Applied Microbiology and Biotechnology

, Volume 93, Issue 1, pp 273–283

Biosynthesis of polyhydroxyalkanoates containing 2-hydroxybutyrate from unrelated carbon source by metabolically engineered Escherichia coli

Authors

  • Si Jae Park
    • Chemical Biotechnology Research CenterKorea Research Institute of Chemical Technology
  • Tae Woo Lee
    • Metabolic and Biomolecular Engineering National Research Laboratory, Department of Chemical and Biomolecular Engineering (BK21 Program)Center for Systems and Synthetic Biotechnology, and Institute for the BioCentury
  • Sung-Chul Lim
    • Corporate R&D, LG Chem Research Park
  • Tae Wan Kim
    • Marine Biotechnology Research CenterKorea Ocean Research & Development Institute
  • Hyuk Lee
    • Medicinal Chemistry Research CenterKorea Research Institute of Chemical Technology
  • Min Kyung Kim
    • Metabolic and Biomolecular Engineering National Research Laboratory, Department of Chemical and Biomolecular Engineering (BK21 Program)Center for Systems and Synthetic Biotechnology, and Institute for the BioCentury
  • Seung Hwan Lee
    • Chemical Biotechnology Research CenterKorea Research Institute of Chemical Technology
  • Bong Keun Song
    • Chemical Biotechnology Research CenterKorea Research Institute of Chemical Technology
    • Metabolic and Biomolecular Engineering National Research Laboratory, Department of Chemical and Biomolecular Engineering (BK21 Program)Center for Systems and Synthetic Biotechnology, and Institute for the BioCentury
    • Department of Bio and Brain Engineering, Department of Biological SciencesBioProcess Engineering Research Center and Bioinformatics Research Center
Applied genetics and molecular biotechnology

DOI: 10.1007/s00253-011-3530-x

Cite this article as:
Park, S.J., Lee, T.W., Lim, S. et al. Appl Microbiol Biotechnol (2012) 93: 273. doi:10.1007/s00253-011-3530-x

Abstract

We have previously reported in vivo biosynthesis of polylactic acid (PLA) and poly(3-hydroxybutyrate-co-lactate) [P(3HB-co-LA)] employing metabolically engineered Escherichia coli strains by the introduction of evolved Clostridium propionicum propionyl-CoA transferase (PctCp) and Pseudomonas sp. MBEL 6-19 polyhydroxyalkanoate (PHA) synthase 1 (PhaC1Ps6-19). Using this in vivo PLA biosynthesis system, we presently report the biosynthesis of PHAs containing 2-hydroxybutyrate (2HB) monomer by direct fermentation of a metabolically engineered E. coli strain. The recombinant E. coli ldhA mutant XLdh strain expressing PhaC1Ps6-19 and PctCp was developed and cultured in a chemically defined medium containing 20 g/L of glucose and varying concentrations of 2HB and 3HB. PHAs consisting of 2HB, 3HB, and a small fraction of lactate were synthesized. Their monomer compositions were dependent on the concentrations of 2HB and 3HB added to the culture medium. Even though the ldhA gene was completely deleted in the chromosome of E. coli, up to 6 mol% of lactate was found to be incorporated into the polymer depending on the culture condition. In order to synthesize PHAs containing 2HB monomer without feeding 2HB into the culture medium, a heterologous metabolic pathway for the generation of 2HB from glucose was constructed via the citramalate pathway, in which 2-ketobutyrate is synthesized directly from pyruvate and acetyl-CoA. Introduction of the Lactococcus lactis subsp. lactis Il1403 2HB dehydrogenase gene (panE) into E. coli allowed in vivo conversion of 2-ketobutyrate to 2HB. The metabolically engineered E. coli XLdh strain expressing the phaC1437, pct540, cimA3.7, and leuBCD genes together with the L. lactis Il1403 panE gene successfully produced PHAs consisting of 2HB, 3HB, and a small fraction of lactate by varying the 3HB concentration in the culture medium. As the 3HB concentration in the medium increased the 3HB monomer fraction in the polymer, the polymer content increased. When Ralstonia eutropha phaAB genes were additionally expressed in this recombinant E. coli XLdh strain, P(2HB-co-3HB-co-LA) having small amounts of 2HB and LA monomers could also be produced from glucose as a sole carbon source. The metabolic engineering strategy reported here should be useful for the production of PHAs containing 2HB monomer.

