Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Components of a New Human Protein Kinase Signal Transduction Pathway

Components of a New Human Protein Kinase Signal Transduction Pathway Vol. 270, No. 21, Issue of May 26, pp. 12665-12669, 1995 THE JOURNAL OF BIOLOGICAL CHEMISTRY Printed in U.S.A. © 1995 by The American Society for Biochemistry and Molecular Biology, Inc. Components of a New Human Protein Kinase Signal Transduction Pathway* (Received for publication, March 27, 1995) Gaochao Zhou, Zhao Qin Bao, and Jack E. Dixont From the Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, Michigan 48109-0606 motif, where "X" can be Glu, Pro, or Gly (1). There are also a We have identified two components of a new protein growing number of MEKs (15, 16). In several cases, specific kinase signaling cascade, MAPKIERK kinase 5 (MEK5) and extracellular signal-regulated kinase 5 (ERK5). The MEKs have been shown to phosphorylate specific ERKs in a MEK5 cDNA was isolated by degenerate PCR and en­ given pathway. MEKI and MEK2 phosphorylate ERKI and codes a 444-amino acid protein, which has approxi­ ERK2 in a mitogenesis or differentiation response; MEK3 and mately 40% identity to known MEKs. ERK5 was identi­ p38 show coordinate activation in the cytokine response, as do fied by a specific interaction with the MEK5 mutants MEK4 (i.e. SEKl, the murine homolog ofMEK4) and the JNKs S311Afr315A and K195M in the yeast two-hybrid system. in the stress response. Sequential protein kinase cascades have The proteins were found to interact in an in vitro bind­ been studied extensively in yeast, where genetic tools have ing assay as well. ERK5 did not interact with MEKI or been invaluable in assigning kinase cascades to specific signal MEK2. ERK5 is predicted to contain 815 amino acids and transduction pathways (12-14). In higher eukaryotic systems, is approximately twice the size ofall known ERKs. The C less is known about the specific stimuli responsible for activa­ terminus of ERK5 has sequences which suggest that it tion of individual signal transduction pathways. Although may be targeted to the cytoskeleton. Sequences located there are clearly multiple genes encoding ERKs and MEKs, we in the N terminus ofMEK5 may be important in coupling have no good idea about how many pathways there are in GTPase signaling molecules to the MEK5 protein kinase eukaryotic cells. cascade. Both MEK5 and ERK5 are expressed in many We have identified a new MEK, which we will refer to as adult tissue and are abundant in heart and skeletal mus­ MEK5. When MEK5 was employed in the yeast two-hybrid cle. A recombinant GST-ERK5 kinase domain displays system, we identified a new member of the ERK family of autophosphorylation on Ser!l'hr and Tyr residues. protein kinases, which we have named ERK5. MEK5 and ERK5 interacted specifically with one another and did not interact with kinases in the MEKI/ERKI signaling pathway, Some cell surface receptors respond to extracellular stimuli suggesting that the MEK5/ERK5 protein cascade represents a via a sequential protein kinase cascade (1, 2). The kinase sig­ novel signaling pathway in higher eukaryotes. MEK5 contains naling cascade involves multiple protein kinases, two of which unique amino acid sequences in its N terminus, which suggest are extracellular signal-regulated kinase (ERK\ Refs. 3 and 4)2 that this molecule may interact with GTPases such as CDC42. and MAPKIERK kinase (MEK; Refs. 5 and 6).3 Several ERK In addition, the C terminus of ERK5 also contains sequences related proteins have been identified in vertebrates including which may target this kinase to the cytoskeletal elements c-Jun N-terminal kinase (JNK; Refs. 7 and 8)4 and p38 (9-11). within the cell. A variety of similar proteins have also been identified in yeast (12-14). All of the ERK kinases are activated by phosphoryla­ MATERIALS AND METHODS tion on Tyr and Thr residues within a TXY phosphorylation peR and cDNA Screening-Two degenerate PCR primers were syn­ thesized based on the conserved amino acid sequences present in MEKs from mammals and yeast (6). The sense primer was 5'-CTTGGATC­ * This work was supported by National Institutes of Health NIDDKD QTA(CIT)ATA(CIT)GTNGGNTT(CIT)TA-3' and the antisense primer Grant 18024 and a postdoctoral fellowship from the American Diabetes was 5'-CTTGGATCC(G/T)TCNGGN(C/G)(AIT)CAT(NG)TA-3', which Association (to G. Z,). The costs of publication of this article were corresponds to amino acid sequences YIVGFY and YMSPER, respec­ defrayed in part by the payment of page charges. This article must tively. The underlined BamHI sites were used for cloning the PCR therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. product. Plasmid DNA from a AZAP human fetal brain cDNA library was used as template. The PCR was performed for 35 cycles at 94 DC for The nucleotide sequencers) reported in this paper has been submitted to the GenBanFM I EMBL Data Bank with accession numberts) U25265 1 min, 50 DC for 2 min, and 72 DC for 2 min. The resultant 0.35-kb PCR and U25278. products were resolved by 3% Nusieve-agarose gel electrophoresis and :j:To whom correspondence should be addressed: Dept. of Biological subcloned into a M13 vector and sequenced. Chemistry, University of Michigan Medical School, Rm. 5416, Medical A human fetal brain cDNA library was screened (Stratagene), using Science I, Ann Arbor, MI 48109-0606. Tel.: 313-764-8192; Fax: 313-763­ the cloned MEK5 PCR fragment labeled with a multiprimer DNA labeling kit. Hybridization was carried out at 42 DC in 6 X SSC (1 X 1 The abbreviations used are: ERK, extracellular-regulated kinase; SSC is 0.15 M NaCI, 0.015 M sodium citrate, pH 7.0), 1 X Denhardt's MAP, mitogen-activated protein; MAPK, MAP kinase; MAPKK or solution (1 X Denhardt's solution is 0.02% Ficoll400, 0.02% bovine serum MKK, MAP kinase kinase; MEK, MAPKIERK kinase; JNK, c-Jun N­ albumin, 0.02% polyvinylpyrrolidone), 0.1 mg/ml denatured salmon terminal kinase; kb, kilobase pair'(s); PCR, polymerase chain reaction; sperm DNA, 50% formamide. Both strands ofthe DNA corresponding to PAGE, polyacrylamide gel electrophoresis; GST, glutathione S-trans­ the clone designated pBluescriptSK~ MEK5 were sequenced. ferase. The Yeast Two-hybrid Screen-The yeast strain Y190 (MATa, 2 ERK is also referred to as mitogen-activated protein kinase (MAPK) leu2-3, 112, ura3-52 trpl-901 his3-Ll.200 ade2-101 gal4 tiga180Ll. URA in the literature. We will employ the ERK nomenclature. 3 GAL-lacZ LYS GAL-HIS3 cyh'') was a generous gift of Stephen 3 MEK is also referred to as mitogen-activated protein kinase kinase (MAPKK or MKK) in the literature. We will employ the MEK Elledge, Departments of Biochemistry and Human and Molecular Genetics, Baylor College of Medicine. The NcoI fragment encoding amino nomenclature. 4 JNK is also referred to as stress-activated protein kinase (SAPK) in acid 12 through the stop codon of MEK5 were subcloned into pAS1­ the literature. CYH2 vector. Y190 was transformed with the pAS1-CYH2-MEK5, This is an Open Access article under the CC BY license. 