Characteristics of a New Enantioselective Thermostable Dipeptidase from Brevibacillus borstelensis BCS-1 and Its Application to Synthesis of a d-Amino-Acid-Containing Dipeptide

Abstract

A new thermostable dipeptidase gene was cloned from the thermophile Brevibacillus borstelensis BCS-1 by genetic complementation of the d-Glu auxotroph Escherichia coli WM335 on a plate containing d-Ala-d-Glu. Nucleotide sequence analysis revealed that the gene included an open reading frame coding for a 307-amino-acid sequence with an M_r of 35,000. The deduced amino acid sequence of the dipeptidase exhibited 52% similarity with the dipeptidase from Listeria monocytogenes. The enzyme was purified to homogeneity from recombinant E. coli WM335 harboring the dipeptidase gene from B. borstelensis BCS-1. Investigation of the enantioselectivity (E) to the P₁ and P₁′ site of Ala-Ala revealed that the ratio of the specificity constant (k_cat/K_m) for l-enantioselectivity to the P₁ site of Ala-Ala was 23.4 ± 2.2 [E = (k_cat/K_m)_l,d/(k_cat/K_m)_d,d], while the d-enantioselectivity to the P₁′ site of Ala-Ala was 16.4 ± 0.5 [E = (k_cat/K_m)_l,d/(k_cat/K_m)_l,l] at 55°C. The enzyme was stable up to 55°C, and the optimal pH and temperature were 8.5 and 65°C, respectively. The enzyme was able to hydrolyze l-Asp-d-Ala, l-Asp-d-AlaOMe, Z-d-Ala-d-AlaOBzl, and Z-l-Asp-d-AlaOBzl, yet it could not hydrolyze d-Ala-l-Asp, d-Ala-l-Ala, d-AlaNH₂, and l-AlaNH_2. The enzyme also exhibited β-lactamase activity similar to that of a human renal dipeptidase. The dipeptidase successfully synthesized the precursor of the dipeptide sweetener Z-l-Asp-d-AlaOBzl.

Many biologically active peptides, including antibiotics (6), peptide sweeteners (13), enkephalins (19), and prodrugs used in chemotherapy, contain d-amino acid residues. Although the kinetically controlled enzymatic synthesis of peptides containing d-amino acids has been investigated with several proteases, including α-chymotrypsin and subtilisin (14, 23, 24), the l-specificity of these enzymes makes it impossible to stereospecifically synthesize peptides that contain d-amino acids. d-stereospecific peptidases are recognized as suitable enzyme catalysts for the synthesis of peptides containing d-amino acids, due to the ability of these enzymes to use racemates as acyl group donors and acceptors, as well as the minimal protection and deprotection step of the protective group. Despite the potential of d-stereospecific peptidases, there have been relatively few reports on the production of peptides containing d-amino acids, except for the carboxypeptidase-catalyzed synthesis of a tripeptide containing d-Ala-d-Ala (11) and the d-aminopeptidase-catalyzed synthesis of a d-Ala oligomer (14) in organic solvents. The major problem with these systems is the difficulty in screening and cloning the enzymes, inasmuch as certain dipeptidases act on dipeptides containing d-amino acids.

The VanX protein, which is involved in vancomycin resistance, has been reported to hydrolyze the peptide bond of d-Ala-d-Ala, although it does not hydrolyze N- or C-terminal-protected dipeptides, such as N-acetyl-d-Ala-d-Ala and d-Ala-d-Ala-o-methyl ester (25), making it inappropriate for the synthesis of d-amino-acid-containing peptides. A dipeptidase from Acinetobacter calcoaceticus (ACDP) has also been found to hydrolyze dipeptides with a d-amino acid at the C-terminus, but it does not hydrolyze C-terminal modified dipeptides (5). To overcome these problems, we recently developed a simple and rapid screening method for microorganisms producing d-stereospecific peptidases (17), and we used this method to identify a thermophile that produces a d-stereospecific peptidase.

