Introduction

In recent years, biocatalysis has been attracting much attention as an environment-friendly method for the production of enantiomerically pure building blocks1,2,3. Both whole cells and the related isolated enzymes have been used in biocatalytic production of chiral chemicals, which have emerged as important pharmaceutical and agrochemical intermediates. In comparison with the isolated enzymes, whole-cell catalysts can be prepared readily and inexpensively and thus offer economical and environmental advantages over isolated enzymes3. However, some problems, such as the unpredictability of reactions, the utilization of opportunistic pathogen and the side reactions causing degradation of products, prevent the large-scale use of whole-cell catalysts3,4.

Pseudomonas putida is one of the best studied species of the metabolically versatile, genetic plastic and ubiquitous genus of Pseudomonads5. The typical strain of P. putida, KT2440 (ATCC47054), is a microorganism Generally Recognized as Safe (GRAS certified) and thus has been extensively used in a wide range of biotechnological applications including bioremediation of contaminated areas, production of bioplastics, biocatalytic production of chiral chemicals and so on5,6. Since the complete sequence of P. putida KT2440's genome has been released7, the knowledge related to the strain has significantly increased and various “-omics” data sets such as transcriptomic8,9,10, proteomic10,11 and fluxomic data12,13 have also become available. Genome-scale reconstruction and analysis of the P. putida KT2440 metabolic network based on those data have facilitated the applications of the strain in biocatalysis6.

Lactate has 2 optical isomers: l-lactate and d-lactate. Compared with d-lactate, l-lactate is an organic acid with more versatile applications in food, pharmaceutical, textile and chemical industries. More importantly, l-lactate is also an indispensable monomer for the synthesis of poly-l-lactic acid, a bio-degradable polymer. Many Pseudomonas strains, such as P. aeruginosa, P. putida and P. stutzeri, can use l-lactate and/or d-lactate as the sole carbon and energy source for growth14,15,16. The lactate utilization system involves three membrane bound proteins: LldP (a lactate permease), NAD-independent l-lactate dehydrogenase (l-iLDH) and NAD-independent d-lactate dehydrogenase (d-iLDH). As a lactate permease, LldP takes up l-lactate and/or d-lactate into the cells17,18. l-iLDH and d-iLDH catalyze the oxidation of l-lactate and d-lactate to pyruvate via a flavin-dependent mechanism, respectively18. l-iLDH and d-iLDH are induced coordinately in all of the reported Pseudomonas strains including P. putida. The two enzymes are not constitutively expressed; the enantiomer of lactate can induce the expression of both enzymes18. Thus, in addition that P. putida KT2440 is regarded as the most promising catalyst for the production of l-2-hydroxy-carboxylates such as l-lactate through d-iLDH catalyzed kinetic resolution, the co-present l-iLDH in the strain would also catalyze the oxidation of l-2-hydroxy-carboxylates and produce 2-oxo-carboxylates as the end product.

In this study, the lactate utilization system in the GRAS P. putida strain KT2440 was analyzed and further reconstructed (Fig. 1). A recombinant strain of P. putida KT2440 was obtained through disruption of l-iLDH encoding gene lldD. Whole cells of the recombinant strain were confirmed to exhibit only d-iLDH activity (catalyzing the oxidation of d-2-hydroxy-carboxylates). Two chiral 2-hydroxy-carboxylates: l-lactate and l-2-hydroxybutyrate were then produced from their racemic mixtures by using the recombinant P. putida KT2440 as the catalyst.

Figure 1
figure 1

Scheme of l-2-hydroxy-carboxylate production by the reconstructed lactate utilization system in P. putida KTM.

The membrane bound LldP takes up l-2-hydroxy-carboxylate and d-2-hydroxy-carboxylate into the cells. The membrane bound l-iLDH and d-iLDH catalyze the oxidation of l-2-hydroxy-carboxylate and d-2-hydroxy-carboxylate to 2-oxo-carboxylates, respectively. Red cross indicates that the l-iLDH catalyzed oxidation of l-2-hydroxy-carboxylate was eliminated. R: CH3, CH2CH3.