12666 A New Human Protein Kinase Signal Transduction Pathway 1 70 MEKS MLWLALGPFPAMENQVLVIRIKIPNSGAVDWTVHSGPQLLFRDVLDVIGQVLPEATTTAFEYEDEDGDRI MKK4 MKK3 MEK2 .. MLARRKPVLPALTINPTIAEGP MEK1 ... MPKKKP .. TPIQLNPA.PDGS SEK1 .MAAPSPSGGGGSGGGGGTPGPI 71 140 MEKS TVRSDEEMKAMLSYYYSTVMEQQVNGQLIEPLQIFPRACKPPGERNIHGLKVNTRAGPSQHSSPAVSDSL MKK4 · MQGKRKALKLNF ANPP FKSTARFTLNPNPTGVQNPH. . . . . . . . . . . . . . .. . I ERL MKK3 · MSKPP APNPTPPRN . MEK2 SPTSEGASEANLVDLQKKLEELELDEQQKKRLEAF . MEK1 AVNGTSSAETNLEALQKKLEll LE LDEQQRKRLEAF . SEK1 GPPASGH PAY S SMQGKRKRLKLNFANPPVK STARFTLNPNTTGVQNPH. . . I ERL 141 I *11 210 MEKS PSNSLKKSSAELKKILANGQMNEQIIRYRDTLIINGITIJAYHVPIKIJ'VIVILLDITLELQKQIM MKK4 RTHSIESSGKLKISPEQHWDFTAE LKDLGEI R AY S MVHKP QI V R RSTVDEKEQKQLL MKK3 .... LDSRTFITIG.DRNFEVEAD LVTISEL R AY V E VRHAQ TI 'V R RATVNSQEQKRLL MEK2 ........... LTQKAKVGELKDD FERISEL NG V VQHRP LI 'R L HLEIKPAIRNQII MEK1 ........... LTQKQKVGELKDD FEKISEL NG V P VSHKP LV R L HLEIKPAIRNQII SEK1 RTHSIESSGKLKISPEQHWDFTAE LKDLGEI R AY S RSTNP QI V R RSTVDEKEQKQLL III IV V 280 MEKS SEIEILIK .. CDSSIIIGIFFVENRISITIFIDGGIJD I GKMIHVIRIAVAV MKK4 MD DVVMR.SSDCP IVQ LFREGDCW M L .ST F KFYKYVYSVLDD .. VI EI KITLAT MKK3 MD DINMRT.VDCF TVT LFREGDVW M L .DT KFYR KVLDKNMTI DI EIAVSI MEK2 RE QVLHE .. CNSP IVG . FYSDGEIS M H DGG Q VLKEAKRI EI KVSIAV MEK1 RE QVLHE .. CNSP IVG 'FYSDGEIS M H DGG Q VLKKAGRI QI KVSIAV SEK1 MD DVVMR.SSDCP IVQ 'LFREGDCW M L .ST F KFYKYVYSVLDD .. VI EI KITLAT MEKS MKK4 VKA H KENLKII I I LDRS NI L I G VD I KTRD CRP IDPSASRQ ~:~~YIW'SLKILIVI 'IMIVNTRIQ)LIVI~I T~V)*jK~Y'ITNAIIV:IIISGEQ~~~ MKK3 VRA EH HSKLSVI V INKE H MIG VD KTMD CKP INPELNQK LRG Y REKHQIM I VNSR EI L V G I NSF. TRS LQGTH . MEK2 MEK1 IKG TY REKHKIM I VNSR EI L G I NSF. TRS S LQGTH . VKA H K.KLEII I I LDRS NI L I G V I KTRD CRP IDPSASRQ SEK1 3S1 IX 420 MEKS MKK4 ~~~~:II~I~~~I~~~~~~~~~~~~~: :::::::::::::::::::::::::::::::::::::" GYNVK L ITMI MAILRFPYESWGT . MKK3 MEK2 .YSVQ I M LSL LAVGRYPIPPPDAKELEAIFGRPVVDGEEGEPHSISPRPRPPGRPVSGHGMDS .YSVQ I M LSL MAVGRYPIPPPDAKELELMFGCQV EGDAAETPPRPRTPGRPLSSYGMDS MEK1 GYDVR L ITL LAT. RFPYPKWNS . SEK1 421 X XI 490 MEKS QGSLMPLQLIQCIIDEDSIVLPVGE .. 'FSEPIVHIITQIRIQPKEIPAPEJM~IIVQFNDGNAAV ..... VFDQ TQV KGDP"QLSNSEEREFSPS IN VNL LT DESK PKYKE LK P ILMYEE.RAVE MKK4 MKK3 · .... PFQQ KQV EEPS "QLPAD RFSPE VD TAQ LR NPAE MSYLE ME P FTLHKT. KKTD MEK2 RPAMAIFEL DYI NEPP"KLPNGV FTPD QE VNK LI NPAE ADLK TN T IKR.SEVEEVD RPPMAIFEL DYI NEPP KLPSGV FSLE QD VNK LI NPAE ADLKQ M IKR.SDAEEVD MEK1 · .... VFDQ TQV KGDP "QLSNSEERSSPPS IN VNL LT DESK PKYKIl LK P ILMYEE. RTVE SEK1 491 S16 MEKS VSMWVCRALEERRTSRGPREAAAGH. VACYVCKILDQMPATPSSPMYVD . MKK4 MKK3 IAAFVKKILGEDS . MEK2 FAGWLCKTLRLNQPGTPTRTAVU . MEK1 FAGWLCSTIGLNQPSTPTHAAGVU .. SEK1 VACYVCKILDQMPATPSSPMYVD ... FIG. 1. Primary structure ofMEK5. The primary sequence ofMEK5 was deduced from the sequence of cDNA clones isolated from human fetal brain library. The PILEUP program (Wisconsin Genetics Computation Group) was used to align the amino acid sequence ofMEK5 with that of MEK1 and MEK2 (6), MKK3 and MKK4 (16), and SEK1 (15). Roman numerals designate protein kinase domains; *, conserved lysine in the ATP binding site and the activating phosphorylation sites of MEK1; shading indicates identity of all six sequences. pASl-CYH2 S311AiT315A MEK5 mutant, or pASl-CYH2 K195M scriptSK- (pBluescriptSK~ ERK5). This plasmid was used as template MEK5 mutant using the lithium acetate procedure (17). Plasmid DNA for in vitro transcription/translation employing T3 polymerase and a from a HeLa cell Match Maker cDNA library was then transformed into TNT kit (Promega). The in vitro synthesized ERK5 was labeled with the yeast strain containing each "bait" plasmid. Transformants were [ SJMet , and incubated with glutathione-agarose-immobilized GST, selected on SC yeast medium without Leu, Trp, His, and with 35 mM GST-MEKl, or GST-MEK5 at 4 °C for 1-2 h in PBST (4 mMNaH 2PO., 3-amino-1,2,4-triazole at 30°C for 7-10 days (17). The transformants 16 mM Na 150 mM NaCl, 0.5% Triton X-100, pH 7.4), 0.05% 2HPO., were lifted onto nitrocellulose membrane and assayed for f3-galactosid­ f3-mercaptoethanol, 10% glycerol, 0.2 mM phenymethylsulfonyl fluoride. ase (17). Plasmids were recovered from f3-galactosidase positive clones Following SDS-PAGE, the gel was treated with Amplify (Amersham and transformed into Escherichia coli strain DH5aF' by electropora­ Corp.) and autoradiographed. tion. The clone containing the longest ERK5 open reading frame was Recombinant Proteins and Kinase Assays-GST-MEK5 was con­ structed by ligating a 2-kb Ncol fragment encoding amino acid 12 named AI. In Vitro Binding Assay-A full-length cDNA encoding the ERK5 through the stop codon of MEK5 into pGEX-KG (18). GST-ERK5 was protein sequence was constructed from two partial cDNAs into pBlue- constructed by ligating the Ncol-BstEll fragment of the cDNA clone 3-1 12667 A New Human Protein Kinase Signal Tra nsduction Pathway and the BstEII-Xhol fragment of clone Al (the Xh ol site is in the vector) MEK2. This region contains phosphorylation sites and nega­ into pGEX-KG digested with N col and S all, GST-ERK5 (residues tiv e regulatory sequences (5, 21). Th e most diver gent region 1- 431) was constructed by subcloning a 1.1-kb Nco l fragment from among the MEKs is upstream of kinase subdoma in 1. MEK5 is pBluescriptSK- ERK5(encoding ERK5 amino acids 1- 431) into pGEX­ distinct from other MEKs in th at it contains a long N-terminal KG. GST-fusion proteins were affinity-purified as described (18). The seque nce. A data bank sea rch with the fir st 150 resi dues of kinase assays were performed at 30 °C for 10- 40 min in a buffer containing 18 mxt Hepes (pH 7.5), 10 mxr magnesium acetate, 50 J.LM MEK5 revealed sequence identity with two proteins that have 3 2 ATP, and 10 J.LCi of [y_ PIATP. Phosphoamino acid analysis was car­ important roles in cell division, mating and morphogen esis, ried out as described previously (6). namely th e yeas t cell divi sion protein 24 (encoded by Saccha ­ romyces cereoisia CDC24 gene (22) a nd the homologous protein RESULTS AND DISCUSSION encoded by the sed ] gene from S chizosaccharomyces pom be Isolati on and Chara cterization of MEK5 eDNA-We utilized (23). Th e amino acid sequenc e alignment of MEK5 with the PCR to identify additional proteins with sequence similarity to encoded proteins of th e sed ] gene and CDC24 gene is shown in MEK. Oligonucleotide primers for PCR amplification were Fig. 2. based on the conserved seque nces from th e yeast STE7 and Genetic and biochemical studies have demonstrated that the byrl genes and the partial amino acid seq uence of the hu man CDC24 encoded protein has GDP release activity. When the MEKI (6). A population ofO.35-kb fragments were pro duced by CDC24 encoded protein binds to the GTPase encoded by PCR using human fetal brain cDNA as a template. Th e PCR CDC42, it enhances the GTP/GDP exchange (24). In a similar products were subcloned into an M13 vector. MEKI cDNA was fashion, the scdl en coded protein enhances the nu cleotid e ex­ used subsequently as a hybridization probe to eliminate MEKI change of th e corresponding S. pombe encoded protein and MEK2 from the PCR clones. Th e non-hybridizing clones (cdc42sp). The binding dom ain of scdl that recognizes cdc42sp were sequenced . Sequence analysis revea led clones encoding is located C-te rminal to amino acid 671 of scdl (23). It is this MEK3 and MEK4 (16 ) as well as a novel MEK homolog. We region of scdl that shows seque nce identity to MEK 5 as well as shall refer to the novel sequenc e as MEK5. The PCR fragment CDC24. Although the function of this sequenc e is unknown , it ofMEK5 was used to screen a human fetal brain cDNA library. is poss ible that the N terminus of MEK5 interacts with th e Twenty-four positive clones were analyzed by restriction map­ mammalian equivalent of CDC42. Th e mammali an equiva lent pin g. The longest insert (2.5 kb ) was seq uenced and yielded a of yeast CDC42 is Rac (25). Th e amino acid seque nces in pr edicted protein sequence of 444 amino aci ds (Fig. 1). As question may provide a mechanism for coupling GTPases shown in Fig. 1, all 11 kinase subdomains are conserved for (i.e. CDC42 and Rae) to down stream protein kinase signaling MEK5 (16). The amino acid sequence identity of MEK5 with cascades. ~nown MEKs is approximately 40%. Th e Raf- l phosphoryla­ A Partnership between MEK5 and ERK5-Mutations in ac­ tion and activation motif of MEKl, S218XXXS222 (19, 20), was tivating phosphorylation sites and the conserved lysin e res idue also conserved as S311XXXT in MEK5, suggesting a similar in the ATP binding site of MEKI result in dominant negative regulatory mechanism. Like MEK3 and MEK4 (15 ), there was mutants that interfere with the kin ase funct ion (26-28). We an amino acid sequence gap between domains IX and X, result­ anticipated that similar modifications in MEK5 migh t enha nce ing in deletion of the proline-rich region seen in MEKI and its interaction with upstream and/or downstream signaling molecu les. To enhance the enzyme-substrate in teraction, mu­ 61 EIE ' - ' . RI T VRI' E 77 MEK5 tations that led to the dominant negative pheno type of MEK I 786 K V . - . 