In this paper we report on the cloning, expression, and characterization of a thermostable d-stereospecific dipeptidase (BDP) from Brevibacillus borstelensis BCS-1. We also tested the industrial applicability of this enzyme in the production of the d-amino-acid-containing dipeptide Z-d-Asp-d-AlaOBzl as a model system for the synthesis of the dipeptide sweetener alitame (d-Asp-d-AlaNH₂) in an organic solvent system.

MATERIALS AND METHODS

Materials.

Pepstatin, EDTA, and antipain were purchased from Roche Molecular Biochemicals (Mannheim, Germany), and ninhydrin, bestatin, iodoacetic acid, and phenylmethylsulfonyl fluoride (PMSF) were purchased from Sigma (St. Louis, Mo.). The various di- and tripeptides, l-AlaNH₂, l-AlaNH₂, and l-AlaOMe, were purchased from Bachem (Bubendorf, Switzerland); 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide (EDC) and N-hydroxybenzotriazole hydrate (HOBt) were purchased from Aldrich (St. Louis, Mo.); and Z-l-AspOH, Z-l-AlaOH, and l-AlaOBzl-p-tosylate were purchased from Bachem and Sigma. All other chemicals were reagent grade and purchased from Sigma or Aldrich. The Hitrap Q, phenyl-Sepharose, and Mono Q HR 5/5 columns were purchased from Amersham Pharmacia (Uppsala, Sweden).

Bacterial strains and plasmids.

B. borstelensis BCS-1 was cultured at 55°C in Luria-Bertani (LB) medium. E. coli strain WM335 (kindly donated by Walter Messer of the Max-Planck Institute of Molecular Genetics, Berlin, Germany), which requires d-Glu for growth, was cultured at 37°C and used as the cloning host strain (10). E. coli XL1-Blue (Stratagene, La Jolla, Calif.) was used as the expression host strain. The plasmid pUC118 BamHI/BAP, purchased from Bohan Biomedicals Co., Seoul, South Korea, was used as the cloning vector for the dipeptidase gene, whereas pTrc99A purchased from Amersham Pharmacia (Piscataway, N.J.) was used as the expression vector for this gene.

Cloning and sequencing of the BDP gene.

Genomic DNA of B. borstelensis BCS-1 was isolated (21) and partially digested with Sau3AI at 37°C for 10 min. The resulting 2- to 13-kb fragments were purified by using a Gene Clean II kit (Bio-Rad); the size-fractionated DNAs were ligated into BamHI-cleaved, dephosphorylated pUC118 at 14°C for 12 h by using T4 DNA ligase, and the resulting plasmid was electrotransformed into E. coli WM335. After the transformants were cultivated overnight on an LB plate containing 0.2 mM d-Ala-d-Glu, ampicillin, and streptomycin, and in the absence of added d-Glu, the colonies exhibiting d-Ala-d-Glu hydrolytic activity were picked and grown. The resulting plasmid, pBCS8, which contains the open reading frame (ORF) of the full-length dipeptidase gene, was used as the sequencing template, and the resulting nucleotide and protein sequences were aligned by the MEGALIGN (DNASTAR, Inc.) and National Center for Biotechnology Information BLAST search programs.

Construction of a BDP expression plasmid.

The ORF of the cloned BDP gene was amplified from the recombinant plasmid pBCS8 by PCR. The upstream primer, 5′-AACATCATAGATTTTCACTG-3′, was designed for blunt-end ligation with pTrc99A after treatment with NcoI and E. coli DNA polymerase I, whereas the downstream primer, 5′-CCTCTAGAGGATCCTTATTAACGCGGGCGCCGCTG-3′, containing a BamHI restriction site (underlined), was designed from the terminal sequences of the proenzyme. The amplified DNA fragments of approximately 0.9 kb were digested with BamHI and then ligated into pTrc99A such that the BDP gene was under the control of an isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible tac promoter. The resulting plasmid, named pBDP, was used to transform E. coli XL1-Blue, and positive transformants were selected as recombinant BDP producers.

