A microbial factory for lactate-based polyesters

using a lactate-polymerizing enzyme

Seiichi Taguchi
a,1
, Miwa Yamada
a
, Kenichiro Matsumoto
a
, Kenji Tajima
a
, Yasuharu Satoh
a
, Masanobu Munekata
a
,
Katsuhiro Ohno
b
, Katsunori Kohda
b
, Takashi Shimamura
b
, Hiromi Kambe
c
, and Shusei Obata
c
a
Division of Biotechnology and Molecular Chemistry, Graduate School of Engineering, Hokkaido University, N13-W8, Kita-ku, Sapporo 060-8628, Japan;
b
Biotechnology Laboratory, Sustainable Materials Division, Materials Department, Toyota Central Reasearch and Development Laboratories, Nagakute,
Aichi 480-1192, Japan; and
c
Bio Research Laboratory, Future Project Division, Toyota Motor Corporation, Toyota-cho, Toyota, Aichi 471-8572, Japan
Edited by Alexander M. Klibanov, Massachusetts Institute of Technology, Cambridge, MA, and approved by the Editorial Board September 26, 2008
(received for review June 11, 2008)
Polylactate (PLA) is synthesized as a representative bio-based
polyester by the chemo-bio process on the basis of metal catalyst-
mediated chemical polymerization of lactate (LA) supplied by
microbial fermentation. Toestablishthe one-stepmicrobial process
for synthesis of LA-based polyesters, we explored whether poly-
hydroxyalkanoate (PHA) synthase would exhibit polymerizing ac-
tivity toward a LA-coenzyme A (CoA), based on the fact that PHA
monomeric constituents, especially 3-hydroxybutyrate (3HB), are
structurally analogous to LA. An engineered PHA synthase was
discovered as a candidate by a two-phase in vitro polymerization
system previously developed. An LA-CoA producing Escherichia
coli strain with a CoA transferase gene was constructed, and the
generation of LA-CoAwas demonstrated by capillary electrophore-
sis/MS analysis. Next, when the engineered PHA synthase gene
was introduced into the resultant recombinant strain, we con-
rmed the one-step biosynthesis of the LA-incorporated copoly-
ester, P(6 mol% LA-co-94 mol% 3HB), with a number-average
molecular weight of 1.9 10
5
, as revealed by gel permeation
chromatography, gas chromatography/MS, and NMR.
lactate coenzyme A polyhydroxyalkanoate synthase
substrate specicity CoA transferase enzyme engineering
T
he current polymer materials in common use are nearly all
derived from petrochemical sources, and the industry is a
significant contributor to greenhouse gas emissions, particularly
during the processes of production and incineration of plastics.
At present, the development of nonpetrochemical sources for
plastic has focused on renewable resources, such as sugars, plant
oils, and even CO
2
to replace diminishing supplies of fossil fuel.
Polylactate (PLA) is a representative bio-based polyester, which
is chemically synthesized by ring-opening polymerization of a
cyclic diester (lactide) of lactic acid (LA), produced by microbial
fermentation (the left portion in Fig. 1) (1, 2). By introducing
variations in molecular weight and crystallinity, PLA is turned
into highly valuable materials for biomedical, food, and general-
purpose applications, as described in numerous patents. Thus,
PLA combines inexpensive large-scale fermentation with chem-
ical processing capacity to produce a value-added polymer
product. However, as the chemo-process of PLA can be carried
out via harmful metal catalysts with high reaction velocities, it
often leaves chemical residues that are subject to health and
safety concerns. The paradigm shift from the chemo-process to
the bio-process for PLA production is thus preferable to over-
come this problem.
The complete biosynthesis of PLA is an enormous challenge
for both academic research and industry. For this purpose, a
LA-polymerizing enzyme, which can function as an alternative
to a metal catalyst, would be desired to establish the bio-process,
as shown in Fig. 1. The simplest strategy would be the discovery
of a PLA-producing micro-organism, but this approach has not
succeeded yet. Thus, we focused on the microbial biosynthetic
system for polyhydroxyalkanoates (PHAs), which are natural
polyesters, as energy storage materials. PHAs are also biode-
gradable and biocompatible bio-based polyester (3, 4). We
attempted to convert the present chemo-process of PLA to the
one-step bio-process based on the well-established metabolic
strategy for the synthesis of PHAs. PHA synthase functions as a
key enzyme for polymerization of various monomers, depending
on its substrate specificity. In this study, we presumed that if a
PHA synthase capable of LA-polymerization were to be ob-
tained, the recombinant microbe transformed with this PHA
synthase gene would synthesize LA-based polyester. Namely, it
would imply the recruitment from PHA synthase to a LA-
polymerizing enzyme. This hypothesis, based on the substrate
specificity of the enzyme, arises from the fact that monomeric
constituents of PHA share the common chemical structure,
hydroxy acid, with 2-hydroxypropionate (the same as LA).
