Aminopeptidase from Sphingomonas capsulata *

A novel aminopeptidase with unique substrate specificity was purified from a culture broth ofSphingomonas capsulata. This is the first reported aminopeptidase to demonstrate broad substrate specificity and yet release glycine and alanine with the highest efficacy. On a series of pentapeptide amides with different N-terminal amino acids, this enzyme efficiently releases glycine, alanine, leucine, proline, and glutamate with the lowest turnover value of 370 min−1 for glutamate. At pH 7.5 (pH optimum) and 25 °C, the kinetic parameters for alaninepara-nitroanilide were found to bek cat = 7600 min−1 andK m = 14 mm. For alanine β-naphthylamide, they were k cat = 860 min−1 and K m = 6.7 mm. Polymerase chain reaction primers were designed based upon obtained internal sequences of the wild type enzyme. The subsequent product was then used to acquire the full-length gene from an S. capsulata genomic library. An open reading frame encoding a protein of 670 amino acids was obtained. The translated protein has a putative signal peptide that directs the enzyme into the supernatant. A search of the amino acid sequence revealed no significant homology to any known aminopeptidases in the available data bases.

A novel aminopeptidase with unique substrate specificity was purified from a culture broth of Sphingomonas capsulata. This is the first reported aminopeptidase to demonstrate broad substrate specificity and yet release glycine and alanine with the highest efficacy. On a series of pentapeptide amides with different N-terminal amino acids, this enzyme efficiently releases glycine, alanine, leucine, proline, and glutamate with the lowest turnover value of 370 min ؊1 for glutamate. At pH 7.5 (pH optimum) and 25°C, the kinetic parameters for alanine para-nitroanilide were found to be k cat ‫؍‬ 7600 min ؊1 and K m ‫؍‬ 14 mM. For alanine ␤-naphthylamide, they were k cat ‫؍‬ 860 min ؊1 and K m ‫؍‬ 6.7 mM. Polymerase chain reaction primers were designed based upon obtained internal sequences of the wild type enzyme. The subsequent product was then used to acquire the fulllength gene from an S. capsulata genomic library. An open reading frame encoding a protein of 670 amino acids was obtained. The translated protein has a putative signal peptide that directs the enzyme into the supernatant. A search of the amino acid sequence revealed no significant homology to any known aminopeptidases in the available data bases.
The rising interest, both basic and applied, in aminopeptidases (EC 3.4.11) from different sources has led to the discovery of a number of enzymes that differ from each other in cellular location, catalytic mechanism, and substrate specificity (1)(2)(3). The majority of bacterial monoaminopeptidases are intracellular or membrane-bound metalloenzymes (1). Based on substrate specificity, bacterial monoaminopeptidases can be divided into two basic categories, specific aminopeptidases, which release only a limited number of amino acids, and those that are able to liberate a relatively broad spectrum of Nterminal amino acid residues (1).
Proline and glycine are among the most difficult residues for aminopeptidases to hydrolyze because of their unique structures. Proline is unusual because of its cyclic structure, and glycine is identified by the lack of a side chain. Nature has developed a family of enzymes that recognize proline exclusively (4). A monoaminopeptidase that preferentially releases glycine with high efficiency has not yet been described and thus would be of high interest.
In this report we describe a novel extracellular monoaminopeptidase from Sphingomonas capsulata. This enzyme has a clear preference for N-terminal glycine and alanine. Because of this characteristic, this monoaminopeptidase has the potential to significantly enhance the degree of protein hydrolysis (5) when used as a supplement to endoproteases and other exopeptidases.

