UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The Mekl-Erk1 node : A place of convergence Charest, David Laurent 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-271188.pdf [ 18.47MB ]
Metadata
JSON: 831-1.0088745.json
JSON-LD: 831-1.0088745-ld.json
RDF/XML (Pretty): 831-1.0088745-rdf.xml
RDF/JSON: 831-1.0088745-rdf.json
Turtle: 831-1.0088745-turtle.txt
N-Triples: 831-1.0088745-rdf-ntriples.txt
Original Record: 831-1.0088745-source.json
Full Text
831-1.0088745-fulltext.txt
Citation
831-1.0088745.ris

Full Text

The Mekl-Erkl node: A place of convergence by DAVID LAURENT CHAREST B.Sc.(Honours), Simon Fraser University, 1987 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF DOCTOR OF PHILOSOPHY Department of Medicine We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA March 1998 © David Laurent Charest, 1998  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department  or by his  or her representatives.  It is  understood  that  copying or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department The University of British Columbia Vancouver, Canada  DE-6 (2/88)  ABSTRACT  At present, five mitogen-activated protein (MAP) kinase families have been identified in vertebrates. At least three of these MAP kinases have been shown to operate within highly conserved intracellular signalling modules composed of sequentially activating protein kinases. Cell surface receptors in response to extracellular stimuli invoke MAP kinase cascades to regulate gene expression within the nucleus. The best characterized MAP kinases are the extracellular signal-regulated kinases (Erks), the jun N-terminal kinases (Jnks) and the high osmolarity glycerol kinases (Hogs). Mitogenic signals that stimulate the G protein Ras activate the Raf->Mek->Erk->Rsk module. In turn, the Nik->Mekkl->Mkk4->Jnk and the Mkk3->Hog->MAPKAPK modules respond to environmental stress signals. Recent progress in MAP kinase signalling has revealed that MAPKs respond to a variety of environmental cues that promote cell growth, differentiation and death.  The experimental work described in this doctoral thesis was initiated six years ago as an independent study into the regulation of MAP kinase. At the time, purified MAP kinase was shown to be a prolyl-directed seryl/threonyl kinase that was regulated by phosphorylation on tyrosyl and threonyl residues. In addition, a partial cDNA sequence for a MAP kinase termed Erkl was shown to be homologous to Fus3 (for fusion), a kinase that regulated the mating pathway in Saccharomyces  cerevisiae.  Results from  genetic epistasis experiments revealed that Fus3 was positioned downstream in a cascade that involved several protein kinases. Consequently, we hypothesized that MAP kinase probably functioned as a key intermediary in the transmission of information from the plasma membrane to the nucleus.  ii  To understand the function M A P kinase plays in regulating the flow of information from plasma membrane to the transcriptional machinery in the nucleus, a cDNA clone of Erkl was isolated from a human liver library. The mRNA message encoded a 379-amino acid protein which was 12 residues longer (MAAAAAQGGGGG) than Erkl described previously from rat and, therefore, represented the full-length sequence. By using recombinant Erkl protein, a MAP kinase activator activity was purified from 1-MeAde-treated sea star oocytes. Immunoblotting with antipeptide antibodies directed against mouse Mekl (MAP/Erk kinase) and yeast Ste7 (Sterile 7) revealed that the sea star activator protein was a Mekl homologue. Unexpectedly, the purified activator was able to increase Erkl MBP phosphotransferase activity without directly phosphorylating the enzyme. This contrasts results obtained by other researcher groups that showed direct phosphorylation by the protein kinase activator Mekl leads to Erkl activation. Therefore, the sea star Mek-like protein may activate human Erkl by a different mechanism. The data presented here indicates that sea star Mek may activate human Erkl via allosteric modulation.  M A P kinase family subgroups Erk, Jnk and Hog are activated by phosphorylation on threonyl and tyrosyl residues that are located within a tripeptide ThrXaa-Tyr (where the intervening residue Xaa codes for Glu in Erkl, Gly in Hog and Pro in Jnk) regulatory motif. Phosphorylation of these sites in MAP kinases is performed by one of two specific pairs of MAP kinase kinases. Erkl and Erk2 are the only known substrates for Mekl and Mek2. A similar narrow specificity has been observed for Jnks and Hogs by their immediate activators Mkk4/Mkk7 and Mkk3/Mkk6, respectively.  A systematic analysis of the TEY tripeptide sequence in Erkl was undertaken to understand the importance each residue plays in the substrate specificity of mammalian Mekl.  Constitutively active mouse Mekl was used to phosphorylate a battery of  iii  threonyl and tyrosyl phosphorylation site mutants. Mutation of either Thr-202 or Tyr204 reduced the efficiency at which Mekl was able to phosphorylate the Erkl mutant proteins. In fact, serine substitution for Thr-202 was the only allele that retained MBP kinase activity. The intervening residue located between the regulatory threonyl and tyrosyl phosphorylation sites of MAP kinase may also serve as an important specificity determinant to prevent inappropriate cross-phosphorylation by MAP kinase kinases from parallel modules. The replacement of Erkl TEY motif with the Jnk (TPY) and Hog (TGY) regulatory sites also reduced Mekl specificity and markedly decreased Erk MBP phosphotransferase activity. Taken together, these results indicate that the TEY motif in Erkl is an important consensus sequence for Mekl recognition.  iv  T A B L E OF CONTENTS ABSTRACT  ii  T A B L E OF CONTENTS  v  NOMENCLATURE AND ABBREVIATIONS  xii  LIST OF TABLES  xv  LIST OF FIGURES  xvi  INTRODUCTION  1  1.  HISTORICAL PERSPECTIVE  1  2.  MAP KINASE SIGNALLING MODULE: DISCOVERY  3.  4.  OF A SIGNAL TRANSDUCTION PARADIGM  5  2.1  Lessons from studies in glycogen metabolism.  5  2.2  Dissection of a ubiquitous signalling pathway.  6  2.3  Lessons from Saccharomyces cerevisiae mating pathway.  11  REGULATION OF MAP KINASE  12  3.1  Identification of MAP kinases Erkl and Erk2.  12  3.2  Regulation of Erk phosphotransferase activity.  13  3.3  Receptor-mediated stimulation of MAP kinases Erkl and Erk2.  15  3.4 Substrates of Erk protein kinase. MAP KINASE ACTIVATORS 4.1  5.  6.  16 17  Identification of MAP kinase kinase isoforms Mekl and Mek2.  17  4.2  Mek protein kinases as regulators of MAP kinase activity.  19  4.3  Mek/Erk protein complexes and distribution within the cell.  20  REGULATION OF MEK PROTEIN KINASE  21  5.1  Stimulation of Mek protein kinase activity.  21  5.2  Mek/Erk signalling module.  22  ACTIVATION OF THE ERK PROTEIN KINASE MODULE  23  6.1  23  The Rafl nodal point.  7.  8.  9.  10.  11.  6.1.1 The Ras-Raf 1 regulated pathway.  24  6.1.2 cAMP-Raf 1 regulated pathway.  27  6.1.3 Rapl-RafB regulated pathway.  29  6.1.4 The Ksrl-Rafl regulated pathway.  31  PARALLEL MAP KINASE MODULES IN MAMMALS  35  7.1  35  MAP kinase superfamily of proline-directed kinases.  STRESS-ACTIVATED MAP KINASE SIGNALLING PATHWAY 8.1  Identification of Jun protein kinase in the regulation of stress signalling.  8.2 39  Stimulation of the Jun kinase pathway.  8.3  Effectors of Jnk protein kinase signalling.  36 36  40  REGULATION OF MKK4/MKK7 PROTEIN KINASES  43  9.1  43  Jun protein kinase activators.  ACTIVATION OF THE JNK PROTEIN KINASE MODULE  44  10.1 The Mek kinase (Mekk) node  44  10.2 The rnixed lineage kinase (Mlk) node.  46  10.3 Orphan MAPK kinase kinase nodes.  48  10.4 The Rho monomeric G-protein node.  49  10.5 The Ste20-related kinase node.  50  10.5.1  Pak regulation of the Jnk protein kinase module.  50  10.5.2  Gck regulation of the Jnk protein kinase module.  52  HIGH OSMOLARITY MAP KINASE SIGNALLING PATHWAYS  57  11.1 The Hog MAP protein kinase in Saccharomyces cerevisiae.  57  11.2 Identification of Hog protein kinase in the regulation of stress signalling in mammals. 11.3 Activation of Hog protein kinases. 11.4 Hog effector protein kinase signalling.  58 60 61 vi  12.  13.  11.4.1  Regulation of gene expression.  61  11.4.2  Regulation of MAPKAPK2 activity.  62  11.5 Hog protein kinase activators.  64  ORPHAN MAP KINASE SIGNALLING MODULES  67  12.1 Erk3 signalling module.  67  12.2 Erk5 signalling module.  68  REGULATION OF MAP KINASE IN MATURING SEA STAR OOCYTES  71  13.1 Identification of MAP kinase in sea star oocytes.  71  HYPOTHESIS  73  OBJECTIVES  74  MATERIALS AND METHODS  75  1.  MATERIALS  75  2.  CELL MANIPULATIONS  75  2.2  Oocyte isolation and cell culture.  75  2.2.1 Oocyte preparation.  75  2.2.1.1  Mechanical disruption.  75  2.2.1.2  1-Methyladenine injection.  82  2.2.2 Cell culture. 3.  82  BIOCHEMICAL TECHNIQUES  83  3.1  Determination of protein concentrations.  83  3.2  Column chromatography of sea star cytosolic extracts.  85  3.2.1 Anion exchange chromatography.  85  3.2.2 Cation exchange chromatography.  85  3.2.3 Gel filtration chromatography.  87  3.4  Purification of a sea star MAP kinase kinase-like protein.  87  3.5  Protein kinase assays.  93  vii  3.6  93  3.5.2 MAP kinase activator assay  94  3.5.3 Autophosphorylation.  97  3.5.3.1  Recombinant proteins.  97  3.5.3.2  Purified sea star activator.  97  Protein phosphatase assays.  98  3.6.1 Protein tyrosyl phosphatase assays.  98  3.6.2 Protein seryl/threonyl phosphatase assays.  98  Antibody production.  100  3.7.1 Antigen preparation.  100  3.7.2 Antibody purification and quantitation.  101  3.7.3 Other antibody sources.  101  3.8  One dimensional gel electrophoresis.  102  3.9  Protein visualization.  102  3.9.1 Coomassie blue staining.  102  3.9.2 Silver staining.  103  3.9.3 Western immunodetection.  104  3.7  4.  3.5.1 Single-step seryl/threonyl protein kinase reaction.  3.10 Phosphoamino acid analysis.  105  3.11 Two-dimensional phosphopeptide mapping.  106  MOLECULAR BIOLOGY TECHNIQUES  106  4.1  RNA isolation.  108  4.2  PCR amplification of a partial human Erkl cDNA.  109  4.3  Cloning and sequencing of a full-length human Erkl cDNA.  Ill  4.3.1 Preparation of plating bacteria.  Ill  4.3.2 in situ hybridization of bacteriophage X plaques.  Ill  4.3.2.1  Plating.  112  4.3.2.2  Immobilization of bacteriophage X.  112 viii  4.3.2.3  Fixation of bacteriophage DNA to nitrocellulose.  4.3.2.4  Hybridization of immobilized A, DNA with a P-labelled probe. 32  113  113  4.4  Sub-cloning human Erkl into the pGEX-2T vector.  115  4.5  Chromosomal assignment.  117  4.6  Primer extension.  117  4.7 PCR amplification of a murine Mekl cDNA.  119  4.8 Prokaryotic expression of recombinant GST-fusion proteins.  120  4.9  4.8.1 Protein expression in bacterial culture.  122  4.8.2 Preparation of bacterial cytoplasmic sonicates.  122  4.8.3 Purification of the fusion proteins.  123  4.8.4 Elution of glutathione bound GST-fusion proteins.  124  4.8.5 Thrombin cleavage of glutathione bound GSTfusion proteins.  124  Oligonucleotide-mediated mutagenesis.  125  4.9.1 Mutagenesis by the megaprimer method.  125  4.9.2 Mutagenesis by the double-primer method.  128  RESULTS  130  1.  130  ISOLATION OF A HUMAN ERK1 cDNA FROM HEPG2 CELLS  1.1  Amplification of a partial cDNA encoding a human MAP kinase.  130  1.2  Detection of a MAP kinase in human Hep G2 cells.  131  1.3  Cloning of a full-length cDNA encoding the human Erkl protein.  134  2.  CHARACTERIZATION OF RECOMBINANT ERK1 PROTEIN  2.1  Expression of human Erkl in E. coli as a recombinant  2.2 2.3  145  GST-fusion protein.  145  Characterization of autophosphorylated Erkl. MBP phosphotransferase activity of GST-Erkl.  148 161  ix  2.4  Activation of GST-Erkl by a MAP kinase kinase.  3.  PURIFICATION AND CHARACTERIZATION OF A SEA STAR MAP KINASE ACTIVATOR  164  169  3.1  Detection of MAP kinase activator during oocyte maturation.  169  3.2  Purification of MAP kinase activator.  176  3.3  Identification of the purified MAP kinase activator.  182  3.4  Mechanism of human Erkl activation by the sea star MAP kinase activator in vitro. MOLECULAR ANALYSIS OF THE TEY REGULATORY PHOSPHORYLATION SITES WITHIN THE LI2 ACTIVATION LIP  4.  185 192  4.1  Phosphorylation and activation of Erkl by activated Mekl (EE).  192  4.2  Phosphorylation and activation of Erkl bye activated (AN3EE).  