l-Amino Acid Ligase from Pseudomonas syringae Producing Tabtoxin Can Be Used for Enzymatic Synthesis of Various Functional Peptides

Abstract

Functional peptides are expected to be beneficial compounds that improve our quality of life. To address the growing need for functional peptides, we have examined peptide synthesis by using microbial enzymes. l-Amino acid ligase (Lal) catalyzes the condensation of unprotected amino acids in an ATP-dependent manner and is applicable to fermentative production. Hence, Lal is a promising enzyme to achieve cost-effective synthesis. To obtain a Lal with novel substrate specificity, we focused on the putative Lal involved in the biosynthesis of the dipeptidic phytotoxin designated tabtoxin. The tabS gene was cloned from Pseudomonas syringae NBRC14081 and overexpressed in Escherichia coli cells. The recombinant TabS protein produced showed the broadest substrate specificity of any known Lal; it detected 136 of 231 combinations of amino acid substrates when dipeptide synthesis was examined. In addition, some new substrate specificities were identified and unusual amino acids, e.g., l-pipecolic acid, hydroxy-l-proline, and β-alanine, were found to be acceptable substrates. Furthermore, kinetic analysis and monitoring of the reactions over a short time revealed that TabS showed distinct substrate selectivity at the N and C termini, which made it possible to specifically synthesize a peptide without by-products such as homopeptides and heteropeptides with the reverse sequence. TabS specifically synthesized the following functional peptides, including their precursors: l-arginyl-l-phenylalanine (antihypertensive effect; yield, 62%), l-leucyl-l-isoleucine (antidepressive effect; yield, 77%), l-glutaminyl-l-tryptophan (precursor of l-glutamyl-l-tryptophan, which has antiangiogenic activity; yield, 54%), l-leucyl-l-serine (enhances saltiness; yield, 83%), and l-glutaminyl-l-threonine (precursor of l-glutamyl-l-threonine, which enhances saltiness; yield, 96%). Furthermore, our results also provide new insights into tabtoxin biosynthesis.

INTRODUCTION

Peptides are expected to be among the most promising compounds that are beneficial for improving our quality of life (1–3). Research on functional peptides has been carried out in various fields, including food science, medicine, and cosmetics. New findings are frequently reported. Dipeptides are discussed in this report, although longer peptides also possess unique physiological functions and physical properties that the constituent individual amino acids do not exhibit. In the field of food science, Nippon Suisan Kaisha, Ltd., has applied for a patent on a saltiness-strengthening agent obtained by adding a dipeptide containing l-glutamic acid (Glu, E) such as l-glutamyl-l-threonine (Glu-Thr) (M. Shimono and K. Sugiyama, Japanese patent WO2009/113563, 2009). In addition, Kao Corporation has applied for a patent on a similar function of l-leucyl-l-serine (Leu-Ser), which also enhances saltiness (M. Koike, Japanese patent JP 2012-165740, 2012). Reduction of the salt content of foods is particularly important for people with high blood pressure. Furthermore, Ohinata et al. recently reported that l-arginyl-l-phenylalanine (Arg-Phe) decreased blood pressure and food intake in rodents (4). Arg-Phe does not act as an angiotensin I-converting enzyme-inhibitory peptide (5), as is frequently reported, and it shows vasorelaxing activity. In the medical field, l-glutamyl-l-tryptophan (Glu-Trp), which is named oglufanide, was reported to show antiangiogenic properties and inhibition of tumor growth in in vivo preclinical models (6). It passed phase I and II trials, but the phase III randomized, double-blind, placebo-controlled trial failed to show significant antitumor activity. Nitta et al. reported that l-leucyl-l-isoleucine (Leu-Ile) induced brain-derived neurotrophic factor in cultured neuronal cells and may act as an antidepressant (7). l-Histidyl-β-alanine (His-β-Ala), which is the reverse sequence of β-alanyl-l-histidine (β-Ala-His, carnosine), was reported to induce sedative and hypnotic effects (8). Carnosine has been well studied and has been shown to be useful as an antioxidant (9). The structural differences between β-Ala-His and His-β-Ala are small, but their bioactivities are quite different. In the cosmetic field, Kao Corporation has applied for a patent on the hair growth-inhibitory effect of l-phenylalanyl-β-alanine (Phe-β-Ala) toward elastase (N. Tsuji and S. Moriwaki, Japanese patent JP 2001-226232, 2001). The functional peptides described above are newly discovered. Furthermore, dipeptides have also been reported to act as catalysts by themselves (10, 11). Taken together, these previous studies indicate that the development of a peptide-manufacturing process is important for addressing the growing need for these functional peptides.