Results

Identification of the lactate utilization operon in P. putida KT2440

In a previous study, the lldPDE operon, which encodes lactate permease (LldP), l-iLDH and d-iLDH, was studied in P. aeruginosa XMG18. Comparative genomics revealed that this operon was also found in most Pseudomonas strains including P. putida KT24407. A BLAST search of the genome sequence of P. putida KT2440 with the lldPDE operon of P. aeruginosa XMG as the probe revealed a sequence encoding three proteins with strikingly high homology to LldP (84% sequence identity), l-iLDH (88% sequence identity) and d-iLDH (82% sequence identity). As shown in Fig. 2A, lldD (encoding l-iLDH), lldP (encoding lactate permease) and lldE (encoding d-iLDH) are located adjacent to the regulator LldR encoding gene lldR. Since the l-iLDH- and d-iLDH-encoding genes are in the same lactate utilization operon and are controlled by the same regulator, reconstruction of the lactate utilization operon lldPDE in P. putida KT2440 thus become a feasible method for utilization of the GRAS strain in l-2-hydroxy-carboxylates production.

Figure 2
figure 2

Construction of P.putida KTM.

(A) Organizations of lactate utilization operon in different Pseudomonas species including P. aeruginosa XMG40, P. aeruginosa PAO141, P. putida KT24407, P. putida S1642, P. stutzeri SDM43 and Pseudomonas entomophila L4844. Genes were functionally annotated following the color code indicated. The numbers below the genes indicate the percentage similarities to the P. putida KT2440 protein sequences. The percentages represent amino acid homology of related enzymes. (B) Diagram illustrating the sequence analysis of the disruption of the lldD. (C) Analysis of PCR fragments to confirm lldD disruption. Lane M: molecular mass standard (λDNA/HindIII); Lane 1: product amplified with water as the template (negative control). Lane 2: product amplified with P. putida KT2440 genomic DNA as the template; Lane 3: product amplified with P. putida KTM genomic DNA as the template. The PCRs were performed with primers LkF and LkR. (D) Results of the RT-PCR assay in detecting the transcription of lldD. (E) Results of the RT-PCR assay in detecting the transcription of lldP. (F) Results of the RT-PCR assay in detecting the transcription of lldE. Lane 1, control reaction in which genomic DNA of P. putida KTM was used as the template; lane 2, reaction in which cDNA of P. putida KTM was used as the template; lane 3, control reaction in which mRNA of P. putida KTM was used as the template; lane 4, control reaction in which ddH2O was used as the template. Lane M: molecular mass standard (λDNA/HindIII). Numbers on the left represent the sizes of the markers (in bp).

Reconstruction of the lactate utilization system in P. putida KT2440

To eliminate the oxidation of l-2-hydroxy-carboxylates by l-iLDH, the lldD gene in P. putida KT2440 was disrupted using homologous recombination (Fig. 2B). Suicide plasmid pK18mobsacB was utilized as described previously19. PCR was used to verify the disruption event of gene lldD using primer set LkF plus LkR. The mutant with a second crossover event (named P. putida KTM) was selected. The result in Fig. 2C shows that the PCR using the primer set generated products of the expected sizes.

The transcription of lldP, lldD and lldE in P. putida KTM was also monitored by reverse transcription (RT)-PCR assays. RT-PCR fragments of the lldP and lldE were obtained from cells of P. putida KTM grown in MSM containing 5 g 1−1dl-lactate (Fig. 2E and Fig. 2F), thus indicating that the disruption of lldD in the P. putida KT2440 (Fig. 2D) would not influence the transcription of lldP and lldE (Table S1). LldP and d-iLDH still work in the transport and oxidation of d-2-hydroxy-carboxylates, respectively. With these considerations in mind, whole cells of P.putida KTM with reconstructed lactate utilization system could be used in the biocatalytic production of l-2-hydroxy-carboxylates (Fig. 1).

Utilization of dl-lactate by P.putida KT2440 and P.putida KTM

P.putida KT2440 and P. putida KTM were cultured with dl-lactate as the sole carbon source. As shown in Fig. 3A, both d-lactate and l-lactate in the medium would be utilized by P.putida KT2440. Since l-iLDH is required for l-lactate utilization in Pseudomonas, P. putida KTM with the mutant of lldD could not use l-lactate as the sole carbon source (Fig. S1). However, P. putida KTM still use d-lactate in the medium for growth (Fig. 3B).

Figure 3
figure 3

Time-course study of P.putida KT2440 (A) and P.putida KTM (B) growth in medium with dl-lactate.