197  4.3  Mutational analysis of Erkl regulatory phosphorylation sites.  202  4.4  Analysis of autophosphorylation and basal MBP phosphotransferase activities of Erkl regulatory phosphorylation site alleles.  205  4.5  Mekl (EE) phosphorylation and activation of Erkl regulatory phosphorylation site alleles.  211  4.6  Mekl (AN3EE) phosphorylation and activation of Erkl regulatory phosphorylation site alleles.  214  Mutational analysis of the intervening glutamic acid residue in the TEY of Erkl.  220  Analysis of autophosphorylation and basal MBP phosphotransferase activities of Erkl amino acid 203 mutant alleles.  220  Mekl (EE) phosphorylation and activation of Erkl amino acid 203 mutant alleles.  223  4.7 4.8  4.9 4.10  Mekl (AN3EE) phosphorylation and activation of Erkl amino acid 203 mutant alleles.  228  DISCUSSION  234  BIBLIOGRAPHY  251  APPENDICES  298  Appendix 1  Peptide conjugation to K L H  298  Appendix 2  Antigen preparation  299  Appendix 3  Antibody isolation, purification and quantitation  300  Appendix 4  Peptide affinity column preparation  302  Appendix 5  Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE)  303  Appendix 6  Silver staining  306  Appendix 7  Gel transfer and Western immunoblotting  307  Appendix 8  Phosphoamino acid analysis  309  Appendix 9  Two-dimensional phosphopeptide mapping  311  Appendix 10  RNA isolation  313  Appendix 11  Detection of DNA fragments by ethidium bromide staining of agarose gels  315  Appendix 12  Digestion of DNA with restriction enzymes  317  Appendix 13  DNA ligation into plasmid vectors  318  Appendix 14  Transformation of recombinant plasmid vectors into E. coli cells  320  Appendix 15  Preparation of 2 x YT and NZY bacterial medium  322  Appendix 16  DNA purification  323  Appendix 17  DNA sequencing  327  Appendix 18  Synthesis of radiolabeled probes  331  Appendix 19  Buffer solutions  333  Appendix 20  Construction of a constitutively active Mekl  335  xi  N O M E N C L A T U R E A N D  1.  °c Ci g h kDa mA mg ml M mmol mM min nM nm sec  ABBREVIATIONS  Measurements degrees Celsius Currie, 2.22 x 10 disintegrations per minute gram times the force of gravity hour kilodalton milliamp milligram mllilitre moles/litre millimol millimolar minute nanomole (10~9 mole). 12  nanometre (10~ metre) second micrometre (10~ metres) mircorgram (10~ gram) 9  6  u.m Lll  uM  6  microlitre (10" litre) micromole (10~ mole) 6  6  xii  2.  Amino acids  Name Alanine Arginine Asparagine Aspartate Cysteine Glutamate Glutamine Glycine Histidine Isoleucine Leucine Lysine Methionine" Phenylalanine Proline Serine Threonine Tryptophan Tyrosine Valine  Three letter abbreviation Ala Arg Asn Asp Cys Glu Gin Gly His Iso Leu Lys Met Phe Pro Ser Thr Trp Tyr Val  One letter symbol A R N D C E Q G H I L K M F P S T W Y V  Characteristics non-polar basic polar acidic polar acidic polar polar basic non-polar non-polar basic nonpolar non-polar non-polar polar polar basic (weak) polar (weak) non-polar  3  Nucleotides  Adenine Cytosine Guanine Thymine Deoxyadenosine 5'-triphosphate Deoxycytosine 5'-triphosphate Deoxyguanine 5'-triphosphate Deoxythymidine 5'-triphosphate  A C G T dATP dCTP dGTP dTTP  LIST OF T A B L E S  Table 1:  Research materials and their commercial sources.  76  Table 2:  Specificity of protease inhibitors.  81  Table 3:  Sources of kinases, phosphatases and proteases.  84  Table 4:  Chromatography resins used for protein analysis and purification.  86  Table 5:  Polyclonal antibodies.  99  Table 6:  DNA and RNA modifying enzymes.  107  Table 7:  Erkl sequencing oligonucleotide primers.  116  Table 8:  Mekl sequencing oligonucleotide primers.  121  Table 9:  Erkl and Mekl mutant oligonucleotide primers.  126  Table 10: Erkl activator purification summary.  181  xv  LIST OF FIGURES  Figure 1:  Regulation of Mek-MAP kinase signalling module by distinct biological processes in mammals and yeast mating in Saccharomyces  Figure 2: Figure 3:  cerevisiae.  33  Diverse proximal inputs regulate the Mkk4-Jnk stress signalling module.  55  Hog-dependent modules in osmosensing in Saccharomyces cerevisiae  and stress signalling in mammals.  65  Figure 4:  MAP kinase specific signalling modules in mammalian cells.  69  Figure 5:  Mono Q and Mono S column elution profiles.  88  Figure 6:  Purification of sea star MAP kinase activator.  91  Figure 7:  MAP kinase activator assay.  95  Figure 8:  Detection of MAP kinase activity in Hep G2 cells.  132  Figure 9:  Nucleotide and predicted amino acid sequence of human Erkl cDNA from HepG2 cells. Figure 10: Primer extension analysis of the 5'-terminal region of the human Erkl mRNA.  138  Figure 11: Sequence comparison of several MAP kinase isoforms.  141  Figure 12: pGEX-2T prokaryotic expression vector of the full-length human Erkl cDNA cloned in-frame with the glutathione-S-transferase (GST) protein.  146  Figure 13: Expression and immunoreactivity of recombinant Erkl protein from E. coli to MAP kinase antibodies.  149  Figure 14: Autophosphorylation and phosphoamino acid analysis of the Erkl recombinant protein.  152  Figure 15: Two-dimensional tryptic phosphopeptide mapping and phosphoamino acid analysis of autophosphorylated recombinant Erk 1.  156  135  Figure 16: Two-dimensional tryptic phosphopeptide mapping and phosphoamino acid analysis of p56' ^ phosphorylated c  recombinant Erkl. Figure 17: Phosphatase treatment of recombinant Erkl. Figure 18: MAP kinase kinase phosphorylation and activation of recombinant Erk 1.  159 162 166 xvi  Figure 19: Sequence comparison between Erkl cDNA from human and an Erk-like cDNA from sea star oocytes.  170  Figure 20: Detection of Erkl activator activity in 1-methyladenine-treated sea star oocytes.  173  Figure 21: Purification of Erkl activator from mature sea star oocytes.  177  Figure 22: Analysis of Erkl activator purification.  183  Figure 23: Gel filtration chromatography of the Erkl activator.  186  Figure 24: Sea star Erkl activator phosphorylation and activation ofGST-Erkl. Figure 25: Time course of Mekl (EE) phosphorylation, activation and phosphoamino acid analysis of wild type Erkl.  190 194  Figure 26: Time course of Mekl (AN3EE) phosphorylation, activation and phosphoamino acid analysis of wild type Erkl.  199  Figure 27: Sequence alignments of MAP kinase isoforms.  203  Figure 28: Expression, immunodetection and quantitation of recombinant Erkl proteins purified from bacteria. Figure 29: Mekl (EE) phosphorylation of Erkl regulatory phosphorylation site mutant alleles.  206 209  Figure 30 Mekl (EE) activation of Erkl regulatory phosphorylation site mutant alleles.  212  Figure 31: Mekl (AN3EE) phosphorylation of Erkl regulatory phosphorylation site mutant alleles.  215  Figure 32: Mekl (AN3EE) activation of Erkl regulatory phosphorylation site mutant alleles.  218  Figure 33: Expression, immunodetection and quantitation of recombinant Erkl protein purified from bacteria.  221  Figure 34: Mekl (EE) phosphorylation of Erkl intervening amino acid 203 mutant alleles.  224  Figure 35: Mekl (EE) activation of Erkl intervening amino acid 203 mutant alleles.  226  Figure 36: Mekl (AN3EE) phosphorylation of Erkl intervening amino acid 203 mutant alleles.  229  XVll  Figure 37: Mekl (AN3EE) activation of Erkl intervening amino acid 203 mutant alleles.  231  Figure 38: Nucleotide and predicted amino acid sequence of mouse Mekl cDNA from liver.  336  Figure 39: Prokaryotic expression vector of the full-length mouse Mekl protein cloned downstream of the glutathione-S-transferase (GST) gene.  339  Figure 40: Expression and immunoreactivity of recombinant Erkl protein from E. coli to MAP kinase antibodies.  341  Figure 41: Sequence alignments of several protein seryl/threonyl kinases.  344  Figure 42: Expression and analysis of a constitutively active Mekl allele.  346  xviii  INTRODUCTION  1.  HISTORICAL PERSPECTIVE  For more than twenty-five years, oncogenes have been the subject of intense investigation. The ability of these mutant genes to mediate unregulated proliferation resulting in tumour growth was, in part, the result of excessive protein phosphorylation within the cell (reviewed in Cantley et al, 1991). In support of this notion, some oncogene products possess intrinsic enzymatic activities that control the catalysis of protein phosphorylation (reviewed in Bishop, 1991). Previously, it was demonstrated that phosphorylation cascades were essential for regulating a variety of metabolic activities in the cell and it was later postulated that similar signalling relay systems may modulate many other biological activities (reviewed in Krebs, 1983). The premise that the cellular counterparts of oncogenes, the proto-oncogenes, were key players in intracellular communication systems has been supported by a large body of genetic and biochemical research.  The vast majority of oncogene kinases, whether retrovirally-induced or naturally occuring in mammals, were found to encode a class of enzymes that transfer the yphosphate moiety of ATP onto the hydroxyl group of tyrosyl residues (Hanks et al, 1988). This subfamily of kinases exist as membrane-spanning receptor or membraneassociated protein-tyrosyl kinases. The EGF-encoded receptor oncogene ErbB- and the v-Src-family kinases are two examples from the tyrosyl-directed family of protein kinases. Very few seryl/threonyl-directed protein kinases (for example Rafl, Mos, Piml, Akt for A K R mouse thymoma (alias PKB) and Tpl2 (for tumour progression locus 2 alias Cot for cancer Osaka thyroid) are known to be cancer-causing oncogenes. This is surprising given that more than 99.9% of phosphate ester-linkages in cells occurs on  serine and threonine (Cooper et al, 1981). It is believed that persistent elevated levels of protein tyrosyl phosphorylation has very strong oncogenic potential.  A possible  explanation for the large number of oncogene-encoded tyrosyl kinases may be their location high upstream within signal transduction pathways. By and large, these enzymes direct signalling messages to a variety of regulatory elements within different cascades that exist downstream of them. In addition, the complex architectural design of many of these enzymes, such as critical autoinhibitory domains for example, may make them more susceptible to transforming mutations. The fact that unrelated classes of oncogenes such as extracellular ligands, eg. Neu; membrane-spanning and -associated tyrosyl kinases, e.g. ErbB and Src family; guaninenucleotide-binding (GTP-binding) proteins, e.g. Ras family proteins located at the cell membrane; cytoplasmic seryl/threonyl kinases, e.g. Rafl and Tpl2; and transcription factors present in the nucleus, e.g. Jun and Fos that differed vastly in their structure and function were able to cause a similar transformation phenotype in a variety of immortalized cell types implied that they may activate similar signal transduction pathways. Several lines of evidence in different model systems indicate this to be the situation in the living cell. The relationship between each of these specialized signal transduction molecules has been the subject of intense research during the past two decades.  Some of our earliest insights into the relationship between these disparate cell signalling molecules as well as their position within cell signalling cascades occured from studies into the regulation of Ras activity. In quiescent cells, Ras was shown to exist in an inactive GDP-bound state. However, certain point mutations could maintain Ras in a constitutively active GTP-bound state. Microinjection of activated GTP-Ras protein caused the transient transformation of a number of mammalian cell lines and induced  germinal vesicle breakdown (GVBD) in Xenopus oocytes (Stacey et al, 1984; Feramisco et al, 1984; Birchmeier et al, 1985). Later, dominant inhibitory mutants of Ras were used to block growth in NIH 3T3 cells and differentiation in PC 12 and 3T3-L1 cell lines (Cai et al, 1990; Szeverenyi et al, 1990; Ogiso et al, 1990 and Benito et al, 1991). Furthermore, engagement of platelet-derived growth factor (PDGF) or epidermal growth factor (EGF) cell surface recepetors caused a rapid but transient increase in biologically active Ras-GTP complexes. Similarly, transformation of the activated forms of these receptors also elicited the same effect. These findings indicated that wild type Ras was a downstream transducer of these mitogens (Satoh et al, 1990; Satoh et al, 1990; Gibbs et al, 1990).  