Effects of Halophilic Peptide Fusion on Solubility, Stability, and Catalytic Performance of $\small{D}$-Phenylglycine Aminotransferase

Abstract

$\small{D}$-Phenylglycine aminotransferase ($\small{D}$-PhgAT) from Pseudomonas stutzeri ST-201 is useful for enzymatic synthesis of enantiomerically pure $\small{D}$-phenylglycine. However, its low protein solubility prevents its application at high substrate concentration. With an aim to increase the protein solubility, the N-terminus of $\small{D}$-PhgAT was genetically fused with short peptides ($A_1$ ${\alpha}$-helix, $A_2$ ${\alpha}$-helix, and ALAL, which is a hybrid of $A_1$ and $A_2$) from a ferredoxin enzyme of a halophilic archaeon, Halobacterium salinarum. The fused enzymes $A_1$-$\small{D}$-PhgAT, $A_2$-$\small{D}$-PhgAT, and ALAL-$\small{D}$-PhgAT displayed a reduced pI and increased in solubility by 6.1-, 5.3-, and 8.1- fold in TEMP (pH 7.6) storage, respectively, and 5-, 4.5-, and 5.9-fold in CAPSO (pH 9.5) reaction buffers, respectively, compared with the wild-type enzyme (WT-$\small{D}$-PhgAT). In addition, all the fused $\small{D}$-PhgAT displayed higher enzymatic reaction rates than the WT-DPhgAT at all concentrations of L-glutamate monosodium salt used. The highest rate, $23.82{\pm}1.47$ mM/h, was that obtained from having ALAL-$\small{D}$-PhgAT reacted with 1,500 mM of the substrate. Moreover, the halophilic fusion significantly increased the tolerance of $\small{D}$-PhgAT in the presence of NaCl and KCl, being slightly in favor of KCl, where under the same condition at 3.5 M NaCl or KCl all halophilic-fused variants showed higher activity than WT-$\small{D}$-PhgAT.

Introduction

The purification and characterization of D-phenylglycine aminotransferase (D-PhgAT; E.C. 2.6.1.72), a stereo-inverting pyridoxal-5’-phosphate (PLP)-dependent enzyme from Pseudomonas stutzeri ST-201 with the ability to catalyze the reversible transamination of D-phenylglycine (D-Phg) with α-ketoglutaric acid (Fig. 1), was first reported in 1997 [24]. Crystallography and 3D studies revealed that D-PhgAT is a homodimer enzyme, where each monomer possesses three domains namely a C-terminus, an N-terminus, and a cofactor binding domain with 453 amino acid residues and 47.5 kDa molecular mass [11]. Owing to its stereo-inverting property and the ability to utilize L-glutamic acid as a cheap amino group donor, D-PhgAT is an invaluable catalyst in the single-step synthesis of D-Phg and D-4-hydroxyphenylglycine (D-4-OHPhg), which are essential side-chain moieties for manufacturing semisynthetic antibiotics in the penicillin and cephalosporin families. Currently, a two-step enzyme reaction employing hydantoinase and carbamoylase enzymes is used to produce these side-chain moieties; however, owing to the low solubility of the substrates, the reaction rate was low [13]. Another approach is to employ D-amino-acid aminotransferases for these purposes, but again owing to the low activity and high specificity of the enzymes toward only D-amino acids which are costly, their use was quite limited. Some researchers solved the problem of requiring only D-amino acids by incorporating an amino acid racemase and L-amino acids in the reaction to provide the required D-amino acids, but this may baffle the aminotransferase activity [9]. Therefore, up to now, there was no effective enzymatic reaction for the production of these side-chain compounds.

Fig. 1.Stereo-inverting and reversible reaction of D-PhgAT, which can be used for synthesis of D-Phg from benzoylformic acid and L-glutamate.

It is known that most enzymes are susceptible to inactivation, denaturation, and precipitation during production, storage, and application in industries [8]. Some may have an additional problem of low solubility, and that also happened with D-PhgAT in our study. The low in vitro solubility of D-PhgAT was partially solved via introduction of double site-directed mutations (N439D and Q444E), such that the solubility increased up to 5.9-fold [3]; however, it is still precipitated if used in highly concentrated solution or in the presence of high salts, which are occasionally unavoidable in industrial applications.