Expression and purification of recombinant BDP.

E. coli XL1-Blue/pBDP was grown in 1 liter of LB medium containing ampicillin (100 μg ml⁻¹) and tetracycline (10 μg ml⁻¹) to an absorbance of 0.4 at 600 nm. The culture was incubated with 0.4 mM IPTG to induce expression of BDP. The cultured cells were harvested by centrifugation at 10,000 × g for 10 min, suspended in 0.1 M Tris-HCl buffer (pH 8.0), and passed through a French press under a pressure of 12,000 lb/in². Cell debris was removed by centrifugation at 10,000 × g for 20 min, and the cell lysate was incubated at 55°C for 30 min to remove any heat-labile E. coli proteins. The denatured E. coli proteins were removed by centrifugation at 15,000 × g for 30 min, and the crude enzyme solution was dialyzed against the same buffer. The dialysate was loaded onto a HiTrap Q column (height, 30 mm; inner diameter, 16 mm) equilibrated with the same buffer, and the bound proteins were eluted with 200 ml of 0.1 M Tris-HCl (pH 8.0) by using a linear gradient of 0.0 to 1.0 M NaCl. The active fractions (20 ml) were pooled, dialyzed, and loaded onto a phenyl-Sepharose column (height, 10 mm; inner diameter, 16 mm) equilibrated with 0.1 M Tris-HCl (pH 8.0)-0.5 M ammonium sulfate, and the enzyme was eluted by using a descending linear gradient of 0.5 to 0 M ammonium sulfate. The active fractions (5 ml) were collected, dialyzed, and reloaded onto a Mono Q HR 5/5 column (height, 50 mm; inner diameter, 5 mm), and the proteins were eluted by using a descending linear gradient of NaCl. The active fractions were collected, dialyzed against a 0.1 M Tris-HCl buffer (pH 8.0), and concentrated by using an Amicon PM-10 ultrafiltration membrane. Protein concentrations were determined by the Bradford method (8) with bovine serum albumin as the standard.

Determination of molecular mass.

The subunit molecular mass of the purified BDP was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under denaturing conditions (18) and using reference proteins (low-molecular-weight electrophoresis protein standards; Bio-Rad).

BDP assay.

Enzyme activity was assayed at 55°C by quantitatively measuring the d-amino acids liberated from the dipeptides with an automated amino acid analyzer (L-8500A; Hitachi, Tokyo, Japan). The enzyme reaction was carried out in a 100-μl reaction volume at 55°C with 50 mM Tris-HCl (pH 8.0) as the buffer, except where otherwise noted.

One unit of enzyme activity was defined as the amount of enzyme that released 1 μmol of d-amino acids from d-amino acid-containing dipeptides per min at 55°C. The kinetic parameters were obtained from the Michaelis-Menten equation with substrate concentrations ranging from 0.1 to 500 mM. The β-lactamase activity was determined by measuring the consumption of substrate in a reaction mixture containing 10 mg of penicillin G in 50 mM Tris-HCl buffer (pH 8.0), and the activity was measured spectrophotometrically at 240 nm (22).

Effect of metal ions and inhibitors.

The effect of metal ions on purified BDP was measured after overnight treatment with 1 mM EDTA at 4°C. The metal-chelated enzyme solution was preincubated with 3 mM concentrations of various metal ions in 0.1 M Tris-HCl (pH 8.0). After incubation for 30 min, the reaction was started by the addition of 10 mM l-Ala-l-Ala at 55°C. To determine the effect of enzyme inhibitors, the purified dipeptidase was preincubated with various protease inhibitors, including EDTA, antipain, bestatin, iodoacetic acid, pepstatin, and PMSF.

Chemical synthesis of Z-l-Asp-l-AlaOBzl and Z-l-Ala-l-AlaOBzl.