In a manner similar to the control of Ras by membrane receptors, the seryl/threonyl kinase Raf was shown to be regulated by similar growth factors in many different mammalian cells (reviewed in Rapp, 1991).  The ability of activated  recombinant v-Raf to transform microinjected NIH 3T3 cells supports a role for Raf as an important intracellular messenger (Smith et al, 1990; Kolch et al, 1991). Several lines of evidence indicated that Raf acts downstream of Ras in growth factor signal transduction. The constitutively active form of Raf was able to reverse growth-factor arrest caused by overexpression of dominant negative Ras or microinjection of neutralizing Ras antibodies (Smith et al, 1986; Huleihel et al, 1986). Also, transfection of the kinase-inactive Raf 1 as well as expression of the anti-sense RNA or N-terminal regulatory domain of this enzyme was sufficient to block cells transformed with activated Ras or ligand-activated receptors (Heideker et al, 1990; Kolch et al; Bruder et al, 1992).  One of the first consequences of growth factor signal transduction in cells is the rapid increase in transcription of early-response genes like c-Fos, which was first  identified through its oncogenic counterpart v-Fos (Curran et al, 1983; Greenberg and Ziff, 1994). In these studies, constitutively active Rafl was shown to be capable of increasing c-Fos promotor transactivation. This clearly positioned Raf as an intermediary regulator since it was shown to act downstream of growth factor receptors and upstream of Fos gene expression (Jamal and Ziff, 1990).  Taken together these results placed several receptor tyrosine kinases (eg. PDGF and EGF) and their ligands, intracellular transducers (eg. Ras, Raf) and nuclear effectors (eg. Fos) as well as their oncogenic homologues (eg. v-Sis, v-ErbB, v-Ras, v-Raf and v-Fos) within the same signalling system. Until recently, little was known about the intervening steps following the engagement of receptor tyrosine kinases by their ligands and the ensuing mechanisms that activated cytoplasmic signal transducers and the consequent elevation of gene expression in the nucleus.  For those researchers seeking to understand the regulatory mechanisms that operate within the cell at a molecular level, several outstanding questions remained to be answered concerning each step in the signalling process from the cell membrane to the nucleus. What regulated the switch from tyrosyl phosphorylation at the cell membrane to seryl/threonyl phosphorylation within the cytoplasm? How does ligand-induced receptor stimulation activate transcription factors in the nucleus to increase gene expression? Using the combinatory approach of biochemical reconstitution and genetic disruption new signalling proteins were positioned within signalling pathways. In situations where genetic manipulations were not amenable in biochemical models or biochemical assays too difficult to develop for genetic systems, molecular biology provided the necessary tools to advance the research. Isolation of new genes by the yeast two-hybrid analysis or site-directed mutagenesis of known genes using recombinant DNA technology enabled biologists to uncover key molecular interactions and determine essential regulatory sites  within the individual signalling proteins. It has become apparent from the diverse research efforts of geneticists, biologists and biochemists involved in the study of such diverse organisms as yeast, flies, nematodes and mammals that the cell has the molecular equivalent to electronic circuitry or a cellular intranet (Egan and Weinberg, 1993; Pelech, 1996).  It has become evident from research conducted during the past few years that the prototypical MAP kinase defined by the extracellular-signal regulated kinases 1 and 2 (Erkl and Erk2 ) isoforms operate within a specific cell signalling pathways. In addition, Erks were shown to be regulated by over 50 different stimili in a variety of model systems (Pelech and Charest, 1995). More recently, several other MAP kinase family homologues have been discovered that function within distinct, non-overlapping signal transduction pathways (Pelech and Charest, 1995). In fact, as will be discussed later, the level at which MAP kinases operate within the signal transduction pathways (i.e. intermediary between the cell membrane and nucleus) means that these enzymes may act as critical convergent points or nodes through which the flow of information must traverse. The remainder of this review will discuss the MAP kinase superfamily  2.  MAP KINASE SIGNALLING MODULE: DISCOVERY OF A SIGNAL TRANSDUCTION PARADIGM.  2.1  Lessons from studies in glycogen metabolism.  Cell biologists have been fascinated with the complexity and regulation of important biological processes such as cell growth, division, differentiation and death in response to specific stimuli. To uncover the molecular events that control essential cellular activities it has been necessary to define ways to assay changes observed in  response to external cues. Alterations in cellular protein phophorylation in response to hormone treatment was one such approach, that in the middle of the 1950's, was demonstrated to be essential for regulating glycogen metabolism (reviewed in Krebs, 1983; Cohen, 1983). Phosphorylase is a key regulatory enzyme required for glycogen breakdown. The interconversion of phosphorylase between active and inactive states is a dynamic process involving enzyme phosphorylation-dephosphorylation reactions. Several groups exploited this hormone regulated phosphorylation event to successfully detect and purify the phosphorylase regulatory factor. After many years of experimental investigation, the first protein phosphorylation cascade activated as a consequence of hormone-dependent stimulation was shown to cause an increase in glycogen breakdown This important final metabolic step in glycogen metabolism was the starting point to begin working backward in a step wise fashion toward the epinephrine receptor located at the cell surface. Both Drs. Edwin Krebs and Edmond Fisher received the Nobel Prize in Medicine in 1992 for their pioneering work in the area of protein phosphorylation research.  2.2  Dissection of a ubiquitous signalling pathway.  Insulin- and growth factor-dependent stimulation also cause dynamic changes in the phosphate content of many cellular proteins in a manner similar to that observed during studies into glycogenolysis (reviewed in Mailer et al, 1990; Pelech et al, 1990). One consequence of cell exposure to polypeptide ligands such as insulin is a robust increase in protein synthesis (Krieg et al, 1988). Phosphorylation of multiple seryl residues within the 36-kDa S6 protein of the 40S ribosome subunits was reported to be an essential regulatory step for modulating translation of specific mRNAs (Gressner and Wood, 1974; Halselbacher et al, 1979; Bommer et al, 1980; Kozma et al, 1989). Subsequently, phosphorylation of 40S ribosomes in vitro identified S6 kinase activities in  cytosolic extracts obtained from serum- and ligand-induced activation of mammalian cells and progesterone-induced Xenopus oocyte maturation (Rosen et al, 1981; NovakHofer and Thomas; 1985; Tabarini et al, 1985; Erikson and Mailer, 1986; Cobb, 1986; Blenis et al, 1987; Pelech and Krebs, 1987). Subsequent purification and cloning experiments identified two distinct classes of S6 kinases. The 85- to 92-kDa ribosomal S6 kinase (Rsk) family isolated from Xenopus was demonstrated to be structurally unique from the 70-kDa S6 kinase isolated from mammals (Jones et al, 1988; Bannerjee et al, 1990). It is now widely accepted that the 70-kDa S6 kinase and not the Rsk kinases is responsible for modulating protein synthesis within cells (Chung et al, 1992).  During this same period, several research groups initiated studies into the cellular regulation of insulin and growth factor signalling in mammalian cells as well as GVBD in maturing Xenopus and echinoderm oocytes (Sturgill and Ray, 1986; Ray and Sturgill, 1987; Cicirelli et al, 1988; Pelech et al, 1988; Hoshi et al, 1998). A protein kinase activity, distinct from the S6 protein kinase described previously, was uncovered in insulin-treated murine fibroblast cells that phosphorylated the exogenous substrate mjcrotubule-associated protein 2 (MAP2) (Ray and Sturgill, 1987; Hoshi et al, 1988). Pelech and co-workers also described an enzyme with a similar molecular mass that was stimulated to phosphorylate the substrate myelin basic protein (MBP) in vitro in response to hormone-induced GVBD in immature frog and sea star oocytes (Cicirelli et al, 1988; Pelech et al, 1988). This novel enzyme termed MAP kinase (for mitogen-activated protein kinase) was shown to be identical to pp42, a protein that becomes transiently phosphorylated on tyrosyl residues in response to a variety of mitogens, including EGF, PDGF,  12-0-tetradecanoylphorbol-13-acetate  (TPA),  thrombin,  insulin  and  oncogenically transformed cells (Rossomando et al, 1989 and references 1-10 therein).  The fact that MAP kinase and S6 kinase were acutely regulated by insulin and that the time course of activation for MAP kinase preceded that of S6 kinase in 3T3-L1 cells led Drs. Sturgill and Mailer and their colleagues to investigate the relationship between the two enzymes. The S6 kinase from Xenopus oocytes was inactivated in vitro by incubation with seryl/threonyl-specific phosphatases 1 and 2A. The dephospho-S6 kinase phosphotransferase activity was partially restored in vitro after preincubating the enzyme with insulin-stimulated MAP kinase that was partially purified from Xenopus oocytes (Sturgill et al, 1988). Support for the notion of an insulin- and growth factorregulated cascade involving MAP kinase and S6 kinase in the same pathway was demonstrated by other researcher groups using partially purified enzymes and reconstitution experiments (add-back) from fractionated cell extracts (Gregory et al, 1989; Ahn and Krebs, 1990; Kyriakis and Avruch, 1990; Chung et al, 1991a; Chung et al, 1991b). However, MAP kinase was unable to activate the 70-kDa S6 kinase in similar experiments and therefore the two proteins were recognized to operate within distinct signalling pathways (Ballou et al, 1991; Mukhopadhyay et al, 1992). These MAP kinase enzymes were later termed Erkl and Erk2.  Erk protein kinases are regulated in mammalian cells by many growth factors and their cognate receptor tyrosyl kinases (Pelech et al, 1990; Cobb et al, 1991). Activation of the monomeric guanine nucleotide-binding protein Ras is accomplished by engaging many of these same receptors (Downward et al, 1990; Gibbs et al, 1990; Satoh et al, 1990; Burgering et al, 1991). Leevers and Marshall (1992) were the first researchers to demonstrate that Ras and Erk kinases lie on the same signalling pathway. In the absence of growth factor activation, constitutively active Ras (Val-12) introduced into Swiss 3T3 cells by scrape loading activated MAP kinase within minutes. Furthermore, in a NIH3T3 cell line overexpressing the insulin receptor, a dominant interfering Ras (Asn-17) blocked insulin-induced activation of Erk2 without affecting phosphatidylinositol-3-  kinase (PI3-kinase) activity in these same cells (de Vries-Smits et al, 1992). In Rat-1 cells, the activity was inhibited by the same Ras mutant after PDGF treatment but remained active after exposure to phorbol ester (Leever and Marshall, 1992; de VriesSmits etal, 1992).  Concurrently, experiments using Swiss 3T3 and NIH-3T3 fibroblastic cells overexpressing activated forms of Raf 1 led to the activation of Erkl and Erk2 in a Rasand receptor tyrosyl kinase-independent manner (Howe et al, 1992; Kyriakis et al, 1992; Dent et al, 1992). As further support, Erk2 was stimulated by TP A and EGF in a COS-1 cell line overexpressing full-length Rafl, while Rafl and Ras were shown to cooperate in the activation of Erkl in coexpression experiments in baculovirus Sf9 insect cells (Howe et al, 1992; Williams et al, 1993). MAP kinases are also directly activated by regulatory kinases, since isoforms of this enzyme family isolated as recombinant proteins or present in unstimulated cells are activated with partially purified fractions from NGF- TPA- or EGF-stimulated cells (Gomez and Cohen, 1991; Ahn et al, 1991; Adams and Parker, 1991). MAP kinase kinase isoforms Mekl and Mek2 (MAPK Erkl kinase) in turn are regulated by direct phorphorylation, since they become inactivated upon treatment with the protein-seryl/threonyl phosphatase PP2A (Gomez and Cohen, 1991; Matsuda et al, 1992; Crew and Erikson, 1992).  