l-Amino acid ligase (Lal) is a microbial enzyme that catalyzes dipeptide synthesis from unprotected amino acids by hydrolysis of ATP to ADP and phosphate (P_i) (12). Lal belongs to the ATP-dependent carboxylate-amine/thiol ligase superfamily (13) and therefore catalyzes ligation in an ATP-dependent manner through an aminoacyl-phosphate reaction intermediate (see Fig. S1 in the supplemental material) (14). This condensation process is achieved in one step, and degradation of reaction products does not occur. Since amino acid substrates and ATP can be supplied in microbial cells, fermentative production of dipeptides has been developed (15). Therefore, enzymatic peptide synthesis using Lal could be very efficient. To increase the variety of peptides that can be synthesized, we have screened a new Lal that exhibits unique substrate specificity, and Lals have been obtained from various microorganisms (16–23). In the course of these examinations, our results have also provided new insight into the biosynthesis of peptidic secondary metabolites, e.g., rhizocticin (24) and phaseolotoxin (25). However, there are more reports on the biosynthesis of peptidic secondary metabolites involving nonribosomal peptide synthetase than reports describing Lals (26). Rhizocticin and phaseolotoxin are di- or tripeptide antibiotic and tripeptide phytotoxins, respectively, that contain an unusual amino acid, and two ATP-grasp enzymes were identified in each biosynthetic gene cluster (27, 28). In rhizocticin biosynthesis, our results revealed that two Lals (RizA and RizB) combined three amino acids, yielding tripeptides similar to rhizocticin (19, 20). RizA catalyzed the condensation of amino acids, and RizB combined the resulting dipeptide and another amino acid. Rhizocticin biosynthesis might be sequentially preceded by two Lals. In phaseolotoxin biosynthesis, we demonstrated that one Lal synthesized l-alanyl-l-homoarginine, which is a component of phaseolotoxin, but the ligase activity of the other enzyme could not be tested because of the formation of the insoluble form when the corresponding gene was overexpressed in Escherichia coli cells (21).

We have recently focused on a tabtoxin phytotoxin to obtain a Lal showing novel substrate specificity. Tabtoxin is a dipeptidic phytotoxin produced by Pseudomonas syringae strains (29). This compound is composed of tabtoxinine-β-lactam (TβL) and l-threonine (Thr, T); TβL is located at the N terminus, and Thr is at the C terminus, resulting in TβL-Thr. TβL-Thr is hydrolyzed, and the resulting TβL irreversibly inhibits glutamine synthetase (EC 6.3.1.2), causing characteristic chlorosis in plants. Willis et al. cloned the biosynthetic gene cluster of tabtoxin from P. syringae ATCC 11528 (equal to BR2) and analyzed each gene (30). tblF was identified as an ATP-grasp enzyme-encoding gene. Arrebola et al. hypothesized that TblF is involved in the condensation of TβL and Thr in their review (31). Subsequently, Walsh et al. demonstrated the role of TblF, combining TβL and Thr as self-protection from TβL toxicity (32). They further showed that TβL is spontaneously isomerized to tabtoxinine-δ-lactam (TδL), and TblF did not recognize TδL as a substrate, yielding no TδL-Thr. Some characteristics were also provided by them, but only from the perspective of tabtoxin biosynthesis.

In this study, we cloned the homolog of tblF from P. syringae NBRC14081 (ATCC 27881) and investigated its enzymatic properties in detail from the aspect of a peptide-synthesizing enzyme. Surprisingly, this protein was useful for the synthesis of various functional peptides with high selectivity. Furthermore, our results provide new insights into tabtoxin biosynthesis.

MATERIALS AND METHODS

Materials.

P. syringae NBRC14081 (29) was purchased from the NITE Biological Resource Center (Chiba, Japan). E. coli BL21(DE3) cells and the pET-28a(+) vector were purchased from Merck (Darmstadt, Germany). All of the other chemicals used in this study are commercially available and are of chemically pure grade.

Genetic manipulation and preparation of recombinant proteins.