For utilization as a dipeptide substrate and standard material, Z-l-Asp-l-AlaOBzl was synthesized by the EDC-HOBt method (16). Five mmol of Z-l-AspOH and 10 mmol of l-AlaOBzl-p-tosylate were dissolved in 1 ml of N,N-dimethylformamide, and 0.55 mmol of N-hydroxybenzotriazole hydrate and 5 ml of EDC were added. The solution was stirred for 10 h, a few drops of water were added, and the solution was allowed to stand at 4°C. The solution was evaporated, and the precipitate Z-l-Asp-l-AlaOBzl was collected.

Z-l-Ala-l-AlaOBzl was synthesized by the same method, but with Z-l-AlaOH and l-AlaOBzl-p-tosylate as the starting materials. The reaction products were purified by reverse phase high-pressure liquid chromatography with an isocratic elution of methanol-water (85:15, vol/vol).

Enzymatic synthesis of Z-l-Asp-l-AlaOBzl.

To synthesize dipeptides containing l-amino acids by using BDP, chemically protected Z-l-AspOH and l-AlaOBzl-p-tosylate were used as the acyl group donor and acceptor, respectively, in the model reaction system. The enzyme reaction was performed in several organic solvents, including water-saturated methyl acetate, ethyl acetate, butyl acetate, diethyl ether, and hexane. BDP was suspended in 0.1 M Tris-HCl (pH 8.0) containing 0.3 mM PMSF and 10 mM Z-l-AspOH and lyophilized. The reaction was initiated by adding the lyophilized enzyme (0.3 U ml⁻¹ for d-Ala-d-Ala) into the water-saturated organic solvents (10 ml of total volume) containing 30 mM d-AlaOBzl-p-tosylate and then continued with vigorous shaking (180 rpm) at 40°C. To analyze the synthesis of Z-l-Asp-d-AlaOBzl, aliquots (50 μl) were withdrawn from the reaction mixture, mixed with 1 N HCl to stop any further enzymatic reaction, and analyzed by high-pressure liquid chromatography with an octadecylsilica column and by measuring absorbance at 254 nm. The elution was carried out with a solvent A-solvent B gradient of 90:10 to 1:100 within 20 min, where solvent A consisted of 0.1% (vol/vol) aqueous trifluoroacetic acid and solvent B consisted of CH₃CN containing 0.1% (vol/vol) trifluoroacetic acid.

Nucleotide sequence accession number.

The nucleotide sequence of the thermostable BDP was deposited in the GenBank nucleotide sequence database under accession number AAF97793.

RESULTS AND DISCUSSION

Cloning of BDP gene from B. borstelensis BCS-1.

DNA fragments from B. borstelensis BCS-1 were partially digested with Sau3AI and ligated into the BamHI site of pUC118 and subsequently used to transform the d-Glu auxotroph E. coli WM335. After overnight cultivation on a plate containing d-Ala-d-Glu, five resultant colonies were grown on a plate containing LB medium; one colony was found to hydrolyze d-Ala-d-Glu. The resulting plasmid pBCS8, containing B. borstelensis BDP, was isolated and sequenced, and an ORF of 921 nucleotides was identified. This ORF was found to encode a polypeptide of 307 amino acid residues, with a deduced molecular mass of 35 kDa.

Alignment with the GenBank database and the BLAST program revealed that the deduced amino acid sequence for B. borstelensis BDP exhibited strong sequence homology to dipeptidases from various other organisms, with the highest similarity to the dipeptidase from Listeria monocytogenes (52.1%; GenBank accession number NP465985) (data not shown).

Linear alignment of the BDP from B. borstelensis BCS-1 and related proteins from other organisms revealed several homologous regions (Fig. 1). Glu125 and His219 residues were highly conserved, being found in other microbial dipeptidases and in renal membrane dipeptidases from higher organisms (1, 2, 3, 4, 12). In addition to active site residues, the RHIDH motif (residues 241 to 246) was also found to be highly conserved, and these residues in renal membrane dipeptidases from pigs and humans are thought bind zinc (4, 15, 19). Most mammalian dipeptidases are zinc-requiring, membrane-bound enzymes. In contrast, BDP activity from B. borstelensis BCS-1 was detected only in cytoplasmic fractions, suggesting that the enzyme is a thermostable dipeptidase that differs from other zinc-requiring dipeptidases.