Materials
Chemicals used as buffers and reagents were commercial products of at least reagent grade. para-Nitroanilides of L-amino acids and peptide substrates were from Sigma or Bachem. Pentapeptide amides were synthesized at the Core Laboratories (Louisiana State University). S. capsulata strain IFO 12533 was purchased from the Institute for Fermentation (Osaka, Japan). A Whatman glass microfiber 2.7-m filter and Nalgene Filterware equipped with a 0.45-m filter were used for filtering buffers and supernatants. Protein purification was performed on an Amersham Pharmacia Biotech fast performance liquid chromatography device with column supports and resins from the same. Ultrafiltration units (10-, 180-, and 350-ml) and membranes were from Amicon. The Tricine 1 gels and polyvinylidene difluoride membranes used in the peptide separation and sequencing process were from Novex. The molecular weight of proteins was estimated using Novex Multi-Mark pre-stained and Mark 12 SDS-PAGE markers. Endoproteinase Glu-C (V8 protease) was obtained from Roche Molecular Biochemicals. Assays were performed on a THERMOmax microplate reader, Shimadzu spectrophotometer UV160U, or Hewlett Packard Series 1050 HPLC system with column supports from Vydac, Inc. The protein sequencer used was from Applied Biosystems (model 476A). The sequencing reagents were purchased from PerkinElmer Life Sciences.

Purification of 66-kDa Aminopeptidase
Cultivation of S. capsulata strain IFO 12533 was performed for 15 h at 31°C, 250 rpm, and initial pH value of 7.45 in 1.5 liters of medium composed per liter of 10 g of bactopeptone, 5 g of yeast extract, 3 g of NaCl, 2 g of K 2 HPO 4 , 0.1 g of MgSO 4 ⅐7H 2 O, and 5 g of glucose (autoclaved separately).
The culture broth supernatant (ϳ1 liter) was obtained by initial centrifugation followed by filtration using a Whatman glass microfiber and Nalgene Filterware 0.22-m filters consecutively. The filtrate was concentrated using an Amicon spiral ultrafiltration system equipped with a PM-10 ultrafiltration membrane. The sample was equilibrated with 10 mM phosphate buffer, pH 6.0, until the conductivity and pH value were equal to the loading buffer, 50 mM MES, pH 6.0. The filtered solution was loaded onto a 24 ϫ 390-mm column containing ϳ180 ml of SP-Sepharose fast flow, pre-equilibrated with 50 mM MES buffer, pH 6.0. Protein with aminopeptidase activity was eluted with a 240-ml gradient from 0 to 0.2 M NaCl in 50 mM MES buffer, pH 6.0. Fractions with enzymatic activity toward Ala-pNA were pooled, desalted using a PM-10 membrane, and equilibrated with 20 mM phosphate buffer, pH 7.0.
The pooled solution was then loaded onto a 20-ml Amersham Pharmacia Biotech Mono Q Bead column pre-equilibrated with 20 mM phosphate buffer, pH 7.0. Protein with aminopeptidase activity did not bind to the column and was collected in the flow-through. The flow-through * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The concentrated flow-through was loaded onto a 1.0-ml Amersham Pharmacia Biotech Mono S column that had been pre-equilibrated with 50 mM MES buffer, pH 6.0. The aminopeptidase was eluted with a 60-ml gradient from 0 to 0.2 M NaCl in 50 mM MES buffer, pH 6.0. The fractions with significant activity were then pooled, concentrated, and equilibrated with 50 mM phosphate buffer, pH 7.0, containing 0.5 M (NH 4 ) 2 SO 4 using a PM-10 membrane as above.
Finally, the concentrated sample was loaded onto an Amersham Pharmacia Biotech phenyl Superose 5/5 pre-packed 7 ϫ 50-mm column pre-equilibrated with 50 mM phosphate buffer, pH 7.0, containing 0.5 M (NH 4 ) 2 SO 4 . Protein with aminopeptidase activity was then eluted with a 30-ml gradient from 0.5 to 0 M (NH 4 ) 2 SO 4 in 50 mM phosphate buffer, pH 7.0. Fractions containing aminopeptidase activity were analyzed by SDS-PAGE and then pooled.

