation of suitable
surfactant-producing organisms from nature; (iii) utilise classical
techniques of mutation and selection, as well as recombinant DNA
technology, in strain improvement programmes; (iv) exploit unique
biological specificity for producing specific formulations; (v) produce
materials which are non-toxic. Along with the advantages come a number of
problems which must be overcome if biosurfactant production and
technology is to become a reality. These may include: (a) the high cost of
fermentation products in comparison to those which are chemically
synthesised; (b) the difficulty of introducing new products to compete with
materials already on the market; (c) the high cost of health risk assessments
for new products introduced into certain industries.
Kosaric et aI., (1987) and Kachholz and Schlingmann (1987) have

considered many of the properties of potential biosurfactants as well as the
industries where they may be applied. These properties include
emulsification, de-emulsification, wetting, spreading and penetration,
solubilisation, dispersal of solids, air-entrapment, foaming, de-foaming,
detergency, antistatic and corrosion inhibition. The industries include
agriculture, building and construction, elastomer and plastics, foods and
beverages, industrial cleaning, leather, metals, paper, paint and protective
coating, petroleum and petrochemical products, and textiles. However, we
have found only a few published reports summarising the results of large-
428
scale field experiments with biosurfactants. Such experiments require that:
(i) the fermentation and production technology be developed to the point
where sufficient product can be produced for testing; (ii) sufficient
information is available so that the biosurfactant product can be suitably
formulated for the particular application to be tested; (iii) enough is known
about the particular application so that it can be tested on a large scale using
suitable equipment, storage conditions, means of monitoring results etc.;
(iv) prior governmental approval for certain tests is obtained.
Developmental programmes of this magnitude are usually sponsored within
the industry itself and, as such, the results are not always available.
The Use of Emulsan in Stabilisation of Heavy-Oil-in-Water

Emulsions
Although high molecular mass extracellular bioemulsifiers such as

emulsan are not particularly effective in reducing interfacial tension, they do
have the property of binding tightly to an interface. The practical outcome
of such a property is the ability to utilise relatively small amounts of
biopolymer to stabilise oil-in-water emulsions. This section will describe
two field experiments illustrating potential applications of emulsion
stabilisation which represented part of a programme by Petroferm USA to
explore two technologies (Hayes et aI., 1986): (a) pipeline transportation of
oil-in-water emulsions of heavy viscous crude oils, and (b) direct combustion
(without de-watering) of oil-in-water emulsions containing viscous
hydrocarbons.
Heavy oil transportation. A pipeline test was carried out in

which about 40 barrels of Venezuelan crude oil (viscosity of about 200,000
cp) were emulsified to form a 70% oil-in-water emulsion with a commercial
surfactant-emulsan formulation at a ratio of one part surfactant package to
500 parts oil. The emulsion was pumped a physical distance of
approximately 380 miles at ambient temperature through a 3.125 in. half-
mile pipeline loop. The apparent viscosity of the emulsion was reduced to
70 cpo During the test the flow was stopped for threedays to simulate a pump
failure. When pumping was resumed there were no effects observed on
pressure drop, flow rates, or emulsion characteristics. It was calculated that
when pipe diameters, pump diameters, flow rates, frequency of pump
transits, etc., were considered, the emulsion was subjected to stresses and
shear forces comparable to having been pumped a distance of 26,000 miles in
a commercial pipeline. Conditions of high stress during pump transit
. generally give rise to inversions of oil-in-water emulsions stabilised by
classical surfactants.
Direct combustion of heavy fuel emulsions. Injection of

small amounts of water into fuel to form water-in-oil emulsions improve
combustion characteristics of the fuel. It was of interest, therefore, to
determine whether emulsan-stabilised oil-in-water emulsions could be
combusted directly. In this test a high-asphaltene no.6 fuel oil-in-water
emulsion (70:30 w /w) was prepared one week prior to the combustion test.
429
The results of the test can be summarised as follows: (i) ignitability and
stability of the biopolymer-stabilised emulsion were comparable to the
unemulsified no.6 oil; (ii) minimum excess air levels of less than 2% were
obtained without any visible smoke or CO in the flue gas; (iii) flame
appearance was the same as that of the no.6 fuel oil itself; (iv) carbon
burnout and light-off were also similar to the parent oil; (v) axial flame
temperatures were lower by 100-150°C along the length of the combustion
chamber; (vi) nitrogen oxide gases (Nox emissions) were significantly
reduced during combustion of the emulsions. The direct combustion test of
emulsan-stabilised emulsion fuels suggested several advantages including:
(a) ease of transportation; (b) uniform flow behaviour for various fuels - the
hydrocarbon fuel itself does not significantly influence the emulsion
viscosity; (c) ease of handling cumbersome heavy fuels such as high-
asphaltene vacuum residuals; (d) upgrading of low cost fuels - cutter stock
need not be added in order to burn the lower quality fuel. This technology
has been tested on a large scale and shown to be feasible. Its commercial
viability will depend on the relative cost of crude oil compared to the savings
associated with upgrading a heavy fuel by emusifying it in water.
The Use of Emulsan in the Treatment of Dental Plaque
A patent filed by Eigen et al. (US Patent No. 4,619,825), on behalf of

the Colgate-Palmolive Company, describes the application of emulsan to
improve oral hygiene by removing dental plaque. Dental plaque is a deposit
of a mucilaginous gelatinoid-type material which forms on the surface of
teeth and which is readily produced by and/or invaded by bacteria.
Different organic acids produced by the plaque-forming bacteria can cause
decalcification of the dental enamel, leading to dental caries. Other
microbial products are thought to promote gingivitis by attacking soft gum
tissue. Streptococcus mutans is generally recognised as the major
aetiological source of dental caries. Emulsan solutions (0.1 %) were shown
to reduce dental plaque formation by S. mutans in vitro by about 81%.
Similar results were obtained with S. salivarius SS2 and S. sanguis FC4.
In an animal model employing hamsters whose mouths were infected with S.
mutans, it was found that: (a) pre-treatment of the infecting organisms with
apoemulsan resulted in about a 61% decrease in the recoverable S. mutans
after feeding the hamsters for five weeks on a diet designed to induce dental
caries; (b) when the infection was carried out with S. mutans which were
not pre-treated, but with emulsan (1%) included in the drinking water, the
redu ction was 77 %; (c) a large decrease in den tal caries was observed
following either pre-treatment with emulsan solutions (47%), or inclusion in
drinking water (63%). Finally, after infection, hamster teeth were swabbed
twice daily with solutions containing either sodium fluoride or emulsan (0.25
or 1% respectively ). At the end of a 12-week period the hamster teeth were
examined for S. mutans infections and dental caries. In the case of
fluoride, the S. mutans population actually increased (+21%), while the
emulsan treatment resulted in a 53% reduction compared with a control
430
containing only water. However, fluoride treatment did bring about a 90%
reduction in dental caries compared with 61 % for the emulsan treatment.
The authors described emulsan-containing formulations for either mouth-
wash or toothpaste which combine the anti-caries activity of fluoride and the
activity of emulsan in reducing plaque formation. The results with emulsan
were somewhat surprising in view of the finding that emulsan interactions
with microbes are generally hydrophobic in nature (Rosenberg et aI., 1981;
Rosenberg and Kjelleberg, 1986), while the adhesion of S. mutans is
thought to be primarily via lectin-like interactions (Gibbons, 1980; Gibbons
and Etherden, 1983). The patent authors explain the strong effect of
emulsan in the reduction of dental caries and plaque formation as an
interaction of the galactosamine residues in the apoemulsan polysaccharide
backbone with lectins on the surface of the S. mutans cells, thereby
preventing their adherence. If this is true, than de-esterified emulsan, which
is no longer amphipathic, should also be effective.
The Use of Biodispersan in Grinding Limestone and Making Paper
Limestone (calcium carbonate) is a component of paper, the quality of

which is determined to some extent by the size, shape and homogeneity of the
limestone particles. The major cost in the processing of limestone powder is
grinding, since chemical dispersants are frequently not used because of their
adverse effects on the quality of the paper, as well as the difficulties of
disposal. When Biodispersan (0.1%) was tested in this system it was found
to reduce the time necessary for grinding from 6 h to 3 h. After 9 h about
70% of the particles were about 2Jtm in diameter. Production of
Biodispersan was scaled-up and several trials showed that paper blended
with Biodispersan-ground limestone was of high quality. It was concluded
that, in addition to its properties as a flocculating surfactant, Biodispersan is
also a "surfractant". The mechanism of action of this interesting
biopolymer is currently unknown. Biodispersan has been patented in a
number of countries (e.g. Rosenberg and Ron, Swiss Patent No. SW122
672500).
Drag Reduction
The turbulent flow of fluids can be significantly impaired by the

