Role of amino acid residues in the active site of rat liver mercaptopyruvate sulfurtransferase. CDNA cloning, overexpression, and site-directed mutagenesis.

A complete amino acid structure of rat liver mercaptopyruvate sulfurtransferase (MST, EC 2.8.1.2) was determined by sequence analysis of cDNA and purified enzyme. The enzyme consists of 296 amino acid residues with a calculated molecular mass of 32,808 Da. Sequence identity in cDNA and the deduced amino acid sequence are 65 and 60% respectively, between rat MST and rhodanese. By their entire sequence similarity MST and rhodanese are confirmed to be evolutionarily related enzymes (Nagahara, N., Okazaki, T., and Nishino, T. (1995) J. Biol. Chem. 270, 16230-16235). The conversion of MST to rhodanese was attempted, and the role of amino acid residues was studied by site-directed mutagenesis with the isolated cDNA of rat liver MST. There is a strong possibility that Cys247 is a catalytic site of MST. Arg187 is suggested to be a binding site of both mercaptopyruvate and thiosulfate in MST. Arg196, which is missed in rhodanese, is important for catalysis in MST. On the other hand, the substitution of Arg for Gly248 or Lys for Ser249 facilitates catalysis of thiosulfate in MST. It is concluded that Arg187 and Arg196 of rat MST are critical residues in determining substrate specificity for mercaptopyruvate. On the other hand, Arg185, Arg247, and Lys248 of rat rhodanese are critical residues in determining substrate specificity for thiosulfate.

In the reaction the outer sulfur atom of thiosulfate is transferred to the Cys residue of the enzyme molecule to form a persulfide intermediate, which is subsequently attacked by cyanide anion to give thiocyanide. However, the enzyme has rather wide substrate specificity. Certain other sulfur compounds such as thiosulfonate (2) or persulfides (3,4) may substitute for thiosulfate; and sulfite (5), sulfinates (5), or various thiol compounds (6) may substitute for cyanide. This enzyme was reported to be widely distributed in prokaryote and eukaryote mitochondria (7)(8)(9). Although the physiological role of this enzyme is not well understood, the enzyme is well characterized. The enzyme was first isolated from bovine liver (10) and subsequently from rat liver (11). Primary structures of the enzyme from various sources were determined from protein or deduced from cDNA, e.g. bovine liver (12) and adrenal (13), chicken liver (14), human liver (15), rat liver (16), hamster ovary (17), and mouse liver (18). Further, recombinant bovine liver (13) and adrenal (19), rat liver (20), hamster ovary (17), and mouse liver (18) rhodanese were overexpressed in Escherichia coli and characterized. The enzyme from bovine liver was crystallized (10,21), and the three-dimensional structure was determined. (22)(23)(24). On the other hand, mercaptopyruvate sulfurtransferase (MST, 1 EC 2.8.1.2) catalyzes the following reaction.
The enzyme was discovered in rat liver quite a long time ago (25)(26)(27)(28), but compared with rhodanese the enzyme was not well characterized. Although the enzyme responsible for this reaction was at first assumed to be rhodanese, subsequent investigation with crystallized rhodanese showed that rhodanese did not catalyze this reaction and that another enzyme, mercaptopyruvate sulfurtransferase, was responsible for the reaction (28). Similar to rhodanese, compounds other than cyanide may function as sulfur acceptors in the reaction catalyzed by this enzyme (29 -32), and its physiological role is a matter of discussion (33,34). This enzyme was reported to be distributed in both prokaryotes and eukaryotes and to be located in the cytosol of eukaryotic cells (26,27,34,35), but the existence of this enzyme in mitochondria is also reported (36). Recently rat liver MST was purified to homogeneity, and the partial amino acid sequence around the active site of the enzyme was determined (20). Although the two enzymes were considered to be different, it was found that MST and rhodanese possessed both MST and rhodanese activities, but the ratios of their activities differed greatly. Further, a partial amino acid sequence of MST and the deduced primary structure of rhodanese showed striking similarity in sequence around the active site, but two amino acid residues following a catalytic site Cys were substituted. In a mutagenesis study with rat rhodanese cDNA, replacement of the two amino acid residues in the active site of rhodanese (Arg and/or Lys) with MST type residues (Gly and/or Ser) increased MST activity and decreased rhodanese activity, indicating that rhodanese was partly converted to MST (20). Based on these findings, it was proposed that MST and rhodanese were evolutionarily related enzymes. In this paper, to confirm the proposal, we attempted to convert MST to rhodanese by site-* 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 nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) D50564.