An attempt to construct a microbial production system for an
LA-based polyester has been triggered by the discovery of an
engineered PHA synthase with acquired LA-polymerizing ac-
tivity, as revealed by an in vitro polymerization system (5). Based
on this finding, we have established a recombinant Escherichia
Author contributions: S.T. and S.O. designed research; M.Y., K.T., Y.S., K.O., K.K., T.S., and
H.K. performed research; S.T., K.M., and S.O. contributed newreagents/analytic tools; S.T.,
M.Y., K.M., K.T., M.M., K.O., K.K., andT.S. analyzeddata; andS.T. andK.M. wrotethepaper.
The authors declare no conict of interest.
This article is a PNAS Direct Submission.
1
To whom correspondence should be addressed. E-mail: staguchi@eng.hokudai.ac.jp.
This article contains supporting information online at www.pnas.org/cgi/content/full/
0805653105/DCSupplemental.
2008 by The National Academy of Sciences of the USA
Biomass
Sugar
Lactate
Lactyl-CoA
LA-based
polyester
Bio-process
COOH HO
Lactide
PLA
(LA-based polyester)
Lactate
Chemo-process
O
O
O O PCT PCT
LA-polymerizing
enzyme
LA-polymerizing
enzyme
Metal
catalyst
Metal
catalyst
Microbial
factory
Microbial
factory
substitution
Biomass
Sugar
Lactate
Lactyl-CoA
LA-based
polyester
Bio-process
COOH HO COOH HO
Lactide
PLA
(LA-based polyester)
Lactate
Chemo-process
O
O
O O
O
O
O O PCT PCT PCT PCT
LA-polymerizing
enzyme
LA-polymerizing
enzyme
LA-polymerizing
enzyme
LA-polymerizing
enzyme
Metal
catalyst
Metal
catalyst
Metal
catalyst
Metal
catalyst
Microbial
factory
Microbial
factory
Microbial
factory
Microbial
factory
substitution
Fig. 1. Microbial factory for LA-basedpolyester. PLAhas beensynthesizedby
thechemo-bioprocess. Copolymerizationof LAwithother monomer units has
also been performed by basically the same chemo-synthetic procedures. In
bio-process withmicrobes, LAmonomer substrate is recognizedas a CoAform
(LA-CoA) by LA-polymerizing enzyme that is recruited from microbial PHA
synthase. A candidate PHA synthase with acquired LA-polymerizing activity
was selected based on the water-organic solvent two-phase in vitro polymer-
ization system (5).
www.pnas.orgcgidoi10.1073pnas.0805653105 PNAS November 11, 2008 vol. 105 no. 45 1732317327
A
P
P
L
I
E
D
B
I
O
L
O
G
I
C
A
L
S
C
I
E
N
C
E
S
coli that allows the synthesis of LA-based polyester by introduc-
ing the gene encoding PHA synthase discovered here. The
bio-process serves a versatile platform as a microbial factory
for the one-step production of the LA-based polyester from
renewable biomass (see Fig. 1).
Results
Selection of a PHA Synthase with LA Polymerization Activity. PHA
synthase is thought to synthesize polyester via continuous trans-
esterification of coenzyme A(CoA) esters of typically 3-hydroxy-
alkanoates (3HAs), including 3-hydroxybutyrate (3HB) (6).
Thus, the most promising precursor for PLA biosynthesis is the
CoAester of lactate, lactyl-CoA(LA-CoA). To explore the PHA
synthase capable of polymerizing the LA moiety in LA-CoA, we
used a modified water-organic solvent two-phase in vitro system
developed previously (5). In examining LA-polymerizing activ-
ity, an in vitro system is advantageous, because the targeted
monomer substrates can be controllably generated without
the aid of monomer-supplying metabolic pathways for LA-CoA
and 3HB-CoA. The polymerizing activity of PHA synthase
was primarily judged by the generation of the polymer-like
precipitation.
