Engineering of Porcine Pepsin

The S1 substrate specificity of porcine pepsin has been altered to resemble that of fungal aspartic proteinase with preference for a basic amino acid residue in P1 by site directed mutagenesis. On the basis of primary and tertiary structures of aspartic proteinases, the active site-flap mutants of porcine pepsin were constructed, which involved the replacement of Thr-77 by Asp (T77D), the insertion of Ser between Gly-78 and Ser-79 (G78(S)S79), and the double mutation (T77D/G78(S)S79). The specificities of the mutants were determined using p-nitrophenylalanine-based substrates containing a Phe or Lys residue at the P1 position. The double mutant cleaved the Lys-Phe(4-NO2) bonds, while wild-type enzyme digested other bonds. In addition, the pH dependence of hydrolysis of Lys-containing substrates by the double mutant indicates that the interactions between Asp-77 of the mutant and P1 Lys contribute to the transition state stabilization. The double mutant was also able to activate bovine trypsinogen to trypsin by the selective cleavage of the Lys6–Ile7 bond of trypsinogen. Results of this study suggest that the structure of the active site flap contributes to the S1 substrate specificity for basic amino acid residues in aspartic proteinases.

The S 1 substrate specificity of porcine pepsin has been altered to resemble that of fungal aspartic proteinase with preference for a basic amino acid residue in P 1 by site directed mutagenesis. On the basis of primary and tertiary structures of aspartic proteinases, the active site-flap mutants of porcine pepsin were constructed, which involved the replacement of Thr-77 by Asp (T77D), the insertion of Ser between Gly-78 and Ser-79 (G78(S)S79), and the double mutation (T77D/G78(S)S79). The specificities of the mutants were determined using p-nitrophenylalanine-based substrates containing a Phe or Lys residue at the P 1 position. The double mutant cleaved the Lys-Phe(4-NO 2 ) bonds, while wild-type enzyme digested other bonds. In addition, the pH dependence of hydrolysis of Lys-containing substrates by the double mutant indicates that the interactions between Asp-77 of the mutant and P 1 Lys contribute to the transition state stabilization. The double mutant was also able to activate bovine trypsinogen to trypsin by the selective cleavage of the Lys 6 -Ile 7 bond of trypsinogen. Results of this study suggest that the structure of the active site flap contributes to the S 1 substrate specificity for basic amino acid residues in aspartic proteinases.
Aspartic endopeptidases (EC 3.4.23._) comprise a group of enzymes whose proteolytic activities are dependent on two aspartyl residues, Asp-32 and Asp-215, in pepsin numbering (1). Mammalian and fungal enzymes have been extensively characterized, and their three-dimensional structure has been determined at high resolution (2)(3)(4)(5)(6)(7). The enzymes of this family are bilobal with a deep and extended cleft which can accommodate at least seven amino acid residues in the S 4 -S 3 Ј subsites 1 (8). The "flap," an antiparallel ␤-hairpin loop comprising residues 72 to 82 (pepsin numbering), projects across the cleft forming a channel into which a substrate binds. Although the enzymes are quite similar in their three-dimensional structures, there are drastic differences in the catalytic properties, especially in substrate specificities. There also have been many studies on the tertiary structures of aspartic proteinases with peptide-derived inhibitors (4, 5, 9 -14), and the relationship between their structures and substrate specificities has been demonstrated (4, 5, 9 -13, 15, 16). According to these reports, although the hydrogen bonding pattern between the main chain of inhibitor and enzyme is well conserved in all inhibitor-enzyme complexes, the differences in the size of subsites, of which residues make van der Waals contacts with the side chain of inhibitor, may control the substrate specificities in aspartic proteinases. Several attempts to examine the structural determinants of substrate specificities of aspartic proteinases by site-directed mutagenesis have been made recently (17)(18)(19)(20)(21)(22)(23). These provided direct evidence that the subtle differences in the structures of substrate binding sites of aspartic proteinases were sufficient to alter their substrate specificities.
