Amphipathic property of free thiol group contributes to an increase in the catalytic efficiency of carboxypeptidase Y Joji Mima, Giman Jung, Takuo Onizuka, Hiroshi Ueno and Rikimaru Hayashi Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Japan Cys341 of carboxypeptidase Y, which constitutes one side of the solvent-accessible surface of the S1 binding pocket, was replaced with Gly, Ser, Asp, Val, Phe or His by site-directed mutagenesis. Kinetic analysis, using Cbz-dipeptide sub- strates, revealed that polar amino acids at the 341 position increased K m whereas hydrophobic amino acids in this position tended to decrease K m . This suggests the involve- ment of Cys341 in the formation of the Michaelis complex in which Cys341 favors the formation of hydrophobic inter- actions with the P1 side chain of the substrate as well as with residues comprising the surface of the S1 binding pocket. Furthermore, C341G and C341S mutants had significantly higher k cat values with substrates containing the hydropho- bic P1 side chain than C341V or C341F. This indicates that the nonhydrophobic property conferred by Gly or Ser gives flexibility or instability to the S1 pocket, which contributes to the increased k cat values of C341G or C341S. The results suggest that Cys341 may interact with His397 during cata- lysis. Therefore, we propose a dual role for Cys341: (a) its hydrophobicity allows it to participate in the formation of the Michaelis complex with hydrophobic substrates, where it maintains an unfavorable steric constraint in the S1 subsite; (b) its interaction with the imidazole ring of His397 contri- butes to the rate enhancement by stabilizing the tetrahedral intermediate in the transition state. Keywords: amphipathic property; carboxypeptidase Y; substrate-binding site; tetrahedral intermediate; thiol group. Carboxypeptidase Y from Saccharomyces cerevisiae,which is localized in the vacuole and involved in the C-terminal processing of peptides and proteins, belongs to the serine carboxypeptidase family. It has a catalytic triad (Ser, His, Asp) which constructs the charge-relay system at the active center and exhibits peptidase and esterase activities with broad substrate specificity [1–4]. Chemical modification studies have assigned the essential serine and histidine residues to positions 146 and 397, respectively [1–3]. A third member of the catalytic center, aspartic acid, was putatively assigned to position 338 based on structural homology with carboxypeptidase II from wheat (CPW-II) [4]. Cys341 is the only one of the 11 Cys residues of carboxypeptidase Y that is present as a free thiol group. This single cysteine residue is conserved among the single- chain serine carboxypeptidases; however, its role in the catalytic mechanism and substrate binding remains unclear and it has been the target of various chemical modification studies [1,5,6]. Iodoacetate and iodoacetamide do not react with Cys341 unless the enzyme is denatured [1]. On the other hand, p-hydroxymercuribenzoate is able to react with Cys341 to give an inactive enzyme that is no longer reactive with di-isopropyl phosphorofluoridate, a modifier of the essential Ser146. The effects of alkyl and aromatic mercurial compounds on carboxypeptidase Y activity have raised the question whether or not Cys341 plays an essential role in the hydrolytic reaction [5,6]. Attempts have been made to clarify the role of Cys341 by using site-directed mutagenesis techniques [7,8]. Winther & Breddam [7] prepared mutant enzymes in which Cys341 was replaced by Ser, Gly, Gln, Glu, His or Lys. A number of their mutant enzymes exhibited reduced activity toward a wide range of dipeptide and ester substrates. These chemical modification and site-directed mutagen- esis studies indicate that Cys341 is located at the substrate- binding site in the hydrophobic environment, and also suggest that Cys341 may be involved in