friction of the fluids. This phenomenon, known as drag, has been shown to
be reduced by viscous polymers. Both synthetic and biopolymers have been
shown to be effective drag reducers (Hoyt, 1985). Sar and Rosenberg (1987)
reported that the isolated capsular polysaccharide from BD4 was an effective
drag-reducing polymer. A three-fold higher drag-reducing activity was
observed when the capsule was tightly bound to the cell surface (3.6 x 109
cells Iml reduced the friction of turbulent water by 55%). Cells from a
capsule-deficient mutant did not have any significant drag-reducing activity.
The cells apparently provide a natural system for immobilisation of the
polymers.
431
PHOSPHATE REMOVAL FROM WASTE-WATER
BY ACINETOBACTER
Phosphate removal from waste-water is an essential feature of any

sewage treatment facility because of the threat of eutrophication. Although
the addition of various agents, such as salts of calcium, iron or aluminium, to
activated sludge can be used to remove phosphates from the waste-water by
settling, the treatment causes the accumulation of large quantities of
chemical sludge. In addition, the initial cost of the materials required for
chemical treatment has induced investigators to seek alternative biological
methods for phosphate removeal (e.g. Levin and Shapiro, 1965; Fuhs and
Chen, 1975; Buchan, 1983; Marais et al., 1983; Hao and Chang, 1987). The
unique role of Acinetobacter in this process of phosphate removal will be
discussed briefly below. Fixter and Sherwani (this volume) provide a critical
assessment of the evidence for the function of polyphosphate as an energy
reserve and a phosphate reserve in acinetobacters.
Accumulation of Polyphosphate by Microorganisms
Any practical application of metabolic uptake to remove dissolved

orthophosphate (Pi) from sewage effluent depends on the ability of the
sludge organisms to take up the phosphate in quantities exceeding those
required for normal metabolic activities. Polyphosphate (Pn), the form in
which excess phosphate is frequently stored, has been detected in a wide
variety of microorganisms (Kulaev, 1979; Kulaev and Vagabov, 1983) and is
usually stored either in the periplasm (Umnov et al., 1975; Tijssen and van
Steveninck, 1984) or in the form of large cytoplasmic volutin granules
(Harold, 1966; Kulaev, 1979; Miller, 1984).
There are generally thought to be two distinct physiological conditions

leading to excess Pi accumulation by microorganisms. The so-called
"phosphate overplus" phenomenon (Harold, 1966) involves the uptake of
excess Pi which occurs when phosphate-starved microorganisms are
transferred to a phosphate-rich growth medium. Sewage treatment plants
are unlikely to be limited for phosphate and so this form of accumulation is
probably not applicable. The second mode of phosphate accumulation,
which is often referred to as "luxury uptake" (Harold, 1966; Deinema et aI.,
1980, 1985), involves the uptake of Pi by certain microorganisms under
conditions of either balanced growth or, more frequently, under conditions
of nutrient limitation.
A number of reports have demonstrated that effective phosphate

removal from activated sludge effluents requires alternating aerobic and
anaerobic cycles (Fuhs and Chen, 1975; Timmerman, 1979; Deinema et al.,
1985; Fukase et al., 1985; Suresh et al., 1985). Examination of activated
sludge samples from these sewage treatment plants for prevalent strains
demonstrated that the dominant organism is Acinetobacter,either
calcoaceticus or lwoffii (Fuhs and Chen, 1975; Buchan, 1983; Marais et
aI., 1983; Lotter, 1985; Lotter and Murphy, 1985; Murphy and Lotter, 1986;
432
Hao .and Chang, 1987). Under aerobic conditions the Acinetobacter
strains store the accumulated phosphate as Pn, which is subsequently broken
down and released as Pi in response to the anaerobic treatment. Suresh et
al. (1985) studied a strain of A. twoffii isolated from a sewage treatment
plant, using 31P_NMR spectroscopy, and demonstrated that both the
phosphate released during the anaerobic phase and that originally present in
the feed were accumulated when the cells were subsequently exposed to
aerobic conditions, confirming the previous observations (Fuhs and Chen,
1975; Buchan, 1983; Hao and Chang, 1987). However, NMR analysis
(Suresh et al., 1985) monitors only the surface (periplasmic) water-soluble
Pn granules which are dispersed by membrane impermeable chelators such
as EDTA (Halvorson et al., 1987) and cannot detect the large cytoplasmic
volutin granules. It was hypothesised (Suresh et al., 1985), and subsequently
confirmed (Halvorson et al., 1987), that the Pn is maintained in two distinct
pools, and that only the periplasmic polyphosphate is mobilised and released
during the aerobic-anaerobic transition. The cytoplasmic volutin granules
remained relatively inactive metabolically during this period. Surface Pn
breakdown occurs under metabolic stress, e.g. ATP limitation, which would
be expected to occur under anaerobic conditions (Acinetobacter strains
are strictly aerobic). According to Fuhs and Chen (1975), during the aerobic
period a natural enrichment for Acinetobacter occurs because of their
ability to grow on fermentation products such as ethanol, organic acids and
alcohols, which accumulate as a result of glucose metabolism during the
anaerobic cycle. Recently, van Groenestijn et al. (1989) determined the
optimum conditions for Pn accumulation in various Acinetobacter spp.
isolated from sewage treatment plants. Pn accumulation was stimulated by
streptomycin and inhibited by energy poisons. Maximum accumulation
occurred when the energy source was in excess and the cells were limited for
sulphur at low growth rates, or when nitrogen or sulphur was depleted during
the stationary phase. Moreover, Pn accumulation appeared to be relatively
insensitive to changes in pH over a range of 6.5-9.
CONCLUDING REMARKS
In this review we have presented several lines of evidence pointing to

the advantages (both potential and actual) of Acinetobacter as an
industrial microorganism. The unique cell surface properties of many
acinetobacters, together with their ability to compete in Nature once a
selective advantage is presented (e.g. sewage treatment plants), offers the
possibility of isolating surface active polymers, many of which exhibit specific
interactions at interfaces. Moreover, experience with large-scale
production at both pilot and industrial levels has demonstrated that
Acinetobacter is easily handled in the factory, quite stable in its growth and
production characteristics, and lends itself to strain improvement. Finally,
the development of suitable recombinant techniques for use with strains of
industrial importance is likely to lead to an understanding of the underlying
mechanisms of product synthesis and regulation, and offers an opportunity
for producing second and third generation products.
433
ACKNOWLEDGEMENTS
The authors are indebted to Rina Avigad for excellent technical and
administrative assistance for a large portion of the work carried out in the
laboratory of DLG. The work done in Tel Aviv was supported in part by a
grant from Petroleum Fermentations, Inc.
REFERENCES
Atlas, R.M., ed., 1984, "Petroleum Microbiology," Macmillan, New York.