Periplasmic Extraction of S. capsulata
A culture broth was grown in a 500-ml shake flask under the conditions described above. A sample was centrifuged, and the whole cells were separated from the supernatant. The harvested cells were then resuspended in 40 ml of 10 mM phosphate buffer, pH 6.0, which contained 1 mg/ml of lysozyme and 0.1% Triton X-100, and stirred for 2 h. The resulting sample was centrifuged, and the supernatant was analyzed.

N-terminal and Internal Amino Acid Sequences of Wild Type Aminopeptidase
SDS-PAGE electrophoresis and transblotting were performed according to the Novex instruction manual. Stained protein bands were excised and sequenced following the instruction manual from PerkinElmer Life Sciences. Data were collected and analyzed on a Macintosh IIsi using Applied Biosystems 610 Data Analysis Software.
To generate peptide fragments of the enzyme to obtain internal sequences, the protein was cleaved with cyanogen bromide by reconstituting a dried sample of the purified protein in 70% formic acid with a few crystals of cyanogen bromide and incubating for 18 h at room temperature in the dark. The peptide fragments were separated by SDS-PAGE electrophoresis using a 10 -20% Novex Tricine gel and sequenced as described above. To generate additional peptide fragments, the purified aminopeptidase was digested using endoproteinase Glu-C (V8 protease). A 200-g sample of purified aminopeptidase was equilibrated in 0.125 M Tris-HCl, pH 6.7, to a final volume of 60 l, and 7.5 l of 0.125 M Tris-HCl, pH 6.7, containing 2.5% w/v of SDS was added. The sample was boiled for 2 min and then allowed to cool to room temperature. Then 10 l of a 400 g/ml solution of endoproteinase Glu-C was added, and the resulting sample was incubated at 37°C for 2 h. To this solution, 45 l of Novex 2ϫ Tricine sample buffer was added. The peptide fragments were separated by SDS-PAGE electrophoresis using a 10 -20% Novex Tricine gel and sequenced as described above.

Metal Content Analysis
The metal content was determined by atomic absorption spectroscopy using a PerkinElmer Life Sciences 2380 atomic absorption spectrophotometer equipped with an HGA-400 graphite furnace. Each measurement was performed in triplicate. The concentration was determined by running a standard curve of known metal concentration.

Physicochemical Characterization
Assay-Aminopeptidase activity was monitored using Ala-pNA as the substrate. A stock solution of 100 mg/ml Ala-pNA in dimethyl sulfoxide was diluted with 50 mM phosphate buffer, pH 7.5, to a concentration of 2 mg/ml. The reaction of the aminopeptidase with the para-nitroanilide was initiated when a 10 -50-l aliquot of the enzyme solution was added to 200 l of the substrate solution in a microtiter plate well. Initial rates of hydrolysis of the para-nitroanilide were monitored at 405 nm at room temperature using a THERMOmax microplate reader.
Substrate Specificity and Inhibition Study-Stock solutions of paranitroanilides of various amino acids in dimethyl sulfoxide (100 mg/ml) were diluted with 50 mM MOPS buffer, pH 7.5, to concentrations of 2 mg/ml. Where the substrates were incompletely soluble, their suspensions were used. The reaction of the S. capsulata aminopeptidase with each para-nitroanilide was initiated when an aliquot (10-l) of the enzyme solution in 50 mM phosphate, pH 7.0, was added to 190 l of a substrate solution in a 96-well microtiter plate. Hydrolysis of the para-nitroanilides was monitored at 405 nm and 25°C.
Enzymatic hydrolysis of pentapeptides (see structures in Table II) was performed at pH 7.5 in 50 mM MOPS buffer at 21°C. Concentrations of the peptides in the incubation mixtures were between 1.10 and 1.14 mM. To stop the reactions, 50-l aliquots of the incubation mixture were mixed with 50 l of 0.1 M HCl. The resulting samples were analyzed for free amino acids by reverse-phase HPLC (6). The concentrations of the peptide solutions were determined using an extinction coefficient of 1440 M Ϫ1 cm Ϫ1 at 280 nm for the tyrosine residue.