Fig. 2.Amino-acid sequence diagrams showing the construction of D-PhgAT fusion with halophilic peptides from Hs Fdx, as indicated with line-connected boxed amino acid sequences. Acidic residues (D and E) in the extra N-terminus domain of Hs Fdx are shown in gray highlights. The start codon methionine residue added is shown as the first M at the N-terminus of each peptide. Sequence numbering is given at the top. Alpha-helix and beta-sheet regions, α’, α’’, β1, and β2 are indicated below the related sequences. Only the first 20 N-terminus residues of D-PhgAT are shown in this figure.

Currently, various techniques have been reviewed, addressed, and practiced successfully to increase solubility and stability of enzymes [1, 2, 4]. Amongst them, introducing mutations via site-directed mutagenesis to modify and engineer mesophilic enzymes was highly researched in the past few years; however, only a limited number of mutations could lead to a large elevation in solubility and stability [8]. Recent investigations revealed a biased surficial amino acid signature in halophilic archaea as a common adaptation strategy to increase their protein stability in their highly ionic natural habitats [6]. It was found that elevated contents of glutamate, aspartate, and a number of small hydrophobic residues with decreased lysine residues not only enabled these halophilic proteins to withstand the highly ionic habitats but also to warrant higher stability owing to the reduction of their water activity [10]. One of the most interesting halophilic adaptation strategies could be seen with a ferredoxin enzyme from Halobacterium salinarum (Hs Fdx), which was shown to possess an extra N-terminus domain of 30 residues compared with their non-halophilic homologs. The extra domain was composed of two α-helixes and two loops in which 13 residues (~44%) were acidic amino acids (glutamate and aspartate) [14]. The addition of an extra domain is considered to be a rapid adaptation mechanism of halophiles to treat the acquiring genes from their non-halophilic counterparts, so that expression of the genes with the extra domain leads to active proteins in the new environments.

In this study, D-PhgAT was used as a model protein to fuse with each of the two α-helixes (A1 and A2 and a hybrid of them, namely ALAL, from Hs Fdx) at the N-terminus of the enzyme molecule (Fig. 2). It was shown that the solubility and stability of D-PhgAT in high salts (NaCl and KCl) and high concentration of the substrate (L-glutamate) increased.

 

Materials and Methods

Construction of Plasmids for Expression of D-PhgAT with Halophilic Peptide Fusions

All enzymes in this study were purchased from New England BioLabs Inc., USA, unless otherwise noted. Oligonucleotide sequences of A1, A2, and ALAL were obtained from the amino acid sequence of the H. salinarum ferredoxin enzyme (Hs Fdx; PDB: 1E0Z_A) and optimized for expression in Escherichia coli host [19]. Full-length A1 and A2 oligonucleotides (A1-US: 5’-P~TATGGG CTACGAAACCCTGGACGACCAGGG-3’; A1-LS: 5’-P~TACCCT-GGTCGTCCAGGGTTTCGTAGCCCA-3’; A2-US: 5’-P~TATGGG CCTGTTCGAAAAAG-CGGCGGACGCGGG-3’; and A2-LS: P~TACCCGCGTCCGCCGCTTTTTCGAACAGGCC-CA) were synthesized with modified sticky ends suitable for ligation with NdeI-digested vector, and ALAL was synthesized as two half-length oligonucleotides with an 18 nucleotides complementary overlapping region (ALAL-US: P~TGGGCTACGAAACCCT-GGACGACCAGGGTTGGGACATGGACGACGACGACCTGTTC; and ALAL-LS: P~CCG-TCTTCACCGTCCAGACCCGCGTCCGCCGCTTTTTCGAACAGGTCGTCGTCGTC) and followed by one single-step extension with pfu DNA polymerase (Promega, USA) to prepare the full-length ALAL oligonucleotide with the following conditions: denaturation at 95℃ for 30 sec, annealing at 51℃ for 1 min, extension at 72℃ for 5 min. Plasmid pET-17b carrying the dpgA gene (GenBank: AY319935.1) from a previous work [3] was digested with NdeI restriction endonuclease, followed by dephosphorylation with antarctic phosphatase. Furthermore, blunt-ended plasmid was prepared by using T4 DNA polymerase for ligation with ALAL. E. coli XL-10 gold was transformed with the prepared plasmids by the heat shock method [23]. A colony PCR method was applied to verify the correct insertion and orientation of each insert by sets of primers as follow: A1F: 5’-TATGGGCTACGA AACCCTGGACGACCAGGG-3’; A1R: 5’-CATGACCGACCGA GCG-CGTTGTG-3’; A2F: 5’-GGCCTGTTCGAAAAAGCGG-CG-3’; A2R: 5’-GGTTAC-GCCGT-CGGGCATGA-3’. These primers were designed to amplify 106, 117, and 172 bp products from both inserts and the N-terminus region of the dpgA gene of A1, A2, and ALAL, respectively. The PCR condition was 30 cycles of denaturation temperature 95℃ for 20 sec, annealing temperature 63℃ for A1 and ALAL and 62℃ for A2 for 20 sec, and extension temperature 72℃ for 20 sec. Furthermore, DNA sequencing (BioDesign Co., Ltd., Thailand) of the selected plasmid was conducted with a T7 promoter primer (5’-AATACGACTCACTATAGG-3’) to reconfirm the result before being used for transformation of E. coli tuner (DE3) pLysS, which was used as the expression host.

