Bioscience Reports, Vol. 8, No.

6, 1988

Polyhydroxybutyrate: an Intriguing
Biopolymer
Edwin A. Dawes

The microbial polymer poly-3-hydroxybutyrate (PHB) and related poly-hydroxyalkanoates, such as

poly-3-hydroxyvalerate and poly-3-hydroxyoctanoate, are unique biodegradable thermoplastics of
considerable commercial importance. The structure, properties and regulation of synthesis and
degradation of PHB are reviewed and the microbial production of copolymers of 3-hydroxybutyrate
and 3-hydroxyvalerate, with properties varying according to copolymer composition, is discussed.

KEY WORDS: polyhydroxybutyrate, biopolymer.

INTRODUCTION

Poly-3-hydroxybutyrate (PHB) is a linear polyester of D(-)-3-hydroxybutyric acid

which was first discovered in bacteria by Lemoigne in 1925. It is accumulated in
intracellular granules by a wide variety of Gram-positive and Gram-negative
organisms under conditions of a nutrient limitation other than the carbon source
(Dawes and Senior, 1973). The polymer, which serves as a reserve of carbon and
energy, is now known to be but one example, albeit the most abundant, of a
general class of compounds referred to as polyhydroxyalkanoates and possessing
the general formula,

O--CH--CH2--C-~

where R = C H 3 in PHB. The principal other polyesters currently identified have

R = CH3CH2 (polyhydroxyvalerate, PHV) and R = CH3(CH2)4 (polyhydroxy-
octanoate, PHO).
The molecular weight of PHB differs according to organism, conditions of
growth and method of extraction, and can vary from about 50,000 to well over a
million. The polymer possesses the important properties of thermoplasticity and

Department of Biochemistry, University of Hull, Hull HU6 7RX, U.K.

537
0144-8463/88/1200-0537506.00/0O 1988 Plenum Publishing Corporation
538 Dawes

biodegradability and, in consequence, has attracted considerable commercial

interest, to which I shall refer in greater detail later.
PHB is an ideal carbon reserve material since it exists in the cell in a highly
reduced state as a virtually insoluble polymer exerting negligible osmotic
pressure. Marchessault and his colleagues reported that the PHB molecule from
Rhizobium is a compact right-handed helix with a two-fold screw axis and a fibre
repeat of 59.6 nm (Okamura and Marchessault, 1967; Cornibert and Marches-
sault, 1972). The granules are generally spherical and vary in size according to the
organism, e.g. in Bacillus megaterium from 0.2 to 0.7 #m diameter (Ellar et al.,
1968). Electron microscopy of native granules isolated from cell-free extracts by
density-gradient centrifugation shows that they are bounded by a membrane
which is not a typical fluid mosaic membrane and which has associated with it the
final enzyme of polymerization, PHB synthase or polymerase. In some, but not
all, organisms studied, the initial enzyme system for PHB degradation, the
depolymerase, is also entirely associated with the granule membrane.

BIOSYNTHESIS OF PHB A N D ITS R E G U L A T I O N

When Azotobacter beijerinckii is grown in batch culture fixing atmospheric

nitrogen with glucose as the carbon source, it displays no microscopically
perceptible granules in the early exponential phase; by mid-exponential phase
granules of PHB start to accumulate and by the time the stationary phase is
attained the cells are packed with granules, the polymer representing about 75
per cent of the biomass. Chemostat experiments revealed that an oxygen
limitation yields a much greater polymer content than a nitrogen limitation, and
the amount of polymer accumulated is inversely related to the growth rate
(Senior et al., 1972).
When an oxygen limitation is imposed upon a steady-state chemostat culture
which is nitrogen limited, a period of unbalanced growth ensues during which the
initially very low PHB content of the cells increases until it attains a value
characteristic of the new steady state (some 50 per cent of the biomass in the
experiments conducted). Conversely, when an oxygen-limited culture has that
limitation relaxed to give a dissolved oxygen tension of 2 per cent of air
saturation, the PHB content of the cells declines and eventually reaches a much
lower value determined by the new steady state. If a redox electrode is inserted
into cultures undergoing such transitions of dissolved oxygen tension, interesting
observations may be made. Thus when an oxygen limitation is imposed upon a
nitrogen-limited culture operating at a dissolved oxygen tension of 10 per cent of
air saturation, the redox potential plummets from +15 mV to - 5 0 mV. But as
PHB synthesis commences in response to the oxygen limitation, the redox
potential increases rapidly and soon exhibits an Eh of over +30 mV; we thus have
the paradoxical situation of an oxygen-limited culture displaying a redox potential
higher than that of a culture growing in the presence of excess oxygen. To explain
these observations we suggested that PHB synthesis serves as an electron sink for
the reducing power which accumulates as a consequence of an oxygen limitation;
electrons are no longer able to traverse the electron transfer chain to oxygen at
Po,lyhydroxybutyrate 539

