Molecular Basis for Substrate Specificity of Protein-tyrosine Phosphatase 1B*

Protein-tyrosine phosphatases can exhibit stringent substrate specificity in vivo, although the molecular basis for this is not well understood. The three-dimensional structure of the catalytically inactive protein-tyrosine phosphate 1B (PTP1B)/C215S complexed with an optimal substrate, DADEpYL-NH2, reveals specific interactions between amino acid residues in the substrate and PTP1B. The goal of this work is to rigorously evaluate the functional significance of Tyr46, Arg47, Asp48, Phe182, and Gln262 in substrate binding and catalysis, using site-directed mutagenesis. Combined with structural information, kinetic analysis of the wild type and mutant PTP1B usingp-nitrophenyl phosphate and phosphotyrosine-containing peptides has yielded further insight into PTP1B residues, which recognize general features, as well as specific properties, in peptide substrates. In addition, the kinetic results suggest roles of these residues in E-P hydrolysis, which are not obvious from the structure of PTP1B/peptide complex. Thus, Tyr46 and Asp48recognize common features of peptide substrates and are important for peptide substrate binding and/or E-P formation. Arg47 acts as a determinant of substrate specificity and is responsible for the modest preference of PTP1B for acidic residues NH2-terminal to phosphotyrosine. Phe182 and the invariant Gln262 are not only important for substrate binding and/or E-P formation but also important for the E-P hydrolysis step.

Protein-tyrosine phosphorylation is a universal mechanism employed for the regulation of cellular processes such as proliferation, differentiation, motility, cell-cell interactions, metabolism, gene transcription, and the immune response (1,2). The propagation and termination of signaling events controlling these cellular processes are determined by the level of phosphorylated proteins in a cell. The phosphorylation level, in turn, is maintained in an exquisite balance by the reciprocal activities of protein-tyrosine kinases and phosphatases. Thus, in addition to the study of protein-tyrosine kinases, one can appreciate the need to further characterize the dephosphoryl-ation reaction catalyzed by the protein-tyrosine phosphatases (PTPases). 1 Much is known about the catalytic mechanism of the PT-Pases (3). However, the molecular basis for PTPase substrate specificity is not well understood and remains a major unresolved issue in the field. The PTPase family is presently composed of approximately 100 enzymes, which can be either transmembrane (receptor-like) or intracellular (cytoplasmic). Membership in this family of enzymes requires the presence of the PTPase signature motif, (H/V)CX5R(S/T), housed within the catalytic domain. Outside this shared catalytic domain are various targeting and localization domains, which may be utilized for controlling and restricting PTPase substrate specificity. There have been relatively few biochemical analyses of the mechanisms that govern PTPase substrate specificity, although recent genetic and biochemical evidence suggests that in vivo PTPases can exhibit extremely stringent substrate specificity (4 -9). Importantly, it seems that, at least for PTP-PEST and the Yersinia PTPase, the catalytic domains alone exhibit very high intrinsic substrate specificity (5,7). Further understanding of the specific functional roles of PTPases in cellular signaling requires definition of physiological substrates for each individual member of the PTPase family and detailed understanding of structural features that control PTPase substrate specificity.
