Cs

Ya

b

a

ARRAA

KPCSSS

1

tbnfilloaieaeioTa

TT

0h

Enzyme and Microbial Technology 56 (2014) 60–66

Contents lists available at ScienceDirect

Enzyme and Microbial Technology

jo ur nal ho me pa ge: www.elsev ier .com/ locate /emt

ritical residues of class II PHA synthase for expanding the substratepecificity and enhancing the biosynthesis of polyhydroxyalkanoate

i-Jr Chena,b,1, Pei-Chien Tsaia,1, Chun-Hua Hsua, Chia-Yin Leea,∗

Department of Agricultural Chemistry, National Taiwan University, Taipei 10617, TaiwanDepartment of Nursing, Chang Gung University of Science and Technology, Tao-Yuan 33333, Taiwan

r t i c l e i n f o

rticle history:eceived 2 September 2013eceived in revised form 2 January 2014ccepted 7 January 2014vailable online 16 January 2014

eywords:olyhydroxyalkanoates

a b s t r a c t

This study describes protein model of type II Pseudomonas putida GPo1 synthase (PhaC1Pp) and usingsingle or multiple points mutagenesis to identify the beneficial amino acid residues that change the PHAaccumulation and the substrate chain-length specificity of type II PHA synthase. The P. putida GPp104PHA− was used as a host for evaluating the substrate specificity and PHA yield of the mutated PhaC1Pp.The evolved PhaC1Pp were coexpressed with �-ketothiolase (phbARe) and the acetoacetyl-CoA reduc-tase (phbBRe) to supply sufficient short-chain length (R)-3-hydroxyacyl-CoA as a substrate. A single pointmutation at L484V remarkably enhanced the monomer ratio of (R)-3-hydroxybutyrate in a PHA accumu-

lass II PHA synthaseaturation mutagenesisite-directed mutagenesisubstrate specificity

lation experiment. Saturation mutagenesis experiment at 484 concluded that Val is the most favorableamino acid in PhaC1Pp for incorporating (R)-3-hydroxybutyrate unit synthesis. In addition, a single muta-tion at Q481M, S482G and A547V obviously increased PHA yields. Q481M and S482G enhanced the(R)-3-hydroxyhexanoate monomer composition in the PHA accumulation by P. putida GPp104 PHA−. Thisis the first data that spotlighted the important effect of Leu484 on substrate specificity of PHA synthase

ccum

and Ala547 on the PHA a

. Introduction

Polyhydroxyalkanoates (PHAs) are a series of natural polyestershat are accumulated in a wide variety of microorganisms as car-on and energy stores when carbon is in excess and other essentialutrients for cell growth are limited [1]. PHAs can be classi-ed by the length of their carbon chain monomers: short-chain

ength (SCL) PHA monomers containing C3–5, and medium-chainength PHA (MCL) monomers containing C6–14 [2]. Copolymersf SCL and MCL PHA and homopolymers of PHA are classifiedccording to the type of the constituent monomer incorporatednto the polymer chain [3]. The properties of PHAs are influ-nced by their monomer composition. For example, SCL PHAsre more thermoplastic and stiff, while MCL PHAs are morelastic and sticky [4]. PHAs are biodegradable and biocompat-ble, giving them the potential to be utilized in a wide range

f applications, such as medicine, agriculture, and industry [5].he type and relative abundance of PHA monomers that serves substrates for PHA synthase strongly determine the polymer

∗ Corresponding author at: Department of Agricultural Chemistry, Nationalaiwan University, 1, Sec. 4, Roosevelt Road, Taipei, 10617 Taiwan.el.: +886 2 33664812, fax: +886 2 23660581.

E-mail address: clee@ntu.edu.tw (C.-Y. Lee).1 Authors are equal contribution.

141-0229/$ – see front matter © 2014 Elsevier Inc. All rights reserved.ttp://dx.doi.org/10.1016/j.enzmictec.2014.01.005

ulation.© 2014 Elsevier Inc. All rights reserved.

composition of PHAs produced by microorganisms. The carbonsource via metabolic pathways of microorganisms provides thePHA monomers [6]. �-ketothiolase (encoded by phbA) combinesthe SCL monomers such as (R)-3-hydroxybutyrate synthesizedfrom two acetyl-CoAs to form acetoacetyl-CoA, then NADPH-dependent acetoacetyl-CoA reductase (encoded by phbB) reducesacetoacetyl-CoA to (R)-3-hydroxybutyryl-CoA [2]. MCL monomersare synthesized from �-oxidation of fatty acids and fatty acid denovo synthesis from carbohydrates [7]. PHA synthase, a key enzymein the PHA polymerization process, uses the (R)-3 hydroxyacyl-CoA as a substrate for synthesis of PHA [8,9]. PHA synthases aredivided into four classes based on their substrate specificity andconstituent subunits. Classes I, III, and IV use SCL monomers,while class II uses MCL monomers [10]; classes I and II com-prise only one subunit PhaC, whereas classes III and IV comprisetwo different subunit types: PhaC and PhaE, PhaC and PhaR,respectively.