Black solid circle, dry cell weight; blue up triangle, concentration of d-lactate; red down triangle, concentration of l-lactate. The initial dl-lactate concentration is 5 g l−1. The performance was analyzed at 0, 1.5, 3, 4.5, 6, 7.5 and 9 h. Results are means ± SD of three parallel replicates.

P. putida KTM was complemented with a broad-host-range plasmid pBBR1MCS-5 harboring the lldD gene. The resulting complementary strain was designated as P. putida KTM (pBBR-lldD). The activities of l-iLDH and d-iLDH in crude extract of P.putida KT2440, P. putida KTM and P. putida KTM (pBBR-lldD) were also assayed. As shown in Table 1, when cultured in the medium containing dl-lactate, both l-iLDH and d-iLDH activities were detected in P.putida KT2440 and P. putida KTM (pBBR-lldD). However, P. putida KTM exhibited only d-iLDH activity when cultured with dl-lactate as the sole carbon source.

Table 1 Activities of l-iLDH and d-iLDH in crude extract of P.putida KT2440, P. putida KTM and P. putida KTM (pBBR-lldD)a

Substrate specificity of d-iLDH in P.putida KTM

Since P. putida KTM could use d-lactate in racemic lactate and left l-lactate in the medium (Fig. 3B), it might have potential for the production of other l-2-hydroxy-carboxylates. Substrate specificity of d-iLDH in P.putida KTM was then examined with different d-2-hydroxy-carboxylates as the test substrates. As shown in Table 2, only d-lactate and d-2-hydroxybutyrate could be oxidized by d-iLDH in P.putida KTM. d-iLDH from P. stutzeri has been reported to show similar narrow substrate specificity20. Only d-lactate and d-2-hydroxybutyrate were attacked by the enzyme. Similar results have also been observed in other microorganisms including E. coli21.

Table 2 Substrate specificity of d-iLDH from P.putida KTM

The rate of dehydrogenation of d-lactate and d-2-hydroxybutyrate catalyzed by d-iLDH in P.putida KTM followed Michaelis-Menten kinetics. Double-reciprocal plots of the initial rates plotted against the concentrations of d-lactate and d-2-hydroxybutyrate were linear at a fixed concentration of dichlorophenol-indophenol (DCPIP) (0.2 mM) and yielded Km values of 0.16 ± 0.03 mM and 0.44 ± 0.03 mM, respectively. Vmax was estimated to be 101.7 ± 18.2 nmol min−1 mg−1 for d-lactate and 96.7 ± 6.3 nmol min−1 mg−1 for d-2-hydroxybutyrate with DCPIP as the electron acceptor.

Production of l-lactate and l-2-hydroxybutyrate using whole cells of P.putida KTM

Although d-iLDH in P.putida KTM seemed to have narrow substrate specificity, it could attack the two most important members of 2-hydroxy-carboxylates, lactate and 2-hydroxybutyrate. The biocatalytic oxidation of the racemic lactate (100 mM) and 2-hydroxybutyrate (100 mM) with whole cells of P. putida KTM was carried out in distilled water in the presence of 20 mM ethylenediaminetetraacetic acid (EDTA). As shown in Fig. 4, the d-isomers in racemic lactate (100 mM) and 2-hydroxybutyrate (100 mM) could be oxidized into pyruvate and 2-oxobutyrate in 6 h. The final concentrations of l-lactate and l-2-hydroxybutyrate were 49.5 mM and 49.9 mM, respectively. These results revealed that the biocatalytic production of l-lactate and l-2-hydroxybutyrate proceeds in a high degree of enantioselectivity by preferential oxidation of d enantiomers in racemic mixtures with whole cells of P. putida KTM.

Figure 4
figure 4

Kinetic resolutions of racemic lactate (A) and 2-hydroxybutyrate (B) by whole cells of P.putida KTM.

Red down triangle, concentrations of l-2-hydroxy-carboxylates; blue up triangle, concentrations of d-2-hydroxy-carboxylates. The initial dl-2-hydroxy-carboxylate concentration is 100 mM. The performance was analyzed at 0, 2, 4, 6 and 7 h. Results are means ± SD of three parallel replicates.