Using synthetic pTyr-containing peptides that correspond to natural phosphorylation sites in proteins, several groups have demonstrated that PTPases display a range of k cat /K m values for these relatively short peptide substrates (10 -19). We have shown that the undecapeptide DADEpYLIPQQG, modeled on an autophosphorylation site (Tyr 992 ) of the epidermal growth factor receptor (EGFR) is an optimal peptide substrate for both the Yersinia PTPase and the mammalian PTP1 (13,14). In fact, the k cat /K m values for this peptide approach the diffusional limit and are 3-4 orders of magnitude higher than that of phosphotyrosine alone, suggesting that amino acid residues flanking the pTyr moiety contribute to high affinity binding. Furthermore, the Yersinia PTPase and PTP1 display a preference for acidic residues at positions NH 2 -terminal to the phosphorylated tyrosine (14) and require a minimum of six amino acid residues (DADEpYL) for efficient binding and catalysis (17). The recently solved crystal structure of the catalytically inactive PTP1B/C215S complexed with this minimal peptide substrate reveals specific interactions between amino acid residues in the substrate and PTP1B (20). These results suggest that a number of residues in PTP1B are important for peptide substrate recognition and that the sequence surrounding the pTyr residue plays a key role in determining its recognition by PTPases. The structural observations serve as the basis for further functional studies of PTP1B substrate specificity and catalytic mechanism. In order to clarify the role of primary structure in substrate recognition and to test experimentally the importance of specific amino acid residues in PTP1B for substrate recognition, we have mutated a number of residues in PTP1B depicted in the crystal structure to interact with the peptide substrate ( Fig. 1). Furthermore, because PTPase catalysis involves a covalent thiol phosphate intermediate (3), it is not immediately obvious whether residues implicated in substrate binding would have an effect on the hydrolysis of the phosphoenzyme intermediate. The role of these active site amino acids in substrate recognition and processing has been studied by kinetic analyses using several pTyr-containing peptides and p-nitrophenyl phosphate as substrates. In combination with the previously made structural observations, our kinetic data and PTPase amino acid sequence alignment have greatly enhanced our understanding of the function of several conserved PTPase residues that recognize common features in pTyr-containing peptides and of some less conserved residues that may serve as specificity determinants for substrate selectivity.
Protein Purification-The wild type PTP1B (residues 1-321) was expressed in Escherichia coli and purified to homogeneity as described (21). The PTP1B/Y46F transformed cells were grown for a 4-h period instead of overnight as for the wild type. All mutant forms except R47A and R47E were purified in manner similar to wild type PTP1B. The pellet from the 2-liter cultures of R47A and R47E were instead resuspended in 40 ml of 50 mM 3,3-dimethylglutarate, pH 5.7, containing 1 mM EDTA and 1 mM dithiothreitol. The supernatant was loaded onto a CM-Sephadex column (100-ml bed volume), which was pre-equilibrated with the same buffer. The column was washed with the same buffer until the elute had an absorbance at 280 nm of less than 0.03. R47A and R47E were eluted from this column with a salt gradient consisting of the buffer in one chamber and buffer plus 500 mM sodium chloride in the other. Protein purity was checked by 12% SDS-gel electrophoresis, which showed a single band at 37 kDa. The protein concentration was determined from absorbance measurement at 280 nm using an absorbance coefficient of A 1 mg/ml 280 ϭ 1.24.

Determination of Kinetic Constants Using pNPP as a Substrate-The
PTPase activity was assayed at 30°C in a reaction mixture (0.2 ml) containing pNPP concentrations ranging from 0.2 to 5 K m . The following buffered solution was used for activity measurements: at pH 7.0, 50 mM 3,3-dimethylglutarate, 1 mM EDTA, and the ionic strength of 0.15 M, adjusted by addition of NaCl. Initial rate measurements for the enzyme catalyzed hydrolysis of pNPP were conducted as described previously (22). Michaelis-Menten kinetic parameters were determined from a direct fit of the data to the Michaelis-Menten equation using the nonlinear regression program KINETASYST (IntelliKinetics, State College, PA).
Determination of Kinetic Constants Using pTyr-containing Peptides as Substrates-All assays were performed at 30°C in the following buffer: at pH 7, 50 mM 3,3-dimethylglutarate, 1 mM EDTA, ionic strength of 0.15 M. A continuous spectrophotometric assay described previously was employed to determine k cat , K m (13), and k cat /K m (19) for the pTyr-containing peptides. The dephosphorylation reaction can be monitored by either an increase in absorbance at 282 nm or an increase in fluorescence at 305 nm (13). Fluorometric and absorbance determinations were performed on a Perkin-Elmer LS50B fluorometer and a Perkin-Elmer Lambda 14 spectrophotometer, respectively. The instruments were equipped with a water-jacketed cell holder, permitting maintenance of the reaction mixture at the desired temperature (30°C).
Inhibition Studies-The procedure for the preparation of stock solution of sodium orthovanadate was described previously (19). The inhibition constants for vanadate and arsenate were determined for the wild type PTP1B and the mutant PTPases in the following manner. At various fixed concentrations of the oxyanion, the initial rate at various pNPP concentrations was measured by following the production of p-nitrophenol as described above. The inhibition constant and inhibition pattern were evaluated using the direct curve-fitting program KINETASYST (IntelliKinetics) to appropriate equations.