Thus far, the crystal structure of PHA synthase is still unknown;therefore, the functional data obtained through molecular model-ing and manipulation of enzyme amino acid residues can help toimprove understanding of this enzyme [11]. The predicted three-dimensional model of Class II PHA synthase from Pseudomonas

aeruginosa PhaCPa and Pseudomonas sp USM4-55 (PhaC1P. sp USM)were reported using mouse epoxide hydrolase and human gas-tric lipase as the template, respectively [12]. The model of ClassII PHA synthase demonstrated that this enzyme belongs to the

icrobial Technology 56 (2014) 60–66 61

˛XsirCidfetstcP13PCrgKff[

tPsaoac4pFtstrseaseaevbvP[aIfsad

2

2

ptu

Table 1Bacterial strains and plasmids used in this study.

Strain or Plasmid Relevant characteristics Source

Pseudomonas putidaGPp104 PHA−

PHA-negtive mutant ofPseudomonas putida KT2440, mt-2,hsaR1(r-m+)

[17]

Ralstonia eutropha H16 Wild type, PHA+ ATCC17699E. coli DH10B F− mcrA �(mrr-hsdRMS-

mcrBC)�80lacZ�M15 �lacX74deoR recA1 endA1 ara�139�(ara,leu)7697 galU galK �− rpsL (Strr)nupG

Invitrogen

E.coli XL10-Goldultracompetent cells

lac Hte �(mcrA)183�(mcrCB-hsdSMR-mrr)173 tetr

F′[proAB lacIqZ�M15 Tn10(Tetr)Amy Cmr)]

Stratagene

pBluescript II KS(+) Ampr, lacOPZ’; cloning vector StrategenpBlue-REab pBluescript II KS(+) derivative;

ˇ-ketothiolase gene of R. eutrophaH16 (phbARe) and acetoacetyl-CoAreductase gene of R. eutropha H16(phbBRe) (4,958bp)

This study

pBBR1MCS-2 Kmr, lacPOZ’, Mob+,broad-host-range plasmid(5,144bp)

[27]

pBBR1MCS-2-ab pBBR1MCS-2 derivative; phbA andphbB of R. eutropha H16 (7,163bp)

This study

pBHR-POC1-29347 pBBR1MCS-2 derivative; phaC1 ofP. putida GPo1 (6,824bp)

[5]

pBHR-POC1-29347-ab pBHR-POC1-29347 derivative;phbA and phbB of R. eutropha H16(8,821bp)

This study

pPhaC-A295V-ab pBHR-POC1-29347-ab derivative;with site-directed mutation inphaC1Pp

This study

pPhaC-Q481M-ab pBHR-POC1-29347-ab derivative;with site-directed mutation inphaC1Pp

This study

pPhaC-S482G-ab pBHR-POC1-29347-ab derivative;with site-directed mutation inphaC1Pp

This study

pPhaC-L484X-aba pBHR-POC1-29347-ab derivative;with site-directed mutation inphaC1Pp

This study

pPhaC-A547V-ab pBHR-POC1-29347-ab derivative;with site-directed mutation in

This study

Y.-J. Chen et al. / Enzyme and M

/ ̌ hydrolase superfamily, containing a lipase-like box (G-X-C--G) and which is located at residue positions 294–298. Multipleequence alignment has led to the identification of a lipase-like boxn all PHA synthases [2]. In Class II PHA synthases, the amino acidesidues Cys296, Asp451, and His479 form a catalytic triad, withys296 as the nucleophile [12]. This Cys296 residue participates

n a nucleophilic attack on the substrate (R)-3-hydroxyacyl-CoAuring substrate loading and polymerization; Ser297 residue canorm an oxyanion hole to support the catalytic reaction of thenzyme [12]. In a previous study, directed evolution of PHA syn-hase has been applied in class II PHA synthase from Pseudomonasp. 61–3 (PhaC1Ps). The residues Ser325, Ser477 and Gln481 affecthe substrate specificity of the enzyme; their positions may belose to the side-chain binding pocket of the substrate [13,14].oint mutation occurred at C296S or H453Q of the PHA synthase

from P. aeruginosa increasing the affinity for incorporating (R)--hydroxyhexanoyl-CoA and (R)-3-hydroxydodecanoyl-CoA intoHA, but none of the mutants could use (R)-3-hydroxybutyryl-oA [15]. These point mutations are all in the ˛/ ̌ hydrolase foldegion of the PHA synthase family. Recently, site-directed muta-enesis of PHA synthases from Pseudomonas sp. 61-3, 6-19, P. putidaT2440, P. chlororaphis, P. resinovorans, and P. aeruginosa PAO1 at

our residues of Glu130, Ser325, Ser477 and Gln481 have beenound to affect substrate specificities of PHA synthases in E. coli16].