Baumann, P., 1968, Isolation of Acinetobacter from soil and water,
1. Bacteriol., 96:34.
Bayer, E.A., Skutelsky, E., Goldman, S., Rosenberg, E., and Gutnick, D.L.,
1983, Immunochemical identification of the major cell surface
agglutinogen of Acinetobacter calcoaceticus RAG-92,
1. Gen. Microbiol., 129:1109.
Beardmore-Gray, M., and Anthony, c., 1983, The absence of quinoprotein
alcohol dehydrogenase in Acinetobacter calcoaceticus,
1. Gen. Microbio!., 129:2979.
Beerstecher, E., 1954, "Petroleum Microbiology," Elsevier, New York.
Belsky, I., Gutnick, D.L., and Rosenberg, E., 1979, Emulsifier of
Arthrobacter RAG-I, determination of emulsifier-bound fatty
acids, FEBS Lett., 101:175.
Buchan, L., 1983, Possible biological mechanism of phosphorous removal,
WaterSci. Tech., 15:87.
Cain, R., 1987, Biodegradation of anionic surfactants,
Biochem. Soc. Trans., 15S:7S.
Deinema, M.H., Habets, L.H.A., Scholten, J., Turkstra, E., and Webers,
H.A.A.M., 1980, The accumulation of polyphosphate in
Acinetob ac ter spp., F EM S Micro bioI. Lett., 9:275.
Deinema, M.H., van Loosdrecht, M., and Scholten, A., 1985, Some
physiological characteristics of Acinetobacter spp. accumulating
large amounts of phosphate, Water Sci. Tech., 17:119.
Duine, J.A., and Frank, J., 1981, Quinoprotein alcohol dehydrogenase from a
non-methylotroph Acinetobacter calcoaceticus, 1. Gen.
Microbiol., 122:20l.
Duine, J .A., Frank, J., and van Zeeland, J .K., 1979, Glucose dehydrogenase
from Acinetobacter calcoaceticus: a quinoprotein, FEBS
Lett., 108:443.
Fattom, A., and Shilo, M., 1984, Hydrophobicity as an adhesion mechanism
of benthic Cyanobacteria, Appl. Environ. Microbio!., 47:135.
Fixter, L.M., and Nagi, M.N., 1984, The presence of an NADP-dependent
alcohol dehydrogenase activity in Acinetobacter
calcoaceticus, FEMS Microbiol. Lett., 22:297.
Fuhs, G.W., and Chen, M., 1975, Microbiological basis of phosphate removal
in the activated sludge process for the treatment of wastewater,
Microb. Ecol., 2:119.
434
Fukase, T., Shibata, M., and Miyaji, Y., 1985, The role of an anaerobic stage
on biological phosphorous removal, Water Sci. Tech., 17:69.
Gibbons, R.J., 1980, Adhesion of bacteria to the surfaces of the mouth, in
"Adsorption of Microorganisms to Surfaces," p.351, R.C.W.
Berkeley, J.M. Lynch, J. Melling, P.R. Rutter and B. Vincent, eds.,
Ellis Horwood, Chichester.
Gibbons, R.J., and Etherden, 1., 1983, Comparative hydrophobicities of oral
bacteria and their adherence to salivary pellicIes, I nf ec t.
Immun., 41:1190.
Goldman, S., Shabtai, Y., Rubinovitz, c., Rosenberg, E., and Gutnick, D.L.,
1982, Emulsan in Acinetobacter calcoaceticus RAG-I,
distribution of cell-free and cell-associated cross reacting
material, Appl. Environ. Microbiol., 44:165.
Goosen, N., Vermaas, A.M., and van de Putte, P., 1987, Cloning of the genes
involved in synthesis of coenzyme pyrroloquinoline-quinone from
Acinetobacter calcoaceticus, J. Bacterio!., 169:303.
Goosen, N., Horsman, H.P.A., Huinen, R.G.M., and van de Putte, P., 1989,
Acinetobacter calcoaceticus genes involved in biosynthesis of
the coenzyme pyrrolo-quinoline-quinone: nucleotide sequence and
expression in Escherichia coli K-12, 1. Bacteriol., 171:447.
Gutnick, D., 1984, Biosurfactants in the oil industry, World Biotech
Report, 2:645.
Gutnick, D.L., 1987, The emulsan polymer: Perspectives on a microbial
capsule as an industrial product, Biopolymers, 26S:223.
Gutnick, D.L., and Minas, W., 1987, Perspectives on microbial surfactants,
Biochem. Soc. Trans., 15S:22.
Gutnick, D.L., and Rosenberg, E., 1977. Oil tankers and pollution, a
microbiological approach, Ann. Rev. Microbiol., 31:379.
Gutnick, D.L., and Shabtai, Y., 1987, Exopolysaccharide emulsifiers, in
"Biosurfactants and Biotechnology," p.211, N. Kosaric, W.L. Cairns
and N.C.C. Gray, eds., Marcel Dekker, New York.
Gutnick, D.L., Reddy, P.G., Allon, R., Rusansky, S., and Avigad, R., 1987,
Cloning and expression in E. coli of Emulsan related genes from
the oil-degrading bacterium Acinetobacter calcoaceticus
RAG-I, in "Genetics of Industrial Microorganisms. Proceedings
of the Fifth International Symposium on the Genetics of Industrial
Microorganisms," p.313, M.Alacevic, D. Hranueli and Z. Toman,
eds., Ognjen Prica, Karlovac.
Halvorson, H.O., Suresh, N., Roberts, M.F., Coccia M., and Chikarmane,
H.M., 1987, Metabolically active surface polyphosphate pool in
Acinetobacter lwoffi, in "Phosphate Metabolism and Cellular
Regulation in Microorganisms," p.220, A. Torriani-Gorini, F.G.
Rothman, S. Silver, A. Wright and E. Yagil, eds., American Society
for Microbiology, Washington, D.C.
Hao, O.J., and Chang, C.H., 1987, Kinetics of growth and phosphate uptake
in pure culture studies of Acinetobacter Species, Biotech.
Bioeng., 24:819.
Harold, F.M., 1966, Inorganic polyphosphates in biology: structure,
metabolism and function, Bacteriol. Rev., 30:772.
435
Hayes, M.E., Hrebenar, K.R., Murphy, P.L., Futch, L.E., and Deal, J.F.,
1986, Combustion of viscous hydrocarbons, US Patent
4,618,348.
Hinchliffe, E., and Vivian, A., 1980, Naturally occurring plasmids in
Acinetobacter calcoaceticus: pAV1, a P class R factor of
restricted host range, 1. Gen. M icrobiol., 116:75.
Horowitz, A., Gutnick, D.L., and Rosenberg, E., 1975, Sequential growth of
bacteria on crude oil, Appl. Environ. Microbiol., 30;10.
Hoyt, J.W., 1985, Drag reduction in polysaccharide solutions, Trends
Biotech., 3:17.
Hunger, M., Schmucker, R., Kishan, V., and Hillen, W., 1990, Analysis and
nucleotide sequence of an origin of DNA replication in
Acinetobacter calcoaceticus and its use for Escherichia coli
shuttle plasmids, Gene, 87:45.
Juni, E., and Janick, A., 1969, Transformation of Acinetobacter calco-
aceticus (Bacterium anitratum), 1. Bacteriol., 98:281.
Kacholz, T., and Schlingmann, M., 1987, Possible food and agricultural
application of microbial surfactants, an assessment, in
"Biosurfactants and Biotechnology," p.183, N. Kosaric, W.L. Cairns
and N.C.C. Gray, eds., Marcel Dekker, New York.
Kaplan, N., and Rosenberg, E., 1982, Exopolysaccharide distribution of and
bioemulsifier production by Acinetobacter calcoaceticus BD4
and BD413, Appl. Environ. M icrobiol., 44:1335.
Kaplan, N., and Silverman, M., 1989, Analysis of cloned exopolysaccharide
genes and of "archived" DNA from Acinetobacter calco-
aceticus BD4, Abst. Ann. Meet. Amer. Soc. Microbiol.,
H130.
Kaplan, N., Rosenberg, E., Jann, B., and lann, K., 1985, Structural
studies of the capsular polysaccharide of Acinetobacter
calcoaceticus BD4, Eur. 1. Biochem., 152:453.
