IMMUNODIAGNOSTIC TECHNIQUES FOR BACTERIAL INFECTIONS

I E. A. EDWARDS

M. W. RYTEL

R. L. HILDERBRAND

Lo .91
REPORT NO. 77-39 4

:::> NAVAL HEALTH RESEARCH CENTER

C-:l P. 0. BOX 85122
L. SAN DIEGO, CALIFORNIA 92138

NAVAL MEDICAL RESEARCH AND DEVELOPMENT COMMAND

S82 2 08 0
BETHESDA, 034,
02MARYL0

A.
4
-Z
 IMMUNODIAGNOSTIC TECHNIQUES FOR BACTERIAL INFECTIONS-

Earl A. Edwards, M.S.t tt

Michael W. Rytel, M. D.tt

Richard L. Hilderbrand, Ph.D.t

Naval Health Research Center

P.O. Box 8S122
San Diego, California 92138

and

The Medical College of Wisconsin

Milwaukee, Wisconsin 53226

Published in: A Balows & WJ Hausler, Jr. (eds), Diagnostic Procedures

for BacteriaZ, Mycotic, and Parasitic Infections, 6th Ed. Wash-
ington, DC: American Public Health Assn, 1981. pp 767-790.

* Report No. 77-39, supported by the Naval Medical Research and Development
Command, under research work unit M009S.PN.002-5044. The views presented in
this paper are those of the authors. No endorsement by the Department of

the Navy has been given or should be inferred.

t Mr. Edwards is Head of the Biological Sciences Department, Naval Health Research

Center, and Dr. Hilderbrand (LT MSC USN) was head of the Biochemistry Branch,
Biological Sciences Department, Naval Health Research Center.
ttDr. Rytel is Head, Section of Infectious Diseases, The Medical College of

Wisconsin.
 OVERVIEW

Immunodiagnostic techniques for detection of antigen or antibody immediately

imply a variety of applications of immunological and immunochemical methods to
the diagnosis of microbial diseases. Immunodiagnosis lends itself to the study
of human diseases because of the sensitivity and specificity that specific immune
sera (the immune response) provides. The early use of immunological diagnosis
was to demonstrate circulating antibodies to disease producing microbial agents.
This was done directly (Widal test) and by showing a rising titer between serum
samples taken in the acute phase compared to those taken during convalescence.
The data from such testing and their reliability in diagnosis of various micro-
bial diseases has formed a solid cornerstone from which a much broader and more
sophisticated technology has developed.

The initial application of precipitin reaction in the diagnosis of infectious

diseases goes back to 1909 when Vincent and Bellot described detection of
meningococcal antigen in cerebrospinal fluids of patients with meningitis.(1)
Dochez and Avery in 1917 were the first to identify pneumococcal antigen in sera
(2)
and urines of patients with lobar pneumonia.
However, its been the data collected over the past 2 decades that has shown
that specific antigen-antibody complexing can be a powerful tool in identifying
either antigen or antibody. A thorough description of this reaction was made by
Heidelberger and Kendall in the mid 1930's. Landsteiner and Pauling also played
a major role in our understanding of the precipitin reaction in the late 30's
and early 40's. That the precipitin reactions could occur and be visualized in
gels was described by Ouchterlony and by Oudin in the late 40's.

The technique described by Ouchterlony has been widely used to identify

multiple antigen and antibody systems. The method also allows on- to show the
immunological relationship between two antigens by the line confluence when two

I
antigen preparations are used. The Ouchterlony technique has been extensively
used to determine antigen or antibody purity, detect antigen in biological
secretions and to quantitate either antigen or antibody when needed. Elik was
one of the first to use the gel/precipitin reaction as a diatnJSLic tool to

identify pathogenic strains of diptheria.

,1 __ _ • " ""i -- I HI I 1
Edwards, et al.

This chapter will deal with several of the technologies that have been
developed and have proven useful during the past decade in identifying either
antigen or antibody. It will give more emphasis to identifying antigens because
of the urtency of early identification of microbial agents for proper chemotherapy.
In addition, the sensitivity of these techniques allows identification of pico-
gram amounts of antigen, which are the levels that are frequently found in
biological secretions during acute illnesses.

Techniques to be described will include the double diffusion test (Ouchter-

lony), counterimmunoelectrophoresis (CIE), rocket immunoelectrophoresis (Laurell),
specifically sensitized Protein A-containing Staphylococcus aureus (Kronvall),
radioimmunoassay (Catt), and enzyme-linked immunosorbent assay (Engvall).

DOUBLE DIFFUSION OR OUCHTERLONY TEST

The double diffusion test described by Ouchterlony makes use of the fact
that soluble antigens and their antibodies will precipitate in a gel and the
precipitate is permeable to all other antigens and antibodies that have no points
of antigenic similarity with the precipitating pair. Thus, identical antigens
or antibodies diffusing from two or more wells will form a precipitating band
with an angle with one another that will fuse (lines of identity). By the same
reasoning, antigens of different specificities will form an angle with one another
that will cross (non-identity) or will partly cross and partly fuse (partial
identity) (See Fig. 1 on pg 3).

The Ouchterlony test has been used to study microbial antigens and for the
detection of antibodies. The test is relatively insensitive compared to other
tests available and has not had broad clinical application but has proven very
useful in research. It also suffers from the fact that it takes at least 24-28
hours for the results to become available. However, use has been made of the
test for presumptive diagnosis and therefore a methodology for the macro double-
diffusion test is described.

Materials:
1. Agarose
2. 0.15 M sodium chloride (saline)
3. Microscope slide (cleaned with Baboo and then with 95% ethanol,
towel dry)
4. Well punch set (#51450 Gelman or individual well cutters 3.0 mm
diameter)

2
 Edwards, et al.

AcCO
t-r
t
" For -
5. Pasteur pipettes or capillary tubes
T.13
6. Moist chamber

ry~ ~1-------

2 AvtS

Fig. 1 A graphic preservation of patterns of precipitation that may be seen

Iusing theOuchterlony technique to identify or compare antigens. Well #2 represents
antiserum; wells 1, 4, 5, and 6 represent antigen preparations. Top: reaction
of non-identity; middle: reaction of identity; bottom: reaction of partial
identity.

3
Edwards, et al.

Method:

1. Prepare 1% agarose in saline, bring to boil to dissolve agarose,

overlay 2-2.5 ml onto a clean microscope slide.

2. Allow agarose to solidify (1 hr in moist chamber).

3. Punch desired number of wells in desired pattern. The distance

between wells range from 205 mm. A circular pattern around a central well lends
itself to many testing requirements.

4. Carefully remove punched plugs by suction.

5. Fill the central well with immune serum and the surrounding wells
with test material (spinal fluid, serum, sputum, etc.).

6. Place in a moist chamber at room temperature for 24-72 hours.

7. Read the slides by use of oblique illumination. A hand lens 3-7X

is highly recommended.
8. The precipitin bands may be easier to interpret after appropriate
staining. Should this be necessary, the following steps
are recommended.
a. Rinse or soak the slides (step 7 above) in a saline solution
at room temperature for 2-3 days, making daily changes of the saline bath. This
procedure removes the escess antigen and serum proteins.
b. Soak in distilled water for 1-2 hours.
c. Place the washed slides in either amido black or coomassie
blud for 30 minutes.
d. Destain until precipitin bands are readily differentiated from
the background.
e. Alternative to step c. Place slides at 37*C until completely
dry. Proceed with step c.

Staining Solution:

Amido Black or Coomassie Brilliant Blue R-250 5.0 gram

Ethanol 95% 450.0 ml
Glacial Acetic Acid 100.0 ml
Distilled Water 450.0 ml

4
Edwards, et al.

Destaining Solution:

Ethanol 95% 450.0 ml

Glacial Acetic Acid 100.0 ml
Distilled Water 450.0 ml

Caution: Unless the glass surface of the slides is carefully cleaned, the
agarose may float off the slide during step 8.a. It is advisable to
precoat the clean slide with a thin layer of a .2% agarose, allow to
air dry and further super dry the slides for several days in a jar

with CaCl 2 . This provides a bonding for the fresh agarose layer
applied for double diffusion testing and reduces the chance of having
the agarose float off the glass surface upon washing.