Keywords

Polyhydroxyalkanoate (PHA)2-hydroxybutyrate (2HB)PHA synthaseRecombinant E. coli

Introduction

Polyhydroxyalkanoates (PHAs) are biodegradable and biocompatible polyesters consisting of various hydroxycarboxylic acids, which accumulate in many bacteria as carbon (energy) and/or reducing power storage material (Lee 1996; Madison and Huisman 1999). PHAs are polymerized by PHA synthase, which uses various hydroxyacyl-CoAs as its substrates. If hydroxyacyl-CoAs contain an asymmetric center at the carbon position of hydroxyl group, they are all in the (R)-configuration. Among various hydroxyacyl-CoAs, natural PHA synthases generally accept 3-hydroxyacyl-CoAs (3HA-CoAs) as the most favorable substrates, but 4-, 5-, and 6-hydroxyacyl-CoAs can also be accepted (Lee 1996; Madison and Huisman 1999). However, there has been no report on the in vivo biosynthesis of PHAs containing 2-hydroxyacid as monomer by natural PHA-producing bacteria. Natural PHA synthases screened to date have shown very low or negligible in vitro activities towards 2-hydroxyacyl-CoAs compared with other HA-CoAs (Yuan et al. 2001; Zhang et al. 2001). Recently, we developed metabolically engineered E. coli strains capable of producing polylactic acid (PLA) and its copolymers consisting of lactate and other monomers including 3-hydroxypropionate (3HP), 3-hydroxybutyrate (3HB), 3-hydroxyvalerate (3HV), 3-hydroxyhexanoate (3HHx), 3-hydroxyoctanoate (3HO), 3-hydroxydecanoate (3HD), 3-hydroxydodecanoate (3HDD), and 4-hydroxybutyrate (4HB) by a direct fermentation (Jung et al. 2010; Jung and Lee 2011; Park et al. 2008a, b; Yang et al. 2010, 2011). In this metabolically engineered E. coli strain, PLA is synthesized by in vivo generation of (d)-lactyl-CoA followed by its polymerization. We have demonstrated that (d)-lactyl-CoA can be synthesized in recombinant E. coli by employing the engineered Clostridium propionicum propionyl-CoA transferase (PctCp) that converts (d)-lactate to (d)-lactyl-CoA using acetyl-CoA as a CoA donor. For the polymerization of (d)-lactyl-CoA to (d)-PLA, Pseudomonas sp. 6-19 PHA synthase (PhaC1Ps6-19) was engineered to efficiently utilize (d)-lactyl-CoA as a substrate because natural PHA synthases have negligible activity towards (d)-lactyl-CoA (Park et al. 2008a, b; Yang et al. 2010; Jung et al. 2010). Also, the efficiency of PLA synthesis could be enhanced by genome engineering of E. coli XL1-Blue to strengthen (d)-lactyl-CoA biosynthesis flux based on the systems-level metabolic pathway analysis and by enhancing the PctCp activity by random mutagenesis (Jung et al. 2010; Yang et al. 2010). In a different study, Taguchi and colleagues also reported the biosynthesis of poly(3-hydroxybutyrate-co-lactate) [P(3HB-co-LA)] using a similar system (Shozui et al. 2010; Taguchi et al. 2008; Yamada et al. 2009, 2010, 2011). As the engineered PhaC1Ps6-19 has a much broad substrate specificity and lactate is also the member of 2-hydroxyacid, it was reasoned that other 2-hydroxyacyl-CoAs might also be used as substrates for PhaC1Ps6-19 resulting in the biosynthesis of novel PHAs containing 2-hydroxyacid monomer. Because 2-hydroxybutyrate (2HB) can be derived from the amino acid biosynthetic pathway, it was thought that 2HB might not have toxic effects to E. coli. Thus, 2HB was examined as a proof-of-concept monomer among various 2-hydroxyacids for the biosynthesis of PHAs containing 2-hydroxyacids.

Recently, the citramalate pathway for the production of 1-propanol and 1-butanol from glucose, which directly converts pyruvate to 2-ketobutyrate, was developed by employing the evolved Methanococcus jannaschii citramalate synthase (CimA) (Atsumi and Liao 2008a). The precursor for 2HB, 2-ketobutyrate, could be produced from this metabolic pathway. Thus, E. coli was metabolically engineered to produce 2-hydroxybutyryl-CoA (2HB-CoA), the substrate for PhaC1Ps6-19, by additional expression of the 2-hydroxyacid dehydrogenase gene and propionyl-CoA transferase gene together with the genes corresponding to citramalate pathway (Fig. 1). Using this system, we presently report for the first time the microbial biosynthesis and characterization of novel PHA copolymers containing 2HB as monomer by metabolically engineered E. coli.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-011-3530-x/MediaObjects/253_2011_3530_Fig1_HTML.gif
Fig. 1

Metabolic pathways for the production of PHAs containing 2HB monomer developed in this study. The overall metabolic pathway is shown together with the introduced metabolic pathways for the production of PHAs containing 2HB monomer in E. coli. Cross marks represent blocked pathway