‡ To whom correspondence should be addressed. Tel.: 81-3-3822-2131; Fax: 81-3-5685-3054. directed mutagenesis of some amino acid residues around the proposed active site. The complete primary structure of the rat liver MST was determined after cDNA cloning. Wild type and six mutant MSTs were then constructed and overexpressed in E. coli. Based on kinetic studies for purified wild type and mutant MSTs, some amino acid residues that participate in the catalysis of mercaptopyruvate are discussed.

MATERIALS AND METHODS
Chemicals-Mercaptopyruvate for kinetic study was synthesized essentially by the method of Kun (37) as described previously (20). Other chemicals were of analytical grade.
Isolation of a cDNA Clone for Rat Liver MST and Construction of an Expression Vector for Wild Type MST-PCR primers were designed with reference to the cDNA sequence of human rhodanese (14), because the deduced amino acid sequence of human rhodanese was found to be strikingly similar to the primary structure of purified rat MST by partial peptide sequencing (20). 5Ј-Sense primer (CACAGAATTCGAC-CCCGCCTTCATCAAG; 506 -523 bp of human rhodanese cDNA with an EcoRI restriction site) and 3Ј-antisense primer (GATCAAGCTTCAT-GTACCACTCCACCCA; 857-874 bp with a HindIII site) were synthesized according to the DNA sequence of human liver rhodanese. Singlestranded cDNA for a template was synthesized with poly(A) ϩ RNA as a template, which was obtained from rat liver by a modification of the method of Chomczynski and Sacchi (38). Reverse transcriptase PCR was performed under low stringency annealing conditions (annealing at 43°C for 2 min, extension at 72°C for 3 min, and denaturation at 94°C for 80 s; 10 cycles) and then high stringency annealing conditions (annealing at 57°C for 2 min; 30 cycles) with AmpliTaq DNA polymerase (Takara Shuzo Co., Ltd.) adding Taq Extender TM PCR additive (Stratagene). 369 bp of PCR product was obtained, and its deduced amino acid sequence was identical to the amino acid sequence of rat MST (20). The PCR product was labeled with [␣-32 P]dCTP by means of a random-primed DNA labeling kit (Boehringer Mannheim).
A rat liver cDNA library (ZAP II, Stratagene) was screened by plaque hybridization with the PCR product as a probe. Three positive plaques were obtained after 1.7 ϫ 10 6 plaque hybridization. They were 940-bp clones, which contained the 5Ј-untranslated region (209 bp) but did not contain the terminal codon. This 5Ј-region clone in pBs(skϪ) was obtained by in vitro excision from the ZAP II vector containing cloned insert cDNA (39). This plasmid was designated as p241. Another rat liver cDNA library (gt11, Clontech Laboratory, Inc.) was then screened. Two identical positive plaques were obtained out of 1.3 ϫ 10 6 plaques. They contained a terminal codon with a 3Ј-untranslated region (200 bp), but this sequence did not extend to an initiation codon. Overlapped sequences of the two different clones were completely identical, and BstXI and EcoRI sites were coded in that order. The 5Ј-region clone was digested with EcoRI, and the digested fragment was subcloned into pBs(skϪ) between the EcoRI sites (designated pEIs). These inserted clones in p241 and pEIs were ligated at the BstXI site to construct full-length cDNA for rat liver MST containing 3Ј-and 5Јuntranslated regions in pBs(skϩ) between the HindIII and BamHI sites.
cDNA for MST (from the ATG coding initiation Met to the XhoI site in pBs) was inserted into the pET-15b vector (NOVAGEN) between the NcoI and XhoI sites for expression.