COUNTERIMMUNOELECTROPHORESIS (CIE)

Tiselius pioneered the separation of proteins by their moving boundries

which was carried out in a liquid phase system, however, his (Tiselius and Flodein)
zone electrophoresis technology in the early 1950's set the stage for the immuno-
electrophoresis techniques which were to follow in the late 50's and throughout
the 60's. Immunoelectrophoresis was first described by Williams and Grabar in
the mid 50's.( Application of this technique allowed one to separate and
identify mixtures of antigens and was especially used to study serum proteins
and their mobilities in an electrical field. It became customary to carry out
electrophoresis of serum proteins at pH 8.6 during this early stage of studying
protein separation because most proteins are negatively charged at this pH and
will migrate to the anode. The exception to this was the case of the three major
classes of immunoglobulins. IgM and IgA remained either at the point of appli-
cation or migrated toward the cathode. IgG, being made up of a heterogenous group
of protein moieties, shows a broad range of movement, moving both anodal and
cathodal. This (cathodal movement) was found to be due to the impurities in the
gel (agaropectins) which have a strong negative charge. When in an electrical
field, these "fixed" negative charges surround themselves with positive charges
from the anode, thusmoving the liquid buffer from the anode toward the cathode.
This movement of buffer is called electroosmotic flow or endosmosis, and is
effected by agar purity, concentration, thickness, buffer molarity and pH as
well as by the voltage at which the electrophoresis is performed. These aspects
will be addressed in greater detail below.

5i
Edwards, et al.

The endosmotic effect "carries" the relatively "neutral" charge of the

immunoglobulins toward the cathode. This characteristic was used in developing
the counterimmunoelectrophoresis (CIE) test as first described by Brussard in
19S9.(4) He called the test l'electrosynerse. The CIE test developed by
Brussard has the advantage of using an electric current (and resulting endos-
mosis) as a migratory force, which increases the speed and sensitivity of the
precipitin reaction. The technique was used sparingly until Gocke and Howe
published their paper on its use to detect Australia antigen in 1970(5) and
a paper by edwards in 1971(6) to detect bacterial antigens. Since then, there
has been widespread use of the technique to detect both bacterial, viral, fungal,
(7,8,9,10,11,12)
protozoan, and various protein constituents associated with diseases.
Current uses of CIE are summarized in Table 1.

Table 1. Current Uses of CIE in Infectious Diseases

1. Detection of antigen in body fluids.

2. Detection of antibodies.
3. Prognostic usefulness, i.e., positive correlation between
presence and amount of antigen and severity of illness
and prognosis.
4. Possible role of antigens in disease pathogenesis.
5. Identification and typing of isolates in clinical micro-
biology laboratory.

Although CIE is a variation of immunoelectrophoresis and agar gel diffusion,

the various parameters involved seem much more critical. An illustration of a
simple immunoelectrophoresis chamber is shown in Figure 2 (see page 7). Since
the test has not been standardized with respect to variation in antigen compo-
sition, several parameters of the technique which influence sensitivity will be
brought to the attention of the reader/user.

One of the critical points is to have the correct current/voltage so that

the precipitin band is formed between the two wells (Figure 3, see page 7). A
number of factors contribute to this success, i.e., well size, distance between
wells, composition of agar, depth of agar, composition of buffer, concentration
of antigen and antibody. These parameters will eventually be standardized as

6
 Agarose covered Slicle

Electrode connecting wicks

Electrode vessel

ANODE

I
CATHODE

,I,I,I
DC
SUPPLY
Figure 2. A simply designed electrophoresis chamber which is easily adapted to
a single microscope slide, multiple slides, or Kodak glass plates.

ANODE (+) CATHODE (-)

ANTIBODY ANTIGEN Figure 3. Antigen-antibody

precipitin patterns observed

in counterimmunoelectro-
phoresis. The upper pattern
represents the optimal

# pattern, not often observed

when testing clinical
------
- - samples. The lower two
patterns are frequently
seen in clinical samples.
----
------- The bottom pattern indicates

- -- antigen excess or a support

medium with low endosmotic
flow.
.-- - - - - ---

ENDOSMOTIC FLOW ELECTRICAL "CURRENT"

FLOW

7
Edwards, et al.

commercially prepared reagents for CIE testing become available. However, until
that time occurs, those points, which bear directly on the sensitivity of the
test, will be discussed.

Well Size: For the antigens that have so far been successfully detected
by CIE, a 3 mm diameter well separated by 2-4 mm (edge to edge) has proven
satisfactory. This size well will hold from 5-7 p1 fluid. There are variations
that can be substituted for this pattern. If the antibody used to detect anti-
gen is relatively weak, the "antibody" well size can be increased to 5-8 mm in
diameter, thus allowing a larger quantity of antibody to be used, at the same
time, keeping the antigen well at 3 mm diameter and vice versa. Variations in
well size patterns should be established with each lot of antisera to insure
the sensitivity needed.

Distance Between Wells: It is recommended that when trying a "new"

antigen-antibody system, a series of wells be spaced, starting at 2 mm (edge
to edge) and increasing the distance between wells by 2 mm through 5 wells.
Such a pattern will reveal an optimal range of distances which will give a
precipitin band with the expected antigen concentration.

Composition and Depth of Agar: The less pure the agar, the greater the
negative charge, and in turn, the greater the endosmotic flow. This can have
catastrophic consequences on the CIE test. Agarose, although the quality
varies from company to company and even from batch to batch, appears to give
the most consistent results with the antigens so far detected. Each agarose
batch should be checked for sensitivity. The depth of the agarose layer over
the slide is another variable that has definite influence on the sensitivity
of the test. Our experience indicates that agar thicknesses of 3 nun or greater
are generally unsatisfactory. Optimal thickness ranges from 1-2½ mm.

Buffer Composition: A wide variety of buffers have bcen used in CIE.

However, barbital buffers seem to be the buffer of choice for most antigens.
One exception is the use of a borate buffer when testing for pneumococci types
7 and 14.(13) Barbital buffer (0.05-0.1 M) at pH 8.4-8.6, has been used success-
fully for most antigens since they are negatively charged at this pH. The
concentration of the buffer in the agarose layer is of importance since the
speed of migration of a given substance decreases as the ionic strength of the
surrounding liquid increases. It should be pointed out that for buffers con-
ststing of weak monobasic acids (such as barbital) the ionic strength (u) is

8
Edwards, et al.

equal to the concentration (M) of the salt. Some authors have used the P and m
interchangeably which has led to confusion. The buffer concentration has a
direct influence on the endosmotic flow in relation to the impurities of the
agar layer. It (endosmosis) is greater in less concentrated buffers and in
buffers with a pH which favors the ionization of the COOH groups of the agarose,
i.e., more alkaline buffers. It is greater with thick layers of agar than with
thin layers of agar. These physiochemical variables are important considerations
in selecting buffers and agar, both of which are critical to reproducibility
and optimal immunological reactivity.

Concentration of antibody and antigen: The sensitivity of the CIE test,

assumin-, all other parameters are optimal, depends upon the potency of the
antisera. It has been the authors experience that this is the limiting factor
in detecting small amounts of antigen. The test is also subject to prozone
effects due to antigen excess. If a potent antiserum is used, prozones have
not been observed in clinical material, even to the extent of 20 pl/ml concen-
trations of meningococcal antigen in one spinal fluid examined. Our experience
indicates that potent antisera is the critical ingredient. The authors have
not observed prozones due to antibody excess even under experimental conditions.
However, when an attempt is made to detect antibody in a clinical specimen,
several antigen concentrations should be tried, as prozones with antigen excess
3
have been seen. ( )

With potent antisera, pneumococcal and meningococcal antigen can be detected

at from 1-2 pg/ml concentration. Since only 5-7 pl. of body fluid (spinal fluid
for expniple) is added to a well, this gives a sensitivity of 5-10 nanograms of
antigen. This sensitivity has allowed several quantitative studies to be made
in which the severity of illness or prognosis of the acute illness can be
predicted. It has been found that the greater the antigen concentration, the
more severe and prolonged is the course of recovery. Antigen quantitation in
spinal fluids and blood sera has become an important adjunct to antigen detection
due to the consistent empirical relationship between antigen level and prognosis.
This aspect is discussed in greater detail below.