'FI T IT " 0 8 02 scd1 696 K Q. - . ' F V V LG • - 0 715 COC24 were ma de in MEK5, i.e. S311A1T315A and K195M (20). Usi ng the yeast two-hybrid system , three "bait" plasmids containing FIG. 2. Sequence align ment of MEK5 with the scd l gen e and CDC24 gene encode d proteins. sequences encoding wild-type MEK5, MEK5 S311A1T315A, or FIG. 3. Specific interaction between I ' MEK5 and ERK5. The ERK5-containing plasmid Al (ERK5 fusion with gal4 acti­ vation domain) was cotransformed into yeast stra in Y190 with vector pASl­ CYH2 or "bait" plasmids of MEKl, MEK2, MEK5, MEK5 S311A!T315A, and MEK5 K195M. A , transformants were re­ streaked to SC medium minus Leu Trp and His, plus 3-amino-l ,2,4-triazole (35 rnxt) to select for interaction. B, the trans­ formants were also restr eaked to SC me­ dium minus Leu and Trp and screened for J3-galactosidase activity. C, diagram showing the orientation of each tran sfer­ MWkb mant. D, in vitro translated ERK5 (1 ) was incubated with glutathione-agarose-im­ 200 - mobilized GST (2), GST-MEKI (3), GST­ MEK5 (4), and GST-MEK5(KI95M) (5) followed by washing and SDS-PAGE. Af­ ter treatment with Amplify, the SDS gel was autoradiographed with Kodak x-ray film . 30 - 12668 A New Human Protein Kinase Signal Transduction Pathway 1 ****** 70 ERK5 MAEPLKEEDGEDGSAEPPAREGRTRPHRCLCSAKNLALLKARSFDVTFDVGDEYEIIETIGNGAYGVVSS ERKI ........ MAAAAAQGGGGGEPRRTEGVGPGVPGEVEMVKGQ PFDVGPRYTQLQYIGEGAYGMVSS ERK2 ........ MAAAAAAGP EMVRGQ VFDVGPRYTNLSYIGEGAYGMVCS JNKI '" MSRSKRDNNFYSVEIGDSTFTVLKRYQNLKPIGSGAOQIVCA Hp36 .............................. MSQERPTFYRQELNKTlWEVPERYQNLSPVGSGAYGSVCA 71 ***** 140 ERK5 ARRRLTGQOVAIKKIPNAFDVVTNAKRTLRELKILKHFKHDNIIAIKDILRPTVPYGEFKSVYVVLDLME ERKI AYDHVRKTRVAIKKIS.PFEHQTYCQRTLREIQILLRFRHENVIGIRDILRAS.TLEAMRDVYIVQDLME ERK2 AYDNLNKVRVAIKKIS.PFEHQTYCQRTLREIKILLRFRHENIIGINDllRAP.TIEQMKDVYIVQDLME JNKI AYDAILERNVAIKKLSRPFQNQTHAKRAYRELVLMKCVNHKNIIGLLNVFTPQKSLEEFQDVYIVMELMD Hp36 APDTKTGLRVAVKKLSRPFQSI lHAKRTYRELRLLKHMKHENVIGLLDVFTPARSLEEFNDVYLVTHLMG 141 210 ERK5 SDLHQIIHSSQPLTLEHVRYFLYQLLRGLKYMHSAQVIHRDLKPSNLLVNENCELKIGDFGMARGLCTSP ERK1 TDLYKLL.KSOQLSNDHICYFLYQILRGLKYIHSANVLHRDLKPSNLLINTTCDLKICDPGLAR.IADPE ERK2 TDLYKLL.KTQHLSNDHICYPLYQILRGLKYIHSANVLHRDLKPSNLLLMTTCDLKICDFGLAR.VADPD JNK1 ANLCQVIQM.. ELDHERHSYLLYQMLCGIKHLHSAGIIHRDLKPSNIVVKSDCTLKILDPGLARTAGTS. Hp36 ADLNNIV.KCQKLTDDHVQPLIYQILRGLKYIHSADIIHRDLKPSNLAVNEDCELKILDPGLARHTDDE. 211 f\f\f\ *** 280 ERK5 AEHQYFMTEYVATRWYRAPELHLSLHEYTQAIDLWSVGCIFGEMLARRQLFPGKNYVHQLQLIMMVLGTP ERK1 HDHTGPLTEYVATRWYRAPEIMLNSKGYTKSIDIWSVGClLAEMLSNRPIFPGKHYLDQLNHILGILGSP ERK2 HDHTGFLTEYVATRWYRAPEIMLNSKGYTKSIDIWSVGClLAEMLSNRPIFPGKHYLDQLNHILGILGSP JNK1 .... PMMTPYVVTRYYRAPEVILG.MGYKENVDLWSVGCIMGEMVCHKILFPGRDYIDQWNKVIEQLGTP FIG.4. The primary structure of Hp38 ...... MTGYVATRWYRAPEIMLNWMHYNQTVDIWSVGClMAELLTGRTLFPGTDHIDQLKLILRLVGTP ERK5. A, the deduced amino acid se­ quence of ERK5 was aligned with that of 281 350 ERKl and ERK2, JNKl and human ERK5 SPAVIQAVGAERVRAYIQSLPPRQPVPWETVYPGA DRQALSLLGRHLRPEPSARISAA p38.*, lysine residue in the conserved ERK1 SQEDLNCIINMKARNYLQSLPSKTKVAWAKLPPKS DSKALDLLDRMLTFNPNKRITVE ATP binding site and important catalytic ERK2 amino acid residues; A, activating phos­ SQEDLNCIINLKARNYLLSLPHKNKVPWNRLPPNA DSKALDLLDKMLTPNPHKRIEVE JNK1 phorylation motif. Proline-rich region 1 CPEPMKKL.QPTVRTYVENRPKYAGYSPEKLPPDVLPPADSEHNKLKASQARDLLSKMLVIDASKRISVD Hp38 and proline-rich region 2 are underlined. GAELLKKISSESARNYIQSLTQMPKMNFANVPIGA NPLAVDLLEKMLVLDSDKRITAA 351 420 ERKS AALRHPFLAKYHDPDDEPDCAPPP.DPAFDREALTRERlKEAIVAEIEDPHARREGIROQIRFQPSLQPV ERK1 EALAHPYLEQYYDPTDEPVAEEPF.TFAMELDDLPKERLKELIFQETA RPQPGVLEA ERK2 QALAHPYLEQYYDPSDEPIAEAPF.KFDMELDDLPKEKLKELIFEETA RPQPGYRS. JNK1 EALQHPYINVWYDPSEAEAPPPKIPDKQLDEREHTIEEWKELIYKEVMDLEER.. TKNGVIRGQPSPLAQ Hp38 QALAHAYFAQYHDPDDEPVAD.PY.DQSFESRDLLIDEWKSLTYDEVISFVPPPLDQEEMES . 421 P Rich 1 490 ERK5 ASEPGCPDVEMPSPWAPSGDCAMESPPPAPPPCPGPAPDTIDLTLQPPPPYSEPAPPKKDGAISDNTKAA ERK1 P.. ERK2 JNK1 VQQ Hp38 491 560 ERK5 LKAALLKSLRSRLRDGPSAPLEAPEPRKPVTAQERQREREEKRRRRQERAKEREKRRQERERKERGAGAS ERK5 GGPSTDPLAGLVLSDNDRSLLERWTRMARPAAPALTSVPAPAPAPTPTPTPYOPTSPPPGPLAQPTGPOP P Rich 2 ERK5 OSAGSTSGPVPOPACPPPGRAPHPTGPPGPIPYPAPPOIATSTSLLAAQsLVPPPGLPGSSTP9YLPYFP ERK5 PGLPPPPAGGAPQSSMSESPDVNLVTQOLSKSQVEDPLPPVFSGTPKGSGAGYGVGFDLEEPLNQSPDMG ERK5 VADGPQDGQADSASLSASLLADWLEGHGMNPADIESLQREIQMDSPMLLADLPDLQDP MEK5 K195M were used to screen a HeLa cell library. From a useful in the identification of additional upstream as well as 6 6 6 pool of 4.2 X 10 transformants (2 X 10 from MEK5, 2 X 10 downstream signaling proteins in similar cascades. from S311Afl'315A mutant, and 0.2 X 10 from K195M mu­ To obtain the full-length clone for ERK5, the 2.5-kb insert of tant), 6 clones were identified that encoded a novel ERK, which Al was used to screen a human fetal brain eDNA library. From we shall refer to as ERK5. Five clones in this pool were ob­ 60 positive clones, two were used to complete sequence analy­ tained with the S311Afl'315A mutant of MEK5 and one with sis. The 2828-base pair sequence contained an open reading the K195M mutant. All clones contained the TEY activation frame of 2445 nucleotides encoding 815 amino acids (Fig. 4). A motif. The product encoded by the longest ERK5 clone, AI, number of residues important for the kinase activity of ERKs are conserved, including the TEY activation motif (1). ERK5 interacted with MEK5 in the yeast two-hybrid system, but not with MEKI or MEK2 (Fig. 3, A-C). In addition, MEK5 does not contains a 400-amino acid C-terminal domain, which houses interact with p38 or Raf A, B, or C in the yeast two-hybrid two proline-rich regions. The first proline-rich region, consist­ system. This result supports the highly specific nature of the ing of 32 amino acids, contains 16 prolines. The second proline­ rich region (124 amino acids with 44 prolines) contained sev­ interactions between MEK5 and ERK5. It should be noted that wild-type MEK5 also interacted with ERK5, although we did eral small Pro-Ala repeats, (PA)3' followed by Pro-Thr repeats, not isolate any ERK5 clones from the initial yeast two-hybrid (PT)3' and three repeats of PPGP. The (PA)n repeat is present screen. We anticipate that dominant negative MEKs may be in myosin light chain kinase, and this sequence has been shown A New Human Protein Kinase Signal Transduction Pathway 12669 moval of the C-termina l domain resul ts in the ability of the fun ctional ERK5 kinase domain to displ ay its kinase activity. The fact that ERK5, like ERK1 , has a TEY motif an d that phospho-Ser/Thr and Tyr are observed in the ana lysis of la­ ~lWkb beled ERK5, suggests that the catalytic properties of the two kin ases may be similar. 