Site-directed mutagenesis and construction of an expression vector for mutant MST replacement of Arg 187 (initiation Met was not counted) by Gly (designated as R187G), Arg 196 by Gly (R196G), Cys 247 by Ser (C247S), Gly 248 by Arg (G248R), Ser 249 by Lys (S249K), and Ser 249 by Ala (S249A) were performed according to the method of Kunkel (40). p241 containing 3Ј-region cDNA of MST between EcoRI sites and pEIs containing 5Ј-region cDNA of MST between EcoRI sites were used as templates of the M13K07 helper phage. In the present study, six mutagenic oligonucleotides, GCAGCTGGCGGTTTCCAAGGC, CCAGAAC-CCGGAGATGGCATC, GGTAGCCACAAGCGGCTCCGGTG, AGCCA-CATGCAGATCCGGTGTCA, CACATGCGGCAAGGGTGTCACAG, CA-CATGCGGCGCCGGTGTCACAG, were synthesized for the R187G, R196G, C247S, G248R, S249K, and S249A mutants, respectively (underlined codons indicate mutagenized sites). The accuracy of mutagenesis was checked by DNA sequencing of the mutagenized MST cDNA. Each mutagenic fragment between the BamHI and BstXI sites in the p241 plasmid was ligated with a fragment between the BstXI and HindIII sites of the pEIs plasmid; other mutagenic fragments located between the BstXI and HindIII site in pEIs were ligated with a fragment between the BamHI and BstXI sites of p241 plasmid. Each ligated fragment was subcloned into pBs(skϩ) between the BamHI and HindIII sites. These cDNA constructs were then inserted into pET-15b expression vectors between the NcoI and XhoI sites for expression.
Expression and Purification of Wild Type and Mutant MST-These constructs were transformed into E. coli strain BL21(DE3). Transformed cells were cultured in 2 liters of LB medium containing 50 mg/ml of ampicillin at 27°C. At an absorbance of 0.8 at 600 nm, expression was induced by adding 1 mM isopropyl ␤-D-thiogalactopyranoside and by increasing the culture temperature to 37°C. After 3 h at 37°C, cells were harvested by centrifugation at 6,000 ϫ g for 5 min. Lysate was obtained essentially by the method of Sambrook et al. (41) except for the addition of 0.5 mg/ml trypsin inhibitor. The lysate was centrifuged, and the supernatant was fractionated with ammonium sulfate (40 -60% saturation). The precipitate was dissolved in a minimal volume of 10 mM Tris-HCl buffer, pH 7.8, containing 0.2 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride and 10 mM ␤-mercaptoethanol. The solution was dialyzed overnight against 1 liter of the same buffer at 4°C. 30 ml of the dialyzed enzyme was loaded onto an Express-ion TM exchanger D (anion exchanger, Whatman) column (1.5 ϫ 17 cm) equilibrated with the same buffer. After the column was washed with 45 ml of the same buffer and then with 45 ml of 50 mM Tris-HCl buffer, pH 7.8, containing 0.2 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride, and 10 mM ␤-mercaptoethanol, the enzyme was eluted with a linear gradient to 200 mM Tris-HCl buffer, pH 7.8, containing 0.2 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride, and 10 mM ␤-mercaptoethanol. 25 ml of the enzyme-containing fractions eluted at approximately 90 mM Tris-HCl were collected, and the ammonium sulfate was added to 65% saturation. The precipitate was dissolved in a minimal volume of 5 mM sodium acetate buffer containing 10 mM ␤-mercaptoethanol, pH 5.4. The solution was dialyzed against 1 liter of the same buffer at 4°C. The dialyzed enzyme was loaded on a Express-ion TM exchanger C (cation exchanger, Whatman) column (0.5 ϫ 12.5 cm) equilibrated the same buffer. After the column was washed with 10 ml of the same buffer and then with 10 ml of 5 mM sodium acetate containing 150 mM sodium chloride, pH 5.4. The enzyme was eluted with a linear gradient to 50 mM sodium acetate containing 400 mM sodium chloride. 12 ml of recombinant MST-containing fractions eluted at approximately 14 mM sodium acetate containing 200 mM sodium chloride, were collected and concentrated with Centricon-10 (Amicon) to more than 10 mg/ml. To the concentrated sample was added one-third by volume of 100% saturated ammonium sulfate solution, and it was stored at Ϫ80°C. Under these conditions, no significant loss of activity was observed for at least 6 months.