Kappeli, 0., and Finnerty, W.R., 1979, Partition of alkane by an extracellular
vesicle derived from hexadecane-grown Acinetobacter, 1.
Kappeli, 0., and Finnerty, W.R., 1980, Characteristics of hexadecane
partition by the growth medium of Acinetobacter, Biotech.
Bioeng.,22:495.
Karavaiko, G.!', Yurchenko, V.A., Remizov, V.I., and Klyushnikova, T.M.,
1986a, Release of hydrogen peroxide by the Acinetobacter
calcoaceticus culture for manganese leaching from oxides,
Microbiol. Zh., 48:41.
Karavaiko, G.!" Yurchenko, V.A., Remizov, V.I., and Klyushnikova, T.M.,
1986b, Manganese dioxide reduction by Acinetobacter
calcoaceticus cell-free extracts, Microbiologiya, 55:709.
Kosaric, N., Gray, N.C.C., and Cairns, W.L., 1987, Introduction:
Biotechnology and the surfactant industry, in "Biosurfactants and
Biotechnology," p.1, N. Kosaric, W.L. Cairns and N.C.C. Gray,
eds., Marcel Dekker, New York.
436
Kulaev, I.S., 1979, "Biochemistry of Inorganic Polyphosphates," John Wiley,
New York.
Kulaev, I.S., and Vagabov, V.M., 1983, Polyphosphate metabolism in
microorganisms, Adv. Microb. Physiol., 24:83.
Kupfer, D., and Zusman, D.R., 1984, Changes in cell surface hydrophobicity
of Myxococcus xanthus are correlated with sporulation-related
events in the developmental program, 1. Bacteriol., 159:776.
Lachica, R.V., and Zink, D.L., 1984, Plasmid-associated cell surface charge
and hydrophobicity of Yersinia enterocolitica, Infect.
Immun., 44:540.
Levin, G. V., and Shapiro, J., 1965, Metabolic uptake of phosphorous by
waste-water organisms, 1. Water Pollut. Control Fed., 37:800.
Levy, c., 1990, Using insertional mutagenesis to study genes involved in
polysaccharide synthesis and glucose utilization in Acinetobacter
calcoaceticus RAG-l and BD4, MSc Thesis, Tel Aviv
University, Israel.
Lotter, L.H., 1985, The role of polyphosphate metabolism in enhanced
phosphorous removal from the activated sludge process, Water
Sci. Tech., 17:127.
Lotter, L.H., and Murphy, M., 1985, The identification of heterotrophic
bacteria in an activated sludge plant with particular reference to
polyphosphate accumulation, Water SA, 11:179.
Marais, G.V.R., Loewenthal, R.E., and Siebritz, J.P., 1983, Observations
supporting phosphate removal by biological excess uptake - a
review, Water Sci. Tech., 115:15.
Miller, J.J., 1984, In vitro experiments concerning the state of
polyphosphate in the yeast vacuole, Can. 1. Microbiol., 30:236.
Minas, W., Sahar, R, and Gutnick, D., 1988, Flow cytometric screening and
isolation of Escherichia coli clones which express surface
antigens of the oil-degrading microorganism Acinetobacter
calcoaceti cus RAG-I, Arch. M i crob iol., 150:432.
Murphy, M., and Lotter, L.H., 1986, The effect of acetate and succinate on
polyphosphate formation and degradation in activated sludge, with
particular reference to Acinetobacter calcoaceticus, Appl.
M icrobiol. Biotech., 24:512.
Ofek, I., Whitnack, E., and Beachey, RH., 1983, Hydrophobic interactions of
group A streptococci with hexadecane droplets, 1. Bacteriol.,
154:139.
Pines, 0., and Gutnick, D.L., 1981, Relationship between phage resistance
and emulsan production; interaction of phages with the cell
surface of Acinetobacter calcoaceticus RAG-I,
Pines, 0., and Gutnick, D.L., 1984a, Specific binding of a bacteriophage at a
hydrocarbon-water interface, 1. Bacteriol., 157:179.
Pines, 0., and Gutnick, D.L., 1984b, Alternate hydrophobic sites on the cell
surface of Acinetobacter calcoaceticus RAG-I, FEMS
437
Pines, 0., and Gutnick, D.L., 1986, A role for emulsan for growth
of Acinetobacter calcoaceticus on crude oil,
Appl. Environ. Microbial., 51:66l.
Pines, 0., Bayer, E.A., and Gutnick, D.L., 1983, Localisation of emulsan-like
polymers associated with the cell surface of Acinetobacter
calcoaceticus, J. Bacterial., 154:893.
Pines, 0., Shoham, Y., Rosenberg, E., and Gutnick, D.L., 1988, Unmasking
of surface components by removal of cell-associated emulsan from
Acinetobacter sp. RAG-I, Appl. Microbial. Biotech., 28:93.
Ramphal, R., and Kluge, R.M., 1979, Acinetobacter calcoaceticus
variety anitratus: an increasing nosocomial problem, Amer. J.
Med. Sci., 277:57.
Reddy, P.G., Allon, R., Mevarech, M., Mendelovitz, S., Sato, Y.,
and Gutnick, D.L., 1989, Cloning and expression in
Escherichia coli of an esterase-coding gene from the oil-
degrading bacterium Acinetobacter calcoaceticus RAG-I,
Gene, 76:145.
Reisfeld, A., Rosenberg, E., and Gutnick, D.L., 1972, Microbial degradation
of crude oil: factors affecting the dispersion in sea water by mixed
and pure cultures, Appl. Microbial., 24:363.
Rosenberg, E., 1986, Microbial surfactants, CRC Crit. Rev.
Biotech., 3:109.
Rosenberg, E., and Gutnick, D.L., 1981, The hydrocarbon-oxidising bacteria,
in "The Procaryotes: A Handbook on Habitats, Isolation &
Identification of Bacteria," p.903, M.P. Starr, H. Stolp, H.G.
Trueper, A. Balows and H.G. Schlegel, eds., Springer Verlag,
Heidelberg.
Rosenberg, M., and Kjelleberg, S., 1986, Hydrophobic interactions: role in
bacterial adhesion, Adv. M icrob. Ecol. 9:353.
Rosenberg, M., and Rosenberg, E., 1985, Bacterial adherence at the
hydrocarbon-water interface, Oil Petrochem. Pollut., 2:155.
Rosenberg, E., Perry, A., Gibson, D., and Gutnick, D.L., 1979a, Emulsifier of
Arthrobacter RAG-I: specificity of hydrocarbon substrate, Appl.
Environ. Microbia!., 37:409.
Rosenberg, E., Zuckerberg, A., Rubinovitz, c., and Gutnick, D.L., 1979b,
Emulsifier of Arthrobacter RAG-I: isolation and emulsifying
properties, Appl. Environ. Microbia!., 37:402.
Rosenberg, M., Gutnick, D., and Rosenberg, E., 1980, Adherence of bacteria
to hydrocarbons: a simple method for measuring cell-surface
hydrophobicity. FEMS l'vlicrobiol. Lett., 9:29.
Rosenberg, M., Perry, A., Bayer, E.A., Gutnick, D.L., Rosenberg, E., and
Ofek, 1., 1981, Adherence of Acinetobacter calcoaceticus
RAG-l to human epithelial cells and to hexadecane, I nf ec t.
Immun., 33:29.
Rosenberg, M., Bayer, E.A., Delarea, J., and Rosenberg, E., 1982, Role of
calcoaceticus on hexadecane, Appl. Environ. Microbia!.,
44:929.
438
Rosenberg, E., Gottlieb, A., and Rosenberg, M., 1983a, Inhibition of
bacterial adherence to epithelial cells and hydrocarbons by
emulsan, Infect. Immun., 39:1024.
Rosenberg, M., Judes, H., and Weiss, E., 1983b, Cell surface hydrophobicity
of dental plaque microorganisms in situ, I nf ect. I mmun., 42:831.
Rosenberg, E., Kaplan, N., Pines, 0., Rosenberg, M., and Gutnick, D., 1983c,
Capsular polysaccharides interfere with adherence of