To obtain the gene corresponding to tblF, the gene fragment was amplified from the genomic DNA of P. syringae NBRC14081 by PCR with the following primers that were designed on the basis of the sequence of the whole tabtoxin-biosynthetic gene cluster derived from P. syringae BR2: PT3 sense (5′-AGTGGTCAATGACCTCTGC-3′) and PT3 anti (5′-TCATGGCGATGATTTCCAGAC-3′). The resulting fragment was sequenced, and the open reading frame was designated tabS. tabS was amplified from the genomic DNA of P. syringae NBRC14081 by PCR with the following primers: tabS_sense (5′-AAGTAGCACATATGATGACGCAAGCCAAGG-3′, NdeI) and tabS_antisense (5′-ATCGAATTCCTACTGGACGCTGAACA-3′, EcoRI). The PCR fragments were digested with NdeI and EcoRI and then ligated into the pET-28a(+) vector. The resulting plasmids were designed to express the genes with a His tag sequence at the N terminus under the control of the T7 promoter, and the plasmid was then introduced into E. coli BL21(DE3) cells.

E. coli BL21(DE3) cells harboring tabS-ligated pET-28a(+) were cultivated in 3 ml of Luria-Bertani (LB) medium (1% Bacto tryptone, 0.5% yeast extract, 1% NaCl) containing 30 μg/ml (final concentration) of kanamycin at 37°C for 5 h with shaking at 160 rpm. The cultivated cells were transferred to 100 ml of fresh LB medium containing the same antibiotics used during precultivation and were further cultivated at 37°C for 1 h with shaking at 120 rpm. Isopropyl-β-d-thiogalactopyranoside (final concentration, 0.1 mM) was then added, and cultivation was continued at 25°C for 19 h with shaking at 120 rpm. The cells were harvested by centrifugation (4,160 × g, 10 min, 4°C), resuspended in 100 mM Tris-HCl buffer (pH 8), and then disrupted by sonication at 4°C. Cellular debris was removed by centrifugation (20,000 × g, 30 min, 4°C), and the supernatant was collected and purified with a HisTrap HP Ni affinity column (GE Healthcare, Buckinghamshire, United Kingdom). The fractions were then desalted with a PD-10 column (GE Healthcare) and equilibrated with 100 mM Tris-HCl buffer (pH 8 or 9). To confirm the presence of protein in the solution, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed. The protein concentration was then determined by the Bradford method, with bovine serum albumin as the standard.

Characterization of TabS.

To test substrate specificity, every combination of one or two amino acids selected from 20 types of proteinogenic amino acids and β-alanine (β-Ala, βA) was examined. The reaction mixture used for assays (0.3-ml total volume) contained 25 mM amino acid substrate, 12.5 mM ATP, 12.5 mM MgSO₄ · 7H₂O, and 0.1 mg/ml of TabS in 100 mM Tris-HCl buffer (pH 8). The reaction mixture was incubated at 30°C for 20 h. When reactions involved two substrates, the concentration of each substrate was 12.5 mM. To detect the activity in the reaction mixtures, the reaction products were analyzed by liquid chromatography-electrospray ionization-mass spectrometry (LC-ESI MS) and the amount of P_i was determined with a Determiner L IP kit (Kyowa Medex, Tokyo, Japan) according to the manufacturer's protocol. To examine the effects of pH, temperature, and thermal stability of the protein, a reaction mixture containing 20 mM l-leucine (Leu, L), 20 mM ATP, 20 mM MgSO₄ · 7H₂O, and 0.1 mg/ml of TabS was prepared. When the effect of pH was examined, reaction mixtures were incubated at 30°C for 60 min in 100 mM Tris-HCl buffer maintained at pH 8, 8.5, 9, 9.5, or 10 or in 100 mM sodium phosphate buffer maintained at pH 7 or 8. When the effect of temperature was examined, reaction mixtures were incubated at 20, 25, 30, 35, 40, 45, or 50°C for 60 min in 100 mM Tris-HCl buffer (pH 9). When thermal stability was examined, TabS was incubated at 20, 25, 30, 35, 40, 45, or 50°C for 15 min prior to initiation of the reactions. The reaction mixture was incubated at 30°C for 20 h in 100 mM Tris-HCl buffer (pH 9). l-Leucyl-l-leucine (Leu-Leu) in these reaction mixtures was analyzed by high-performance liquid chromatography (HPLC). To test the effect of a reducing agent, dithiothreitol (DTT) was added to a reaction mixture containing 25 mM Leu, 12.5 mM ATP, 12.5 mM MgSO₄ · 7H₂O, and 0.5 mg/ml of TabS in 100 mM Tris-HCl buffer (pH 9). The reaction mixture was incubated at 30°C for 30 min, and the reaction product was analyzed by HPLC. To determine the molecular mass of the protein, a HiLoad 16/60 Superdex 200 pg column was used (GE Healthcare). The column was equilibrated with 100 mM Tris-HCl buffer (pH 8) containing 0.15 M NaCl, and the protein was eluted with the same buffer at a flow rate of 0.5 ml/min.