Aspartic proteinases generally show specificity for the cleavage of the bond between hydrophobic residues occupying the S 1 -S 1 Ј subsites. However, fungal enzymes also have preferences for a Lys residue in the P 1 position, which leads to activation of trypsinogen by cleavage of the Lys 6 -Ile 7 bond (24 -27). The S 1 subsite is formed by several hydrophobic residues in the neighborhood of the catalytic Asp-32 and by the residues on the active site flap. The sequence alignments of mammalian and fungal enzymes reveal that Asp-77 and Ser-79 on the flap are conserved in all family members able to activate trypsinogen, but Asp-77 is replaced by Ser or Thr and Ser-79 is deleted in those unable to do so (Table I) Pepsin (EC 3.4.23.1) is a typical aspartic proteinase produced in the gastric mucosa of vertebrates as a zymogen form. This enzyme has been extensively characterized, and its three-dimensional structure has been determined at high resolution (2,3,9). Porcine pepsin, in particular, has been studied as a model to analyze the structure-function relationship of the aspartic proteinases.
The present studies investigated whether the active site flap controls the S 1 subsite specificity of aspartic proteinases, and for this we generated several mutants of porcine pepsin by site-directed mutagenesis and analyzed their enzymatic properties. The substitution of Thr-77 by Asp and the insertion of Ser between Gly-78 and Ser-79, corresponding to the fungal enzymes, conferred upon porcine pepsin the ability to hydrolyze the substrate containing a Lys residue at the P 1 position. These experiments provide positive evidence of the importance of the active site flap of aspartic proteinases in the S 1 subsite specificity.

EXPERIMENTAL PROCEDURES
Materials-Porcine pepsin, bovine trypsinogen, and bovine hemoglobin were from Sigma. Pro-Thr-Glu-Phe-Phe(4-NO 2 )-Arg-Leu was ob-* This work was supported in part by grants-in-aid for scientific research from the Ministry of Education, Science and Culture of Japan. 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.
Synthesis of Porcine Pepsinogen cDNA and Construction of Expression Plasmid-Total RNA from porcine gastric mucosa was extracted and purified by the modified acid guanidinium thiocyanate/phenol/ chloroform RNA extraction method (29). Porcine pepsinogen cDNA was synthesized by reverse transcriptase-polymerase chain reaction (PCR) (30) using the oligonucleotides 5Ј-CGAACATATGCTCGTCAAGGTC-CCGCTGGT-3Ј and 5Ј-GGAGGAATTCAGGCTCAGGCCACGGGAG-CC-3Ј, based on cDNA sequence of porcine pepsinogen reported by Lin et al. (31). Sense primer was designed to delete the signal sequence and introduce a translation initiation codon (Met) for direct expression. Synthesized cDNA was cloned in pUC119 and sequenced to confirm the absence of an undesired mutation. The NdeI/SalI fragment was inserted into pET12a, which was designated as pETPP, to express porcine pepsinogen under the control of a T7 promoter.
