Evidence for an essential histidine in carboxypeptidase Y. Reaction with the chloromethyl ketone derivative of benzyloxycarbonyl-L-phenylalanine.

The possible role of histidine residues in the catalytic function of carboxypeptidase Y from bakers' yeast has been investigated using site-specific reagents. Among the reagents tested, benzyloxy-L-phenylalanylchloromethane (Z-PheCH2Cl) was the most powerful inhibitor of the enzyme. It irreversibly inactivated both the peptidase and esterase activities with an apparent second order rate constant of 3.8 M-minus 1 S-minus 1; the D isomer caused essentially no effect on either activity. Inhibition by L-Z-PheCH2Cl, the reaction retarded by certain competitive inhibitors of the enzyme. Using radioactive L-Z-PheCH2Cl, the reaction with the enzyme was shown to be essentially stoichiometric. Diisopropylphosphorofluoridate (iPr2PF)-inactivated enzyme failed to react with Z-PheCH2Cl, and conversely, the Z-PheCH2Cl-inhibited enzyme failed to react with radioactive iPr2PF. Amino acid analyses of the Z-PheCH2Cl-inactivated enzyme revealed the loss of essentially 1 residue, with a concomitant yield of a 0.62 residue of N-t-carboxymethylhistidine. Since carboxypeptidase Y has a reactive serine at its active center, we concluded from these results that the mechanism involves a charge-relay system in the hydrolysis of peptide and ester substrates, as in chymotrypsin. An -SH group of carboxypeptidase Y was not affected during the reaction with L-Z-PheCH2Cl. The generic name "serine carboxypeptidase" has been proposed for carboxypeptidase Y and for the iPr2PF-sensitive carboxypeptidases from plants, molds, and animal tissues, in order to distinguish them from "metal carboxypeptidase" to which carboxypeptidase A (EC 3.4.12.2) and B (EC 3.4.12.3) belong.

RIKIMARU HAYASHI, YASUO BAI, AND TADAO HATA From The Research Institute for Food Science, Kyoto University, Uji, Kyoto, Japan The possible role of histidine residues in the catalytic function of carboxypeptidase Y from bakers' yeast has been investigated using site-specific reagents. Among the reagents tested, benzyloxy-Lphenylalanylchloromethane (Z-PheCH,Cl) was the most powerful inhibitor of the enzyme. It irreversibly inactivated both the peptidase and esterase activities with an apparent second order rate constant of 3.8 M-' s-l; the D isomer caused essentially no effect on either activity. Inhibition by L-Z-PheCH,Cl was retarded by certain competitive inhibitors of the enzyme. Using radioactive L-Z-PheCH,Cl, the reaction with the enzyme was shown to be essentially stoichiometric.
Diisopropylphosphorofluoridate (iPr,PF)inactivated enzyme failed to react with Z-PheCH,Cl, and conversely, the Z-PheCH,Cl-inhibited enzyme failed to react with radioactive iPr,PF. Amino acid analyses of the Z-PheCH,Cl-inactivated enzyme revealed the loss of essentially 1 residue, with a concomitant yield of a 0.62 residue of N'-carboxymethylhistidine.
Since carboxypeptidase Y has a reactive serine at its active center, we concluded from these results that the mechanism involves a charge-relay system in the hydrolysis of peptide and ester substrates, as in chymotrypsin. An -SH group of carboxypeptidase Y was not affected during the reaction with L-Z-PheCH,Cl.
The generic name "serine carboxypeptidase" has been proposed for carboxypeptidase Y and for the iPr,PF-sensitive carboxypeptidases from plants, molds, and animal tissues, in order to distinguish them from "metal carboxypeptidases" to which carboxypeptidase A (EC 3.4.12.2) and B (EC 3.4.12.3) belong.
Carboxypeptidase Y has been obtained from bakers' yeast and characterized as an enzyme of broad specificity (2)(3)(4)(5). Its ability to release proline is especially useful for structural studies of proteins and peptides (3,6). The active site of this enzyme also appears to be unique. The enzyme has no essential metals (lo), but has a serine hydroxyl at the active center and, thus, differs from the pancreatic carboxypeptidases A and B. The uniqueness of carboxypeptidase Y has also been shown by its strong esterase activity toward the substrates of chymotrypsin, i.e. Ac-Tyr-OEt' (2,3), in contrast to the pancreatic enzymes which hydrolyze only ester substrates with a free carboxyl group in the leaving group, i.e. Bz-Gly-p-phenyllactic acid for A (7), and Bz-Gly-arginic acid for B (8). Thus, carboxypeptidase Y seems to be quite similar to chymotrypsin at its active site and in its mechanism, although the former enzyme is an exopeptidase, whereas the latter is an endopeptidase. When we began this study, however, the following features made it difficult to define unequivocally carboxypeptidase Y as a serine enzyme. First, the enzyme displays activity at acidic pH values (4), whereas serine enzymes usually are active in the alkaline region. Second, as judged from the stoichiometric inhibition by p-HMB and chemical analyses of the enzyme (9), the enzyme has an -SH group, the functional role of which remains to be clarified.