Erkl M B P phosphotransferase  activity and  consequently that of dephosphorylated Meks was restored in vitro  by direct  phosphorylation and activation with immunoprecipitated Rafl from transfected cell lines (Howe et al, 1992; Kyriakis et al, 1992; Dent et al, 1992). Reconstitution experiments involving prokaryotic and eukaryotic expressed proteins supported the hypothesis that Raf-Mek-Erk operate within the same signal transduction pathway (MacDonald et al, 1993).  Induction of neurite-like processes in rat adrenal pheochromocytoma PC 12 cells in response to nerve growth factor (NGF) treatment has been shown in separate experiments to coincide with the stimulation of Ras, Raf, MAP kinase and Rsk activities (Hagag et al, 1986; Blenis and Erickson, 1986; Gotoh et al, 1990; Boulton et al, 1991; Ohmichi et al, 1992; Scimeca, et al, 1992). Expression of dominant inhibitory mutant of Ras (Asp-17) blocked NGF-induced tyrosyl phosphorylation of Erkl and Erk2, and also inhibited Raf 1 hyperphosphorylation (previously used as a marker of Rafl activation) and Rsk activation (Thomas et al, 1992; Wood et al, 1992). The MAP kinase activator activity detected in PC 12 cells was also attenuated by the Ras inactive-mutant (Gomez and Cohen, 1991; Robbins et al, 1992). Catalytically compromised mutants of Rafl, Mekl, and MAP kinase isoform, Erk2, also interfered with the neurological differentiation of PC12 cells (Wood and Roberts, 1993). However, overexpression of constitutively active truncated form of Rafl in PC12 cells did not significantly stimulate MAP or Rsk kinase activities, and neither was Rafl able to activate MAP kinase directly during in vitro reconstitution experiments (Wood et al, 1992; Jaiswal et al, 1994). These results implied that although Ras activates all three enzymes in an NGF-dependent fashion, Raf1 and MAP/Rsk kinases may lie on separate signalling pathways. In light of these results, Jaiswal et al, (1994 and 1996) later revealed that RafB is the principal mediator for Ras activation of the MAP kinase pathway in PC 12 cells.  In Xenopus oocytes, microinjection of oncogenic Ras (Val-12) induced maturation that was slightly delayed from what is observed after progesterone treatment (Birchmeier et al, 1985). As expected, activated Ras protein is able to stimulate quiescent oocytes to undergo GVBD and hence circumvent the need for receptor stimulation by triggering the activation of a protein termed REKS (for Ras p21-dependent Erk-kinase stimulator), along with M k k l (MAP kinase kinase 1) , Erk2, and Rsk (Pomerance et al, 1992; Shibuya et al, 1992; Hattori et al, 1992; Itoh et al, 1993). REKS is a multimeric  protein-complex in which one of the proteins, a 98-kDa protein kinase was shown to be immunologically distinct from Rafl, Mos and Stell (Kuroda et al, 1995). Partial purification of an identical complex from bovine brain cytosol revealed that the active kinase component is the 95-kDa brain-specific RafB (Yamamori et al, 1995).  2.3  Lessons from Saccharomyces cerevisiae mating pathway  The first MAP kinase signal transduction pathway was elucidated from genetic epistasis experiments performed with budding yeast. At present, five distinct MAP kinase signal transduction pathways have been identified in S. cerevisiae (Herskowitz, 1995; Pelech and Charest, 1995). The conservation of the basic MAP kinase signalling module (consisting of a MAPKKK, MAPKK and MAPK) is a testament of its importance in transducing messages in response to environmental cues (Errede and Leven, 1993; Neiman,1993). In yeast, highly related MAP protein kinase modules regulate such disparate biological activities as mating, growth, differentiation and homeostasis (Herskowitz, 1995; Pelech and Charest, 1995).  The M A P kinase pathway was first elucidated from genetic studies into the regulation of mating in haploid yeast (Herskovitz, 1995). Mating between two yeast strains through conjugation initiates cell cycle exit prior to D N A replication. Engagement of unique seven transmembrane receptors (Ste2 and Ste3 for sterile) with the specific mating pheromone ligand of the opposite cell type (MAToc-factor for Ste2 and MATa-factor for Ste3) initiates this process (Jenness et al, 1983; Burkholder and Hartwell, 1985; Nakayama et al, 1985; Hagen et al, 1986). Binding of these small oligopeptide pheromones (a-factor is a 12 residue lipoprotein; a-factor is a 13 residue peptide) leads to the induction of a cascade of biological events that increases matingspecific functions including GI cell cycle arrest before Start (the point at which a cell  commits itself to a further round of mitosis) and changes associated with the conjugation process (Bender and Sprague, 1986; Nakayama et al, 1987; Sprague, 1991). The pheromones a and a interaction with Ste3 and Ste2 promotes the exchange of GDP for GTP on the heterotrimeric guanylyl nucleotide-binding protein G a-subunit (Gpal) and subsequent dissociation of G(3- (Ste4) and Gy- (Stel8) subunits from the receptor (Nakafuku et al, 1987; Jang et al, 1988; Whiteway et al, 1989; Blumer and Thorner, 1990). The liberated G|3-subunit, in turn, forms a complexe with Ste20 and activates the kinase by a poorly defined mechanism (Wu et al, 1995; Leberer et al, 1997). At present, Ste20 appears to be the most proximal kinase to the membrane receptor within the mating-pheromone-regulated MAP kinase pathway. Although the biochemical evidence is sketchy, genetic epistasis studies support a role for Ste20 in the sequential activation of the Stell, Ste7 and the MAP kinases Fus3 and Kssl within a non-branching pathway (Herskovitz, 1995). The non-kinase protein Ste5 may function to maintain fidelity within the module by sequestering Ste4, Stell, Ste7, Fus3 and Kssl into a large signalling,particle (Choi et al, 1994; Marcus et al, 1994; Printen and Sprague, 1994). Another possible function for this platform protein may be to localize the MAP kinase module and more specifically Stell in the vicinity of Ste20 at the cell surface.  3.  REGULATION OF MAP KINASE  3.1  Identification of MAP kinases Erkl and Erk2.  The first purification of a MAP kinase to homogeneity was achieved from sea star oocytes induced to undergo GVBD (Sanghera et al, 1990). Subsequently, the MAP kinase isoform Erkl was purified from several mammalian sources (Boulton, et al, 1991; Northwood et al, 1991). The Erk2 isoform first detected by Ray and Sturgill (1988) in mouse 3T3 cells was purified from mammalian cell culture and mature Xenopus oocytes  (Gotoh et al, 1991; Payne et al, 1991; Barrett et al, 1992). All three kinases display apparent relative molecular masses of 42-44 kDa by gel filtration and sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) in addition to exhibiting similar substrate preference for MAP2 or MBP proteins (Rossomondo et al, 1991). Subsequent isolation of Erk cDNAs revealed a high degree of sequence homology. The predicted amino acid sequences for rat and human Erkl and Erk2 are 85-90% identical (Boulton et al, 1991; Owaki et al, 1992; Gonzalez et al, 1992; Charest et al., 1993). There is also a high degree of conservation across species. Human full-length Erkl is 96% identical to rat Erkl (Marquardt and Stabel, 1992; Owaki et al, 1992; Charest et al, 1993), while the deduced protein sequence of Erk2 from human is 95% identical to Xenopus  Erk2 (aliases Mpkl and Xp42) (Boulton et al, 1991; Posada et al, 1991; Gotoh  et al, 1991). Representatives of the Erk MAP kinase subfamily have been isolated from all four eukaryotic kingdoms (Ferrell, 1996 and references therein).  3.2 Regulation of Erk phosphotransferase activity.  Erk MAP kinases are members of the seryl/threonyl family of protein kinases while themselves being subject to phosphorylation on tyrosine and threonine (Hanks et al, 1988; Rossomondo et al, 1989 and references therein; Anderson et al, 1990; Rossomondo et al, 1991). In cells labeled in vivo with P , Erkl and Erk2 become 32  phosphorylated on tyrosyl, seryl and threonyl residues (Ray and Sturgill, 1988; Robbins and Cobb, 1992). To maintain full enzymatic activity, however, Erkl and Erk2 appear to require phosphorylation on threonyl and tyrosyl residues.  Removal of either of these  phosphoresidues with seryl/threonyl or tyrosyl specific phosphatases completely inactivates these enzymes (Sturgill et al, 1988; Anderson et al, 1990; Ahn et al, 1991; Rossomondo et al, 1991; Pollack et al, 1991). Payne et al (1991) identified the regulatory phosphorylation sites from P-labeled Erk2. Active Erk2 purified from EL4 32  mouse T cell line after stimulation with phorbol dibutyrate (PDB) and subsequently proteolytically treated with trypsin yielded a single phosphopeptide containing phosphorylated threonine and tyrosine. The phosphorylated residues mapped to two neighbouring sites, namely Thr-183 and Tyr-185 (equivalent to Thr-202 and Tyr-204 in Erkl), at positions within kinase subdomains VII and VIII (Hanks et al, 1988; Payne et al, 1991). It has become clear from 3D structural analysis of several seryl/threonyl protein kinases that phosphorylation of residues between subdomains VII and VIII can induce the necessary global conformational changes to increase enzyme activity (Johnson et al., 1996; Canagarajah et al., 1997). The characteristic phosphothreonine and phosphotyrosine motif is separated by an intervening amino acid T X Y (where X is Glu184 in Erk2), which is a hallmark of the superfamily of MAP kinases (Ferrell, 1996). The level of seryl phosphorylation appears to remain unaffected following growth factor stimulation (Robbins and Cobb, 1992).  Erks obtained from rat 1 HTRcB cells autophosphorylate on tyrosyl residues in vitro  (Seger et al, 1991; Scimeca et al, 1991). These same kinases also display tyrosyl  phosphorylation in situ when expressed in Escherichia coli as well as in vitro after their purification as fusion proteins (Seger et al, 1991; Crews et al, 1991; Wu et al, 1991; Robbins et al, 1993; Charest et al, 1993). Since bacteria possess no intrinsic tyrosyl kinase activity, the Erk enzymes probably accumulate tyrosine by an autophosphorylation reaction (Wu et al, 1991; Seger et al, 1991; Charest et al, 1993). This basal autophosphorylating activity on tyrosyl is concentration independent and thus occurs by an intramolecular reaction (Wu et al, 1991; Robbins et al, 1993). The small but detectable increase in tyrosyl autophosphorylation consequently leads to a nominal stimulation in Erk phosphotransferase activity toward exogenous substrates including MBP (Seger et al, 1991; Crews et al, 1991; Wu et al, 1991; Rossomondo et al, 1992; Robbins et al, 1993; Charest et al, 1993). Tyr-185 in Erk2 and Tyr-204 in Erkl are the  sites of autophosphorylation (Rossomondo et al, 1992b, Charest et al, 1993). Only minor amounts of Thr-183 were detected from fusion proteins expressed in bacterial cells or during in vitro experiments and may explain the observed Erk activation. This selfcatalyzed phosphate incorporation prompted several research groups to speculate that, in a manner similar to PKA, (cAMP-dependent protein kinase) autophosphorylation may play a significant role in Erk regulation (Knighton et al, 1991; Seger et al, 1991; Crews et al, 1991; Wu et al, 1991; Robbins and Cobb, 1992). However, catalytically compromised Erk2 when expressed in Xenopus oocytes or CCL39 hamster cells becomes phosphorylated at the regulatory phosphorylation sites after stimulation (Posada and Cooper, 1992; Her et al, 1993). This supported earlier observations that Erks could be activated by a kinase activator without the need for autophosphorylation (Ahn et al, 1991; Gomez and Cohen, 1991; L'Allemain et al, 1992; Rossomondo et al, 1992a; Alessandrini etal, 1992).  3.3  Receptor-mediated stimulation of MAP kinases Erkl and Erk2.  