Investigation of substrate recognition.

For l-glutaminyl-l-threonine (Gln-Thr) synthesis, a standard reaction mixture containing 25 mM l-glutamine (Gln, Q), 25 mM Thr, 50 mM ATP, 25 mM MgSO₄ · 7H₂O, and 0.5 mg/ml of TabS in 100 mM Tris-HCl buffer (pH 9) was prepared and the reaction mixture was incubated for 20 h at 30°C. The reaction products were analyzed by HPLC. To analyze kinetic properties, the reaction mixtures were incubated by varying the concentration of Gln, Thr, or ATP under standard conditions, except for the concentration of TabS. The enzyme concentration used was 0.25 mg/ml, and the reaction time was 60 min. Gln-Thr was analyzed by HPLC. To investigate substrate recognition, homopeptide synthesis and heteropeptide synthesis were examined under standard conditions by varying the amino acids. Reaction mixtures were incubated for 30 min at 30°C, and after the reaction, the amounts of P_i produced in the reaction mixtures were analyzed. To examine the reactivity with l-proline (Pro, P) and its analogs, a reaction mixture containing 12.5 mM l-pipecolic acid, Pro, cis-4-hydroxy-l-proline, trans-4-hydroxy-l-proline, or l-azetidine-2-carboxylic acid; 12.5 mM Thr; 12.5 mM ATP; 12.5 mM MgSO₄ · 7H₂O; and 0.5 mg/ml of TabS in 100 mM Tris-HCl buffer (pH 9) was prepared. The reaction mixture was incubated for 20 h at 30°C. The amount of P_i produced in the reaction mixture was analyzed, and the reaction products were deduced by LC-ESI MS.

Synthesis of functional peptides.

The substrate concentrations used for the reaction were as follows: 50 mM l-arginine (Arg, R) and 12.5 mM l-phenylalanine (Phe, F) for Arg-Phe synthesis, 50 mM l-isoleucine (Ile, I) and 12.5 mM Leu for Leu-Ile synthesis, 50 mM Gln and 12.5 mM l-tryptophan (Trp, W) for l-glutaminyl-l-tryptophan (Gln-Trp) synthesis, and 12.5 mM Leu and 12.5 mM l-serine (Ser, S) for Leu-Ser synthesis. These reaction mixtures also contained 12.5 mM ATP, 12.5 mM MgSO₄ · 7H₂O, and 0.5 mg/ml of TabS in 100 mM Tris-HCl buffer (pH 9). Furthermore, 300 mM KCl was added to the reaction mixture when examining the synthesis of Arg-Phe and Gln-Trp. To synthesize dipeptides containing β-Ala, the reaction mixture contained 20 mM Phe or l-histidine (His, H), 100 mM β-Ala, 50 mM ATP, 25 mM MgSO₄ · 7H₂O, and 0.5 mg/ml of TabS in 100 mM Tris-HCl buffer (pH 9). Those reaction mixtures were incubated for 20 h at 30°C, and all reaction products were analyzed by HPLC.

Analysis of peptides.

HPLC analysis was performed as follows. Amino acids and peptides were derivatized with N-α-(5-fluoro-2,4-dinitrophenyl)-l-alaninamide, which was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). The sample mixture was analyzed with an HPLC system (L-2000 series; Hitachi High Technologies, Tokyo, Japan). LC-ESI MS analysis was performed as follows. The reaction products were centrifuged (20,000 × g, 30 min, 4°C), and the supernatants were analyzed by LC-ESI MS (HPLC, Agilent 1100 series [Agilent Technologies, Santa Clara, CA]; ESI-MS, LCQ Deca [Thermo Scientific, Waltham, MA]). The details of these analytical procedures have been described in our previous publication (20).

