PHB 3
PHB 3
PHB 3
Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech
Universit de Toulouse, INSA, UPS, INP, LISBP, 135 Avenue de Rangueil, F-31077 Toulouse, France
INRA, UMR792 Ingnierie des Systmes Biologiques et des Procds, F-31400 Toulouse, France
c
CNRS, UMR5504, F-31400 Toulouse, France
d
VEOLIA Environnement, Centre de Recherche sur lEau, Chemin de la Digue, BP 76, F-78603 Maisons-Laftte Cedex, France
b
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
a r t i c l e
i n f o
Article history:
Received 20 June 2013
Received in revised form 31 July 2013
Accepted 2 August 2013
Available online 26 August 2013
Keywords:
Cupriavidus necator
Polyhydroxybutyrate (PHB)
Volatiles fatty acids (VFA)
Fed-Batch fermentation
Metabolic Flux Analysis (MFA)
a b s t r a c t
In this study a complementary modeling and experimental approach was used to explore how growth
controls the NADPH generation and availability, and the resulting impact on PHB (polyhydroxybutyrate)
yields and kinetics. The results show that the anabolic demand allowed the NADPH production through
the Entner-Doudoroff (ED) pathway, leading to a high maximal theoretical PHB production yield of
0.89 Cmole Cmole1; whereas without biomass production, NADPH regeneration is only possible via
the isocitrate dehydrogenase leading to a theoretical yield of 0.67 Cmole Cmole1. Furthermore, the maximum specic rate of NADPH produced at maximal growth rate (to full biomass requirement) was found
to be the maximum set in every conditions, which by consequence determines the maximal PHB production rate.
These results imply that sustaining a controlled residual growth improves the PHB specic production
rate without altering production yield.
2013 Elsevier Ltd. All rights reserved.
Corresponding author. Address: Laboratoire dIngnierie des Systmes Biologiques et des Procds INSA, UMR INSA/CNRS 5504 UMR INSA/INRA 792, 135
Avenue de Rangueil, 31077 Toulouse Cedex 4, France. Tel.: +33 (0)5 61 55 94 70.
E-mail addresses: estelle.grousseau@insa-toulouse.fr (E. Grousseau), blanchet.
elise@gmail.com (E. Blanchet), stephane.deleris@veoliaeau.fr (S. Dlris), Maria.
ALBUQUERQUE@veolia.com (M.G.E. Albuquerque), etienne.paul@insa-toulouse.
fr (E. Paul), uribelarrea@insa-toulouse.fr (J.-L. Uribelarrea).
0960-8524/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2013.08.120
1. Introduction
There exists increasing industrial interest in the microbial
production of polyhydroxyalkanoates (PHA). PHA are biobased
and biodegradable polymeric materials that can be an interesting
substitute to petroleum derived plastics. PHA production naturally
31
Nomenclature
ai
DCW
kED
kTCA
nCi
qi
qNADPH
qNADPH
QPHB
RQ
rj
rs
molar
accumulation
of
the
ith
metabolite
(mole Cmole1 h1)
dry cell weight (g L1)
proportional factor of NADPH generation by ED to
NADPH requirement for anabolism
maximum NADPH ux generated by the tricarboxylic
acid cycle (TCA) (mmole Cmole1 h1)
number of carbon of the metabolite i (Cmole mole1)
specic i production rate (Cmole Cmole1 h1)
Xrmax maximum specic rate of NADPH generated
(reached at lmax) (mmole Cmole1 h1)
required for anabolic demand
Xr NADPH
(mmole Cmole1 h1)
overall volumetric PHB productivity (g L1 h1)
respiratory quotient (Cmole mole1)
specic molar ux of the reaction j (mole Cmole1 h1)
substrate consumption rate (Cmole L1 h1)
occurs under the unfavorable growth conditions of nutrient limitation such as oxygen, nitrogen, phosphorus, with adequate availability of carbon (Koller et al., 2010). About 90 monomers of
hydroxyalkanoates have been reported in the literature (Steinbuchel and Valentin, 1995). Among them 3-hydroxybutyrate (3HB)
is the most common naturally occurring from various carbon
sources.
Hundreds of bacterial genera are able to synthesize and store
PHA (Koller et al., 2010). Among them Cupriavidus necator (formerly known as Ralstonia eutropha) has been widely used because
of its industrial interest (Verlinden et al., 2007). This strain is a
gram-negative obligate aerobic bacterium living autotrophically
or heterotrophically using a variety of carbon sources.
