A Novel Dual Specificity Phosphatase SKRP1 Interacts with the MAPK Kinase MKK7 and Inactivates the JNK MAPK Pathway

Mitogen-activated protein kinases (MAPKs) are activated in response to various extracellular stimuli, and their activities are regulated by upstream activating kinases and protein phosphatases such as MAPK phosphatases (MKPs). We report the identification and characterization of a novel MKP termed SKRP1 (SAPK pathway-regulatingphosphatase 1). It contains an extended active site sequence motif conserved in all MKPs but lacks a Cdc25 homology domain. Immunoblotting analysis revealed that SKRP1 is constitutively expressed, and its transcripts of 4.0 and 1.0 kb were detected in almost tissues examined. SKRP1 was highly specific for c-Jun N-terminal kinase (JNK) in vitro and effectively suppressed the JNK activation in response to tumor necrosis factor α or thapsigargin. Endogenous SKRP1 was present predominantly in the cytoplasm and co-localized with JNK. However, SKRP1 does not bind directly to its target JNK, but co-precipitation of SKRP1 with the MAPK kinase MKK7, a JNK activator, was found in vitro and in vivo. Furthermore, we found that SKRP1 did not interfere with the co-precipitation of MKK7 with JNK. Together, our findings indicate that SKRP1 interacts with its physiological substrate JNK through MKK7, thereby leading to the precise regulation of JNK activity in vivo.

Here we have identified and characterized a novel dual specificity MKP that is highly specific for JNK in vitro and negatively regulates the activation of the JNK pathway. We also found that the enzyme interacts physically with the MAPKK MKK7, a JNK activator, but not with JNK. Thus, these findings suggest that the enzyme interacts indirectly with its physiological target JNK through MKK7, consequently leading to the specific regulation of JNK signaling pathway. Therefore, we have termed this novel enzyme SKRP1, as (SAPK pathwayregulating phosphatase 1.

EXPERIMENTAL PROCEDURES
Materials-Unless otherwise noted, routine chemical materials were obtained from either Sigma or Wako (Osaka, Japan). Molecular standards and restriction enzymes were purchased from New England Biolabs (Beverly, MA). * This work was supported by Research for the Future Program. 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) 1. SKRP1 is a novel member of DSP family. A, nucleotide and deduced amino acid sequences of mouse SKRP1 cDNA. The conceptually translated amino acid sequences are presented in single-letter code below the nucleotide sequences. The stop codon is shown by an asterisk. The numbers on the left or the right denote the positions of nucleotides or amino acids, respectively. The active site sequence motif conserved in all DSPs is indicated by a shaded box. The catalytic cysteine is double-circled. Note that the conserved Ser in the active site sequence motif is replaced by Ala (a circle at the amino acid position 156). The peptide sequence spanning residues 80 -97 of the mouse SKRP1 protein was used to generate the rabbit polyclonal antiserum against SKRP1 and is shown by a double underline in A-C. B, alignment of amino acid sequences of mouse and human SKRP1. The numbers denote the amino acid positions from the first Met. The active site sequence motif conserved in all DSPs is indicated by a shaded box. The catalytic cysteine is indicated by an asterisk. C, comparison of the phosphatase catalytic domains of mouse SKRP1 and other members of DSP family (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) using the ClustalX software. To optimize homology, gaps were inserted as denoted by dots. The black boxes or the gray boxes indicate identities or similarities, respectively. The catalytic cysteine is indicated by an asterisk. D, schematic representation of mouse SKRP1 amino acid identity with other members of DSP family. CH2 domains and phosphatase catalytic domains are indicated by black and shaded boxes, respectively. The percentages indicate the amino acid identity with mouse SKRP1 within the phosphatase catalytic domain. E, a radial dendrogram representing the relation between DSP family (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30). amplified from the mouse embryonic cDNA library with a sense primer 5Ј-GTGCTCTTCTCTGGAAGG-3Ј and an antisense primer 5Ј-TCCG-GATCCTTTAAAATGTATTAATTCAGG-3Ј, corresponding to the both ends of cDNA. The coding sequence of SKRP1 was amplified with a sense primer 5Ј-GAATTCGTCGACATGCATTCCCTGAACCAAAAAA-TC-3Ј having a SalI site and an antisense primer 5Ј-GCGGCCGCTG-CAGTCACAGACCATGGGTCGTGTC-3Ј having a NotI site to bear an in-frame fusion to epitope-tagged expression vectors. To isolate the human homologue of SKRP1, we screened human EST data base using murine full-length SKRP1 cDNA as a probe, and we obtained overlapping clones that match the 5Ј-and 3Ј-ends of the SKRP1 sequence.