the catalysis in a manner other than substrate binding. However, the precise type of catalytic event affected by Cys341 remains unclear. X-ray crystallographic studies of Endrizzi et al. [9] have shown that carboxypeptidase Y contains clearly defined substrate-binding sites, S1¢ and S1 [10], each of which binds the C-terminal (P1¢) and penultimate (P1) side chain of the substrate, respectively. The S1¢ and S1 subsites are located between two hydrophobic depressions separated by a Ser146–His397 diad. The crystal structure of p-chloromer- curibenzoic acid-modified carboxypeptidase Y has also revealed that Cys341, along with other hydrophobic residues, comprises the solvent-exposed surface in the S1 subsite. The other residues include Tyr147, Leu178, Tyr185, Tyr188, Trp312, and Ile340 [9] (Fig. 1). The SH group of Cys341 is also situated in the vicinity of the side chains of Leu178, Ile340, and the essential His397 [9]. In these previous studies, the importance of the hydro- phobic property of Cys341 has been neglected. In general, when an SH group is fully protonated, it is able to form hydrophobic interactions with neighboring hydrophobic residues and aids in creating a hydrophobic environment [11,12]. To address this issue, we constructed six mutants by site-directed mutagenesis, in which Val and Phe mutants are Correspondence to J. Mima, Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Kitashirakawa, Sakyo-ku, Kyoto 606-8502, Japan. Fax: + 81 75 7536128, Tel.: + 81 75 7536125, E-mail: mima@kais.kyoto-u.ac.jp Enzyme: carboxypeptidase Y (EC 3.4.16.5). (Received 15 January 2002, revised 13 May 2002, accepted 15 May 2002) Eur. J. Biochem. 269, 3220–3225 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02997.x designed to maintain hydrophobicity and Gly, Ser, Asp and His mutants to provide hydrophilicity. The catalytic roles of Cys341 in carboxypeptidase Y are examined by comparing the kinetic parameters of the Cys341 mutant enzymes. MATERIALS AND METHODS Materials Cbz-Phe-Leu-OH was obtained from Fluka Chemie AG, Buchs, Switzerland. Cbz-Ala-Phe-OH was from Sigma Chemical Company, St Louis, MO, USA. Cbz-Gly-Phe-OH and Cbz-Gly-Leu-OH were from the Peptide Institute Inc., Osaka, Japan. The synthetic oligonucleotides were obtained from Japan Bio Services, Saitama, Japan. Restriction endonucleases and T4 polynucleotide kinase were purchased from Toyobo, Osaka, Japan. The Transformer TM site- directed mutagenesis kit was purchased from Clontech, Palo Alto, CA, USA. The Taq dyedeoxy TM terminator cycle sequencing kit was obtained from Applied Biosystems, Foster City, CA, USA. DEAE-Sephadex A-50 was from Pharmacia Fine Chemicals, Uppsala, Sweden. Hydroxyl- apatite gel was purchased from Bio-Rad, Hercules, CA, USA. Bistris was obtained from Nacalai Tesque, Kyoto, Japan. All other chemicals were of reagent grade and obtained locally. Strains and plasmid DNA The plasmid pTSY3 containing the PRC1 gene coding for carboxypeptidase Y and S. cerevisiae SEY2202 (MATa Dprc1::(LEU2) leu2-3,112ura3-52hi4-519)were kindly provided by Dr Klaus Breddam, Carlsberg Labor- atory, Copenhagen, Denmark. Escherichia coli JM109 (recA1 supE44 endA1 hsdR17 gyrA96 relA1 thiD(lac-proAB) F¢ [traD36 proAB + laqI9 lacZDM15]) was from the in-house collection. Site-directed mutagenesis In vitro mutagenesis was performed with the pTSY3 subclone of the PRC1 gene [13]. Three oligonucleotides were used for mutation: a 25-mer as the mutagenic primer to introduce other amino acids (Gly/His/Ser/Val/Asp/Phe) for Cys341 (sequence 5¢-AAGATTTCATCGGT/CAT/TCT/ GTT/GAT/TTCAACTGGTTGGG-3¢);a26-merasthe selection primer (5¢-ACTACAAAATGAGCTCCCTCGC GCGT-3¢) to introduce a PvuII restriction site into plasmid pTSY3 for selection; and a 20-mer sequence primer for DNA sequencing (5¢-GCCATGGAAGTACGACGA AG-3¢). The mutation was performed with a Transformer site-directed mutagenesis kit as described by