Despite signicant decreases in recent years, PHA production
costs are still high compared to current traditional plastics. High
substrate conversion yield, high productivity, and high PHA content in microbial cells are important industrial criteria to decrease
production costs (Choi and Lee, 1999). The use of low cost, renewable carbon sources is also of major interest for an economic and
sustainable production of PHA (Koller et al., 2010). For example,
the by-products from biodiesel industry (Garcia et al., 2013; Vrana
Spoljaric et al., 2013) or Volatile Fatty Acids (VFA) are both suitable
carbon sources. VFA such as valeric, butyric, propionic, lactic, or
acetic acids, can be obtained from acidogenic fermentation of different kinds of waste streams such as industrial, agricultural or
municipal wastewater (Du et al., 2004; Li and Yu, 2011; Yu,
2001; Marang et al., 2013). PHA biosynthesis has been extensively
investigated on carbohydrates as single substrate and very high
titers, productivities and yields have been achieved (Ryu et al.,
1997). In contrast, limited knowledge is available on VFA substrates utilization for PHA production. In particular, butyric acid
has not been extensively studied (Du et al., 2004; Yu and Si,
2004; Zinn et al., 2003) despite great conversion yield (Shi et al.,
1997; Marang et al., 2013). In order to determine the carbon ux
distribution during PHA biosynthesis from various carbon sources,
stoichiometric and Metabolic Flux Analysis (MFA) approaches have
been proposed (Shi et al., 1997; Yamane, 1993; Yu, 2001). Although
all of these studies pointed out the critical role of NADPH in PHA
synthesis, the inuence of NADPH generation pathway on PHA production has not been intensively investigated. Moreover few
authors have evaluated the effects of sustaining growth during
PHA-production on organic acids (Kobayashi et al., 2000; Shimizu
et al., 1993; Yu and Si, 2004).
S
Xr
YATP,Xr
YP,Xr
YS,i
Y tS;i
Y lim
S;i
Y theo
S;i
%PHA
l
ei;j
substrate
residual biomass
energetic biomass yield (gXr MoleATP1)
overall residual biomass production on phosphorus
yield (Cmole Mole1)
overall i production on substrate yield (Cmole Cmole1)
instantaneous i production on substrate yield
(Cmole Cmole1)
limit i production on substrate yield (Cmole Cmole1)
theoretical
i
production
on
substrate
yield
(Cmole Cmole1)
intracellular PHA content (g g1)
growth rate (h1)
stoichiometric coefcient for the metabolite i in the
reaction j
In this study, well controlled phosphorus limited fed-batch cultures were carried out with butyric acid as substrate under noninhibitory feeding conditions. A stoichiometric model and a kinetic
model of the organism were constructed. The experiments and
modeling were integrated to explore how the microbial growth
sustained by a controlled phosphorus feeding inuences the
NADPH generation pathway and availability, and to explain the
resulting impacts on polyhydroxybutyrate (PHB) yields and
kinetics.
2. Methods
2.1. Micro-organism
The strain C. necator DSM 545 (or H1 G+3) was obtained from
DSMZ (Germany). The strain was stored at 80 C in liquid NB
medium (Nutrient Broth Merck: Meat Peptone 5 g L1, Meat extract 3 g L1) with 20% glycerol (v/v). Strain DSM 545 is a spontaneous mutant of the strain DSM 529 (H1) and is able to
catabolize glucose unlike the wild type C. necator DSM 428 (H16).
2.2. Seed cultures
One glycerol stock was streaked on a NB medium Petri dish (NB
with addition of Agar 15 g L1). The plate was incubated for
2448 h at 30 C. One colony was used to inoculate the rst seed
culture grown for 12 h in tube with 10 mL of Liquid NB medium
(30 C, 200 RPM). The second seed culture was grown for
1215 h with 150 mL of mineral-salt medium in a 1 L Erlenmeyer
ask (30 C, 100 RPM). The mineral salt medium composition is described in the literature by Aragao et al., 1996 except for the carbon
source (Glucose added to a nal concentration of 4 g L1) and for
the phosphorus source (2 mL/L of a solution of 224 g L1 of Na2HPO412 H2O and 37.5 g L1 of KH2PO4). This latter culture was
used to inoculate the bioreactor.