Overexpression and Purification of Different Variants of Recombinant D-PhgAT

The overexpression and purification of recombinant D-PhgAT were done as previously described [3, 11]. Briefly, overnight culture of the recombinant E. coli tuner (DE3) pLysS was used to inoculate 250 ml of Luria Bertani broth (Difco) containing 100 μg/ml of ampicillin (Bio Basic Inc., Canada) and 34 μg/ml of chloramphenicol (Sigma-Aldrich Co. LLC, USA), followed by incubation at 37℃ and 200 rpm until OD600 of 0.6 was reached. IPTG (US Biological, USA) was added to 0.4 mM final concentration and incubation was continued at 20℃ and 100 rpm for the next 16 h. Cells were harvested at 10,000 ×g and 4℃ for 40 min and were resuspended in 9 vol of cold TEMP buffer, pH 7.6 (20 mM Tris-HCl, 1 mM EDTA, 0.01% β-mercaptoethanol, and 2.5 μM PLP). The cell suspension was disrupted by sonication in 20 cycles of 10 sec burst with 10 sec interval cooling each and followed by centrifugation at 12,000 ×g and 4℃ for 45 min. Clarified cell-free lysate was subjected to 25–45% saturation fractional ammonium sulfate precipitation as the first step of purification. Then, the protein pellets were collected by centrifugation at 12,000 ×g and 4℃ for 45 min, and dissolved in TEMP buffer, pH 7.6, containing 1 M ammonium sulfate, and the dissolved protein was applied on a Phenyl Sepharose 6 FF hydrophobic interaction column (Amersham Pharmacia Biotech, Uppsala, Sweden). The hydrophobic chromatography was done with the ÄKTA purifier FPLC system (GE Healthcare Bio-Sciences, Uppsala, Sweden). The bound proteins were eluted with four bed volumes of linear descending gradient of 1–0 M ammonium sulfate at the 2 ml/min flow rate. Active fractions were pooled, concentrated, and desalted by a 50 kDa Amicon Ultra-15 cut off centrifugal filter device (Millipore, Cork, Ireland) and then applied on a DEAE Sepharose FF IEX column (GE Healthcare). The bound proteins were eluted with eight bed volumes of a linear ascending gradient of 0–1 M NaCl at the 2 ml/min flow rate. Active fractions were pooled, concentrated and desalted as mentioned above. Protein purity was checked by the SDS-PAGE method and the protein concentration was determined with the NanoDrop ND-1000 (Thermo Fisher Scientific Inc., USA) at 280 nm based on molar absorption coefficients (ε) of 38,180, 39,670, 38,180, and 54,170 M-1cm-1 for WT-D-PhgAT, A1-DPhgAT, A2-D-PhgAT, and ALAL-D-PhgAT, respectively.

D-PhgAT Activity Assay

D-PhgAT activity was assayed by using a method based on the formation of 4-hydroxybenzoylformic acid (4-OHBZF) upon transamination of D-4-OHPhg as the substrate, with α-ketoglutaric acid monosodium salt as the amino acceptor. The 4-OHBZF was monitored by using a Heλios α spectrophotometer (Spectronic Unicam, Cambridge, UK) at 340 nm and 37℃ for 180 sec. The reaction mixture (1 ml), contained 20 μl of enzyme solution, 10 mM D-4-OHPhg, 10 mM α-ketoglutarate, 50 mM CAPSO buffer (pH 9.5), 5 μM PLP, and 5 μM EDTA. The molar absorption of 24 × 103 M-1cm-1 was used for calculation of the generated 4-OHBZF in 1 min. One unit of D-PhgAT was defined as the amount of enzyme used for transamination of 1 μmol of D-4-OHPhg to 1 μmol of 4-OHBZF per 1 min at 37℃.