Table 2 (see pg 10) summarizes microbial antigens and antibodies which have
been detected in body fluids or in vit-o in the laboratories of the authors and
other investigators. The three types of infections which have been studied the
most extensively by CIE (besides hepatitis B), are thouse caused by meningococcus,

9
Edwards, et al.
'lab it. 2 .De'teCt inn of Mi crobial Ant ipens and Ant ibod jes
by CIF in Various Body Fluids

Ant igcns (in vivo) D~etecte(d in

Viral:
IHBsAg Serum, tears, saliva, urine
Enteroviruses Cerebraisninal fluid (CSF)
Bacterial:
Streptococc~if, pneumoniae Sputum, serum, urine, CSF, pleural,
bullus, joint, peritoneal fluids
Neisseria ineninuitidis
A, C, D, X, Y, Z CSF, serum, joint fluid
liaenmophilus influenzae type 1) CSF, serum, subdural, joint
fluid
Pseudomonas aeruginosa Scrum
Kiebsiella pneumoniae Skerum, CSF, urine, pleural fluid
E. coli K1 CSF, serum
Staphylococcus aureus (tilchoic
acid) CSF, pericardial fluid
Streptococcus group B CSF

Antibodies Detected in

Viral:
Hepatitis B Serum
Influenza A2 Serum
California encephalitis virus Serum
Cytomnegalovirus Serum
Bacterial:
Serratia Tnarcescens Serum
Staphylococcus aureus Serum,
Mycoplasma I erwnnoniae Serum
Fungal:
Candida sp. Serum
Coccidioides irniitis Serum
}Iistopli ,ma capsulatuni Serum
Actinomvcesr israeli Serum
Aspergillus fumiqatus Serum
Protozoan:
Trvnanosoma cruzi Serum
Entameaba histolvtica Serum
Trichinella sviralis Seru7

Antigens (in vitro) Detected in

Viral:
Plant viruses Plant juice, tissue-culture fluid
Influenza A2 Tissue- culture, chic],
ciiorioallantoic fluid
cytomegalovirus Tiss,-ue culture fluid
bacterial:
Strentococcus pneumnniae' Broth culture extract,~
Streptococcusý groups; A-F Broth culture extracts
Poptococcus mapnus, Broth culture extracts,
I~lehsielln pneumonihac. Broth culture extracts,
Bacteroides fracilis Broth culture extracts
Enterobacterial coimmon arti4cn
(P-scheri chlia coli 014) Lr-)th cu lture e':tractn

10
Edwards, et al.

6 10
pneumococcus, and hemophilus.( ' ) Of the various clinical syndromes studied
the highest diagnostic yield was obtained in meningitis caused by these three
organisms (over 90% in most studies). The method has been found to be more
specific and sensitive than the Gram stain, and more rapid than the culture of
cerebral spinal fluid (CSF). It has an additional advantage over the latter in
that it may remain positive in partially treated cases. In pneumococcal pneumonia,
the results have been less satisfactory in that only approximately 50% of
bacteremic cases have detectable antigen in their sera.(10) This figure could
be increased to 64% if both sera and urines (concentrated 20-fold by 95%
ethanol precipitation at 5GC) were studied. Though CIE appears to be less
sensitive(7) than bacterial cultures in pneumococcal pneumonia, it continues
to have the advantage of rapidity, and specificity in relation to sputum
cultures. In this regard, current work on this disease focuses on detection by
CIE of pneumococcal capsular antigens in sputa.(15) The test was reported to
be sensitive (100% [8 of 8] sputa from established cases of pneumococcal pneu-
monia were positive) and specific (saliva of 83 normal individuals was negative).

Two other areas of CIE application will be commended on briefly. One is

the diagnostic application in antibody detection. The clinical situations
where this has been found useful are summarized in Table 2. In general,
presence of precipitin antibodies has been found to correlate with active
10 1 1 14
infection.( ' ' ) The specificity of the method could b4 increased by ob-
taining an antibody titer; a titer of > 1:8 has generally discriminated well
between active infection and "serofast" status or presence of cross-reactive
14
antibodies.( ) An additional application of CIE has been in the identification
16
and typing of microbial isolates obtained in vitro.(9, ) In general, appro-
priate dilutions of sonicates of suspensions of bacterial cultures from plates
or broth have been employed for study. This approach offers an additional
diagnostic dimension in that it may still lead to a more rapid identification
of pathogen, which was present in body fluids (blood, CSF, etc.) in insufficient
quantity to yield enough antigen for direct detection of CIE.

Finally, one of the most intriguing "spin-offs" of the diagnostic application

of CIE has been the demonstration of a direct relationship between the presence
and the amount of antigen in body fluids and the disease morbidity and mortality.
2 17 2 2 )
' '
This relationship has been reported in a number of infectious diseases.(
Specifically, meningococcal disease patients with antigenemia had a higher

11
Edwards, et al.

incidence of disseminated intravascular coagulation (DIC), shock and more pro-

17 there
longed duration of coma.( ) In patients with pneumococcal infections,
was also a direct relationship between the quantity of circulating antigen and
development of DIC, other complications, and mortality rates.(18,19) In both
infections, antigen levels correlated inversely with antibody response and levels
of complement and its components. (17,19) The interpretation of these findings
is a matter of speculation, but they may signify formation of toxic immune
complexes, or impairment of immunologic function which may have an adverse effect
of resistance to the infection. Work is in progress in several laboratories to
elucidate this problem.

Materials Needed

Apparatus required for CIE.

1) Variable DC power source with ampere and voltage meters.
2) Chamger with appropriately spaced electrode vessels.
3) Cleaned lantern or microscope slides. Wash with Baboo (or
equal cleaner), towel dry, then wash in 95% ethanol, towel
dry to remove all residue.
4) Filter paper for electrode wicks. (Eaton=Dykman #301 or
Gelman S1290 were found to provide superior conductivity).
5) Approximately 3 mm diameter needles (Gelman well cutters,
#51466).
6) Agarose (generally 1%) - there are several sources, they
vary in quality from lot to lot. See references 23 & 24.
7) Buffer - a barbital buffer pH 8.6 has proven satisfactory
for many antigen-antibody systems: However, the optimal
buffer and pH should be established for each antigen to
be identified.

The Test Procedure

The technique can briefly be described as follows:

A glass slide or langern slide is covered with 1% agarose in buffer
to a depth of 1-21mm (2 ml per microscope slide/10-14 ml per lantern slide).
After the agarose has solidified, two rows of wells are punched in the agarose,
the rows being approximately 3 mm apart, edge to edge. The plugs are removed
by suction, using a Pasteur pipet attached to a faucet aspirator. The prepared

12
Edwards, et al.

plates are then placed in an electrophoresis apparatus for immediate use or can
be stored in a moist chamber in the cold, for several (but not more than 4) days
until needed. If the slides are not used the same day, a preservative (0.1%
sodium azide final concentration) should be added to the agarose prior to layer-
ing over the glass slides to retard bacterial contamination. The slides are
connected vessles by a paper wick. Pre-soak the filter paper to insure a sat-
isfactory bridge with the agar layer. The sample to be tested (antigen) is
placed in the wells on the cathode (-) side of the electrophoresis set-up and
the antibody on the opposite wells (the anode side). The apparatus is connected
to a power pack and run using from 3-6 mA per microscope slide or from 12-20 mA
per lantern slide, as measured at the power source. For diagnostic detection
14
of antibodies 25 mA has been employed.( ) After one hour, turn off the power,
remove the slides and observe for a precipitin band between the wells by using
an oblique lighting effect, holding the slides against a dark background. A hand
lens 3-7X is highly recommended. A simple and rapid method of intensifying the
precipitin band is to soak the slide in 95% ethanol for 15-20 minutes or in
physiologic saline for several hours. This provides greater contrast so weak
reactions can be more readily observed, but does not alter the final results.

Quantitating antigen (previously identified): Two-fold dilutions of the

specimen (1:1-1:128) are run against undiluted type-specific antiserum. Con-
currently, for reference quantitation, purified type-specific antigen is also
run; the following dilutions are employed: 25, 10, 5, 2.5, 1.0, 0.5, 0.1,
0.05 Pg/ml.

Purified lyophilized antigen (for pneumococcus, see reference 25) should

be reconstituted at concentrations no less than 5 mg/ml (stock) because of the
possibility of polysaccharide absorbing onto glass or plastic at lower concen-
trations. The sock solution is stored at -20'C. Any left-over test dilutions
below stock concentrations should be discarded.

Following is an example of quantitation of a pneumococcal (PNC) antigen:

Quantity of antigen (pg/ml) = specimen titer (dilution) x reference quantita-
tion titer.