The aminopeptidase was incubated for 2.5 h in 50 mM MOPS buffer, pH 7.5, that contained no inhibitor, 1 mM EDTA, 1 mM o-phenanthroline, or 1 mM phenylmethylsulfonyl fluoride. Following the incubation, the enzyme samples were assayed using Ala-pNA as described above.
pH Optimum-An aliquot (20-l) of a stock solution of Ala-pNA in dimethyl sulfoxide (100 mg/ml) was diluted with 980-l aliquots of sodium acetate-Tris-HCl buffer (0.125 M) that had different pH values between 5.0 and 8.5. The resulting pH value of the substrate solutions was measured. A stock solution (0.05 mg/ml)of the aminopeptidase in 50 mM phosphate buffer was diluted 5-fold by 10 mM Tris-HCl buffer, pH 7.5. The reaction mixture contained 200 l of the substrate solution and 10 l of an enzyme solution at room temperature.
Temperature Optimum-An aliquot (970-l) of 50 mM MOPS buffer, pH 7.5, was incubated for 15 min at the chosen temperature, which was maintained using a thermojacket (Shimadzu cell positioner CPS-240A) of the spectrophotometer. Then, 30 l of 100 mg/ml Ala-pNA in Me 2 SO was added. The reaction was initiated by adding 7 l of enzyme solution. Initial velocities were monitored over a 2-min period at 405 nm.
Thermal Stability-An enzyme aliquot (10-l) was added to 190 l of 50 mM phosphate buffer, pH 7.5, which had been preincubated for 30 min at the chosen temperature. The sample was placed on ice after a 20-min incubation. The samples were then assayed at room temperature following the protocol shown above.
Sequential Release of N-terminal Amino Acid Residues from a Natural Peptide-Leucine enkephalin was dissolved in 1 ml of 50 mM MOPS buffer, pH 7.5, to a final concentration of 1 mg/ml. Enzymatic hydrolysis was initiated by 8.3 g of S. capsulata aminopeptidase. After incubation at room temperature (21°C), aliquots of the incubation mixture were added to 0.1 N HCl to terminate the reaction. The free amino acids of the sample were analyzed by reverse-phase HPLC (6).

Cloning
Construction of a Genomic DNA Library-Genomic DNA was isolated from S. capsulata IFO 12533 using a Qiagen Tip-500 column as per the manufacturer's instructions (7). The library was constructed by ligating Sau 3A partially digested (5-7-kilobase) S. capsulata IFO 12533 chromosomal DNA into the BamHI sites of the vector pSJ1678 (8) (Fig. 1) and transformed into Escherichia coli XL1 Blue MR supercompetent cells (Stratagene, Inc.).
Polymerase Chain Reaction Amplification of Aminopeptidase Coding Sequences-The following primers were synthesized based on amino acid sequence data obtained from peptide fragments obtained following cyanogen bromide and V8 protease digestion of the purified aminopeptidase. Forward primer, 5Ј-GCRTCRTANGCRTCNCC-3Ј; reverse primer, 5Ј-ACYTTYTCRTCYTTRTC-3Ј (R ϭ A or G, Yϭ C or T, N ϭ A or G or C or T).
Amplification reactions were prepared in a 50-l volume with 50 pmol of forward and reverse primers, 1 g of S. capsulata IFO 12533 chromosomal DNA as template, 1ϫ polymerase chain reaction buffer (PerkinElmer Life Sciences), 200 M each of dATP, dCTP, dGTP, and dTTP, and 0.5 units of AmpliTaq Gold (PerkinElmer Life Sciences). Reactions were incubated in a Stratagene Robocycler 40 (Stratagene) programmed for 1 cycle at 95°C for 10 min, 35 cycles each at 95°C for 1 min, 44°C for 1 min, and 72°C for 1 min, and 1 cycle at 72°C for 7 min. The resulting product of ϳ190 base pairs was cloned into vector pCR2.1/TOPO as per the manufacturer's instructions (Invitrogen, Inc.).