PHA synthase 1 of P. putida Gpo1 is a classic class II PHA syn-hase, which prefer MCL 3-acyl-CoA as a substrate for synthesizingHA [17]. All types of PHA synthases exhibit rather low sequenceimilarity [18]. The multiple sequence alignment results reveal thatlthough the sequences are quite different between various typesf PHA synthase, six conserved regions, termed F1, F2, F3, F4, F5,nd F6, are present in all PHA synthases. Each conserved regionontains amino acid residues 225-233, 291-299, 376-384, 398-403,78-485, and 518-525, respectively [5,19]. The results of multi-le sequence alignment implied that variants of amino acids in1–6 regions are naturally evolved. Our previous report showedhat there are 23 amino acids distributed throughout F1–6 con-erved regions involved in the protein evolution. Except for F1,he others conserved regions localized on the ˛/ ̌ hydrolase foldegion of the PHA synthase family [5]. Accordingly, the substratepecificity of PHA synthases can be altered by directing naturallyvolved amino acid residues. Based on this idea, we employed

localized semi random mutagenesis substantially changed theubstrate specificity of PhaC1Pp and yielded the evolved mutantnzymes, which exhibited a broad range of substrate specificitynd also conferred the higher PHA production [5]. These notablyvolved enzymes contained multiple point mutations, which wereery diverse. However, no individual mutations were identified aseing associated with a change in substrate specificity [5]. The pre-ious data indicated that evolution in the F3, F4 and F6 regions ofhaC1Pp might reduce enzyme activity and result in low PHA yields5]. Thus, it could be speculated that the variations occurred in F2nd F5 might be responsible of the expanded substrate specificity.n this study, we performed a mutational analysis of the PhaC1Pp,ocusing on some amino acids located in the F2 and F5 regions, usingite-specific and site-saturated mutagenesis. A threading model of

class II PHA synthase was developed and mutation positions wereiscussed.

. Materials and methods

.1. Bacterial strains, plasmids, and culture conditions

Table 1 lists the bacterial strains and plasmids used in this study. Pseudomonasutida GPp104 PHA− (P. putida GPp104 PHA−) was used as a host cell for analyzinghe substrate specificity of PhaC1Pp mutants. E. coli DH10B and E. coli XL10-Goldltracompetent cells were used as host cells for gene cloning and constructing the

phaC1Pp

a X indicates the site- saturation mutagenesis substitution 19 amino acid besidesLeucine.

mutant library, respectively. The vector pBluescript II KS (+) was used for generalgene cloning. The vector pBBR1MCS-2 was used for constructing the coexpressionplasmid containing the phaC1Pp , phbARe , and phbBRe genes. E. coli DH10B and E. coliXL10-Gold ultracompetent cells were cultivated at 37 ◦C in Luria-Bertani medium. P.putida GPp104 PHA− and R. eutropha H16 were cultivated at 30 ◦C in 2 × YT medium.Antibiotics were added as supplementation when required, including kanamycin50 mg/liter.

2.2. DNA manipulation and plasmid construction

Extraction of genomic DNA was achieved using a blood and tissue genomickit (Viogene, Taipei, Taiwan). Plasmid DNA was isolated using the Gene-Spin TM-V3 Miniprep Purification kit (Protech, Woodinville, WA, USA). DNA fragmentswere purified from agarose gels using the QIAquick gel extraction kit (QIAGEN,Chatsworth, CA, USA), and all restriction enzymes were purchased from NewEngland Biolabs (Ipswich, MA, USA). Primer synthesis and sequencing analysis werecarried out by Tri-I Biotech, Inc. (Taipei, Taiwan). The plasmid was introduced intohost cells by electroporation.

The recombinant plasmid pBlue-REab carrying the phbARe and phbBRe genes ofR. eutropha H16 was amplified by PCR with the following primer pair: abBamHF,5′-GTTCCCTCCGGATCCCATTGAA-3′ (BamHI site is underlined), and abXbalR, 5′-AGGCCTCTAGATCAGCCCATATG-3′ (XbalI site is underlined), from the chromosomalDNA of R. eutropha H16. The results were then cloned into the vector pBluescript

II KS(+). To construct the recombinant coexpression plasmids pBBR1MCS-2-ab andpBHR-POC1-29347-ab, the DNA fragments containing phbARe and phbBRe derivedfrom the plasmid pBlue-REab were digested with BamHI and XbalI to produce theinsert fragments, which were cloned into the vectors pBBR1MCS-2 and pBHR-POC1-29347, respectively.