2.3. Fed-Batch culture conditions and bioreactor system
The Fed-Batch culture was performed in a 2 L (working volume)
fermentor (Braun Biotech, Melsungen Germany) with an initial
medium volume of 1.2 L, equipped with pH, dissolved oxygen
(DO), temperature, pressure and Anti-foam controllers. The on-line
32
All Yields are expressed as carbon ratios except for the energetic
yield. Instant Yield was the ratio of specic i production rate and
specic substrate consumption rate: Y tS;i qi =qS . Overall Yield
was the ratio of produced quantities of i and consumed quantities
of substrate during a time interval (t2t1): YS,i = |(it2 it1)/
(St2 St1)|.
2.5.2. Modeling metabolic descriptor
The metabolic descriptor was composed of two equation
systems:
(1) The rst is an anabolic network (53 reactions: Supplementary online material S2) leading to the synthesis of different
macromolecular components of biomass from 28 intracellular intermediate metabolites. This network has been used to
calculate the global molar stoichiometry (reaction r0, Supplementary online material S1) for the production of 1 g of
biomass, assuming the biomass formula (C1H1.77O0.44N0.25,
4% ash, MMr = 25.35 g mole1) and the mass fraction of macromolecular components (Proteins 74.00%, RNA 3.58%, DNA
1.85%, Phospholipids 6.49%, Polysaccharides 4.34%, Peptydoglycane 2.64%, LPS 4.98%, Polyamine 0.11%, water 7.01%,
ashes 3.92%).
(2) The second is a catabolic network (Supplementary online
material S1) which describes carbon assimilation including
both PHB synthesis and the central metabolic pathways such
as glyoxylate shunt (GXS), tricarboxylic acid cycle (TCA), gluconeogenesis (45 reactions, 46 intermediate metabolites:
Supplementary online material S3 and S4).
2.5.3. Modeling metabolic routes
Metabolic routes used in this study were obtained from a bibliographic study and available databases such as KEGGS. It is currently accepted that the metabolic routes from basic precursors
to small molecules are common to a wide variety of organisms.
Where specic data were not available for C. necator, composition
data of Escherichia coli were used instead.
Catabolism of butyric acid: Butyric acid in un-ionized form penetrates into the cells and is activated by the sequential action of
butyrate kinase and phosphate butyryltransferase (r1, Supplementary online material S1). Butyryl-CoA is converted to acetoacetylCoA via b-oxidation (Shi et al., 1997) (r2 to r4). It is then either
decomposed to acetyl-CoA by a b-ketothiolase to produce biomass,
energy and co-factors via the central metabolism or it is incorporated into PHB (r5 to r6).
Central metabolic pathways: The acetyl-CoA is then directed towards either the tricarboxylic acid cycle (TCA, r8 to r15) or towards
the glyoxylate shunt (GXS, r29 to r30). There are two paths for bioconversion of C4 to C3 (Bruland et al., 2010) to feed gluconeogenesis: (i) the malic enzyme (EM, r17) which decarboxylates malate to
pyruvate, (ii) the PEP (Phospho-Enol-Pyruvate) carboxykinase
(PCK, r16) which catalyzes the decarboxylation of oxaloacetate into
P-enolpyruvate. The glycolysis pathway and the PentosePhosphate Pathway (PPP) are incomplete in C. necator because
the key enzymes fructose-1,6-bisphophatase (FBP) and gluconate
6-phosphate dehydrogenase (GND) are lacking. (Pohlmann et al.,
2006). However, the Entner-Doudoroff pathway (ED, r27 and r28)
overcomes this limitation.
Biosynthesis of PHB from acetoacetyl-CoA: acetoacetyl-CoA is reduced to R-hydroxybutyryl-CoA by a NADPH acetoacetyl-CoA
reductase (r5). The last necessary enzyme in PHB production is
the PHA synthase (r6) which polymerises monomers (Braunegg
et al., 1998).
The NADPH consumed by the PHB synthesis reaction (r5) must
be generated by some other biochemical reactions in order to
ensure continuous PHB synthesis. As indicated before, the
33
Fig. 1. Time course of Fed-Batch culture of C. necator with butyric acid as carbon source.
34
Table 1
Comparison of PHA production with C. necator from butyric acid in Fed-Batch mode.