Identification of Aminopeptidase Clones-The genomic S. capsulata IFO 12533 library was screened by colony hybridization using a polymerase chain reaction-generated probe with the Genius chemiluminescent system (Roche Molecular Biochemicals) as per the manufacturer's instructions.
DNA Sequence Analysis of S. capsulata IFO 12533 Aminopeptidase Gene-DNA sequencing of two aminopeptidase-containing clones, pMRT004.1-7 and pMRT004.1-14, was performed with an Applied Biosystems model 373A automated DNA sequencer (Applied Biosystems, Inc.) on both strands using (a) the Primer Island Transposition method (Applied Biosystems, Inc.) as per the manufacturer's instructions and (b) the primer walking technique using dye-terminator chem-istry (9). Oligonucleotide sequencing primers were synthesized by Operon Technologies, Inc.

Localization and Purification of the Native Enzyme-
The aminopeptidase was obtained from the supernatant of S. capsulata IFO 12533. Periplasmic extraction of the whole cells was also performed, but the enzyme was not present in the extract. It is evident from this result that the S. capsulata aminopeptidase is a secreted enzyme. The purification led to a protein that migrated as a single band of 66 kDa on SDS-PAGE (Fig. 2) Physicochemical Properties of the Enzyme-A specific activity of 105 units/mg was determined for Ala-pNA under the condition described above, assuming that the A 280 of a 1 mg/ml solution of the aminopeptidase is 1.89. The theoretical extinction coefficient of the enzyme was calculated based on the deduced protein sequence (10).
We were able to determine the kinetic parameters for Ala-pNA (k cat ϭ 7600 Ϯ 850 min Ϫ1 , and K m ϭ 14 Ϯ 2 mM) and alanine ␤-naphthylamide (k cat ϭ 860 Ϯ 90 min Ϫ1 , and K m ϭ 6.7 Ϯ 1.1 mM). However, the kinetic parameters for the other amino acid para-nitroanilides and ␤-naphthylamides could not be accurately measured. This can be attributed to a combination of large K m values, as well as poor solubility of the synthetic substrates. In terms of relative activity, the S. capsulata aminopeptidase preferably hydrolyzes alanine para-nitroanilide. It also demonstrates high efficacy on para-nitroanilides of leucine, methionine, glycine, and aspartic and glutamic acids ( Table I).
The lower estimation of turnover numbers of the S. capsulata aminopeptidase on a series of pentapeptide amides with different N-terminal amino acids are shown in Table II. The enzyme exhibited the highest "turnover" on the pentapeptide amide with N-terminal glycine, followed by alanine, leucine, glutamate, and proline.
A study of the hydrolysis of several natural peptides, catalyzed by the S. capsulata aminopeptidase, revealed that the enzyme is capable of hydrolyzing a variety of peptide bonds. Among the bonds most readily hydrolyzed was a Gly-Gly bond (Table III). This is very unusual. The Gly-Gly bond is extremely resistant to enzymatic hydrolysis, probably because of a lack of side chain groups on both the N-terminal and penultimate amino acids. The S. capsulata aminopeptidase also hydrolyzed the peptides YAGFL and EALELARGAIFQA-amide with obvious "bottlenecks" at phenylalanine (data not shown). It is im-portant to stress that it also releases N-terminal proline (Table  II). This aminopeptidase, meanwhile, does not split off N-terminal amino acids with a penultimate proline and thus possesses no proline aminopeptidase activity.
Only o-phenanthroline demonstrated an inhibitory effect among the class-specific inhibitors tested. In this case, the residual activity was found to be 4% of the initial. Neither EDTA nor phenylmethylsulfonyl fluoride influenced the performance of the enzyme.