6 icrobi

stp

2

a3bwddiaga

2

Piew[up

3

3

sostpps1hhw1oo5Tca(

wrcFpGt

3p

trwt

2 Y.-J. Chen et al. / Enzyme and M

The point mutations of the PhaC1Pp were generated using a QuickChange II XLite-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) following the manufac-urer’s instructions. pBHR-POC1-29347-ab was used as template. The correspondingrimer pairs used for introducing mutations are listed in Table S1.

.3. Production and analysis of PHA

P. putida GPp104 PHA− Cells harboring PhaC1Pp were cultivated in 30 mL 2 × YTt 30 ◦C for 14 h, then washed with mineral salt medium once and transferred to0 mL mineral salt medium with 0.5% sodium octanoate for an additional 30 h incu-ation at 30 ◦C with 130 rpm [5]. The PHA content and the polymer compositionere analyzed using gas chromatography. Cultures were centrifuged, washed withistilled water, and lyophilized overnight. Lyophilized cells were weighed for cellry weight. Ten milligram of lyophilized cell was methanolyzed at 100 ◦C for 140 min

n a solution consisting of 1 mL chloroform, 0.85 mL methanol, 0.15 mL 98% sulfuriccid, and 2 mg benzoic acid [20]. The methanolyzed monomers were analyzed byas chromatography, as previously described, for measurement of the PHA contentnd the polymer composition.

.4. Molecular modeling

The amino acid sequence was threaded to the library of known folds usingHYRE server [21]. The sequence-structure alignment calculated using the thread-ng method was then used as the input for the development of the 3D modelsmploying MODELLER [22]. The structure refinement procedure with water soakingas executed to regulate bond and angle distortions using the program GROMACS

23]. Finally, validation of PhaC model was evaluated by the Ramachandran plotsing PROCHECK program [24]. All the graphic presentations of the 3D model wererepared using PYMOL.

. Results

.1. Protein model building and structure predictions for PhaC1Pp

To investigate residues that are critical in affecting sub-trate specificity of PhaC1Pp, the three-dimensional (3D) modelf PhaC1Pp (accession AAA25932) was predicted. PSI-BLAST [25]earches of the PhaC1Pp sequence was performed against the pro-ein data bank (PDB); however there are no significant alignmentsroduced. Thus, secondary structure prediction and threadingrocedures were used to obtain the 3D model of PhaC1Pp. Theecondary structure of PhaC1Pp was composed of 23 �-helix and3 �-strands. The structure containing the alternate pattern of �-elices and �-strands was demonstrated which exhibit the ˛/ˇydrolase fold (Fig. 1). The 3D structure covering catalytic domainas built with the structure sequence alignment between residue

83 to residue 499 of PhaC1Pp and residue 227 to residue 544f human soluble expoxide hydrolase (pdb code: 1ZD3) [26]. Thether regions of the PhaC1Pp including residue 1 to 182 and00–559 were not modeled because without satisfied alignments.he alignment for threading showed the result agreed with theonserved catalytic residues of PhaC (Cys296, Asp451, His479)ligned with the catalytic triad of human soluble epoxide hydrolaseAsp333, Asp496, His524).

An overall structure including residues 183–499 of the PhaC1Pp

as displayed in Fig. 2A. This result indicated that the catalyticesidue Cys296 and lipase like box are located in the F2 region,atalytic residue His479 is located in the F5 region, and however,5 region is not far from the catalytic triads. A closer look at relatedositions of amino acid residues in Ala295 that was near Cys296 andln481, Ser482 and Leu484 that were near the His479 of catalytic

riads were presented in Fig. 2B.

.2. Effect of site-specific mutations of PhaC1Pp on PHAroduction

Several residues in F2 and F5 regions were identified as muta-

ion points by semi-random mutagenesis of PhaC1Pp in our previouseport [5]. To evaluate the conserved amino acid residues whichere near the catalytic triads in F2 and F5 regions of PhaC1Pp