Materials and methods

Bacterial strains and plasmids

All bacterial strains and plasmids used in this study are listed in Table 1. E. coli XL1-Blue (Stratagene Cloning Systems, La Jolla, CA, USA) was used for general gene cloning studies. For the production of PHA copolymers containing 2HB, a recombinant E. coli ldhA mutant, XLdh strain, was used as a host strain. Plasmid p619C1437-pct540, which expresses the Pseudomonas sp. 6-19 PHA synthase gene containing quadruple mutations of E130D, S325T, S477G, and Q481K (PhaC1437) and the C. propionicum propionyl-CoA transferase mutant gene containing V193A and four silent nucleotide mutations of T78C, T669C, A1125G, and T1158C (Pct540) under the Ralstonia eutropha PHA biosynthesis operon promoter, has been previously described (Yang et al. 2010). Also, the M. jannaschii cimA3.7 gene (Atsumi and Liao 2008a) was synthesized from GenScript (www.genscript.com) based on the reported sequence. Plasmids pZE12-MCS and pZA31-MCS were purchased from the EXPRESSYS (www.expressys.com).
Table 1

Lists of bacterial strains and plasmids used in this study

Plasmid

Relevant characteristics

Reference or source

Strains

E. coli XL1-Blue

recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F′ proAB lacIqZΔM15 Tn10 (TetR)]

Stratagenea

E. coli XLdh

XL1-Blue ΔldhA

This study

E. coli W3110

F-mcrA mcrB IN(rrnD-rrnE)1λ

KCTCb

C. difficile

Genome sequenced strain, ATCC BAA-1382

ATCCc

L. lactis subsp. lactis Il1403

Genome sequenced strain

KCCMd

R. eutropha NCIMB11599

Natural PHA producer ; GentamicinR

NCIMB

Plasmids

p619C1437-pct540

pBluescript KS II(+) derivative; R. eutropha PHA biosynthesis promoter, the phaC1437Ps6-19 gene, the pct540Cp gene; Apr

Yang et al. 2010

pBBR1MCS2

Broad host range vector; Kmr

Kovach et al. 1995

pZE12-MCS

Expression vector; PLlacO-1 promoter; Apr

EXPRESSYS

pKE12-MCS

Expression vector; PLlacO-1 promoter, R. eutropha PHA biosynthesis genes transcription terminator; Apr

This study

pZA31-MCS

Expression vector; PLtetO-1 promoter; Cmr

EXPRESSYS

pKA32-MCS

Expression vector; PLlacO-1 promoter, R. eutropha PHA biosynthesis genes transcription terminator; Cmr

This study

pKM22-MCS

pBBR1MCS2 derivative ; PLlacO-1 promoter, R. eutropha PHA biosynthesis genes transcription terminator; Kmr

This study

p619C1437LeuBCD

p619C1437-pct540 derivative; R. eutropha PHA biosynthesis promoter, the phaC1437Ps6-19 gene, the E. coli leuBCD gene; Apr

This study

pKA32CimA

pKA32-MCS derivative; the M. jannaschii cimA3.7 gene; Cmr

This study

pKA32CimALeuBCD

pKA32-MCS derivative; the M. jannaschii cimA3.7 gene, the E. coi leuBCD gene; Cmr

This study

pKM22PanE

pKM22-MCS derivative; PLlacO-1 promoter, the L. lactis subsp. lactis Il1403 PanE gene; Kmr

This study

pKM22LdhA

pKM22-MCS derivative; PLlacO-1 promoter, the C. difficile 630 ldhA gene; Kmr

This study

pKM22PanEPhaBA

pKM22-MCS derivative; PLlacO-1 promoter, the L. lactis subsp. lactis Il1403 PanE gene, the R. eutropha phaB and phaA genes; Kmr

This study

aStratagene Cloning System, La Jolla, CA, USA

bKorean Collection for Type Cultures, Daejeon, Republic of Korea

cAmerican Type Culture Collection

dKorean Culture Center of Microorgansims, Seoul, Republic of Korea

Construction of plasmids

All DNA manipulations were performed following standard procedures (Sambrook and Russell 2001). Polymerase chain reaction (PCR) was performed with the C1000 Thermal Cycler (Bio-Rad, Hercules, CA, USA). Primers used in this study (Table 2) were synthesized at Bioneer (Daejeon, Korea). The expression vector, pZE12-MCS, was modified to have multiple-cloning sites (MCS) of pTacLac (Lee et al. 2008) and the transcription terminator of the R. eutropha PHA biosynthesis operon. First, the terminator of the R. eutropha PHA biosynthesis operon was obtained by PCR with the primers 8 and 9 using p619C1437-pct540 as a template. PCR was performed with primers 7 and 9 using this gene fragment as a template to obtain the gene fragment containing RBS and pTacLac MCS, and the R. eutropha transcription terminator. This PCR product was digested with MfeI and AvrII and was cloned into the EcoRI/AvrII digested pZE12-MCS to obtain pKE12-MCS. Plasmid pKA32-MCS was constructed by replacing the PLtetO-1 promoter in pKA31-MCS with the PLlacO-1 promoter obtained from pKE12-MCS by XhoI/EcoRI digestion.
Table 2