Based on the sequences from these clones, we designed a sense primer 5Ј-TTCTCCAACGAAGTCTTATA-3Ј and an antisense primer 5Ј-AGCATTTAATCACACATTAT-3Ј and cloned putative full-length cDNA from 293 cells by reverse transcriptase-PCR.
Site-directed Mutagenesis-SKRP1 C149S mutants were generated in a standard two-step megaprimer PCR with mutated oligonucleotides as primers and full-length SKRP1 complementary DNA as a template. The first step included two PCRs that span the mutation site. Final PCR products were cloned into appropriate vectors, and mutations were verified by DNA sequencing.
Cell Culture and Transfection-COS-7 and 293 cells were maintained in high glucose Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 100 units/ml penicillin, and 100 units/ml streptomycin. NIH3T3 or BALB/3T3 cells were maintained in high glucose Dulbecco's modified Eagle's medium or minimum essential medium, respectively, supplemented with 10% heat-inactivated calf serum, 2 mM L-glutamine, 100 units/ml penicillin, and 100 units/ml streptomycin. For transfection assays, cells were transfected with the appropriate combinations of expression plasmids using LipofectAMINE (Invitrogen), according to the manufacturer's protocol. The total amount of DNA in each transfection was kept constant by supplementation with empty vector DNAs.
In Vitro Dephosphorylation of MAPKs-COS-7 cells were transfected with HA-MAPKs or GST-MAPKs. After culture for 24 h, the cells were left untreated or treated with 250 ng/ml 12-O-tetradecanoylphorbol-13acetate after serum deprivation for 8 h (for ERK2) or 10 g/ml anisomycin (for JNK2 and p38␣) and then incubated for 20 min. Tagged MAPKs were precipitated from the cell lysates by incubation with anti-HA antibody (12CA5, Roche Diagnostics) and protein G-Sepharose beads (Amersham Biosciences), or with glutathione-Sepharose 4B (Amersham Biosciences) for 3 h at 4°C. The precipitates were incubated for 1 h at 37°C with the indicated amounts of His-SKRP1, His-SKRP1 C149S, GST-SKRP1, GST-JSP-1, or GST-MKP-6 in a reaction buffer containing 20 mM Tris-HCl (pH 7.5) and 5 mM MgCl 2 . The reaction products were subjected to immunoblotting using antibodies specific to dually phosphorylated MAPKs (Cell Signaling Technology), anti-phos-phothreonine antibody (PTR-8, Sigma), or anti-phosphotyrosine antibody (SC-508, Santa Cruz Biotechnology).
Immunoblotting-Protein concentration was assessed using the DC Protein Assay kit (Bio-Rad). Proteins were loaded at 50 g per well and separated by SDS-PAGE. After overnight transfer on polyvinylidene difluoride membranes and blocking for at least 5 h in Tris-buffered saline, 5% non-fat skim milk, and 0.1% Tween 20, blots were hybridized and proteins were visualized by enhanced chemiluminescence (ECL, Amersham Biosciences).
Northern Blot Analysis-A 1.1-kb fragment of mouse SKRP1 cDNA was amplified by PCR and labeled with [␣-32 P]dCTP using a Rediprime II DNA labeling system (Amersham Biosciences). Labeled probe was hybridized to a mouse multiple tissue Northern blot (Clontech) in ExpressHyb hybridization solution (Clontech).
Rabbit Polyclonal Antiserum to SKRP1-The rabbit polyclonal antiserum to SKRP1 was produced by immunizing rabbits with a synthetic peptide NH 2 -GCAHDLELLRKHKVTHILNV-COOH (the bold letters correspond to residues 80 -97 of the mouse SKRP1 protein), coupled to keyhole limpet hemocyanin.