Deng & Nickoloff [14]. The DNA sequence reaction was performed with a Taq dyedeoxy terminator cycle sequencing kit. Yeast strain SEY2202 was transformed by the lithium acetate method [15]. E.coliJM109 was transformed by the method of Hanahan [16] using a standard transformation buffer. Purification of wild-type and mutant forms of carboxypeptidase Y Typical purifications were carried out as previously des- cribed [17]. An additional purification step of hydroxyl- apatite column chromatography was performed essentially as described by Bernardi [18]. The preparations, at 4 °C, were loaded on to a hydroxylapatite column equilibrated with 75 m M sodium phosphate buffer, pH 6.8, and both wild-type carboxypeptidase Y and mutant carboxypepti- dase Y were eluted with 150 m M buffer. The enzyme solutions were desalted on ultrafiltration apparatus (Amicon stirred cells model 8050) by repeated concentration and dilution with water. Enzyme activity was assayed with N-benzoyl- L -tyrosine-p-nitroanilide during the purification steps [19]. The purity of the refined enzymes was verified by SDS/PAGE. CD measurements CD spectra were measured on a JASCO J-720 W spectro- polarimeter at room temperature. Ten scans were averaged for the wild-type and mutant forms of carboxypeptidase Y at concentrations of % 3.0 l M in 10 m M sodium phosphate, pH 7.0. Kinetic characterization Peptidase activities were measured in 50 m M Bistris buffer/1 m M EDTA,pH6.5,at25°C. Enzyme concen- trations of the wild-type and mutant carboxypeptidase Y in the hydrolysis reaction were 32.8 n M . Substrate con- centrations were 0.02–20 m M . The reaction was termin- ated and deproteinized by adding 330 lL0.4 M sulfosalicylic acid per mL of the enzyme/substrate solu- tion. The initial rates of hydrolysis of dipeptide substrates were measured by quantitating the amount of C-terminal amino acid released, using a Jeol JLC-300 amino-acid analyzer. The kinetic parameters for the hydrolysis of various Cbz-peptide substrates were derived from Hanes- Woolf plots. Fig. 1. Catalytic triad (Ser146, His397, and Asp338) and S1 binding site of carboxypeptidase Y. Cys341 constitutes the left side of the solvent- accessible surface with Ile340 viewed from the point of the active Ser- Hisdiad.ThesulfuratomofCys341islocatedwithin5A ˚ of the imidazole ring of His397. Ó FEBS 2002 A single thiol group in carboxypeptidase Y (Eur. J. Biochem. 269) 3221 Calculation of energy levels of intermediates and apparent binding energies The hydrolysis of the peptide substrates catalyzed by carboxypeptidase Y is described in Scheme 1. For typical serine protease catalysis, an acylation step is the rate- determining step (k 2 ( k 3 ); therefore, K m % K s and k 2 % k cat are assumed [20]. The energy level, DG,ofeach enzyme state was calculated from the following thermody- namic equations [20] and is shown relative to the free enzyme defined as DG ¼ 0(Table3): DG s ¼ RT lnK s % RT lnK m ; DG ‡ ¼ RT lnðk B T=hÞÀRT lnk 2 % RT lnðk B T=hÞÀRT lnk cat ; DG ‡ T ¼ RT lnðk B T=hÞÀRT lnðk 2 =K s Þ % RT lnðk B T=hÞÀRT lnðk cat =K m Þ; DDG ‡ ¼ DG ‡ ðmutantÞÀDG ‡ ðwild-typeÞ where DG s is the binding energy of the substrate to the enzyme, DG à is the activation energy in the chemical steps of bond making and breaking, DG ‡ T is the activation energy for the free enzyme reacting with the free substrate to give products, R is the gas constant, T is the absolute tempera- ture, k B is the Boltzmann’s constant, h is Planck’s constant. RESULTS Purification of mutant carboxypeptidase Y Mutant carboxypeptidase Y in which Cys341 is replaced by Gly, Ser, Asp, Val, Phe, or His residues was isolated and purified from % 150 g yeast cells (Table 1). Single bands on SDS/PAGE analysis verified the purity of all mutants. The protein yields of C341G, C341S and C341D were similar to that of the wild-type, whereas the yields of C341V, C341F and C341H were