Determination of pH Effect on Enzyme Activity

To determine the effect of pH on enzyme activity at a range from 5.0 to 11.0, 50 mM of the following buffers were used: citrate buffer (pH 5.0-6.0), PIPES buffer (pH 6.0-7.5), Tris-HCl buffer (pH 7.5-8.5), CAPSO buffer (pH 8.5-10.0), and CAPS buffer (pH 10.0-11.0). D-PhgAT activity was measured at 37℃ by using the spectrophotometric method described above.

Determination of Experimental pI Value

Immobiline DryStrips (7 cm, pH 3-10 NL; GE Healthcare Bio-Sciences AB) were used to determine the isoelectric values of variants of D-PhgAT. Strips were rehydrated with 125 μl of rehydration solution (8 M Urea, 4% CHAPS, 100 mM DTT, 2% IPG buffer, and trace amount of bromophenol blue) containing 100 μg of the protein sample for 13 h in ceramic IPGphor strip holders. The first dimension was carried out by using Ettan IPGphor (Amersham Pharmacia Biotech, CA, USA) according to the protocol provided by the supplier. SDS-PAGE was used as the second dimension and the protein spots were visualized by Coomassie Blue staining.

Determination of In Vitro Protein Solubility

In vitro protein solubility was determined at 25℃ in storage buffer (20 mM TEMP buffer, pH 7.6) and reaction buffer of the enzyme (50 mM CAPSO, pH 9.5) by the concentration method as previously described [3, 5, 12]. Briefly, a Microcon YM-30 centrifugal device with 30 kDa cut-off membrane (Millipore, MA, USA) was used to concentrate the enzyme solution at 4,000 ×g with interval quantification of the protein concentration by the NanoDrop ND-1000 at 280 nm as described above, until no further change in concentration was observed.

Determination of the Effects of Substrate Concentrations on D-PhgAT Activity

For this experiment, the concentration of D-phenylglycine produced in the D-PhgAT-catalyzed reaction was assayed by the HPLC method modified from Rojanarata et al. [20]. The reaction mixture, in 1 ml final volume, contained 100 mM CAPSO buffer (pH 9.5), 50 mM benzoylformic acid (pH 9.5), 25 μM PLP, 25 mM EDTA, different concentrations of L-glutamic acid monosodium salt (MSG), and 10 μl of enzyme solution (1 unit). The mixture was incubated at 37℃ for 60 min; at 20 min intervals, 50 μl aliquots were withdrawn, inactivated, and diluted simultaneously with 150 μl of 80℃ preheated distilled water and left standing for 2 min. Then, 20 μl of the diluted solution was injected into a Spherisorb ODS2 column (250 × 46 mm; Waters, Milford, USA), which was connected to an HPLC 1100 series system (Hewlett Packard, USA) and analyzed at 254 nm. Isocratic 50 mM ammonium acetate buffer (pH 5.5) containing 10% methanol was used as the mobile phase at 1.2 ml / min flow rate.

D-PhgAT Activity Assay in the Presence of NaCl or KCl

D-PhgAT activity in the presence of NaCl or KCl (0.5, 1.0, 1.5, 2.2, 2.5, 3.0, and 3.5 M) in the reaction mixture was determined via the spectrophotometric method. Since the pH of a buffered solution changes upon addition of salt and this may significantly affect the enzyme activity of the reaction mixtures, the pH shift (Δ pH) was determined in advance by using thymol blue (Fluka Analytical, USA) as the pH indicator at 595 nm, by a modified method as previously described [15, 17]. The CAPSO buffers with modified pH were used to maintain the final pH at 9.5.

Statistical Analysis

All experiments were carried out in triplicate (n = 3). Analysis of variance was used to interpret the data. Means and standard deviations are shown in the graphs.

 

Results and Discussion

Overexpression and Purification of Different Variants of Recombinant D-PhgAT

Fig. 3A shows the SDS-PAGE analysis of the clarified cell lysates of different successfully overexpressed variants of recombinant D-PhgAT in E. coli tuner (DE3) pLysS after induction with 0.4 mM IPTG. Then, the clarified cell lysates of induced cells were subjected to three steps of purification. Fig. 3B shows the SDS-PAGE analysis of each purified DPhgAT variant with a single band.