The classification of the PHA synthase family was the basis for
the selection of PHAsynthase as a candidate of LA-polymerizing
enzyme. Four major classes of PHA synthases can be distin-
guished with respect to the primary structures deduced from
these sequences, the substrate specificities of the enzymes, and
the subunit composition (7). In this study, we examined four
representative PHA synthases belonging to the individual
classes; synthases derived from Ralstonia eutropha (class I),
Pseudomonas sp. 613 (class II), Synechosystis sp. PCC6803 (class
III), and Bacillus sp. INT005 (class IV), and three engineered
PHA synthases (PhaC1) from Pseudomonas sp. 613. The en-
gineered PHA synthases were two single mutants [Ser325Thr
(ST) and Gln481Lys (QK)] and one double mutant carrying
these mutations (STQK). The two positions (325 and 481) were
closely related to the activity or substrate specificity of the
enzyme (8, 9). These mutants were selected from the huge
mutant library of PhaC1 that has been created through evolu-
tionary engineering directed to the reinforcement of 3HB in-
corporation ability into the PHA polymer chain (10, 11).
The PHA synthases selected for in vitro assay intrinsically
polymerize 3HB-CoA into P(3HB). When LA-CoA was gener-
ated, no clear polymer-like precipitation occurred for any of the
enzymes, suggesting that the PHAsynthases hardly catalyzed any
polymerization of LA-CoA. Thus, we next set out to generate
3HB-CoA together with LA-CoA. This experiment was set up
for facilitating polymerization of LA-CoA by adding the favor-
able substrate 3HB-CoA to consequently synthesize the copol-
ymer, based on the case of wild-type PhaC1 from Pseudomonas
sp. 613 that the coexistence of favorable substrate(s), 3HA-
CoA, leads to the enhanced polymerization of less favorable
substrate, 3HB-CoA (12). When LA-CoA and 3HB-CoA were
supplied, the mutant PhaC1STQKclearly exhibited polymer-like
precipitation. Gas chromatography/MS (GC/MS) analysis re-
vealed that the precipitant consisted of 36 mol% of the LA unit
[supporting information (SI) Fig. S1]. The result strongly sug-
gested that PhaC1STQK polymerized LA-CoA in the presence
of 3HB-CoA. By contrast, the other enzymes did not synthesize
any polymer. Therefore, we selected PhaC1STQKas a promising
candidate of LA-polymerizing enzyme for further investigation.
Construction of a Recombinant E. coli Strain Generating LA-CoA. The
finding of an engineered PHA synthase (PhaC1STQK) with the
capacity to polymerize LA-CoA as a substrate prompted us to
create a microbial production system of LA-based polyesters,
such as PLA and P(LA-co-3HB). Another essential component
for microbial production of LA-based polyesters is an intracel-
lular supplier of LA-CoA. We used propionyl-CoA transferase
(PCT) as a LA-CoA supplier because PCT catalyzes the gener-
ation of LA-CoA from lactate (acceptor) and acetyl-CoA (do-
nor) (13). The ability of PCT to synthesize LA-CoA was exam-
ined using recombinant E. coli harboring the pct gene. By
capillary electrophoresis/MS (CE/MS) analysis using the prepared
recombinant cells, we found a peak, whose retention time and
molecular masses, 838 and 419 (corresponds to bivalent ion), were
identical to those of the LA-CoA standard (Fig. 2A). This is direct
evidence for PCT-catalyzed generation of LA-CoA in vivo.
Microbial Production of LA-Based Polyester Using PhaC1STQK. Based
on the metabolic pathways for generation of the monomer
substrates, LA-CoA and 3HB-CoA (Fig. 2B), we constructed
two plasmids, pTV118NpctC1(STQK) (for PLA) and
pTV118NpctC1AB(STQK) (for copolymer), respectively. The
phaAB genes encode -ketothiolase (PhaA) and acetoacetyl-
m/z = 838
time / min
pTV118Npct
pTV118N
Lactyl-CoA
m/z = 419
Lactyl-CoA
pTV118Npct
A
b
u
n
d
a
n
c
e

pTV118N
time / min
A
pTV118Npct
C1(STQK)
pct
Plac
TRe
phaA phaB
pTV118Npct
C1AB(STQK)
phaC1
STQK
pct
phaC1
STQK
PRe
Plac PRe
C
TRe
Glucose
Acetyl-CoA
Acetoacetyl-CoA
(R)-3-Hydroxybutyryl-CoA
P(LA-co-3HB)
Pyruvate
Lactate
Lactyl-CoA
LDH
PCT
(Glycolysis)
B
PhaA
PhaB
PhaC
Fig. 2. Construction of microbial production system for LA-based polyester.