In spite of these differences, however, the high reactivity of a serine residue is common to and the most prominent feature of both carboxypeptidase Y (10) and the serine enzymes (11). Thus, it was advisable to explore the structure of the active center of carboxypeptidase Y using methods analogous to those applied to serine enzymes.
In the serine enzymes, a serine residue of the active center is linked in a charge-relay system with the imidazole ring of a histidine and the carboxylate anion of an aspartic acid enhancing the nucleophilicity of the serine hydroxyl (12). Reaction with site-specific reagents toward the serine enzyme, i.e. iPr,PF and PhCH,SO,F for the serine residue and chloromethyl ketone reagents such as Tos-PheCH,Cl and Tos-LysCH,Cl (13) for the histidine, has consistently shown the mechanism of the charge-relay system.
To explore the possible role of histidine residues in carboxypeptidase Y, its inhibition by chloromethyl ketone reagents has been studied, especially with respect to the effects on the enzyme of the chloromethyl ketone derivative of benzyloxycarnitrobenzoic acid), was also used in the determination with protein bonyl-L-phenylalanine (Z-PheCH,Cl were determined as cysteic acid and methionine sulfone, respectively, after performic acid oxidation. The oxidation was performed at 0' for 8 hours-using the method of Hirs (22). This method has been recommended as preventing the decomposition of histidine derivatives during oxidation (23). N'-Cm-histidine was determined using the color value of glycine derived during the same amino acid analysis (23).

Effects of Various Chloromethyl
Ketone Reagents on Carboxypeptidase Y Actiuity-The effects of chloromethyl ketone derivatives of Tos-Phe, Z-Phe, Z-Ala-Phe, and Z-Ala-Gly-Phe on the peptidase and esterase activities of carboxypeptidase Y were tested as a 20.fold molar excess of reagents to protein for 8 hours of incubation.
After 8 hours, Tos-PheCH,C1,3 Z-PheCH,Cl, and Z-Ala-Gly-PheCH,Cl showed about I%, 5091, and 25% inactivation by both assays, respectively, while Z-Ala-PheCH,Cl had essentially no effect on either activity. The inhibitions apparently followed first order kinetics. The second order rate constants for inactivation were calculated by dividing the apparent first order rate constants by the inhibitor concentrations.
The constants are shown in Table I and compared with those for chymotrypsin (11). Z-PheCH,Cl was the most effective inhibitor for carboxypeptidase Y. Although the inactivation rate was somewhat slower than that with chymotrypsin, it was much faster than that with a model compound, acetylhistidine (4.5 x 10m5 M-' ss') (25). Replacement of the Z group by Tos resulted in a weaker inhibition of the enzyme.
Inhibition by Z-PheCH,Cl: Effects of Its Enantiomer, Reversible Inhibitors, and pH-The first order rate plot for inactivation by the L isomer of Z-PheCH,Cl is shown in Fig. 1. Here, 85% inactivation was observed after 22 hours. Parallel losses in both peptidase and esterase activities occurred. The inactivation was irreversible, as evidenced by the fact that the inactivated enzyme regained no activity when diluted or dialyzed against water. The D isomer of Z-PheCH,Cl caused essentially no inactivation (Fig. 1). The presence of competitive inhibitors of carboxypeptidase Y, Z-D-Phe-D-Leu, Ac-n-Phe-OEt, and trans.cinnamic acid, at concentrations equal to or greater than their respective K, values (5), reduced the inactivation rate by L-Z-PheCH,Cl in both peptidase and esterase assays (Fig. 2).
The pH dependency of the second order rate constant for the inactivation by L-Z-PheCH,Cl revealed a broad bell-shaped curve with maximum at pH 5.5 to 6.5 (Fig. 3).
Stoichiometry and the Site of the Reaction of Z-PheCH,C&When L-[%]Z-P~~CH,C~ was mixed with carboxypeptidase Y in lo-fold molar excess, the radioactivity incorporated increased with time of incubation and in parallel with the disappearance of enzyme activity (Fig. 4). By extrapolation to total inactivation, the radioactivity bound to the enzyme was estimated as 0.9 mol of radioactive Z-PheCH,Cl/ mol of enzyme. This value essentially accounts for a stoichiometric reaction between Z-PheCH,Cl and the enzyme, when one corrects for the loss of activity due to denaturation during the prolonged incubation (see under "Experimental Procedure"). (The loss has tentatively been corrected as a dashed line in Fig. 4.) Amino acid analyses were performed with native and L-Z-PheCH,Cl-treated carboxypeptidase Y after performic acid oxidation.
