Molecular Cloning and Characterization of a Novel p38 Mitogen-activated Protein Kinase*

The p38 mitogen-activated protein kinases (MAPK) are activated by cellular stresses and play an important role in regulating gene expression. We have isolated a cDNA encoding a novel protein kinase that has significant homology (57% amino acid identity) to human p38 (cid:97) / CSBP. The novel kinase, p38 (cid:100) , has a nucleotide sequence encoding a protein of 365 amino acids with a putative TGY dual phosphorylation motif. Dot-blot analysis of p38 (cid:100) mRNA in 50 human tissues revealed a distribution profile of p38 (cid:100) that differs from p38 (cid:97) . p38 (cid:100) is highly expressed in salivary gland, pituitary gland, and adre- nal gland, whereas p38 (cid:97) is highly expressed in placenta, cerebellum, bone marrow, thyroid gland, peripheral leu- kocytes, liver, and spleen. Like p38 (cid:97) , p38 (cid:100) is activated by cellular stress and proinflammatory cytokines. p38 (cid:100) phosphorylates ATF-2 and PHAS-I, but not MAPK-acti- vated protein kinase-2 and -3, known in vivo and in vitro substrates of p38 (cid:97) . We also observed that p38 (cid:100) was strongly activated by MKK3 and MKK6, while p38 (cid:97) was preferentially activated by MKK6. Other experiments showed that a potent p38 (cid:97) the pCR2.1 vector (Invitrogen, Carlsbad, CA). To generate constitutively active MKK3 (ca-MKK3), site-directed mutagenesis (43) was used to replace Ser-189 and Ser-193 with Glu. This DNA was cloned into the baculovirus transfer vector pVL1392 (Invitrogen) and expressed in Hi-5 cells. Recombinant ca-MKK3 was purified by hy- droxyapatite (Bio-Rad) followed by Phenyl-Sepharose HP (Pharmacia) chromatography. To generate human FLAG-tagged p38 (cid:97) , two primers from the published nucleotide sequences (14) were used in PCR with human peripheral blood leukocyte cDNA as templates. The PCR prod- uct was then cloned into mammalian expression vector pCMVXV (cid:98) 5. HA-tagged MKK6 in pME vector (31) was kindly provided by Dr. Hagi- wara Masatoshi and HA-tagged MKK3 and MKK4 (SEK1) in mammalian expression vector pMT (44) were provided by Dr. James Woodgett. Molecular Cloning of p38 (cid:100) — An expressed sequence tag (EST) (311 base pairs) with homology to p38 was identified in the Amgen EST data base. Gene-specific forward and reverse primers were designed from the EST sequence and used in PCR to clone full-length cDNA with the Marathon-Ready human fetal brain cDNA templates (CLONTECH) following the manufacturer’s protocol. These Marathon-Ready cDNAs have adaptors ligated at the 5 (cid:57) and 3 (cid:57) ends. The gene-specific forward primer (5-GAGCTGTCCAAGACCTACGTGTC-3 (cid:57) ) and an adaptor primer (CLONTECH) were used in combination to amplify the 3 (cid:57) portion of p38 (cid:100) . The gene-specific reverse primer (5 (cid:57) -CTGGGGTGAAGA- CATCCAGG-3 (cid:57) calyculin A, 0.1% (cid:98) -mercaptoethanol, 100 (cid:109) M ATP, 75 (cid:109) g/ml ca-MKK3, and approximately 20 (cid:109) g/ml p38 (cid:100) or p38 (cid:97) . The reaction (60 min at 28 °C) was terminated by washing the p38-bound immunoprecipitates three times with Dulbecco’s phosphate-buffered saline. Substrate specificity studies were performed in kinase buffer (40- (cid:109) l reactions; 30 min at 28 °C) with 2 (cid:109) M cold ATP, 5 (cid:109) Ci of [ (cid:103) - 32 P]ATP (3000 Ci/mmol), no ca-MKK3, approximately 1 (cid:109) g/ml p38 (cid:100) or p38 (cid:97) , and 50 (cid:109) g/ml of either full-length GST-ATF-2, GST-c-Jun, GST-MAPKAP kinase-2, GST-MAPKAP kinase-3, or PHAS-I. To evaluate the sensitivities of p38 (cid:97) and p38 (cid:100) to AMG 2372, 40- (cid:109) l kinase reactions (15 min at 28 °C) were performed as described for substrate reactions, except that ATF-2- (1–109) was used as substrate in the presence of various inhibitor concentrations. Reactions were terminated by boiling in the presence of 2 (cid:51) sample buffer. The reaction products were resolved by SDS-PAGE, visualized by