13
 Edwards, et al.

Dilutions Concentration
of Anti- of Purified Anti-
Cathode Specimen sera PNC (pg/ml) sera Anode
(-) (+)

1:1 0 1O 25.0 0 10
1:2 0 J0 10.0 0 0
0
1:4 0 J 0 5.0 0 j 0
1:8 0 0 2.50 I 0
1:16 0 0 1.0 0 0
1:32 0 0 0.5 01 0
1:64 0 0 0.1 0 0
1:12b 0 0 0.05 0 0
1:256 0 0

Results: vg antigen/ml serum = 16 x 0.5 vg/ml = 8.0 vig/ml.

Source of Reagents for Diagnostic CIE

Pneumococcal antiserum: omniserum, pooled antisera (groups A-I),

specific types: Statens Seruminstitut, Amager Boulevard 80, DK 2300, Copenhagen
S. Denmark (instruct them to mark customs slip as follows: "TSUS 437.76/Free
Antitoxin"). Also available from Difco; however, Statens Seruminstitute rea-
gents have been more reliable for the CIE test.

H. influenzae antiserum: Difco Laboratories, Box 1058A, Detroit,

MI 48232; Hyland Labs, 330 Hyland Ave., PO Box 2214, Costa Mesa, CA 92626;
and Burroughs-Wellcome. PO Box 1887, Greenville, NC 27834.

Meningococcal antiserum: Difco; and Burroughs-Wellcome.

Rocket Immunoelectrophoresis:

This test has not received wide clinical application. However, it seems
to have the potential of being a sensitive method for antigen detection and
quantitation. It has the added advantage of performing several tests on a
single microscope or Kodak slide.

The test is basically a modification of the single-radial-diffusion method

of Mancini but uses an electrical current to force diffusion of the antigen in
a gel metrix containing antibody. The antigen reacts with antibody in the gel,
giving ricket like zones of precipitation. Not only can antigen be detected
but when known concentrations of antigen are added as controls, the rocket
heights are proportional to the concentration of the antigen. Thus by plotting

14
Edwards, et al.

the height of the precipitation formed by the controls on arithmetic graph paper,
the values for the unknown samples can be determined by simple extrapolation.

A limited number of different antibody impregnated agar plates are com-

mercially available. These include IgG, IgA, IgM, C3; haptoglobin, albumin,
alpha 1 antitrypsin and transferrin (ICL Scientific, 18249 Euclid St., Fountain
Valley, CA 92708). Because of the technical difficulties in standardizing and
preparing the plates with uniform agar layer thickness, it is recommended that
no attempt be made (except under experimental conditions) to prepare the plates
in-house for diagnostic use.

Specifically sensitized Protein A-containing Staphylococcus aureus.

The cell wall of most coagulase positive Staphylococcus aureus strains con-
tain a protein, called protein A, which combines with the gamma globulin of most
mammalian species. For many years it was thought that this was a classical anti-
gen-antibody reaction which occurred because of "natural" antibodies due to
frequent exposure to this common bacterium. However, Forsgren and Sjoquist
found that the reaction was mediated by sites on the Fc part of immunoglobulin
IgG and labelled the reaction a "pseudo-immune" reaction. The reaction between
the staphylococcal protein A and the Fc part of IgG provides an example that a
reaction can be due to IgG factors other than specific antigen combining sites.
All known antibody combining sites are located on the Fab-fragment of the IgG
molecule.

Inhibition experiments with isolated heavy and light chans of immunoglobulin

IgG showed the protein A activity to be present only in the heavy chain preparations.
When F(ab') 2, Fab, Fc and F'c fragments were studied, only the Fc fragment
preparation gave inhibition of the precipitate formed between a myeloma globulin
and protein A. These inhibition studies confirm the original studies of
Forsgren and SjUquish that the reactivity of normal human gamma globulin with
staphylococcus protein A resides in the Fc part of the molecule and is analogous
to investigations for rheumatoid factor which also reacts with the Fc fragment
of the IgG molecule.

This spontaneous "uptake" of IgG by the coagulase positive Staphylococcus

aureus by way of the Fc fraction, leaving the antigen binding Fab portion of the
molecule free to combine with antigen has many practical applications. Kronvall
showed that by sensitizing Protein A-containing Staphylococcus aureus with sub-

15

-, -...- . . . ---
Edwards, et al.

agglutinating amounts of rabbit Lti-panumnicoccal typing serum, the staphylococcus

would take on the specificity of the immune serum and be agglutinated only by

the corresponding pneumococcal antigen. Thf antisera used must be specific as

cross reactions occur as in any immunological method. The authors have found

the sensitivity of the test to be greater than the capillary precipitation test

for grouping streptococcus but not as sensitive as the CIE test. However, it

is a rapid test in that a positive agglutination reaction occurs within 1-3

minutes.

The sensitized staph-reagent has been used to group/identify antigens from

broth cultures and also to group/identify antigens directly upon the primary
isolation plate, thus making it possible to give presumptive identification
within hours after having received a test sample compared to 1-4 days by con-
ventional methods. Until further data is gathered on the sensitivity and
specificity of the test using antisera to a broad range of bacterial species,
the use of the test is not to replace accepted standard procedures of microbial
identification,- but rather serve to augment these procedures and if/when so
used, may provide early presumptive identification of an infectious agent.

MATERIALS AND METHODS

1. Staphylococcus aureus, Cowan I Strain, American Type Culture.

2. Trypticase Soy Broth (BBL) TSB
3. Clean 500 ml Erlenmyer flasks and 10x125 text tubes.
4. 250 ml polyethylene screw cap centrifuge tubes or if a 210 ml
centrifuge head is not available, 50 ml polyethylene tubes can be used.
5. Centrifuge with 3-5000 rpm capabilities.
6. Phosphate Buffered Saline 0.03 M pH 7.3 (PBS).

PREPARATION OF STAPHYLOCOCCUS AUREUS

1. Check purity of stock culture by culting 18 hours, 37°C.

2. Pick a colony of pure Staphylococcus aureus and inoculate 2-5 ml
tubes of TSB.
3. After 6-7 hours growth (log phase), gram stain for purity, if pure,
pipette (aseptically) 1 ml into a 500 ml Erlenmyer flask containing

16
Edwards, et al.

250 ml sterile TBS and inoculate a Blood Agar plate to recheck

for purity.
4. Incubate flasks and plates at 37°C for 18-24 hours. (An incubator-
shaker is preferred, however, static incubation can be used but the
yield will be less than when using an incubator-shaker.)
5. As a safety precaution, inactivate the growth by either adding 0.5
ml betapropiolactone, mix and allow to stand at room temperature
for 2 hours or by heating at 56°C for 2 hours.
6. Centrifuge to pellet the bacterial cells and wash 3 times with PBS.
7. Make a 10% suspension of the sediment in 0.5% formaldehyde in PBS.
Allow to stand at room temperature for 3 hours with occasional
shaking.
8. Heat suspension at 80*C in a hot water bath for 1 hour.
9. Wash with PBS three times and store at a 10% suspension in PBS
containing 0.1% sodium azide.

SENSITIZING STAPHYLOCOCCUS AUREUS

1. Remove stock 10% suspension S. aureus from refrigerator. Mix

thoroughly to insure complete suspension and to break up small
clumps that may have developed during storage.
2. Transfer 1 ml (fractions or multiples thereof) of the 10% stock
suspension to a clean test tube.
3. Add 0.1 ml of immune serum, mix. (Ratio of immune serum to stock
cells should remain at 0.1 to 1.0 ml respectively.)
4. Allow to "sensitize" for 1-2 hours at room temperature. Shake
periodically (every 15 minutes) to promote optimal sensitization.
5. Pellet staph, wash pellet at least once with PBS and resuspend to
10 ml with PBS.

TEST PROCEDURE TO IDENTIFY COLONY OF ORGANISMS FROM PRIMARY ISOLATE

I. Make at least 2 parafin rings around suspected colonies. A wire

*Group A,B,C,D streptococci; meningococci group A, C,Y; Salmonella A,B,C,CI, D,

and Shigella sonnei and flexnerii, pneumococcus types 3,4,6,7,8,9,11,12,13,14,

18,19, and 23 have been shown to give a positive test with polyvalent (omniserum)
sensitized staph.
17
Edwards, et al.

ring 12 mm in diameter has been suitable for this purpose. The

parafin has to be hot so a good seal is made with the agar surface.
2. Add a drop of sensitized staph reatent to one ring around a sus-
pected colony(s) and a drop of unsensitized staph reagent around a
similar suspected colony.
3. Rotate the petri dish in a to and fro pattern for at least 1 minute.
Preliminary observations can be made using a dissecting scope. Con-
tinue the rotations for 3 minutes before final reading. Positive
aggregates usually form by 1-2 minutes. The negative (unsensitized
staph) should remain negative. (Continue mixing of the staph reagent
over the colony is very important for development of visible
aggregates.)
4.ý Observe for large aggregates using the dissecting microscope.
Other methods of observing aggregation of staph reagent have not
proven satisfactory.
S. Record results.