Protein Sequencing-Amino acid sequencing of recombinant enzymes was performed by the automated Edman degradation method with a gas phase protein sequencer model 477A (Applied Biosystems).
Western Blotting-Recombinant wild-type and C247S mutant MST were analyzed by SDS-polyacrylamide gel (10%) electrophoresis according to the method of Laemmli (42). Western blotting with anti-MST polyclonal antibody was performed as described previously (20).
Enzyme Assays-Rhodanese and MST activities were measured by a modification of the methods of Vachek and Wood (32) and Sörbo (10), respectively, as described previously (20).
Protein Determination-The protein concentration was determined with a Coomassie protein assay kit (Pierce) with crystalline bovine serum albumin as the standard.
Estimation of Tertiary Structure of Rat Rhodanese and MST-Edmundson wheel analysis was performed for the N-terminal region of rat rhodanese by means of GENETYX (Software Development Co.). The tertiary structure of the rat MST was estimated with QUANTA/ CHARMm (Molecular Simulations Inc.) based on data for bovine rhodanese obtained by x-ray crystallography (22)(23)(24).

Cloning of cDNA of Rat Liver MST and Deduced Amino Acid
Sequence-To clarify the structural and catalytic properties of MST, we cloned cDNA for rat liver MST. The 1094-bp clone was isolated (Fig. 1) under the procedure described under "Materials and Methods." The open reading frame contains 891 bp encoding 297 amino acid residues, which include sequences identical to the partial peptide sequences previously reported for purified rat liver MST (20) (underlined sequences in Fig. 1).
It should be noted that the sequence of human rhodanese is very similar to that of rat MST. The entire sequence identity between rat MST and human rhodanese (15) is 82 and 84% in cDNA and deduced primary structure, respectively. Around the active center, Gly 248 (Fig. 2, #4) and Ser 249 (Fig. 2, #5) of rat MST are conserved in human rhodanese. As described previously, replacement of the two amino acid residues (Arg and/or Lys) of rat liver rhodanese with MST type residues (Gly and/or Ser) increased MST activity and decreased rhodanese activity (20). Furthermore, Arg 196 (Fig. 2, #2) of rat MST is a unique residue that is absent in all rhodaneses except human rhodanese. This observation supports the previous proposal that the reported human rhodanese may in fact be human MST (20). It is intriguing that the primary structure of SseA deduced from the sseA gene of E. coli (43) is similar to those of MST and rhodaneses (Fig. 2). SseA protein was reported to enhance growth inhibition of E. coli when serine was added to the medium. SseA protein lacks the first 50 amino acid residues in the N-terminal region and six amino acid residues in the Cterminal region, compared with sequences of rat MST (Fig. 2). Sequence identity between SseA and rat MST is 53 and 38% for nucleotide and amino acid sequences, respectively, but that between SseA and rat rhodanese is 50 and 35% for nucleotide and amino acid sequences, respectively. Further, SseA contains the typical Cys-Gly-Ser sequence found in the active site of MST, as mentioned above (Fig. 2). These findings suggest that SseA is possibly an E. coli MST and an evolutionary prototype of the two family enzymes. It was reported that rhodanese was sorted to mitochondria in eukaryotic cells (7,8,10), but MST was localized in cytosol and mitochondria (35,36). Bovine rhodanese was reported to contain a noncleavable signal in the N-terminal region for mitochondrial import (13). Residues 11-22 are seen to form an ␣-helix of bovine rhodanese upon x-ray analysis (22,23). This is consistent with the fact that a synthetic peptide (residues 3-20) of the N-terminal region of rat rhodanese is found to form an ␣-helix by means of twodimensional NMR (44). This ␣-helix of these family enzymes is considered to be a signal for mitochondrial import. It is reasonable that SseA, a possible E. coli MST, does not contain this ␣-helix in the N-terminal region (Fig. 2). Further, positively charged Lys 13 and Arg 20 of bovine rhodanese are closely positioned at the outside of the ␣-helix (22,23). Edmundson wheel analysis for rat rhodanese reveals that positively charged Lys 12 and Arg 19 are also positioned very close on one side of the amphipathic helix. Arg 19 of rat rhodanese is conserved in other rhodaneses and MST. On the other hand, Lys 12 of rat rhodanese is conserved in other rhodaneses but is replaced by Gln in MST (Fig. 2). This Lys may make a difference in mitochondrial import between these two enzymes. Miller et al. (13) proposed that the repetitive sequence (Gly-Lys-X) 2 at the C terminus of bovine rhodanese (Fig. 2) was a signal for retention in the mitochondrial matrix, and the last three amino acids, Gly-Lys-Ala, were cleaved in the mature enzyme. This repetitive sequence is modified in MST. SseA of E. coli lacks this C-terminal repetitive region, as shown in Fig. 2.