List of primers used in this study

Primer

Primer sequence

Target gene

Primer 1

CCTGCAGGTTTCACACAGGAAACA ATGTCGAAGAATTACCATATTG

E. coli leuBCD genes

Primer 2

CATATGTTAATTCATAAACGCAGGTTG

Primer 3

GAATTCATGAGAATTACAATTGCCGG

L. lactis subsp. lactis Il1403 panE gene

Primer 4

GGTACCTTATTTTGCTTTTAATAACTC

Primer 5

GAATTCATGAAAATACTAGTATTTGG

C. difficile 630 ldhA gene

Primer 6

GGTACCCTAATTTACTCTATTAGTAG

Primer 7

CAATTGATTAAAGAGGAGAAA GAATTCGAGCTCGGTACCCGGGGATCCTCTAGAGTCGACCTGCAGGCATGCAAGCTT

RBS and MCS for pKE12-MCS

Primer 8

GAGTCGACCTGCAGGCATGCAAGCTT CCTGCCGGCCTGGTTCAACC

R. eutropha PHA biosynthesis genes transcription terminator

Primer 9

CCTAGG GCCTCGCCCCCGCGAGGGCC

Primer 10

AATATT AATTGTGAGCGGATAACAAT

PLlacO-1 promoter, MCS, and R. eutropha PHA biosynthesis genes transcription terminator for pKM22-MCS

Primer 11

AATATT GCCTCGCCCCCGCGAGGGCC

Primer 12

GGTACCTTTCACACAGGAAACA ATGACTCAGCGCATTGCGTA

R. eutropha phaB gene

Primer 13

GGATCCTTATCAGCCCATGTGCAGGC CGCCGTTGAGCGAG

Primer 14

GCAGGCCTGCAGGGGGATCC TTCCCTCCCGTTTCCATTGAAAG

R. eutropha phaA gene

Primer 15

AACTGCAGGTCCACTCCTTGATTGGCTTC

Genes involved in the generation of 2HB-CoA were amplified from the chromosomal DNAs of E. coli W3110, Clostridium difficile 630 and Lactococcus lactis subsp. lactis Il1403, and from the plasmid p619C1437-pct540. R. eutropha phaAB genes were amplified from the chromosomal DNAs of R. eutropha NCIMB11599. Plasmid p619C1437LeuBCD was constructed by replacing the pct540 gene in p619C1437-pct540 with the PCR-amplified E. coli leuBCD genes (using the primers 1 and 2) at SbfI and NdeI sites. Plasmid pKA32-CimA was constructed by cloning the synthesized cimA3.7 gene (Atsumi and Liao 2008a) into pKA32-MCS at EcoRI and SbfI sites. Then, pKA32-CimALeuBCD was constructed by cloning the E. coli leuBCD genes obtained from p619C1437LeuBCD by digestion with SbfI and HindIII into pKA32-CimA. The gene fragment containing PLlacO-1 promoter, MCS, and R. eutropha transcription terminator was amplified from pKA32-MCS using the primers 10 and 11, and then was cloned into the SspI of pBBR1MCS2 to make pKM22-MCS. Plasmids pKM22PanE and pKM22LdhA were constructed by cloning the L. lactis subsp. lactis Il1403 panE gene and the C. difficile 630 ldhA gene into pKM22-MCS at EcoRI and KpnI sites, respectively. Plasmid pKM22PanEPhaB was constructed by cloning the R. eutropha phaB gene at KpnI and BamHI sites. Then, the R. eutropha phaA gene was cloned into pKM22PanEPhaB at BamHI and PstI sites to make pKM22PanEPhaBA.

Culture condition

E. coli XL1-Blue was cultured at 37°C in lysogeny broth (LB) medium (containing per liter 10 g tryptone, 5 g yeast extract, and 5 g NaCl). For the synthesis of PHAs containing 2HB monomer, recombinant E. coli XLdh strains were cultured in MR medium supplemented with 20 g/L of glucose and desired concentrations of 2HB and 3HB at 30°C in a rotary shaker at 250 rpm for 72 h. MR medium (pH 7.0) contains (per liter) 6.67 g KH2PO4, 4 g (NH4)2HPO4, 0.8 g MgSO4·7H2O, 0.8 g citric acid, and 5 ml trace metal solution. The trace metal solution contains (per liter of 0.5 M HCl) 10 g FeSO4·7H2O, 2 g CaCl2, 2.2 g ZnSO4·7H2O, 0.5 g MnSO4·4H2O, 1 g CuSO4·5H2O, 0.1 g (NH4)6Mo7O24·4H2O, and 0.02 g Na2B4O7·10H2O. Glucose, MgSO4·7H2O, 2HB, and 3HB (Acros Organics, Geel, Belgium) were sterilized separately. 2HB and 3HB were added into the culture medium to different concentrations. Recombinant E. coli XLdh strain was grown to an optical density (600 nm) of 0.5 before induction with 1 mM of isopropyl-beta-d-thiogalactopyranoside (IPTG) for the expression of the cimA, leuBCD, phaAB, and panE genes. Ampicillin (Ap, 50 μg/mL), kanamycin (Km, 30 μg/mL), and chloramphenicol (Cm, 34 μg/mL) were added to the medium depending on the resistance marker of the employed plasmid.