Kinase Assays-Cells were lysed on ice in the following buffer: 20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100, 12 mM ␤-glycerophosphate, 1 mM sodium orthovanadate, 2 mM EGTA, 3 mM dithiothreitol, and a protease inhibitor mixture (Complete, Roche Diagnostics). Lysates were clarified at 15,000 ϫ g for 15 min, and either analyzed immediately or stored at Ϫ80°C. Immunoprecipitations or glutathione-Sepharose precipitations from cell lysates were carried out using anti-HA antibody (12CA5, Roche Diagnostics) together with protein G-Sepharose beads (Amersham Biosciences) or using glutathione-Sepharose 4B (Amersham Biosciences), respectively. After incubation for 3 h at 4°C, the complexes were washed three times with lysis buffer and then twice with kinase reaction buffer containing 20 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 12 mM ␤-glycerophosphate, 1 mM sodium orthovanadate, and 1 mM dithiothreitol. The kinase assays were initiated by the addition of each substrate protein (3.0 g) and 10 Ci of [␥-32 P]ATP in a final volume of 25 l. The reactions were terminated after 20 min at 30°C by addition of Laemmli sample buffer. Samples were boiled for 5 min, and the phosphorylation of the substrate proteins was examined by SDS-PAGE with a Fuji BAS2000 image analyzer.
Immunofluorescence Staining-NIH3T3 or BALB/3T3 cells were seeded in 8-well culture slides (BIOCOAT, BD PharMingen) overnight and then fixed in 4% paraformaldehyde in phosphate-buffered saline for 15 min at room temperature, followed by permeabilization in phosphate-buffered saline containing 0.2% Triton X-100. The cells were incubated with primary antibodies in blocking buffer (phosphate-buffered saline containing 0.2% Triton X-100 and 3% bovine serum albumin) overnight at 4°C. After washing three times, the cells were incubated with fluorescein isothiocyanate-conjugated goat anti-rabbit antibodies (Zymed Laboratories Inc. laboratories) for SKRP1 staining and Cy3-conjugated goat anti-mouse antibodies (Zymed Laboratories Inc. laboratories) for JNK2 staining, respectively, to detect bound primary antibodies. Coverslips were mounted on slides with Permafluor (Lipshaw Immunon). Fluorescent images were Kalman-averaged and analyzed by using a MicroRadiance laser-scanning confocal microscope (Bio-Rad).
In Vitro Binding Assays-GST or GST fusion proteins (1.0 g) immobilized on 15 l of glutathione-Sepharose beads were incubated with the indicated His-tagged proteins or cell lysates in lysis buffer for 3 h at 4°C. After extensively washing with the same buffer, glutathione-Sepharose precipitates were immunoblotted for the presence of Histagged proteins or HA-tagged MKK7 derivatives using mouse monoclonal anti-His antibody (Qiagen) or rabbit polyclonal anti-HA antibody (MBL), respectively, followed by Coomassie staining. 1% of the proteins used in binding reactions was also loaded as a positive control (Input).
Metabolic Labeling-Metabolic labeling of cells was performed in methionine-free medium supplemented with 35 S-EasyTag (Perkin-Elmer Life Sciences) to 0.2 mCi/ml for 5 h. Cell lysates were prepared in lysis buffer and precipitated with 1.0 g of GST or GST-SKRP1 fusion proteins. Glutathione-Sepharose precipitates were subjected to 10% SDS-PAGE, followed by autoradiography.
Co-precipitation Analysis-Cell lysates from COS-7 cells transfected with the appropriate combinations of expression plasmids were precipitated using glutathione-Sepharose 4B as described previously (38 -44). The glutathione-Sepharose precipitates were then washed five times with lysis buffer, resolved on SDS-PAGE, and examined by immunoblotting using appropriate antibodies.

Identification of SKRP1 as a Member of MKP Family-DSPs
display a high degree of sequence similarity within an active site sequence motif. Therefore, to identify novel DSPs, we used both degenerate PCR-based methods and EST data bases for cDNA fragments containing this conserved motif. Several genes were isolated from a mouse embryo and human embryonic kidney (293) cells, including JSP-1 (30) and MKP-6 (26). One clone encodes a novel protein, which we have termed SKRP1. The sequence of the mouse SKRP1 cDNA (GenBank TM accession number AB051896) revealed an open reading frame of 220 amino acids with a predicted molecular weight of 25,000 (Fig. 1A). The human (GenBank TM accession number AB063186) SKRP1 cDNA has 83% sequence identity with fulllength mouse SKRP1 cDNA (data not shown) and encodes protein of 217 amino acids that have 81% sequence identity with the mouse SKRP1 protein (Fig. 1B). The SKRP1 protein contains the extended active site sequence motif conserved in all DSPs, within which the cysteine residue is essential for catalytic activity (Fig. 1C) and displayed 33-40% identity to the phosphatase catalytic domains of other members of the DSP family (Fig. 1D). However, the predicted amino acid sequence of SKRP1 does not contain two N-terminal Cdc25 homology domains (CH2 domains), characteristic of almost other members of DSP family (Fig. 1D). The radial dendrogram showed that SKRP1, JSP-1 (30), MKP-5 (24,25), and hVHR (14) constitute a subfamily of DSPs (Fig. 1E).