Acinetobacter calcoaceticus to hydrocarbons, FEMS
Rosenberg, M., Rosenberg, E., Judes, H., and Weiss, E., 1983d, Bacterial
adherence to hydrocarbons and to surfaces in the oral cavity,
Rosenberg, E., Rubinovitz, c., Gottlieb, A, Rosenhak, S., and Ron, E.Z.,
1988a, Production of biodispersan by Acinetobacter
calcoaceticus A2, Appl. Environ. Microbiol., 54:317.
Rosenberg, E., Rubinovitz, c., Legmann, R., and Ron, E.Z., 1988b,
Purification and chemical properties of Acinetobacter
calcoaceticus A2 biodispersan, Appl. Environ. Microbiol.,
54:323.
Rosenberg, E., Schwartz, Z., Tenenbaum, A., Rubinovitz, c., Legmann, R.,
and Ron, E.Z., 1989, A microbial polymer that changes the surface
properties of limestone: effect of biodispersan in grinding
limestone and making paper, 1. Disp. Sci. Tech., 10:241.
Rubinovitz, c., Gutnick, D.L., and Rosenberg, E., 1982, Emulsan production
by Acinetobacter calcoaceticus in the presence of
chloramphenicol, 1. Bacteriol., 152:126.
Rusansky, S., 1985,Growth and emulsification of crude oil by R57, a
multiplasmid-containing strain of Acinetobacter, MSc Thesis,
TelAviv University, Israel.
Rusansky, S., Avigad, R., Michaeli, S., and Gutnick, D.L., 1987, Involvement
Acinetobacter calcoaceticus R57, Appl. Environ.
Microbiol.,53:1918.
Sar, N., and Rosenberg, E., 1983, Emulsifier production by Acinetobacter
calcoaceticus strains, Curro Microbiol., 9:309.
Sar, N., and Rosenberg, E., 1987, Drag reduction by Acinetobacter
calcoaceticus BD4, Appl. Environ. Microbiol., 53:2269.
Shabtai, Y., 1982, Enhancement of release of an amphipathic
exopolys'accharide from Acinetobacter calcoaceticus RAG-l,
PhD Thesis, Tel Aviv University, Israel.
Shabtai, Y., and Gutnick, D.L., 1985a, Exocellular esterase and emulsan
release from the cell surface of Acinetobacter calcoaceticus
RAG-I, 1. Bacteriol., 161:1176.
Shabtai, Y., and Gutnick, D.L., 1985b, Tolerance of Acinetobacter
calcoaceticus RAG-1 to the cationic surfactant
cetyltrimethylammonium bromide (CTAB) - role of the
bioemulsifier emulsan, Appl. Environ. Microbiol., 49:192.
Shabtai, Y., and Gutnick, D.L., 1986, Enhanced emulsan production in
mutants of Acinetobacter calcoaceticus RAG-l selected for
439
resistance to cetyltrimethylammonium bromide, Appl. Environ.
Microbiol., 52:146.
Shabtai, Y., Pines, 0., and Gutnick, D.L., 1985, Emulsan: a case study of
microbial capsules as industrial products,
Develop. I ndust. M icrob iol., 26:29l.
Shoham, Y., and Rosenberg, E., 1983, Enzymatic depolymerization of
emulsan, J. Bacteriol., 156:161.
Shoham, Y., Rosenberg, M., and Rosenberg, E., 1983, Bacterial degradation
of emulsan, Appl. Environ. M icrobiol., 46:573.
Singer, M.E., and Finnerty, W.R., 1984, Microbial metabolism of straight-
chain and branched alkanes, in "Petroleum Microbiology," p.1,
R.M. Atlas, ed., Macmillan, New York.
Singer, M.E., and Finnerty, W.R., 1985, Alcohol dehydrogenase in
Acinetobacter sp. strain HOI-N: Role in hexadecane and
hexadecanol metabolism, i. Bacteriol., 164:1017.
Suresh, N., Warburg, R., Timmerman, M., Wells, J., Coccia. M., Roberts,
M.F., and Halvorson, H.O., 1985, New strategies for the isolation
of microorganisms responsible for phosphate accumulation,
Water Sci. Tech., 17:99.
Taylor, W.H., and Juni, E., 1961a, Pathways for biosynthesis of a bacterial
capsular polysaccharide. L Characterisation of the organism and
polysaccharide, i. Bacteriol., 81:688.
Taylor, W.H., and Juni, E., 1961b, Pathways for biosynthesis of bacterial
capsular polysaccharide. II. Carbohydrate metabolism and
terminal oxidation mechanisms of a capsule producing coccus, i.
Bacteriol.,81:694.
Taylor, W.H., and Juni, E., 1961c, Pathways for biosynthesis of bacterial
capsular polysaccharide. Ill. Synthesis from radioactive
substrates,i. Bioi. Chem., 236:1231.
Tijssen, J.P.F., and van Steveninck, J., 1984, Detection of a yeast
polyphosphate fraction localized outside the plasma membrane by
the method of phosphorous-31 nuclear magnetic resonance,
Biochem. Biophys. Res. Commun., 119:447.
Timmerman, W.M., 1979, Biological phosphate removal from domestic waste
water using anaerobic/aerobic treatment, Devel. Indust.
Microbiol.,20:285.
Umnov, A.M., Steblyak, A.G., Umnova, S., Mansurova, S.E., and Kulaev, I.S.,
1975, Possible physiological role of the high molecular weight
polyphosphate phosphohydrolase system in Neurospora crassa,
Mikrobiologiya, 44:414.
van Groenestijn, J.W., Zuidema, M., Van de Worp, J.J.M., Deinema, M.H.,
and Zehnder, A.J.B., 1989, Influence of environmental parameters
on polyphosphate accumulation in Acinetobacter sp., Ant. v.
Leeuw., 55:67.
van Schie, B.J., van Dijken, J.P., and Kuenen, J.G., 1984, Non-coordinated
synthesis of glucose dehydrogenase and its prosthetic group in
Acinetobacter and Pseudomonas species, FEMS Microbiol.
Lett., 24:133.
440
van Schie, B.J., HeUingwerf, K.J., van Dijken, J.P., Elferink, M.G.L., van
Dijl, J.M., Kuenen, J.G., and Konings, W.N., 1985, Energy
transduction by electron transfer via a pyrroloquinoline-quinone-
Pseudomonas aeruginosa, and Acinetobacter
calcoaceticus, J. Bacteriol., 163:493.
Yurchenko, V.A., Karavaiko, G.I., Remizov, V.I., Klyushnikova, T.M., and
Tarusin, A.D .. 1987, Role of the organic acids produced by
Acinetobacter calcoaceticus in manganese leaching, Prikl.
Biochim. Mikrobiol., 23:404.
ZoBell, C.E., 1973, Microbial degradation of oil: present status, problem
and perspectives, in "The Microbial Degradation of Oil
Pollutants," p.3, Georgia State University, Atlanta.
Zosim, Z., Gutnick, D.L., and Rosenberg, E., 1982, Properties of
hydrocarbon-in-water emulsions stabilized by Aci ne to b a c ter
RAG-1 emulsan, Biotech. Bioeng., 24:281.
Zosim, Z., Gutnick, D.L., and Rosenberg, E., 1983, Uranium binding by
emulsan and emulsanosols, Biotech.Bioeng., 25:1725.
Zosim, Z., Rosenberg, E., and Gutnick, D.L., 1986, Changes in hydrocarbon
emulsification specifity of the polymeric bioemulsifier emulsan:
effects of alkanols, Colloid Polymer Sci., 264:213.
Zosim, Z., Gutnick, D.L., and Rosenberg, E., 1987, Effect of protein content
on the surface activity and viscosity of emulsan, Colloid Polymer
Sci., 265:442.
Zosim, Z., Fleminger, G., Gutnick, D., and Rosenberg, E., 1989, Effect of
protein on the emulsifying activity of emulsan, J. Disp.
Sci. Tech., 10:307.
Zuckerberg, A., Diver, A., Perri, Z., Gutnick, D.L., and Rosenberg, E., 1979,
Emulsifier of Arthrobacter RAG-I, chemical and physical
properties, Appl. Environ. Microbiol., 37:414.
441
INDEX
ACE enzymes, 121-30 Acinetobacter (continued)