(A) CE/MS analysis of LA-CoA generated in recombinant Escherichia coli
harboring the pct gene. The ions of m/z 838 and 419 correspond to
monovalent and bivalent ions of LA-CoA, respectively. pTV118N, control
vector; pTV118Npct, expressionvector for thepct gene. (B) LA-basedpolyester
synthetic pathway in recombinant E. coli. The letters in boxes indicate the
enzyme. LDH, lactate dehydrogenase; PCT, propionyl-CoA transferase; PhaA,
-ketothiolase; PhaB, NADPH-dependent acetoacetyl-CoA reductase; PhaC,
PHA synthase. (C) Construction of the plasmids used in this study. P
Re
and T
Re
denote respectively the promoter and terminator regions of phbCAB operon
from Ralstonia eutropha. P
lac
denotes the lac promoter. The phaC1(STQK)
gene encodes the Ser325Thr/Gln481Lys mutant of PHAsynthase fromPseudo-
monas sp. 613. The phaA and phaB genes encode -ketothiolase (PhaA) and
NADPH-dependent acetoacetyl-CoA reductase (PhaB) derived from R. eutro-
pha, respectively. The pct gene encodes a propionyl-CoA transferase from
Megasphaera elsdenii.
17324 www.pnas.orgcgidoi10.1073pnas.0805653105 Taguchi et al.
CoA reductase (PhaB), both of which are the most typical
supplying pairs for 3HB-CoA generation from acetyl-CoA (see
Fig. 2B). We next extracted polymer from recombinant E. coli
strains harboring each plasmid and analyzed the extracts using
GC/MS. The strain harboring pTV118NpctC1(STQK) did not
produce detectable polymer under the conditions used. How-
ever, the recombinant harboring pTV118NpctC1AB(STQK), in
which the phaAB genes were coexpressed, produced 19 wt% of
polymer. GC/MS analysis revealed the presence of 6 mol%of the
LA unit in the polymer and the 3HB unit as a main component
(Fig. S2), suggesting that the combination of PhaC1STQK and
PCT together with PhaAB does produce the P(LA-co-3HB)
copolymer.
Characterization of LA-Based Polyester Produced in the Recombinant
Strain. To further confirm that the LA unit was incorporated into
the polymer chain, the extracted polymer was applied to pre-
parative gel permeation chromatography (GPC), and the high-
molecular-weight polymer fraction was collected. The molecular
weights of the fractionated sample were determined to be M
n

1.9 10
5
and M
w
4.8 10
5
, respectively, using analytical GPC
(Table 1). The fractionated sample was subjected to GC/MS and
NMR analyses. The mass spectrum of the peak at 6.5 min was
identical to that of the ethyl-LA standard, indicating that the
fraction contains the LA unit (Fig. 3A). Furthermore,
13
C-NMR
analysis revealed that the polymer had typical peaks of 3HB unit
in P(3HB) and LA unit in PLA (Fig. 3B). However, in
1
H-NMR
analysis, small peaks appeared at 1.5 and 5.0 ppm, which slightly
differed from those of PLA (1.6 and 5.1 ppm) (Fig. 3C). The
high-field shifts of the peaks are presumably because of the
presence of the LA unit within the backbone of P(3HB). Indeed,
2d-NMR revealed cross signals between the carbons of LA and
the LA-like peaks of protons (Fig. 3D). Thus, the proton peaks
Table 1. Properties of bio-based polyesters
Sample
Monomer
composition
(mol%)* Molecular weight Thermal properties

LA 3HB M
n
(10
5
) M
w
(10
5
) M
w
/M
n
T
g
(C) T
m
(C)
P(LA-co-3HB) 6 94 1.9 4.8 2.4 2 168
P(3HB) 0 100 2.9 7.2 2.4 1 176
PLA 100 0 nd nd nd 63 173
*Monomer composition was determined by
1
H-NMR. LA, lactate; 3HB, 3-hydroxybutyrate. M
n
, number-average molecular weight; M
w
,
weight-average molecular weight; M
w
/M
n
, polydispersity index; nd, not determined.

Thermal properties are measured by differential scanning calorimetry. T

g
, glass transition temperature; T
m
, melting temperature.
Fig. 3. Characterization of a P(LA-co-3HB) copolymer produced in the recombinant E. coli harboring the plasmid pTV118NpctC1AB(STQK). (A) GC/MS analysis
of a P(LA-co-3HB) copolymer isolated from the recombinant E. coli harboring the plasmid pTV118NpctC1AB(STQK). Ethyl-LA, ethyl lactate; Ethyl-3HB, ethyl
3-hydroxybutyrate. (B)
13
C-NMR spectrum of a P(LA-co-3HB) copolymer isolated from the recombinant E. coli harboring the plasmid pTV118NpctC1AB(STQK).