To determine whether SKRP1 possesses intrinsic phosphatase activity, GST-SKRP1 fusion proteins were assayed for enzymatic activity toward p-nitrophenyl phosphate (pNPP), compared with JSP-1 (30) and MKP-6 (26), two members of DSP family. These GST fusion proteins hydrolyzed pNPP in a dose-dependent manner, and sodium vanadate, a potent inhibitor of tyrosine-specific phosphatase, effectively inhibited this catalytic activity ( Fig. 2A). A point mutation in the catalytic cysteine of SKRP1 (C149S) abolished the phosphatase activity as seen in JSP-1 and MKP-6 ( Fig. 2A), indicating that SKRP1 C149S is a catalytically inactive mutant of SKRP1. Similar phosphatase activities were obtained with His-tagged SKRP1 proteins (data not shown). Furthermore, these studies showed that SKRP1 was 10 -fold less active toward pNPP compared with JSP-1 and MKP-6 ( Fig. 2A). Three amino acids, Asp, Cys, and Ser, in the active site sequence motif (Fig. 1A, shaded box) have been shown to contribute to the catalytic mechanism of DSPs (45)(46)(47)(48)(49). Therefore, to examine whether the natural replacement S156A of SKRP1 protein is responsible for its low phosphatase activity, we compared the catalytic activity of SKRP1 with that of SKRP1 A156S using in vitro phosphatase assay. However, these studies showed that both proteins exhibited the similar enzymatic activities in vitro (Fig. 2B).
Since each DSP has its own substrate specificity for MAP kinases (MAPKs) (31-33), we tested the activity of SKRP1 toward different MAPKs in vitro. Dually phosphorylated HA-MAPKs were immunoprecipitated from the transfected cells that had been treated with 12-O-tetradecanoylphorbol-13-ace-tate (for ERK2 activation), or anisomycin (for JNK2 and p38␣ activation). Each sample was incubated with His-SKRP1 or His-SKRP1 C149S in vitro, and the phosphorylation state of each MAPK was examined by immunoblotting using the indicated anti-phospho-MAPK antibodies. The results showed that SKRP1 is highly specific for JNK2 in vitro, suggesting that SKRP1 is a novel member of MKP family (Fig. 2, C and D). Furthermore, we investigated whether SKRP1 is a DSP that dephosphorylates both phosphothreonine and phosphotyrosine residues. As shown in Fig. 2E, dually phosphorylated JNK was dephosphorylated on both residues by SKRP1. Thus, these findings suggested that SKRP1 is indeed a dual specificity MKP.
Expression and Tissue Distribution of SKRP1-A polyclonal antiserum to SKRP1 was produced by immunizing rabbits with a synthetic peptide encompassing the residues 80 -97 (Fig. 1C, double underline) of the mouse SKRP1 protein, which are not homologous to other DSPs. Immunoblotting analysis of NIH3T3 cells revealed that antiserum to SKRP1 recognizes a protein of 27 kDa co-migrating with an in vitro translated 35 S-labeled SKRP1. In contrast, preimmune serum from the same rabbit did not react with this protein (Fig. 3A). Furthermore, SKRP1 protein was detected by immunoblotting using antiserum to SKRP1 in various cell lines, containing BALB/ 3T3, NRK, PC12, and 293 (data not shown). We next examined the tissue distribution of mouse SKRP1 by Northern blot analysis. This study revealed two transcripts of 4.0 and 1.0 kb, which were expressed at moderate levels in brain, placenta, lung, and small intestine, and highly expressed in heart, liver, kidney and testis, respectively ( Fig. 3B and data not shown).