hat affect the substrate specificity and activity of the enzyme,

al Technology 56 (2014) 60–66

we selected four mutations at residues Ala295, Gln481, Ser482,and Leu484 of PhaC1Pp to perform site-directed mutagenesis. Thecontents and composition of the PHA accumulated were deter-mined by gas chromatography (Table 2). The PhaC-Q481M-abled to relatively higher (R)-3-hydroxyhexanoate monomer frac-tion than wild type, and the PHA yield was effectively enhanced.This result suggested Gln481 has influence on both the substratespecificity and enzyme activity of PhaC1Pp PhaC-S482G-ab hadsimilar mol% value of (R)-3-hydroxybutyrate to wild-type, rela-tively higher (R)-3-hydroxyhexanoate monomer fraction than wildtype, yet the PHA content in PhaC-S482G-ab increased from 10%(wild-type) to 21%, suggesting that position 482 might affect sub-strate specificity and enzyme activity. PhaC-L484V-ab exhibitedmol% values for (R)-3-hydroxybutyrate that was obviously higherthan wild-type BHR-POC1-29347-ab. That indicating that aminoacid residue at 484 in PhaC1Pp might have the important effecton the substrate specificity of the enzyme. Furthermore, the PHAyields of PhaC-L484V-ab did not decrease compared with wild-type, demonstrating that this mutation did not affect enzymeactivity. PhaC-A295V-ab had a PHA content and mol% valuesfor (R)-3-hydroxybutyrate that were similar to wild-type BHR-POC1-29347-ab, but the PHA content in PhaC-A295V/S482G-ab,PhaC-A295 V/L484V-ab, and PhaC-A295V/S482G/L484V-ab dra-matically decreased from 10% (wild-type BHR-POC1-29347-ab)to 0.6–0.8%. The double- and triple-point mutations described inabove have had very low level of PHA content, implying that theposition at A295V combined with mutations such as S482G orL484V might lose the enzyme activity.

In addition, our previous results suggested that one A547Vmutation occurred in the C terminal of PhaC1Pp was likely toinfluence substrate specificity by semi-random mutagenesis [5].Therefore, the residue was also chosen as the target sites for thisstudy. The phaC-A547V-ab shows that the (R)-3-hydroxybutyratemol% was decreased, while the PHA yield improved remarkably.This data agreed with other report that hydrophobic amino acidsat the positions could enhance the PHA biosynthesis [13]. However,due to limited information on the tertiary structure of class II PHAsynthases at present, it is hard to explain why such amino acidschanges could result in these hydrophobic interactions.

3.3. Effect of site-saturated mutagenesis at position 484 inPhaC1Pp on PHA production

Because PhaC-L484V-ab might have the principal influence onthe substrate specificity of PhaC1Pp, site-saturation mutagenesisat position 484 in PhaC1Pp was carried out to observe the effectsof different amino acid substitutions on PhaC1Pp substrate speci-ficity. Fig. 3 showed that only L484V was able to increase the (R)-3-hydroxybutyrate unit content dominantly remarkably amongthe other amino acid substitutions. The PHA content of mutationL484V was not significantly different from that of the wild type.This result demonstrated that among the 19 amino acids, a valinereplacement at L484 lead to produce the maximum amount of (R)-3-hydroxybutyrate unit content, beside valine, serine or cystine,other amino acid substitutions at position L484 in PhaC1Pp havevery lower activities for PHA production. Thus, the leucine residueat position 484 plays an important role in regulation of PHA copoly-mer composition.

4. Discussion

Substrate specificity of the PhaC1Pp can be successfully alteredby single mutation in residue 484. On the basis of 3D model ofPhaC1Pp, the residues 484 was adjacent to the catalytic triad ofthe enzyme, when residue replacement of leucine with valine,

Y.-J. Chen et al. / Enzyme and Microbial Technology 56 (2014) 60–66 63

F ida GPc lignedp rands

tsr

C[iso

FcA

ig. 1. Sequence-structure alignment between PHA synthase 1 of Pseudomonas putatalytic triad, Cys296, Asp451 and His479. The catalytic site for both proteins is aentapeptide motif (lipase like box) G-X-C-X-G. Blue line and pink line denote �-st

he higher hydrophobicity of valine may be lead to orientation ofhort substrates (R)-3-hydroxybutyl-CoA stabilized than leucine,esulted in the substrate specificity of enzyme toward 3HB unit.

The effects of mutation at Q481 have been investigated withlasses II PHA synthase PhaC1 from Pseudomonas sp. 61-3 (PhaC1Ps)13]. Previously, amino acid substitutions in PhaC1 at 481

Ps

ncreased the PHA accumulation and altered the monomer compo-ition of PHA in recombinant E. coli. The residue 481 replacementf glutamine with methionie of Classes II PHA synthase PhaC1Ps

ig. 2. Structural view of the active site of the predicted model PhaC1Pp (A) >Ribbon diagontaining Cys296 of catalytic triad) and purple (F5 region, residue 478–485, containing

sp451 and His479 in PhaC1Pp . The residues that were mutated are shown as stick and co

o1 (PhaC1Pp) and 1ZD3 from residue 183–499. The triangular marker indicated the very well between the proteins. The blank rectangle box indicated the consensus

and �-helices.

can increase the PHA biosynthesis toward poly(3-hydroxybutyrate)[13]. In this study, recombinant strain of P. putida GPp104 PHA−

harboring Q481M single mutation in PhaC1Pp remarkably promotethe PHA yield and enhance the 3HHx monomer fraction in the PHA(Table 2). The similar effect was shown in residue 482 replacementof serine with glysine. These results indicated point mutations at

Q481 and S482 contributed to the alteration of substrate specificityand increasing the PHA yield. On the basis of 3D model of PhaC1Pp,the residues Gln481, Ser482 was located near to the His479 in

ram of the PhaC1Pp structure model are shown in red (F2 region, residue 291–299,His479 of catalytic triad). (B). Three catalytic residues in the active site are Cys296,lored by element.