autoradiography, and quantified using a PhosphorImager (Molecular Dynamics). Distribution expression was examined in a variety of human tissues by Northern blot analysis using a probe derived from the 5 The p38 (cid:100) probe hybridized strongly to a transcript of approximately 1.8 kilobases and weakly to a transcript of 6.0 kilobases, while a probe derived from p38 (cid:97) hybridized to a single transcript of 4.1 The same p38 (cid:100) probe was used to hybridize a human RNA master blot containing poly (cid:49) RNAs from 50 different human tissues. The RNAs in this blot have been normalized to the mRNA levels of eight different housekeeping genes; thus the relative levels of mRNA in different tissues could be determined. Among the tissues examined, strong hybridizing signals were observed in exocrine/ endocrine tissues including human salivary gland, pituitary gland, adrenal gland, and placenta (Fig. 2 A Moderate signals were observed in pancreas, trachea, thyroid gland, stomach, prostate, colon, small intestine, lymph node, kidney, and lung.

p38 was originally identified in lipopolysaccharide (LPS)stimulated mouse macrophages and was found to have substantial homology to the Saccharomyces cerevisiae HOG1 kinase (13,38). The human homologues of p38 were cloned after p38 was identified with a radiophotoaffinity-labeled pyridinyl imidazole compound (14). Inhibition of p38 by this class of compound prevents the production of IL-1 and TNF␣ by human monocytes stimulated with LPS (14). In addition to the original isoform of p38 (now referred to as p38␣), a second p38 kinase member (p38␤) was identified which shows 74% amino acid identity to p38␣ (39). p38␤ also has a TGY motif in kinase subdomain VIII (39). More recently, a third p38 kinase family member with a TGY motif was cloned and is termed p38␥/ ERK6/SAPK3 (40 -42). The amino acid sequence of p38␥/ ERK6/SAPK3 is 60% identical to p38␣ (40).
Here we report the isolation of a novel p38 MAPK (p38␦) with a TGY motif in its activation domain. p38␦ was characterized with regard to tissue distribution, stimulus activation, MKK activation, substrate specificity, and inhibitor sensitivity. These studies reveal interesting similarities as well as differences in the properties of p38␦ as compared with p38␣.

EXPERIMENTAL PROCEDURES
Reagents-Recombinant GST-c-Jun protein was purchased from Upstate Biotechnology Inc. (Lake Placid, NY). Recombinant PHAS-I protein was purchased from Stratagene (La Jolla, CA). ATF-2 was amplified by PCR from human skeletal muscle cDNA using two primers * 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) AF015256.
Molecular Cloning of p38␦-An expressed sequence tag (EST) (311 base pairs) with homology to p38 was identified in the Amgen EST data base. Gene-specific forward and reverse primers were designed from the EST sequence and used in PCR to clone full-length cDNA with the Marathon-Ready human fetal brain cDNA templates (CLONTECH) following the manufacturer's protocol. These Marathon-Ready cDNAs have adaptors ligated at the 5Ј and 3Ј ends. The gene-specific forward primer (5-GAGCTGTCCAAGACCTACGTGTC-3Ј) and an adaptor primer (CLONTECH) were used in combination to amplify the 3Ј portion of p38␦. The gene-specific reverse primer (5Ј-CTGGGGTGAAGA-CATCCAGG-3Ј) and the adaptor primer were used to amplify the 5Ј portion of p38␦. PCR was performed for 30 cycles (95°C for 30 s, 42°C for 30 s, and 72°C for 20 s) followed by an extension at 72°C for 7 min. The resulting PCR product was ligated into the pCR2.1 vector (Invitrogen) and sequenced on both strands. A second murine EST sequence (GenBank TM accession number W53837) that has homology to the Amgen EST sequence was identified in the GenBank data base. This EST fragment was used as a probe to screen a human macrophage library, and two clones were isolated. Sequencing of one of the clones revealed an identical open reading frame as the one cloned by PCR. The clone isolated from human fetal brain library was used for subsequent studies described here.