214 - In summa ry, we have iden tified two novel kin ases that in­ ter act selectively with one another in what appears to repre­ - P-S _GST·ERK5 sent a new mammali an signa l transductio n pathway. Our re­ 111 - sults sugges t th at there are no interactions between the _ P·T protein s in the MEK1/ERK1 pathway an d the MEK5/ERK5 pa th way, suggesting that the two signaling pathways have 74 - _ GST·ERK5(1-431) - p.y eithe r different or complementary functions. Our res ults also suggest that th e two differen t signaling path ways most likely do not have overlapping functions in the cell. We are currently atte mpting to identify both upstream activators of MEK5 as 45 - well as downstream substra tes of ERK5. The unu sual length of the Codomain of ERK5 suggests that the sequenc es located FI G. 5.Au top h osp h oryla t ion of recombinant GST-ERK5 (1­ here may undergo additio na l post-tran slational modificat ions 431). A, autophosp hory lation of purified full-length GST-ERK5 and GST-ERK5 (1-431). The sa mples wer e reso lved on a SDS-PAGE gel as well as play roles in regulating or targe ting of ERK5. followed by autoradiography. Th e arrows indicate the position of the purified GST-ERK5 an d GST-ERK5 0-43 1). B , phos phoa mino acid Acknowledgments-We thank K.-L. Guan for plasmid s encoding ana lysis of autophos phorylated GST-ERK5 (1- 431 shown in A. P-S, MEK1, MEK2 and ERKI. We also th ank B. Yas har and R. Ston e for phosphoseri ne; poT, phosph oth reonine; P-Y, phosph otyrosin e. critical rea ding of the ma nu script. We also acknowledge th e help of J. Clemens, M. Wish art, and C. F. Zhen g with the yeast two-hybrid syste m. to directly interact with actin, targeting the kinase to a specific location in the cell (29). The unique ERK5 C-te rmina l seque nce REF E RE NCES may serve as a locali zation domain an d/or a regulatory domain. Cell 80, 179-185 1. Mar sh all, C. J . (1995) interesting to speculate th at th e (PA)" seque nces which It is 2. Sch lessinger , J ., and Ullr ich, A. (1992) Neuron 9, 383-39 1 3. Boulton, T. G., Nye, S. H., Robbin s, D. J ., Ip, Y. N., Radziejewsk a, E., target myosin light chain kinase to actin may also target ERK5 Morgenbe sser , S., Depinho , R., Pan ayotatos, N., Cobb, M. H., and to a simila r location. Northern blot ana lysis of ERK5 mRNA Yancopoulos, G. D. (1991) Cell 65, 663-675 (3.1 kb) and MEK5 mRNA (2.7 kb ) demonstrated they are 4. Sturgill, T. W., a nd Ray, L. B. (1986 ) Biochem . Biophys. Res. Commun. 134, 565-571 expressed in man y human tissu es and that th e RNAs are most 5. Wu, J ., Harri son, J . K., Vincent, L. A., Haystead , C., Harstead , T. A. J ., Michel, abunda nt in heart and skeletal mu scle (da ta not shown). H., Hunt, D. F., Lynch, K. R., and Sturgill, T. W. (1993 ) Proc. Na tl. Acad. To fur th er docum ent the specific interaction between MEK5 Sci. U. S. A. 90, 173-177 6. Zheng , C. F., a nd Guan , K. L. (1993 ) J . Bioi. Chem. 268, 11435-11439 and ERK5 , in vitro binding assays were conducted (Fig. 3D). In 7. Derijard, B., Hibi , M., Wu, I.-H. , Barrett, T., Su, B., Deng, T., Kari n, M., and this exper iment, full-length ERK5 was synthes ized in an in Davis, R. (1994 ) Cell 76, 1025-1 037 vitro transcription/t ranslation system. As shown in Fig. 3D, 8. Kyriakis, J . M., Ban erjee, P., Nikolak aki , E., Dai , T., Rubie, E. A., Ahmad , M. F., Avruch, J ., and Woodgett, J . R. (1994) Nature 369 , 156 -1 60 ERK5 binds to GST-MEK5 and the GST-MEK5 mutant 9. Han , J ., Lee, J . D., Bibbs, L., a nd Ulevitch, R. J . (1994) Science 265, 808- 811 K195M, bu t not to GST alone or to GST-MEKl. Thus, MEK5 10. Rouse, J. , Cohen , P., Trigon, S., Moran ge, M., Alonso-L1amazares, A., specifically interacted with ERK5 in both our in vitro syste m Zamanillo, D., Hunt, T., a nd Nebreda , A. R. (1994 ) Cell 78, 1027-1037 11. Lee J . C., Layton , J . T., McDonn el, P. C., Gall agher , T. F., Kum ar , S., Green, and in the yeast two-hybrid system. D., McNulty, D., Blum enth al, M. J ., Heys, J . R., Lahd vat ter, S. W., et 01. Protein Kin ase Activity of MEK5 and ERK5-GST fusi on (1994 ) Natu re 372, 739-746 protein containing full-length MEK5 could not be overex­ 12. Herskowitz, I. (1995) Cell 80, 187-197 13. Ammerer, G. (1994 ) Curro Opin . Genet. Dev. 4, 90-95 pressed in E. coli. However , elimination of the N-terminal 11 14. Errede, B., a nd Levin, D. E. (1993) CurroOpin . Cell Bioi. 5, 254- 260 res idues produced a MEK5 fusion protein which could be over­ 15. Sanc hez, I., Hughes, R. T., Mayer, 0. J ., Yee, K., Woodgett, J. R., Avru ch, J ., expressed an d affinity-purified. Th e purified protein showed no Kyri ak is, J. M., and Zon, L. I. (1994) Nature 372, 794 - 798 16. Derij ard , B., Rain gea ud, J ., Barrett, T., Wu, I.-H., Han, J ., Ulevit ch, R. J ., a nd protein kinase activity toward ERK1 , JNK1, or GST-ERK5. Science 267, 682--685 Davis, R. J . (1995 ) However, a weak autophosphorylation activity could be de­ 17. Lundblad , V. (1988) in Current Protocols in Molecular Biology (Aus ubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidm an , J . G., Smith, J . A., and tecte d with the wild-type enzyme but not with the MEK5 K195M St ru hl, K., eds ) pp. 13.7.1- 13.7.5, John Wiley & Sons, New York mutant (data not shown). The se experiments suggested that 18. Guan, K. L., and Dixon, J . E. (199 1) Anal. Biochem. 192, 262-267 additiona l modifications are requi red for full kinase activity. 19. Alessi, D. R., Saito, Y., Ca mpbell, D. G., Cohe n, P., Sith an andam , G., Rapp, V., Ashworth, A., Marshall, C. J ., a nd Cowley, S. (1994 ) EMRO J. 13, Two GST fusio n protein s of ERK5 were created for expres ­ 1610-1619 sion in E. coli, a full-length GST-ERK5 and a truncated GST­ 20. Zheng, C. F., a nd Guan, K. L. (1994) EMBO J . 13, 1123- 1131 ERK5 with the C-te rminal 384 residues removed. Thi s trun­ 21. Rossomando, A. J ., Dent , P., Sturgill, T. W., a nd Marshak, D. R. (1994 ) Mol. Cell. Bioi. 14, 1594 - 1602 cated ERK5 corr esponds to th e kinase domain of th e protein , 22. Miya moto, S., Ohya, Y., Ohsumi, Y., and Anr aku, Y. (1987) Gene (Arnst.) 54, i.e. GST-ERK5(1-431). Both fusion proteins wer e purified on 125-132 glutathione-aga rose affinity columns. The full-length GST­ 23. Chang, E. C., Ba rr, M., Wan g, Y., Jung, V., Xu, H. P., and Wigler, M. H. (1994) Cell 79, 131-141 ERK5 protein had no detectable protein kin ase activity (Fig. 5A 24. Zheng, Y., Ceri one, R., and Bender , A. (1994) J. BioI. Chern. 269, 2369-2372 and data not shown). The purified GST-ERK5(1-431) show ed 25. Manser, E., Leu ng, T., Salihuddin, H., Zhao, Z. S., and Lim, L. (1994) Nature basal autophosphorylation activi ty, as was evident from the 367 , 40- 46 26. Cowley, S., Paterson , H., Kemp, P., a nd Marshall , C. J . (1994) Cell 77, 841-852 radi oactive band of th e predicted molecular weight following 27. Seger R., Seger, D., Reszka, A. A., Mun ar , E. S., Eldar-Finkelm an , H., SDS-PAGE analysis (Fig. 5A). Phosphoamino acid analysis of Dobrowolska, G., J ensen , A. M., Ca mpbell, J . S., Fischer , E. H., and Krebs, the rad ioactive protein revealed phosphotyrosine, phospho­ E. G. (1994 ) J . BioI. Chern. 269, 25699 - 25709 28. Man sour, S. J ., Matten , W. T., Herm an n, A. S., Candia, J . M., Rang, S., threonine, and phospho serine (Fig. 5B). These resu lts suggest Fuka sawa, K., Vande Woude, G. F., and Ahn , N. G. (1994) Science 265, tha t th e kin ase domain of the full-length ERK5 is render ed 966 -970 29. Williamson, M. P. (1994) Biochem. J. 297, 249- 260 inactive by the presence of th e C-te rmina l 384 residues. Re- http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

Components of a New Human Protein Kinase Signal Transduction Pathway

Zhou, Gaochao; Bao, Zhao Qin; Dixon, Jack E.
Journal of Biological Chemistry – May 1, 1995
Download PDF
5 pages

Loading next page...