the same rate and thus are diverted to the reductive step of polymer synthesis
(Senior and Dawes, 1973). This hypothesis received support from measurements
of the intracellular N A D H / N A D § ratio during transition from a nitrogen to an
oxygen limitation. The ratio increased dramatically on imposition of the oxygen
limitation but then, as PHB synthesis commenced, decreased to a lower level but
one which was rather higher than that recorded for the original nitrogen-limited
steady state (Jackson and Dawes, 1976).
Investigations of carbohydrate metabolism in A . beijerinckii showed the
operation of the Entner-Doudoroff pathway with terminal oxidation via the
tricarboxylic acid cycle. We found that metabolism of glucose was subject to
regulation at three points as a consequence of an oxygen limitation. First, at the
g~ucose 6-phosphate dehydrogenase step: this is an allosteric enzyme inhibited by
ATP and, even more powerfully, by N A D P H and NADH. Second, citrate
synthase is subject to inhibition by N A D H , an effect which can be overcome by
1 mM AMP, in keeping with David Weitzman's general observation relating to
Gram-negative bacteria (Weitzman and Jones, 1968). The third enzyme under
control is isocitrate dehydrogenase, inhibited by N A D P H , N A D H and ATP.
Thus in the tricarboxylic acid cycle there are two points of control: citrate
synthase and isocitrate dehydrogenase. The cycle is, of course, generating
intermediates for biosynthesis as well as effecting oxidation and energy gener-
ation. Furthermore, the organism is fixing nitrogen, a process which requires both
reducing power and a significant input of energy, estimated at some 15 ATP per
mole N2 reduced.
Thus, when an oxygen limitation is imposed and the N A D H / N A D § ratio
rises, inhibition of citrate synthase and isocitrate dehydrogenase will occur, the
tricarboxylic acid cycle will slow down and this could have serious effects on the
growth of the organism. However, PHB synthesis, by effectively "mopping up"
some of the excess reducing power, redresses the N A D H / N A D + balance and
relieves the inhibition, enabling the cycle to operate at a rate faster than would
otherwise have been possible (Senior and Dawes, 1971).
Inhibition of citrate synthase, the initial enzyme of the tricarboxylic acid
cycle, leads to an accumulation of acetyl-CoA, the precursor of PHB, and to link
up these two segments of metabolism it is now necessary to consider the pathway
of polymer synthesis.
Early experiments had ruled out the involvement of an acyl carrier protein in
PHB synthesis and had shown that CoA esters were intermediates (Ritchie and
Dawes, 1969). The route of synthesis from acetyl-CoA established involves three
enzymes: 3-ketothiolase, acetoacetyl-CoA reductase and PHB polymerase. Two
molecules of acetyl-CoA are condensed by 3-ketothiolase, with the release of
CoA, to yield acetoacetyl-CoA, which is then reduced by the reductase to
D(-)-3-hydroxybutyryl-CoA (Ritchie et al., 1971) which, in turn, is polymerized
with the release of CoA. The biosynthetic pathway is controlled by 3-ketothiolase
activity. We discovered that the reversible reaction catalysed by this enzyme is
inhibited in the condensation direction by free CoA, while in the direction of
thiolysis high concentrations of acetoacetyl-CoA inhibit but this effect is reversed
by increasing the concentration of the other substrate, CoA (Senior and Dawes,
540 Dawes