RESULTS AND DISCUSSION
The goal of this study was to experimentally evaluate the role of PTP1B residues that are in direct contact with the phosphopeptide substrate DADEpYL-NH 2 , as shown in the crystal structure of the PTP1B/C215S-substrate complex ( Fig.  1) (20). We have focused our effort on residues Tyr 46 , Arg 47 , Asp 48 , Phe 182 , and Gln 262 . Crystallographic structural studies have revealed that Tyr 46 , Arg 47 , and Asp 48 are located in a surface loop between the ␣1 helix and the ␤1 strand and are directly involved in peptide substrate binding (20). Interestingly, these three residues are absent in the dual specificity phosphatases (23). Phe 182 is found in a flexible surface loop (the WPD loop) in the vicinity of the active site and is proposed to be important for the loop closure. The side chain of Gln 262 is observed to interact with the phenyl ring of pTyr and to define a portion of the rim for the pTyr binding pocket shown in the structure of PTP1B/C215S-substrate complex (20).
To gain further insight into the roles of these amino acid residues in the binding and hydrolysis of pTyr-containing peptide substrates by PTP1B, we have undertaken a mutational analysis that relates the resulting kinetic effects to the structure of PTP1B-substrate complex. We have changed Tyr 46 to Phe and Ala, Arg 47 to Glu and Ala, Asp 48 to Ala, Phe 182 to Ala, and Gln 262 to Ala. All of the recombinant proteins were expressed in E. coli and purified to near homogeneity as judged by SDS-polyacrylamide gel electrophoresis. In order to confirm that any mutations in these residues did not perturb the active site structure of PTP1B, inhibition studies were performed on all purified PTP1B mutants. The inhibitors utilized were the PTPase transition state analog vanadate (24,25) and the phosphate mimic arsenate. Both of these small molecules display a competitive inhibition pattern for PTP1B. Vanadate and arsenate competitively inhibited all mutant forms of PTP1B with dissociation constants, K i , ranging between approximately 0.3and 4-fold for vanadate and between approximately 0.2-and 2-fold for arsenate in comparison with the wild type enzyme ( Table I). The K i values for vanadate are slightly higher than published results (26) and can be attributed to the presence of approximately 20 M EDTA in all reactions. EDTA is a known chelator of vanadate and chelation would result in elevated K i values for vanadate. Collectively, these results indicate that alterations in these residues do not lead to significant structural changes in the PTP1B active site.
The effects of substitutions at Tyr 46 , Arg 47 , Asp 48 , Phe 182 , and Gln 262 in PTP1B were evaluated using the low molecular weight aryl phosphate, pNPP, and several pTyr-containing peptides as substrates. The overall mechanism of the PTPasecatalyzed reaction involves a number of steps that are represented schematically in Scheme 1, where ArOPO 3 2Ϫ can be either aryl phosphates or pTyr-containing peptides. The reaction proceeds through a sequence, involving binding of the substrate, which is then cleaved with phosphoryl transfer (k 2 ) to the active site nucleophilic Cys residue. Subsequent base-catalyzed reaction with water cleaves the phosphoenzyme intermediate (k 3 ), and release of phosphate completes the catalytic cycle (3). The kinetic parameter k cat /K m monitors the reaction beginning with binding of the substrate and up to and including the first irreversible step in the kinetic mechanism, which is represented in Scheme 1 as the E-P formation step accompanied by release of phenol or Tyr-containing peptide. For pNPP, k cat /K m is primarily limited by the chemical step (i.e. E-P formation) (27). The k cat /K m for phosphopeptides may be greatly influenced by substrate binding, 2 as they are 3-4 orders of magnitude more efficient substrates than aryl phosphates (17). The k cat term describes the rate-limiting step under saturating concentrations of substrate and is mostly determined by E-P hydrolysis step (17,28).