Enzyme Purification-E. coli BL21(DE3) cells harboring the desired mutant plasmid were inoculated into 500 ml of LB medium with 50 g/ml ampicillin and cultured at 37°C until the OD 660 was 0.6; isopropyl-1-thio-␤-D-galactopyranoside was added to a final concentration of 0.1 mM, and the incubation was continued for 4 h. The cells were harvested by centrifugation and resuspended in 50 ml of TE buffer (50 mM Tris-HCl, pH 8.0, containing 2 mM EDTA). After Triton X-100 was added to the suspension to a final concentration of 1%, the cells were ruptured by sonication. The cell homogenate was centrifuged at 6,000 ϫ g for 15 min. The resulting pellets, which contained mostly porcine pepsinogen as inclusion bodies, were washed by being resuspended in 10 ml of TE buffer followed by centrifugation. The washed inclusion bodies were dissolved in 40 ml of TE buffer containing 8 M urea and allowed to stand at room temperature for 2 h, solid NaCl was added to the solution to a final concentration of 0.5 M, and the pH was adjusted to 10.0 with 1 N NaOH. The solution was dialyzed against 1 liter of 20 mM sodium bicarbonate, pH 10.0, containing 4 M urea, and the urea concentration was lowered stepwise from 4 M to 2, 1, 0.5, and 0 M every 3 h at room temperature. The dialyzing buffer was then exchanged for 2 liters of 20 mM sodium phosphate, pH 7.0, and dialysis was continued at 4°C for 8 h. The dialysate was used as a crude porcine pepsinogen preparation. Five milliliters of 2 M glycine-HCl, pH 2.0, was added to 40 ml of this pepsinogen preparation and incubated at 25°C for 3 h. The solution was dialyzed against 1 liter of 20 mM sodium acetate, pH 5.0, at 4°C. The dialysate was applied to an anion-exchange RESOURCE Q column (Pharmacia Biotech Inc.; 0.64 ϫ 3 cm) and eluted with 20 mM sodium acetate, pH 5.0, with a linear gradient of 0 to 1.0 M NaCl. The active fractions were dialyzed against 10 mM sodium acetate, pH 5.5, and used as purified porcine pepsinogen. Purity of the enzymes was examined by SDS-PAGE (33), and their amino-terminal amino acid sequences were determined by the method of Matsudaira (34).
CD Measurements-The circular dichroism (CD) spectra were measured with a Jasco J-700 spectropolarimeter at room temperature in a 1-mm path length cell. The concentrations of enzymes were adjusted to 0.2 mg/ml in 5 mM sodium acetate, pH 5.0. Contents of the ␣-helix and ␤-structure of the enzymes were calculated according to the SSE-338 program given in Ref. 35.
Kinetic Characterizations of Mutant Porcine Pepsins-Three chromogenic peptide substrates, Pro-Thr-Glu-Phe-Phe(4-NO 2 )-Arg-Leu (peptide A), Pro-Thr-Glu-Lys-Phe(4-NO 2 )-Arg-Leu (peptide B), and Ac-Ala-Ala-Lys-Phe(4-NO 2 )-Ala-Ala-NH 2 (peptide C), were used as substrates for kinetic analyses. Assays with peptides A and B were performed as described by Dunn et al. (37) and with peptide C by the method of Hofmann and Hodges (38). The kinetic constants, K m and V max , were determined from plots of initial rates versus substrate concentrations with ranges of 0.010 -1.000 mM for all substrates. Values for k cat were derived from V max ϭ k cat [E], where [E] is the enzyme concentration.
Product Analysis-Purified enzymes (0.35 g) were incubated in total volume of 100 l of 20 mM sodium acetate, pH 3.5, containing 10 nmol of each peptide substrate (peptides A, B, and C) for 8 h at 37°C (enzyme/substrate, 1/1,000 mol/mol). Cleavage products were separated by reverse-phase high performance liquid chromatography on a TSK ODS-120T column (4.0 ϫ 250 mm, Tosoh, Tokyo, Japan) and confirmed by amino acid sequence analysis.
pH Dependence Studies-The kinetic parameters of the double mutant for three peptide substrates (peptides A, B, and C) were determined at pH 2.55-6.0 in 0.1 M sodium acetate. The enzyme concentrations used in these experiments were 1.0 -115 nM, 28.8 -288 nM, and 28.8 -230 nM for peptides A, B, and C, respectively.

Overexpression and Purification of Recombinant Porcine
Pepsin-The structure of porcine pepsinogen cDNA was reported earlier by Lin et al. (31). Therefore, porcine pepsinogen cDNA was cloned by reverse transcriptase-PCR, using total RNA from porcine gastric mucosa as a template. PCR primers were designed for direct expression in the E. coli pET system. The amplified cDNA was dideoxy-sequenced to verify whether it was synthesized correctly. The sequence was identical with that reported by Lin et al. (31) except for the substitution of Tyr-242 by Asp. Porcine pepsin has been sequenced in several laboratories by protein and cDNA sequencing (39,40), and these studies indicated that there are two variants at position 242 of amino acid sequence (Asp or Tyr). The electrospray mass spectrometry of porcine pepsin also demonstrated the existence of two variants (41). Hence, the cDNA obtained in this study seemed to be correct and available for expression studies.