Reference
This work
Yu (2001)
Du et al. (2004)
Carbon source
Butyric acid
Butyric acid
Limiting element
Process time (h)
DCW (g L1)
PHB (g L1)
PHV (g L1)
%PHA (g g1)
QPHB (g L1 h1)
YS,PHA (Cmole Cmole1)
P
67.4
46.7
38.4
0.0
82%
0.57
0.62
N
46.0
20.0
9.4
0.0
47%
0.20
nr
Acetic, propionic,
butyric and
lactic acid
nr
79.0
22.7
16.0
0.5
73%
0.21
nr
Table 2
Experimental and limit overall yields, Respiratory quotient (RQ), during each stage of
the Fed-Batch culture (strain C. necator).
Fed-Batch culture phase
0.59
0.04
0.64
0.15
0.62
0.58
0.80
0.07
0.67
0.61
0.75
1
Y lim
S;PHB (Cmole Cmole )
spilling by Shi et al., 1997. In our model a yield of 0.6 was reached
with a YATP,Xr of about 10.5 gXr moleATP1 instead of 14.
During phase C, the overall experimental PHB production yield
was 0.62 Cmole Cmole1 (Table 2) occurring simultaneously with a
biomass production. The decrease in biomass production in subsequent phase (D), did not signicantly increase the overall experimental PHB production yield (0.67 Cmole Cmole1, Table 2).
3.2. Sustaining a low residual growth did not alter the PHB production
yield
It is generally accepted in the literature that PHB yield increases
with biomass yield decrease (Shi et al., 1997). In fact, nutrient starvation is often used as a trigger for batch or fed-batch PHA production processes. As depicted in Fig. 2-A, while instantaneous
experimental PHB production yield (Y tS;PHB ) from butyric acid are
clearly increasing when the growth rate decreased from 0.33 h1
down to 0.06 h1, however, below 0.06 h1, this relationship
changes. Our current results shown that sustaining a low residual
growth (l equal or lower than 0.06 h1) via a controlled phosphorus feeding, did not affect signicantly the instantaneous experimental PHB production yield (Y tS;PHB ) from butyric acid. To
understand what happened for this growth rate range, limit PHB
production yields (Y lim
S;PHB ) from butyric acid have been calculated
and are depicted versus growth rate on the Fig. 2-B. This Figure shows that with the decrease in growth rate from 0.06 h1 to
0 h1, the limit PHB yield decreased from about 0.83 to
0.71 Cmole Cmole1. In order to understand the carbon allocation
during PHB production phases (C and D), the theoretical PHB production depending on the NADPH sources (ED, TCA, or EM described in Supplementary online material S1) have been
simulated using the metabolic descriptor. This simulation shows
that the theoretical PHB production yield was greatly dependent
on NADPH generation pathway (Table 3):
(1) If NADPH is generated by the isocitrate dehydrogenase
NADPH linked to the TCA or by the EM, the theoretical yield
of PHB production is 0.67 Cmole Cmole1, and the RQ is
0.67 Cmole Mole1. This is in accordance with Yamane,
1993 and Shi et al., 1997, who respectively calculated 0.66
and 0.68 using similar hypotheses. In our simulation, ATP
is generated in association with PHB production and is not
limiting: the ratio of ATP spilled to Carbon consumed is
between 1330 and 1500 mmole Cmole1 (Table 3) which is
4 to 5 time higher than during growth. As seen in Table 3,
results were the same if NADPH was generated by TCA or
EM. Therefore to simplify the system, EM pathway was
removed.
(2) If NADPH is generated by a cyclic carbon circulation in the
ED pathway, the theoretical PHB production yield is
0.89 Cmole Cmole1 and the RQ is 0.45 Cmole Mole1. In this
case, ATP generation is limiting (0 mmole of spilled ATP per
Cmole, Table 3). Energy spilling reaction, protein turn-over,
maintenance (ionic transports, homeostasis. . .) must be
Table 3
Theoretical Yields of i production on substrate (butyric acid) calculated with stoichiometric modeling ( parameters set).
i
Growth
PHB production
Xr
PHB
PHB
PHB
CO2
Y theo
(Cmole Cmole1)
S;i
EM/TCA/ED
0.640.67
EM or TCA
0.67
ED
0.89
ED
0.82
EM/TCA/ED
1.00
RQ (Cmole Mole1)
Spilled ATP (mmole.Cmole1)
YATP,Xr (g moleATP1)
0.630.61
320330
14
0.67
13301500
0.45
0
0.55
300
0.80
4390
35
Such high yield of PHB production on butyric acid with C. necator has been reported once in the literature and a similar statement
has been used by Shi et al., 1997: these authors suggested the
cyclic use of pentoses-phosphate pathway (which leads to
1
Y theo
). However, from the literature data,
S;PHB = 0.82 Cmole Cmole
the gluconate 6-phosphate dehydrogenase (GND) is lacking in
C. necator (Pohlmann et al., 2006), and therefore, the pentosesphosphate pathway cannot be used. Alternatively such high yields
may be achieved by considering the activity of a transhydrogenase.