Nucleotide sequence accession numbers.

The complete sequence of the P. syringae tabtoxin biosynthetic gene cluster (accession number DQ187985.1), the tblF gene sequence (accession number AAP13071.1), and the tabS gene sequence (accession number AB548153.1) have been deposited in the GenBank database.

RESULTS

Cloning and overexpression of tabs from P. syringae NBRC14081.

The homolog of tblF was amplified from the genomic DNA of P. syringae NBRC14081 by PCR, and the resulting tabS gene was then sequenced. The amino acid sequence of TabS was identical to that of TblF, although silent mutations occurred at three points (see Fig. S2 in the supplemental material).

General enzymatic properties of TabS.

The substrate specificity of TabS was investigated. Reaction mixtures were incubated in 100 mM Tris-HCl buffer (pH 8) at 30°C overnight on the basis of reported conditions used to examine the substrate specificities of known Lals. The results showed that TabS has broad substrate specificity, detecting m/z peaks corresponding to dipeptides in 136 of 231 combinations of amino acid substrates (Fig. 1). The substrate specificity of TabS is the broadest of any known Lal. When d-form amino acids were used as substrates, small amounts of P_i produced by the hydrolysis of ATP after the formation of the amide bond were detected, suggesting that d-form amino acids are not acceptable substrates.

Fig 1.

Overview of the substrate specificity of TabS. Reaction mixtures were analyzed by LC-ESI MS. A filled circle indicates the formation of the corresponding dipeptide. The amino acids used as substrates are shown at the left and along the bottom and are identified by a one-letter code.

The optimal reaction conditions were examined with Leu as a substrate because the maximum amount of P_i was detected when Leu was used in homopeptide synthesis. The concentrations of P_i were as follows: Leu, 4.2 mM; l-methionine (Met, M), 4.1 mM; His, 1.8 mM. The optimal temperature and pH for Leu-Leu synthesis for 1 h were 35°C and 9 to 9.5, respectively (Fig. 2). Thermal stability was then examined. The activity was maintained up to 35°C, but it sharply decreased above 40°C (see Fig. S3A in the supplemental material). These properties of TabS are similar to those of known Lals. In addition, the specific activity was increased by the addition of DTT (see Fig. S3B). The molecular mass was estimated by gel filtration to be 45.4 kDa (see Fig. S3C). This result suggests that TabS is a monomeric enzyme, since the molecular mass calculated on the basis of its amino acid sequence, including the tag sequence, was 48.4 kDa.

Fig 2.

Effects of pH (A) and temperature (B) on the peptide-synthesizing activity of TabS. Reaction mixtures were incubated for 1 h, and reaction products were analyzed by HPLC. In panel A, open circles indicate the use of 100 mM sodium phosphate buffer at pH 7 to 8 and closed circles indicate the use of 100 mM Tris-HCl buffer at pH 8 to 10. The data shown are averages of three measurements.

Substrate recognition of TabS.

TabS should catalyze the ligation of TβL and Thr because it shows 100% homology with TblF, which has been reported to catalyze the condensation of TβL and Thr in an ATP-dependent manner. Thus, TβL may be the most suitable N-terminal substrate for TabS. Indeed, Walsh et al. revealed high affinity between TβL and TblF by examination of the K_m value. On the other hand, TβL was reported to inhibit the activity of glutamine synthetase (33). Hence, Gln and Thr were selected as the model substrates and peptide synthesis was investigated. First, the amount of reaction product was measured by HPLC. The results showed that 23.9 ± 0.60 mM Gln-Thr was synthesized from 25 mM Gln and 25 mM Thr, showing a yield of about 96%. Additionally, 1.2 ± 0.60 mM l-threonyl-l-threonine (Thr-Thr) was detected but l-glutaminyl-l-glutamine (Gln-Gln) and l-threonyl-l-glutamine (Thr-Gln) were not detected. Hence, TabS possessed high selectivity toward amino acid substrates located at the N and C termini. Next, the kinetic properties of Gln-Thr synthesis were examined (see Fig. S4 in the supplemental material). The apparent K_ms for Gln, Thr, and ATP were 70.4 ± 2.3, 2.0 ± 0.4, and 2.5 ± 0.3 mM, respectively. The V_max for Gln-Thr synthesis was 728.8 ± 21.5 nmol · min⁻¹ · mg protein⁻¹.