Acinetobacter, see Taxonomy CCM 2355, 326, 334
A. cyclohexanophilus, 392 CCM 2376, 326, 334, 408
A. inotti, 118 CCM 5572, 326, 334
other named species, see Taxonomy CCM 5593, 260-4
3B-l,358 CCM 5594/5595, 326, 334
2CA2,332 DON2,334
4CB1,374 DY-l,366
6.3,398 EBI0C/WD,334
69-V, 260-6, 303-4, 325-41, 404-8 EBW2, 326, 334
73,357 EBW3,334
80-1,357 EBI04,326-39
210A,283 EBW6,326
A2,421-2 EB113/114, 326, 334
A6E,334 EBF65/61,370
AAC1,184 EBF65/65, 150-5, 193-7,358-63
Ac78,196 EBF65/174,360
ACJ2,150 GB1,11
ADPl, see BD413 H01-N, 8,150-3,162, 197,260,
AH60, 334 264,276-7,280-1,325-41,
ASl,361 422
A TCC 9036, 408 IFO 13006, 408
ATCC 15150, 195 JCI, 11
ATCC 17902, 279 JCI7, 156, 195
ATCC 17907, see NCTC 10306 JW11,283-4
ATCC 23055,9,279 LMD73.1,305
ATCC 31012, see RAG-1 LMD79.39,297
ATCC 33305, see BD413 (see also LMD79.41)
ATCC 39647/39648,407 LMD79.41, 196,296-8,303-5
ATCC 39769/39770, 407 LMD82.3, 305
BD4,420-5 NAV2,375
BD413, 25,150-4,162,172-7,184-5, NCIB 8250, 8-11,193-7,276-91,
192-3,197,325,332-4,357, 326,351,357-8,361,365-6,
417,421-5 370,391
BM2500,157 NCIB 10487,278
CW,392 NCIB 10533,326
443
Acinetobacter (continued) Anticancer agent,
NCIMB 9871, 392-9 see Glutaminase-
NCTC 10305, 265 asparaginase
NCTC 10306, 277 Aromatic amino acids, 375-7
0-16,262,326,340 Aromatic compounds,
P6,374 efficiency of utilisation, 352
RA57, 325, 417 metabolism of, 351-90
RAG-I, 78, 260, 332, 340, pathways for catabolism, 353
412-24,427 Aromatic hydrocarbons, 335
RJE 74,366 A TP production, 8, 304, 357
SAK,134-48 Auxotrophic enrichment, 196-7
TMS 262/266, 80
W45,159-60 Bacteriocin typing, 66
YON,135 Bacteriophages, 39, 65, 153, 195-6,
Acylcarnitines, 404 415,422-3
Adherence, 78, 331-2 Bacteriophage typing,
assay for, 419-20 see Phage typing
Airborne spread, 57 Baeyer-V illiger reaction, 392-9
Alcohol dehydrogenase, 281, 333-8, Base composition, 251
417-9,427 Batch cultnre, 8, 274-7, 283, 330
Aldehyde dehydrogenase, 333-8 Ben genes, 203, 210-1, 214-8
Aldose sugars, Benzaldehyde dehydrogenase, 362
acid formation from, 302 Benzene utilisation, 161,366
Alicyclic compounds, cloning of genes involved in, 163
metabolism of, 391-402 Benzoate dehydrogenase, 202
Aliphatic hydrocarbons, 324-5, 335 Benzoate metabolism, 203, 210-1,
Alkanaldehydrogenase,281 214-5,218,230,360-2,373
Alkane monooxygenase, 333-41 Benzoate oxygenase, 214
n-Alkanes, Benzyl alcohol dehydrogenase, 362
chain length specificity, 325-6 Benzyl alcohol pathway, 358-65
genes for assimilation, 324-5 Betaines, metabolism of, 403-10
metabolism of, 323-50 Betaketoadipate, see B-Ketoadipate
oxidation of, 276, 324-5, 333 Betalactam, see B-Lactam
uptake and assimilation of, Betalactamase, see B-Lactamase
331-42 Biochemical characterisation, 30
Alkanols, 276, 328 Biodeterioration, 366, 373-7
Alkenes, 328 Biodispersan, 421-2, 431
n-Alkyl hydroperoxide, 333 Bioemulsifying agents, 12,329-31,
Aminoglycoside modifying enzymes, 342,353,375,412-21
see Resistance (see also Emulsan)
Aminoglycoside resistance, 100-3, Biomass production, 14, 304
138-46 Biopolymers,411-31
Aminopeptidase, 263 Bioremediation, 366, 373, 377, 413
Amylase, 263 Biosurfactants, 428-31
Aniline, 367 Biosynthetic applications, 342
Anthranilate metabolism, 215, 376 Biotransformations, 12, 342
Antibiogram typing, 46, 65 Biotyping, 38, 65, 69
Antibiotic resistance, Butyrate, 286
see Resistance
444
Cadmium ion transport, 284 Colonisation, see Occurrence
Caprolactone, 392-4, 397-8 Competence, development of, 185
Capsular polysaccharide, 6, 78, Composition of outer membrane,
259,415,421-4 264-6
CARB enzymes, 89, 91 Conjugative transfer,
Carbon monoxide dehydrogenase, 11 of chromosomal DNA, 152, 193-6
L-Carnitine, 13 of plasmids, 150, 193
metabolism of, 13,403-10 Continuous culture, 8, 274, 283, 303
synthesis of, 407-8 Control of transcription, 145
CAT, see Chloramphenicol resistance Crotonobetaine, 403
Cat genes, 203-30 Crude oil degradation,
Catechol 1,2-dioxygenase, 202, see Oil degradation
224-8,356,366 CTX enzymes, 95
Catechol metabolism, 203, 225, Cycloalkanes, 391
334-7, 360-2, 366-7, Cyclohexane metabolism, 391-402
371-2 Cyclohexanecarboxylic acid, 398
CAZ enzymes, 95 Cyclohexanol, 391-3, 398
Cell envelope protein profiles, Cyclohexanone, 391-5, 398
29,43 Cyclohexanone 1,2-monooxygenase,
Cephalosporinase, see B-Lactamase 394-6
Chemotaxis, 353 gene coding for, 396
Chloramphenicol resistance, 137-48 Cycloisomerase, 202
4-Chlorobiphenyl, 161,373 Cyclopentanone, 397
Chlorocatechol, 225 Cytochromes, 9, 298, 325-6, 335-7,
Chlorocatechol oxygenase, 225-6 341-2
l-Chlorohexadecane, 330 b 562' 298
Chromosome mobilisation, 152, 193 P-450, 325-6, 334-7
Chromosome organisation, 7, 194 Cytoplasmic membranes, 340-1
Cinnamate, 370
Citrate synthase, 313 Decarboxylase, 202
Citric acid cycle, 313-31 Dental plaque, treatment of, 430-1
biosynthetic branch points, 319 Detergents, sensitivity to, 262
energy control of, 317 2,4-Dichlorophenoxyacetate, 373
multipoint control of, 320 DNA probes,
Classification, see Taxonomy for AAD(2"), 135
Clinical importance, 4-5, 54-9 for CATl, 134
Cloning, (see also Genes) DNA sequence exchange, 204, 219-28
of antibiotic resistance genes, DNA slippage structures, 204, 222-8
136-8 n-Dodecane, 324
of benzene utilisation genes, 163 Drag reduction, 431
of est genes, 416-9
of glucose dehydrogenase genes, Electron transport, 8, 334, 341
300-2 Emulsan, 331, 340, 412-20,429-30
of pea genes, 162 genes for production of, 416-9
Cloning vectors, see Vector plasmids Emulsification,
Co-factor, see PQQ see Bioemulsifying agents
Codon usage, 206-8, 217 -20, 241, Endemic infection, 54
251-8 Endotoxin, 80
Collagenase, 263 Energy conservation, 8, 303
445
Energy generation, 295 Genetic organisation, 191
Energy reserves, 10, 273, 330 of linkage map, 7, 152, 194
Enrichment culture, 4 of trp genes, 239
Enrichment for auxotrophs, 196-7 Genetic techniques for mutant
Entner-Doudoroff pathway, 302 analysis, 229-31
Environmental pollution, Genetically engineered
see Pollution control microorganisms, see GEMs
Environmental sources, see Sources Genomic species, see Taxonomy
Enzymes, electrophoretic analysis Genospecies, see Taxonomy
of,29-47 Gentisate, 367
Enzymology, 9-11 Glucanase, 263
Epidemic spread, 56 Gluconic acid, 295
Epidemiology, 4, 6, 53-76 Gluconolactone, 295
Est gene, 416-9 Glucose dehydrogenase, 13, 30,
Esterase,29,262,281,340,416 295-312,424
Ethers, 328-9 applications of, 300
Evolution, cloning of gene for, 300-2
of antibiotic resistance, 84 different forms of, 296
of B-ketoadipate genes, 201-37 electron acceptors for, 298, 302
of trp genes, 239 induction of, 305
Evolutionary divergence, genetic m-GDH,301
mechanism of, 221-8 physiological role of, 301-2
Exopolysaccharide biosynthesis, 427-8 properties of, 299
s-GDH,298
Fatty acids, Glutamate dehydrogenase, 11,29
accumulation of, 329 Glutaminase-asparaginase, 15
composition of, 331 Glycerides, 329, 340
production of, 282, 330, 337, 341 Glycogen, 274
respiration of, 336 Glycolipids, 340
Ferredoxin, 335-6 Glyoxylate bypass, 10
Ferredoxin reductase, 335-6 Glyoxylate effect, 316
Fimbriae, 78, 332, 412, 420 Growth energetics, 8
Fluoroquinolone resistance, 103 Growth yields, 8, 352
Food spoilage, 4 improvement of, 303
FPLC of B-Iactamases, 121
Habitats, see Occurrence
Gap repair, 231 Haem biosynthesis, 319
Gasworks liquor, 351, 366 Haemolysin, 262
GC content, 2, 205, 219-20 Haloalkanes, 328
GEMs, 7,169 Halogenated compounds,
Gene regulation, 7, 239 degradation of, 373-4
(see also Regulation) Heavy metal resistance, 159
Gene transfer, n-Heptadecane, 330
see Conjugative Herbicide degradation, 373
transfer, Hexadecan-1-ol,277
Transformation, n-Hexadecane, 324-30, 335,412
Transduction n-Hexadecanol,328-30
Genetic determinants of resistance, 1-Hexadecene, 330
108,133 n-Hexane, 335
446
Hospital epidemiology, Krebs TCA cycle, 10,313-22
see Epidemiology
Hydrocarbons, as growth B-Lactam resis tance, 88-100
substrates, 260 B-Lactamase, 5, 89-100,117-32,
Hydrophobic pathway, 260 263-4
Hydrophobicity, 78 induction of, 129-30
p-Hydroxybenzoate, 203, 229-30, inhibitors of, 95-7,128-9
367 kinetics of hydrolysis, 127-8
p-Hydroxybenzoate hydroxylase, 202 Lactate dehydrogenase, 361
Hydroxyphenylacetate,367-70 Lactone hydrolase, 222, 397-8
4-Hydroxyquinazoline,372 Lactones, 303, 397
B-Hydroxyvalerate,285 hydrolysis of, 304, 397
Hypersensitivity to hydrophobic Lignin compounds,
agents, 260 degradation of, 370-1
Hysteresis phenomenon, 317 Linkage map, 7, 152