1973). The Km for acetyl-CoA is quite high (0.9 raM). Similar behaviour had been
shown for the 3-ketothiolase of Hydrogenomonas eutrophus H16 (Alcaligenes
eutrophus) by Oeding and Schlegel (1973).
We found that acetoacetyl-CoA reductase was five-fold more active with
N A D P H than with N A D H throughout purification. It is a typical thiol enzyme
sensitive to the usual inhibitors. The final enzyme of synthesis is the polymerase
or synthase which is associated with the membrane of the polymer granule. This
enzyme is of crucial importance and the subject of intensive interest although, at
the present time, relatively little has been published. It is a fascinating system
involving a reaction in which soluble components of the cytoplasm react at a
surface to give a water-insoluble product which accumulates in presumably an
extremely hydrophobic environment within the granule. Griebel and Merrick
(1971) proposed that polymerization is a two-stage reaction involving the
formation of an acyl-enzyme intermediate via a functional thiol group on the
enzyme:
D( - )-3-hydroxybutyryl-CoA + Synthase-SH--*
3-hydroxybutyryl-S-Synthase + CoA
3-Hydroxybutyryl-S-Synthase + PHB primer (n units)
PHB primer (n + 1 units) + Synthase-SH.

D E G R A D A T I O N OF PHB A N D THE PHB CYCLE

Degradation of PHB occurs via a pathway which does not involve CoA
esters. In A. beijerinckii a depolymerase associated with the polymer granule
membrane initiates hydrolysis to yield free D ( - )-3-hydroxybutyrate, which is
then oxidized to acetoacetate by an NAD-specific dehydrogenase. This enzyme is
competitively inhibited by NADH, pyruvate and 2-oxoglutarate; N A D P H has no
effect. We discovered that the acetoacetate is next converted to acetoacetyl-CoA
via the action of an acetoacetate : succinate CoA transferase, an important finding
because it meant that PHB metabolism is a cyclic process, acetoacetyl-CoA
serving as a precursor of PHB synthesis and also as a product of its degradation
(Fig. 1).
It is of interest that while A. beijerinckii and Ale. eutrophus both possess an
acetoacetate:succinate CoA transferase, in Zoogloea ramigera acetoacetate is
esterified with CoA by a CoA-synthetase reaction employing ATP (Tomita et al.,
1983).
Acetoacetate + CoA + ATP---~ acetoacetyl-CoA + AMP + PPi

The existence of a PHB cycle demands efficient control to prevent futile

cycling. This is achieved by regulation, first of 3-ketothiolase. Under conditions of
ample oxygen supply pyruvate derived from glucose metabolism is oxidized to
acetyl-CoA which enters the tricarboxylic acid cycle; hence the intracellular
concentration of acetyl-CoA is low, citrate synthesis serving as a sink for acetyl
units and simultaneously releasing CoA. Thus 3-ketothiolase activity in the
Polyhydroxybutyrate 541

PYRUVATE

AcetoacetyI-SCoA ACETYL- SCoA ~ J

k _jt
Acetoocetyl- SCoA
I
I

// :..o,~," \N~ J

NADH

2- Oxoglutarate
ACETOACETATE

~NAOH /'
D(-)-3- H Y O R O X Y B U T Y R Y L - S C O A

O[-)-3-HYDROXYBUTYRATE [POLY-@-HYOROXYBUTYRATE I

Fig. 1. Cyclicmetabolism of PHB and its regulation in Azotobacter beijerinckii.