FIG. 1.

Multiple sequence alignment of B. borstelensis BCS-1 BDP gene and flanking sequences. Highly conserved residues are in black (100% identity), and less strongly conserved residues are in gray (90% identity). The arrows indicate the key active site residues of the enzyme. The proposed zinc-binding motif is indicated within bold round brackets. Proteins (and GenBank accession numbers): BCS-1, thermostable dipeptidase from B. borstelensis BCS-1 (AFF97793); Bhal, dipeptidase from Bacillus halodurans (NP243137); Banth, renal dipeptidase from Bacillus anthracis (NP657741); Lmono, dipeptidase from L. monocytogenes (NP465985); Oih, diepeptidase from Oceanobacillus iheyensis (AC14598); Pab, dipeptidase from Pyrococcus abyssi (NP126634); Ttengco, Zn-dependent dipeptidase from Thermoanaerobacter tengcongensis (NP623690); Atume, diepeptidase from Agrobacterium tumefaciens (NP533190); Mloti, dipeptidase from Mesorhizobium loti (NP104738); Humren, renal membrane dipeptidase from Homo sapiens (NP004404), Ratdip, dipeptidase from Rattus norvegicus (NP446043). The alignment was performed with the alignment software MEGALIGN by the CLUSTAL method. The amino acid sequences were obtained from the GenBank, EMBL, DDBJ, PIR, or SWISS-PROT database.

Overproduction and purification of BDP in E. coli XL1-Blue.

E. coli transformants harboring pBDP were cultivated in LB medium, and production of BDP was induced by IPTG. A 35-kDa protein band, which was anticipated from the deduced molecular weight of BDP, was detected in cell extracts of the transformant. When l-Ala-l-Ala was used as a substrate, the specific activity of the cell extract was 74.8 U mg⁻¹. The dipeptidase activity of wild-type B. borstelensis BCS-1 was about 0.02 U mg⁻¹, indicating that the specific activity of this enzyme was increased about 3,700-fold upon expression in E. coli. We purified 18 mg of BDP to homogeneity, with an overall yield of 33% in terms of activity (Table 1). The specific activity of the enzyme was increased about 4.6-fold compared with that of the cell extract. Purified BDP, which is composed of 307 amino acids with an estimated molecular mass of 34,973 Da, migrated on SDS-PAGE as a single band of about 35 kDa (Fig. 2).

TABLE 1.

Purification of BDP from E. coli XL1-Blue/pBDP^a

Purification step	Total protein (mg · ml−1)	Total activity (U · ml−1)	Specific activity (U · mg−1)	Yield (%)	Purification (fold)
Cell extract	255	19093	74.8	100	1.0
Heat treatment	124	14594	118.2	76	1.6
HiTrap Q column	51	13407	263.4	70	3.5
Phenyl-Sepharose column	33	10863	326.2	57	4.4
Mono Q column	18	6256	343.7	33	4.6

^aAll fractions were assayed by using l-Ala-d-Ala as the enzyme substrate under the conditions described in Materials and Methods.

FIG. 2.

SDS-PAGE of BDP from B. borstelensis BCS-1. The proteins were analyzed by SDS-12% PAGE. Lane M, molecular size marker (phosphorylase b, 94 kDa; albumin, 67 kDa; ovalbumin, 43 kDa; carbonic anhydrase, 30 kDa; and trypsin inhibitor, 20.1 kDa); lane 1, crude extract of E. coli XL1-Blue carrying pBDP; lane 2, soluble fraction of crude extract after heat treatment for 30 min at 55°C; lanes 3, 4, and 5, active fractions eluted from the HiTrapQ, phenyl-Sepharose, and MonoQ columns, respectively.