64 Y.-J. Chen et al. / Enzyme and Microbial Technology 56 (2014) 60–66

Table 2Composition analysis of PHA accumulated with recombinants strains of P. putida GPp104 PHA− harboring single or multiple mutants of the phaC1Pp .

Mutant PHA content (% wt./wt.)a Polymer composition (mol%)b

3-HB 3-HHx 3-HO 3-HD

BHR-POC1-29347-ab 10 ± 0.7 8 ± 2.9 12 ± 0.2 75 ± 5.5 6 ± 2.7BBR1MCS-2-ab 0.3 ± 0.03 ND ND ND 100PhaC-Q481M-ab 35 ± 0.7 8 ± 0.4 22 ± 0.3 69 ± 0.6 2 ± 0.03PhaC-S482G-ab 21 ± 1.0 8 ± 0.9 26 ± 2.3 64 ± 2.9 2 ± 0.3PhaC-L484V-ab 11 ± 1.7 36 ± 5.1 13 ± 0.1 47 ± 5.3 5 ± 0.7PhaC-A295V-ab 11 ± 3.0 7 ± 1.3 7 ± 0.2 83 ± 1.6 4 ± 0.8PhaC-A547V-ab 37 ± 1.3 3 ± 0.4 15 ± 0.1 81 ± 0.4 2 ± 0.1PhaC-A295V/L484V-ab 0.7 ± 1.8 60 ± 5.6 ND ND 40 ± 4.3PhaC-A295V/S482G-ab 0.8 ± 0.2 55 ± 6.2 ND 16 ± 5.2 28 ± 1.8PhaC-A295V/S482G/L484V-ab 0.6 ± 0.2 59 ± 4.3 ND 9 ± 2.4 33 ± 3.2

medio ented

noate;

ct

Pocmaastp

FPp

a Cells cultivated on 30 mL 2xYT at 30 ◦C for 14 h; then washed with mineral saltctanoate for an additional 30 h incubation at 30 ◦C with 130 rpm. The data are presb 3-HHx, 3-hydroxyhexanoate; 3-HO, 3-hydoxyoctanoate; 3-HD, 3-hydroxydeca

atalytic triad that may influence the position of His479 cause ofhe PHA yield increased.

Moreover, site-saturated mutagenesis analysis of leuine 484 inhaC1Pp substituted with valine yielded the greatest incorporationf the (R)-3-hydroxybutyrate monomer into PHA. This result wasonsistent with findings for the residues valine and isoleucine, theost conserved residue at the corresponding positions of class I

nd III synthases based on a comparison using multiple sequencelignment (Fig. 4). Both class I and III PHA synthases prefer the

hort-chain-length (R)-3-hydroxyacyl-CoA as substrate [14]. Fur-hermore, except for valine, almost amino acid substitutions atosition 484 in PhaC1Pp would decrease PHA yield. It is possible

ig. 3. Intracellular 3-HB unit contents in total PHA from recombinant P. putida GPp104 PHA synthase 1. Cells cultivated conditions were the same as the method described in Tresented as mean ±standard deviations of three independent experiments. 3-HB, 3-hyd

um once and transferred to 30 mL mineral salt medium (pH7.4) with 0.5% sodium as mean ±standard deviations of three independent experiments.

ND, not detected.

that the residue was close to the His479 in catalytic triad, mutationat 484 residue could influence the conformation of catalytic triad.

The important effect of pH on PHA production during batchcultivation was observed in our previous studies as the PHA accu-mulation was promoted when pH was 8.5 [unpublished data]. Dueto this point of view, mineral salt medium with pH 8.5 was usedin saturation mutagenesis experiment at residue 484. As shown inthe Fig. 3, the PHA contents of the P. putida GPp104 PHA− strainsharboring phaC1-ab and phaC1-L484V-ab were 30.27% CDW and

22.13% CDW, respectively. Both of them was relatively higher thanthat in Table 2, which was 11% CDW and 10% CDW when strainscultivated in pH 7.4 mineral salt medium.

HA− harboring site-saturation mutants at amino acid position 484 of P. putida GPo1able 2, except for using the 30 mL mineral salt medium with pH8.5. The data areroxybutanoate; phaC1-ab, pBHR-POC1-29347-ab.