The latter enzyme lost about 90% of its activity during incubation with Z-PheCH,Cl. Results are shown in Table II. Essentially no difference was observed with respect to the amino acid content of the two enzymes, except for the content of histidine. Carboxypeptidase Y appeared to lose 1 of the 8 histidine residues through the reaction with Z-PheCH,Cl. This was further confirmed as follows.
The yield is much less than 1 mol/mol of the enzyme, since oxidation of the ketone group may occur at either side of the cy carbon atom (26). Performic acid oxidation of Z-PheCH,Cl-inhibited carboxypeptidase Y gave 0.62 residue of N'-Cm-histidine (Table II). Neither N"-nor N""-Cm-histidine was detected on amino acid analyses. Along with the loss of approximately 1 histidine residue, this constitutes evidence that inactivation of the enzyme by Z-PheCH,Cl proceeds through the alkylation of a single histidine residue at the N' position. The enzyme and L-Z-PheCH,Cl were incubated with or without competitive inhibitors. Aliquots were removed to assay peptidase (A) and esterase (Ac-Tyr-OEt, pH-stat method) (B) activities (see Table  I  The second order rate constant for the inactivation of peptidase activity was determined as described in Fig. 1. Buffers used were 0.09 M sodium acetate (pH 4 to 6), sodium phosphate (pH 6 to S), and sodium borate (over pH 9.0).
A methionine residue of cu-chymotrypsin is alkylated by chloromethyl ketone reagents (27). However, amino acid analyses of Z-PheCH,Cl-treated carboxypeptidase Y after performic acid oxidation revealed that the methionine content was practically unchanged, as determined from methionine sulfone ( Table II). The reaction site of chloromethyl ketone reagents has also been reported to be a cysteine residue in thiol enzymes (28)(29)(30)(31). However, the -SH group of carboxypeptidase Y was not affected during the reaction with L-Z-PheCH,Cl, as judged from the Ellman reaction and from determination of S-Cmcysteine (Table II). This was further confirmed by the constant content of cysteic acid in the native and Z-PheCH,Cl enzymes. The tyrosine content was also not affected during the reaction with Z-PheCH,Cl, as judged from amino acid analyses performed on samples without performic acid oxidation.
Amino acid analysis of carboxypeptidase Y from Oriental Yeast showed a slight difference from that of the enzyme from Fleischmann's yeast (3). One less histidine and 2 less methionine residues were found in the former enzyme. with iPr,PF, the reaction of L-['T JZ-Phe-CH,Cl was prevented. If p-HMB was added, or, if the enzyme was heat-denatured, the reaction with radioactive Z-PheCH,Cl was also largely prevented. The appreciable amounts of radioactivity incorporated into the heat-denatured enzyme may be accounted for by nonspecific reactions of [14C]Z-PheCH,C1 with various side chains of the denatured protein (13). DISCUSSION The reaction of Z-PheCH,Cl and Tos-PheCH,Cl toward carboxypeptidase Y is somewhat slower than that toward chymotrypsin; unlike the serine proteases (14,16,(33)(34)(35), the chain length of the chloromethyl ketone reagents (peptide chloromethyl ketones) is not effective in enhancing the inactivation (see Table I). Nevertheless, the following evidence shows that Z-PheCH,Cl reacts with the active center of carboxypeptidase Y in an enzymatically promoted reaction. (a) The optical isomer of L-Z-PheCH,Cl has essentially no effect on enzyme activity, in accord with the stereospecificity of the enzyme (4, 5). (b) Inactivation by Z-PheCH,Cl is reduced in the presence of competitive inhibitors.
(In spite of the small K, value, Leu-Phe had no effect on reducing the inactivation rate, for unknown reason(s).) (c) The pH dependence of the inactivation is similar to the pH dependence of the enzymatic  (13). (e) The iPr,PF-inactivated enzyme fails to react with Z-PheCH,Cl, and, conversely, the Z-PheCH,Cl-inhibited enzyme fails to react with iPr,PF. In addition, the reaction site of Z-PheCH,Cl was unequivocally shown to be the N' position of a histidine residue of the enzyme.