Genome engineering

Deletion of the ldhA gene in the chromosome of E. coli XL1-Blue to reduce lactate synthesis was carried out using the one-step inactivation method as previously reported (Jung et al. 2010; Datsenko and Wanner 2000).

Polymer analysis

The content and monomer composition of the synthesized polymer were determined by gas chromatography (GC) (Braunegg et al. 1978). Polymers were purified from the cells by the solvent extraction method (Jacquel et al. 2008). The structure, molecular weight, and thermal properties of the polymers were determined by nuclear magnetic resonance (NMR) spectroscopy, gel permeation chromatography (GPC), and differential scanning calorimetry (DSC) as previously described (Yang et al. 2010; Jung et al. 2010; Lim et al. 2008).

Results

Biosynthesis of PHAs containing 2HB monomer in recombinant E. coli XLdh by adding 2HB and 3HB into the culture medium

We previously reported the production of P(3HB-co-LA) in recombinant E. coli employing the engineered PhaC1Ps6-19 and PctCp (Yang et al. 2010; Jung et al. 2010). As the engineered PhaC1Ps6-19 was able to accept lactate (2-hydroxypropionate) and exhibited rather broad substrate specificity, it was reasoned that 2HB-CoA can also be used as its substrate. To facilitate the incorporation of 2HB monomer into PHA in recombinant E. coli, 3HB was selected as the second monomer since 3HB-CoA acts as an initiator for lactate containing polymer synthesis (Jung et al. 2010; Taguchi et al. 2008; Yamada et al. 2009; Yang et al. 2010). Among the variants of PhaC1Ps6-19 and PctCp, PhaC1437 and Pct540 were chosen because these enzymes efficiently produce P(3HB-co-LA) copolymers with the highest lactate fraction in recombinant E. coli (Yang et al. 2010). It was thought that this could be also the case in the incorporation of 2HB monomer into poly(2-hydroxybutyrate-co-3-hydroxybutyrate) [(2HB-co-3HB)]. To make P(2HB-co-3HB) without incorporation of lactate monomer, E. coli XLdh strain, in which the E. coli ldhA gene has been completely deleted from the chromosome, was constructed and used as host strain for polymer synthesis. Recombinant E. coli XLdh (p619C1437-pct540) was cultured in MR medium supplemented with 20 g/L of glucose and different concentrations of 2HB and 3HB to examine the synthesis of P(2HB-co-3HB). Recombinant E. coli XLdh expressing PhaC1437 and Pct540 was able to produce PHA random copolymers consisting of 2HB, 3HB, and a small fraction of lactate. As shown in Fig. 2, PHAs with different 2HB monomer fractions, ranging from 10 to 60 mol%, were synthesized depending on the culture condition. The highest 2HB fraction in the copolymer was obtained by adding 2 g/L of 2HB and 0.5 g/L of 3HB to the culture medium, which resulted in the production of P(61 mol%2HB-co-36 mol%3HB-co-3 mol%LA) with the PHA content of 26.7 wt.%. Even though the ldhA gene was completely deleted in the chromosome of E. coli, GC analysis revealed that lactate was incorporated into the polymer up to 6 mol% depending on the culture condition. Monomer compositions of the PHA copolymer could be modulated by changing the concentrations of 2HB and 3HB added to the culture medium. The mole fraction of 2HB in the copolymer was proportional to the concentration of 2HB added to the culture medium. The polymer content was proportional to the total concentration of added 2HB and 3HB, and the highest polymer content of 68 wt.% was obtained when 2 g/L each of 2HB and 3HB were added. Biosynthesis of PHAs consisting of 2HB and a small amount of LA was also examined by the cultivation of recombinant E. coli XLdh (p619C1437-pct540) in MR medium supplemented with 20 g/L of glucose and 2 g/L of 2HB. However, PHA was not produced suggesting that an initiator for polymer synthesis such as 3HB-CoA is still necessary for 2HB containing polymer. Thus, 2HB-CoA is also not a favorable substrate for PHA synthase, as with lactyl-CoA.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-011-3530-x/MediaObjects/253_2011_3530_Fig2_HTML.gif
Fig. 2

Biosynthesis of PHAs consisting of 2HB, 3HB, and a small fraction of lactate in recombinant E. coli XLdh (p619C1437-pct540) by addition of 2HB and 3HB into the culture medium

Construction of metabolic pathways for the production of PHAs containing 2HB monomer from glucose