The E. coli cells harboring pETPP produced pepsinogen as an inclusion body under the control of the T7 promoter. The inclusion body was purified by washing with buffer and then re- Porcine pepsin folded by solubilization with 8 M urea and subsequent dialysis under alkaline pH as described under "Experimental Procedures." Porcine pepsin was purified from the refolded pepsinogen preparation by acidification and subsequent chromatography on a RESOURCE Q column and gave a single band on SDS-PAGE (Fig. 1, lane 2). The NH 2 -terminal sequence of recombinant pepsin was found to be predominantly Ile-Gly-Asp-Glu-Pro-Leu-Glu-Asn-. . . which is known as the NH 2terminal sequence of porcine pepsin (39), and a minor sequence of Ala-Ala-Leu-Ile-Gly-Asp-Glu-Pro-. . . was also detected; the same results were reported in the activation of native porcine pepsinogen by Kageyama and Takahashi (42). The secondary structures of native and recombinant porcine pepsin were analyzed by CD spectrometry. The CD spectral data showed that the spectrum of the recombinant enzyme was essentially superimposable on that of the native enzyme (data not shown). The native and recombinant enzymes were almost equal in specific activity for hydrolysis of acid-denatured hemoglobin (82 and 83 millikatals/kg, respectively).

Molecular Properties of Mutant Porcine Pepsins-
The mutant porcine pepsins, T77D, G78(S)S79, and T77D/G78(S)S79, were purified by the same method as wild-type pepsin, and the purities of the enzymes were judged by SDS-PAGE (Fig. 1). The NH 2 -terminal sequences of the mutants were the same as that of wild-type enzyme. The secondary structures of recombinant wild-type and mutant pepsins were analyzed by CD spectrometry to determine whether localized or global changes of structures were induced by the mutations. The CD spectra data showed that the spectra of the mutants were essentially superimposable on that of the wild-type enzyme (data not shown). These results suggest that no major conformational alterations occurred in the mutant enzymes.
Enzymatic Properties of Mutant Porcine Pepsins for Protein Substrates-Proteolytic activities for acid-denatured hemoglobin and trypsinogen-activating activities due to limited proteolysis of the Lys 6 -Ile 7 bond of trypsinogen were measured with wild-type and mutant pepsins. The specific activities of mutant enzymes, T77D, G78(S)S79, and T77D/G78(S)S79, for hemoglobin hydrolysis were determined to be 53, 8, and 50 millikatals/ kg, respectively. Each mutant enzyme effectively hydrolyzed hemoglobin although the activity of G78(S)S79 mutant was 10 times lower than that of wild-type enzyme. The specific activities of wild-type, T77D, and T77D/G78(S)S79 enzymes for trypsinogen activation were 0.04, 0.01, and 18.5 microkatals/ ml, respectively, and no detectable activity was found for G78(S)S79 enzyme, which indicates that the activity of trypsinogen activation was significantly increased only by the double mutant. However, it was still 300-fold less than the activity of aspergillopepsin I. SDS-PAGE analysis showed that trypsinogen was converted to trypsin by T77D/G78(S)S79 mutant alone (Fig. 2) and the cleavage site was found to be Lys 6 -Ile 7 by NH 2 -terminal sequencing. These results indicated that the double mutation in the active site flap altered substrate specificity of porcine pepsin to those of fungal aspartic proteinases, aspergillopepsin, penicillopepsin, and rhizopuspepsin.