(C)
1
H-NMR spectrum of a P(LA-co-3HB) copolymer isolated from the recombinant E. coli harboring the plasmid pTV118NpctC1AB(STQK). (D) 2d-NMR spectrum
of a P(LA-co-3HB) copolymer isolated from the recombinant E. coli harboring the plasmid pTV118NpctC1AB (STQK).
Taguchi et al. PNAS November 11, 2008 vol. 105 no. 45 17325
A
P
P
L
I
E
D
B
I
O
L
O
G
I
C
A
L
S
C
I
E
N
C
E
S
at 1.5 and 5.0 ppm were assigned to LA, but were differentiated
from PLA. LA molar fraction in the copolymer was 6 mol%
based on the
1
H-NMR.
Based on the number-average molecular weight (1.9 10
5
)
and LA molar fraction (6 mol%), the number of LA units was
estimated to be much higher than that of the case in which the
LA unit is incorporated as a terminator into the end of the
polymer chains. Thus, the LA unit should be incorporated into
the internal region of the polymer chain. In addition, the peak
positions of
1
H-NMR corresponding to the LA unit were not
identical to those of PLA. This suggests that most of the LA units
are distributed separately from each other and connected with
the 3HB unit. Furthermore, the P(LA-co-3HB) has lower melt-
ing temperature (T
m
) than P(3HB) that is probably because of
the lowered crystallinity caused by random copolymerization of
LA and 3HB units (see Table 1). In fact, the decrease in T
m
agrees with the result that chemically synthesized P(LA-co-3HB)
random copolymers have lower T
m
compared with P(3HB) (14).
Therefore, the P(LA-co-3HB) would be synthesized in a random
manner rather than a blend or block copolymerization of PLA
and P(3HB). From these results, we concluded that the polymer
produced by recombinant strain was P(6 mol% LA-co-94 mol%
3HB) with a possible random sequence.
Discussion
In this study, the substitution of metal catalysts with an LA-
polymerizing enzyme was a key methodology to convert the
chemo-process of LA-based polyesters into a microbial bio-
process (see Fig. 1). The desired LA-polymerizing enzyme has
been recruited from PHA synthase, which is an essentially
participating in biosynthetic pathway in the microbial polymer
production system. The general structural formula of PHAs is
presented as [-O-CHR-(CH
2
)
X
-CO-]
n
. R basically presents hy-
drogen or alkyl chains of up to C
13
in length, and x can range from
1 to 3 or more. These variations in the length and composition
of the side and main chains are the basis for the diversity of the
PHA family and the material properties. When R is a methyl
group and x 1, the polyester is P(3HB). When R is a methyl
group and x 0 (equals to LA), the polymeric product is PLA.
In the naturally occurring micro-organisms, LA has not been
identified to date as a monomeric constituent in the PHAs. This
fact suggests the absence or very scarce distribution of PLA
synthases.
To date, the substrate specificity of the PHA synthases has
been mainly discussed based on the alkyl chain length of the side
chain built at the 3 position of hydroxy acids (7, 10, 11). On the
other hand, the LA-reactivity of PHA synthase should be related
to the substrate specificity toward the main chain length or
conformation of hydroxyl groups. Wild-type PhaC1 from
Pseudomonas sp. 613 has especially broad side-chain substrate
specificity and its mutant (STQK) has enhanced activity toward
3HB-CoA. Surprisingly, PhaC1STQK incorporated the LA unit
into the polymer chain. Other PHA synthases, which have higher
substrate specificity toward 3HB-CoA than PhaC1STQK, did
not polymerize LA-CoA, although 3HB is most structurally
analogous to LA. This fact suggests that PHA synthase, which
has considerable tolerance in recognizing structurally diverse
substrates, like PhaC1, would be rather preferable to expand
substrate specificity toward LA-CoA than the other types of
PHAsynthases with strict substrate specificity, such as class I and
class III enzymes (15). In other words, the STQK double
mutation was found to be closely related to not only the
side-chain substrate specificity, but also the main-chain substrate
specificity of hydroxyacyl-CoAs, including LA-CoA.