As described above, TabS did not synthesize Gln-Gln from Gln, indicating that Gln is not recognized as the C-terminal substrate. In contrast, Thr is strongly recognized as the C-terminal substrate, although a trace amount of Thr-Thr was synthesized. Therefore, when the reaction was conducted with Gln and Xaa (Xaa, arbitrary amino acid) as the substrate, the detection of significant amounts of P_i indicated that Xaa was preferentially located at the C terminus. In contrast, when the reaction was conducted with Xaa and Thr as the substrates, the detection of significant amounts of P_i indicated that Xaa is preferentially located at the N terminus. These reactions were then carried out for 30 min. The results showed that TabS preferably recognized Gln, Arg, l-lysine (Lys, K), l-tyrosine (Tyr, Y), l-asparagine (Asn, N), Pro, Phe, His, Met, and Leu as N-terminal substrates, while Thr, l-valine (Val, V), Ile, Ser, l-alanine (Ala, A), l-cysteine (Cys, C), Trp, and glycine (Gly, G) were recognized as C-terminal substrates (Fig. 3). l-Aspartic acid (Asp, D), Glu, and β-Ala showed low reactivity. P_i analysis also showed that Leu, Met, and His seemed to be recognized as C-terminal substrates because significant amounts of P_i were detected when Xaa was used as the sole substrate. Further examination showed that TabS could synthesize 0.27 ± 0.03 mM l-prolyl-l-proline (Pro-Pro) from 100 mM Pro. Pro seemed to be acceptable at the C terminus, but its location at the N terminus is strongly preferred. In addition, when the reaction mixtures were incubated with Pro and Ala, Phe, Leu, or Lys, l-prolyl-l-alanine (Pro-Ala), l-prolyl-l-phenylalanine (Pro-Phe), l-prolyl-l-leucine (Pro-Leu), and l-prolyl-l-lysine (Pro-Lys) were detected, suggesting that the N-terminal orientation of Pro is stronger than that of Phe, Leu, and Lys.

Fig 3.

Measurement of P_i production in reaction mixtures. Black bars show the result of homopeptide synthesis, in which an arbitrary amino acid (Xaa) was used as the sole substrate. White bars show the result of heteropeptide synthesis with Glu and Xaa. Gray bars show the result of heteropeptide synthesis with Xaa and Thr. BLK shows the result when the reaction mixture was prepared without amino acid substrates. The data shown are averages of three measurements. The dashed vertical lines divide the areas, showing the preferred substrates recognized at the N and C termini.

It was reported that TβL was spontaneously isomerized to TδL, which has a six-membered ring. The reaction with l-pipecolic acid and Thr was performed because l-pipecolic acid also has a six-membered ring and is similar to TδL. In addition, cis-4-hydroxy-l-proline, trans-4-hydroxy-l-proline, and l-azetidine-2-carboxylic acid were used for the reaction as a Pro analog. LC-ESI MS analysis showed that m/z peaks corresponding to heterodipeptides containing each Pro analog were detected (see Fig. S5 in the supplemental material), and P_i analysis suggested that l-pipecolic acid is more suitable as a substrate for TabS than Pro is (Table 1).

Table 1.

Analysis of reaction products catalyzed by TabS with Pro analogs as substrates

Substrate		Avg phosphatea concn (mM) ± SD	[M + H]+
1	2		Calculatedb	Foundc
l-Pipecolic acid	Thr	7.6 ± 0.26	231.13	231.16
l-Proline (Pro)	Thr	6.9 ± 0.13	217.12	217.13
cis-4-Hydroxy-l-proline	Thr	6.8 ± 0.16	233.11	233.13
trans-4-Hydroxy-l-proline	Thr	4.1 ± 0.11	233.11	233.15
l-Azetidine-2-carboxylic acid	Thr	4.1 ± 0.09	203.10	203.11

^aAverages of three measurements are shown. When reactions were performed using 12.5 mM substrate 1 alone without Thr (substrate 2), the amounts of phosphate produced in the reaction mixtures were less than 0.5 mM.

^b[M + H]⁺ values shown were calculated on the basis of the chemical formula.

^c[M + H]⁺ values shown are for when the m/z peak corresponding to the deduced peptide was detected.

Functional peptides synthesized by TabS.