The kinetic parameters for the wild type and mutant PTP1B catalyzed hydrolysis of pNPP are summarized in Table II. A continuous spectrophotometric assay described previously (13) was used to follow the dephosphorylation of tyrosine on phosphopeptides. Both k cat and K m can be obtained from direct fitting of the complete time course of the reaction to the integrated form of the Michaelis-Menten equation using a nonlinear least-squares algorithm (13). The k cat /K m value can also be determined directly by monitoring the reaction at substrate concentrations Ͻ ϽK m , so that the reaction is first order with respect to [S] (19). In every case, k cat /K m values determined by this approach agreed very well with those calculated from the ratio of k cat and K m determined from the integrated form of the Michaelis-Menten equation. Values of k cat , K m , and k cat /K m for the wild type and mutant PTP1B phosphatases catalyzed dephosphorylation of peptide substrates are listed in Tables III  and IV. Tyr 46 -The side chain of Tyr 46 is engaged in interactions with the main chain atoms and the aromatic ring of pTyr in peptide substrates (Fig. 1). A hydrogen bond is also observed between the OH group of Tyr 46 and the side chain of Ser 216 , which may be required for the stabilization of the conformation of Tyr 46 and its interaction with the substrate (20). To probe the role of the OH group in Tyr 46 , it was mutated to Phe, which eliminates the OH group while retaining the aromatic ring. Using pNPP as a substrate, the k cat for Y46F was slightly lower that that of the wild type, whereas the K m value was similar (Table II). Using the EGF receptor peptide DADEpYLIPQQG (EGFR 988 -998 ) as a substrate, the k cat value for Y46F was only 2-fold lower than that of PTP1B, whereas the K m was 2-fold higher (Table III). It appears that the OH group of Tyr 46 may not be essential for activity, and a Phe residue can adequately substitute for Tyr at position 46. Indeed, in the Yersinia PTPase (Yop51), which is the most active PTPase, and in human PTPD1 (Fig. 3), a Phe residue, instead of a Tyr, occupies the corresponding position. It is possible that the OH group on Tyr 46 may be required for PTP1B protein stability, because the recombinant Y46F protein was degraded in E. coli during normal overnight culture after isopropyl-1-thio-␤-D-galactopyranoside induction (data not shown). In order to prepare sufficient amount of Y46F protein and minimize protein degradation, a shorter incubation time was employed (see under "Experimental Procedures").    To further study the function of Tyr 46 , it was changed to an Ala. The substitution of the Tyr side chain by a methyl group at position 46 resulted in a 7-fold decrease in k cat , a 9-fold increase in K m , and a 66-fold decrease in k cat /K m with pNPP as a substrate. Furthermore, Y46A displayed a 2-fold decrease in k cat , whereas its K m increased 380 times against the EGFR 988 -998 peptide, resulting in a 780-fold decrease in k cat /K m when compared with the wild type enzyme. Our results are similar to a previous study in which Tyr 46 was changed to a Ser or a Leu (9). The Y46S and Y46L mutants displayed a 15-fold decrease in k cat and a 17-fold increase in K m using the tyrosine phosphorylated lysozyme as a substrate. The combined effect was a 240-fold lower k cat /K m for Y46S or Y46L. Collectively, these results indicate that Tyr 46 is more important for substrate binding and/or E-P formation than E-P hydrolysis and suggest that the aromaticity of Tyr 46 , not the hydrophobicity, is important for the recognition of pTyr and optimal substrate orientation. This is consistent with the crystal structure, which shows that the side chain of Tyr 46 is involved in ainteraction with the phenyl ring of pTyr and contributes to the formation of the approximately 9-Å-deep pTyr binding pocket, which restricts substrate hydrolyses to pTyr-containing peptides for PTP1B (20).
Arg 47 -It has been shown that PTP1 (the rat homolog of human PTP1B) displays moderate selectivity toward acidic amino acids NH 2 -terminal to the pTyr in peptide substrates (14). The structure of PTP1B/C215S in complex with DADEpYL-NH 2 revealed that the guanidinium group of Arg 47 forms salt bridges with the carboxylate groups at the Ϫ2 and Ϫ1 positions of the peptide substrate and a long hydrogen bond with the main chain carbonyl at the Ϫ4 position (Fig. 1, 20). Thus, Arg 47 may be responsible for the preference of PTP1B for peptides with acidic residues NH 2 -terminal to pTyr. In addition, the main chain nitrogen of Arg 47 is hydrogen bonded to the main chain carbonyl oxygen of the peptide substrate residue at the Ϫ2 position. Because Arg 47 does not interact directly with pTyr (or pNPP) and is involved in interaction with acidic residues away from the pTyr site, it was anticipated that mutations at residue 47 would not lead to any alteration in the rate of PTP1B catalyzed pNPP hydrolysis. Indeed, R47A and R47E exhibited k cat and K m values similar to those of the wild type PTP1B with pNPP as a substrate (Table II), indicating that Arg 47 is not directly involved in the hydrolysis of small aryl phosphates.