lower. The specific activities of all the mutants were decreased by more than 10-fold compared with the wild-type, with C341D and C341F showing more than a 100-fold decrease and C341H was nearly inactive. Properties of mutant carboxypeptidase Y The effect of the amino-acid substitution at Cys341 on secondary structure was evaluated by analyzing the CD spectra (Fig. 2). All mutant enzymes except C341H had a spectrum identical with that of the wild-type enzyme. The different CD spectrum for C341H suggests that the introduction of a positive charge at 341 causes some alteration in the secondary structure. Because of the poor yield and altered CD spectrum, C341H mutant was not analyzed further. Kinetic properties of mutant carboxypeptidase Y The effects of amino-acid substitution on the enzymatic properties of the Cys341 mutant enzymes were investigated using two sets of substrates, Cbz-X 1 -Leu-OH and Cbz-X 2 - Phe-OH, where X 1 was a Gly or Phe residue and X 2 was a Gly or Ala residue (Table 2). Mutant enzymes had reduced k cat /K m values with all four substrates compared with the wild-type enzyme, except that C341G had a somewhat increased value with Cbz-Ala-Phe. Mutant enzymes exhibited similar k cat /K m values with Cbz-Gly-Leu and Cbz-Gly-Phe, although C341G exhibited higher k cat /K m values than the other mutants with Cbz-Gly- Leu. The k cat values of C341G, C341S and C341F were largely higher than that of the wild-type. A nearly identical profile was observed for K m values with Cbz-Gly-Leu and Scheme 1. Table 1. Yields of wild-type carboxypeptidase Y and Cys341 mutants on purification from 150 g yeast cells. Enzymes were purified by the method of Hayashi et al. [17]. Activity toward N-benzoyl- L -tyrosine- p-nitroanilide was determined [19]. ND, not detected. Enzyme Total protein (mg) Total activity (units) 10 )3 · Specific activity (unitsÆmg )1 ) Wild-type 3.9 3.4 860 C341G 3.3 0.24 72 C341S 4.9 0.31 62 C341D 4.6 0.014 3.0 C341V 1.0 0.032 32 C341F 1.0 0.0053 5.1 C341H 0.41 ND – Fig. 2. CD spectra of wild-type carboxypeptidase Y and Cys341 mutants. Protein concentrations of carboxypeptidase Y and its mutants were % 3.0 l M in 10 m M sodium phoshate, pH 7.0. Condi- tions for measurements: band width 1.0 nm; sensitivity 50 millidegrees; response 0.5 s; wave length 190–250 nm; scan speed 100 nmÆmin )1 ; step resolution 1 nm; 10 measurements were made. (d) Wild-type carboxypeptidase Y; (m) C341G; (h) C341S; (j) C341D; (r) C341V; (s) C341F; (e) C341H. 3222 J. Mima et al.(Eur. J. Biochem. 269) Ó FEBS 2002 Cbz-Gly-Phe. This suggests that, when glycine occupies the S1 subsite, the binding preference of the S1¢ subsite for hydrophobic amino acid is not absolute, but is still affected by the mutation at position 341. Mutant enzymes with a hydrophobic residue at the 341 position, i.e. C341V and C341F, had lower K m values with Cbz-Phe-Leu and Cbz-Ala-Leu than the wild-type enzyme. This suggests that a hydrophobic interaction at the S1 pocket is important for substrate binding. On the other hand, hydrophobic residues at position 341 significantly decreased the k cat values (20–90-fold decreased compared with the k cat value of the wild-type carboxypeptidase Y), while values for C341G and C341S were decreased only 1.5-fold to sixfold over that of the wild-type enzyme. C341G exhibited higher k cat /K m values with all substrates than the other mutants. This value was even higher than the wild-type when Cbz-Ala-Phe was used as a substrate. The characteristics of the C341S mutant were similar to those of C341G but its binding ability was slightly less. Both C341G and C341S maintained a relatively high enzyme activity. These results for C341G and C341S with hydrophobic dipeptide substrates are in agreement with the results obtained in previous work [7]. DISCUSSION Effect of replacement of Cys341 on kinetic constants The introduction of Gly at position 341 