Recent studies reported the production of higher alcohols via the keto-acid intermediates of native amino-acid pathways in microorganisms (Atsumi and Liao 2008a, b; Atsumi et al. 2008; Shen and Liao 2008). Among the various recombinant pathways for higher alcohols, the 1-propanol biosynthetic pathway using 2-ketobutyrate as a precursor is of particular interest because 2-ketobutyrate is also the possible precursor for 2HB. Between two possible pathways—the threonine degradation pathway and the citramalate pathway—the latter was chosen to synthesize 2-ketobutyrate because it has been suggested to be the most direct route for 2-ketobutyrate (Atsumi and Liao 2008a). The 2-ketobutyrate synthesis pathway was constructed by the expression of the evolved M. jannaschii cimA3.7 gene and E. coli leuBCD genes as previously described (Atsumi and Liao 2008a). To construct the metabolic pathway for the synthesis of copolymer containing 2HB monomer from glucose, 2-ketobutyrate synthesized through the citramalate pathway should be converted into 2HB. There have been reports on the molecular characterization of (d)-2-hydroxyacid dehydrogenases from several microorganisms, which catalyze the enantioselective reduction of branched chain 2-keto acids to (d)-2-hydroxyacids using NADH as a cofactor (Bernard et al. 1994; Chambellon et al. 2009; Kim et al. 2005). (d)-2-Hydroxyacid dehydrogenase has broad specificities to substrates including 2-ketoisocaproate, 2-ketoisovalerate, 2-ketocaproate, 2-ketovalerate, 2-ketobutyrate, and mandelate. Among various (d)-2-hydroxyacid dehydrogenases, two 2-hydroxyacid dehydrogenases encoded by the L. lactis subsp. lactis Il1403 panE gene (Chambellon et al. 2009) and the C. difficile 630 ldhA gene (Kim et al. 2005) were examined for the synthesis of 2HB. However, the expression of the C. difficile ldhA gene that encodes a putative (d)-2-hydroxyacid dehydrogenase, especially 2-ketoisocaproate dehydrogenase in the l-leucine fermentation pathway of C. difficile (Kim et al. 2005), did not allow production of PHAs containing the 2HB monomer in recombinant E. coli. Thus, the conversion of 2-ketobutyrate into 2HB was carried out by expressing the L. lactis Il1403 panE gene encoding the 2HB dehydrogenase. The metabolically engineered E. coli XLdh strain expressing the phaC1437, pct540, cimA3.7, and leuBCD genes together with the L. lactis Il1403 panE gene successfully produced PHAs consisting of 2HB, 3HB, and a small fraction of lactate from glucose and 3HB added to the culture medium (Fig. 3). This result could be justified by the previous study showing that PanE has a substrate specificity toward 2-ketobutyrate by in vitro enzyme activity measurement (Chambellon et al. 2009). To synthesize PHA copolymers containing 2HB and 3HB as monomers using this recombinant E. coli strain, in which 2HB was endogenously generated from glucose, 3HB was added to the culture medium to provide 3HB-CoA. As the 3HB concentration increased, both the 3HB monomer fraction and the polymer content increased (Fig. 3). When this recombinant E. coli XLdh strain was cultured on glucose as a sole carbon source, no PHA was produced, suggesting that 2HB-CoA cannot be efficiently incorporated into PHA (as in the lactyl-CoA case) without an initiator for PHA synthesis such as 3HB-CoA. Finally, we constructed a pathway to generate 3HB-CoA from glucose in E. coli by employing β-ketothiolase and acetoacetyl-CoA reductase encoded by R. eutropha phaAB genes. The metabolically engineered E. coli XLdh strain expressing the phaC1437, pct540, cimA3.7, and leuBCD genes together with the L. lactis Il1403 panE gene and R. eutropha phaAB genes produced P(3 mol%2HB-co-96 mol%3HB-co-1 mol%LA) with the PHA content of up to 74 wt.% from glucose as a sole carbon source.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-011-3530-x/MediaObjects/253_2011_3530_Fig3_HTML.gif
Fig. 3

Biosynthesis of PHAs consisting of 2HB, 3HB, and a small fraction of lactate in recombinant E. coli XLdh (p619C1437-pct540 + pKA32CimALeuBCD + pKM22PanE) by addition of 3HB into the culture medium

Characterization of PHAs containing 2HB and 3HB monomers synthesized in recombinant E. coli

The composition and the sequence distribution of PHA consisting of 22 mol% 2HB, 76 mol% 3HB, and a small fraction of lactate (2 mol%) were investigated with 1D (1H and 13C) NMR and 2D (1H–1H) COSY NMR spectroscopy (Fig. 4a–c). The exact lactate mole fraction in the copolymer could not be resolved in 1H NMR analysis due to the low integral value of lactate and peaks overlap of lactate and 2HB, even though the methyl proton of lactate was observed at 1.6 ppm, and the oxymethine proton of lactate was observed at around 4.9–5.2 ppm. The mole fractions of 2HB and 3HB units in the copolymer obtained from 1H NMR were in good agreement with those from GC analysis except the lactate fraction. The complete assignments of 1H NMR spectra in CDCl3 for P(22 mol%2HB-co-76 mol%3HB-co-2 mol%LA) are shown in Fig. 4a. The methyl protons and methylene protons of 2HB are assigned at the regions of δ0.7–1.2 and 1.6–2.1 ppm, respectively. The methylene protons of 3HB are assigned at the region of δ2.3–2.9 ppm, while the oxymethine protons of 2HB and 3HB are at around δ4.5–5.7 ppm.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-011-3530-x/MediaObjects/253_2011_3530_Fig4_HTML.gif
Fig. 4