 
/lp/unpaywall/components-of-a-new-human-protein-kinase-signal-transduction-pathway-OgiMoaUPa5

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.270.21.12665
Publisher site
See Article on Publisher Site

Abstract

Vol. 270, No. 21, Issue of May 26, pp. 12665-12669, 1995 THE JOURNAL OF BIOLOGICAL CHEMISTRY Printed in U.S.A. © 1995 by The American Society for Biochemistry and Molecular Biology, Inc. Components of a New Human Protein Kinase Signal Transduction Pathway* (Received for publication, March 27, 1995) Gaochao Zhou, Zhao Qin Bao, and Jack E. Dixont From the Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, Michigan 48109-0606 motif, where "X" can be Glu, Pro, or Gly (1). There are also a We have identified two components of a new protein growing number of MEKs (15, 16). In several cases, specific kinase signaling cascade, MAPKIERK kinase 5 (MEK5) and extracellular signal-regulated kinase 5 (ERK5). The MEKs have been shown to phosphorylate specific ERKs in a MEK5 cDNA was isolated by degenerate PCR and en­ given pathway. MEKI and MEK2 phosphorylate ERKI and codes a 444-amino acid protein, which has approxi­ ERK2 in a mitogenesis or differentiation response; MEK3 and mately 40% identity to known MEKs. ERK5 was identi­ p38 show coordinate activation in the cytokine response, as do fied by a specific interaction with the MEK5 mutants MEK4 (i.e. SEKl, the murine homolog ofMEK4) and the JNKs S311Afr315A and K195M in the yeast two-hybrid system. in the stress response. Sequential protein kinase cascades have The proteins were found to interact in an in vitro bind­ been studied extensively in yeast, where genetic tools have ing assay as well. ERK5 did not interact with MEKI or been invaluable in assigning kinase cascades to specific signal MEK2. ERK5 is predicted to contain 815 amino acids and transduction pathways (12-14). In higher eukaryotic systems, is approximately twice the size ofall known ERKs. The C less is known about the specific stimuli responsible for activa­ terminus of ERK5 has sequences which suggest that it tion of individual signal transduction pathways. Although may be targeted to the cytoskeleton. Sequences located there are clearly multiple genes encoding ERKs and MEKs, we in the N terminus ofMEK5 may be important in coupling have no good idea about how many pathways there are in GTPase signaling molecules to the MEK5 protein kinase eukaryotic cells. cascade. Both MEK5 and ERK5 are expressed in many We have identified a new MEK, which we will refer to as adult tissue and are abundant in heart and skeletal mus­ MEK5. When MEK5 was employed in the yeast two-hybrid cle. A recombinant GST-ERK5 kinase domain displays system, we identified a new member of the ERK family of autophosphorylation on Ser!l'hr and Tyr residues. protein kinases, which we have named ERK5. MEK5 and ERK5 interacted specifically with one another and did not interact with kinases in the MEKI/ERKI signaling pathway, Some cell surface receptors respond to extracellular stimuli suggesting that the MEK5/ERK5 protein cascade represents a via a sequential protein kinase cascade (1, 2). The kinase sig­ novel signaling pathway in higher eukaryotes. MEK5 contains naling cascade involves multiple protein kinases, two of which unique amino acid sequences in its N terminus, which suggest are extracellular signal-regulated kinase (ERK\ Refs. 3 and 4)2 that this molecule may interact with GTPases such as CDC42. and MAPKIERK kinase (MEK; Refs. 5 and 6).3 Several ERK In addition, the C terminus of ERK5 also contains sequences related proteins have been identified in vertebrates including which may target this kinase to the cytoskeletal elements c-Jun N-terminal kinase (JNK; Refs. 7 and 8)4 and p38 (9-11). within the cell. A variety of similar proteins have also been identified in yeast (12-14). All of the ERK kinases are activated by phosphoryla­ MATERIALS AND METHODS tion on Tyr and Thr residues within a TXY phosphorylation peR and cDNA Screening-Two degenerate PCR primers were syn­ thesized based on the conserved amino acid sequences present in MEKs from mammals and yeast (6). The sense primer was 5'-CTTGGATC­ * This work was supported by National Institutes of Health NIDDKD QTA(CIT)ATA(CIT)GTNGGNTT(CIT)TA-3' and the antisense primer Grant 18024 and a postdoctoral fellowship from the American Diabetes was 5'-CTTGGATCC(G/T)TCNGGN(C/G)(AIT)CAT(NG)TA-3', which Association (to G. Z,). The costs of publication of this article were corresponds to amino acid sequences YIVGFY and YMSPER, respec­ defrayed in part by the payment of page charges. This article must tively. The underlined BamHI sites were used for cloning the PCR therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. product. Plasmid DNA from a AZAP human fetal brain cDNA library was used as template. The PCR was performed for 35 cycles at 94 DC for The nucleotide sequencers) reported in this paper has been submitted to the GenBanFM I EMBL Data Bank with accession numberts) U25265 1 min, 50 DC for 2 min, and 72 DC for 2 min. The resultant 0.35-kb PCR and U25278. products were resolved by 3% Nusieve-agarose gel electrophoresis and :j:To whom correspondence should be addressed: Dept. of Biological subcloned into a M13 vector and sequenced. Chemistry, University of Michigan Medical School, Rm. 5416, Medical A human fetal brain cDNA library was screened (Stratagene), using Science I, Ann Arbor, MI 48109-0606. Tel.: 313-764-8192; Fax: 313-763­ the cloned MEK5 PCR fragment labeled with a multiprimer DNA labeling kit. Hybridization was carried out at 42 DC in 6 X SSC (1 X 1 The abbreviations used are: ERK, extracellular-regulated kinase; SSC is 0.15 M NaCI, 0.015 M sodium citrate, pH 7.0), 1 X Denhardt's MAP, mitogen-activated protein; MAPK, MAP kinase; MAPKK or solution (1 X Denhardt's solution is 0.02% Ficoll400, 0.02% bovine serum MKK, MAP kinase kinase; MEK, MAPKIERK kinase; JNK, c-Jun N­ albumin, 0.02% polyvinylpyrrolidone), 0.1 mg/ml denatured salmon terminal kinase; kb, kilobase pair'(s); PCR, polymerase chain reaction; sperm DNA, 50% formamide. Both strands ofthe DNA corresponding to PAGE, polyacrylamide gel electrophoresis; GST, glutathione S-trans­ the clone designated pBluescriptSK~ MEK5 were sequenced. ferase. The Yeast Two-hybrid Screen-The yeast strain Y190 (MATa, 2 ERK is also referred to as mitogen-activated protein kinase (MAPK) leu2-3, 112, ura3-52 trpl-901 his3-Ll.200 ade2-101 gal4 tiga180Ll. URA in the literature. We will employ the ERK nomenclature. 3 GAL-lacZ LYS GAL-HIS3 cyh'') was a generous gift of Stephen 3 MEK is also referred to as mitogen-activated protein kinase kinase (MAPKK or MKK) in the literature. We will employ the MEK Elledge, Departments of Biochemistry and Human and Molecular Genetics, Baylor College of Medicine. The NcoI fragment encoding amino nomenclature. 4 JNK is also referred to as stress-activated protein kinase (SAPK) in acid 12 through the stop codon of MEK5 were subcloned into pAS1­ the literature. CYH2 vector. Y190 was transformed with the pAS1-CYH2-MEK5, This is an Open Access article under the CC BY license. 