Since TabS showed distinct substrate specificity at both the N and C termini, this enzyme should be applicable to highly selective peptide synthesis that does not generate by-products. As described above, TabS synthesized Gln-Thr with a 96% yield. Gln-Thr can be converted to Glu-Thr by using N-terminal amidase (34), and the resulting peptide enhances the saltiness of foods. Hence, Gln-Thr is usable as a precursor of Glu-Thr, which is an important peptide for people with high blood pressure, for reducing the salt levels in foods. Furthermore, TabS could selectively synthesize various functional peptides, such as Arg-Phe, Leu-Ile, and Gln-Trp, as precursors of Glu-Trp and Leu-Ser (Table 2). These peptides were the main products, while the levels of by-products such as homopeptides and other heteropeptides were low or not detected. Gln-Trp can also be converted to a Glu-Trp functional peptide by N-terminal amidase. All of these peptides have beneficial effects for humans.

Table 2.

Functional peptides synthesized by TabS

Peptide (reaction product)	Function	Avg producta concn (mM) ± SD	Yieldb (%)
Arg-Phe	Antihypertensive effect	7.7 ± 0.06	62
Leu-Ile	Antidepressive effect	9.6 ± 0.09	77
Gln-Trp (precursor of Glu-Trp)	Antiangiogenic activity (Glu-Trp)	6.8 ± 0.02	54
Leu-Ser	Enhancement of saltiness	10.5 ± 0.18	83

^aReaction products were analyzed by HPLC. Averages of three measurements are shown.

^bYield was calculated based on the amount of ATP added to the reaction mixture. Hence, 12.5 mM reaction product is the maximum.

TabS also synthesized a functional peptide containing β-Ala, e.g., β-Ala–His, His–β-Ala, and Phe–β-Ala. Since the reactivity with β-Ala is low, the reaction mixture was prepared with an excess amount of β-Ala compared to that of His or Phe. HPLC analysis showed that 0.05 ± 0.00 mM β-Ala–His, 5.7 ± 0.06 mM His–β-Ala, 8.5 ± 0.07 mM l-histidyl-l-histidine (His-His), and 0.23 ± 0.00 mM β-alanyl-β-alanine (β-Ala–β-Ala) were synthesized from 20 mM His and 100 mM β-Ala. For Phe and β-Ala, 0.89 ± 0.06 mM Phe-β-Ala, 0.95 ± 0.00 mM β-alanyl-l-phenylalanine (β-Ala–Phe), 0.64 ± 0.05 mM l-phenylalanyl-l-phenylalanine (Phe-Phe), and 0.99 ± 0.04 mM β-Ala–β-Ala were synthesized from 20 mM Phe and 100 mM β-Ala. The standard deviation values of β-Ala–His, β-Ala–β-Ala, and β-Ala–Phe were less than 0.00 mM. Consequently, functional peptides containing β-Ala were also synthesized but the yields and selectivity were low because of low reactivity with β-Ala.

DISCUSSION

Our results revealed that TabS can synthesize various functional peptides by using only one enzyme with high selectivity. The synthesis of Gln-Thr, Arg-Phe, Leu-Ile, Gln-Trp, and Leu-Ser was demonstrated, and these peptides were found to be important compounds that can improve the quality of human life. TabS showed V_max values in Gln-Thr synthesis similar to that of YwfE in l-alanyl-l-glutamine (Ala-Gln) synthesis, and the YwfE protein is used for the commercial supply of Ala-Gln (12). Hence, TabS could be used for peptide production on an industrial scale, which might lead to the establishment of a commercial supply of various functional peptides at low cost.

TabS possessed some unusual properties compared to those of known Lals. Most known Lals show a strict specificity for N-terminal substrates, but the substrate specificity at the C terminus was broader. Hence, by-products such as homopeptides and heteropeptides that had the reverse sequence were generated when peptide synthesis was examined. In contrast, TabS could specifically synthesize one peptide. YwfE also showed specificity at both the N and C termini, and Ala-Gln was synthesized as the main product when Ala and Gln were used as substrates. YwfE preferably accepts amino acids with small residues at the N terminus and amino acids with bulky residues at the C terminus. In other words, amino acids that are categorized into groups with the same characteristics are accepted at each terminus. In contrast, TabS allowed amino acids with various characteristics at both the N and C termini. TabS should allow the selective synthesis of various peptides with various combinations of Gln, Arg, Lys, Tyr, Asn, Pro, Phe, His, Met, or Leu as the N-terminal substrate and Thr, Val, Ile, Ser, Ala, Cys, Trp, or Gly as the C-terminal substrate. Some novel substrate specificities were found: TabS accepted Lys as an N-terminal substrate, β-Ala as an N- or C-terminal substrate, and Pro and its structural analogs as N-terminal substrates. Pro was also acceptable at the C terminus, although the affinity was low.