The ubiquitous distribution of the MAP kinase family proteins Erkl and Erk2 may explain why these enzymes are activated by a bewildering array of extracellular stimuli (Pelech et al, 1990; Boulton et al, 1991; Cobb et al, 1991; Megan and Cobb, 1997). Besides mitogens such as PDGF and NGF which activate tyrosyl specific cell surface receptors, the MAP kinase module is stimulated by a variety of other ligands. On the one hand, these include peptides (eg. angiotensin and bradykinin) that activate seven transmembrane spanning (7TM) receptors that are coupled to heterotrimeric G proteins. Interleukins, on the other hand, interact with their cognate single transmembrane (1TM) receptors which stimulate the activities of the Src (sarcoma) and Jak (Janus kinase) family of receptor-associated tyrosyl kinases. Steroid hormones (e.g. estradiol) which  have high affinity nuclear receptors also increase MAP kinase activity by a pathway that remains poorly understood (Migliaccio et al, 1996).  3.4  Substrates of the Erk protein kinase. '  The microtubule-associated protein 2 and myelin basic protein kinases were the substrates first utilized to detect activation of MAP kinases (Ray et al, 1987; Hoshi et al, 1988; Cicicelli et al, 1988). Thr-97 in MBP was identified as the phosphorylation site (Sanghera et al, 1990b; Erickson et al, 1990). Synthetic peptides patterned after this site were used to determine the optimal substrate recognition sequence for the Erk family to be Pro-(basic/neutral)-(Ser/Thr)-Pro motif (Clark-Lewis et al, 1991). This result was later confirmed using a random peptide expression library (Songyang et al, 1996). A large number of proteins have been identified as in vitro substrates for Erkl and Erk2 including metabolic enzymes acetyl-CoA carboxylase (Pelech et al, 1991), tyrosine hydroxylase (Haycock et al, 1992) and phospholipase A2 (Lin et al, 1993; Nemehoff et al, 1993; Sa et al, 1995); structural proteins caldesmon (Childs et al, 1992; Adam and Hathaway, 1993) and tau (Drewes et al, 1992; Ledesma et al, 1992; Goedert et al, 1992) ; signalling proteins EGF receptor (Northwood et al, 1991; Theroux et al., 1992), the guanine nucleotide exchange protein Sos (for son-of-sevenless) (Cherniack et al., 1995; Waters et al, 1995; Rozakis-Adkcock et al, 1995; Laszlo et al., 1995; Porfir and McCormick, 1996; Corbalan-Garcia et al, 1996; Klarlund et al, 1996), Rafl (Anderson et al, 1991; Lee et al, 1992) Mekl (Saito et al, 1994), Rksl-3 (Zhao et al, 1996) and Mnkl-2 (Waskiesicz et al, 1997; Fukugana and Hunter, 1997); transcription factors Tall (Cheng et al, 1993) and Elkl (Gille et al, 1992; Janknecht et al, 1993; Marais et al, 1993) ; and finally the nuclear estradiol steroid hormone receptor (Kato et al, 1995; Bunone et al, 1996).  4.  MAP KINASE ACTIVATORS  4.1  Identification of MAP kinase kinase isoforms Mekl and Mek2.  A number of groups studying M A P kinase signal transduction worked backwards using recombinant Erk proteins as activatable substrates to purify MAP kinase kinases from a variety of model systems including Xenopus and sea star oocytes induced to undergo GVBD (Matsuda et al, 1992; Charest et al, unpublished data), insulin-treated rabbit skeletal muscle (Nakielny et al, 1992a; Wu et al, 1992), as well as from phorbol ester- and EGF-stimulated human and murine cell lines (Seger et al, 1992a; Shirakabe et al, 1992; Crews and Erikson, 1992). Meks exhibit apparent molecular masses of between 45-46-kDa as observed on SDS-PAGE. Partial protein sequence of Mek revealed that it contains sequence homology to MAP kinase kinases Ste7 (Sterile 7) and Byrl (Bypass of Rasl), two kinases that are regulated by pheromone-dependent MAP kinase signalling pathways in Saccharomyces pombe  cerevisiae  and  Schizosacchraomyces  (Kosako et al, 1992; Nakielny et al, 1992b; Wu et al, 1992; Crew and Erikson,  1992). The purified Mek proteins possess identical functional and immunological properties as those MAP kinase activators described in previous in vitro studies (Ahn et al, 1991; Gomez and Cohen 1991; Matsuda et al, 1992; L'Allemain et al, 1992; Rossomondo et al, 1992; Alessandrini etal, 1992).  Several Mek cDNAs were obtained by using the polymerase chain reaction (PCR) technique and degenerate oligonucleotides based on peptide sequences obtained from purified proteins (Crew et al, 1992; Ashworth et al, 1992; Seger et al, 1992b; Kosako et al, 1993; Wu et al, 1993a; Dbring et al, 1993; Zheng and Guan, 1993; Yashar et al, 1993). Mammalian Mekl is a 393 amino acid long polypeptide displaying an approximate molecular mass of 43-kDa (Crew et al, 1992; Ashworth et al, 1993;  Seger et al, 1992b; Wu et al, 1992a; Doring et al, 1993; Zheng et al, 1993). Sequence comparisons between Mekl isolated from murine pre-B cells and human T cells revealed homologies of 97% and 91%, respectively, with the rat brain isoform. These same mammalian proteins are 30-40% homologous with yeast Ste7 and Byrl (Crews et al, 1992; Seger et al, 1992b; Doring et al, 1993). Recombinant Mekl purified as a GST (glutathione-S-transferase) fusion protein phosphorylates Erkl in vitro while its overexpression in COS cells causes a 3-fold greater activity after TPA treatment than non-transfected cells (Crew et al, 1992; Seger et al, 1992b).  Dr. Edwin Kreb's research team also obtained a second clone with a coding region identical to Mekl except for a 78 nucleotide deletion located at position 471-584 in subdomain V (equivalent to 26 amino acids) (Seger et al, 1992b). This alternative mRNA transcript which exists in A431 cells might therefore explain the presence of 45and 46-kDa Mek isoforms described at the protein level in these same cells (Seger et al, 1992a). What unique role this 45-kDa isoform performs in the cells awaits further investigation. Zheng and Guan (1993) isolated a second M A P kinase kinase family member, termed Mek2, from several human libraries. Human Mekl and Mek2 are 80% identical at the amino acid level. Bacterially expressed GST-Mek2 incubated with serum-treated NIH-3T3 cell extract phosphorylated and activated recombinant human Erkl in vitro (Zheng and Guan, 1993). Mek2 has also been cloned from rat (Otsu et al, 1993; Largaespada et al, 1993; Wu et al, 1993b).  A 45-kDa MAP kinase activator similar to Mekl (alias M A P K K in frog) was isolated from Xenopus and was shown to be expressed predominantly in ovarian tissue (Kosako et al, 1993).  Two other Meks distinct from Mekl described during  progesterone-induced oocyte maturation were cloned from frog (Yashar et al, 1993). Mek2 is also expressed early in the developing central nervous system including brain,  spinal neurons and the eye (Yashar et al, 1993). A developmentally-regulated MAP kinase with a similar expression pattern has been observed in tissue from the nervous system and, therefore, may define a new MAP kinase signal transduction pathway (Zaitsevskaya and Cooper, 1993). A third distinct MAP kinase kinase, Mkk3, (alias Mek3) mRNA transcript was shown to be expressed in the later stages of the developing frog zygote (Yashar et al, 1993).  4.2  Mek protein kinases as regulators of MAP kinase activity.  Mekl and Mek2 are the only known physiological activators described for Erkl and Erk2. These same enzymes are able to phosphorylate catalytically-compromised Erkl and Erk2 mutants. However, active Mek2 displays a higher phosphotransferase activity toward both Erk isoforms than does Mekl (Zheng and Guan, 1993). Furthermore, Mekl and Mek2 exhibit a high degree of specificity toward these MAP kinase enzymes, since neither can phosphorylate denatured Erk proteins nor synthetic TEY peptides patterned after the physiologically relevant regulatory phosphorylation sites (TEY peptide: PEHDHTGFLTEYVAWATR WYR) (Seger et al, 1992a). Also, these kinases possess a very narrow substrate specificity such that common and commercially available proteins (e.g. MBP, histones, casein) act as poor in vitro substrates (Seger et al, 1992a).  MAP kinase activators that are capable of phosphorylating Erk enzymes also stimulate their phosphotransferase activity in vitro (Ahn et al, 1991; Gomez and Cohen 1991; Matsuda et al, 1992; L'Allemain et al, 1992; Rossomondo et al, 1992; Alessandrini et al, 1992). To eliminate the possibility that these MAP kinase activators are cellular factors that stimulate Erks by enhancing their autophosphorylating activity, several research laboratories used kinase-inactive Erkl and Erk2 to demonstrate that  these Meks were bona fide kinases. In fact, Meks activate Erks by phosphorylating them on the physiologically relevant sites both in vitro and in vivo (L'Allemain et al, 1992; Rossomondo et al, 1992a; Alessandrini et al, 1992; Posada and Cooper, 1992; Her et al, 1993). These combined results contrast earlier observations that MAP kinases act as points of integration for seryl/threonyl and tyrosyl signalling pathways (Anderson et al, 1990). Because Meks are capable of phosphorylating both seryl/threonyl and tyrosyl residues, they are considered to be dual-specificity protein kinases. Mek protein kinases phosphorylate Erks on their regulatory sites in an ordered manner (Robbins and Cobb, 1992; Hay stead et al, 1992). Under conditions of limiting ATP concentrations in vitro Tyr-185 phosphorylation accumulates before the appearance of Thr-183 in Erk2 (Haystead et al, 1992). Moreover, only the diphospho form of Erk2 appears to be fully active toward MBP. In fact, in NGF-treated PC 12 cells, the dual phosphorylated form is coincident with Erkl and Erk2 attaining maximal enzymatic activity (Robbins and Cobb, 1992).  4.3  Mek/Erk protein complexes and distribution within the cell.  Spatial distribution of the Mek/Erk complexes has been the subject of intense research and debate. In quiescent cells, both Mek and Erk isoforms are located exclusively in cytoplasm (Chen et al, 1992; Sanghera et al, 1992; Seth et al, 1992; Lenormand et al, 1993; Zheng and Guan, 1994). After growth factor stimulation, the Erk isoforms undergo a massive redeployment and subsequently translocate from the cytoplasm into the nucleus where they are presumed to regulate gene expression (Chen et al, 1992; Sanghera et al, 1992; Seth et al, 1992). However, the Mek activator remains localized within the cytoplasm (Lenormand et al, 1993; Zheng and Guan, 1994). Recently, Nishida's laboratory discovered that a short amino acid sequence located in the amino-terminal region (residues 32-44) that is responsible for exporting the Mek protein  from the nucleus into the cytoplasm (Fukuda et al, 1996). This nuclear export signal (NES) is related to the leucine-rich a-helical conformation present in other NES sequence-containing proteins including the protein kinase inhibitor of cAMP-dependent protein kinase (PKA) and the human immunodeficiency virus, type I-coded Rev protein (Fakuda et al, 1996). In addition, a 32 amino acid Erk-binding site peptide located at the very amino-terminal of the Mek protein and contiguous with the NES region, promotes the specific association between Mek and Erk proteins (Fukuda et al, 1997). Mutations at any of the critical leucine residues in the NES region enabled Erk that was microinjected into the cytoplasm to diffuse passively into the nucleus. Hence, The MAP kinase-binding and NES sequences in Mek are thought to act cooperatively as a cytoplasmic anchoring region for Erk (Fukuda et al, 1997). However, Jaaro et al, (1997) recently showed that Mek does massively translocate to the nucleus along with Erk upon cell stimulation, but it is rapidly exported back to the cytoplasm. Therefore, the rapid exclusion of Mek from the nucleus by an active process probably explains the observed subcellular distribution after cell activation (Lenormand et al, 1993; Zheng and Guan, 1994; Fukura et al, 1997).  The mechanisms that promotes the Erk/Mek  translocation to the nucleus and the subsquent dissociation of the molecular complex remain to be established at the molecular level.  5  REGULATION OF MEK PROTEIN KINASE  5.1  Stimulation of Mek protein kinase activity.  