Full-length p38␦ cDNA was cloned into a mammalian expression vector PCR3.1 (Invitrogen) by PCR using two primers (5Ј-ACCAT-GGACTACAAGGACGACGATGACAAGAGCCTCATCCGGAAAAAGG-GCTTCTACAAG-3Ј and 5Ј-ACCTGCAGGCGATTCTCCAGAT-3Ј). The first primer added a FLAG epitope at the 5Ј end. PCR site-directed mutagenesis (43) was used to create a p38␦ mutant (AGF) by substituting Thr-180 and Tyr-182 with an Ala and a Phe, respectively. The inserts were completely sequenced to make sure that no PCR errors were introduced.
Northern and Dot-Blot Analysis of p38␦ mRNA-A Northern blot filter containing poly(A) ϩ RNA from multiple tissues and a normalized Master blot filter containing mRNA from 50 different tissues (CLON-TECH) were probed with a 32 P-labeled DNA fragment generated from the 5Ј portion of the coding region of p38␦ (nucleotides 1 to 550). Hybridization was performed at 68°C in ExpressHyb Buffer (CLON-TECH) followed by two washes in 0.1 ϫ SSC, 0.1% SDS at 55°C. Blots were exposed overnight at Ϫ70°C. The same Northern blot was then probed with a 32 P-labeled DNA fragment generated from the 5Ј portion of the coding region of p38␣ using identical hybridization and washing conditions.
Immunoprecipitation and Western Blot Analysis-Immunoprecipitation was performed as described previously (45). Briefly, cells were dislodged into lysis buffer (20 mM Tris-HCl, pH 7.5, 1% Triton X-100, 0.5% Igepal, 150 mM NaCl, 20 mM NaF, 0.2 mM Na 3 VO 4 , 1 mM EDTA, 1 mM EGTA) and sedimented (15,000 ϫ g for 60 min) to remove insoluble debris. Total protein in cell lysates was quantified by the Bradford method using a protein assay kit (Pierce). Supernatants containing 100 g of protein were immunoprecipitated with 5 g of anti-HA mAb 12CA5 (Berkeley Antibody Co., Berkeley, CA) or anti-FLAG M2 mAb (Sigma) and protein A-Sepharose CL-4B beads (Pharmacia). For Western blot analysis, lysates containing equal amounts of total protein were resolved by 4 -20% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and electroblotted onto nitrocellulose membranes. The blots were then probed with mAb M2, followed by biotinylated rabbit antimouse IgG (Amersham Life Science Inc.) and developed using the enhanced chemiluminescence (ECL) detection system (Amersham Life Science Inc.).
p38 Kinase Assay-Cells transfected with FLAG-tagged p38␦ or p38␣ were lysed, and recombinant protein was immunoprecipitated using mAb M2 and protein A-Sepharose CL-4B beads. Beads were washed three times with lysis buffer, once with kinase buffer (25 mM HEPES, pH 7.4, 25 mM ␤-glycerophosphate, 25 mM MgCl 2 , 25 mM dithiothreitol, 0.1 mM Na 3 VO 4 ), and resuspended in 40 l of kinase buffer. The beads were then incubated with ATF-2 and 1 l of [␥-32 P]ATP (3000 Ci/mmol) at 30°C for 30 min. Reaction mixtures were then resuspended in 2 ϫ sample buffer (125 mM Tris, pH 6.8, 6% SDS, 20% glycerol) and boiled for 3 min. Phosphorylated proteins were resolved by SDS-PAGE, after which the gels were dried and exposed to radiographic film.