would be expected to reduce any structural constraints at the S1 subsite because of elimination of the bulky SH group. It has been suggested that eliminating steric constraints present in the S1 subsite of the wild-type carboxypeptidase Y would increase the activ- ity of the Leu178 mutant carboxypeptidase Y toward substrates with the basic P1 side chains [21]. It was assumed that the lack of a side chain at position 341 would make the S1 pocket unstable because of the elasticity introduced. The size of the S1 subsite may also be reduced as the result of shrinkage of the hydrophobic side chains in the surface of the S1 pocket which is exposed to solvent. We were able to engineer the P1 preference of carboxypeptidase Y from a bulky hydrophobic side chain, i.e. Phe, to a small hydro- phobic residue, i.e. Ala, on the C341G mutant carboxy- peptidase Y. It was evident that the k cat /K m value with Cbz-Ala-Phe-OH increases up to 556-fold relative to the hydrolysis of Cbz-Gly-Phe-OH, whereas only a 27-fold increase was obtained with Cbz-Phe-Leu-OH relative to Cbz-Gly-Leu-OH (Table 2). It is also shown that C341G narrows the P1 amino-acid preference as it no longer possesses a wild-type-like preference for a bulky hydropho- bic amino acid. Of the mutant enzymes, C341G had the highest k cat values with respect to substrates Cbz-Gly-Leu and Cbz-Ala-Phe (Table 2). Although detailed structural evaluation is needed, the reason for C341G exhibiting high k cat values is probably the increased elasticity at the S1 subsite. C341S behaves in a similar manner to C341G in its kinetic profile, except that it tends to have higher K m .This low affinity of C341S can be attributed to the water molecule(s) co-ordinated to the solvent-accessible surface of the S1 subsite via the hydroxy group. This additional interaction with water molecule(s) may inhibit the forma- tion of the Michaelis complex (Scheme 1), which results in increased K m values (Table 2). It was also assumed in a previous report [7] that the higher K m values of C341S are due to co-ordination of water molecule(s) around the hydrophilic side chain. When the Michaelis complex is formed with substrates that have hydrophobic P1 side chains, the surface of the S1 subsite may become inaccessible to the solvent, which would cause reorientation of the side chain of Ser directly away from the hydrophobic surface of the S1 subsite. In the transition state, the steric environment of the S1 pocket in C341S may be almost identical with that of C341G, which explains the similar catalytic characteris- tics of C341S and C341G (Table 2). A hydrophilic group, such as Asp or Ser, at the S1 subsite has a tendency to reduce its affinity for substrates. The substitution of the Cys341 with a negatively charged amino acid led to an increase in K m . Mutant enzymes in which Cys341 was replaced with Glu and Gln had higher K m values than wild-type carboxypeptidase Y [7]. This suggests that formation of hydrogen bonds or electrostatic interac- tions at the S1 subsite may not be involved in the substrate- recognition mechanism. In general, a thiol group does not undergo deprotona- tion when it is in a hydrophobic environment [11,12]. Therefore, we postulate that Cys341, with a fully proto- nated SH group, maintains hydrophobic interactions with neighboring amino-acid residues. C341V, which has a hydrophobic residue almost identical with cysteine in size at position 341, would be predicted to show kinetic constants similar to the wild-type carboxypeptidase Y. Indeed our results show that C341V and wild-type Table 2. Kinetic parameters of wild-type carboxypeptidase Y and Cys341 mutants for hydrolysis of Cbz-dipeptide. ND, not determined because K m values exceeded the applicable substrate concentration range. Enzyme Substrate k cat (s )1 ) K m (m M ) k cat /K m (s )1 Æm M )1 ) Wild-type Cbz-Gly-Leu 3.2 0.83 3.8 Cbz-Phe-Leu 92 0.19 