Y.-J. Chen et al. / Enzyme and Microbial Technology 56 (2014) 60–66 65

Fig. 4. Partial sequence alignment of class II PHA synthase and corresponding amino acid from class I and III PHA synthase. The mutated site and catalytically important residuesa mophid i, RastR latensi

ocefpeoasw

5

ePSbagahLspu

A

vP0C

A

i2

[

[

[

[

[

[

re star- and triangle-labeled, respectively. Pseudomonas putida, Pseudomonas entroomonas aeroginosa, Aeromonas caviae, Alcaligenes latus, Burkholderia pseudomallehodobacter sphaeroides, Synechococcus sp. MA19, Chlorogloea fritschii, Arthrospira p

The engineered PHA synthases were capable of synthesisf PHAs containing different monomer compositions and PHAontents in the P. putida GPp104 PHA− (Table 2). In our studies, wengineered the PHA synthase PhaC1 from P. putida, and the vectoror expression of PhaC is pBBR1MCS2, which is a low copy numberlasmid and thus displayed a few amounts of proteins. However, toxamine these copolymers forming abilities of the strains whetherr not related with their PHA synthase activities, the PHA synthasectivities towards 3-hydroxyacyl-CoAs and PHA synthase expres-ion levels of engineered PHA synthases developed in this studyill be evaluated in the further investigations.

. Conclusion

In summary, mutation at L484V of PhaC1Pp remarkablynhanced the (R)-3-hydroxybutyrate monomer composition in theHA accumulation experiment. Besides, mutant enzymes Q481M,482G and A547V obviously increased PHA yields. Among them,oth Q481M and S482G increase 3HHx mole fractions in PHAccumulation in the P. putida GPp104 PHA− cells. Saturation muta-enesis at 484 residue demonstrated that Val is the most beneficialmino acid in PhaC1Pp for enhancing the synthesis of poly(3-ydroxybutyrate). This is the first report indicating that Ala547 andeu484 residues of PhaC1Pp are crucial for polymer yield and sub-trate specificity, respectively. These findings can serve as a startingoint for improving PHA yield and regulating PHA compositionsing engineered PHA synthase.

cknowledgments

The authors thank Drs. M. E. Kovach and K. M. Peterson for pro-iding the plasmid pBBR1MCS-2 and B. Witholt for P. putida GPp104HA−. This work was supported in part by grants NSC 94-2313-B-02-027 and NSC 99-2324-B-002-006 from the National Scienceouncil, Taipei, Taiwan.

ppendix A. Supplementary data

Supplementary data associated with this article can be found,n the online version, at http://dx.doi.org/10.1016/j.enzmictec.014.01.005.

[

la, Pseudomonas sp. 61-3, Pseudomonas chlororaphis, Pseudomonas fluorescens, Pseu-onia eutropha, Paracoccus denitrificans, Ralstonia metallidurans, Rhodococcus ruber,s.

References

[1] Budde CF, Mahan AE, Lu J, Rha C, Sinskey AJ. Roles of multiple acetoacetyl coen-zyme A reductases in polyhydroxybutyrate biosynthesis in Ralstonia eutrophaH16. J Bacteriol 2010;192:5319–28.

[2] Tan Y, Neo P-C, Najimudin N, Sudesh K, Muhammad TST, Othman AS, et al.Cloning and characterization of poly(3-hydroxybutyrate) biosynthesis genesfrom Pseudomonas sp. USM 4-55. J Basic Microbiol 2010;50:179–89.

[3] Gao X, Chen J-C, Wu Q, Chen G-Q. Polyhydroxyalkanoates as a source of chem-icals, polymers, and biofuels. Curr Opin Biotechnol 2011;22:768–74.

[4] Scheller Jg Conrad U. Plant-based material, protein and biodegradable plastic.Curr Opin Plant Biol 2005;8:188–96.

[5] Sheu D-S, Lee C-Y. Altering the substrate specificity of polyhydroxyalkanoatesynthase 1 derived from Pseudomonas putida GPo1 by localized semirandommutagenesis. J Bacteriol 2004;186:4177–84.

[6] Li Q, Chen Q, Li M-J, Wang F-S, Qi Q-S. Pathway engineering results thealtered polyhydroxyalkanoates composition in recombinant Escherichia coli.New Biotechnol 2011;28:92–5.

[7] Zinn M. Biosynthesis of medium-chain-length poly[(R)-3-hydroxyalkanoates].In: Chen G-Q, editor. Plastics from bacteria: natural functions and applications.Heidelberg, Berlin: Springer; 2010. p. 213–36.

[8] Prieto M, Eugenio Ld, Galàn B, Luengo J, Witholt B. Synthesis and degrada-tion of polyhydroxyalkanoates. In: Ramos J-L, Filloux A, editors. Pseudomonas.Netherlands: Springer; 2007. p. 397–428.