12666 A New Human Protein Kinase Signal Transduction Pathway 1 70 MEKS MLWLALGPFPAMENQVLVIRIKIPNSGAVDWTVHSGPQLLFRDVLDVIGQVLPEATTTAFEYEDEDGDRI MKK4 MKK3 MEK2 .. MLARRKPVLPALTINPTIAEGP MEK1 ... MPKKKP .. TPIQLNPA.PDGS SEK1 .MAAPSPSGGGGSGGGGGTPGPI 71 140 MEKS TVRSDEEMKAMLSYYYSTVMEQQVNGQLIEPLQIFPRACKPPGERNIHGLKVNTRAGPSQHSSPAVSDSL MKK4 · MQGKRKALKLNF ANPP FKSTARFTLNPNPTGVQNPH. . . . . . . . . . . . . . .. . I ERL MKK3 · MSKPP APNPTPPRN . MEK2 SPTSEGASEANLVDLQKKLEELELDEQQKKRLEAF . MEK1 AVNGTSSAETNLEALQKKLEll LE LDEQQRKRLEAF . SEK1 GPPASGH PAY S SMQGKRKRLKLNFANPPVK STARFTLNPNTTGVQNPH. . . I ERL 141 I *11 210 MEKS PSNSLKKSSAELKKILANGQMNEQIIRYRDTLIINGITIJAYHVPIKIJ'VIVILLDITLELQKQIM MKK4 RTHSIESSGKLKISPEQHWDFTAE LKDLGEI R AY S MVHKP QI V R RSTVDEKEQKQLL MKK3 .... LDSRTFITIG.DRNFEVEAD LVTISEL R AY V E VRHAQ TI 'V R RATVNSQEQKRLL MEK2 ........... LTQKAKVGELKDD FERISEL NG V VQHRP LI 'R L HLEIKPAIRNQII MEK1 ........... LTQKQKVGELKDD FEKISEL NG V P VSHKP LV R L HLEIKPAIRNQII SEK1 RTHSIESSGKLKISPEQHWDFTAE LKDLGEI R AY S RSTNP QI V R RSTVDEKEQKQLL III IV V 280 MEKS SEIEILIK .. CDSSIIIGIFFVENRISITIFIDGGIJD I GKMIHVIRIAVAV MKK4 MD DVVMR.SSDCP IVQ LFREGDCW M L .ST F KFYKYVYSVLDD .. VI EI KITLAT MKK3 MD DINMRT.VDCF TVT LFREGDVW M L .DT KFYR KVLDKNMTI DI EIAVSI MEK2 RE QVLHE .. CNSP IVG . FYSDGEIS M H DGG Q VLKEAKRI EI KVSIAV MEK1 RE QVLHE .. CNSP IVG 'FYSDGEIS M H DGG Q VLKKAGRI QI KVSIAV SEK1 MD DVVMR.SSDCP IVQ 'LFREGDCW M L .ST F KFYKYVYSVLDD .. VI EI KITLAT MEKS MKK4 VKA H KENLKII I I LDRS NI L I G VD I KTRD CRP IDPSASRQ ~:~~YIW'SLKILIVI 'IMIVNTRIQ)LIVI~I T~V)*jK~Y'ITNAIIV:IIISGEQ~~~ MKK3 VRA EH HSKLSVI V INKE H MIG VD KTMD CKP INPELNQK LRG Y REKHQIM I VNSR EI L V G I NSF. TRS LQGTH . MEK2 MEK1 IKG TY REKHKIM I VNSR EI L G I NSF. TRS S LQGTH . VKA H K.KLEII I I LDRS NI L I G V I KTRD CRP IDPSASRQ SEK1 3S1 IX 420 MEKS MKK4 ~~~~:II~I~~~I~~~~~~~~~~~~~: :::::::::::::::::::::::::::::::::::::" GYNVK L ITMI MAILRFPYESWGT . MKK3 MEK2 .YSVQ I M LSL LAVGRYPIPPPDAKELEAIFGRPVVDGEEGEPHSISPRPRPPGRPVSGHGMDS .YSVQ I M LSL MAVGRYPIPPPDAKELELMFGCQV EGDAAETPPRPRTPGRPLSSYGMDS MEK1 GYDVR L ITL LAT. RFPYPKWNS . SEK1 421 X XI 490 MEKS QGSLMPLQLIQCIIDEDSIVLPVGE .. 'FSEPIVHIITQIRIQPKEIPAPEJM~IIVQFNDGNAAV ..... VFDQ TQV KGDP"QLSNSEEREFSPS IN VNL LT DESK PKYKE LK P ILMYEE.RAVE MKK4 MKK3 · .... PFQQ KQV EEPS "QLPAD RFSPE VD TAQ LR NPAE MSYLE ME P FTLHKT. KKTD MEK2 RPAMAIFEL DYI NEPP"KLPNGV FTPD QE VNK LI NPAE ADLK TN T IKR.SEVEEVD RPPMAIFEL DYI NEPP KLPSGV FSLE QD VNK LI NPAE ADLKQ M IKR.SDAEEVD MEK1 · .... VFDQ TQV KGDP "QLSNSEERSSPPS IN VNL LT DESK PKYKIl LK P ILMYEE. RTVE SEK1 491 S16 MEKS VSMWVCRALEERRTSRGPREAAAGH. VACYVCKILDQMPATPSSPMYVD . MKK4 MKK3 IAAFVKKILGEDS . MEK2 FAGWLCKTLRLNQPGTPTRTAVU . MEK1 FAGWLCSTIGLNQPSTPTHAAGVU .. SEK1 VACYVCKILDQMPATPSSPMYVD ... FIG. 1. Primary structure ofMEK5. The primary sequence ofMEK5 was deduced from the sequence of cDNA clones isolated from human fetal brain library. The PILEUP program (Wisconsin Genetics Computation Group) was used to align the amino acid sequence ofMEK5 with that of MEK1 and MEK2 (6), MKK3 and MKK4 (16), and SEK1 (15). Roman numerals designate protein kinase domains; *, conserved lysine in the ATP binding site and the activating phosphorylation sites of MEK1; shading indicates identity of all six sequences. pASl-CYH2 S311AiT315A MEK5 mutant, or pASl-CYH2 K195M scriptSK- (pBluescriptSK~ ERK5). This plasmid was used as template MEK5 mutant using the lithium acetate procedure (17). Plasmid DNA for in vitro transcription/translation employing T3 polymerase and a from a HeLa cell Match Maker cDNA library was then transformed into TNT kit (Promega). The in vitro synthesized ERK5 was labeled with the yeast strain containing each "bait" plasmid. Transformants were [ SJMet , and incubated with glutathione-agarose-immobilized GST, selected on SC yeast medium without Leu, Trp, His, and with 35 mM GST-MEKl, or GST-MEK5 at 4 °C for 1-2 h in PBST (4 mMNaH 2PO., 3-amino-1,2,4-triazole at 30°C for 7-10 days (17). The transformants 16 mM Na 150 mM NaCl, 0.5% Triton X-100, pH 7.4), 0.05% 2HPO., were lifted onto nitrocellulose membrane and assayed for f3-galactosid­ f3-mercaptoethanol, 10% glycerol, 0.2 mM phenymethylsulfonyl fluoride. ase (17). Plasmids were recovered from f3-galactosidase positive clones Following SDS-PAGE, the gel was treated with Amplify (Amersham and transformed into Escherichia coli strain DH5aF' by electropora­ Corp.) and autoradiographed. tion. The clone containing the longest ERK5 open reading frame was Recombinant Proteins and Kinase Assays-GST-MEK5 was con­ structed by ligating a 2-kb Ncol fragment encoding amino acid 12 named AI. In Vitro Binding Assay-A full-length cDNA encoding the ERK5 through the stop codon of MEK5 into pGEX-KG (18). GST-ERK5 was protein sequence was constructed from two partial cDNAs into pBlue- constructed by ligating the Ncol-BstEll fragment of the cDNA clone 3-1 12667 A New Human Protein Kinase Signal Tra nsduction Pathway and the BstEII-Xhol fragment of clone Al (the Xh ol site is in the vector) MEK2. This region contains phosphorylation sites and nega­ into pGEX-KG digested with N col and S all, GST-ERK5 (residues tiv e regulatory sequences (5, 21). Th e most diver gent region 1- 431) was constructed by subcloning a 1.1-kb Nco l fragment from among the MEKs is upstream of kinase subdoma in 1. MEK5 is pBluescriptSK- ERK5(encoding ERK5 amino acids 1- 431) into pGEX­ distinct from other MEKs in th at it contains a long N-terminal KG. GST-fusion proteins were affinity-purified as described (18). The seque nce. A data bank sea rch with the fir st 150 resi dues of kinase assays were performed at 30 °C for 10- 40 min in a buffer containing 18 mxt Hepes (pH 7.5), 10 mxr magnesium acetate, 50 J.LM MEK5 revealed sequence identity with two proteins that have 3 2 ATP, and 10 J.LCi of [y_ PIATP. Phosphoamino acid analysis was car­ important roles in cell division, mating and morphogen esis, ried out as described previously (6). namely th e yeas t cell divi sion protein 24 (encoded by Saccha ­ romyces cereoisia CDC24 gene (22) a nd the homologous protein RESULTS AND DISCUSSION encoded by the sed ] gene from S chizosaccharomyces pom be Isolati on and Chara cterization of MEK5 eDNA-We utilized (23). Th e amino acid sequenc e alignment of MEK5 with the PCR to identify additional proteins with sequence similarity to encoded proteins of th e sed ] gene and CDC24 gene is shown in MEK. Oligonucleotide primers for PCR amplification were Fig. 2. based on the conserved seque nces from th e yeast STE7 and Genetic and biochemical studies have demonstrated that the byrl genes and the partial amino acid seq uence of the hu man CDC24 encoded protein has GDP release activity. When the MEKI (6). A population ofO.35-kb fragments were pro duced by CDC24 encoded protein binds to the GTPase encoded by PCR using human fetal brain cDNA as a template. Th e PCR CDC42, it enhances the GTP/GDP exchange (24). In a similar products were subcloned into an M13 vector. MEKI cDNA was fashion, the scdl en coded protein enhances the nu cleotid e ex­ used subsequently as a hybridization probe to eliminate MEKI change of th e corresponding S. pombe encoded protein and MEK2 from the PCR clones. Th e non-hybridizing clones (cdc42sp). The binding dom ain of scdl that recognizes cdc42sp were sequenced . Sequence analysis revea led clones encoding is located C-te rminal to amino acid 671 of scdl (23). It is this MEK3 and MEK4 (16 ) as well as a novel MEK homolog. We region of scdl that shows seque nce identity to MEK 5 as well as shall refer to the novel sequenc e as MEK5. The PCR fragment CDC24. Although the function of this sequenc e is unknown , it ofMEK5 was used to screen a human fetal brain cDNA library. is poss ible that the N terminus of MEK5 interacts with th e Twenty-four positive clones were analyzed by restriction map­ mammalian equivalent of CDC42. Th e mammali an equiva lent pin g. The longest insert (2.5 kb ) was seq uenced and yielded a of yeast CDC42 is Rac (25). Th e amino acid seque nces in pr edicted protein sequence of 444 amino aci ds (Fig. 1). As question may provide a mechanism for coupling GTPases shown in Fig. 1, all 11 kinase subdomains are conserved for (i.e. CDC42 and Rae) to down stream protein kinase signaling MEK5 (16). The amino acid sequence identity of MEK5 with cascades. ~nown MEKs is approximately 40%. Th e Raf- l phosphoryla­ A Partnership between MEK5 and ERK5-Mutations in ac­ tion and activation motif of MEKl, S218XXXS222 (19, 20), was tivating phosphorylation sites and the conserved lysin e res idue also conserved as S311XXXT in MEK5, suggesting a similar in the ATP binding site of MEKI result in dominant negative regulatory mechanism. Like MEK3 and MEK4 (15 ), there was mutants that interfere with the kin ase funct ion (26-28). We an amino acid sequence gap between domains IX and X, result­ anticipated that similar modifications in MEK5 migh t enha nce ing in deletion of the proline-rich region seen in MEKI and its interaction with upstream and/or downstream signaling molecu les. To enhance the enzyme-substrate in teraction, mu­ 61 EIE ' - ' . RI T VRI' E 77 MEK5 tations that led to the dominant negative pheno type of MEK I 786 K V . - . 