480 Cbz-Gly-Phe 0.96 0.41 2.3 Cbz-Ala-Phe 150 1.1 140 C341G Cbz-Gly-Leu 7.0 3.2 2.2 Cbz-Phe-Leu 15 0.26 59 Cbz-Gly-Phe 1.1 3.7 0.30 Cbz-Ala-Phe 95 0.57 170 C341S Cbz-Gly-Leu 1.9 5.2 0.37 Cbz-Phe-Leu 18 0.32 57 Cbz-Gly-Phe 2.4 6.7 0.36 Cbz-Ala-Phe 93 4.2 22 C341D Cbz-Gly-Leu 1.4 9.1 0.15 Cbz-Phe-Leu 2.1 3.0 0.71 Cbz-Gly-Phe ND ND 0.15 Cbz-Ala-Phe ND ND 3.4 C341V Cbz-Gly-Leu 0.48 0.89 0.54 Cbz-Phe-Leu 4.6 0.16 29 Cbz-Gly-Phe 0.16 0.45 0.36 Cbz-Ala-Phe 8.0 1.1 7.5 C341F Cbz-Gly-Leu 2.3 8.4 0.27 Cbz-Phe-Leu 1.2 0.070 25 Cbz-Gly-Phe 5.5 19 0.28 Cbz-Ala-Phe 2.2 0.98 2.3 Ó FEBS 2002 A single thiol group in carboxypeptidase Y (Eur. J. Biochem. 269) 3223 carboxypeptidase Y have similar K m values with all substrates examined (Table 2). This provides support for the hypothesis that Cys341 is not only located at the S1 binding site [5–7,9], but the hydrophobicity of the thiol group also plays a role in substrate binding and interacting with the hydrophobic residues of the S1 subsite. However, the k cat values of C341V were much lower than those of the other mutant carboxypeptidase Y, i.e. C341G or C341S (Table 2). It is possible that the hydrophobic interaction of Val with other side chains in the S1 pocket or P1 side chain of the substrates in the Michaelis complex inhibits the acylation step in the hydrolysis reaction (Scheme 1). It can be assumed that the hydrophobic interaction of Val in the S1 subsite is stabilized so that the rate of acylation becomes signifi- cantly reduced. In the case of the mutant enzymes with bulky hydrophobic residues at position 341, i.e. C341F, the hydrophobicity at position 341 appears to be import- ant for substrate binding, although it may not necessarily have any effect on the rate of acylation. Roles of Cys341 in the catalytic mechanism of carboxypeptidase Y Before the formation of the Michaelis complex, the free thiol group of Cys341 participates in a hydrophobic bond network on the solvent-accessible surface of the S1 subsite, which is constructed of Tyr147, Leu178, Tyr185, Tyr188, Trp312, and Ile340 (Fig. 1). Thus, the thiol group of Cys341 may participate in controlling the depth and width of the S1 solvent-accessible cavity, where a bulky hydrophobic P1 side chain such as phenylalanine can be accommodated. At the time the Michaelis complex is established, the free thiol group of Cys341 is located in close proximity to the P1 side chain and the hydrophobic interaction with hydrophobic P1 side chain is in effect. In fact, our results for C341V and C341F provide support for a scenario in which the hydrophobicity of the SH group of Cys341 is important for substrate binding and maintaining the solvent-accessible cavity at the S1 subsite. However, in the case of the transition state in the acylation reaction, the role of Cys341 as a hydrophobic residue should be altered because C341V, which is solely hydrophobic in nature, exhibits a lowered k cat . This suggests an additional role for this thiol group. Free-energy parameters derived from the kinetic analysis aresummarizedinTable3forthewild-typeandC341V mutant enzymes. This result can be visualized in the form of a diagrammatic scheme shown in Fig. 3. A characteristic of C341V is its increased activation energy, DG à , at the stage where the tetrahedral intermediate is formed, the step that has the most influence on the rate of acyl enzyme formation. We are concerned as to why C341V exhibits an increased DG à compared with the wild-type enzyme. As the properties of Val, such as size and hydrophobicity, are similar to those of Cys and the affinity of C341V for the substrates tested are similar to that of the wild-type enzyme, there must be some explanation for the difference in DG à . What is the specific role of Cys341 in the transition state? In the tetrahedral transition state, Cys341 is located adjacent to His397 of the catalytic center [9], the imidazole nitrogen of