condensation direction is prevented both by the low acetyl-CoA concentration (as

previously noted, the Km for this substrate is relatively high) and the high CoA
concentration, which inhibits the condensation reaction. PHB synthesis does not
occur therefore under these conditions.
When, however, an oxygen limitation intervenes, citrate synthase is inhibited
by the increased N A D H / N A D § ratio, the acetyl-CoA concentration in conse-
quence increases, while the CoA concentration correspondingly decreases. These
conditions promote 3-ketothiolase activity in the condensation of acetyl-CoA to
acetoacetyl-CoA and the accumulated reducing power is used in the succeeding
reductive step, leading to PHB synthesis.
Turning to the control of PHB degradation, while intuitively one might
expect that the depolymerase will be subject to some control, insufficient is
known yet about this enzyme to support this belief. In the absence of such
information it is still possible to explain control of polymer degradation at the
next enzymic step, catalysed by D ( - )-3-hydroxybutyrate dehydrogenase. This
enzyme is competitively inhibited by N A D H and thus under an oxygen limitation,
when the N A D H / N A D § ratio will be high and polymer synthesis occurring, the
dehydrogenase will be inhibited and polymer degradation will not proceed.
However, even if the oxygen limitation is relaxed, provided there is still an ample
supply of carbon source the intracellular concentrations of pyruvate and
2-oxoglutarate would be expected to inhibit the enzyme and prevent overall
polymer degradation. Such control makes for cellular economy since physiologi-
cal evidence supports the role of PHB as a carbon and energy storage compound,
542 Dawes

and there would be no need to break down an intracellular reserve in the

presence of an exogenous carbon supply. But also, PHB serves as an admirable
redox regulator in the bacterial cell.
Our own researches on A. beijerinckii in Hull had been carried out in parallel
with the work of Hans Schlegel's group in G6ttingen using Alc. eutrophus H16.
The metabolic pathways and their regulation in both organisms were shown to be
similar although Alc. eutrophus was able to accumulate rather higher amounts of
PHB, in some cases up to 90% of the biomass.

COMMERCIAL EXPLOITATION OF POLYHYDROXYALKANOATES

Now I should like to turn to a consideration of the commercial interest that

has been aroused by PHB on account of its very important physical and chemical
characteristics. It is, of course, biodegradable and is broken down relatively
rapidly by soil micro-organisms. It is thermoplastic, melting at about 180~ and
can be moulded, cast into films, spun into monofilaments and produced in a
cotton wool-like form. It was in the early 1970s that Imperial Chemical Industries
PLC in Britain began work on the polymer. They evaluated three possible
organisms for the commercial production of PHB and finally settled for Alc.
eutrophus (see Byrom, 1987). Prior to a Biotechnology Conference at East-
bourne, England, in 1981, New Scientist presented the "leaked" information that
ICI would be announcing a new polymer that could be used for textile spinning
and which would be biodegradable. PHB was duly marketed by ICI under the
generic trade name of 'Biopol', and is offered in two weight-average molecular
weight forms: standard (423,000) and low (67,700). It is sold as pellets but for
customers who wish to cast it as a film or to spin monofilaments, it is also
available in solution.
The polymer is biocompatible as well as biodegradable and, of course, its
degradation product, 3-hydroxybutyrate is a normal mammalian metabolite. The
use of capsules of PHB for controlled drug release, filaments as surgical sutures,
and the "cotton wool" product as swabs has attracted medical attention.
Pure PHB is a highly crystalline thermoplastic and in consequence is rather
brittle and not sufficiently flexible for some purposes. To secure greater flexibility
the degree of crystallinity has to be decreased and it was discovered by Holmes et
aL (1981) that Alc. eutrophus could synthesize a range of copolymers, based on
3-hydroxybutyrate (HB) and 3-hydroxypentanoate (hydroxyvalerate, HV) mono-
mer units, which had these more desirable characteristics. As the HV content of
the copolymer rose so did its degree of crystallinity decrease and its flexibility and
toughness increase. An amusing anecdote seems appropriate here. Earlier this
year (1987) I appeared on a B.B.C. TV science programme, "Take Nobody's
Word for It", in the course of which I talked about, and showed some examples
of Biopol. My reward was a letter from a retired sewage chemist whose working
life until the availability of the contraceptive pill had been complicated by the
disposal of condoms. Now, after some twenty years, the advent of AIDS had
resulted in both medical and governmental authorities urging the use of condoms,
and so the old problem was obviously reasserting itself. The solution? Biodegrad-
able condoms! Here the question of flexibility and toughness is clearly relevant!
Polyhydroxybutyrate 543