Fig. 3.SDS-PAGE analysis of (A) expression of D-PhgAT variants and (B) purified D-PhgAT variants. M, molecular weight marker, N: non-expressed cells, 1: WT-D-PhgAT; 2: A1-D-PhgAT; 3: A2-D-PhgAT; and 4: ALAL-D-PhgAT.

Effect of pH on Variants of D-PhgAT

The effect of pH on the activity of D-PhgAT variants was determined by the spectrophotometric method. All variants of D-PhgAT showed relatively low transamination activity at the acidic and neutral pH, whereas the activity of all variants gradually increased as the pH progressively shifted toward the alkaline condition until reaching a maximum at pH 9.5 (in CAPSO buffer). Further increase in pH from 9.5 resulted in a decreasing of D-PhgAT activity, and no significant activity was observed at pH 11.0 and above. All D-PhgAT variants displayed a similar pH-activity profile to that of WT-D-PhgAT, indicating that there was no significant structural change of the catalytic site and probably of the overall enzyme molecule upon fusion with the halophilic peptides.

Determination of pI value of Variants of D-PhgAT

The isoelectric pH (pI) values of D-PhgAT variants were determined by the isoelectric focusing method using Immobiline DryStrips immobilized pH gradient gel. Fusion of A1, A2, or ALAL peptide, which contained 3, 2, or 13 negatively charged residues, respectively, at the N-terminus of D-PhgAT significantly reduced the pI values compared with that of WT-D-PhgAT. As shown in Fig. 4, the pI values of A1-D-PhgAT, A2-D-PhgAT, and ALAL-D-PhgAT were 5.53, 5.76, and 5.28, respectively, and that of the WT-DPhgAT was previously reported to be 6.2 [3]. This is as expected because the pI value of a given protein is actually related to its net charge, properties of its amino acid composition, and the pK of its ionizable groups, where the increase in net negatively charged residues would result in the decrease of the protein pI value [21].

Fig. 4.Position of different variants of D-PhgAT after being focused on 7 cm Immobiline DryStrips pH 3-10 NL. (A) A1-D-PhgAT, (B) A2-D-PhgAT, and (C) ALAL-D-PhgAT. M indicates 45 kDa molecular mass marker.

Effect of Halophilic Peptide Fusion on In Vitro Solubility of D-PhgAT

The in vitro solubility of D-PhgAT variants was determined at 25℃ in TEMP (pH 7.6) and CAPSO (pH 9.5) buffers by the concentration method. As expected and shown in Table 1, fusion of negatively charged halophilic peptides at the N-terminus domain of D-PhgAT increased the solubility of A1-D-PhgAT, A2-D-PhgAT, and ALAL-D-PhgAT in TEMP (pH 7.6) storage buffer by 6.1-, 5.3-, and 8.1-fold, respectively, and in CAPSO (pH 9.5) reaction buffers by 5-, 4.5-, and 5.9-fold, respectively, compared with that of the WT-D-PhgAT. The protein solubility is the function of its total net charge and the buffer pH [18], where at the pH higher than the pI, deprotonation of the residues occurrs, leading to stronger repulsive forces between the protein molecules, which consequently prevent aggregation and lead to higher solubility. The other factor that can largely affect protein solubility is the number of hydrophobic residues on the protein surface, such as that with D-PhgAT. Predominantly, the hydrophobic force pushes these residues into the protein core, but formation of hydrophobic patches on the surface of hydrophobic proteins is usually found. This phenomenon leads to a higher tendency of proteins to aggregate, precipitate, and display low solubility. Our results attest the previous study [12] that fusion with a negatively charged peptide can increase the solubility of DPhgAT. The fusion peptides having a high content of polar residues interact with water molecules, resulting in lowering protein-protein interactions and thereby favoring protein solubility.

Table 1.Summary of the solubility properties of different variants of D-PhgAT in 20 mM TEMP (pH 7.6) storage and 50 mM CAPSO (pH 9.5) reaction buffers (n = 3).