Kinetic Properties of Mutant Pepsins for Peptide Substrates-To probe the effects of the mutations on substrate specificity for the P 1 residue, we used the peptide substrates, Pro-Thr-Glu-Phe-Phe(4-NO 2 )-Arg-Leu (peptide A), Pro-Thr-Glu-Lys-Phe(4-NO 2 )-Arg-Leu (peptide B), and Ac-Ala-Ala-Lys-Phe(4-NO 2 )-Ala-Ala-NH 2 (peptide C). Peptide A was cleaved at the Phe-Phe(4-NO 2 ) bond by each enzyme, which indicated that mutant enzymes also prefer hydrophobic residues in P 1 and P 1 Ј in the substrate. The T77D/G78(S)S79 mutant hydrolyzed Lys-Phe(4-NO 2 ) bonds in peptides B and C, while wild-type enzyme cleaved mainly at Glu-Lys and Phe(4-NO 2 )-Ala bonds in peptides B and C, respectively. The T77D mutant cleaved at the Lys-Phe(4-NO 2 ) bond in peptide B, but at the Phe(4-NO 2 )-Ala bond in peptide C. There was no cleavage in peptides B and C by G78(S)S79 mutants. These results indicate that the T77D/ G78(S)S79 mutant acquired the ability to recognize a Lys residue at the P 1 position although the enzyme still exhibited a preference for hydrophobic residues at the P 1 position. From kinetic determination of the k cat /K m (catalytic efficiency) one can obtain the second-order rate constant for conversion of a substrate to a product (43). Combining effects due to substrate binding and transition state stabilization, this parameter is useful for assessing altered substrate specificity. Differences in log(k cat /K m ) provide an accurate measure of the lowering of the transition state activation energy (⌬G T ‡ ). The kinetic parameters for hydrolysis of Phe-Phe(4-NO 2 ) and Lys-Phe(4-NO 2 ) bonds were determined from the initial rate measurements, and the results are listed in Table II. The mutations led to a 4to 130-fold decrease of k cat /K m compared to wild-type enzyme for peptide A. However, the k cat /K m value for the double mutant was larger than those of single mutants. Although the k cat /K m values of the double mutant for peptides B and C were 50 and 30 times smaller than that for peptide A, respectively, the double mutation led to hydrolysis of the Lys-Phe(4-NO 2 ) bonds in the former two peptides. The double mutant showed a similar ratio of k cat /K m for peptide C over peptide A to aspergillopepsin I, although the preference for peptide B was significantly different between the double mutated pepsin and aspergillopepsin I. These results indicated that fungal aspartic proteinase-like specificity can be introduced into porcine pepsin.
pH Activity Profile for the T77D/G78(S)S79 Double Mutant-To obtain information on recognition of a Lys residue at the P 1 position by the T77D/G78(S)S79 double mutant, the kinetic parameters for this double mutant were determined at various pH values using three peptide substrates containing a Phe or Lys in P 1 (Fig. 3). In all substrates, the K m values were pH independent, and above pH 4 the values of k cat and k cat /K m were controlled by the dissociation of a carboxyl group with a pK a of about 5. Below pH 4, however, the k cat and k cat /K m values for the P 1 Lys substrates (peptides B and C) decreased as pH declined, while there was no significant change in k cat and k cat /K m values for the P 1 Phe substrate (peptide A).
The pH-dependent changes of substrate preference are shown in Fig. 4. The P 1 Lys/Phe preference increased as pH rose, which indicates that the dissociation of a carboxyl group as Asp-77 may affect the preference of a Lys residue at the P 1 position.