POLYHYDROXYALKANOATE COPOLYMER SYNTHESIS

To set the scene on copolymers it is necessary to go back to 1974 when
Wallen and Rohwedder first reported the presence of two 3-hydroxyacids in
polymers isolated from activated sewage sludge. Subsequently David White and
his colleagues recovered the polymer from marine sediments and by gas-liquid
clhromatography demonstrated the presence of at least 11 short chain 3-
hydroxyacids in polymer hydrolysates; they also showed that batch-grown
Bacillus megaterium accumulated a polymer which consisted of 95% 3-
hydroxybutyrate, 3% 3-hydroxyheptanoate, 2% of an 8-carbon 3-hydroxy acid
and trace amounts of three other 3-hydroxy acids including 3-hydroxypentanoate
(Herron et al., 1978; Findlay and White, 1983). Further, in 1983 de Smet et al.
reported the presence of granules of poly-3-hydroxyoctanoate in Pseudornonas
oleovorans grown on 50% (v/v) octane. This class of compound which, as
previously noted, is given the generic name of polyhydroxyalkanoate, is attracting
world-wide interest on account of the commercial potential for novel biodegrad-
~tble plastics.
Our recent work in association with ICI has been an investigation of the
regulation of polymer and copolymer synthesis in Alc. eutrophus to secure a
thorough understanding of the basis of hitherto empirical methods for Biopol
production. First, we were able to show that the organism possesses two
constitutive 3-ketothiolase isoenzymes (designated A and B) which were separ-
ated on DEAE-cellulose or DEAE-Sepharose CL-6B and purified by hydroxy-
apatite chromatography and Sephacryl S-200 filtration (Haywood et al., 1988).
]Enzyme A is active only with acetoacetyl-CoA and 3-ketopentanoyl-CoA
whereas enzyme B is active with all 3-ketoacyl-CoA substrates tested (C4 to C10).
The characteristics of the two enzymes are recorded in Table 1. The condensation
:reaction catalysed by both enzymes is potently inhibited by CoA. The charac-
teristics of enzyme A resemble those of the 3-ketothiolase partially purified from
Alc. eutrophus H16 by Oeding and Schlegel in 1973 although at that time neither
the substrate specificity nor the presence of isoenzymes was recorded. In

Table 1. 3-Ketothiolases of Alcaligenes eutrophus

ENZYME A
Subunit Mr 44,000, Tetramer MT170,000
Optimum pH (thiolysis) 8.0; (condensation) 7.75-8.0
Isoelectric point 5.0
K~ (thiolysis): aeetoacetyl-CoA 44/tM
CoA 16/zM
K~. (condensation): acetyl-CoA 1.1 mM
Competitively inhibited by CoA
ENZYME B
Subunit Mr 46,000, Tetramer 168,000
Optimum pH (thiolysis) 8.0; (condensation) 7.75-8.0
Isoelectric point 6.4
Km (thiolysis): acetoacetyl-CoA 394 ~tM
CoA 93/~M
Km (condensation): acetyl-CoA 230/~M
Complex inhibition with CoA
544 Dawes

possessing two 3-ketothiolase isoenzymes, Alc. eutrophus resembles Zoogloea

ramigera (Tomita et aL, 1983). While the concept of the presence of biosynthetic
and degradative ketothiolases in prokaryotic and eukaryotic cells is well-
established, we were able to show that both enzyme A and enzyme B can
function in polymer synthesis and, as noted, both are subject to regulation by
CoA. Obviously enzyme A, together with the appropriate reductase and
polymerase, would permit the formation of polymers containing H V units.
The next enzyme in the biosynthetic sequence, acetoacetyl-CoA reductase,
was also found to be present as two constitutive isoenzymes, one N A D H -
dependent and the other requiring N A D P H . The enzymes have been separated,
partially purified and characterized. The N A D H - d e p e n d e n t enzyme was active
with all the L( + ) and D ( - )-3-hydroxyacyl-CoA substrates tested (Ca to Cto)
whereas the N A D P H - d e p e n d e n t enzyme was specific for D( - ) substrates of C4
to C6 chain length and with decreasing activity (Table 2).