Certain inorganic anions that form salts with zinc of low solubility were found to have inhibitory properties. Among the  compounds examined were phosphate, ferrocyanide, and iodate. The K i values for these ions were determined to be 3.0, 4.2, and 11 mM, respectively. There is a direct correlation between the K i value and the solubility product of the particular anion with zinc.
A plot of the relative activity of S. capsulata aminopeptidase in the hydrolysis of Ala-pNA gave a typical bell-shaped pH dependence with a sharp optimum at pH 7.2-7.4. An incubation of the enzyme for 20 min at 45 and 50°C led to losses of 40 and 100% activity, respectively. The optimal temperature for the activity was determined to be 43°C.
Sequencing of the Wild Type S. capsulata Aminopeptidase-The N-terminal sequence of the 66-kDa homogeneous protein was determined to be blocked. Digestion of the protein with cyanogen bromide resulted in fragments with molecular masses of 42, 30, 17, 15, 10, 6, and 4 kDa. The N-terminal sequence of the 10-kDa fragment was determined to be AVNG-DAYDADKLKGAITNAKTGNPGAGRPI. The N-terminal sequences of the other bands were inconclusive. The digestion with endoproteinase Glu-C resulted in the generation of peptides with molecular masses of 40, 30, 25, 22, 20, 17, 10, 6, 5, and 4 kDa. The sequence FKDEPNPYDKARMADAKV-LSLFNSLGVTLDKDGKV was obtained from the 22-and 17-kDa peptide fragments. The other bands were either not sequenced, or the results were inconclusive. The obtained protein sequences did not demonstrate homology to any known aminopeptidases.
Sequence Analysis of the DNA Encoding S. capsulata Aminopeptidase-Screening of the S. capsulata IFO 12533 genomic library produced five colonies that exhibited strong hybridization signals with the probe. Two plasmids carrying the aminopeptidase gene were sequenced, and one was confirmed to contain the entire aminopeptidase gene. Sequence analysis of the plasmid containing the aminopeptidase gene revealed an open reading frame of 2010 nucleotides, encoding a protein of 670 amino acids. The GϩC content of this open reading frame was 65%. Based on the rules of von Heijne (11), the first 32 amino acids likely comprise a secretory signal peptide that directs the nascent polypeptide into the periplasm.
The calculated molecular mass of the primary translation product determined from the deduced amino acid sequence of the S. capsulata aminopeptidase was 70.6 kDa, which is consistent with the estimation of 66 kDa based on the mobility of the purified protein on SDS-PAGE. The zinc binding motif HEXXH, which is present in a number of metallopeptidase families (12), was identified in the primary sequence (Fig. 3). A BLAST search of the S. capsulata aminopeptidase against known data bases (EBI, Swiss-Prot, GenBank TM , and GEN-ESEQ) revealed no significant homology to any known aminopeptidases. The highest percent identity (23%) was observed with a hypothetical 67-kDa protein from Synechocystis sp.
The presence of one atom of zinc per enzyme molecule was detected by atomic absorption spectroscopy. This method also indicated four iron atoms per enzyme molecule in a homogeneous preparation of the S. capsulata aminopeptidase. There were no detectable levels of cobalt or calcium.