If Arg 47 plays a role in controlling PTP1B substrate specificity by interacting directly with acidic amino acids in pTyrcontaining peptides, then abrogation of this interaction should lead to a decrease in k cat /K m , the substrate specificity constant, toward peptide substrates. Indeed, the k cat /K m values for the R47E and R47A mutants decreased by 8-fold and 2-fold, respectively, for the DADEpYLIPQQG peptide when compared with PTP1B (Table III). This decrease can be attributed to increases in K m , as k cat remained almost unchanged for R47E and R47A.
The lack of substitutional effect on k cat (which follows E-P hydrolysis) at residue 47 is understandable because the dephosphorylated tyrosyl peptide has already left the active site during E-P hydrolysis and Arg 47 is positioned away from the site where E-P hydrolysis takes place. It is worth noting that the effect of charge reversal (Arg to Glu mutation) at residue 47 is more severe than simple removal of the positive charge (Arg to Ala substitution). The fact that R47E displayed a greater loss in k cat /K m than R47A may be explained by charge repulsion between the Glu residue at position 47 and the acidic residues in the substrate. This is consistent with the notion that Arg 47 specifically interacts with acidic residues in the substrate.
To further probe the role of Arg 47 in substrate recognition, we also analyzed the kinetics of PTP1B-catalyzed hydrolysis of Ala-substituted EGFR 988 -998 peptides (Table IV). Wild type PTP1B suffered a 4-fold and 2-fold decrease in k cat /K m when the acidic residue at the Ϫ1 or Ϫ2 position, respectively, was changed to an Ala. A 17-fold decrease in k cat /K m was observed for the wild type PTP1B when all four NH 2 -terminal residues were Ala residues. These results are consistent with the crystal structure that depicts Arg 47 interacting with both the Ϫ2 and Ϫ1 acidic residues of the substrate. The kinetic data also suggest that the interaction between Arg 47 and the Ϫ1 residue is stronger than that between Arg 47 and the Ϫ2 residue. Interestingly, the effect of the Arg 47 to Glu mutation on k cat /K m was mitigated as the overall negative charge NH 2 -terminal to pTyr in the substrate was reduced or removed with the substitution of Ala at the Ϫ1, Ϫ2 or all three NH 2 -terminal acidic residues of the EGFR 988 -998 peptide (Table IV). Moreover, R47A exhibited little preference for the native and Ala-substituted EGFR 988 -998 peptides. These results suggest that Arg 47 is critical for promoting selection of negatively charged residues at positions amino-terminal to pTyr. Finally, when KRSpYEEHIP (modeled after the phosphorylation site Tyr 1316 in insulin receptor) was used as a substrate, a reversal of substrate specificity was observed for PTP1B, R47A, and R47E (Table IV). Thus, KRSpYEEHIP was processed by R47E 5-and 2-fold more efficiently than the wild type PTP1B and R47A, respectively. Based on the results described above, one would predict a more pronounced reversal of specificity if a peptide with a basic residue at the Ϫ1 position were used as a substrate. Collectively, the observed effects of mutations at Arg 47 in PTP1B or substitutions of Ϫ1 and Ϫ2 acidic residues by Ala in peptide substrates on k cat /K m are consistent with the disruption of the observed hydrogen-bonds between Arg 47 and the side chains at the Ϫ1 and Ϫ2 positions seen in the crystal structure of PTP1B and the DADEpYL-NH 2 peptide. These results provide strong evidence that amino acid at position 47 is an important substrate specificity determinant for PTP1B that recognizes residues at the Ϫ1 and Ϫ2 positions of pTyr-containing peptides. However, it is important to point out that the contribution from Arg 47 in determining the in vivo substrate specificity is unknown. Unlike PTP-PEST, which exhibits stringent substrate specificity (5), the substrate specificity of PTP1B is more relaxed in vivo when its COOH-terminal localization domain is removed (5,9).