Characteristics of BDP from E. coli XL1-Blue.

When we assayed the purified BDP by using l-Ala-d-Ala as a substrate, we found that enzyme activity increased with increasing temperature, being optimal at 65°C (Fig. 3). At higher temperatures, however, enzyme activity dramatically decreased. When we assayed the thermostability of this enzyme by measuring residual BDP activity after incubation for 30 min at various temperatures, we found that BDP was stable at temperatures of up to 55°C, but only 50% of the enzyme activity remained after incubation for 30 min at 65°C (Fig. 3). From these results, we determined that BDP enzyme activity was optimal at 65°C, but it was unstable at this temperature. Therefore, all further experiments were carried out at 55°C.

FIG. 3.

Thermal properties of BDP from B. borstelensis BCS-1. Effect of temperature on BDP activity (•) was measured in a 50 mM Tris-HCl buffer (pH 8.0). For determining thermostability of BDP (○), the enzyme activity was measured in 50 mM Tris-HCl (8.0) buffer after heat treatment for 30 min at different temperatures. The values are shown as the relative activity, and the maximum relative activity is indicated as 100%. Each experiment was performed in duplicate.

The pH dependence of enzyme activity was determined between pH 5.5 and pH 10 by using buffers consisting of 50 mM of bis-Tris, Tris-HCl, and 3-(cyclohexylamino)ethanesulfonic acid adjusted to the appropriate pH values. BDP exhibited optimal activity at pH 8.5 but had a relatively broad pH optimum of 7.5 to 10. At less than pH 7.5, however, the activity of this enzyme decreased with decreasing pH, with only 2% of enzyme activity remaining at pH 5.5 (data not shown).

We found that the enzyme activity of BDP was strongly inhibited by metalloprotease inhibitors, such as bestatin and EDTA, but was unaffected by other protease inhibitors, including antipain, iodoacetic acid, pepstatin, and PMSF (data not shown). When we tested the effect of various divalent cations on the ability of BDP to hydrolyze l-Ala-d-Ala, we found that this enzyme was activated about three- to sevenfold by divalent cations, such as Co²⁺ and Mn²⁺, similar to results for mammalian renal dipeptidases (7, 15). In addition, since the mammalian dipeptidases have been found to exhibit β-lactamase activity (9, 20), we tested the ability of BDP to hydrolyze penicillin G. We found that BDP exhibited β-lactamase activity (0.73 U mg⁻¹) on penicillin G without metal ions and that the β-lactamase activity of this enzyme was increased about sixfold (4.3 U mg⁻¹) by the addition of 5 mM MnCl₂ to the reaction mixture.

Kinetic properties and substrate specificity.

When we tested the specificity constant (k_cat/K_m) of BDP on four diastereomers of Ala-Ala, we found that BDP preferentially cleaved l-Ala-d-Ala rather than the other diastereomers (Table 2). The enantioselectivities (E) of the P₁ and P₁′ sites of BDP were determined as the ratio of the specificity constants k_cat and K_m for each site. We found that the k_cat/K_m ratio of l-Ala-d-Ala to d-Ala-d-Ala was 23.4 ± 2.2 [E = (k_cat/K_m)_l,d/(k_cat/K_m)_d,d], but the d-enantioselectivity of the P₁′ site of Ala-Ala was only 16.4 ± 0.5 [E = (k_cat/K_m)_l,d/(k_cat/K_m)_l,l]. Similarly, the d-enantioselectivity of the P₁ site was higher than the d-enantioselectivity of the P₁′ site when dl-Ala-dl-Ala was used as the substrate (data not shown).

TABLE 2.