Effects of Substrate and Salt Concentrations on D-PhgAT Activity

The effects of various concentrations of MSG salt as a substrate on the activity of D-PhgAT to produce D-Phg were determined by the HPLC method. Increasing concentrations of MSG salt in the reaction mixture resulted in a proportional increase in activity of D-PhgAT and its variants up to a certain value for each, but beyond that the activity started to drop (Fig. 5). The highest reaction rate for WT-D-PhgAT was 12.56 ± 1. 16 mM/h with 1,250 mM of MSG and those for A1-D-PhgAT, A2-D-PhgAT, and ALAL-D-PhgAT were 17.81 ± 0.67, 20.17 ± 1.1, and 23.82 ± 1.47 mM/h, respectively, with 1,500 mM of MSG. At higher concentrations than 1,250 mM for WT-D-PhgAT and 1,500 mM for the halophilic fused variants, the reaction rates diminished gradually. Surprisingly, the activity of A2-D-PhgAT was higher than that of the A1-D-PhgAT at all MSG concentrations, which has higher solubility compared with A2-D-PhgAT. The reason for this incoherence is yet to be investigated. The decrease in activity with the high concentrations of the substrate might be due to the effects of substrate inhibition or the lower substrate diffusion to the active site of the enzyme, as previously suggested [20]. At 2,000 mM MSG, visible precipitation of WT-D-PhgAT was observed, whereas no precipitation was detected with those halophilic fused variants (data not shown). In addition to this, after the reaction mixture was diluted up to the final concentration of 1,500 mM MSG, the activity of the halophilic fused variants was partially retrieved suggesting that the decrease in activity was due to substrate inhibition rather than denaturation. Of these, ALAL-D-PhgAT displayed the highest reaction rate and the fusion with halophilic peptide improved the solubility in high concentration of MSG.

Fig. 5.Effects of different proportions of L-glutamate monosodium salt on the reaction rate of WT-D-PhgAT (●), A1-D-PhgAT (▲), A2-D-PhgAT (▼), and ALAL-D-PhgAT (■) at 50 mM benzoylfomaric acid. Bars represent the standard deviation (n = 3).

When NaCl or KCl was included in the reaction mixture with WT-D-PhgAT and all its variants, only a slight decrease in D-PhgAT activity was observed for both salts at the concentration up to 1,500 mM. Further increase in salt concentrations significantly inhibited the enzyme activity (Fig. 6). However, the activity was still detected even in the presence of 3,500 mM salts. Determination of the C50 value, which is the value of salt concentration that inhibits 50% of the enzyme activity, revealed that all variants displayed slightly higher tolerance for KCl than NaCl, and all variants with the halophilic peptide fusions were more active than the WT-D-PhgAT (Table 2). Based on the Hofmeister series, NaCl is a kosmotropic and KCl a chaotropic salt; thus, they manifest different enzyme inactivation mechanisms, which can possibly be explained by affecting the physical states of water around the enzyme molecules [16]. Kosmotropic salts, which are known as a water maker, will interact with bulk water molecules in the solvent more than the hydrophobic surficial residues of the enzyme. This phenomenon is referred to as preferential exclusion and will force to bury these nonpolar residues into the core of the enzyme at the expense of protein flexibility and thereby inactivation of the enzyme. In contrast, chaotropic salts, which are referred as water structure breakers, are pushed away from bulk water and will interact with the aqueous shell region of the enzyme or even bind to it. This phenomenon is known as preferential binding and will lead to weakened hydrophobic interactions, increased enzymesolvent accessible area, and subsequently denaturation of the enzyme. The negatively charged residues on A1, A2, and ALAL α-helixes bind to notable amounts of hydrated ions and thus decrease surficial hydrophobicity; thus, the lower tendency to aggregate unless at higher salt concentrations. To the best of our knowledge, this is the first study that successfully applies the use of the halophilic peptide fusion to increase the solubility and stability of a hydrophobic enzyme, D-PhgAT, which is an enzyme of invaluable biotechnological and pharmaceutical potentials. Overall, the results of our in vitro study clearly indicate that the method not only could increase the solubility of D-PhgAT but also could interestingly increase its tolerance to high concentrations of MSG as the substrate of the D-PhgAT reaction. The discovery indicates the high potential of employing this new knowledge for production of D-phenylglycine and D-4-hydroxyphenylglycine, the essential side-chain moieties for manufacture of highly demanded semisynthetic antibiotics such as penicillins and cephalosporins.

Table 2.C50 values of the wild-type and variants of D-PhgAT in salts (n = 3).

Fig. 6.Activities of WT-D-PhgAT (●), A1-D-PhgAT (▲), A2-DPhgAT (▼), and ALAL-D-PhgAT (■) at different proportions of NaCl (A) and KCl (B). Bars represent the standard deviation (n = 3).