Carboxypeptidase Y has 1 serine residue which is reactive toward iPr,PF at twice the rate of an active serine of chymotrypsin (9, 10) (see also Table I). In addition, the behavior presented above closely parallels the behavior of chymotrypsin. Thus, Z-PheCH,Cl reacts with the enzyme via a noncovalent enzyme-substrate complex to produce an alkylated histidine, as has been explained in detail for chymotrypsin by Baker (36). Furthermore, it is reasonable to assume the presence of a -SH group of carboxypeptidase Y is placed near the active site; hence reaction with p-HMB leads to inhibition due to interference with the substrate binding.' This is supported by the fact that the p-HMB-inhibited enzyme incorporated ["C]Z-PheCH,Cl in appreciable amounts during the prolonged incubation (Table III). Carboxypeptidases from plants (41)(42)(43)(44)(45)(46)(47) and molds (48-50) have recently been shown to be inhibited by iPr,PF and/or PhCH,SO,F, but not by metal chelators. Catheptic enzymes from animal tissues (51,52) also share some of these properties. An active serine has been identified in the carboxypeptidases from the French bean (phaseolain) (53) and from cotton seed (46). Inactivation by Tos-PheCH,Cl has been described with carboxypeptidases from Aspergillus saitoi (48) and oryzae (49). Inactivation by photo-oxidation has been described with a carboxypeptidase from Penicillium (54). Thus, as in carboxypeptidase Y, the serine protease mechanism may also operate in the catalytic process of these enzymes.
The above carboxypeptidases display activity in the acidic pH region; therefore, they seem to belong to the family of "acid carboxypeptidases" originally described by Zuber and Matile (42). The enzymes utilizing the serine protease mechanism should not be referred to as "acid carboxypeptidases", since this term is closely associated with acid proteases. Acid proteases, i.e. pepsin, renin, and a number of mold enzymes, have recently been shown to have a common catalytic center, aspartic acid (55). According to the classification of proteases by Hartley (56) (serine, thiol, metal, and acid proteases), carboxypeptidase Y should be a serine protease, but neither a metal nor an acid protease. The enzyme also differs from thiol proteases. Thus, carboxypeptidase Y and probably the family of enzymes described above should more appropriately be referred to by the generic name "serine carboxypeptidases" to distinguish them from "metal carboxypeptidases" to which carboxypeptidases A and B belong. Carboxypeptidase Y (5 x 10mB~) in 0.1 M sodium phosphate, pH 7.0, was incubated with lo-' M iPr,PF or p-HMB for 30 min. Heat inactivation was performed by immersing the tube in a boiling water bath for 10 min. After these treatments, enzyme activity was completely lost. The enzyme was incubated with 5 x 1Om5 M L-Z-PheCH,Cl in 0.09 M sodium phosphate, pH 7.0, containing 9.1% methanol for 40 hours. The remaining activity was 9%. After these pretreatments, the mixtures were successively incubated with ["C]iPr,PF (1000 cpm/ nmol .of iPr,PF) or L-["C]Z-PheCH,Cl(108 cpm/nmol of Z-PheCH,Cl) under the above conditions. All incubations were performed at 25" in a total volume of 1 ml. Radioactivity was determined after excess inhibitors had been removed by dialysis against water for 30 hours at 5'. Protein concentrations, after dialysis, were determined with a loo-p1 aliquot using the ninhydrin reaction (32) after hydrolysis with 6 N HCl in a vacuum at 110' for 22 hours. charge-relay system in carboxypeptidase Y. This charge-relay system in the enzyme would operate in the enzymatic hydrolysis of both peptide and ester substrates, since inactivations by iPr,PF (10) and Z-PheCH,Cl closely paralleled loss of the two activities in many respects.
Interestingly, chymotrypsin is an endopeptidase while carboxypeptidase Y is an exopeptidase. This difference may arise from features of the active site other than the catalytic center. As an inherent characteristic of carboxypeptidases, carboxypeptidase Y may have a specific binding site for the terminal carboxylate anion of the substrates (4, 5), corresponding to arginine 145 of carboxypeptidase A (37). The pH dependence for the enzymatic hydrolysis of the peptide differs between carboxypeptidase Y and chymotrypsin. Carboxypeptidase Y has two ionizable groups with pK values of 4.4 and 6.5 for the hydrolysis of the peptide substrate, and an ionizable group with pK 5.9 for the hydrolysis of the ester substrate, while chymotrypsin has two ionizable groups with pK values of around 6.8 and 8.5 for the hydrolysis of Ac-Trp-OEt and Ac-Trp-NH, (38). These features are also reflected in the optimum pH for inactivation by Z-PheCH,Cl: pH 5.5 to 6.5 for carboxypeptidase Y, and 7.2 for chymotrypsin (39). These differing features of the pH dependence probably arise from a different environment around the charge-relay system (40). Certain functions have been considered for the -SH group of carboxypeptidase Y; the enzyme is completely inactivated by a stoichiometric reaction of p-HMB (9). The -SH group is not available to react with either iodoacetate or iodoacetamide unless the protein is denatured (9). Furthermore, Z-PheCH,Cl exhibited no reactivity toward the -SH group. These properties are in marked contrast to those of -SH groups of thiol proteases, which are easily alkylated by these reagents. The