Inactivation of JNK Pathway by SKRP1 in Vivo-SKRP1 displayed a restricted substrate specificity in vitro, with preference for JNK2. Therefore, to investigate whether JNK can be a substrate for SKRP1 in mammalian cells, we examined the effect of SKRP1 on the JNK activation. In this experiment, GST-JNK2 was co-expressed in COS-7 cells with increasing amounts of Myc-SKRP1. Cells were then left untreated or treated with distinct stimuli, such as thapsigargin (50), anisomycin, or ultraviolet (UV) radiation, known to activate the JNK pathway, and the activities of GST-JNK2 at 1 h after treatment were examined by in vitro kinase assays as described previously (50). This study showed that SKRP1 expression effectively inhibits the basal and thapsigargin-induced activities of JNK in a dose-dependent manner (Fig. 4A). Furthermore, to analyze the kinetics of the JNK activation in cells expressing SKRP1, Myc-SKRP1 was co-expressed in 293 cells with GST-JNK2, and cells were treated with distinct stimuli for the indicated times. The activities of GST-JNK2 at different times after treatment were then examined by in vitro kinase assays as described above. The kinetics analysis revealed that cotransfection of SKRP1 had only a small influence on the activation of JNK induced by anisomycin and UV radiation, but strongly inhibited both TNF␣-and thapsigargin-induced JNK activation (Fig. 4, B and C). These data indicate that SKRP1 is involved in the regulation of the JNK pathway in response to specific extracellular stimuli.
Co-localization of SKRP1 and JNK-To examine the subcellular localization of exogenously expressed SKRP1, NIH3T3 cells were transiently transfected with Myc-SKRP1, followed by indirect immunofluorescence staining using anti-Myc antibody. The result showed that exogenously expressed SKRP1 was localized to the cytoplasm (Fig. 5A, right panel). We next investigated the localization of endogenous SKRP1 in NIH3T3 cells using preimmune serum or antiserum to SKRP1 described in Fig. 3A, and we found that endogenous SKRP1 was also detected predominantly in the cytoplasm (Fig. 5A, left and  middle panels). Furthermore, together with the result as shown in Fig. 3A, these findings revealed that SKRP1 was constitutively expressed.
As shown in Fig. 4, SKRP1 suppressed the activation of the JNK pathway. Therefore, to examine whether endogenous JNK can be a physiological target for endogenous SKRP1, NIH3T3 and BALB/3T3 cells were immunostained for their subcellular localizations using anti-SKRP1 and anti-JNK2 antibodies. These studies showed that endogenous SKRP1 co-localized with the endogenous JNK2 in both cells (Fig. 5B). To test whether SKRP1 interacts directly with JNK2, we performed in vitro binding assays using recombinant proteins, and we found that SKRP1 did not bind directly to JNK2 (Fig. 5C). Thus, SKRP1 co-localizes with but does not interact directly with JNK2. These findings raised the possibility that SKRP1 interacts indirectly with JNK2 through unidentified molecules in vivo.
Interaction of SKRP1 with the MAPKK MKK7-To identify molecules related to the JNK pathway that associate with SKRP1, metabolically labeled extracts were precipitated with GST or GST-SKRP1 fusion proteins, followed by SDS-PAGE analysis. This study showed that three major proteins of 46, 48, and 54 kDa were specifically precipitated by GST-SKRP1 and GST-SKRP1 C149S fusion proteins but not by GST protein alone (Fig. 6A).
We therefore searched for such candidates related to JNK, with the molecular mass of 46 -54 kDa. One candidate is the JNK activator MKK7 (40,(51)(52)(53)(54)(55)(56)(57)(58)(59), which is a MAPKK recognized as 47-48-kDa protein by immunoblotting in various mammalian cells containing NIH3T3 cells (52,53). We then examined the ability of SKRP1 to interact with MKK7. When His-MKK7 was incubated with GST, GST-SKRP1, or GST-SKRP1 C149S in vitro, His-MKK7 did not interact with GST untreated (UT) or treated with thapsigargin (1.0 M), anisomycin (10 g/ml), or UV radiation (80 J/m 2 ), and incubated for 1 h. GST-JNK2 was precipitated from the cell lysates, and the kinase activity was examined using GST-cJun as a substrate. The phosphorylated GST-cJun was detected after SDS-PAGE with a Fuji BAS2000 image analyzer. The amounts of GST-JNK2 and Myc-SKRP1 in the cell lysates were monitored by immunoblotting. B and C, kinetics of the JNK activation by distinct stimuli in SKRP1-expressing cells. 293 cells transfected with either control or Myc-SKRP1 together with the same amounts of GST-JNK2 were treated with TNF␣ (100 ng/ml), thapsigargin (1.0 m), anisomycin (10 g/ml), or UV radiation (80 J/m 2 ) and incubated for the indicated times. Glutathione-Sepharose precipitates were then processed for in vitro kinase assays as described in A. The phosphorylated GST-cJun was detected after SDS-PAGE with a Fuji BAS2000 image analyzer (B) and then quantitated (C). The amounts of GST-JNK2 and Myc-SKRP1 in the cell lysates were monitored by immunoblotting (B), and it was also verified that Myc-SKRP1 was comparable between co-expressions with GST-JNK2 in each experiment. Shown is a typical result of an experiment repeated three times. protein alone but did interact with GST-SKRP1 and GST-SKRP1 C149S (Fig. 6B).