Table 2. The substrate specificities of NADH- and NADPH-

acetoacetyi-CoA reductases of Alcaligenes eutrophus
Relative activity (%)
Substrate NADH-reductase NADPH-reductase
D( - )-3-hydroxyacyl-CoAs
C4 9.7 100
C5 31 48
C6 17 3.6
C7 14 0
C8 40 0
Clo 22 0
L( + )-3-hydroxyacyl-CoAs
c~ lOO o
C5 35 0
C6 50 0
C7 99 0
C8 146 0
C10 60 0
100% Activity= 13.5 U (mg protein)-~ for NADH-acetoacetyl-CoA
reductase
100% Activity= 13.0U (mg protein)-~ for NADPH-acetoacetyl-CoA
reductase

We have been able to test the activities of the individual isoenzymes of

3-ketothiolase and acetoacetyi-CoA reductase in overall PHB-synthesizing syst-
ems by measuring the incorporation of [1-14C]acetyl-CoA into polymer. We
found that while either of the 3-ketothiolases functioned well in acetyl-CoA
incorporation, only the N A D P H - d e p e n d e n t acetoacetyl-CoA reductase permitted
polymer synthesis to occur. We are now studying the P H B synthase system of
Alc. eutrophus, the specifcity of which is of considerable importance in relation
to copolymers which can be synthesized by the organism. To date only polymers
containing HB and H V have been reported and the specificities of the
3-ketothiolases and the N A D P H - d e p e n d e n t acetoacetyl-CoA reductase can
account for these observations.
Polyhydroxybutyrate 545

Recently Doi et al. (1986) in Japan have examined the route by which the
hydroxyvalerate unit is derived when Alc. eutrophus is incubated with [2-
13C]acetate or [1-13C]propionate in the absence of a nitrogen source. Polymer
isolated from the organism after 48h incubation was analysed by ~3C N.M.R. in
each case and the results accorded with PHB synthesized from [2-~3C]acetate
being labelled in the 2 and 4 positions of the HB units, while in the case of the
[1-~3C]propionate substrate the following scheme was proposed to account for the
labelling pattern:

CH3CH2COO- " > CH3CH2CH.CH2COO-

COPOLYMER

CH3COO- ~ CH3CH.CH2COO-
I
OH

The random copolymer was labelled only in the C-3 of the HV units. The HB
units are unlabelled because the label is lost as CO2 in the conversion of
propionate to acetate.
We were interested to know whether any turnover of polymer occurs in the
bacterial cells. Thus we grew Alc. eutrophus on [U-14C]glucose in a chemostat
and, after a steady state was attained, the labelled substrate was replaced by
unlabelled glucose and the loss of radioactivity from the whole cells and from
isolated PHB was compared. Both mirrored the washout rate and we concluded
that no significant turnover was occurring.

COPOLYMER P R O D U C T I O N BY PSEUDOMONAS EXTORQUENS

As part of our programme we have also examined copolymer production in

other organisms and here one rather interesting example must suffice--copolymer
production by Pseudomonas extorquens (Fig. 2). This organism grows on
methanol as the carbon source and ammonium as the nitrogen source. In this
experiment growth ceased due to exhaustion of the ammonium but then, in the
]presence of excess methanol, PHB synthesis commenced and the polymer
continued to accumulate in the cells until the methanol was exhausted, when it
started to be utilized. Now at the time of ammonium exhaustion valeric acid was
added to the culture; it was utilized from the outset and PHV synthesis started
and continued in parallel with PHB accumulation. However, in contrast to PHB,
net synthesis of which halted when the methanol was exhausted, PHV continued
to be synthesized from valerate throughout the remainder of the experiment.
Obviously very interesting biochemical and physiological relationships exist
between PHB and PHV synthesis in P. extorquens which should repay further
investigation.
5,~ Dawes