Expression and Purification of Wild Type and Mutant Rat Liver MST-To clarify the role of amino acid residues of Arg 187 FIG. 1. Nucleotide sequence of rat liver mercaptopyruvate sulfurtransferase cDNA and its deduced amino acid sequence. Nucleotides are numbered in the 5Ј to 3Ј direction, beginning with the first residue of the ATG coding the initiation Met. The peptide sequences determined by Edman degradation (see "Materials and Methods") are underlined. (Fig. 2, #1), Arg 196 (Fig. 2, #2), Cys 247 (Fig. 2, #3), Gly 248 (Fig.  2, #4) and Ser 249 (Fig. 2, #5) in catalysis, we constructed six mutant MSTs, as shown in Table II. The mutant and wild type enzymes were then overexpressed and purified as described under "Materials and Methods." The 20 amino acids sequences in the N terminus of all of the expressed proteins agreed with the cDNA sequences. All expressed proteins were preceded by Ala (not Met), as expected from construction of the expression system described under "Materials and Methods," indicating that purified proteins were recombinant MST rather than native E. coli enzymes with MST activity.
Recombinant MSTs are unstable but can be stabilized by adding ammonium sulfate as was also observed for the native enzyme (20). Ammonium sulfate was therefore added to purified recombinant enzyme solutions before freezing. The amount of the overexpressed enzymes obtained from 1 liter of culture varied from 0.2 to 2 mg depending on the mutant. The specific activity of the purified wild type enzyme shows 3015 units/mg of protein and about a 20-fold increase compared with that of the lysate. SDS-polyacrylamide gel electrophoresis during the purification of wild type and mutant MST is shown in Fig. 3. Because neither MST nor rhodanese activity was detected in the C247S mutant (see below), this purified protein was confirmed to be a MST mutant by cross-reacting with anti-MST antibody (Fig. 4). SDS-polyacrylamide gel electrophoresis shows that wild type and mutant enzymes are 34-kDa molecules (Fig. 3), which is in reasonable agreement with the subunit size calculated from the deduced primary structure of rat MST (molecular mass 32,808 Da).