Asp 48 -The pTyr residue in the peptide substrate adopts a right-handed ␣-helical conformation in the PTP1B-bound state. It is proposed, based on the crystal structure, that the primary determinant of the peptide conformation at the pTyr site is Asp 48 because it forms two hydrogen bonds with the main chain nitrogens of pTyr and the ϩ1 residue (Fig. 1). Thus, substitution at Asp 48 is not expected to have any adverse effects on the hydrolysis of pNPP, which lacks the corresponding main chain nitrogens. Indeed, D48A exhibited k cat and K m values similar to those of the wild type PTP1B with pNPP as a  (Table II). However, when the EGFR 988 -998 peptide was used as a substrate, the D48A mutant exhibited a 22-fold decrease in k cat /K m and only a slightly lower k cat as compared with the wild type enzyme (Table III). The more severe effect on k cat /K m is consistent with the loss of two hydrogen bonds to the main-chain nitrogens at the pTyr and the ϩ1 residues as seen in the crystal structure of PTP1B and the DADEpYL-NH 2 peptide. The D48A mutation has little effect on E-P hydrolysis, because the tyrosyl peptide with which the side chain of Asp 48 interacts with is no longer present in active site during E-P hydrolysis. Thus, elimination of the carboxyl group at residue 48 primarily reduces the affinity of peptide substrates and/or the rate of the phosphoryl group transfer from the peptide substrate to the PTP1B active site Cys, whereas it has little effect on the hydrolysis of the phosphoenzyme intermediate. These kinetic results are consistent with the structural data and suggest that Asp 48 plays an important role in positioning the peptide substrates in an optimal conformation for peptide binding and/or the initial nucleophilic attack by the active site Cys residue. Phe 182 -Phe 182 resides in a flexible surface loop (the WPD loop) that harbors the general acid/base Asp 181 (20, 29 -32). In the unliganded PTP1B and Yersinia PTPase structures, the WPD loop was observed in an open conformation such that Asp 181 (Asp 356 in Yersinia PTPase) is greater than 10 Å away from the phosphate binding site (30,33). However, upon substrate binding, the WPD loop adopts the closed conformation that covers the active site like a flap, which places the general acid/base Asp 181 carboxylate close to the phenolic oxygen of pTyr ( Fig. 1) (20). This movement shifts the side chain of Phe 182 12 Å toward the pTyr binding site, allowing the formation of hydrophobic stacking interactions with the phenyl ring of the pTyr residue of the substrate. In the Yersinia PTPase, the WPD loop adopts a closed conformation in the presence of oxyanions (30,31). In contrast, WPD loop closure is not observed in PTP1B when tungstate is bound (33). The interaction between pTyr and Phe 182 is therefore proposed to be crucial for the formation of the WPD loop closed PTP1B conformation (20), which should be catalytically competent.
The K m values for the F182A-catalyzed reactions were not significantly different from those of the native enzyme (Tables  II and III). In contrast, the k cat /K m for the F182A-catalyzed pNPP and EGFR 988 -998 peptide was 9-and 13-fold lower than those of the wild type enzyme. These results suggest that the hydrophobic interactions between Phe 182 and the phenyl ring of pTyr, as observed in the crystal structure of the PTP1B/ substrate complex, are required for substrate binding and/or E-P formation (Scheme 1) by bringing Asp 181 into proximity for proton donation to the leaving group phenolic oxygen. Interestingly, the k cat for the F182A catalyzed pNPP and peptide substrate hydrolysis was also decreased by 16-and 19-fold, respectively, as compared with the wild type PTP1B. These are somewhat surprising results because the tyrosyl part of the substrate, which interacts with the side chain of Phe 182 , has already left the active site during the hydrolysis of E-P. Our kinetic results indicate that Phe 182 also plays an important role in the hydrolysis of E-P. Because Asp 181 functions as a general base to activate the nucleophilic water for E-P hydrolysis, its precise location in the active site is crucial to the dephosphorylation activity. It is possible that alteration of Phe 182 structure affects the conformation of the WPD loop and therefore the exact positioning of Asp 181 . Interestingly, although Phe 182 plays a key role in both the E-P formation and E-P hydrolysis, it is not well conserved within the PTPase family (29). For example, the structurally equivalent residue is Gln in the Yersinia PTPase, Phe in PTP1B, His in LAR, and Met in yeast  (34). Thus, the residue immediately following the catalytic Asp in the WPD loop may be important for the attainment of the catalytic competent active site geometry. Variations in the structure of the WPD loop may modulate the dynamic and catalytic properties of the PTPases (20,31,32). Further study on the effect of loop dynamics on catalytic efficiency for PTPases should clarify the exact functional role of the WPD loop residues.