PROCEDURE FOR IDENTIFYING ORGANISMS FROM A BROTH CULTURE

1. Transfer suspected organisms to 1-5 ml TSB, or in the case of Sal-

monella, inoculate Dulcitol-Selenite enrichment broth, incubate
12-18 hours or until growth is evident.
2. Transfer 1 drop of the growth onto a marked area of a microscope
slide and a drop of uninoculated growth medium to another area of
the slide, add 1 drop of specific sensitized staph reagent to both
the test and control test area.
3. Mix thoroughly with an applicator stick. Rotate slide for 1-2 minutes.
4. Observe for agglutination. The control should remain negative
throughout the observation period.

18
Edwards, et al.

ENZYME-LINKED IMMUNOSORBENT ASSAY

Introduction
In 1971, Engvall and Perlmann( 2 6 ) and van Weeman and Schuurs independ-

ently reported the use of enzymes conjugated to antibodies for the detection
and assay of biological molecules. This technique was the outgrowth of the use
of enzyme labeled antibodies to detect and localize cellular antigen in light-
(28,29)
and electron-microscopy studies2' and is an alternative to the radioimmuno-
assay (RIA) method. Enzyme-linked immunosorbent assay (ELISA) has been designated

variously as enzyme- or enzymo-immunoassay (EIA) and immuno-enzymatic assay.

Enzyme-linked immunospecific assay (also ELISA) has been proposed by Engvall
as a title with a broader range of application.

The procedures used in enzyme immunoassays are directly analogous to the

techniques developed for RIA and are of the same range of sensitivity due to
the amplification provided by the enzyme. In comparison with radioimmunoassay,
ELISA uses only spectrophotometric or visual detection rather than the expensive
instrumentation necessary for quantitation of radioactivity. In addition, with
ELISA the need for special training and procedures for handling and disposal
of reagents is obviated. RIA and ELISA have been used in similar manners for
serodiagnosis (e.g. 30). However, RIA has not been widely applied in diagnostic
microbiology although direct and indirect methods for bacterial identification
31
have been developed.( ) RIA has also been used for identification of staphyl-
ococcal enterotoxins A, B, and C, and for clostridial toxin. (32,33) The prin-
ciple use of ELISA has been in serology by the use of conjugates directed
against a particular antibody.

Several different ELISA procedures have been used to measure antigen

and/or antibody. In the "competitive" assay for antigen, labeled antigen and

unlabeled antigen compete for a limited quantity of antibody bound to a solid

support. The excess antigen is eluted and either the free or bound enzyme is
quantitated and related to standard concentrations of antigen.

In the "homogenous" assay an enzyme bound to a hapten is inactivated when

an antibody specific for that hapten is present. The addition of free unlabeled
hapten, as either a standard or test, will compete for the antibody binding

19
Edwards, et al.

site freeing. the hapten and thus the enzyme in proportional quantities. The

enzyme can then be quantitated, producing an assay which requires no separation

step. This procedure works only for haptens.

For "immunoenzymometric" assay antigen is reacted with excess labeled anti-

body. Excess solid phase antigen is then added to remove any free labeled

antibody and separated. The labeled antibody-antigen complex remaining in

solution is then quantitated.

Another method used for quantitating antibody is performed by attaching

antigen to a solid support and adding the antibody to be measured. Following
incubation an excess of enzyme labeled antibody with specificity for the first
antibody is added followed by incubation and quantitation of the bound enzyme.

The method described in this chapter is an example of a "sandwich" assay

(see Figure 4, pg 21). The method requires that the antigen (materioan to be
detected) have at least two antibody binding sites.

Summary of Methodology for "Sandwich Assay"

1. Specific antibody is adsorbed to a solid support (polystyrene or

polypropylene) and the support washed.
2. Appropriate antigen in standards or tests is complexed to the
specific antibody adsorbed on the solid support and non-bound
material removed by washing.
3. The same specific antibody as in 1. which has been chemically
conjugated to an enzyme is incubated with the antigen. Non-
bound material is again removed by washing.
4. Substrate is added to allow colorimetric quantitation of the
enzyme label bound to antigen. This value is proportional to
the quantity of antigen present in standards allowing one to
quantitate the color produced over the range of standards and
to relate this to the concentration of antigen in test materials.

If the second antibody is added unlabeled and a third incubation with excess
labeled antibody with specificity for the second antibody rather than for the
antigen is added, the assay is termed a "double sandwich". This last technique
has the advantage that one labeled antibody can be used to quantitate a number
of specific antibodies from a single species.

20
 0
0 0 0

So0 0

00
000

01 2
0Oo
00

4 .5 6

Fig. 4 Schematic of "sandwich" ELISA. 1) Specific antibody is added and

2) absorbed to solid support. 3) Antigen to be measured is added, bound by the
specific antibody, and the unbound antigen washed out. 4) Specific antibody
conjugated to an enzyme is added and bound to the antigen remaining on solid
support antibody. 5) Enzyme substrate is added and 6) the optical density of
the product assayed visually or spectrophotometrically.

21
Edwards, et al.

Below is a procedural outline for commonly used methods of adsorption,

conjugation, and immunoassay. Alternative procedures are included, however, a
master list of required materials has not been presented. Materials needed
will inclufe polystyrene or polypropylene tubes (12x75 mm, Sarstedt), Tween 20
(J. T. Baker), Glutaraldehyde (Sigma), and either Alkaline Phosphatase (Sigma)
and p-Nitrophenyl phosphate (Sigma) or Horseradish peroxidase (Sigma) and 2.2'-
Azino-di[3-ethyl-benzthiazoline sulfonate (6)] (ABTS, Boehringer-Mannheim).
PROCEDURE

I. Adsorption of Specific Antibody to Solid Support

(Select one of the two methods).
A. The most frequently used solid support for serodiagnostic methods has
been polystyrene(34) in the form of either tubes or plates. The pro-
cedure for the adsorption of protein to polystyrene tubes is as follows:

Dilute specific antibody material to be adsorbed just

prior to use with 0.1 Na 2 CO3 (pH 9.8). For dilution
see Section on titration of solid phase antibody. Add
1.0 ml of this solution to a polystyrene tube (12x75 mm)
and allow it to set at 37°C for three hours. An additional
period (overnight) at 4°C may also be used, but adsorption
seems essentially complete within 3 hours. Wash the
tubes three times with gentle swirling with 2 ml 0.9%
NaCi containing 0.05% Tween 20 (Saline-Tween). The Saline-
Tween wash as used throughout this discussion is intended
to reduce non-specific adsorption of antigen or antibody
in subsequent steps without removing the solid phase
antibody. If nonspecific adsorption is still a problem
the wash procedure may incorporate 10% aged human serurm
or 4% BSA or 0.5 M NaCl with the 0.05% Tween 20. The
tubes prepared with adsorbed antibody can be stored at
40 C for several weeks.

B. The following method using polypropylene tubes has been reported as

superior to the polystyrene method because glutaraldehyde is used to
link the protein to the solid support. There is apparently little or
no "leakage" of protein during wash or assay procedures.

22

,.. ' " . . . .. . .. . , ,,,' ' • . .

E~dwards, et al.

Incubate polypropy'lene tubes (12x7S mm, Sarstedt) with

1.0 ml freshly prepared 0.1% glutaraldehyde in 0.1 N! carbo-

ate buffer (pH 9,0) for 3 hours at 56°C and iash thoroughly
with deionized water. Incubate tubes for 20 hours at 4 0 C
using 1.0 ml of an appropriate dilution of antiserum in 0.02
M phosphate 0.9% NaCl (pHl 7.2) and 0.02% NaN3 as preservative.

The diluted antiserum can be left in the tubes at 4°C and

the tubes used for up to eight weeks. (35)

II. Conjugation of Enzyme-label to antibody.

A. Alkaline Phosphatase (AP).

Combine 1.0 mg of the specific antibody to be conjugated

and 3.0 mg of the AP (Sigma, Type VII) in 0.2 ml total vol-
ume of 0.1 M phosphate buffer (pH 6.8). If either material
contains (NH4 ) 2 SO 4 dialyze prior to conjugation procedure.