Effect of Mutagenesis, Possible Role of Amino Acid Residues in Catalysis-The double reciprocal plots of recombinant wild type MST show a mixed inhibition pattern when MST activity is measured (Fig. 5) as observed for the native rat liver MST (20). It is suggested that recombinant MST follows a sequential kinetic pattern (45). There is no significant variation in kinetic pattern among wild type and mutant MSTs except for the C247S mutant. On the other hand, the reaction mechanism of bovine liver rhodanese was reported to follow a ping-pong pattern (46,47). An intermediate product (persulfide formation at a catalytic site Cys) in the rhodanese catalytic process was also identified (22,24). In the present study, however, the double reciprocal plots of velocity versus potassium cyanide concentration did not show a straight line when rhodanese activities were measured with wild type and mutant MST, as also observed for native MST (20). Apparent K m values for thiosulfate were therefore determined with a constant concentration of potassium cyanide of 60 mM. The present study strongly suggests that Cys 247 (Fig. 2, #3) is a catalytic site of MST, because C247S mutant (Table I) loses both MST and rhodanese activities (Table II). In bovine rhodanese, a Cys corresponding to Cys 247 of MST was identified as a catalytic site forming persulfide in the process of catalysis (22)(23)(24)48). Cys 247 of MST is therefore proposed to form persulfide in catalysis of the rhodanese activity. In a previous study (20), it was shown that rhodanese had weak MST activity, and replacement of two amino acid residues of rat rhodanese (R248G and/or K249S) FIG. 2. Comparison of partial primary structure of rat liver MST with sequences of rhodaneses of human, rat, hamster, bovine and avian liver, and E. coli SseA. HUM RHOD, deduced primary structure of human liver rhodanese (15); RAT RHOD, deduced primary structure of rat liver rhodanese (16); HAM RHOD, deduced primary structure of hamster ovary rhodanese (17); MOU RHOD, deduced primary structure of mouse liver rhodanese (18); BOV RHOD, primary structure of purified bovine liver rhodanese (12); AVI RHOD, amino acid sequence from purified chicken liver rhodanese (14); SseA, deduced primary structure of E. coli sseA gene (43); shaded box, identical amino acid residues: arginine 187 (#1), arginine 196 (#2), cysteine 247 (#3), arginine 248 (#4), and lysine 249 (#5). AAQ, the order of these three amino acid residues was not determined (14); ‫,ށދއ‬ these three amino acids were not detected by protein sequencing of purified mature rhodanese from bovine liver (12) but were contained in the deduced primary structure of bovine adrenal rhodanese cDNA (13). could increase its MST activity. In the present study, to attempt to convert MST to rhodanese reversely, G248R, S249K, and R196G mutants were constructed. Arg substituted for Gly 248 (Fig. 2, #4) and Lys substituted for Ser 249 (Fig. 2, #5) are typical residues for rhodanese, whereas Arg 196 is a unique residue for MST that is absent from rhodanese, as mentioned above. These mutant MSTs, the kinetic properties of which are discussed below, show an increase in the ratio of rhodanese to MST activity (Table II). This indicates that MST is partly converted to rhodanese, confirming the previous conclusion that the difference in the ratio of their activities is caused by specific differences in amino acid residues (20).
To define the role of five amino acid residues of MST in utilization of mercaptopyruvate and thiosulfate, a kinetic study of wild type, R187G, R196G, G248R, S249K, and S249A mutant MST (Table I) was performed. Arg 187 of MST corresponds to the residue of bovine rhodanese that was proposed to serve as a binding site of thiosulfate (22-24, 48, 49). As shown in Table II, for rhodanese activity in R187G, k cat and k cat /K m are decreased to about 1 ⁄5 and 1 ⁄30, respectively, of those in wild type, and K m for thiosulfate is increased to about 5-fold that in wild type, consistent with the previously reported results for the corresponding substitution in rhodanese (20). On the other hand, R187G mutant shows a much greater decrease in k cat and k cat /K m for MST activity to about 3-order lower values and a noticeable increase in K m for mercaptopyruvate to about 60-fold that in wild type (Table II). Arg 187 (Fig. 2, #1) therefore appears to be a binding site of mercaptopyruvate and a critical residue for catalysis in MST. A possible role in catalysis is speculated to be that a positively charged side chain of Arg 187 interacts with the oxygen atom in the carbonyl group of mercaptopyruvate to cause an electrostatic interaction, as shown in Fig. 6. Since cysteine is not a substrate of this