Lipase, 263, 341
Identification, see Taxonomy selection of mutants, 186
Imipenem resistance, 97 Lipid A, 80
Inclusions, 280, 333, 422 Lipids,
Industrial applications, 323, 411-41 accumulation of, 329
Infections, 58 (see also Occurrence) metabolism of, 333, 340-1
Inhibitors of B-Iactamase, 95-7, Lipopolysaccharide, 6, 80, 100,260,
128-9 264-6, 331, 422
Insertional mutagenesis, 183-9,423-4
Introdiol dioxygenases, 223-5 Malate dehydrogenase, 29, 337
Invasiveness, 78 Malonate, 10
Iron uptake, 9, 79 Mandelate dehydrogenase, 357, 361
ISIS, 157 Mandelate pathway, 358-65
IS50, 184 regulation of, 362-3
Iso-alkanes, 328 Map of chromosome, 194
Iso-citrate lyase, 337 Marker plasmids, 170
Isoelectric focusing of B-Iactamases, stability of, 171
89,118-21 Marker rescue, 7
Isolation, 3-4 Mechanisms of resistance, 83-132
to aminoglycosides, 100-3, 139
KDO,265-6 to chloramphenicol, 137
B-Ketoadipate enol-lactone, to fluoroquinolones, 103-8
202-3,229 genetic determinants of, 108-10,
B-Ketoadipate pathway, 201-37, 255, 133
352,356-7 to heavy metals, 159
mutations in, 212-5 to imipenem, 97-100
selection of mutants in, 229-30 to B-lactams, 88-100, 117-32
structural genes of, 202-3 Membrane vesicles, 266
B-Ketoadipate succinyl CoA Mercury resistance, see Heavy metal
transferase, 202, 212-3, 218 resistance
B-Ketoadipyl CoA, 203 Metabolic plasmids, 160
B-Ketoadipyl CoA thiolase, 202, 212-3 Metabolism, 9
Ketone body utilisation, 319 Methylation, 375
Ketones, 328 MICs, 73, 86, 96-9,103-8,126,261
447
Modification, see Restriction- Oxygenative reactions, 352
modification system
Monooxygenases, 356, 394 (see also P78 bacteriophage, 196
Alkane monooxygenase) Particulate surfactant, 422
Morphology, 2 Patent literature, 12,329,412-3,
Motility, 2 430-1
Mu bacteriophage, 153, 196 Pathogenicity, 77
Muconate, 205, 214-5 Pathogenic role, 5
Muconolactone isomerase, 202-3 Pea genes, 153, 162, 197,203,
Multipoint control of citric acid 208-14,217-21,229-31
cycle, 320 PCBs, see Polychlorinated biphenyls
Multiresistance to antibiotics, 88 Penicillinase, see B-Lactamase
Mutagenesis, Peptidase, 263
chemical, 196 Peripheral pathways, 353-4, 357-8,
insertional, 183-9, 197,423-4 362
UV,196 Periplasmic enzymes, 262
Mutarotase, 255 Permeability of outer membrane,
81,99,152,260
Nomenclature, see Taxonomy Pesticide degradation, 373-4
Nonane, 327 Petroleum degradation,
Nosocomial infection, see Oil degradation
see Epidemiology Phage typing, 39,65
Nucleic acid metabolism, 11 Phagocytosis, 81
Nutrient limitation, 274 Phenols,
Nutritional studies, 25, 37 degradation of, 352, 366-7, 373-4
methylation of, 375
Occurrence, Phenylacetate, 367-70
in clinical samples, 4, 32-3, Phenylalanine, 367-70, 377
37,57 Phenylalkanes, 327-8, 367-70
in foodstuffs, 4, 33 Phenyldodecane, 370
in sewage, 3, 33, 282, 288 Phenylglyoxylate decarboxylase, 362
on skin, 4, 33, 54, 57, 64 Pheny lpropionate, 370
in soil, 3, 32-3 Phosphate overplus phenomenon,
in water, 3 275,432
Octadecan-1-ol,277 Phosphate removal, 432-3
n-Octadecane, 330 Phospholipase, 262, 340
1-0ctadecene,30 Phospholipids, 280, 329-31, 337, 340
Oil degradation and stabilisation, Physiological characterisation, 30
161,324-5,342,375, Plasmid profile analysis, 45, 66, 72,
416, 419-20, 429-30 154
Outer membrane, 81, 99,152,259, Plasmids
332,340 behaviour and stability of, 6, 149,
Outer membrane proteins, 151, 169, 184, 193
electrophoresis of, 29, 99 ColE 1, 183-4,423
OXA enzymes, 91 conferring antibiotic resistance,
Overflow metabolites, 273, 278, 305 108, 154
Oxidative phosphorylation, 8 conferring heavy metal resistance,
Oxoglutarate dehydrogenase, 317 159
Oxygenases, see Monooxygenases cryptic, 424-5
448
Plasmids (continued) Plasmids (continued)
involved in oil degradation, 325 R1822, 150
as markers, 170 RP 1, 100, 150-2
metabolic, 160,357,366-7,373 RP4, 110, 150-2, 158, 193-6
pA10cat2,134 RSF1010, 184
pAV1, 156, 194 P / 0 ratio, 8, 304, 352
pA V2, 150, 195 Pob genes, 203, 210-1, 229-30
pAV5/51/52, 156 Pollution control, 12, 160,282,323,
pAVI0/11,196 366,373-5,432-3
pBEE7/10/104,154 Polychlorinated biphenyls,
pBR322, 162, 185, 423, 427 degradation of, 160,373-4
pGSHI01,135 PolY-B-hydroxyalkanoates, 274
pGSH106,140 POlY-B-hydroxybutyrate, 10,273-5,
pGSH108/110, 138-42 278-9, 285-8
pGSH201, 137 metabolism of, 286
pHLl7/124/125/227/234/434,423 Polyphosphate, 10,273,282-8,411,
pIB1341/1343/1345/1362,215-6 432-3
pIP1031, 109, 157 accumulation of, 283, 432-3
p1P1841, 103, 109-10, 158 cellular distribution of, 283
pJB4JI,153 enzymes involved in metabolism
pKFl,160 of,284
pKK223-3, 396 Polysaccharide capsule, 6, 78, 259
pKLl, 159-60 PQQ, 11,30,296-307,424
pKT230, 162, 171, 178,217 assay for, 300
pKT231/248,162 Promoters,
pLV21,163 for Cat genes, 216-7
pLVI010/1011/1013/1016-7,170-8 regulation of, 142, 145
pQM17,160 for Trp genes, 244
pRA17,416-7 Pristane, 329
pRETl,417-9 Protein electrophoresis patterns,
pRK290, 162 27,43,71,99
pRK415,216 Protein export, 262
pRK2013, 154, 184 Protocatechuate, 211-8, 225, 229,
pRP3a/6a/7a, 418 354-7.363,367,371
pSRI-4,161 Protocatechuate 3,4-dioxygenase,
pSR4,325 202,212,217-8,224-30,356
pSS50, 161 Pyrroloquinoline quinone, see PQQ
pSUP2021,184 Pyruvate dehydrogenase, 318
pTB107,162
pUC4K,185 Quinate, 211, 229
pUC19,425 Quinoline, 372
pUN65,151 Quinolone resistance, 103-8
pUN308, 153, 196 Quinoprotein, 296 (see also PQQ)
pUZ12/13,109
pWM4/5/6,425 Rates of infection, 58
pWW174, 161,366 Recombinant DNA, see GEMs
R6K/R57b/R64, 151 Recovery of DNA by gap repair, 231
R68.45,193 Regulation,
R124/R388/R478/R821a, 151 of Ben genes, 214-5
449
Regulation (continued) Structural genes (continued)
of Cat genes, 205-10, 215-7 evolution of, 201-37
of promoter activity, 142, 145 Structure of outer membrane, 264-6
of T rp genes, 245 Substrate inhibition, 8
of wax ester content, 282 Substrate range, 352
Repair of mutations, Succinate thiokinase, 318
by DNA sequence exchange, 219 Suicide vector, 153
Resistance to antibiotics, 5, 46, Supraoperonic clusters, 204, 208-12,
73, 83-132, 158-9 228
(see also Mechanisms Surface-active compounds,
of resistance) see Bioemulsifying agents
Respiratory chain, 341 Surface polymers,
Respiratory efficiencies, 8 applications of, 428-31
Respiratory tract infection, 55-6, 64 Surfactants, see Biosurfactants
Restriction endonuclease analysis, Susceptibility to antibiotics, 86
46, 72 (see also MICs)
Restriction-modification system, 15
Ring cleavage, 353-4, 357-8 Taxonomy, 2-4, 25-36, 288
rRNA gene restriction patterns, 30 TEM enzymes, 89-91, 95, 99, 109
Rubredoxin, 335-6 Temperature for growth, 4, 30-1
Rubredoxin reductase, 335-6 n- Tetradecane, 330
Thioether, 328
Salicylate, 366 Toxigenicity, 79
Selection of mutants, Toxins, 80
see Auxotrophic Transcriptional control and
enrichment regulation, 145, 204-6, 210-1
Sequence exchange, see DNA Transcriptional repressor, 255
sequence exchange Transcriptional terminator, 216
Serological cross-reactions, 29 Transduction, 195
Serotyping, 43, 74 Transfer,
Sewage sludge ecology, 282, 432-3 of chromosomal DNA, 152
Sewage treatment, 282, 432-3 of plasmids, 150
Shikimate pathway, 211-3, 229, 375 Transformation, 185, 192,229-31,
Shuttle vectors, 162,239,424-8 274,421-2
(see also Vector selection of deficient mutants, 186
plasmids) Transport mechanisms, 8, 284, 304,
SHV enzymes, 95 354,358
Single cell protein, 342 Transposons,
S-layers, 259 behaviour of, 6, 149, 153
Slippage structures, conferring antibiotic resistance,
see DNA slippage 145, 158
structures indigenous, 158
Sources, see Occurrence mutagenesis with, 153, 184
Spoilage of foodstuffs, 4 Tn1, 152
Stability of plasmids, 151, 169 Tn5, 153-5,162,184-5,197
(see also Plasmids) Tn7, 153-5
Strain improvement, 419 Tnl0,154-5
Structural genes, Tn3171, 153-5, 196
codon usage in, 251-4 n-Tridecane, 324
450
Trimethylammonium compounds, Vibrio strain 01, 351 (see also
metabolism of, 403-1 ° Acinetobacter NClB 8250)
Trp genes, Virulence, 6, 77-82
genetic organisation of, 192, 239 Volutin granules, 282
nucleotide sequence of, 242
promoters of, 244 Waste disposal, 323, 375
regulation of, 245 (see also Pollution control)
Tryptophan, Wax esters, 10, 13,273,276-82,
biosynthesis of, 377 287-8,323,329-30,
metabolism of, 376 333,342
Typing, 6,37-51, 65, 69 accumulation of, 330
Tyrosine biosynthesis, 377 composition of, 276
degradation of, 279
Ubiquinone, 302-4 metabolism of, 281
Urinary tract infection, 58 regulation of content, 282
Vector plasmids, 153, 162-3, 184, X ylE marker system, 171