Add 8.0 pl of 25% aqueous glutaraldehyde whicn has been

diluted with phosphate buffer (l:4,v:v) for a final glutaral-
dehyde concentration of approximately 0.2%. Allow to set
for 2 hours at room temperature, dialyze exhaustively at 4 0 C
against 0.05 M tris(hydrox)ymethyl) aminomethane-hydrocl uloride
(Tris-HC1), pH 8.0, and chromatograph on Sephadex G-200
(1.5 x 90 cm) in the same buffer. The conjugate will elute
in the void volume and unconjuga*ed material will be re-
tained. The void volume fraction of AP conjugates can then
be stabilized by addition of 4% human serum albumin and 0.02%
NaN3 stored at 4°C and used for immunoassay. If necessary,
a centrifugation at 10,000 xg for several minutes may be
used to removc aggregates. Another procedure for higher
specific activity AP is also published. (36)

B. Horseradish peroxidase (HRP).

The procedure of Nakane(37) allows greater efficiency of

coupling of HRP to either lgG or to F(ab') 2 fragments than
either a one or two step glutaraldehyde procedure. The
procedure is not as simple as the glutaraldehyde but has
provided better results in our work.

23
Edwards, et al.

2
Add 0. 1 of 1% 1-fluoro- .4-dinit robencne ir absolute ethanol
to S mg of tlRP (Sigma, Type VI) dissolved in 1.0 ml of freshly
1
made 0.3 M sodium bicarbonate (p' 8.1). Mi., gently for 1 hour
(room temperature) and add 1.0 ml of 0.06 M NaTO 4 in water. Mix
gently for 30 minutes at room temperature. Add 1.0 ml of 0.16 M
ethylene glycol in water and mix gently for 1 hour at room
temperature. Dialyze this solution against 3 x I liter changes of
0
0 01 N sodium carbonate (pH 915) at 4 C. Add 5 mg IgG in 1 ml of
carbonate buffer to the dialysate a,id mix gently for 2-3 hours at
room temperature. To the IgG-ttRP conjugate add S mg NaBH 4 and allov
to stand at 4°C overnight. Dialyze this at 4 0 C against 0.02 M
phosphate 0.9% NaCi (pH 7.2) (PBS) and remove any precipitate by
centrifugation. Purify the conjugate by Sephadex G-200 column
(1.5 x 90 cm) chromatography in PBS and take the void volume fraction.
HRP conjugate can be stabilized with 1% BSA and stored in aliquots
at 020°C without azide.

For procedure A. or B. a purified antibody fraction (such as 45% (NH 4)2so4

fraction, an IgG fragment F(ab') 2 or an antib- Ay purified by affinity chromoto-

graphy) can be used for conjugation to improve sensitivity or specificity.

III. Immunoassay

A. Alkaline Phosphatase.
1. Wash the antibody coated tubes 3 times with 2 ml Saline-Tween.
2. Dilute standards and serum to be tested with 0.02 M phosphate,
0.9% NaCl, 0.05% Tween 20 and 0.02% NaN3 , pH 7.2 (PBS-Tween).
Add ' mi of diluted standard or serum to each tube and incubate
at about 30°C with mixing for 3-6 hours. Wash 3 times with
Saline-Tween.
3. Dilute enzyme-antibody conjugate to optimal concentration with
0.05 Tris-HCl pH 8.0 and add 1 ml of conjugate to each tube.
Allow to incubate with gentle mixing at about 30 0 C 3-16 hours.
Wash 3 times with Saline-Tween.
4. Add 1 ml of 0.05 M sodium carbonate 1 mM MgCl 2 buffer (pH 9.8)
containing 1 mg/ml p-Nitrophenyl phosphate.(36)
5. Incubate for a suitable time at room temperature (e.g. 30 minutes).
Add 0.1 ml 1N NaOH to Stop the reaction and read aS 4*00 nm. Plot
increase in absorbance units vs. standards.
24
Edwards, et al.

B. Horseradish Peroxidase (IIRP)

1. Wash the antibody coated tubes 3 times with 2 ml Saline-Tween.

2. Dilute standards and serum to be tested with 0.02 M phosphate,
0.9% NaC1, 0.05% Tween 20, pl 7.2 (PBS-Tween). Add 1 ml of di-
luted standard or serum to each tube and incubate at about 30*C
with mixing for 306 hours. Wash 3 times with Saline-Tween.
3. Dilute enzyme-antibody conjugate to optimal concentration with
PBS-Tween and add 1 ml of conjugate to each tube. Allow to in-
cubate with gentle mixing at about 30'C 3-16 hours. Wash 3
times with Saline-Tween.
4. Add 1 ml of ABTS reagent. Make this by adding 2 mg ABTS per
50 ml of 0.03 M citric acid, 0.04 M phosphate, pH 4.0. Immediately
before use add 25 pl of 30% H202 per 50 ml of prepared ABTS
solution. (38)
5. Incubate at 25'C for 10 minutes. Stop the reaction by addition
of 0.05 ml 0.3# NaN3 or by pouring off solid phase. Read optical
density at 415 nm.
or alternatively for HRP
4. add 1 ml of substrate for peroxidase. This is prepared by adding
80 mg of 5-aminosalicylic acid to 100 ml deionized water at 70°C
and cooling to room temperature. Directly prior to use adjust
the pH to 6.0 with 1 N NaOH. To 9 ml 5-aminosalicylic acid add
39 )
1 ml of 0.05% Hydrogen peroxide.(
5. Incubate at room temperature for exactly 60 minutes stopped by
addition of 0.1 ml 1 N NaOH and read at 449 nm. Plot increase
in absorbance units vs. standards.

IV. Dilution of Reagents

A. Titration to solid phase Antibody
To determine the dilution of antiserum to adsorb to the solid support
to give maximal uptake of antigen, take serial 10 fold dilutions of antiserum
up to 10,000 fold. Adsorb each dilution to the solid support as described.
Then add the highest concentration of antigen which you expect to use in your
assay to each dilution. Complete the assay as described above and select the
solid phase antibody dilution which gives the greatest enzymatic activity.
B. Titration of Conjugated Antibody
To a set of tubes coated with the optimal concentration of specific

25

II
Hdwards, et al.

antibody add the highest concentation of antigen which you expect to use. Incu-

bate, wash, and complete the assay using dilutions (e.g., 1:40, 1:80, 1:160,
1:32, 1:640) of conjugated antiserum. Select the optimal dilution.

DISCUSSION

Developmental work increasing the sensitivity and improving separation

methods of ELISA is moving rapidly ahead. In addition, improvements are being
made in methods for determination of the end point of the assay and in automation.
The application of enzyme kinetic analysis will aid in both automation and
quantitation. ELISA faces the limitations which are common to any immunoassay
in that the characteristics of the antibody determine sensitivity and specificity
of the assay. Also, for initial assays, the components of the assay must be
prepared and titrated to maximize the assay. Some technical problems are that
the titrations of antibody and conjugated antibody must be completed for each
new lot of tubes that is used, each new lot of antiserum and each new conjugate
preparation. Antibody specificity and accuracy in measurement are of utmost
importance in quantitation,due to the amplification provided by the enzyme.

In the selection of an enzyme, there are several factors which are of

prime consideration.

1. The end product of the enzymatic reaction should be visible to the

naked eye if ELISA is to be used for screening purposes, absorb in either visible
or ultraviolet regions for spectrophotometric quantitation, or be fluorescent
for fluorimetric determination.

2. The end product should have a high molar extinction coefficient.

3. The enzyme should have high specific activity.

4. The enzyme should be stable, easily obtainable, and not inhibited by

constituents of the material to be tested.

Alkaline Phosphatase (AP) meets the the enzyme requirements for labeling
and has been used frequently. AP is preferable to peroxidase for quantitative
ELISA and can be stored at 4°C with 0.02% NaN 3 as preservative for several
months. The substrate of choice for AP is p-Nitrophenyl phosphate. Since in-
organic phosphate is an end product of the enzymatic reaction, one must avoid
the use of phosphate buffers in the enzyme substrate solution which could alter
the rate of hydrolysis of the substrate.