enzyme (data not shown), the carbonyl group is essential for catalysis. Polarization of the carbonyl group may accelerate catalysis by introducing electron strain in mercaptopyruvate. A nucleophilic attack on the sulfur of mercaptopyruvate by Cys 247 may then be facilitated. An alternative explanation is that this Arg residue stabilizes a persulfide intermediate as postulated for Arg 248 in rhodanese (22)(23)(24). However, this is unlikely, because much more pronounced effects on MST activity than on rhodanese activity have been observed, findings inconsistent with the possible common mechanism that persulfide forms as an intermediate. Arg 196 of MST is a unique residue (Fig. 2, #2), which is absent from rhodanese as described above. Based on the tertiary structure of bovine rhodanese (22)(23)(24), a bulky side chain of Arg 196 of MST is modeled as covering the pocket of the active site by computer simulation analysis with QUANTA/ CHARMm as described under "Materials and Methods" (data not shown). The R196G mutant shows an increase in K m for mercaptopyruvate to about 10-fold that in wild type and a decrease in k cat and k cat /K m for MST activity to about 1 ⁄4 and 1 ⁄30, respectively of those in wild type (Table II). For rhodanese activity, on the other hand, the R196G mutant shows almost the same K m for thiosulfate as that in wild type, but k cat and k cat /K m for rhodanese activity are increased to about 3-fold those in wild type (Table II). These findings indicate that Arg 196 facilitates catalysis of mercaptopyruvate in MST but hinders catalysis of thiosulfate in MST. A possible role in the catalytic mechanism is that the positively charged side chain of Arg 196 interacts with the carboxyl group of mercaptopyruvate and assists in polarizing the substrate.
The G248R mutant shows only a small difference in K m for mercaptopyruvate from that of wild type but a marked decrease in k cat and k cat /K m for MST activity to about 1 ⁄1000 and 1 ⁄1500, respectively, of those in wild type (Table II). This behavior may be explained by steric hindrance. Although Gly 248 may not play an important role in the catalysis of mercaptopyruvate, such a small residue as a Gly may be advantageous for catalysis. A bulky side chain of replaced Arg probably hinders catalysis of mercaptopyruvate, and MST activity is decreased as a consequence. For rhodanese activity, the G248R mutant  shows a decrease in K m for thiosulfate to about one-fifth of that in wild type, but the k cat and k cat /K m values for rhodanese activity were observed to be about 1 ⁄15 and 1 ⁄4, respectively, of those in wild type (Table II). The decrease in rhodanese activity caused by substituting Arg for Gly 248 is inconsistent with the previous finding indicating that this Arg facilitated catalysis of thiosulfate in rhodanese (20). Possibly, the substituted Arg cannot be positioned suitably for rhodanese activity in the pocket of an active site in this case, because the substituted Arg and Arg 196 are modeled as interfering with optimal positioning of each other in the homology-modeled tertiary structure of MST (data not shown). The S249K mutant shows a K m for mercaptopyruvate that is almost the same as that in wild type, suggesting that Ser 249 is not a binding site for mercaptopyruvate. Although the k cat and k cat /K m values for MST activity of the same mutant are decreased to about 1 ⁄16 and 1 ⁄10, respectively, of those in wild type (Table II), the residue may not play a critical role in catalysis. Another mutant (S249A) shows no significant difference in K m for mercaptopyruvate from that in wild type and a smaller decrease in the k cat and k cat /K m values for MST activity to about 1 ⁄10 and 1 ⁄3, respectively, of those in wild type (Table II). In the S249K mutant, a substituted Lys that contains a bulky side chain may interfere with the catalysis of mercaptopyruvate, and MST activity would be decreased as a consequence. In any case, the replacement of Ser 249 with other residues causes a slight decrease in MST activity. It can be speculated that this Ser is of adequate residue size and maintains a suitable conformation of the enzyme for catalysis of mercaptopyruvate. For rhodanese activity, replacement of a Ser with a Lys (S249K) decreased K m for thiosulfate to about 1 ⁄10 of that in wild type and increased k cat /K m for rhodanese activity to about 10-fold that in wild type (Table II). The smaller substrate of thiosulfate may not be influenced by a bulky side chain, but rather the positive charge of this group may be important for catalysis of rhodanese. This is also consistent with the conclusion that the substituted Lys is a binding site of thiosulfate in rhodanese (20, 22-24, 49, 50).