Based on the in vitro polymerization assay, PhaC1STQK has
the intrinsic potential of incorporating the LA unit at a level of
at least up to 36 mol% into the polymer. However, 3HB-CoA
dependent LA-polymerization limits the production of LA-
enriched copolyesters or PLA homopolymer. Similar phenom-
enon was observed for biosynthesis of P(3HB-co-3HA) using
wild-type PhaC1, in which 3HA-CoAs accelerate the polymer-
ization of 3HB-CoA (12). It may be interpreted that 3HAs act
as a priming unit for the polymerization of 3HB-CoA. Through
evolutionary engineering, we succeeded in creating beneficial
mutants of PhaC1 that can synthesize 3HB-enriched copolymers
(10, 11). In the same way, LA-enriched copolymers would be
synthesized by engineered PhaC1 enzymes.
A wide variety of monomers yields PHAs, with diverse
material properties that depend on polymer composition. For
example, P(95 mol% 3HB-co-5 mol% 3HAs) random copoly-
ester has properties similar to low-density polyethylene (3).
Furthermore, wide-ranging changes in the thermal and physical
properties of 3HB-based PHA copolymers, P(3HB-co-3HA),
have been achieved by altering substrate specificity of wild-type
PhaC1 through enzyme evolution (10, 11). In this context, the
introduction of the LA unit into the monomer constitution of
PHAs is very attractive in that it further expands the diversity in
combination of monomers available for polymer production.
This concept is applicable for creating LA-based polyesters using
the microbial factory established here.
In conclusion, the microbial factory will give a versatile
platform for the one-step synthesis of the LA-based copolyesters
with various monomeric compositions, as well as the PLA from
renewable carbon sources. In the future, this prototype system
will be advanced by improvement of the LA-CoA supplying
pathway. The flux of LA-CoA in vivo should be a limiting factor
for the production of LA-enriched copolymer, because in vitro
the system produced P(LA-co-3HB) with higher LA fraction.
The intracellular generation level of LA-CoA should be affected
by two factors: generation level of LA itself and the LA-CoA
converting efficiency of PCT. Because LA is known to efficiently
produce under the anaerobic condition, fine tuning of such a
condition would optimize LA production. In addition, the total
activity of PCT could be increased by enhancing expression level
and enzyme engineering of PCT.
Materials and Methods
Selection of PHASynthases by Using a Water-Organic Solvent Two-Phase in Vitro
PolymerizationSystem. Thiophenyl (R)-3HB((R)-3HB-TP) andthiophenyl (R)-LA
((R)-LA-TP) were synthesized as described in ref. 5. (R)-3HB-TP and (R)-LA-TP,
precursors of CoA derivatives, were dissolved in 5 ml of hexane (organic
solvent phase) togive 5 mM, respectively. The water phase contained100-mM
sodium phosphate (pH 7.5) and 1-mM CoA. We combined the two solutions
prepared above to make the two-phase system. The polymerization was
started by adding PHA synthase to the water phase, and the mixture was
incubatedat 30Cfor 72 h. SevenPHAsynthases were examinedas mentioned
in Results. All synthases were puried using a His-tag system (16). Candidate
of PHA synthase with LA polymerization activity was chosen based on the
amount of the generated polymer-like precipitant.
Plasmids. The pct gene from Megasphaera elsdenii (17) was amplied from
genomic DNA by PCR using following primers: 5-ATGAGAAAAGTAGAAAT-
CATTAC-3 and 5-TTATTTTTTCAGTCCCATGGGACCGTCCTG-3. A 1.6-kb frag-
ment of pct gene was inserted into pTV118N (Takara, Japan) using TOPO PCR
cloningkit (Invitrogen). Subsequently, the Ser325Thr/Gln481Lys mutatedPHA
synthase gene from Pseudomonas sp. 613 (18) and the phaAB genes from R.
eutropha withits original promoter (19) were insertedintoa HindIII site of the
plasmid to yield pTV118NpctC1AB(STQK). The phaAB genes were eliminated
by amplication using following primers: 5-GCCTGGTTCAACCAGTCGG-3
and5-GGAACCTGCAGAGATCCAAC-3. Thefragment was self-ligatedtoyield
pTV118NpctC1(STQK).
CE/MS Analysis. The pct gene was inserted into pTV118N using EcoRI and PstI
to yield pTV118Npct. E. coli JM109 (20) harboring the resultant plasmid was
grown on LB medium at 37C for 3 h. Then, the nal 10-mM isopropyl
-D-thiogalactopyranoside was added to induce the expression of pct and
cells were cultivated at 30C for 6 h. The cells were collected by centrifugation
and transferred to M9 medium containing 1.5% glucose, 10-mM MgSO
4
, and
10-mM calcium pantothenate, and cultivated at 37C for 18 h. The harvested
17326 www.pnas.orgcgidoi10.1073pnas.0805653105 Taguchi et al.
cells were washed with water twice and extracted with methanol. Next 1.6 ml
of chloroformwas added to the same amount of methanol extract, and 640 l
of water was addedandmixed. The solutionwas centrifugedandupper phase
was applied to ultraltration (5,000 Da) to remove contaminants with higher
molecular weight. The path-through fraction was lyophilized and applied to
CE/MS as describedinref. 21. Theions of m/z 838and419, whichcorresponds
to monovalent of LA-CoA and its bivalent ion, respectively, were monitored.