DISCUSSION
Bacteria hydrolyze different proteins to acquire essential amino acids from the pool of free amino acids and peptides. Certain organisms, such as the nutritionally fastidious Lactococcus lactis, apply a non-direct mechanism and employ a cascade of endo-and exopeptidases to release N-terminal glycine (1,13). Others, such as Xanthomonas citri, have developed a more rational method, producing aminopeptidases with broad substrate specificity (14). In this report, we show that S. capsulata secretes a unique enzyme that preferably liberates Nterminal glycine. Because of a combination of large K m values for both para-nitroanilides and ␤-naphthylamides coinciding with low solubilities for these compounds, it is impossible to carry out a comprehensive study of the substrate specificity of the S. capsulata aminopeptidase utilizing kinetic data for these artificial substrates. Nonetheless, a few remarks can be made. This enzyme apparently discriminates similar amino acids effectively releasing leucine but not isoleucine or valine (Table I); results for alanine para-nitroanilide and ␤-naphthylamide unequivocally show that the structure of the leaving group for a substrate affects both k cat and K m . Taking this into account and expecting that the substrate preference of an aminopeptidase toward derivatives of amino acids, such as para-nitroanilides, and natural peptides can be substantially different (15), the study of the hydrolysis of non-protected peptides catalyzed by the S. capsulata aminopeptidase was warranted. Comparative analysis of the Pro-pNA and Pro-Ala-Pro-Tyr-Lys-amide highlights the misleading role of the para-nitroanilide group. Apparently, it hinders the binding of at least some amino acid residues, for example proline, by the enzyme. More importantly, the S. capsulata aminopeptidase demonstrates an unusual substrate pattern with the order of preference Gly Ͼ Ala Ͼ Leu in terms of relative activity (Table II). A plausible explanation for these features could be a catalytic pocket that is not deep and exhibits very limited flexibility.
There are a few bacterial aminopeptidases described in the literature that demonstrate a reasonable ability to release alanine. The alanine-specific aminopeptidase N from E. coli (17) was not shown to release N-terminal glycine. The bimolecular constant for the thiol aminopeptidase from X. citri was almost 40-fold greater for alanine ␤-naphthylamide in comparison to glycine ␤-naphthylamide (14). It is highly unlikely that this enzyme is capable of cleaving a Gly-Gly bond.
Aminopeptidase from S. capsulata occupies a unique niche among bacterial proteases. This is the first reported enzyme a Turnover was calculated based on an initial velocity of the enzymatic hydrolysis of the pentapeptides and is not equivalent to k cat in this case. It can be considered only as a lower estimation of k cat . Because we do not know values of K m , the concentration of peptides (1.10 -1.14 mM) may be insufficient to provide maximal velocity. that hydrolyzes natural peptides with bimolecular constant values that are similar for glycine and alanine or probably even higher for glycine. This extraordinary substrate preference is undoubtedly exhibited in the hydrolysis of leucine enkephalin catalyzed by the S. capsulata aminopeptidase (Table III). A quick and complete release of tyrosine and both glycine residues from this peptide was observed after a 1-h reaction, yet the Phe-Leu bond was nearly untouched. This clearly demonstrates the substantially higher catalytic efficacy of this enzyme for amino acids with small, rather than large, side chains. A high k cat value, at least 5400 min Ϫ1 , for releasing glycine is also uncommon. Another interesting feature of the S. capsulata aminopeptidase is that, for an unknown reason, it is able to distinguish between similar amino acid residues, like tyrosine and phenylalanine, in the case of non-protected peptides (Table III). Metalloaminopeptidases are predominant in bacteria (1). There are several indirect indications that this is a zinc metalloenzyme. First, effective inhibition of the S. capsulata aminopeptidase by o-phenanthroline, but not by phenylmethylsulfonyl fluoride or p-chloromercuribenzoic acid, was observed. The enzyme is also inhibited by certain anions, whose zinc salts have low solubility product value. We assume that EDTA, another strong chelator of transition metals, shows practically no inhibitory effect because of its voluminous structure and polyanionic nature resulting in its inability to penetrate close enough to the zinc of the catalytic site. A neutral pH optimum (18) for the S. capsulata aminopeptidase and a putative zinc binding domain HEXXH (15) in its amino acid sequence are both indications of a zinc metalloenzyme.
Atomic absorption spectroscopy confirms the presence of one atom of zinc per molecule of enzyme. In addition, four atoms of iron were also detected. A plausible explanation might be nonspecific binding of this metal to the protein molecule.
The results that have been presented prove the scientific novelty of the S. capsulata aminopeptidase. We believe that this enzyme will receive significant attention from the food industry, as well. It has an "industrial" pH optimum, high specific activity, and great performance in releasing alanine and glycine, two amino acids that give considerably strong sweetness (19), in natural peptides.