26
Edwards, et al.

Another frequently used enzyme with the same range of sensitivity as AP is

horseradish peroxidase (11RP). IIRP is a smaller molecule than AP and is thus
useful in histochemistry because diffusion into tissue is enhanced. For glu-
taraldehyde conjugation with IIRP, one may use a one step (as described here
for conjugation of AP) or two step procedure. (40) HRP has the unusual charac-
teristic of reacting with only one glutaraldehyde molecule which limits self
coupling of HRP and allows excess glutaraldehyde to be dialyzed out. The
"activated" HRP can be added directly to the IgG which increases coupling
efficiency and decreases self coupling of the second protein. Due to simplicity
and time, the one step procedure has been used most frequently. The IIRP conjugates
can be stored sterilized by filtration and frozen in small aliquots.

The substrate for HRP is H2 0 however, there are a number of chromagens for
HRP of which one can be selected for use with the substrate. Diaminobenzidine
(DAB) is available as a chromagen but is more useful for histochemistry than for
ELISA because the end product forms a polymeric compound which is insoluble in
aqueous solution. 5-Aminosalicylic acid (S-AS) and O-Phenylenediamine (OPD)
are good substrates but OPD is difficult to handle because it is somewhat sensitive
to light. The chromagen of choice for greatest sensitivity is ABTS.

Beta-Galactosidase is a much larger enzyme that is also being used. This

enzyme can be used with 4-methylumbelliferone to produce a fluorescent product
which has proven to be the most sensitive ELISA assay to date. (41)

APPLICATION

ELISA has been used to detect antibodies to viral, bacterial, and parasitic
antigens and bacterial antigens themselves. The simplicity and speed of the
technique has led to an application in veterinary medicine for detecting pathogenic
organisms in meat on line at slaughterhouses. Saunders el al. have developed cap-
ability for multiple screening of serum antibodies to bacterial pathogens.
Initial reports identify Trichinella spiralis, Brucella abortus(42) and hog
cholera (43) and show good correlation with previous methods and in most cases
improved sensitivity. In addition, Saunders has reported the detection of Staph-
ylococcal Enterotoxin A to a level of 3.2 ng toxin/ml of prepared food product
in 1-3 hours test time.(44) Ruitenberg has used ELISA for the identification
of T. spiralis infection in pigs at the slaughterhouse. (45)

27
Edwards, et al.

A second major application has been in disease screening for infectious

disease in humans in field studies in remote areas. The characteristic of the

assay that gives a yes or no answer by visual or crude colorimetric methods makes
this application possible. Voller et al6) have demonstrated the potential of
the test in a seroepidemiological study in an area of Columbia where malaris is

endemic. Using microtiter trays and using visual and spectrophotometric methods
they demostrated the serological differences in the population of endemic area and

area where antimalarial measures have been taken. Voller et al.(47) and Ruitenberg
48
and Buys( ) have applied ELISA micro-techniques to the seroepidemiological
studies in African trypansomiasis.

49
*Other
(50) applications have included serodiagnosis of syphilis( ) and schistoso-
.(1

miasis, the detection of streptococcal M protein antibodies(51) and detection

. (52)
of hepatitis B surface antigen.

SUMMARY

Enzyme-linked immunosorbent assay (ELISA) is a recently developed technology

that has sensitivity and accuracy in the range of radioimmunoassay (RIA) tech-
nology and has the same specificity. The enzyme label can be stable for months
and the enzymatic activity can be quantitated spectrophotometrically. The general
features for radio- and enzyme-immonoassay are the same. However, in ELISA,
separation is usually provided by the use of a polystyrene or polypropylene tube
or microtiter trays as a solid support. Alkaline phosphatase (AP) and horse-
radish peroxidase (HRP) are the most commonly used enzymes and are linked to IgG
or antigen by glutaraldehyde or NaIO 4 . The method can be used as a rapid screen-
ing device visually or quantitated by a spectrophotometer. The methodology can
be automated and adapted to centrifugal analyses and can be performed without
the need fo a separation step in some cases (Homogeneous assay technique).
Kinetic analysis of enzymatic activity may allow increased accuracy and broader
applications.

Two recently written reviews (53,54) have covered discussion of principles,

methods, and applications very extensively.

28
 REFERENCES

1. Vincent, M.H. and Bel]ot, M. Observations a I'occasion du

proces-verbal. Diagnostic de la meningite cere'bro-spinale

a meningocoques par la "precipito-rcaction". Bull. Acad.

Med. (Paris) 61:326-332, 1909.

2. Dochez, A.R. and Avery, O.T. The elaboration of specific

soluble substance by pneumococcus during growth. J. Exp.

Med. 26:477-493, 1917.

3. Williams, C.A.,Jr. and Grabar, P. Immunoelectrophoretic

studies on serum proteins. I. The antigens of human serum.

J. Immunol. 74:158-169, 1955.

4. Bussard, A. Description d'une technique combinant simultanement

l'electrophorese et al pr~cipitation immunologique dans un gel:

1'e'ectrosynerese. Biochim. Biophys. Acta 34:258-260, 1959.

5. Gocke, D.J. and Howe, C. Rapid detection of Australia antigen

by counterimmunoelectrophoresis. J. Immunol. 104:1031-1032, 1970.

6. Edwards, E.A. Immunological investigations of meningococcal

disease. I. Group-specific Neisseria meningitidis antigens

in the serum of patients with fulminant meningococcemia.

J. Immunol. 106:314-317, 1971.

7. Dorff, G.J., Coonrod, J.D. and Rytel, M.W. Detection by

immunoelectrophoresis of antigen in sera of patients with

pneumococcal bacteremia. Lancet 1:578-579, 1971.

8. Gordon, M.A., Almy, R.E., Greene, C.H. and Fenton, J.W., Jr.

Diagnostic mycoserology by immunoelectroosmophoresis: A general,

rapid, and sensitive microtechnique. Am. J. Clin. Path.

56:471-474, 1971.

29
18. Ryte] , M.W., Dee, T.H. , FerstenfeId, J.E. and HensIey, G.T.

Possible pathogenetic role of capsular antigens in fulminant

ppneumococcal disease with disseminated intravascular coagulation

(DIC). Am. J. Med. 57:889-896, 1974.

19. Dee, T.H., Schiffman, G., Sottile, M.1. and Rytel, M.W.

Immunologic studies in pneumococcal disease. J. Lab. Clin.

Med. 89:1198-1207, 1977.

20. Shackelford, P.G., Campbell, J. and Feigin, R.D. Countercurrent

immunoelectrophoresis in the evaluation of childhood infections.

J. Peds. 85:478-481, 1974.

21. Pollack, M. Significance of circulating capsular anticen in

Klebsiella infections. Infect. Immun. 13:1543-1548, 1976.

22. McCracken, G.H.,Jr., Sarff, L.D., Glode, M.P., Mize, S.G.,

Schiffer, M.S., Robbins, J.B., Gotschlich, E.C., 0rskov, I.

and 0rskov, F. Relation between Escherichia coli K1 capsular

polysaccharide antigen and clinical outcome of neonatal

meningitis. Lancet ii:246, 1974.

23. Hibrawi, H., Garrison, F.D. and Smith, H.J. A comparison of

various agarose preparations in a counter-immunoelectrophoresis

(CIE) system for assaying urinary myoglobin. J. Immunol.

Methods 14:59-63, 1977.

24. Tripodi, D., Kochesky, R., Lyons, S. and Davis, 0. Immuno-

chemistry of counterimmunoelectrophoresis and the effect of

electroendosmotic flow on reactivity. J. Immunol. Methods

4:1-10, 1974.

25. Lund, E. Laboratory diagnosis of pneumococcus infections.

Bull. WHO 23:5-13, 1960.

31
26. Enyvall, E. and Perlmann, P. njizymie linked immuriosorbent

assay (ELISA) Quantitative assay of immunoglobu]in G.

Ir;jnuriochem. 8:871-874, 1971.

27. van Weeman, B.K. and Schuurs, A.H.W.M. Irrnunoassay using

antigen-enzyme conjugates. FEBS Letters 15:232-236, 1971.

28. Nakane, P.K. and Pierce, G.B. Enzyme labeled antibodies:

Preparation and application for the localization of antigens.

J. Histochem. Cytochem. 14:929-931, 1967.

29. Avrameas, S. and Uriel, J. Methode de Marquage d'antigenes

et d'anticorps avec des enzymes et son application en

immunodiffusion. Compt. Rend. 262:2543, 1966.

30. Tew, J.G., Burmeister, J., Greene, E.J., Pflaumer, S.K. and

Goldstein, J. A radioimmunoassay for human antibody specific

for microbial antigens. J. Imm. Methods 14:231-241, 1977.