NMR analysis (1D 1H, 2D COSY, and 1D 13C) of PHAs consisting of 2HB, 3HB, and a small fraction of lactate synthesized in recombinant E. coli XLdh (p619C1437-pct540). a 500 MHz 1H NMR spectrum, b1H–1H COSY spectrum, and c 125 MHz 13C NMR spectrum with the chemical shift assignments and an expanded spectrum of carbonyl carbons region

Figure 4b shows the COSY spectrum, which highlights the intra-three bond coupling of protons, and as a result, the delicate configurational structure of P(22 mol%2HB-co-76 mol%3HB-co-2 mol%LA). The signal ① between the oxymethine proton of 2HB and the methylene protons of 2HB appears at δ4.9 ppm/2.0 ppm. The coupling ② between the methylene protons of 2HB and the methyl protons of 2HB can be assigned as a cross peak at δ1.8 ppm/1.0 ppm. The coupling ③ between the oxymethine proton of 3HB and the methylene protons of 3HB can be seen as a cross peak at δ5.4 ppm/2.6 ppm, while the coupling ④ between oxymethine proton of 3HB and the remaining methyl protons of 3HB is seen at δ5.2 ppm/1.3 ppm.

Figure 4c shows the 125-MHz 13C-NMR spectrum of P(22 mol%2HB-co-76 mol%3HB-co-2 mol%LA) with the chemical shift assignments and an expanded spectrum of carbonyl carbons region. This region (168.0–170.1 ppm) is clearly resolved into three groups of peaks showing different diad bonding sequences of 2HB and 3HB. The peak at 168.7 ppm is assigned to carbonyl carbons in the 2HB*–2HB sequence. The peak at 169.1 ppm is assigned to the carbonyl resonance in the 3HB*–3HB and 3HB–3HB* sequences. The peak at 169.5 ppm is assigned to the 3HB*–2HB + 2HB–3HB* sequences. The diad-sequence distribution of 2HB and 3HB units was determined from the peak areas of carbonyl resonances (Table 3). The 2HB–3HB diad-sequence data were compared with the Bernoullian statistics applicable to a statistically random copolymerization.
Table 3

Diad sequence distributions of P(23 mol%2HB-co-77 mol%3HB)

Relative intensity

Diad sequence distribution

3HB–3HB

(3HB–2HB + 2HB–3HB)

2HB–2HB

Observed

57

37

6

Calculated

60

35

5

DSC and GPC studies on P(47 mol%2HB-co-50 mol%3HB-co-3 mol%LA) showed that amorphous polymer was synthesized by recombinant E. coli. The molecular weights of P(47 mol%2HB-co-50 mol%3HB-co-3 mol%LA) were 20,000 (Mn) and 33,800 (Mw) with the polydispersity index of 1.69. The glass transition temperature was 9.7°C, but the melting temperature and the enthalpy of fusion of P(47 mol%2HB-co-50 mol%3HB-co-3 mol%LA) were not detected.

Discussion

PLA is a carbon-neutral alternative to petroleum-based plastic because it can be produced from renewable biomass. PLA has the potential in a wide range of environmentally friendly applications due to its biodegradable, biocompatible, compostable properties. To improve the material properties of PLA, copolymerization with a range of other monomers including glycolide, ε-caprolactone, δ-valerolactone, and trimethylene carbonate has been employed (Rasal et al. 2010). Also, chemical syntheses of poly(phenyllactate) and polymandelate, which contain aromatic rings as side-chain residues, were performed to provide novel PLA analogs with improved material properties (Liu et al. 2007; Simmons and Baker 2001). Recently, poly(2-hydroxybutyrate)[P(2HB)], the analog of PLA in which the methyl group of PLA is substituted with an ethyl group, was used for the generation of P[(S)-2HB]/P[(R)-2HB] stereocomplex to enhance thermal and material properties (Tsuji and Okumura 2009). Therefore, development of a microbial system for the production of PHA containing 2HB as a monomer should be useful for the synthesis of a polymer with novel material property.

As previously reported, efficient production of PHAs can be achieved by enhancing the metabolic capacity of host strains to supply enough precursors for PHA synthesis along with the expression of PHA synthase having high substrate specificity towards generated substrates (Fukui et al. 1999; Park et al. 2005). Activation of hydroxyacid into hydroxyacyl-CoA by enzymes such as CoA synthetase and CoA transferase has been employed for the production of PHA from structure-related carbon sources added into culture medium. PHAs containing lactate, 4HB and 3-hydroxypropionate, and 3-mercaptopropionate as a monomer constituent have successfully been synthesized by this strategy (Hein et al. 1997; Liu and Steinbüchel 2000; Lütke-Eversloh et al. 2002; Valentin et al 2000; Yang et al 2011). At first, recombinant E. coli strains expressing PhaC1437 and Pct540 were cultured in the medium containing 2HB following this strategy to provide enough precursors for 2HB-CoA. 3HB was also added to the medium to facilitate the incorporation of 2HB monomer into PHA as an initiator for 2HB containing polymer synthesis. As shown in Fig. 2, PHAs having different 2HB monomer compositions were synthesized depending on the concentration of 2HB and 3HB added to the medium. 2HB monomer fractions in PHA copolymer were increased up to 60 mol% by enhancing the metabolic capacity of E. coli strain to provide more 2HB-CoA rather than 3HB-CoA, which was modulated by increasing the concentration of 2HB and reducing the concentration of 3HB added to the culture medium. Because PhaC1437 can also accept lactyl-CoA as a substrate, an E. coli mutant strain in which the ldhA gene is completely deleted in the chromosome was used as a host strain to prevent lactate incorporation into PHA copolymer. However, lactate was incorporated up to 6 mol% in the PHA copolymer depending on the culture condition. This seemed to be due to the synthesis of lactate by the remaining lactate dehydrogenase activities present in E. coli, despite the complete deletion of the ldhA gene as shown previously (Hong and Lee 2001; Zhang et al. 2009).