'FI T IT " 0 8 02 scd1 696 K Q. - . ' F V V LG • - 0 715 COC24 were ma de in MEK5, i.e. S311A1T315A and K195M (20). Usi ng the yeast two-hybrid system , three "bait" plasmids containing FIG. 2. Sequence align ment of MEK5 with the scd l gen e and CDC24 gene encode d proteins. sequences encoding wild-type MEK5, MEK5 S311A1T315A, or FIG. 3. Specific interaction between I ' MEK5 and ERK5. The ERK5-containing plasmid Al (ERK5 fusion with gal4 acti­ vation domain) was cotransformed into yeast stra in Y190 with vector pASl­ CYH2 or "bait" plasmids of MEKl, MEK2, MEK5, MEK5 S311A!T315A, and MEK5 K195M. A , transformants were re­ streaked to SC medium minus Leu Trp and His, plus 3-amino-l ,2,4-triazole (35 rnxt) to select for interaction. B, the trans­ formants were also restr eaked to SC me­ dium minus Leu and Trp and screened for J3-galactosidase activity. C, diagram showing the orientation of each tran sfer­ MWkb mant. D, in vitro translated ERK5 (1 ) was incubated with glutathione-agarose-im­ 200 - mobilized GST (2), GST-MEKI (3), GST­ MEK5 (4), and GST-MEK5(KI95M) (5) followed by washing and SDS-PAGE. Af­ ter treatment with Amplify, the SDS gel was autoradiographed with Kodak x-ray film . 30 - 12668 A New Human Protein Kinase Signal Transduction Pathway 1 ****** 70 ERK5 MAEPLKEEDGEDGSAEPPAREGRTRPHRCLCSAKNLALLKARSFDVTFDVGDEYEIIETIGNGAYGVVSS ERKI ........ MAAAAAQGGGGGEPRRTEGVGPGVPGEVEMVKGQ PFDVGPRYTQLQYIGEGAYGMVSS ERK2 ........ MAAAAAAGP EMVRGQ VFDVGPRYTNLSYIGEGAYGMVCS JNKI '" MSRSKRDNNFYSVEIGDSTFTVLKRYQNLKPIGSGAOQIVCA Hp36 .............................. MSQERPTFYRQELNKTlWEVPERYQNLSPVGSGAYGSVCA 71 ***** 140 ERK5 ARRRLTGQOVAIKKIPNAFDVVTNAKRTLRELKILKHFKHDNIIAIKDILRPTVPYGEFKSVYVVLDLME ERKI AYDHVRKTRVAIKKIS.PFEHQTYCQRTLREIQILLRFRHENVIGIRDILRAS.TLEAMRDVYIVQDLME ERK2 AYDNLNKVRVAIKKIS.PFEHQTYCQRTLREIKILLRFRHENIIGINDllRAP.TIEQMKDVYIVQDLME JNKI AYDAILERNVAIKKLSRPFQNQTHAKRAYRELVLMKCVNHKNIIGLLNVFTPQKSLEEFQDVYIVMELMD Hp36 APDTKTGLRVAVKKLSRPFQSI lHAKRTYRELRLLKHMKHENVIGLLDVFTPARSLEEFNDVYLVTHLMG 141 210 ERK5 SDLHQIIHSSQPLTLEHVRYFLYQLLRGLKYMHSAQVIHRDLKPSNLLVNENCELKIGDFGMARGLCTSP ERK1 TDLYKLL.KSOQLSNDHICYFLYQILRGLKYIHSANVLHRDLKPSNLLINTTCDLKICDPGLAR.IADPE ERK2 TDLYKLL.KTQHLSNDHICYPLYQILRGLKYIHSANVLHRDLKPSNLLLMTTCDLKICDFGLAR.VADPD JNK1 ANLCQVIQM.. ELDHERHSYLLYQMLCGIKHLHSAGIIHRDLKPSNIVVKSDCTLKILDPGLARTAGTS. Hp36 ADLNNIV.KCQKLTDDHVQPLIYQILRGLKYIHSADIIHRDLKPSNLAVNEDCELKILDPGLARHTDDE. 211 f\f\f\ *** 280 ERK5 AEHQYFMTEYVATRWYRAPELHLSLHEYTQAIDLWSVGCIFGEMLARRQLFPGKNYVHQLQLIMMVLGTP ERK1 HDHTGPLTEYVATRWYRAPEIMLNSKGYTKSIDIWSVGClLAEMLSNRPIFPGKHYLDQLNHILGILGSP ERK2 HDHTGFLTEYVATRWYRAPEIMLNSKGYTKSIDIWSVGClLAEMLSNRPIFPGKHYLDQLNHILGILGSP JNK1 .... PMMTPYVVTRYYRAPEVILG.MGYKENVDLWSVGCIMGEMVCHKILFPGRDYIDQWNKVIEQLGTP FIG.4. The primary structure of Hp38 ...... MTGYVATRWYRAPEIMLNWMHYNQTVDIWSVGClMAELLTGRTLFPGTDHIDQLKLILRLVGTP ERK5. A, the deduced amino acid se­ quence of ERK5 was aligned with that of 281 350 ERKl and ERK2, JNKl and human ERK5 SPAVIQAVGAERVRAYIQSLPPRQPVPWETVYPGA DRQALSLLGRHLRPEPSARISAA p38.*, lysine residue in the conserved ERK1 SQEDLNCIINMKARNYLQSLPSKTKVAWAKLPPKS DSKALDLLDRMLTFNPNKRITVE ATP binding site and important catalytic ERK2 amino acid residues; A, activating phos­ SQEDLNCIINLKARNYLLSLPHKNKVPWNRLPPNA DSKALDLLDKMLTPNPHKRIEVE JNK1 phorylation motif. Proline-rich region 1 CPEPMKKL.QPTVRTYVENRPKYAGYSPEKLPPDVLPPADSEHNKLKASQARDLLSKMLVIDASKRISVD Hp38 and proline-rich region 2 are underlined. GAELLKKISSESARNYIQSLTQMPKMNFANVPIGA NPLAVDLLEKMLVLDSDKRITAA 351 420 ERKS AALRHPFLAKYHDPDDEPDCAPPP.DPAFDREALTRERlKEAIVAEIEDPHARREGIROQIRFQPSLQPV ERK1 EALAHPYLEQYYDPTDEPVAEEPF.TFAMELDDLPKERLKELIFQETA RPQPGVLEA ERK2 QALAHPYLEQYYDPSDEPIAEAPF.KFDMELDDLPKEKLKELIFEETA RPQPGYRS. JNK1 EALQHPYINVWYDPSEAEAPPPKIPDKQLDEREHTIEEWKELIYKEVMDLEER.. TKNGVIRGQPSPLAQ Hp38 QALAHAYFAQYHDPDDEPVAD.PY.DQSFESRDLLIDEWKSLTYDEVISFVPPPLDQEEMES . 421 P Rich 1 490 ERK5 ASEPGCPDVEMPSPWAPSGDCAMESPPPAPPPCPGPAPDTIDLTLQPPPPYSEPAPPKKDGAISDNTKAA ERK1 P.. ERK2 JNK1 VQQ Hp38 491 560 ERK5 LKAALLKSLRSRLRDGPSAPLEAPEPRKPVTAQERQREREEKRRRRQERAKEREKRRQERERKERGAGAS ERK5 GGPSTDPLAGLVLSDNDRSLLERWTRMARPAAPALTSVPAPAPAPTPTPTPYOPTSPPPGPLAQPTGPOP P Rich 2 ERK5 OSAGSTSGPVPOPACPPPGRAPHPTGPPGPIPYPAPPOIATSTSLLAAQsLVPPPGLPGSSTP9YLPYFP ERK5 PGLPPPPAGGAPQSSMSESPDVNLVTQOLSKSQVEDPLPPVFSGTPKGSGAGYGVGFDLEEPLNQSPDMG ERK5 VADGPQDGQADSASLSASLLADWLEGHGMNPADIESLQREIQMDSPMLLADLPDLQDP MEK5 K195M were used to screen a HeLa cell library. From a useful in the identification of additional upstream as well as 6 6 6 pool of 4.2 X 10 transformants (2 X 10 from MEK5, 2 X 10 downstream signaling proteins in similar cascades. from S311Afl'315A mutant, and 0.2 X 10 from K195M mu­ To obtain the full-length clone for ERK5, the 2.5-kb insert of tant), 6 clones were identified that encoded a novel ERK, which Al was used to screen a human fetal brain eDNA library. From we shall refer to as ERK5. Five clones in this pool were ob­ 60 positive clones, two were used to complete sequence analy­ tained with the S311Afl'315A mutant of MEK5 and one with sis. The 2828-base pair sequence contained an open reading the K195M mutant. All clones contained the TEY activation frame of 2445 nucleotides encoding 815 amino acids (Fig. 4). A motif. The product encoded by the longest ERK5 clone, AI, number of residues important for the kinase activity of ERKs are conserved, including the TEY activation motif (1). ERK5 interacted with MEK5 in the yeast two-hybrid system, but not with MEKI or MEK2 (Fig. 3, A-C). In addition, MEK5 does not contains a 400-amino acid C-terminal domain, which houses interact with p38 or Raf A, B, or C in the yeast two-hybrid two proline-rich regions. The first proline-rich region, consist­ system. This result supports the highly specific nature of the ing of 32 amino acids, contains 16 prolines. The second proline­ rich region (124 amino acids with 44 prolines) contained sev­ interactions between MEK5 and ERK5. It should be noted that wild-type MEK5 also interacted with ERK5, although we did eral small Pro-Ala repeats, (PA)3' followed by Pro-Thr repeats, not isolate any ERK5 clones from the initial yeast two-hybrid (PT)3' and three repeats of PPGP. The (PA)n repeat is present screen. We anticipate that dominant negative MEKs may be in myosin light chain kinase, and this sequence has been shown A New Human Protein Kinase Signal Transduction Pathway 12669 moval of the C-termina l domain resul ts in the ability of the fun ctional ERK5 kinase domain to displ ay its kinase activity. The fact that ERK5, like ERK1 , has a TEY motif an d that phospho-Ser/Thr and Tyr are observed in the ana lysis of la­ ~lWkb beled ERK5, suggests that the catalytic properties of the two kin ases may be similar. 