Discussion

Optically active 2-hydroxy-carboxylates are important building blocks for glycols, halo esters and epoxides compounds, which are important intermediates of pharmaceuticals22,23. Chemical processes for 2-hydroxy-carboxylates production result in a racemic mixture of both stereospecific forms. Many routes, such as high-performance ligand exchange chromatography24, enzymatic resolution25 and asymmetric hydrolysis26, have been developed for the resolution of the racemic 2-hydroxy-carboxylates. Due to its excellent stereo-selectivity, high product yield and environmental friendly process, biocatalytic oxidative resolution of the racemate has emerged as a desirable technique for the production of optically active 2-hydroxy-carboxylates22,25. Until now, all of the reported 2-hydroxy-carboxylates resolution processes utilized NAD-independent l-2-hydroxy-carboxylate dehydrogenases such as glycolate oxidase25, l-lactate oxidase27 and membrane bound l-iLDH28,29 as the biocatalysts and could only produce d-2-hydroxy-carboxylates. Thus, there is a demand for searching of biocatalysts that could be utilized in the resolution of racemic 2-hydroxy-carboxylates to produce l-2-hydroxy-carboxylates.

Although d-iLDH might catalyze the oxidation of d-2-hydroxy-carboxylates, the co-present l-iLDH would also catalyze the l-2-hydroxy-carboxylates into 2-oxo-carboxylates30,31,32. After analysis of the lactate utilization system through comparative genomics, the lldPDE operon was identified in P. putida KT2440. l-iLDH encoding gene lldD was disrupted using homologous recombination. As we expect, the resulting recombinant strain P. putida KTM exhibited only d-iLDH activity. The specific activity of d-iLDH in P. putida KTM was 65.0 ± 3.7 nmol min−1 mg−1, which was higher than that of P. aeruginosa XMG (20 ± 4 nmol min−1 mg-1)32 but lower than that of P. stutzeri SDM (132 ± 5 nmol min−1 mg−1)20. Similar to d-iLDH in other Pseudomonas strains such as P. stutzeri SDM20, d-iLDH in P. putida KTM also exhibited narrow substrate specificity and only catalyzed the oxidation of d-lactate and d-2-hydroxybutyrate. Optically active l-lactate and l-2-hydroxybutyrate could be used in production of polylactate and poly(2-hydroxybutyrate), which can be utilized as a biodegradable material for biomedical, pharmaceutical and environmental applications33,34. Then, whole-cells of P. putida KTM were used to catalyze the resolution of racemic 2-hydroxybutyrate and lactate. The d-lactate and d-2-hydroxybutyrate could be oxidized into pyruvate and 2-oxobutyrate, respectively. l-Lactate and l-2-hydroxybutyrate, which were not oxidized in the biocatalytic process, accumulated with enantiomeric excess higher than of 99% (Fig. S2 and Fig. S3).

In a previous study, l-iLDH in P. stutzeri SDM was rationally re-designed on the basis of sequence alignment and the active site structure of a homologous enzyme; a new biocatalyst with high catalytic efficiency toward an unnatural substrate (l-mandelate) was successfully constructed35. Although the crystal structure of d-iLDH in E. coli was described36, the position of the active site is still unknown. On the other hand, there is only 28% sequence identity between d-iLDHs from P. putida KT2440 and E. coli. Thus, at this stage, we did not attempt to expand the application range of d-iLDH in P. putida KTM due to the difficulty in the rational re-design of this membrane bound enzyme. However, if its structure is clarified, reconstruction of d-iLDH to improve its activity towards other straight long aliphatic or aromatic 2-hydroxy-carboxylic acids and synthesis of other valuable l-2-hydroxy-carboxylates might be successful in the future.

In summary, racemic 2-hydroxy carboxylic acids were first utilized in the production of l-2-hydroxy-carboxylates through enantioselective oxidation. A novel catalyst with lactate utilization system in P. putida KT2440 was reconstructed; it exhibited high bio-catalytic activities for production of l-2-hydroxy-carboxylates. The reconstruction of the lactate utilization system in P. putida KT2440 resulted in a novel catalyst that has exhibited high biocatalytic activities for l-2-hydroxy-carboxylates production. This study is a good example for the application of the GRAS P. putida KT2440 in the bio-catalysis through genomic analysis and recombination. Other applications of P. putida KT2440 would also be possible through further understanding and reconstruction of this biotechnologically important organism.