Next, to determine whether SKRP1 specifically interacts with MKK7 in vivo, GST-SKRP1 was co-expressed in COS-7 cells together with either HA-MKK4, another MAPKK for JNK, or HA-MKK7, followed by co-precipitation analysis using glutathione-Sepharose as described previously (38 -44). This study showed that MKK7 was specifically detected in SKRP1 precipitates whereas MKK4 was not (Fig. 6C). These findings suggest that SKRP1 interacts selectively with the MAPKK MKK7 in vivo.
Binding of MKK7 to SKRP1 Independently of JNK2-Because MKK7 binds to JNK (40,54), the SKRP1-MKK7 direct interaction (Fig. 6B) raised the possibility that SKRP1 interacts indirectly with JNK through MKK7. We therefore tested the effect of SKRP1 on the physical interaction between MKK7 and JNK2. When GST-MKK7 was used as affinity reagents, JNK2 and SKRP1 mixed together bound to GST-MKK7 with more efficiency than that of each individual protein alone (Fig.  7A). Consistent results were obtained from co-transfection experiments, in which cells were transfected with MKK7, JNK2, and increasing amounts of SKRP1. Expression of SKRP1 did not interfere with the co-precipitation of MKK7 with JNK2 (Fig. 7B). Thus, JNK2 and SKRP1 did not compete for binding to MKK7. To confirm this conclusion further, we mapped the domains in MKK7, which are necessary for SKRP1 binding. Analysis of various MKK7 derivatives demonstrated that MKK7 residues 128 -214 were required for interaction with SKRP1 (Fig. 7C). Previous study (40) showed that three isoforms (␣, ␤, and ␥) of MKK7 with different N termini are created by alternative splicing, and its substrate JNK binds to the N-terminal extension that is present in MKK7␤ but not MKK7␣ isoforms. This region is also present in the residues 1-89 of MKK7␥ isoform used in our study and therefore can account for the interaction with JNK (Fig. 7, A and B). Furthermore, these findings indicate that SKRP1 and JNK2 can interact independently with MKK7 to form ternary complexes. DISCUSSION To date many MKPs have been identified, and their precise regulation is considered to confer the specificity of MAPK sig-naling pathways. They share the extended active site sequence motif (DX 28 VHCXAGXSRS, where X is any amino acid), and previous studies (45)(46)(47)(48)(49) showed that three amino acids, Asp, Cys, and Ser, in this motif contribute to the catalytic mechanism of them. The predicted amino acid sequence of mouse and human SKRP1 revealed that one of their critical residues, Ser, is naturally replaced by Ala at amino acid positions 156 and 157, respectively (Fig. 1B). Therefore, to test whether this replacement affects the SKRP1 catalytic activity, we performed in vitro phosphatase assay of GST-SKRP1 and GST-SKRP1 A156S using pNPP as a substrate. These studies showed that both proteins exhibited the similar enzymatic activities in vitro (Fig. 2B). In vitro studies as described previously (45) showed that a substitution of Ala for Ser conserved in the active site sequence motif had essentially no effect on substrate binding, intermediate formation, but dramatically affected the rate of intermediate hydrolysis. Therefore, these findings suggest that the rate-limiting step in the catalytic mechanism is completely converted from intermediate formation in the native enzyme to intermediate hydrolysis in the SA mutant. However, the physiological significance of Ala-156 in SKRP1 protein remains to be known. We next compared the in vitro enzymatic activity of recombinant SKRP1 with two other DSP members, JSP-1 and MKP-6, which have been isolated in the cloning process. These studies showed that SKRP1 is 10-fold less active toward pNPP compared with JSP-1 and MKP-6 ( Fig. 2A). In addition, unlike almost other MKPs, SKRP1 is constitutively expressed in cells (Figs. 3A and 5A), and its inducibility in response to various stresses was not detected (data not shown). Together, these findings suggest that as yet unknown mechanism may exist to regulate the phosphatase catalytic activity of SKRP1 in vivo.