The MAP kinase kinases are regulated by reversible protein phosphorylation. The seryl/threonyl-specific protein phosphatase 2A (PP2A) has been shown to inactivate impure Mek preparations but not with tyrosyl specific phosphatases like CD45 (Gomez and Cohen, 1991; Matsuda et al, 1992; L'Allemain et al, 1992; Rossomondo et al,  1992). Mek phosphorylation occurs on serine and threonine residues (Gomez and Cohen, 1992; Matsuda et al, 1993). However, stimulation of Mek enzyme activity is achieved by phosphorylation of two seryl residues located within the kinase activation loop. Ser218 and Ser-222 phosphorylation sites are conserved in all isoforms of MAP kinase kinases (Figure 41). Exchange of either of the regulatory residues with a nonphosphorylatable amino acid like alanine inactivates the Mek enzyme catalytic activity (Zheng and Guan, 1994; Pages et al, 1994; Alessi et al, 1994; Seger et al, 1994; Gotoh et al, 1994; Yan et al, 1994). In contrast, substitution of one or both of the serine sites with acidic residue(s) (e.g. aspartic or glutamic acid) and/or deletion of the NES peptide causes Mek to become constitutively active (Zheng and Guan, 1994; Pages et al, 1994; Alessi et al, 1994; Brunet et al, 1994; Seger et al, 1994; Gotoh et al, 1994; Yan et al, 1994; Cowley et al, 1994; Mansour et al, 1994). Injection of transformed cells expressing activated forms of Mek into mice promoted the formation of solid tumours (Mansour et al, 1994).  5.2  Mek/Erk signalling module.  The combination of Mek's narrow substrate specificity for the Erk protein kinase, the ability of these two enzymes to form stable protein complexes, and their dynamic protein distribution pattern within the cell, implies that these two cytoplamic kinases act as a critical conversion point for cell signalling. Mek protein kinases appear to function as nodal points since these enzymes can be phosphorylated and activated by several upstream kinases including seryl/threonyl-specific proto-oncoproteins Rafl, RafB, Tpl2, Mos, and the budding yeast Stell-related mammalian Mekk. Therefore, the tight regulation of the Mek/Erk module serves as an important convergence point for many inputs emanating from different ligands and their cognate receptors.  6  ACTIVATION OF THE ERK PROTEIN KINASE MODULE  6.1  The Raf Nodal point.  The Raf family of kinases which include Rafl (74-kDa), RafA (68-kDa) and RafB (95-kDa), exhibit more than 70% amino acid sequence identity in their kinase catalytic domain (Storm et al, 1990). Both RafA and RafB transcripts are highly expressed in just a few non-proliferating tissues such as kidney and brain, respectively, whereas Rafl displays a more ubiquitous pattern of expression (Storm et al., 1990). Raf protein kinases share a common architectural design composed of three conserved regions. The amino-terminal non-catalytic CR1 domain contains both the monomeric G protein Ras effector binding and the cysteine-rich zinc finger motif subdomains that have the potential to interact with cellular lipid cofactors. The CR2 is a subdomain rich in seryl and threonyl residues that may be phosphorylated. These phosphorylation events may be important for regulation of enzyme activity of Raf. These two regulatory subdomains control Raf kinase catalytic domain CR3 presumably by folding over and obstructing the active site. Truncation of this region renders the enzyme constitutively active provided that the enzyme is suitably phosphorylated. The precise mechanism involved in the activation of Raf has been difficult to ascertain, since the kinase is regulated by different cell surface receptors and each receptor regulates a set of unique downstream effectors. Furthermore, full enzyme activity may only be achieved through the regulation one or more critical molecular events including translocation, phosphorylation, complex oligomerization and lipid interaction (Hall, 1994; Morrison, 1994; Pelech and Charest; 1995; Gosh et al, 1996; Marshall, 1996).  6.1.1  The Ras-Rafl regulated pathway.  While the MAP kinase pathway was being dissected at the molecular level, other researchers were investigating the mechanisms related to the transduction of the signalling information across the cell's selectively permeable plasma membrane. It is now possible to trace the flow of information from engagement of transmembrane receptor activation with its specific ligand to activation of monomeric Ras GTP-binding protein and subsequent stimulation of cytoplasmic kinases such as MAP kinases and ultimately gene expression within the nucleus.  Ligand interactions (e.g. epidermal growth factor - EGF) with specific binding sites located on the extracellular portion of cell surface receptor kinases trigger a myriad of biochemical events just beneath the surface of the plasma membrane (Panayotou and Waterfield, 1993). In addition to the amino-terminal ligand binding domain, the proteintyrosyl kinase receptor also possesses a hydrophobic transmembrane domain and a carboxy-terminal cytoplasmic catalytic domain (Fantle et al., 1993). Placement of the ligand between the polypeptide receptor chains promotes the oligomerization of these proteins. Juxtaposition of receptors such as PDGF and EGF, allows the cytoplasmic kinase catalytic subdomains to promote phosphorylation on tyrosyl residues via an intermolecular autophosphorylation reaction (Heldin, 1995).  The details of the  mechanism has been confirmed recently from X-ray crytallographic studies (Hubbard et al., 1994). These phosphorylated tyrosines are embedded within specific consensus recognition sequences that together serve as high affinity binding sites for tyrosyldirected peptide binding domains like the Src homology domain 2 (SH2) (Panayotou and Waterfield, 1993). This domain is present in many receptor binding proteins including Grb2 (for growth receptor binding-protein 2), the tyrosyl kinase Src, PI 3-kinase (phosphatidyl-inositol-4,5-bis-phosphate 3-kinase), the Ras GTPase-activating protein  (GAP), and phospholipase Cy (PLCy). Many of these effectors can evoke more than one signalling pathway within the cell. In receptors for such mitogens as insulin, hepatocyte growth factor (HGF) and fibroblast growth factor receptor (FGF), autophosphorylation leads to the increase in catalytic activity and subsequent phosphorylation of other SH2containing proteins like insulin receptor substrate-1 (IRS-1), She and the FGF receptor substrate 2 (FRS2) which can then serve as binding sites for tyrosyl-directed interacting proteins (Skolnik et al, 1993; Kouhara et al, 1997).  On the cytoplamic side of the plasma membrane, Grb2 indirectly binds to these receptors by interaction with the tyrosyl phosphorylated forms of these docking proteins. The two SH3 domains in Grb2 then bind the proline-rich region of the nucleotide exchange protein Sos. Although cellular GTP is more abundant than GDP, the dissociation of GDP from Ras is a rate-limiting step. Once recruited to the cell membrane, however, the guanine nucleotide exchange protein, Sos, stimulates the exchange of GDP for GTP in the G protein Ras (Shou et al, 1992; Wei et al, 1992; Martegani et al, 1992; Botwell et al, 1992; Chardin et al, 1993). The intrinsic GTPase activity regulates the duration of the Ras effector signal. Upon ligand stimulation, conversion of Ras to the active confirmation by Sos induces the mobilization of different cellular effectors.  The link that couples events that begin on the cytoplasmic side of the tyrosyl receptor complex to the seryl/threonyl phosphorylation signalling apparatus within the cytoplasm occurs at the interface between the Ras and Raf proteins. Precisely how Ras activation by receptor kinases stimulates Raf activity remains i l l defined.  Cell  stimulation promotes complex formation between Ras and Raf proteins both in vitro and in vivo (Moodie et al, 1993; Zhang et al, 1993; Warne et al, 1993; Vojtek et al, 1993; Van Aelst et al, 1993). Interaction of the amino-terminal domain of Rafl (residues 52-  132) occurs with the effector binding domain of Ras (residues 26-48) in a GTPdependent manner. The biologically inert Ras (Ala-38) effector domain mutant and the Ras (Asn-17) dominant negative mutant fail to interact with Raf kinase (Moodie et al, 1993; Zhang et al, 1993; Warne et al, 1993; Vojtek et al, 1993; Van Aelst et al, 1993; Traverse et al, 1993; Levers et al, 1994; Stokoe et al, 1994). It was postulated that the Ras-Raf interaction may lead to Raf translocation to the plasma membrane where it becomes activated by a mechanism that still remains poorly defined at the molecular level. It appears that Raf activation might require other elements besides protein-protein interaction with GTP-Ras (Traverse et al, 1993; Zhang et al, 1993). Direct translocation to the plasma membrane can in fact bypass the requirement for Ras. Incorporation of a C A A X isoprenylation signal motif [C, cysteine; A, aliphatic amino acid; X , any amino acid] followed by six polybasic lysine residues in the amino-terminal region of Raf was sufficient to target this enzyme to the inner surface of the plasma membrane. Membrane localization of the CAAX-Raf chimera permitted its constitutive stimulation in the absence of Ras activation (Leevers et al, 1994; Stokoe et al, 1994). Hence, it appears that one important function for Ras during cell signalling is the recruitment of Raf enzyme within the vicinity of its activators located on the inner surface of the plasma membrane. This indicated that other factors located at the cell surface may be required for Rafl activation.  Stokoe and McCormick (1997) devised an in vitro system to examine K-Ras activation of Rafl.  Using reaction conditions that prohibited phosphorylation,  farnesylated GTPyS-K-Ras alone or present in membrane preparations potently activated Rafl in a fashion similar to purified membranes harbouring the activated mutant form of RasG12V. This is in contrast previous studies where other modification events, namely, seryl/threonyl phosphorylation, was implied to be required for Rafl activation after translocation of this protein to the plasma membrane (Kolch et al, 1993; Morrison et al,  1993; Dent et al, 1995; Jelinek et al, 1996). In fact, addition of Mg +-ATP to H2  RasG12V-activated Raf resulted in a rapid autophosphorylation and concomitant inactivation of Rafl activity toward Mekl (Stokoe and McCormick, 1997). Moreover, these data are consistent with earlier findings which demonstrated that an intact zinc finger was necessary for Ras-dependent Raf activation and that this interaction may be facilitated by the farnesyl moiety of Ras (Brtva et al, 1995: Hu et al, 1995; Luo et al, 1997).  6.1.2.  cAMP-Raf 1 regulated pathway.  Production of the second messenger cyclic adenosine 3', 5' monophosphate (cAMP) follows hormone stimulation of heterotrimeric G protein-linked by a seven s  membrane spanning receptor such as the (3 adrenergic receptor. The intracellular increase in cAMP is regulated by G cc subunit (Ga -GTP) activation of adenylyl cyclase. The s  s  cAMP subsequently causes the dissociation of the cAMP-dependent protein kinase (alias protein kinase A, PKA) catalytic subunits from the regulatory subunits. There are numerous reports of cell-type specific effects produced by cAMP. In thyroid cells, hepatocytes, epithelial and Swiss 3T3 cells, cAMP promotes cell division. In contrast, proliferation in cells of lymphoid, fibroblastic or neuronal origin as well as Ras- and Srctransformed cell lines are all inhibited by elevated cAMP levels (Hordijk et al, 1993 and references therein). These results were obtained with the use of stable analog mimetics of cAMP (e.g. 8-bromo-cAMP, dibutyryl cAMP), factors that elevate intracellular cAMP levels through activation of adenylyl cyclase (e.g. forskolin), inhibition of phosphodiesterase (e.g. 3-isobuiyl-l-methylxanthine, IBMX), or expression of constitutively active G cc subunit. Furthermore, these same agents also produce similar s  inhibitory effects on receptor protein-tyrosyl kinase-dependent activation (insulin, PDGF, EGF) and phorbol 12-myristate 13-acetate (PMA) activation of MAP kinases Erkl and  Erk2 (Wu et al, 1993c; Cook and McCormick, 1993; Graves et al, 1993; Stevetson et al, 1993; Burgering et al, 1993; Hordjik et al, 1993).  PKA has no direct inhibitory effect on either of the Erk isoforms or the immediate M A P kinase activators Mekl and Mek2 (Graves et al, 1993).  