214 - In summa ry, we have iden tified two novel kin ases that in­ ter act selectively with one another in what appears to repre­ - P-S _GST·ERK5 sent a new mammali an signa l transductio n pathway. Our re­ 111 - sults sugges t th at there are no interactions between the _ P·T protein s in the MEK1/ERK1 pathway an d the MEK5/ERK5 pa th way, suggesting that the two signaling pathways have 74 - _ GST·ERK5(1-431) - p.y eithe r different or complementary functions. Our res ults also suggest that th e two differen t signaling path ways most likely do not have overlapping functions in the cell. We are currently atte mpting to identify both upstream activators of MEK5 as 45 - well as downstream substra tes of ERK5. The unu sual length of the Codomain of ERK5 suggests that the sequenc es located FI G. 5.Au top h osp h oryla t ion of recombinant GST-ERK5 (1­ here may undergo additio na l post-tran slational modificat ions 431). A, autophosp hory lation of purified full-length GST-ERK5 and GST-ERK5 (1-431). The sa mples wer e reso lved on a SDS-PAGE gel as well as play roles in regulating or targe ting of ERK5. followed by autoradiography. Th e arrows indicate the position of the purified GST-ERK5 an d GST-ERK5 0-43 1). B , phos phoa mino acid Acknowledgments-We thank K.-L. Guan for plasmid s encoding ana lysis of autophos phorylated GST-ERK5 (1- 431 shown in A. P-S, MEK1, MEK2 and ERKI. We also th ank B. Yas har and R. Ston e for phosphoseri ne; poT, phosph oth reonine; P-Y, phosph otyrosin e. critical rea ding of the ma nu script. We also acknowledge th e help of J. Clemens, M. Wish art, and C. F. Zhen g with the yeast two-hybrid syste m. to directly interact with actin, targeting the kinase to a specific location in the cell (29). The unique ERK5 C-te rmina l seque nce REF E RE NCES may serve as a locali zation domain an d/or a regulatory domain. Cell 80, 179-185 1. Mar sh all, C. J . (1995) interesting to speculate th at th e (PA)" seque nces which It is 2. Sch lessinger , J ., and Ullr ich, A. (1992) Neuron 9, 383-39 1 3. Boulton, T. G., Nye, S. H., Robbin s, D. J ., Ip, Y. N., Radziejewsk a, E., target myosin light chain kinase to actin may also target ERK5 Morgenbe sser , S., Depinho , R., Pan ayotatos, N., Cobb, M. H., and to a simila r location. Northern blot ana lysis of ERK5 mRNA Yancopoulos, G. D. (1991) Cell 65, 663-675 (3.1 kb) and MEK5 mRNA (2.7 kb ) demonstrated they are 4. Sturgill, T. W., a nd Ray, L. B. (1986 ) Biochem . Biophys. Res. Commun. 134, 565-571 expressed in man y human tissu es and that th e RNAs are most 5. Wu, J ., Harri son, J . K., Vincent, L. A., Haystead , C., Harstead , T. A. J ., Michel, abunda nt in heart and skeletal mu scle (da ta not shown). H., Hunt, D. F., Lynch, K. R., and Sturgill, T. W. (1993 ) Proc. Na tl. Acad. To fur th er docum ent the specific interaction between MEK5 Sci. U. S. A. 90, 173-177 6. Zheng , C. F., a nd Guan , K. L. (1993 ) J . Bioi. Chem. 268, 11435-11439 and ERK5 , in vitro binding assays were conducted (Fig. 3D). In 7. Derijard, B., Hibi , M., Wu, I.-H. , Barrett, T., Su, B., Deng, T., Kari n, M., and this exper iment, full-length ERK5 was synthes ized in an in Davis, R. (1994 ) Cell 76, 1025-1 037 vitro transcription/t ranslation system. As shown in Fig. 3D, 8. Kyriakis, J . M., Ban erjee, P., Nikolak aki , E., Dai , T., Rubie, E. A., Ahmad , M. F., Avruch, J ., and Woodgett, J . R. (1994) Nature 369 , 156 -1 60 ERK5 binds to GST-MEK5 and the GST-MEK5 mutant 9. Han , J ., Lee, J . D., Bibbs, L., a nd Ulevitch, R. J . (1994) Science 265, 808- 811 K195M, bu t not to GST alone or to GST-MEKl. Thus, MEK5 10. Rouse, J. , Cohen , P., Trigon, S., Moran ge, M., Alonso-L1amazares, A., specifically interacted with ERK5 in both our in vitro syste m Zamanillo, D., Hunt, T., a nd Nebreda , A. R. (1994 ) Cell 78, 1027-1037 11. Lee J . C., Layton , J . T., McDonn el, P. C., Gall agher , T. F., Kum ar , S., Green, and in the yeast two-hybrid system. D., McNulty, D., Blum enth al, M. J ., Heys, J . R., Lahd vat ter, S. W., et 01. Protein Kin ase Activity of MEK5 and ERK5-GST fusi on (1994 ) Natu re 372, 739-746 protein containing full-length MEK5 could not be overex­ 12. Herskowitz, I. (1995) Cell 80, 187-197 13. Ammerer, G. (1994 ) Curro Opin . Genet. Dev. 4, 90-95 pressed in E. coli. However , elimination of the N-terminal 11 14. Errede, B., a nd Levin, D. E. (1993) CurroOpin . Cell Bioi. 5, 254- 260 res idues produced a MEK5 fusion protein which could be over­ 15. Sanc hez, I., Hughes, R. T., Mayer, 0. J ., Yee, K., Woodgett, J. R., Avru ch, J ., expressed an d affinity-purified. Th e purified protein showed no Kyri ak is, J. M., and Zon, L. I. (1994) Nature 372, 794 - 798 16. Derij ard , B., Rain gea ud, J ., Barrett, T., Wu, I.-H., Han, J ., Ulevit ch, R. J ., a nd protein kinase activity toward ERK1 , JNK1, or GST-ERK5. Science 267, 682--685 Davis, R. J . (1995 ) However, a weak autophosphorylation activity could be de­ 17. Lundblad , V. (1988) in Current Protocols in Molecular Biology (Aus ubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidm an , J . G., Smith, J . A., and tecte d with the wild-type enzyme but not with the MEK5 K195M St ru hl, K., eds ) pp. 13.7.1- 13.7.5, John Wiley & Sons, New York mutant (data not shown). The se experiments suggested that 18. Guan, K. L., and Dixon, J . E. (199 1) Anal. Biochem. 192, 262-267 additiona l modifications are requi red for full kinase activity. 19. Alessi, D. R., Saito, Y., Ca mpbell, D. G., Cohe n, P., Sith an andam , G., Rapp, V., Ashworth, A., Marshall, C. J ., a nd Cowley, S. (1994 ) EMRO J. 13, Two GST fusio n protein s of ERK5 were created for expres ­ 1610-1619 sion in E. coli, a full-length GST-ERK5 and a truncated GST­ 20. Zheng, C. F., a nd Guan, K. L. (1994) EMBO J . 13, 1123- 1131 ERK5 with the C-te rminal 384 residues removed. Thi s trun­ 21. Rossomando, A. J ., Dent , P., Sturgill, T. W., a nd Marshak, D. R. (1994 ) Mol. Cell. Bioi. 14, 1594 - 1602 cated ERK5 corr esponds to th e kinase domain of th e protein , 22. Miya moto, S., Ohya, Y., Ohsumi, Y., and Anr aku, Y. (1987) Gene (Arnst.) 54, i.e. GST-ERK5(1-431). Both fusion proteins wer e purified on 125-132 glutathione-aga rose affinity columns. The full-length GST­ 23. Chang, E. C., Ba rr, M., Wan g, Y., Jung, V., Xu, H. P., and Wigler, M. H. (1994) Cell 79, 131-141 ERK5 protein had no detectable protein kin ase activity (Fig. 5A 24. Zheng, Y., Ceri one, R., and Bender , A. (1994) J. BioI. Chern. 269, 2369-2372 and data not shown). The purified GST-ERK5(1-431) show ed 25. Manser, E., Leu ng, T., Salihuddin, H., Zhao, Z. S., and Lim, L. (1994) Nature basal autophosphorylation activi ty, as was evident from the 367 , 40- 46 26. Cowley, S., Paterson , H., Kemp, P., a nd Marshall , C. J . (1994) Cell 77, 841-852 radi oactive band of th e predicted molecular weight following 27. Seger R., Seger, D., Reszka, A. A., Mun ar , E. S., Eldar-Finkelm an , H., SDS-PAGE analysis (Fig. 5A). Phosphoamino acid analysis of Dobrowolska, G., J ensen , A. M., Ca mpbell, J . S., Fischer , E. H., and Krebs, the rad ioactive protein revealed phosphotyrosine, phospho­ E. G. (1994 ) J . BioI. Chern. 269, 25699 - 25709 28. Man sour, S. J ., Matten , W. T., Herm an n, A. S., Candia, J . M., Rang, S., threonine, and phospho serine (Fig. 5B). These resu lts suggest Fuka sawa, K., Vande Woude, G. F., and Ahn , N. G. (1994) Science 265, tha t th e kin ase domain of the full-length ERK5 is render ed 966 -970 29. Williamson, M. P. (1994) Biochem. J. 297, 249- 260 inactive by the presence of th e C-te rmina l 384 residues. Re-

Journal

Journal of Biological Chemistry – Unpaywall

Published: May 1, 1995

There are no references for this article.