Pseudomonas extorquen$ : growth on methanol in batch culture

Valeril acid ~ Cell density

I % ,o o I -
/ Methanol . " ~ ' l - - i ' n - r-I- 1 - 1 ~

o -l-I /
6
o

"5

..--o //\ f l

0 8 "E 4 o,O Ammonia \

, , , , , 9 ,. , , , II ,
0 8 16 24 32 40 48 II 72
Time (h)

Fig. 2. Accumulation of PHB and PHV by Pseudomonas extorquens growing in batch culture with
methanol as carbon source and limiting ammonium as nitrogen source. At the time of exhaustion of
the ammonium (indicated by solid arrow), valeric acid was added to the culture.

To conclude, I find it fascinating that a bacterial cell component, which made

its first unobtrusive entry into the wings of science over sixty years ago as a
curiosity, now occupies a position centre stage in the star role of being a unique
biodegradable plastic and the focus of immense international attention. As
pollution of the environment with petrochemical-based plastics continues apace,
and attempts to dispose of them produce, in turn, further alternative and
unacceptable pollution, biodegradable plastics produced by micro-organisms offer
a shining hope for the amelioration of this global problem.

REFERENCES

Byrom, D. (1987) Trends in Biotechnology 5:246-250.

Cornibert, J. and Marchessault, R. H. (1972) J. Mol. Biol. 71:735-756.
Dawes, E. A. and Senior, P. J. (1973) Adv. Microbial Physiol. 10:135-266.
de Smet, M., Eggink, G., Witholt, B., Kingma, J. and Wynberg, H. (1983) J. Bacterial. 154:870-878.
Doi, Y., Kunioka, M., Nakamura, Y. and Saga, K. (1986) J. Chem. Sac., Chem. Commun.
1696-1697.
Ellar, D., Lundgren, D. G., Okamura, K. and Marchessault, R. H. (1968) J. MoL BioL 35:489-502.
Findlay, R. H. and White, D. C. (1983) Appl. Environ. Microbial 45"71-78.
Griebel, R. J. and Merrick, J. M. (1971) J. Bacterial. 1@8:782-789.
Haywood, G. W., Anderson, A. J., Chu, L. and Dawes, E. A. (1988) FEMS Microbial. Letts.
52:91-96.
Herron, J. S., King, J. D. and White, D. C. (1978) Appl. Environ. Microbial. 35:251-257.
Holmes, P. A., Wright, L. F. and Collins, S. H. (1981) European Patent Specification 0 0052 459.
Jackson, F. A. and Dawes, E. A. (1976) J. Gen. Microbial 97:303-312.
Polyhydroxybutyrate 547

Lemoigne, M. (1925)Ann. Inst. Pasteur, Paris 39:144.

Oeding, V. and Schlegel, H. G. (1973) Biochem. J. 134:239-248.
()kamura, K. and Marchessault, R. H. (1967) In: Conformation of Biopolymers (G. M. Rarnachan-
dran, ed.) Vol. 2, pp. 709-720. Academic Press: New York.
RJtchie, G. A. F. and Dawes, E. A. (1969) Biochem. J. 112:803-805.
RJtchie, G. A. F., Senior, P. J. and Dawes, E. A. (1971) Biochem. J. 121:309-316.
Senior, P. J. and Dawes, E. A. (1971) Biochem. J. 125:55-66.
Senior, P. J., Beech, G. A., Ritchie, G. A. F. and Dawes, E. A. (1972) Biochem. J. 128:1193-1201.
Senior, P. J. and Dawes, E. A. (1973) Biochem. J. 134:225-238.
Tomita, K., Saito, T. and Fukui, T. (1983) In: Biochemistry of Metabolic Processes (eds. D. L. F.
Lennon, F. W. Stratman and R. N. Zahlten) pp. 353-366. Elsevier Science Publishing, N.Y.
Wallen, L. L. and Rohwedder, W. K. (1974) Environ. Sci. Technol. 8:576-579.
Weitzman, P. J. D. and Jones, D. (1968) Nature 219:270-272.