DISCUSSION
The double mutation in porcine pepsin, the substitution of Thr-77 by Asp, and the insertion of Ser between Gly-78 and Ser-79 (T77D/G78(S)S79), successfully altered the S 1 subsite specificity of pepsin to those of fungal aspartic proteinases. Some fungal aspartic proteinases, aspergillopepsin I, penicillopepsin, rhizopuspepsin, and others, are able to activate trypsinogen by cleavage of its Lys 6 -Ile 7 bond, showing an affinity    and Lys-Phe  bonds in the peptide substrates by wild-type and mutant pepsins Each reaction was carried out in 0.1 M sodium acetate, pH 4.5, at 37°C. for the anionic lysine side chain in S 1 (26). It is known that Asp-77 (pepsin numbering) of these enzymes is the binding site to the Lys residue in P 1 by means of site-directed mutagenesis of aspergillopepsin I (20, 21) and rhizopuspepsin (22), as pointed out for the case of inhibitor binding to penicillopepsin (11). In porcine pepsin, the hydrophobic side chain can be accommodated in the S 1 subsite, while basic residue cannot. As expected, the double mutant pepsin was able to accept the lysine residue in P 1 , which led to trypsinogen activation, although it still showed a preference for the hydrophobic residue in P 1 . However, the single mutants, T77D and G78(S)S79, were not able to hydrolyze the substrates containing a Lys residue in P 1 and also exhibited lower catalytic efficiency (k cat /K m ) for the P 1 Phe substrate than those of wild-type and double mutant enzymes. The double mutant exhibited a similar preference for peptide C as did aspergillopepsin I, while for peptide B the catalytic efficiency of double mutated pepsin was about 20-fold lower than that of aspergillopepsin I; this was mainly due to the smaller K m values of aspergillopepsin I for peptides A and B than those of the double mutated pepsin. The differences in the substrate affinity may be a consequence of alterations in the structures of other substrate binding sites (S 4 , S 3 , S 2 , S 2 Ј, and S 3 Ј) than the S 1 and S 1 Ј subsites. Fig. 5 shows the structures of S 1 subsites of human pepsin and penicillopepsin complexed with pepstatin and statinebased inhibitor, Iva-Val-Val-LySta-OEt (Iva, isovaleryl; LySta, 4S,3S-4,8-diamino-3-hydroxyloctanoic acid), respectively. The hydroxy group of the transition state analogue, statine (4S,3S-4-amino-3-hydroxyl-6-methylheptanoic acid) or LySta, is hydrogen-bonding to the oxygen atoms of two catalytic aspartyl side chains, and the positions of the analogue and catalytic apparatus are essentially superimposable in both complexes despite the differences in the flap structures. The flap structure of human pepsin is included in the typical "class 2" ␤-hairpin loop, while that of penicillopepsin is an anomalous structure, which has been included in the "class 3" ␤-hairpin loop according to Milner and Poet (44). Schematic models of the flap structure of human pepsin and penicillopepsin show that Ser-79 of penicillopepsin is inserted while human pepsin is not (Fig. 6). However, the hydrogen bonds between the main chain of the inhibitor and the enzyme are well conserved in both complexes, which may contribute to the proper orientation of the scissile peptide bond. The 77th amino acid, Thr in pepsin and Asp in penicillopepsin, forms hydrogen bonds to the P 2 NH and the carbonyl group of main chain in the inhibitor. In the penicillopepsin complex, the side chains of Asp-77 and Ser-79 form additional hydrogen bonds to the side chain of the Lys residue at P 1 .