Methods

Chemicals and biochemicals

l-Lactate (L7022), dl-lactate (71720), d-3-phenyllactate (376906), d-2-hydroxybutyrate (54917), dl-2-hydroxyisovaleric acid (219835), dl-2-hydroxyisocaproic acid (219819), dl-2-hydroxyoctanoic acid (H7396), DCPIP (33125), dithiothreitol (DTT, 43815), bovine serum albumin (BSA, B8667) and phenylmethanesulfonyl fluoride (PMSF, P7626) were purchased from Sigma (Sigma-Aldrich China Inc. Shanghai, China). d-Lactate (71716) was purchased from Fluka (Sigma-Aldrich China Inc. Shanghai, China). dl-2-Hydroxybutyrate (H0387) and dl-glycerate (G0232) were purchased from TCI (Tokyo, Japan). The restriction enzymes were obtained from Fermentas Bio Inc. (Beijing, China). FastPfu DNA polymerase was purchased from TransGen Biotech (Beijing, China). T4 DNA ligase was obtained from New England Biolabs (Beijing, China). All of the biochemicals mentioned above were of analytical grade reagents and are commercially available.

Bacterial strains and culture conditions

Bacterial strains, plasmids and oligonucleotide primers used in this study were listed in Table 3. P. putida KT2440 was cultured in 500-ml baffled shake flasks containing 50 ml minimal salt medium (MSM) at 30°C and 120 rpm for 9 h18. MSM was supplemented with 5.0 g l−1dl-lactate or pyruvate as the sole carbon source. E. coli DH5α was cultured in 500-ml baffled shake flasks containing 50 ml Lysogeny broth (LB) and maintained at 37°C and 120 rpm for 12 h. Isolation of vectors, restriction enzyme digestion, agarose gel electrophoresis and other DNA manipulations were carried out according to the standard methods37.

Table 3 Strains, plasmids and oligonucleotide primers used in this study

Gene knockout procedure

Genomic DNA of P. putida KT2440 was extracted through the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA). A fragment from P. putida KT2440 genome containing whole length of lldD was amplified by PCR using LkF and LkR as the primers. The PCR reaction mixture contained 2.5 μl of each primer at 10 μM, 5.0 μl of 10 × PCR buffer, 2.0 μl of 25 mM MgCl2, 2.5 μl of 2 mM dNTPs mix, 2 μl of enzyme solution containing 1 U FastPfu DNA polymerase in 33 μl of ddH2O. Genomic DNA of P. putida KT2440 (0.5 μl at a concentration of 0.5 g l−1) was used as the template. Thirty amplification cycles were performed on a Mastercycler® pro PCR thermal cycler (Eppendorf, Germany), each consisting of a denaturation of 30 s at 94°C, an annealing of 30 s at 50°C and an extension of 60 s at 72°C. Final extension was performed at 72°C for 5 min. The PCR product was digested by EcoRI and HindIII and then inserted in pK18mobsacB to form a new plasmid pKLK0119. Plasmid pKLK01 was completely digested by AvaI and then the large fragment was self-ligated to form pKLK02. As a result, only partial lengths of lldD were inserted into pK18mobsacB in pKLK02. Plasmid pKLK02 was transformed into P. putida KT2440 by electroporation. Integration of the plasmid pKLK02 into the chromosome of P. putida KT2440 by the first crossover was selected on LB plate supplemented with 50 μg ml−1 kanamycin. The second crossover cells were selected by culture on LB plates containing 10% (w/v) sucrose. The resulting deletion mutant was designated as P. putida KTM.

All the constructed strains were validated by PCR and DNA sequencing.

Complementation of P. putida KTM

A fragment containing whole length of lldD of P. stutzeri SDM was amplified by PCR using the primers LcF and LcR, digested with HindIII and PstI and then inserted in pBBR1MCS-5 to form a new plasmid pBBR-lldD. Plasmid pBBR-lldD was transformed into P. putida KTM by electroporation. The resulting complementary strain was designated as P. putida KTM (pBBR-lldD).