Some members of MKP family display the specificity for inactivating MAPKs. It has been reported that the N-terminal CH2 domain of each MKP mediates the interaction with its preferred MAPK (31), which then induces the catalytic activation of MKP (23,27,60,61) and subsequent dephosphorylation of its bound MAPK. Furthermore, MAPK-docking site of MKP has been mapped to the consensus sequence motif containing more than two consecutive positively charged amino acids (62). However, SKRP1 possesses neither the N-terminal CH2 domain nor a sequence motif homologous to the consensus MAPK-docking sequence, and concomitantly we could not find any significant interaction between SKRP1 and MAPKs (data not shown).
In the present study, we demonstrated that SKRP1 interacts directly with the MAPKK MKK7 (Fig. 6B), but not directly with JNK2 (Fig. 5C)  It was also verified that GST (Control) was precipitated using glutathione-Sepharose beads (data not shown). Similar results were obtained in three different experiments. C, analysis of SKRP1 binding domains in MKK7. Schematic representation of various MKK7 derivatives used in this study is shown in left panel. In vitro binding assays were performed using either GST (Ϫ) or GST-SKRP1 (ϩ), and cell lysates from COS-7 cells transfected with control or the indicated HA-tagged MKK7 derivatives (a-h) (right panel). Glutathione-Sepharose precipitates of SKRP1 were immunoblotted for the presence of various HA-MKK7 derivatives using anti-HA antibody (top), followed by Coomassie staining (middle panel). The amounts of various HA-MKK7 derivatives in the cell lysates were examined by immunoblotting using anti-HA antibody (bottom). Similar results were obtained in two different experiments. which ␤ and ␥ isoforms possess N-terminal JNK-interacting domain at amino acid residues 1-73 and 1-89, respectively (40,63). As shown in Fig. 7, A and B, we actually observed that JNK2 interacted with the MKK7␥2 isoform that we have employed in this study and that JNK2 binding on MKK7 was not competed by SKRP1 (Fig. 7B). In line with this result, we have also determined SKRP1-binding domain on MKK7 as shown in Fig. 7C, which differs from JNK-binding domain on MKK7. These results indicated that the selective interaction between SKRP1 and MKK7 within the cells contributes to the specificity in JNK dephosphorylation by SKRP1. However, 10 JNK isoforms have been isolated so far (64); therefore, SKRP1 may have differential dephosphorylation efficiency on these JNK isoforms, according to the affinity of MKK7-JNK interaction that varies among the combinations of use of MKK7 and JNK isoforms. Furthermore, SKRP1 strongly inhibited the activation of JNK2 induced by TNF␣ or thapsigargin, whereas SKRP1 had only a small influence on the JNK2 activation induced by anisomycin or UV radiation (Fig. 4). These findings are, to our knowledge, the first evidence to show that MKP interacts directly with a MAPKK and preferentially regulates the particular MAPK pathway.
As shown in Fig. 6A, in vitro binding analysis using both metabolically labeled extracts and GST-SKRP1 fusion proteins as affinity reagents revealed that only three major proteins were detectable in GST-SKRP1 precipitates. However, there is a possibility that SKRP1 can associate with proteins other than them, which may contribute to the regulation of its subcellular localization and function. Interestingly, the cytosolic SKRP1 has been found to translocate to the nucleus in response to osmotic stress, a potent activating stress for the p38 MAPK pathway (data not shown), and we also found that SKRP1 suppresses the basal and stress-induced activities of p38 MAPK in vivo (data not shown). Thus, different compartments of SKRP1 might exist in the cell, which would involve different signaling pathways. Therefore, SKRP1 may function as a target site for cross-talk with other signaling pathway.
Finally, although MKP has been implicated in the negative regulation of MAPK pathways, the general importance in terminating MAPK, the molecular mechanisms that may control its phosphatase catalytic activity, and its signaling specificity for inactivating the particular MAPK pathway signaling remain to be determined. Here our results may provide a hint on these issues and illustrate a new tier of the precise regulation of MAPK signaling pathways.