In this study, the insertion of Ser between Gly-78 and Ser-79 of porcine pepsin may have caused an alteration in the direction of the Thr-77 side chain and have disabled the side chain of Thr-77 from hydrogen bonding to the main chain of the substrate. These would result in overall decreases in catalytic efficiencies for all substrates in the G78(S)S79 mutant. In the T77D mutant, the presence of the longer Asp-77 side chain may allow the flap to bind more loosely to the substrate and result in a decrease in the number of van der Waals contacts between enzyme and substrate; this would lead to a reduction in catalytic efficiency. However, the T77D mutant enzyme hydrolyzed the Lys-Phe(4-NO 2 ) bond in Pro-Thr-Glu-Lys-Phe(4-NO 2 )-Arg-Leu (peptide B) although its catalytic efficiency was very low, whereas the Glu-Lys bond was the main cleavage site in wildtype pepsin. This would be due to the electrostatic repulsion between Glu of the substrate and Asp-77 of the T77D mutant enzyme, and it would not be the positive recognition of anionic Lys residue in P 1 . The wild-type and T77D mutant enzymes cleaved the Phe(4-NO 2 )-Ala bond in Ac-Ala-Ala-Lys-Phe(4-NO 2 )-Ala-Ala-NH 2 (peptide C) due to the strong interaction between the Phe(4-NO 2 ) residue of the substrate and S 1 hydrophobic pocket of the mutants. The double mutated pepsin exhibited higher catalytic efficiency (k cat /K m ) for the P 1 Phe substrate than did the single mutants. The double mutant would have a flap structure and hydrogen bond network between Asp-77 and the main chain of a substrate similar to those of the fungal aspartic proteinases. In the double mutant, the scissile peptide bond of the substrate may lie at a more stable position. Furthermore, the double mutant effectively hydrolyzed the Lys-Phe(4-NO 2 ) bond in peptides B and C, due to the additional two hydrogen bonds; the side chains of Asp-77 and inserted Ser in the mutant may be hydrogen-bonded to the ⑀-amino group of the Lys residue in P 1 as fungal enzymes. In a previous study on aspergillopepsin I (21), the effects of the substitution of Ser-79 by Ala (S79A) and the deletion of Ser-79 (⌬S79; corresponding to porcine pepsin) on the substrate specificity were analyzed. The P 1 Lys/Phe preference of the ⌬S79 mutant was reduced more significantly than the S79A enzyme. Therefore, in mutant pepsins the insertion of Ser is important for the orientation of the side chain at position 77, although this Ser may also contribute to the hydrogen acceptor in the hydrogen-bonding interaction.
The effects of the dissociation of the ␤-carboxyl group of Asp-77 in double mutated pepsin on the recognition of Lys residue at the P 1 position were examined by pH-dependent hydrolysis of two types of peptide substrates containing Phe or Lys at P 1 . The major differences in pH activity profiles were the k cat /K m values below pH 4. These values for P 1 Lys substrates (peptides B and C) decreased as the pH dropped, while there was no significant change for P 1 Phe substrate (peptide A). The K m values were pH independent in all cases. These results are different from those of the studies that the existence of Lys or Arg residues at the P 4 , P 3 , P 2 , P 3 Ј, P 4 Ј, and P 5 Ј positions in the substrates influences the K m values with little effect on k cat in porcine pepsin (45,46); the ionic interactions between the basic residues in the substrate and the side-chain carboxylates in the S 4 , S 3 , S 2 , S 3 Ј, S 4 Ј, and S 5 Ј subsites contribute to substrate affinity. In double mutated pepsin, it is likely that the dissociation of Asp-77 side chain contributes to the stabilization of the transition state when substrates contain Lys residue at the P 1 position rather than substrate affinity. The electrostatic interaction between Asp-77 side chain and P 1 Lys of the substrates could play an important role in the recognition of Lys residue at the P 1 position. In aspergillopepsin I, however, the substitution of Asp-77 by Asn hardly affected the P 1 Lys/Phe preference compared with wild-type enzyme, whereas the substitution of Asp-77 by other amino acids (Glu, Ser, and Thr) and the deletion of Ser-79 reduced the P 1 Lys/Phe preference with the decrease in the k cat values for the P 1 Lys substrates. This indicates that the recognition of basic P 1 side chain is not dependent on the specific electrostatic interactions between P 1 side chain and Asp-77 side chain of the enzyme (21). From a site-directed mutagenesis study of rhizopuspepsin, Lowther et al. (22) concluded that the presence of Asp at position 77 has the potential to establish an extensive hydrogen-bonding network between the enzyme and the substrate containing Lys residue at the P 1 position. In the double mutated pepsin, although the dissociation of the ␤-carboxyl group of Asp-77 would be dispensable to accommodate the Lys residue in the S 1 pocket, it enhances the stabilization of the transition state complex when substrates contain a Lys residue at the P 1 position. This may be due to the formation of the electrostatic interaction or the changes of the hydrogen-bonding pattern between Lys residue in P 1 and Asp-77 of the mutant pepsin as the pH is higher.