239 (see also Suicide expression and regulation of, 176
vector)
Vesicles, 331-2, 340, 422 Yield, see Growth yields
451

The Biology of Acinetobacter Taxonomy, Clinical Importance, Molecular Biology, Physiology, Industrial Relevance by K. J. Towner, E. Bergogne-Bérézin, C. A. Fewson (Auth.), K. J. Towner, E. Bergogne-B

Uploaded by

Copyright:

Available Formats

The Biology of Acinetobacter Taxonomy, Clinical Importance, Molecular Biology, Physiology, Industrial Relevance by K. J. Towner, E. Bergogne-Bérézin, C. A. Fewson (Auth.), K. J. Towner, E. Bergogne-B

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Biology of Acinetobacter Taxonomy, Clinical Importance, Molecular Biology, Physiology, Industrial Relevance by K. J. Towner, E. Bergogne-Bérézin, C. A. Fewson (Auth.), K. J. Towner, E. Bergogne-B

Uploaded by

Copyright:

Available Formats

The Biology of

1990 • MICROBIOLOGY AND BIOCHEMISTRY OF STRICT ANAEROBES

1990 • DENITRIFICATION IN SOIL AND SEDIMENT

1991 • CANDIDA AND CANDIDAMYCOSIS

1991 • MICROBIAL SURFACE COMPONENTS AND TOXINS

1991 • GENETICS AND PRODUCT FORMATION IN STREPTOMYCES

1991 • THE BIOLOGY OFACINETOBACTER: Taxonomy, Clinical Importance,

Springer Science+Business Media, LLC

Based on the proceedings of a symposium held under the auspices of the

ISBN 978-1-4899-3555-7 ISBN 978-1-4899-3553-3 (eBook)

©1991Springer Science+Business Media New York

All rights reserved

The 1st International Workshop on Acinetobacter was held on 6th

In addition to sponsorship by the Federation of European

Acinetobacter: Portrait of a Genus 1

Hospital Epidemiology of Acinetobacter Infection 53

Epidemiology of Acinetobacter Strains Isolated from

Molecular Epidemiological Analysis of Nosocomial

Factors Influencing the Virulence of Acinetobacter 77

Antibiotic Resistance Mechanisms in Acinetobacter 83

The Chromosomal B-Lactamases of the Genus Acinetobacter:

The Use of Molecular Techniques for the Location and

Plasmid Stability and the Expression and Regulation

Random Insertional Mutagenesis in Acinetobacter

Genetic Organisation of Acinetobacter

Evolution of Genes for the B-Ketoadipate Pathway

Organisation, Potential Regulatory Elements and

Codon Usage in Acinetobacter Structural Genes

The Outer Membrane of Acinetobacter: Structure-

Energy Reserves in Acinetobacter

Energy Generation and the Glucose Dehydrogenase

Acinetobacter - Citric Acid Cyclist with a Difference

Metabolism of Alkanes by Acinetobacter

Metabolism of Aromatic Compounds by Acinetobacter

Metabolism of Cyclohexane and Related Alicyclic

Applications of Acinetobacter as an Industrial

K.J.Towneru, E.Bergogne-Berezinband C.A.Fewson'

'Department of Microbiology & PHLS Laboratory

The genus Acinetobacter originally proposed by Brisou and Prevot

Why the interest in this genus? Acinetobacters are ubiquitous

IDENTIFICATION AND TAXONOMY

Morphological and Biochemical Identification

Members of the genus Acinetobacter are short, plump, Gram-

Members of the genus Acinetobacter have suffered a long history of

OCCURRENCE AND ISOLATION

Strains of Acinetobacter can be obtained easily from soil, water and

Isolation of acinetobacters can be achieved on standard laboratory

CLINICAL IMPORTANCE OF ACINETOBACTER SPP.

Extent of the Problem

Strains of Acinetobacter are distributed widely in the hospital

The severity and difficulties associated with Acinetobacter

The virulence of Acinetobacter has been studied primarily in animal

Typing and Epidemiology

Analysis of the cell structure of acinetobacters has also provided the

GENETICS AND MOLECULAR BIOLOGY

Plasmid and Transposon Behaviour

There is still a relative paucity of information about the detailed role

Chromosome Organisation and Gene Regulation

Acinetobacter has been shown to have a circular chromosomal