The  mechanism of cAMP-dependent inhibition of the MAP kinase pathway appears to occur at the level of Rafl or above (Burgering et al, 1993; Cook et al, 1993; Graves et al, 1993; Sevetson et al, Wu et al, 1993; Hafner et al, 1994; Russell et al, 1994; Mischak et al, 1996). Rafl phosphorylated by PKA displays a lower binding affinity for Ras protein (Wu et al, 1993; Chuang et al, 1994; Hafner et al, 1994). Therefore, Rafl lies at a crucial interface for two fundamental cell signalling pathways and has the task of sensing and integrating these opposing effects into an appropriate cell response. The inhibitory effect elicited by the increase in cAMP levels and consequently PKA activity on the MAP kinase pathway may be due to the phosphorylation of Ser-43, which is located near the Ras binding site in the CR1 regulatory region (Wu et al, 1993). In fact, treatment of Rati cells with the pharmacological agent forskolin or direct phosphorylation with PKA increased Ser-43 phosphorylation which dramatically reduced Ras-Rafl protein complex formation.  However, other studies showed that P K A  phosphorylation of a Rafl Ser-43 to alanine mutant protein still caused a dramatic reduction in Raf interaction with Ras. This PKA-dependent inhibition of Raf activity was reversible by PKCoc phosphorylation (Hafner et al, 1994). These results imply that there may be other subdomains in Rafl that are equally important for Ras interaction. Furthermore, a similar PKA-dependent inactivation also occured in Raf mutant proteins that are unable to bind Ras due to a point mutation in the Ras-binding-site (Arg-89) or truncation of the regulatory amino-terminal (Hafner et al, 1994; Mischak et al, 1996). This inhibitory phosphorylation mapped to residue Ser-621 located within the CR3 kinase catalytic domain. The surrounding tryptic peptide sequence RSASEPSLHR  conformed to the minimal consensus sequence (RXXS where X is any amino acid) for PKA (Pearson and Kemp, 1991). Previously, Ser-621 was shown to be constitutively phosphorylated in situ and alteration of this residue by mutagenesis resulted in catalytic inactivation of Rafl (Morrison et al, 1993). Indeed, an intact Ser-621 is required for Raf activity, however, phosphorylation by PKA correlates with kinase inactivation which can be reversed by treatment with seryl/threonyl-specific protein phosphatases 1 and 2A (Mischak et al., 1996, Sprenkle etal, 1997). Furthermore, autophosporylation of Rafl at Ser-621 may lead to its inactivation in vivo (Mischak et al, 1996; Stokoe and McCormick, 1997).  6.1.3  Rapl-RafB regulated pathway  In the rat adrenal medullary-derived pheochromocytoma PC 12 cell line, elevated levels of cAMP stimulated by the NGF ligand causes increased gene expression and neuronal differentiation (Marshall, 1995). Consistent with these results, NGF activation of Erk protein kinase also requires cAMP-dependent activation of PKA to promote these same morphological changes (Frodin et al, 1994; Young et al, 1994; Pan et al, 1994).  There has been much debate as to effectors involved in the regulation of neurite outgrowth formation by NGF. Initial studies supported a role for Ras and Rafl in NGF activation of the MAP kinase pathway. Induction of neurite-like processes in PC 12 cells in response to NGF treatment has been shown in separate experiments to coincide apparently with the stimulation of Ras, Raf, Erk and Rsk catalytic activities (Hagag et al, 1986; Blenis and Erickson, 1986; Gotoh et al, 1990; Boulton et al, 1991; Ohmichi et al, 1992; Scimeca et al, 1992). Expression of dominant inhibitory mutant Ras (Asp-17) attenuated NGF-induced phosphorylation and activation of Erk and Mek protein kinases (Gomez and Cohen, 1991; Robbins et al, 1992; Thomas et al, 1992; Wood et al, 1992).  Also, atalytically compromised forms of Raf, Mek and Erk also interfered with proper neurological differentiation of PC 12 cells, whereas the constitutive-active forms promoted their differentiation (Robbins et al, 1992; Wood et al, 1992; Cowley et al, 1992; Wood and Roberts, 1993). However, in a more rigorous experimental analysis of the involvement of Rafl in PC 12 cell signalling, overexpression of an activated form of Rafl did not stimulate MAP kinase (Wood et al, 1993). Moreover, partially purified Rafl from NGF-treated cells failed to activate the Mek-Erk module (Jaiswal et al, 1994). These results nullify the earlier studies that relied on changes in Rafl phosphorylation state and retarded electrophoresis on SDS-PAGE as a measure of kinase activation. Therefore, a separate signalling pathway originating from the NGF receptor must converge at the Mek-Erk node.  RafB is also expressed in neuronal cells (Vaillancourt et al, 1994). Similar to Rafl, growth factor activation of RafB and its subsequent interaction with Ras-GTP is inhibited by elevated cAMP levels present in the cell (Wu et al, 1993; Moodie et al, 1994; Vaillancourt et al, 1994; Peraldi et al, 1995). However, Erhardt et al (1995) have argued that RafB displays no difference in sensitivity toward cAMP and that an additional signalling component may be required for the kinase in PC 12 cells. The GTPbinding protein Rapl, which has the same effector domain structure as Ras, interacts with RafB in cell free extracts and stimulates the enzymes phosphotranferase activity toward the Mek (Ohtsuka et al, 1996). The exchange of GDP for GTP is mediated by PKA phosphorylation of Rapl. Indeed, PC 12 cells are able to differentially employ RafB and not Rafl in a cAMP-PKA-dependent manner by increasing GTP-loading of Rapl, while inhibiting Ras effector function (Vossler et al, 1997). Consequently, stimulation of Erk and Mek activities in differentiating neuronal cells requires cAMP-dependent activation of PKA, Rapl and RafB.  6.1.4  The Ksrl-Raf 1 regulated pathway  In the nematode Caenorhabditis  elegans,  hyperactive Ras causes pluripotent  precursor cells to assume vulval cell fates. Similarly, nonneuronal cells in the fruit fly Drosophila  melanogaster,  ectopic expression of activated Ras can induce nonneuronal  cells to differentiate into specialized R7 photoreceptor cells. Ksr (kinase suppressor Of Ras) was initially identified in genetic screens for suppressors of constitutively active forms of Ras in these genetic model systems (Therrien et al., 1995; Sundaram and Han, 1995; Kornfeld et al, 1995). The proteins are predicted to range in size from 95-kDa in C. elegans,  100-kDa in mouse and 115-kDa in D. melanogaster.  The isolation of amino-  terminal splice variants for human Ksr indicate that the mammalian gene may add another level of complexity in the regulation of the activity of this enzyme (Therrien et al, 1995). Within the catalytic domain, Rafl appears to be the closest relative of the Ksr protein. One unusual feature found in mammalian Ksr is the substitution of an arginyl residue in place of the highly conserved lysyl-amino acid in catalytic subdomain II that is essential for the phosphotransferase reaction (Therrien et al, 1995). The amino-terminal region contains five conserved areas (CAI to CA5) (Therrien et al, 1995). C A I is a conserved domain unique to Ksr. The CA2 domain is a prolyl-stretch that is reminiscent of an SH3-domain binding site. A cytseinyl-rich domain (CA3) is similar to the lipidbinding domain in PKC and to CR1 domain in Rafl and thus may associate with lipid derivatives. CA4 is a region rich in seryl and threonyl residues (similar to CR2 in Rafl) that contains a MAP consensus phosphorylation site. Finally, CA5 domain in Ksr, like that of Raf CR3, encodes the kinase catalytic domain.  Recently, Ksr has been shown to be related to the ceramide-activated protein (CAP) kinase (Zhang et al, 1997). CAP kinase is a membrane-bound enzyme that is activated in a variety of cell types by natural or synthetic analogues of ceramide (Mathias et al,  1991). The cytokine TNFa, which stimulates sphingomyelinase to release ceramide through hydrolysis of sphingomyelin, can also stimulate the activation of CAP kinase (Dressier et al, 1992). Biochemical evidence indicates that the two kinases may be one and the same protein (Zhang et al., 1997). Both enzymes display the same molecular mass after renaturation in SDS-PAGE (Lui et al, 1994). In a manner similar to CAP kinase, Ksr autophosphorylation was stimulated after overexpressing cells with TNFa or after  in vitro  in vivo  treatment of COS-7 Ksr  exposure of purified membranes to  ceramide (Zhang etal., 1997).  Rafl was also shown to be activated in a sphingomyelinase-dependent manner in response to TNFa receptor engagement (Belka et al, 1995). Moreover, TNFa or ceramide stimulation of endogenous CAP kinase in myeloid FJL-60 cells, or transient overexpression of Ksr in COS-7 cells led to the complex formation and activation of Rafl (Yao et al, 1995; Zhang et al, 1997). Ksr/CAPK phosphorylation of Rafl on Thr-269 was prolyl-directed and led to the stimulation of its phosphotransferase activity toward Mekl (Yao et al, 1995; Zhang et al, 1997). This regulatory phosphorylation site conforms to minimal substrate recognition motif Thr-Leu-Pro for CAPK that was identified using synthetic peptide substrates (Joseph et al, 1993). Therefore, Rafl is probably an important intermediary between engagment of the TNFa receptor and activation of cytoplasmic effectors via the MAP kinase module (Raines et al, 1993; Winston et al, 1995). Complete details of the Erk MAP kinase pathway are summarized in Figure 1.  Figure 1: Regulation of the Mek-MAP kinase signalling module by distinct biological processes in mammals and in yeast Saccharomyces cerevisiae.  33  7.  7.1  PARALLEL MAP KINASE MODULES IN M A M M A L S  MAP kinase superfamily of proline-directed kinases.  Several research groups have identified MAP kinases that are distinct from the p42/p44 isoforms first detected in growth factor-treated quiescent (G ) mammalian cells Q  and hormone-induced germinal vesical breakdown (G2-M) in Xenopus or sea star oocytes (Pelech and Sanghera 1992; Mordret 1993). These MAP kinase isoforms have been identified by molecular cloning studies (Boulton et al, 1991; Gonzalez et al, 1992; Zhu et al, 1994, Lee et al, 1995; Zhou et al, 1995), biochemical analysis (Kariakis and Avruch, 1990; Adler et al, 1992; Hibi et al, 1993; Han et al, 1993; Guesdon et al, 1993; Minshull et al, 1994 and Heider et al. 1994) or immunological detection (Boulton and Cobb, 1991; Sanghara et al, 1992) and have been shown to range in size from between 38-110 kDa. Because so much attention has focused on delineating the Ras —> Raf —> Mek —> MAP kinase regulatory pathway, the importance of these new isoforms in various cellular processes via signal transduction mechanisms has only now begun to be elucidated at the molecular level. It appears that each of these MAP kinase isoforms operate independently in cells within distinct signalling modules with their cognate MAP kinase kinase activators (Zanke et al, 1996a). The notion of distinct M A P kinase pathway regulating unique cellular processes has been demonstrated widely with yeast (Errede and Levin, 1993; Ammerer, 1994; Herskovitz, 1995). Indeed, these recently identified MAP kinase isoforms are members of the MAP kinase superfamily of prolyldirected, seryl/threonyl-specific kinases (Ferrell, 1996).  8  STRESS-ACTIVATED MAP KINASE SIGNALLING PATHWAYS  8.1  Indentification of Jun protein kinase in the regulation of stress signalling.  A second Mjcrotubule-associated protein 2 (MAP2 later called SAPK for stressactivated protein kinase or Jnk for Jun N-terminal kinase) kinase was detected in rat liver after intraperitoneal injection of the protein synthesis inhibitor cyclohexamide (Kyriakis and Avruch, 1990).  Polypeptide misfolding caused by cycloheximide treatment  stimulated a 54-kDa MAP2 kinase that was distinct from the insulin-treated p42 MAP kinase Erk2 originally observed from 3T3-L1 cells (Ray and Sturgill, 1987). Besides the obvious difference in molecular mass of the two proteins, both kinases had distinguishable substrate specificity. Although both Erk2 and MAP2 kinases are proline directed seryl-threonyl kinases (Kyriakis and Avruch, 1990) that require threonyl and tyrosyl phosphorylation for activity (Kyriakis et al. 1991) only Erk2 was capable of phosphorylating and activating the Xenopus Rsk in vitro.  Furthermore, Erk2 was  activatable by ligands that regulate growth and differentiation, while MAP2 kinase was induced by chemical stress. This implied that MAP2 kinase may be involved in a unique signalling pathway (Kyriakis and Avruch, 1990).  