which is positively charged (Fig. 4). The thiol group of Cys341 is located within 5 A ˚ of the imidazole nitrogen of His397. Therefore, it is reasonable to assume that the sulfur atom of Cys341 becomes polarized by the positive charge on the imidazole nitrogen of His397. This newly created electrostatic interaction may cause redirection of the cystei- nyl side chain from the surface of the S1 subsite toward the charged imidazole nitrogen (Fig. 4). SuchanalterationinthesidechainofCys341inthe tetrahedral intermediate may give structural flexibility to the S1 subsite, which reduces steric constraint at the S1 subsite: elimination of the side chain of Cys341 from the S1 solvent- accessible surface weakens a hydrophobic bond network, which maintains the depth and width of the S1 subsite cavity. The flexibility introduced wouldbe used to create asubstrate- Table 3. Gibbs free energies (kJÆmol )1 ) of complexes of wild-type carboxypeptidase Y and C341V mutant. DG S is algebraically negative and DG à and DG ‡ T positive. The activation energy (DG à ) of wild-type carboxypeptidase Y is lowered by DDG à compared with that of C341V. Substrate Enzyme DG ‡ s DG à DG ‡ T DDG à Cbz-Phe-Leu Wild-type )21 62 40 – C341V )22 69 48 7.5 Cbz-Ala-Phe Wild-type )17 59 41 – C341V )17 66 49 7.5 Fig. 3. Free-energy profiles for the formation of the tetrahedral inter- mediate by wild-type (energy level in heavy line) and C341V (in light line) carboxypeptidase Y. Fig. 4. Model of alteration in the S1 subsite in the transition state through interaction between Cys341 and His397 making it comple- mentary with the P1 side chain of the substrate. 3224 J. Mima et al.(Eur. J. Biochem. 269) Ó FEBS 2002 binding pocket that is complementary in size to various hydrophobic P1 amino acids (Fig. 4) and thereby increase the interaction energy of the tetrahedral intermediate with the substrate. Because C341V lacks the ability to interact with the protonated His397, this mutant enzyme is inefficient at utilizing the increased interaction energy to stabilize the tetrahedral intermediate in the transition state (Fig. 3). The described flip-flap motion of Cys341 explains our experimental data and leads us to propose that Cys341 may have two distinct roles in the catalytic mechanism: (a) the hydrophobicity of the thiol group (Cys341) is involved in substrate binding at the S1 subsite and in maintaining the width and depth of the S1 subsite; (b) the rearrangement of the S1 subsite induced by the interaction between the SH group and the imidazole nitrogen of His397 stabilizes the transition state. X-ray crystallography of the mutant enzymes and their complexes with the transition state analog to confirm the hypothesis of a dual role for Cys341 is underway. We suggest that the proposed function is common to free thiol groups adjacent to active histidine residues found in the monomeric serine carboxypeptidases including carb- oxypeptidase S1 from Penicillium janthinellum, carboxy- peptidase MIII from barley malt, and Kex1p from S. cerevisiae [22]. As a number of subtilisin-like serine endoproteases, e.g. proteinase B from S. cerevisiae,prot- einase K from Tritirachium album, and thermitase from Thermoactinomyces vulgaris, have free thiol groups in the vicinity of the catalytic triad (Ser, His, Asp) [7,23], these cysteine residues may also have the putative dual role in the hydrolysis reactions. The present hypothesis provides fur- ther insights, with the revelation of the new residues at subsites that contribute to catalytic efficiency in the transition state. REFERENCES 1. Hayashi, R., Moore, S. & Stein, W.H. (1973) Serine at the active center of yeast carboxypeptidase. J. Biol. Chem. 248, 8366–8369. 2. Hayashi, R., Bai, Y. & Hata, T. (1975) Further confirmation of carboxypeptidase Y as a metal-free enzyme having a reactive serine residue. J. Biochem. (Tokyo) 77, 1313–1318. 