But this hypothesis seemed less plausible because there is no
objective reason to correlate this enzyme activity to anabolism contrary to the Entner-Doudoroff pathway. High yields with butyric acid
have also been reported in microbial enrichment cultures instead of
axenic cultures of C. necator (Marang et al., 2013). These results
emphasized the need of optimization of waste-streams digestion to
promote butyrate yield over other acids in resulting mixture.
To note, there is a debate in literature on isocitrate dehydrogenase regarding the existence of NAD+-dependent activity: two
genes (icd1 and icd2) coding for isocitrate dehydrogenase were
identied (Wang et al., 2003), but no NAD+-dependent isocitrate
dehydrogenase activity was detected in cells extracts. Nevertheless
an older study cited by Wang et al. (2003) revealed an isocitrate
dehydrogenase with NAD+-dependent activity. Two isocitrate
dehydrogenases with respective afnity to NADP+ and NAD+ were
separated and partially puried. Sequence analysis identied an
additional gene coding for an isocitrate dehydrogenase icd3 (locus
H16_B1016) in C. necator (Pohlmann et al., 2006) although not
experimentally studied. This third gene could encode for a NAD+
dependent enzyme. No studies have been reported on regulation
of isocitrate dehydrogenase isoenzymes with respect to NADH or
NADPH generation. Considering all these informations, both
NAD+- and NADP+-dependent activities were considered in the
stoichiometric model.
(Y tS;PHB )
36
Table 4
Experimental and simulated data (in brackets) of the Fed-Batch of C. necator with butyric acid as carbon source.
t = 30,4 h
t = 32,2 h
t = 37,7 h
t = 43 h
t = 48,6 h
t = 57,3 h
t = 58,8 h
t = 65,3 h
t = 67,5 h
l (h1)
0,33a
0,056
0,048
0,045
0,038
0,034
0,014
0,013
0,012
0,007
0,004
(14)
(9,54)
(3,94)
(3,57)
(3,21)
(4,9)
(1,63)
(1,49)
(1,49)
(0,95)
(0,49)
(316)
(206)
(575)
(635)
(855)
(653)
(1137)
(938)
(888)
(1009)
(1275)
0,65
(0,65)
0,14
(0,15)
0,12
(0,11)
0,12
(0,11)
0,13
(0,13)
0,17
(0,17)
0,08
(0,08)
0,06
(0,06)
0,06
(0,06)
0,04
(0,04)
0,025
(0,026)
0,26
(0,23)
0,26
(0,29)
0,25
(0,27)
0,18
(0,18)
0,12
(0,12)
0,11
(0,11)
0,14
(0,15)
0,14
(0,15)
0,14
(0,12)
0,12
(0,11)
0,64
(0,64)
0,67
(0,67)
0,66
(0,66)
0,60
(0,60)
0,58
(0,58)
0,62
(0,62)
0,68
(0,68)
0,70
(0,70)
0,70
(0,70)
0,68
(0,67)
QR (Cmole Mole1)
0,63
(0,63)
0,56
(0,57)
0,56
(0,58)
0,57
(0,58)
0,62
(0,62)
0,64
(0,60)
0,68
(0,64)
0,60
(0,62)
0,63
(0,61)
0,66
(0,62)
0,65
(0,65)
(45.5)
(72.0)
(44.9)
(40.2)
(20.1)
(18.1)
(6.7)
(14.0)
(15.1)
(10.1)
(4.0)
(55.6)b
(3.1)
(15.9)
(41.6)
(35.5)
(21.4)
(24.8)
(26.8)
(25.2)
(22.5)
(23.5)
10
Total NADPH
(mmole Cmole1 h1)
(101.1)
(75.1)
(86.2)
(81.8)
(55.6)
(39.5)
(31.5)
(40.9)
(40.4)
(32.6)
(27.5)
11
(101.1)c
12d
13
lmax.
kTCA.
qNADPH#Xr max .