Molecular
Cloning of p38␦-To identify novel MAPKs, we searched the Amgen EST data base with the p38 nucleotide sequences as the query sequences. One partial human cDNA sequence (311 base pairs) was identified that has homology to p38␣. Through PCR and hybridization techniques, we isolated a full-length cDNA corresponding to this EST sequence. The novel cDNA is predicted to encode a protein of 365 amino acids with a molecular mass of 42 kDa (Fig. 1A). The deduced amino acid sequence predicts a protein kinase with all 11 characteristic kinase subdomains (Fig. 1B). GenBank TM and European Molecular Biology Laboratory Data base searches identified p38␣/CSBP2 (14), p38␤ (39), and p38␥/ERK6 (40 -42) as the most closely related molecules. p38␦ displays 57, 55, and 62% amino acid identity to p38␣/CSBP2, p38␤, and p38␥/ERK6, respectively. p38␦ has a putative dual phosphorylation motif (TGY) in kinase subdomain VIII similar to that found in other p38 family members. The amino acid sequence alignment of p38␦ with human p38␣/CSBP2, p38␤, and p38␥/ERK6 is shown in Fig. 1B.
Tissue Distribution of p38␦ mRNA-The expression of p38␦ was examined in a variety of human tissues by Northern blot analysis using a probe derived from the 5Ј end of p38␦. The p38␦ probe hybridized strongly to a transcript of approximately 1.8 kilobases and weakly to a transcript of 6.0 kilobases, while a probe derived from p38␣ hybridized to a single transcript of 4.1 kilobases (14) (data not shown). The same p38␦ probe was used to hybridize a human RNA master blot containing poly (A) ϩ RNAs from 50 different human tissues. The RNAs in this blot have been normalized to the mRNA levels of eight different housekeeping genes; thus the relative levels of mRNA in different tissues could be determined. Among the tissues examined, strong hybridizing signals were observed in exocrine/ endocrine tissues including human salivary gland, pituitary gland, adrenal gland, and placenta ( Fig. 2A). Moderate signals were observed in pancreas, trachea, thyroid gland, stomach, prostate, colon, small intestine, lymph node, kidney, and lung. Probing the master blot with p38␣ DNA revealed a different tissue distribution profile. Strong hybridizing signals were found in placenta, cerebellum, bone marrow, thyroid gland, peripheral leukocyte, liver, and spleen. Moderate signals were found in occipital lobe, fetal liver, pituitary gland, adrenal gland, aorta, uterus, stomach, lymph node, cerebral cortex, hippocampus, and thymus (Fig. 2B). Probing the master blot with p38␤ DNA found that p38␤ is abundantly expressed in brain tissues such as hippocampus, frontal lobe, cerebral cor-tex, cerebellum, caudate nucleus, medulla oblongata, whole brain, and fetal brain (Fig. 2C). Interestingly, probing the master blot with p38␥ DNA found that it has a very limited tissue distribution profile. p38␥ was highly expressed in skeletal muscle, while the expression in other tissues appears to be low (Fig. 2D).
Substrate Specificity of p38␦-Full-length p38␦ cDNA and p38␣ were cloned into mammalian expression vectors with a FLAG epitope sequence added at the 5Ј end and transfected into 293 cells. Transfected cell lysates were subjected to immunoprecipitation with a FLAG mAb. The immunoprecipitated p38␦ and p38␣ were activated using recombinant ca-MKK3, washed, and used in immune complex kinase assays with various substrates. As shown in Fig. 3, p38␦ and p38␣ phosphorylated full-length ATF-2 (lanes 1 and 6) and PHAS-I (lanes 5 and 10), but not c-Jun (lanes 2 and 7), a known substrate for JNK (6). We also observed that p38␦ showed minimal phosphorylating activity against MAPKAP kinase-2 and -3 (Fig. 3,  lanes 3 and 4), while p38␣ phosphorylated these substrates efficiently (Fig. 3, lanes 8 and 9). Control lysates did not phosphorylate any of the substrates (data not shown).