3. Jung, G., Ueno, H., Hayashi, R. & Liao, T H. (1995) Identifi- cation of the catalytic histidine residue participating in the charge- relay system of carboxypeptidase Y. Protein Sci. 4, 2433–2435. 4. Liao,D I.,Breddam,K.,Sweet,R.M.,Bullock,T.&Remington, S.J. (1992) Refined atomic model of wheat serine carboxy- peptidase II at 2.2-A ˚ resolution. Biochemistry 31, 9796–9812. 5. Bai, Y. & Hayashi, R. (1979) Properties of the single sulfhydryl group of carboxypeptidase Y. Effects of alkyl and aromatic mer- curials on activities toward various synthetic substrates. J. Biol. Chem. 254, 8473–8479. 6. Breddam, K. (1983) Modification of the single sulfhydryl group of carboxypeptidase Y with mercurials. Influence on enzyme specificity. Carlsberg Res. Commun. 48, 9–19. 7. Winther, J.R. & Breddam, K. (1987) The free sulfhydryl group (Cys341) of carboxypeptidase Y: functional effects of mutational substitutions. Carlsberg Res. Commun. 52, 263–273. 8. Olesen, K. & Kielland-Brandt, M.C. (1993) Altering substrate preference of carboxypeptidase Y by a novel strategy of muta- genesis eliminating wild type background. Protein Eng. 6, 409–415. 9. Endrizzi, J.A., Breddam, K. & Remington, S.J. (1994) 2.8-A ˚ structure of yeast serine carboxypeptidase. Biochemistry 33, 11106–11120. 10. Schechter, J. & Berger, B. (1967) On the size of the active site in proteases. I. Papain. Biochem. Biophys. Res. Commun. 27, 157– 162. 11. Heitmann, P. (1968) Reactivity of sulfhydryl groups in micelles. A model for protein. Eur. J. Biochem. 3, 305–315. 12. Heitmann, P. (1968) A model for sulfhydryl groups in proteins. Hydrophobic interactions of the cysteine side chain in micelles. Eur.J.Biochem.3, 346–351. 13. Bech, L.M. & Breddam, K. (1989) Inactivation of carboxy- peptidase Y by mutational removal of the putative essential his- tidyl residue. Carlsberg Res. Commun. 54, 165–171. 14. Deng, W.P. & Nickoloff, J.A. (1992) Site-directed mutagenesis of virtually any plasmid by eliminating a unique site. Anal. Biochem. 200, 81–88. 15. Ito, H., Fukuda, Y., Murata, K. & Kimura, A. (1983) Transfor- mation of intact yeast cells treated with alkali cations. J. Bacteriol. 153, 163–168. 16. Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166, 557–580. 17. Hayashi, R., Bai, Y. & Hata, T. (1975b) Evidence for an essential histidine in carboxypeptidase Y. Reaction with the chloromethyl ketone derivative of benzyloxycarbonyl- L -phenylalanine. J. Biol. Chem. 250, 5221–5226. 18. Bernardi, G. (1971) Chromatography of proteins on hydroxyl- apatite. Methods Enzymol. 22, 325–339. 19. Hayashi, R. (1976) Carboxypeptidase Y. Methods Enzymol. 45, 568–572. 20. Fersht, A. (1985) Enzymes Structure and Mechanism, 2nd edn, pp. 84–102, 244–273. WH Freeman, New York. 21. Olesen, K., Mortensen, U.H., Aasmul-Olsen, S., Kielland-Brandt, M.O., Remington, S.J. & Breddam, K. (1994) The activity of carboxypeptidase Y toward substrates with basic P1 amino acid residues is drastically increased by mutational replacement of leucine 178. Biochemistry 33, 11121–11126. 22. Olesen, K. & Breddam, K. (1995) Increase of the P1 Lys/Leu substrate preference of carboxypeptidase Y by rational design basedonknownprimaryandtertiarystructuresofserinecar- boxypeptidases. Biochemistry 34, 15689–15699. 23.Moehle,C.M.,Tizard,R.,Lemmon,S.K.,Smart,J.&Jones, E.W. (1987) Protease B of the lysosome-like vacuole of the yeast Saccharomyces cerevisiae is homologous to the subtilisin family of serine proteases. Mol. Cell. Biol. 7, 4390–4399. Ó FEBS 2002 A single thiol group in carboxypeptidase Y (Eur. J. Biochem. 269) 3225 . Amphipathic property of free thiol group contributes to an increase in the catalytic efficiency of carboxypeptidase Y Joji Mima, Giman Jung, Takuo. doi:10.1046/j.1432-1033.2002.02997.x designed to maintain hydrophobicity and Gly, Ser, Asp and His mutants to provide hydrophilicity. The catalytic roles of Cys341 in carboxypeptidase Y are examined