Gln 262 -The role of Gln 262 in PTP1B catalysis was also investigated, because the side chain of Gln 262 interacts with the phenyl ring of pTyr and defines a portion of the rim for the pTyr binding pocket shown in the structure of PTP1B/C215S-substrate complex (Fig. 1). Previously, it was shown that the k cat /K m and k cat for the PTP1B/Q262A catalyzed hydrolysis of phosphorylated lysozyme were reduced by 7-and 83-fold, respectively (9). Similarly, Q262A exhibited 2-and 13-fold decrease in k cat /K m for pNPP and the EGFR 988 -998 peptide, respectively (Tables II and III), suggesting that Gln 262 may help to align the phenyl ring of pTyr in the E-P formation step. In addition, the k cat for the Q262A catalyzed hydrolysis of pNPP and EGFR 988 -998 peptide was reduced 80 -150-fold, suggesting that Gln 262 may also participate in the E-P hydrolysis step. Work on the structural equivalent Gln 446 in the Yersinia PTPase (35) and Gln 262 in PTP1B (36) 3 suggests that the invariant glutamine residue is important for the optimal positioning of the nucleophilic water molecule for efficient E-P hydrolysis. These observations are consistent with the dramatic decrease in K m for the Q262A mutant (Tables II and III  and Ref. 9), which is indicative of increased accumulation of the E-P intermediate in the Q262A-catalyzed reactions.
In summary, results from kinetic studies of the wild type PTP1B and site-directed mutants of Tyr 46 , Arg 47 , Asp 48 , Phe 182 , and Gln 262 are consistent with conclusions reached from the crystal structure of PTP1B/C215S bound with DADEpYL-NH 2 (20) regarding the role of these residues in substrate binding and/or the E-P formation step. In addition, the kinetic results reveal functional significance of these residues in E-P hydrolysis, which is not obvious from the structure of PTP1B/substrate complex. Thus, Tyr 46 forms part of the pTyr binding pocket, and Asp 48 helps to define the main chain peptide conformation at the pTyr site. Both residues recognize common features of the peptide substrates and are important for peptide substrate binding and/or the E-P formation step. These two residues are not expected to interact with the phosphoryl moiety of the substrate and therefore have little influence on the E-P hydrolysis step. Tyr 46 (or Phe) and Asp 48 (or Asn) are highly conserved throughout the PTPase family (Fig.  2), suggesting that results described in this paper should be applicable to most PTPases. However, there are exceptions, most notably the cytoplasmic EC-PTP and LC-PTP, and the receptor-like PTPases, PTP, IA-2, and IA-2␤. It is not clear whether these proteins are active tyrosine phosphatases, because there are also sequence deviations in the WPD loop region in PTP, IA-2, and IA-2␤. Arg 47 acts as a determinant for substrate specificity and is responsible for the modest preference of PTP1B for acidic residues NH 2 -terminal to pTyr. There are more sequence variations at Arg 47 (Fig. 2), suggesting that PTPases with different residues at this position may show different sequence selectivity. The interaction between the benzene ring of Phe 182 and pTyr in the substrate may be important for positioning the general acid Asp 181 in the WPD loop for effective protonation of the tyrosine phenolate leaving group. Surprisingly, our data indicate that Phe 182 also plays a role in the E-P hydrolysis, possibly by maintaining the general base Asp 181 and the WPD loop in a catalytically competent form. It is possible that residues in the WPD loop may play a role in controlling the activity of PTPases. Finally, the invariant Gln 262 is not only involved for peptide substrate binding and/or E-P formation, but also in E-P hydrolysis. Evidence suggests that Gln 262 is important for maintaining the strictly hydrolytic activity in PTPases (35,36). 3