Activation of p38␦ by Extracellular Stimuli-The p38 group of kinases can be activated by a variety of stress stimuli and proinflammatory cytokines (13,14,18). Because p38␦ is closely related to p38␣, we determined whether similar stimuli could activate p38␦ kinase activity. 293 cells were transiently trans- fected with either p38␦ or p38␣ cDNA and treated with various stimuli. p38␦ and p38␣ activity was measured by their ability to phosphorylate ATF-2-(1-109) in an immune complex assay. As shown in Fig. 4A, p38␦ was strongly activated by H 2 O 2 , UV, NaCl, and Na 3 VO 4 and moderately activated by anisomycin, IL-1␤, TNF␣, and epidermal growth factor. p38␣ was strongly activated by UV, NaCl, H 2 O 2 , and anisomycin and moderately activated by TNF␣, IL-1␤, and epidermal growth factor (C). A notable difference is that Na 3 VO 4 strongly activated p38␦ (Fig. 4A, lane 5) but not p38␣ (Fig. 4C, lane 5). To eliminate the possibility that changes in p38 kinase activity are due to the variations in protein expression, Western blot analysis was performed. Fig. 4, B and D, shows that similar amounts of p38 were expressed under all conditions tested.
p38␦ Is Activated by Phosphorylation at the Dual Phosphorylation TGY Motif-MAPKs are activated by dual phosphorylation at the Thr-Xaa-Tyr motif within kinase subdomain VIII (18). To determine whether this motif is required for p38␦   FIG. 2. Expression pattern of human p38␦ and p38␣ mRNA. Filters containing poly(A) ϩ RNA from the indicated tissues were hybridized with radioactive p38␦ (A), p38␣ (B), p38␤ (C), or p38␥ (D) probes as described under "Experimental Procedures." Autoradiographs were scanned using densitometry, and individual spots were quantitated using ImageQuant software (Molecular Dynamics). The numbers on the y axis denote arbitrary units. activation, we generated a mutant p38␦ by replacing the Thr-Gly-Tyr motif with Ala-Gly-Phe (AGF mutant) and tested whether this mutant could be activated. Wild type p38␦ phosphorylated ATF-2 when activated by UV irradiation (Fig. 6A,  lane 4) or by co-transfection with MKK6 (Fig. 6A, lane 7). However, the AGF mutant was unresponsive to UV stimulation (Fig. 6A, lane 6) or activation by upstream kinase MKK6 (Fig. 6A, lane 8). Western blot analysis demonstrated that the AGF mutant was expressed to comparable levels as wild type p38␦ (Fig. 6B).
Effect of AMG 2372 on p38␦ and p38␣ Kinase Activity in Vitro-p38␦ and p38␣ from transfected cell lysates were immunoprecipitated with mAb M2 and activated in vitro using purified ca-MKK3. Excess ca-MKK3 was used to ensure that p38␦ and p38␣ were maximally activated. Kinase assays were then performed in the presence of the indicated concentrations of AMG 2372. AMG 2372 only weakly inhibited the kinase activ-ity of p38␦ (Fig. 7A), whereas p38␣ kinase activity was inhibited in a dose-dependent manner (Fig. 7B). At 1 M inhibitor concentration, there was 98% inhibition of p38␣ (Fig. 7B, lane  3), but less than 25% inhibition of p38␦ (Fig. 7A, lane 3). Similar results were obtained in three separate experiments. DISCUSSION In this report, we describe the cloning and characterization of a novel member of the p38 group of protein kinases. p38␦ has significant homology at the amino acid level to p38␣, -␤, and -␥ and contains the dual phosphorylation TGY motif that is found in this p38 group of kinases (13,18). Mutation of the Thr and 293 cells were transfected with p38␦ (A) or p38␣ (C) and stimulated for 30 min with 0.5 M NaCl, 500 M H 2 O 2 , 1 mM Na 3 VO 4 , for 20 min with 50 ng/ml anisomycin, 100 ng/ml IL-1␤, 50 ng/ml TNF␣, and for 10 min with 20 ng/ml epidermal growth factor, or cells were irradiated with 857 J/cm 2 UV light and then lysed 30 min later. p38␦ and p38␣ were then immunoprecipitated, and their kinase activities were assayed using ATF-2-(1-109) as substrate. B and D, Western