Polymer Extraction from Recombinant Cells. Recombinant E. coli JM109 har-
boring pTV118NpctC1(STQK) and pTV118NpctC1AB(STQK) were grown on
100-ml LBmediumcontaining2%glucose, 10-mMcalciumpantothenate, and
100-g/L ampicillin at 30C for 72 h. The harvested cells were lyophilized and
polymer was extracted with 5-ml chloroform at 100C for 3 h. Cell debris was
removed by passing through a PTFE lter. Then, 100-ml methanol was added
to precipitate polymer. Solubility of the polymer was very similar to that of
P(3HB) or PLA: that is, soluble in chloroform and dichloromethane but insol-
uble in methanol and hexane. The precipitant was dried in vacuo to measure
the weight of polymer, and polymer content was calculated based on cell dry
weight. The polymer was subjected to further analyses.
GC/MS, GPC, and NMR Analyses. Of the polymer-like precipitant, 12.5 mg
recoveredfromE. coli harboringpTV118NpctC1AB(STQK) was dissolvedin250
l of chloroform(50 g/ml), and the solution was applied to GC/MS analysis as
described in ref. 22. The recombinant E. coli harboring pTV118NpctC1(STQK)
did not obviously produce polymer, thus all chloroform soluble fraction was
dried up, dissolved in 250-l chloroform, and applied to GC/MS analysis. The
precipitant was also applied to preparative GPC LC2901 (JAI, Japan) equipped
with JAIGEl-1H column (JAI, Japan). The high molecular weight fraction was
collected and dried up to yield a white lm. The fractionated polymer was
further applied to analytical GPC (Shimadzu) equipped with TSKgel Super
HZM-H (Tosoh). The molecular weight was estimated using polystyrene stan-
dard (Waters). Of the fractionated lm, 12.5 mg was dissolved in 250 l of
chloroform (50 g/ml), and the solution was applied to GC/MS analysis again.
The fractionated lm was dissolved in CDCl
3
and analyzed by
1
H-NMR and
13
C-NMR spectroscopy.
Differential Scanning Calorimetry Analysis. Differential scanning calorimetry
data were recorded in the temperatures ranging from 90 to 210C on a
Perkin-Elmer Pyris 1 instrument equipped with a cooling accessory under a
nitrogen ow rate of 20 ml/min. The solvent-cast lms (10 mg) were encap-
sulated in aluminum pans and heated from 50 to 210C at 20C/min (rst
heating scan). The melt samples then followed by rapid quenching at 90C,
and maintained at 90C for 5 min. They were heated from90 to 210C at
20C/min (second heating scan). The glass-transition temperature (T
g
) was
taken as midpoint of the heat capacity change. The melting temperature (T
m
)
was determined from positions of the endothermic peaks. P(3HB) was pro-
duced using recombinant E. coli and PLAwas purchased fromToyota (Japan).
ACKNOWLEDGMENTS. We thank Dr. Toshifumi Satoh, Hokkaido University,
for his useful comments on NMR analysis. This study was partly supported by
a Grant-in-Aid for Scientic Research of Japan fromthe Ministry of Education,
Culture, Sports, Science, and Technology-Japan Grant 70216828 (to S.T.), the
Nagase Science Technology Foundation (to S.T.), and the Global Center of
Excellence Program (Project No. B01: Catalysis as the Basis for Innovation in
Materials Science) from the Ministry of Education, Culture, Sports, Science,
and Technology-Japan.
1. Tsuji H (2002) Biopolymers (Wiley-VCH, Weinheim).
2. Auras R, Harte B, Selke S (2004) An overview of polylactides as packaging materials.
Macromol Biosci 4:835864.
3. Sudesh K, Abe H, Doi Y (2000) Synthesis, structure and properties of polyhydroxyal-
kanoates: Biological polyesters. Prog Polym Sci 25:15031555.
4. Steinbu chel A, Hein S (2001) Biochemical and molecular basis of microbial synthesis of
polyhydroxyalkanoates in microorganisms. Adv Biochem Eng Biotechnol 71:81123.