RT-PCR of lldP, lldD and lldE

Cells of P.putida KT2440 and P. putida KTM grown in MSM containing 5 g 1−1dl-lactate were harvested when they reached the mid-exponential phase. Total mRNA was prepared using a Qiagen RNeasy total RNA kit. DNA contamination was eliminated by a DNase I treatment at 37°C for 1 h. RNA integrity was checked by agarose gel electrophoresis. After quantification, 0.2 mg of purified total RNA was used to prepare cDNA by using Superscript II RT (Invitrogen) and reagents supplied by the manufacturer (42°C, 50 min). PCR were carried out using FastPfu DNA polymerase (TransGen Biotech, Beijing, China). Primer sequences for the lldP, lldD and lldE are listed in Table S1. Samples were first denatured by heating at 94°C for 3 min. Thirty amplification cycles were performed on a Mastercycler® pro PCR thermal cycler (Eppendorf, Germany), each consisting of a denaturation of 30 s at 94°C, an annealing of 30 s at 50°C and an extension of 60 s at 72°C. Final extension was performed at 72°C for 5 min. Control reactions in which mRNA was used as the templates ensured that DNA products resulted from the amplification of cDNA rather than from DNA contamination. PCR products were separated by electrophoresis at 80V on 1.3% Tris Borate EDTA agarose gel, followed by ethidium bromide staining.

Preparation of crude extracts of P.putida KT2440, P. putida KTM and P. putida KTM (pBBR-lldD)

P.putida KT2440, P. putida KTM and P. putida KTM (pBBR-lldD) were maintained on LB agar slants. The slants were incubated at 30°C for 12 h and then stored at 4°C. Cells of P.putida KT2440, P. putida KTM and P. putida KTM (pBBR-lldD) from freshly prepared slants were inoculated into 300-ml baffled shake flasks containing 50 ml of LB. The cultivation was conducted at 30°C and 120 rpm for 9 h. Then cells of P.putida KT2440, P. putida KTM and P. putida KTM (pBBR-lldD) were collected by centrifugation at 4500 rpm for 10 min, washed twice with MSM, then diluted (at approximately OD620 0.05) in MSM containing 5.0 g 1−1dl-lactate as the sole carbon source. The cultivation was conducted in 500-ml baffled shake flasks at 30°C and 120 rpm for 9 h. Cells of P.putida KT2440, P. putida KTM and P. putida KTM (pBBR-lldD) were harvested from 500 ml MSM, washed, resuspended in 50 ml of 50 mM Tris–HCl (pH 8.0) containing 10 mM DTT, 20 mM KCl, 5 mM MgSO4 and 1 mM PMSF and disrupted by sonication (Sonics 500 W/20 KHz, USA) in an ice bath. The disrupted cells were centrifuged for 10 min at 10,000 × g and the supernatants were used as the crude cell extracts. Proteins were determined by the Markwell variation of the Lowry method, with BSA as standard38.

Enzymatic activity assays

The activities of iLDHs were determined at 30°C in 1 ml of 50 mM Tris–HCl, pH 7.5, 0.2 mM DCPIP and 50 μl of crude cell extract (0.2 mg ml−1). The reaction was started by addition of 20 mM l- or 20 mM d-lactate and the rate of DCPIP reduction was determined by measuring the absorbance changes at 600 nm using a UV/visible spectrophotometer (Ultrospec 2100 pro, Amersham Biosciences, USA)39.

Kinetic resolution by whole cells of P.putida KTM

Optical density of P. putida KTM was assayed at 620 nm by a 2100 spectrophotometer (Shanghai Exact Scientific Apparatus CO., LTD, China) and was converted to dry cell weight using the following equations: P.putida KTM, dry cell weight (g l−1) = 0.35 × OD620nm. The reaction was carried out at 30°C and 120 rpm in distilled water (pH 7.4) containing 6 g dry cell weight l−1 of P.putida KTM, 20 mM EDTA and 100 mM racemic 2-hydroxy-carboxylates. The concentrations of 2-hydroxy-carboxylates and 2-oxo-carboxylates in the reaction mixtures were quantitatively analyzed by high-performance liquid chromatography (HPLC).

HPLC analyses

Concentrations of 2-hydroxy-carboxylates and 2-oxo-carboxylates were measured by HPLC (Agilent 1100 series, Hewlett-Packard, USA) using an Aminex HPX-87H column (Bio-Rad) and the eluent (10 mM H2SO4) at 1 ml min−1 flow-rate. Stereoselective assays of 2-hydroxy-carboxylates were performed by HPLC equipped with a chiral column (MCI GEL CRS10W, Japan) and the eluent at 1 ml min−1 flow-rate. The mobile phase was 2 mM CuSO4 for lactate and a mixture of water and acetonitrile (90:10), containing 2 mM copper sulphate for 2-hydroxybutyrate. The enantiomeric excess (ee) value of 2-hydroxy-carboxylates was defined as follows .