[9] Bhubalan K, Chuah J-A, Shozui F, Brigham CJ, Taguchi S, Sinskey AJ,et al. Characterization of the highly active polyhydroxyalkanoate synthaseof Chromobacterium sp. strain USM2. Appl Environ Microbiol 2011;77:2926–33.

10] Tomizawa S, Hyakutake M, Saito Y, Agus J, Mizuno K, Abe H, et al. Molecu-lar weight change of polyhydroxyalkanoate (PHA) caused by the PhaC subunitof PHA synthase from Bacillus cereus YB-4 in recombinant Escherichia coli.Biomacromolecules 2011;12:2660–6.

11] Taguchi S. Current advances in microbial cell factories for lactate-basedpolyesters driven by lactate-polymerizing enzymes: towards the fur-ther creation of new LA-based polyesters. Polym Degrad Stabil 2010;95:1421–8.

12] Wahab H, Ahmad Khairudin N, Samian M, Najimudin N. Sequence analysis andstructure prediction of type II Pseudomonas sp. USM 4-55 PHA synthase and aninsight into its catalytic mechanism. BMC Struct Biol 2006;6:23.

13] Takase K, Matsumoto K, Taguchi S, Doi Y. Alteration of substrate chain-length specificity of type II synthase for polyhydroxyalkanoate biosynthesisby in vitro evolution: in vivo and in vitro enzyme assays. Biomacromolecules2004;5:480–5.

14] Matsumoto K, Aoki E, Takase K, Doi Y, Taguchi S. In vivo and in vitro char-acterization of Ser477X mutations in polyhydroxyalkanoate (PHA) synthase 1from Pseudomonas sp., 61-3: effects of beneficial mutations on enzymatic activ-ity, substrate specificity, and molecular weight of PHA. Biomacromolecules2006;7:2436–42.

15] Amara AA, Rehm BHA. Replacement of the catalytic nucleophile cysteine-

296 by serine in classII polyhydroxyalkanoate synthase from Pseudomonasaeruginosa-mediated synthesis of a new polyester: identification of catalyticresidues. Biochem J 2003;374:413–21.

16] Yang TH, Jung YK, Kang HO, Kim TW, Park SJ, Lee SY. Tailor-made type II Pseu-domonas PHA synthases and their use for the biosynthesis of polylactic acid

6 icrobi

[

[

[

[

[

[

[

[

[

[

6 Y.-J. Chen et al. / Enzyme and M

and its copolymer in recombinant Escherichia coli. Appl Microbiol Biotechnol2011;90:603–14.

17] Huisman GW, Wonink E, Meima R, Kazemier B, Terpstra P, Witholt B.Metabolism of poly(3-hydroxyalkanoates) (PHAs) by Pseudomonas oleovoransIdentification and sequences of genes and function of the encoded pro-teins in the synthesis and degradation of PHA. J Biologic Chem 1991;266:2191–8.

18] Rehm BHA. Polyester synthases: natural catalysts for plastics. Biochem J2003;376:15–33.

19] Rehm BHA, Steinbüchel A. Biochemical and genetic analysis of PHA syn-thases and other proteins required for PHA synthesis. Int J Biologic Macromol1999;25:3–19.

20] Wu HA, Sheu D-S, Lee C-Y. Rapid differentiation between short-chain-lengthand medium-chain-length polyhydroxyalkanoate-accumulating bacteria withspectrofluorometry. J Microbiol Meth 2003;53:131–5.

21] Kelley LA, Sternberg MJE. Protein structure prediction on the Web: a case studyusing the Phyre server. Nat Protoc 2009;4:363–71.

[

al Technology 56 (2014) 60–66

22] Sali A, Blundell TL. Comparative protein modeling by satisfaction of spatialrestraints. J Mol Biol 1993;234:779–815.

23] Hess B, Kutzner C, van der Spoel D, Lindahl E. GROMACS4: Algorithms for highlyefficient, load-balanced, and scalable molecular simulation. J Chem Theor Com-put 2008;4:435–47.

24] Laskowski RA, MacArthur MW, Moss DS, Thornton JM. PROCHECK: a programto check the stereochemical quality of protein structure. J Appl Crystallogr1993;26:283–91.

25] Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, et al. GappedBLAST and PSI-BLAST: a new generation of protein database search programs.Nucleic Acid Res 1997;25:3389–402.

26] Gomez GA, Morisseau C, Hammock BD, Christianson DW. Human soluble

epoxide hydrolase: structural basis of inhibition by 4-(3-cyclohexylureido)-carboxylic acids. Protein Sci 2006;15:58–64.

27] Kovach ME, Elzer PH, Hill DS, Robertson GT, Farris MA, Roop 2nd RM, et al. Fournew derivatives of the broad-host-range cloning vector pBBR1MCS, carryingdifferent antibiotic-resistance cassettes. Gene 1995;166:175–6.