blot analysis of p38␦ (B) and p38␣ (D). Lysates containing equal amounts of total protein were resolved by 4 -20% SDS-PAGE and transferred to nitrocellulose membranes. The blots were probed with mAb M2 followed by rabbit antimouse IgG. The blots were developed using the ECL system. Lysates containing equal amounts of total protein were immunoprecipitated with mAb M2 followed by kinase assay using ATF-2 as substrate. B and D, lysates containing equal amounts of total protein from transfected 293 cells were resolved by 4 -20% SDS-PAGE and transferred to nitrocellulose membranes. The blots were probed with mAb M2 followed by rabbit anti-mouse IgG. The blots were developed using the ECL system.  (lanes 1 and 2), p38␦ wild type (lanes 3 and 4), p38␦ AGF mutant (lanes 5 and 6), p38␦ wild type plus MKK6 (lane 7), p38␦ AGF mutant plus MKK6 (lane 8), or MKK6 alone (lane 9). Transfected cells were either not exposed (lanes 1, 3, and 5) or exposed to UV irradiation (lanes 2, 4, and 6). p38␦ activity was determined as described under "Experimental Procedures." B, lysates containing equal amounts of total protein were resolved by 4 -20% SDS-PAGE and transferred to nitrocellulose membranes. Blots were probed with mAb M2 followed by rabbit anti-mouse IgG. The blots were developed using the ECL system.
Tyr residues in the TGY motif abolished the kinase activity of p38␦ and blocked UV or MKK6-induced activation. Thus, like other MAPKs, p38␦ requires phosphorylation at the Thr and/or Tyr in the TGY motif for its activation.
The tissue distribution of p38␦ was examined in 50 different human tissues. The pattern of expression of p38␦ mRNA is distinct from that of p38␣, -␤, and -␥. Very high levels of expression of p38␦ mRNA were observed in human gland tissues, while p38␣ was abundantly expressed in placenta, brain (cerebellum), and lymphoid tissues. p38␤ is most abundantly expressed in brain tissues, while p38␥ appears to have a limited tissue distribution. These differences in mRNA expression suggest that p38␣, p38␤, p38␥, and p38␦ may have tissuespecific functions.
Similar to p38␣, p38␦ is activated in 293 cells by a diverse array of cellular stresses and proinflammatory cytokines. However, the degree of activation by various stimuli is different for p38␦ as compared with p38␣. Most notable was the strong activation of p38␦, but not of p38␣ by Na 3 VO 4 . Because Na 3 VO 4 inhibits protein tyrosine phosphatase activity, our data suggest that such phosphatases differentially regulate the basal activity of p38␦ and p38␣.
Differences in activation of p38␦ versus p38␣ were also observed at the MKK level. In cell transfection experiments, p38␦ is strongly activated by MKK3 and MKK6, whereas p38␣ is preferably activated by MKK6. These data suggest that regulators of p38␦ overlap. Like p38␣, it is likely that the dominant activator of p38␦ in a given cell type will reflect the unique cellular environment. For example, it has been observed that MKK6 is the dominant activator of p38␣ in monocytes and KB cells, while MKK3 is the dominant activator of p38␣ in PC-12 cells (46). p38␣/CSBP has been directly linked to inflammatory cytokine production through the use of inhibitors that block its function. We tested one compound that blocks p38␣ activity and found that it was relatively inactive against p38␦. Thus, other compounds will have to be developed to determine if p38␦ is involved in cytokine production. The critical substrates phosphorylated by p38␣ leading to cytokine production have not yet been elucidated, although several candidates have been discovered. Of this group, we showed that p38␦ phosphorylated ATF-2 and PHAS-I, but not MAPKAP kinase-2 and -3. Additional studies are required to identify in vivo p38␦ substrates and to determine if these substrates are involved in cytokine production or other p38␦-mediated processes.