5. Tajima K, et al. (2004) Chemoenzymatic synthesis of poly(3-hydroxybutyrate) in a
water-organic solvent two-phase system. Macromolecules 37:45444546.
6. Stubbe J, Tian J (2003) Polyhydroxyalkanoate (PHA) hemeostasis: The role of PHA
synthase. Nat Prod Rep 20:445457.
7. Rehm BH (2003) Polyester synthases: Natural catalysts for plastics. Biochem J 376(Pt
1):1533.
8. Takase K, Taguchi S, Doi Y (2003) Enhanced synthesis of poly(3-hydroxybutyrate) in
recombinant Escherichia coli by means of error-prone PCR mutagenesis, saturation
mutagenesis, and in vitro recombination of the type II polyhydroxyalkanoate synthase
gene. J Biochem 133(1):139145.
9. Takase K, Matsumoto K, Taguchi S, Doi Y (2004) Alteration of substrate chain-length
specicity of type II synthase for polyhydroxyalkanoate biosynthesis by in vitro evo-
lution: in vivo and in vitro enzyme assays. Biomacromolecules 5:480485.
10. Taguchi S, Doi Y(2004) Evolutionof polyhydroxyalkanoate(PHA) productionsystemby
enzyme evolution: Successful case studies of directed evolution. Macromol Biosci
4(3):146156.
11. Nomura CT, Taguchi S (2007) PHA synthase engineering toward superbiocatalysts for
custom-made biopolymers. Appl Microbiol biotechnol 73:969979.
12. Matsumoto K, et al. (2001) Biosynthesis of poly(3-hydroxybutyrate-co-3-hydroxyal-
kanoates) copolymer from sugars by recombinant Ralstonia eutropha harboring the
phaC1Ps and the phaGPs genes of Pseudomonas sp. 613. Biomacromolecules 2:934
939.
13. Selmer T, Willanzheimer A, Hetzel M (2002) Propionate CoA-transferase from Clos-
tridiumpropionicum. Cloningof geneandidenticationof glutamate324at theactive
site. Eur J Biochem 269:372380.
14. Abe H, Doi Y, Hori Y, Hagiwara T (1997) Physical properties and enzymatic degrad-
ability of copolymers of (R)-3-hydroxybutyric acid and (S,S)-lactide. Polymer 39(1):59
67.
15. Yuan W, et al. (2001) Class I and III polyhydroxyalkanoate synthases from Ralstonia
eutropha and Allochromatium vinosum: Characterization and substrate specicity
studies. Arch Biochem Biophys 394(1):8798.
16. Satoh Y, Tajima K, Tannai H, Munekata M (2003) Enzyme-catalyzed poly(3-
hydroxybutyrate) synthesis fromacetate with CoArecycling and NADPHregeneration
in vitro. J Biosci Bioeng 95:335341.
17. Elsden SR, Gilchrist FM, Lewis D, Volcani BE (1956) Properties of a fatty acid forming
organism isolated from the rumen of sheep. J Bacteriol 72:681689.
18. Matsusaki H, et al. (1998) Cloning and molecular analysis of the Poly(3-hydroxybu-
tyrate) and Poly(3-hydroxybutyrate-co-3-hydroxyalkanoate) biosynthesis genes in
Pseudomonas sp. strain 613. J Bacteriol 180:64596467.
19. Peoples OP, Sinskey AJ (1989) Poly-beta-hydroxybutyrate biosynthesis in Alcaligenes
eutrophus H16. Characterization of the genes encoding beta-ketothiolase and aceto-
acetyl-CoA reductase. J Biol Chem 264:1529315297.
20. Yanisch-Perron C, Vieira J, Messing J (1985) Improved M13 phage cloning vectors and
host strains: Nucleotide sequences of the M13mp18 and pUC19 vectors. Gene
33(1):103119.
21. Soga T, et al. (2002) Pressure-assisted capillary electrophoresis electrospray ionization
mass spectrometry for analysis of multivalent anions. Anal Chem 74:62246229.
22. Arai Y, et al. (2002) Synthesis of a novel class of polyhydroxyalkanoates in Arabidopsis
peroxisomes, and their use in monitoring short-chain-length intermediates of beta-
oxidation. Plant Cell Physiol 43:555562.
Taguchi et al. PNAS November 11, 2008 vol. 105 no. 45 17327
A
P
P
L
I
E
D
B
I
O
L
O
G
I
C
A
L
S
C
I
E
N
C
E
S