Kinetic parameters of BDP purified from E. coli XL1-Blue/pBDP

Substrate	kcat (s−1)	Km (mM)	kcat/Km (s−1 mM−1)
l-Ala-d-Ala	760.7 ± 14	3.7 ± 0.2	205.6 ± 12
l-Ala-l-Ala	225.4 ± 9.1	18 ± 1.0	12.5 ± 1.1
d-Ala-d-Ala	271.6 ± 9.8	31 ± 1.2	8.8 ± 0.6
d-Ala-l-Ala	NDa	ND	ND

^aND, not detected.

The ability of BDP to hydrolyze various dipeptides and amides was also determined (Table 3). We found that BDP preferentially hydrolyzed dipeptides with an ld configuration, but not those with a dl configuration, such as d-Ala-l-Asp and d-Ala-l-Ala. BDP was also inactive towards d-AlaNH₂, l-AlaNH₂, and d-Ala-d-Ala-d-Ala, although a very low enzyme activity was detected towards N- or C-terminally protected dipeptides. These results indicate that BDP is an ld-stereoselective dipeptidase.

TABLE 3.

Substrate preference of purified BDP on d-peptides

Substratea	Relative activity (%)b
l-Ala-d-Ala	100
l-Asp-d-Ala	37.8
l-Ala-l-Ala	8.3
l-Asp-d-AlaOMe	5.4
d-Ala-d-Ala-d-Ala	0.2
d-Ala-d-Ala	2.9
d-Ala-d-Glu	2.1
l-Asp-l-Asp	1.1
Z-l-Asp-d-AlaOBzlb	0.002
Z-d-Ala-d-AlaOBzl	0.001
d-Ala-l-Asp	0
d-Ala-l-Ala	0
d-AlaNH2	0
l-AlaNH2	0

^aZ, benzyloxycarbonyl; OBzl, benzyl ester; OMe, methyl ester.

^bThe activity towards l-Ala-d-Ala, corresponding to 326.2 U·mg⁻¹, was taken as 100%.

To determine whether BDP could act on N- or C-terminally protected dipeptides, the dipeptide hydrolysis activity of BDP towards Z-l-Asp-d-AlaOBzl was measured at 55°C. In contrast to other dipeptidases that act on dipeptides containing d-amino acids, including VanX (24) and ACDP (5), BDP did exhibit some hydrolytic activity, but it was very low and took a long period of time to develop (data not shown). These findings suggest that the binding pocket of BDP is flexible, making BDP a potential biocatalyst for the synthesis of the dipeptide sweetener, alitame (l-Asp-d-AlaNH₂).

Enzymatic synthesis of ld-dipeptides with BDP.

The BDP-catalyzed synthesis of Z-l-Asp-d-AlaOBzl in an organic solvent was used as a model for the synthesis of a dipeptide sweetener. For the synthesis of Z-d-Asp-l-AlaOBzl, chemically protected Z-l-AspOH and d-AlaOBzl-p-tosylate were used as substrates. In several organic solvent systems, including methyl acetate, butyl acetate, and diethyl ether, Z-l-Asp-d-AlaOBzl was effectively synthesized, with a yield after 48 h of 14 to 27% (data not shown). The resulting product was identified by electrospray ionization mass spectrometry [Z-l-Asp-d-AlaOBzl, m/z 428 (M + H)⁺ and 451 (M + Na)⁺] (data not shown). Among these solvents, diethyl ether was the most efficient, but the productivity of Z-l-Asp-d-AlaOBzl was too low (approximately 0.2 μmol U⁻¹ h⁻¹). We postulated that this low productivity was due to the low reactivity of BDP on chemically protected substrates. In spite of the low productivity of Z-l-Asp-d-AlaOBzl, BDP from B. borstelensis BCS-1 would appear to be a new thermostable metallopeptidase with β-lactamase activity that may have commercial applications as an enzymatic biocatalyst for the production of dipeptides containing d-amino acids under high temperature reaction conditions (7, 9, 19); we are developing a method for the synthesis of l-Asp-d-AlaOMe from the non-chemically protected substrates l-AspOH and d-AlaOMe by using BDP from B. borstelensis BCS-1 as a biocatalyst.