31. Strange, R.E. and Hambleton, P. An indirect immunoradiometric

assay of microbes. J. Gen. Microbiol. 93:401-404, 1976.

32. Robern, H., Dighton, M., Yano, Y. and Dickie, N. Double-

antibody radioimmunoassay for staphylococcal enterotoxin C 2 .

Appl. Microbiol. 30:525-529, 1975.

33. Boroff, D.A. and Shu-Chen, G. Radioimmunoassay for type A

toxin of Clostridium botulinum. Appl. Microbiol. 25:545-549, 1973.

34. Catt, K. and Tregear, G.W. Solid-phase radioimmunoassay

in antibody-coated tubes. Science 158:1570-1573, 1967.

35. Boenisch, T. Improved enzymeimmunoassay for trace proteins

in protides of the bioloqical fluids 24th colloquium, Peeters,

H., ed., p. 743-749, Pergamon Press, New York 1976.

32
36. Engvall, E., Jonsson, K. and Perlmann, P. Enzyme linked

imrimunosorbent assay., II. Quantitative assay of protein

antigen, imrnunoglobulin G, by means of an enzyme labeled

antigen and antibody coated tubes. Biochim.Biophys. Acta

251:427-434, 1971.

37. Nakane, P. K. and Kawaoi, A. Peroxidase labeled antibody.

A new method of conjugation. J. Histochem. Cytochem.

22:1084-1091, 1974.

38. Cawley, L. P., Minard, B. J. and Gillissen, J. A. Enzyme-

linked immunosorb assay (ELISA). Paper presented at the

American Society of Clinical Pathologists Workshop on ELISA,

7-9 September 1977, Chicago, Illinois 60612.

39. Ruitenberg, E.J., Steerenberg, P.A., Brosi, B.J.M. and Buys, J.

Serodiagnosis of Trichinella spiralis infections in pigs by

enzyme-linked immunosorbent assays. Bull. WHO 51:108-109, 1974.

40. Avrameas, S. and Ternynck, T. Peroxidase labeled antibody and

Fab conjugates with enhanced intracellular penetration.

Immunochem. 8:1175-1179, 1971.

41. Kato, K., Hamaguchi, Y., Fukui, H. and Ishikawa, E. Enzyme-

linked immunoassay. Conjugation of rabbit anti(human immuno-

globulin G) antibody with ý-D-galactosidase from Escherichia

coli and its use for human immunoglobulin G assay. Eur. J.

Biochem. 62:285, 1976.

42. Saunders, G.C. and Clinard, E.H. Rapid micromethod of screening

for antibodies to disease agents using the indirect enzyme-

labeled antibody test. J. Clin. Microbiol. 3:604-608, 1976.

43. Saunders, G.C. and Wilder, M.E. Disease screening with enzyme-

labeled antibodies. J. Infect. Dis. 129:362-364, 1974.

33
44. Faupners, G.C. and Bartlett, M.L. Double antibody solid phase

enzyme immunoassay for the detection of staphylococcal entero-

toxin A. Report LA-UR 77-285 Revised, Los Alamos Scientific

Laboratory, Los Alamos, N..M. 87545. Submitted to: Applied and

Environmental Microbiology.

45. Ruitenberg, E.J., Steerenberg, P.A., Brosi, B.J.M. and Buys,.7.

Reliability of the enzyme-linked immunosorbent assay (ELISA)

for the serodiagnosis of Trichinella spiralis infections in

conventionally raised pigs. J. Immunol. Mlethods 10:67-83, 1976.

46. Voller, A., Bidwell, D., Huldt, G. and Engvall, E. A micro-

plate method of enzyme-linked immunosorbent assay and its

application to malaria. Bull. WHO 51:209-211, 1974.

47. Voller, A., Bidwell, D.E. and Bartlett, A. A serological

study of human Trypanosoma rhodesiense infections using a

microscale enzyme-linked immunosorbent assay. Tropenmed.

Parasitol. 26:247-251, 1975.

48. Ruitenberg, E.J. and Buys, J. Application of the enzyme-

linked immunosorbent assay (ELISA) for the serodiagnosis

of human African trypanosomiasis (sleeping sickness).

Am. J. Trop. Med. Hyg. 26:31-36, 1977.

49. Veldkamp, J. and Visser, A.M. Application of the enzyme-

linked immunosorbent assay (ELISA) in the serodiagnosis

of syphilis. Brit. J. Vener. Dis. 51:227-231, 1975.

50. Huldt, G., Lagerquist, B., Phillips, T., Draper, C.C. and

Voller, A. Detection of antibodies in schistosomiasis by

enzyme-linked immunosorbent assay (ELISA). Annals of Trop.

Med. Parasitol. 69:483-488, 1975.

34
51. Russell, H., Facklam, R.R. and Edwards, L.R. Enzyme-linked

iwmaunosorbent assay for streptococcal M protein antibodies.

J. Clin. Microbiol. 3:501-505, 1976.

52. Wolters, G., Kuijpers, L., Kacaki, J. and Schuurs, A. Solid-

phase enzyme-immunoassay for detection of hepatitis B surface

antigen. J. clin. Path. 29:873-879, 1976.

53. Scharpe, S.L., Cooreman, W.M., Blomme, W.J. and Laekeman, G.M.

Quantitative enzyme immunoassay: Current status. Cdin. Chem.

22:733-738, 1976.

54. Wisdom, G.B. Enzyme-immunoassay. Clin. Chem. 22:1243-1255,

1976.

35

~L
 UNCLASSIFIED
SECURITY CLASSIFICATION OF TIIIS PAGE 'Hh-n D).i Et-,rd)

REPORT DOCUMENTATION PAGE READ INSTRUCTION.

BEFORE COMPI- ETING FORM
I kL;EUT NUiMLUR 21• OVT ACCESSION NO 3 EIP
NT'S CATALOG NUMbER

4 TITLE (e-d ýhutltll) 5 TYPE OF REPORT 6 PERIOD CCVENED

Immunodiagnostic Techniques for

Bacterial Infections. Final
6 PERFORMING ORG. REPORT NUMBER

7 AUTHOR(.) B CONTRACT OR GRANT NUMBER(*)

Edwards, E. A., Rytel, M.W. and

Hilderbrand, R.L.
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASK
Naval Health Research Center AREA A WORK
AE OKUI UNIT NUMBERS
UBR

P.O. Box 8S122 M0095.PN.002-5044

San Diego, CA 92138
11. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

Naval Medical Research & Development Command 31 October 1977

National Naval Medical Center 13. NUMBER OF PAGES
Bethesda, ýD) 20814 35
14. MONITORING AGENCY NAME 8 AbDRESS(lf different from Controlling Office) IS. SECURITY CLASS. (of fthi report)

Bureau of Medicine and Surgery

Department of the Navy UNCLASSIFIED
Washington, D. C. 20372 ,15. DECLASSIFICATION,
SCHEDULE
DOWNGRADING

16. DISTRIBUTION STATEMENT (of this Report)

"Approved for public release; distribution unlimited.

17. DISTRIBUTION STATEMENT (of the abstract entered In Block 20, II different from Report)

IS. SUPPLEMENTARY NOTES

In: A Balows & WJ Hausler, Jr. (eds), Diagnostic Procedures for BacteriaZ,
Mycotic, and Parasitic Infectiono, 6th Ed. Washington, DC: American Public
Health Assn, 1981. pp 767-790.
19. KEY WORDS (Continue on reverse mide If necessary ind Identify by block number)

Immunodiffusion; Counterimmunoelectrophoresis;
Staphylococcus coagglutination; Rocket Electrophoresis;
Enzyme-Linked Immunosorbent Assay.
20. ABSTRACT (Continue on reverse side It neceeeary and identify by block number)

Methods are described which can be used in immunodiagnosis of

infectious disease. Included are methods for immunodiffusion,
counterimmunoelectrophoresis, rocket immunodiffusion, staph-
lococcus coagglutination test, and enzyme-linked imnunosorbent
assay. The theory of each test system and clinical conditions
in which each test procedure has been used or would be
potentially useful is described.

DD I 1473 EDITION OF I NOV GS IS OBSOLETE UNCLASSIFIED

S/N OO12LF 0146•601
SECURITY CLASSIFICATION OF THIS PAGE (167., Dar. Entered)

•,•,, • ,;: ,.••'4y