It is important to supply precursors for PHA monomers from renewable resources to achieve economical production of PHAs having different monomer composition. Thus, metabolic pathways of E. coli were engineered to provide 2HB monomer from glucose for the efficient production of PHAs containing 2HB monomer. In this study, we employed the citramalate pathway, the most direct route for the 2-ketobutyrate precursor of 2HB, by the expression of the evolved M. jannaschii cimA3.7 gene and E. coli leuBCD genes as previously described (Atsumi and Liao 2008a). Conversion of 2-ketobutyrate into 2HB is mediated by (d)-2-hydroxyacid dehydrogenase of the L. lactis subsp. lactis Il1403 (Chambellon et al. 2009). The monomer composition of PHA copolymer was also dependent on the ratio of the precursor concentrations for the 2HB-CoA and 3HB-CoA. As the 3HB concentration in the culture medium increased, the 2HB monomer fraction decreased (Fig. 3), similar to that observed during P(3HB-co-LA) production studies we previously reported ( Jung et al. 2010; Yang et al. 2010). It was also found that expression of the R. eutropha phaAB genes to supply 3HB-CoA from glucose significantly reduced the 2HB monomer fraction in the copolymer, even though a high PHA content of 74 wt.% was achieved. The significant reduction of 2HB monomer fraction in PHA copolymer was likely due to the fact that expression of R. eutropha phaAB genes supplying 3HB-CoA from acetyl-CoA accelerated incorporation of 3HB-CoA, the most preferred substrate of PHA synthase, into PHAs; this is due to that the metabolic pathways for supplying 2HB, 3HB, and lactate monomer compete for acetyl-CoAs, as shown in Fig. 1. The relatively lower PHA content when 2HB was the major monomer suggests that strategies for supplying more 2HB monomers from glucose and improving the substrate specificity towards 2HB-CoA should be developed to further enhance the production of PHA-containing 2HB monomer. For example, Atsumi et al. (2008a) reported that removal of IlvI and IlvB, which compete for 2-ketobutyrate, resulted in increased production of 1-propanol and 1-butanol in recombinant E. coli. These strategies will be examined in our future studies.

Very recently, chemo-enzymatic synthesis of 2HB-containing PHA by R. eutropha PHA synthase was reported (Han et al. 2011). Because PhaC1437 used in this study has the ability to accept lactyl-CoA as a substrate, PHA copolymer consisting of only 2HB and 3HB could not be synthesized in recombinant E. coli. Based on this aforementioned study, application of R. eutropha PHA synthase in our system should be useful for the production of P(2HB-co-3HB) with a very minute fraction of lactate monomer for the characterization of polymer properties because R. eutropha PHA synthase cannot efficiently accept lactyl-CoA as substrate compared with the type II PHA synthase mutants from Pseudomonas sp. (Taguchi et al. 2008; Yang et al. 2010, 2011).

Presently, we report for the first time the biosynthesis of PHAs containing 2HB monomer from glucose by metabolically engineered E. coli strains. PHA copolymers of 2HB and 3HB could be produced by feeding 3HB, which can be easily generated by the β-ketothiolase and reductase pathways. With the increase of the fraction of 3HB, the most favorable substrate for PHA synthase, the PHA content also increased, reaching a comparable level to that obtained by recombinant E. coli producing P(3HB) homopolymer. The strategies described here should be useful for the development of new strategies for the production of various PHAs consisting of other 2-hydroxyacids by direct microbial fermentation from renewable resources.

Acknowledgements

This work was supported by the Korean Systems Biology Research Project (20090065571) of the Ministry of Education, Science and Technology (MEST) through the National Research Foundation (NRF) of Korea. Further support by World Class University program (R32-2008-000-10142-0) and by the Advanced Biomass R&D Center of Korea (ABC-2010-0029799) through the Global Frontier Research Program of MEST is appreciated. B.K.S, S.H.L, and S.J.P appreciate the financial supports from the R&D Program of MKE/KEIT (10032001) and KRICT.

Copyright information

© Springer-Verlag 2011