Average of row 12: kED = 3 1.
t = 25,6
30,4 h
b
c
Growth
(128.2)
(17.2)
(14.7)
(13.8)
(11.6)
(10.4)
(4.3)
(4.0)
(3.7)
(2.2)
(1.2)
(4.2)
(3.1)
(2.9)
(1.7)
(1.7)
(1.6)
(3.5)
(4.1)
(4.7)
(3.3)
(91.1)
(106.7)
(102.4)
(73.2)
(50.1)
(43.8)
(54.2)
(53)
(43.8)
(39.2)
Fig. 3. Relationship between specic PHB production rate (qPHB) and growth rate
(l) considering simulated and experimental data during Fed-Batch culture of
C. necator on butyric acid.
for very low growth rate may be further rened by considering the
reduction of catalytic capacity with phosphorus limitation and to
the operation of physical constraints with PHB accumulation as
has been shown in previous studies (Heinzle and Lafferty, 1980).
To take these phenomena into account kTCA and kED could decrease
with the growth rate instead of being held constant.
This model shows that maximal specic PHB production rate is
dened by the maximum specic rate of NADPH produced, which
depends on anabolism of the strain. It therefore follows that the
higher the maximal growth rate of the strain, the higher is the specic PHB production rate. Since PHA production is closely related
to growth of the strain, the study of growth capacity should not
be neglected in order to optimize the PHA production process.
Moreover, the choice of a PHB production microorganism can be
principally driven by this criterion.
Among existing models on PHB production (Gahlawat and Srivastava, 2013; Shi et al., 1997; Vrana Spoljaric et al., 2013; Yamane,
1993; Yu and Si, 2004), this work used a stoichiometric model
combined with a simple kinetic model. This strategy proved to
be a useful tool to understand biological phenomenon and to guide
in designing operating strategies to further improve PHB
production.
4. Conclusion
In this work, the complementary modeling and experimental
approach validates that sustaining a residual growth do not alter
PHB production yield and improves PHB specic production rate
during cultivation of C. necator strain with volatile fatty acids as
sole substrate. Further, that a specic growth rate of about
0.05 h1 is the optimum set point. Under these conditions, high
specic productivity of 0.35 Cmole Cmole1 h1, high PHB production yield of 0.67 Cmole Cmole1 and by consequence high PHB
content (about 85%) could be simultaneously reached. Moreover
this paper highlights the importance of the knowledge of the strain
anabolism to optimize PHB production.
Acknowledgements
This work was supported by Veolia Environment Research and
Innovation, by grants from the French National Research Agency
(ANR-08-ECOT-O17-001) and the French National Research and
Technology Association (ANRT, Cifre). We thank Dr. Nathalie
Gorret, Dr. Stphane Guillouet, and Dr. Dores Cirne for the critical
review of this manuscript; Dr. Anne-Sophie Lepeuple and
Dr. Pierre-Alain Hoffmann for their support throughout the course
of this study. And we also kindly thank Mr. John W. Quimby for his
help in correcting the English version of the manuscript.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.biortech.2013.08.
120.
References
Allen, R.J.L., 1940. The estimation of phosphorus. Biochemical Journal 34, 858865.
Ampe, F., Leonard, D., Lindley, N.D., 1996. Growth performance and pathway ux
determine substrate preference of Alcaligenes eutrophus during growth on
acetate plus aromatic compound mixtures. Applied Microbiology and
Biotechnology 46, 562569.
Aragao, G.M.F., Lindley, N.D., Uribelarrea, J.L., Pareilleux, A., 1996. Maintaining a
controlled residual growth capacity increases the production of PHA
copolymers by Alcaligenes eutrophus. Biotechnology Letters 18, 937942.
Brandl, H., Gross, R.A., Lenz, R.W., Fuller, R.C., 1988. Pseudomonas oleovorans as a
source of poly(beta-hydroxyalkanoates) for potential applications as
37
38
Wang, Z.X., Bramer, C., Steinbuchel, A., 2003. Two phenotypically compensating
isocitrate dehydrogenases in Ralstonia eutropha. Fems Microbiology Letters 227,
916.
Yamane, T., 1993. Yield of poly-D(-)-3-hydroxybutyrate from various carbon
sources a theoretical study. Biotechnology and Bioengineering 41, 165170.
Yu, J., 2001. Production of PHA from starchy wastewater via organic acids. Journal of
Biotechnology 86, 105112.
Yu,