Our results provide new information about tabtoxin biosynthesis through enzymology. In nature, TβL-Thr is observed, but Thr-TβL, whose sequence is the reverse of that of tabtoxin, is not. When Thr was used as the sole substrate, a trace amount of Thr-Thr was synthesized but Gln-Thr was synthesized in high yield when Gln and Thr were used in the reaction, indicating that Thr is not suitable as an N-terminal substrate. Structural information about YwfE may also support these results. It was revealed that YwfE possesses distinct substrate pockets for the recognition of each N- and C-terminal amino acid; the size, charge, and hydrophobicity in these pockets should determine the acceptance of amino acid substrates at the N and C termini (35). The substrate specificity of TabS might be determined by similar factors. Hence, Thr is suitable at the C terminus but not at the N terminus. In contrast, TβL is suitable at the N terminus but not the C terminus. Furthermore, Walsh et al. showed that the K_m value of TblF, which has the same amino acid sequence as TabS, for TβL is 1.6 ± 0.1 μM and the affinity for TβL is much higher than that for Gln, 70.4 ± 2.3 mM (32). Therefore, TβL should be recognized only as an N-terminal substrate and Thr-TβL and TβL-TβL were not generated. On the other hand, our results indicate that other TβL-Xaa dipeptides could be generated, such as TβL-Ser, TβL-Ala, TβL-Gly, TβL-Cys, TβL-Val, TβL-Ile, and TβL-Trp. TβL-Ser occurs naturally, but others have not been reported, which may be due to lower affinity than that for Thr. Walsh et al. demonstrated 3-fold and 85-fold decreases in catalytic efficiency for Ser and Ala compared to Thr, respectively. In addition, the P_i analysis shown in Fig. 3 indicates that the catalytic efficiency for the other amino acids might be lower than that for Thr, Ser, and Ala. Furthermore, Durbin et al. reported the accumulation of Thr in Woolley's medium after cultivation of P. syringae pv. tabaci (ATCC 11528) (33). We also experimentally confirmed the same phenomenon by using P. syringae NBRC14081. Therefore, the cultivation conditions in Woolley's medium may induce an excess amount of Thr, which increases the tendency to accept Thr at the C terminus of tabtoxin.

Our results also showed that l-pipecolic acid is acceptable as an N-terminal substrate in the reaction at pH 9. TδL is spontaneously generated from TβL and it shares the fundamental structure of a six-membered ring with l-pipecolic acid. Walsh et al. reported that TδL-Thr was not synthesized in HEPES buffer (pH 7), suggesting that TδL is not suitable as a substrate (32). However, the results of the effect of pH (Fig. 2A) showed that the activity at pH 7 is much lower than that at pH 9. Hence, this result is due to low activity of TblF at pH 7, and TblF may potentially accept TδL as a substrate, as well as l-pipecolic acid.

In conclusion, TabS synthesizes dipeptides with high selectivity and accepts various amino acids, including nonproteinogenic amino acids and unnatural amino acids, as a substrate. Since an amino acid that has a β-lactam ring is acceptable, the use of TabS may lead to the generation of new antibiotics. The synthesis of several functional dipeptides was demonstrated. On the basis of structural information that was reported very recently (35, 36), we have attempted to change and improve the substrate recognition of TabS. In the course of screening effective mutations, we obtained mutated TabS that could synthesize Arg-Phe and Gln-Trp in much higher yields. These results will be reported in a future publication. We believe that the need for functional peptides is growing. Hence, the development of a peptide-manufacturing process to address these needs is important. The physiological function of peptides was featured in this paper, but a new use of dipeptides as catalysts was also reported. For example, Pro-Phe, which can be synthesized by TabS, catalyzes an asymmetric aldol reaction (11). The potential of peptides is unlimited. Our findings should make it possible to synthesize various dipeptides at low cost, which then encourages the study of the application of peptides to open the window on new peptide discoveries.