During this same period, the c-Jun transcription factor was shown to be phosphorylated by several distinct kinase activities in response to stress factors (Pulverer et al, 1991; Alverez et al, 1991; Alder et al, 1992a; Baker et al, 1992; Hibi et al, 1993). However, upon closer scrutiny only a few of these protein kinase activities could phosphorylate c-Jun within its N-terminal activation domain. Also, a phorbol esterstimulated proline-directed kinase was detected in the human leukemic cell line U937 (Adler et al, 1992a). This c-Jun kinase was affinity purified using a glutathione Stransferase-c-Jun fusion protein linked to Sepharose beads and was shown to be a 67 kDa  protein (Adler et al, 1992b). Hibi et al. (1993) also identified two protein kinases activated in response to UV and transforming oncogenes that were able to bind tightly to a GST-c-Jun affinity resin. The major and minor forms of the two proline-directed seryl/threonyl Jnks had molecular masses of 46 kDa and 55 kDa, respectively. Like Jnks, MAP2 kinase had been shown to be a c-Jun amino-terminal transcription factor phophorylating kinase (Pulverer et al. 1991; Kyriakis et al, 1994). Later cloning experiments would show Jnk and MAP2/SAP to be homologous proteins, mmunoblotting with Erk specific polyclonal antibodies revealed that Jnk/SAP kinases were novel MAP protein kinases that may function within distinct MAP kinase signalling module.  cDNAs have been isolated for human Jnk (Derijard et al, 1994; Kallunki et al, 1994; Sluss et al, 1994) and rat SAP protein kinases (Kyriakis et al, 1994). Jnkl and Jnk2 are closely related at the amino acid level (85-90% identity) with a predicted molecular masses of 44-kDa and 48-kDa, respectively. Similarly, rat SAP kinase a and (3 isoforms display a high degree of conservation (88-90% identity) and are predicted to have a molecular mass of ~48-kDa. Kyriakis et al. (1994) also reported the isolation of additional shorter SAPK clones identical to SAPK(3 cDNA. Divergence at the amino acid level was observed when one of the clones was identified as having a 5-nucleotide insertion at Ser-379 resulting in a reading frame shift and a premature stop. The forty residue truncation was replaced with the sequence AQVQQ-stop. Interestingly, these same authors reported that Jnkl (Derijard et al, 1994) was identical to the partial SAP kinase y amino acid sequence through the first 379 amino acids (Leu-379) after which it terminated with the sequence AQVQQ-stop observed with the shorter SAPK(3. Heterogeneity of mRNA transcipts as observed by Northern hybrization analysis, the variance in the carboxy-terminal sequence and the apparent lower predicted molecular mass of the translated protein product than the observed size of the purified 54-kDa on  SDS-PAGE led to further analysis of Jnk at the genomic level. Gupta et al. (1996) identified at least 10 isoforms of Jnk in adult human brain cDNA libraries. Analysis of the nucleotide sequence revealed that three genes: Jnkl, Jnk2 and Jnk3 yielded the multiple isoforms through alternative processing of their transcripts. The four isoforms of Jnkl and Jnk2 are derived from splice variations in (i) the protein kinase subdomains IX and X and (ii) the carboxy-terminal domain. Two Jnk3 isoforms possess two unique carboxy-terminal sequence that are generated by differential processing of the mRNA. Although Jnk3 showed no variation within the protein kinase catalytic domains IX and X, this kinase did code for a unique amino-terminal sequence that was contiguous with the initiating methionine for Jnkl and Jnk2. The shorter Jnk isoforms were recognized as 46to 48-kDa proteins by Western blot analysis, while the larger isoform counterparts migrated in the range of 55- to 57-kDa consistent with these enzymes having longer amino- and carboxy-termini. It remains to be determined if these Jnk splice variants are differentially regulated in response to stimuli. To date, no homologues of the Jnk protein kinase family have been identified in yeast.  Sequence comparisons between the catalytic domains of the Jnk subfamily of protein kinases reveal that these enzymes share -45% homology at the amino acid level with their Erkl and Erk2 counterparts (Kyriakys et al, 1994; Derijard et al, 1994). Moreover, Jnk isoforms display sequence conservation at sites that are essential for MAP kinase regulation (Ferrell, 1996). A dual phosphorylation site motif similar to the ThrGlu-Tyr (TEY) site critical for Erk protein kinase activation, is present in Jnk protein kinases (Kyriakys et al, 1994; Derijard et al., 1994). Replacement of these sites by sitedirected mutatgenesis with non-phosphorylatable residues prevented the activation of Jnkl in UV irradiated cells (Derijard et al, 1994). However, the tripeptide motif in Jnk, i.e. Thr-Pro-Tyr (TPY), differs slightly from the Erk subfamily members (Kyriakys et al, 1994; Derijard et al, 1994). The Thr-Xaa-Tyr regulatory sequence is located in a peptide  loop between kinase catalytic subdomains VII and VIII, a region that harbours the regulatory phosphorylation and activation domain for many known kinases (Johnson et al., 1996). This linker loop structure, termed L12, in Erk2 when phosphorylated by Mek protein kinase is presumed to undergo critical changes in conformation that promote activation of the MAP kinase enzyme (Zhang et al, 1994). In the case of Jnk protein kinases, the length of the L12 loop between the conserved DFG and the APE subdomains is four amino acids shorter than the comparable region in Erk (Farrell, 1996). It is hypothesized that the length of the linker loop structure in combination with the change in the intervening amino acid separating the regulatory phosphorylation sites, plays a crucial role in recognition by specific activators (Han et al, 1994; Butch and Guan, 1996).  8.2 Stimulation of the Jun kinase pathway.  Jun kinases are activated primarily by stimuli that cause cellular stress, and as such may have a protective role by enabling cells to activate damage repair systems or trigger cell death by apoptosis. Jnk is stimulated in response to a variety of stresses, including environmental stresses such as short-wavelength U.V. radiation (Hibi et al, 1993) , fluctuations in osmolality (Galcheva-Gargova et al, 1994; Sluss et al.,. 1994), (Kyriakis et al, 1994) elevations in temperature (Kyriakis et al, 1994), metabolic inhibitors like ribotoxic (Kyriakis and Avruch, 1990; Kyrakis et al, 1994; Iordonov et al, 1997), and genotoxic agents (Kharbanda et al, 1995; Zanke et al, 1996b; Liu et al, 1996), pro-inflammatory cytokines IL-1 and TNF-a (Kyriakis et al, 1994; Sluss et al, 1994) as well as ischemia and reperfusion (Pombo et al, 1994). In haemopoietic cells, Jun kinase display a modest activation by the tumour promoter TPA, while no activation of Jnk is observed in fibroblasts and epithelial cells (Pulverer et al, 1991; Pulverer et al, 1992; Adler et al, 1992a; Adler et al, 1992b; Sluss et al, 1994). Moreover, Jnk  activation in some cell types may require the integration of inputs from two different stimuli (Su et al, 1994). In T lymphocytes, co-stimulation by TPA and C a  2 +  is  necessary for maximal Jun protein kinase activation and IL-2 production (Su et al, 1994).  8.3  Effectors of Jnk protein kinase signalling.  c-Jun is a component of the activator protein-1 (AP-1) transcription factor complex (reviewed in Angel and Karin, 1991). Heterodimerization between c-Jun and cFos bind specific promoter regions and regulate gene expression in an AP-1-dependent manner. The c-Jun transcription factor binds short, cis-acting DNA sequence known as a promoter binding site. The c-Jun transcription factor is composed of two domains; an amino-terminal trans-activation domain contiguous with a carboxy-terminal DNAbinding domain. Protein phosphorylation controls the transcriptional activity of c-Jun (Hunter and Karin, 1992; Karin and Smeal, 1992). Phosphorylation of two critical residues, Ser-63 and Ser-73, located within the amino-terminal transactivation domain promote transcription of AP-1 responsive genes (Smeal et al, 1991; Smeal et al, 1994). Both residues conform to the consensus sequence for MAP kinases: a phosphorylatable seryl/threonyl amino acid followed by prolyl residue at the P - l position. The critical importance of Ser-73 phosphorylation in the regulation of c-Jun actvitity was elegantly demonstrated in experiments where the sequence surrounding this phosphoacceptor site was converted to a cAMP-dependent protein kinase consensus site (Smeal et al, 1994). The transcriptional potential of this modified c-Jun was acutely regulated by signals that normally affected PKA. Indeed, both Erk and Jnk family members phosphorylated c-Jun activation sites in vitro. However, both Erkl and Erk2 phosphorylated Ser-246, a site located in the carboxy-terminal DNA-binding domain that causes inhibition of gene expression (Alverez et al, 1991; Baker et al, 1992; Minden et al, 1994). In addition, the  Erk protein kinases could only phosphorylate the activating phosphorylation sites of the carboxy-terminal deleted c-Jun protein. As described earlier, Hibi et al. (1993) the Jnk protein kinases interacted specifically with GST-c-Jun chromatography resin and phosphorylated full-length c-Jun transcription factor on the two amino-terminal activating seryl sites. Like the Erk family of MAP kinases, the Jnk protein kinase translocated to the nucleus after cell stimulation with U.V. irradiation; where these enzymes are postulated to have an important role in the regulation of gene expression (Cavigelli etal, 1995).  ATF-2 is also a target of protein phosphorylation and is linked to specific signal transduction pathways (Hunter and Karin, 1992). The transcriptional activation of the cJun promoter by ATF-2 homodimers or ATF-2/c-Jun heterodimers is induced in vivo by cellular stresses including genotoxic agents, and proinflammatory cytokines (Livingstone et al, 1995 and references therein). These same stimuli also activate Jnk protein kinases (Sluss et al, 1994; Kharbanda et al, 1995). Phosphorylation has no effect on the DNA binding properties of ATF-2 in nuclear extracts (Gupta et al, 1995). However, several putative MAP kinase consensus phosphorylation sites were localized in the aminoterminal transactivation domain (Livingstone et al, 1995). Tryptic phosphopeptide mapping localized the Jnk phosphorylation sites within the amino-terminal transactivation domain (Gupta et al, 1995). This was subsequently confirmed by replacement of Thr-69 and Thr-79 with alanines by in vitro mutagenesis that caused a significant reduction in phosphorylation of ATF-2 as well as transcriptional activation of the reporter genes (P-galactosidase or chloramphenicol acetyl-transferase) and c-Jun induction in cells (Livingstone et al, 1995; Gupta et al, 1995; van Dam et al, 1995).  The inducible nuclear localization of Jnk protein kinases in murine FR3T3 cells is consistent with the function of these enzymes in the regulation of transcription factors  c-Jun and ATF-2 (Derijard et al, 1995). Further evidence for the role of Jun kinase in the stimulation of c-Jun and ATF-2 transcriptional activity derives from analysis of the enzyme's ability to complex with specific transcription factor regulatory domains. Jnk binds with high affinity to GST-c-Jun (Hibi et al, 1993; Derijard et al, 1994; Kyriakis et al. 1994). Analysis of the Jnk/c-Jun interaction revealed that Jnk complexes to c-Jun more effectively than the truncated v-Jun oncoprotein despite the fact that both transcription factors retained the regulatory Ser-63 and Ser-73 phosphorylation sites (Adler et al, 1992a; Hibi et al, 1993). c-Jun phosphorylation by Jnk requires a small 30 amino acid region (the 8-subdomain) located amino-terminal to the Ser-63 and Ser-73 (Adler et al, 1992b; Hibi et al, 1993; Derijard et al, 1994). A pseudo-binding 5-domain peptide prevents Jnk interaction and phosphorylation of c-Jun (Adler et al, 1994). Furthermore, Jun kinase phosphorylation of c-Jun promotes the dissociation of the complex (Hibi et al, 1993). A similar Jun kinase docking site was identified in ATF-2 (Gupta et al, 1995; Livingstone et al, 1995).  A specificity-determining region that promotes efficient binding of Jnk to the transcription factor c-Jun has been located in subdomains IX and X of the kinase catalytic core (Kallunki et al, 1994; Sluss et al, 1994). This region contains alternative amino acid sequences or docking sites that mediate substrate specificity for the binding of the various Jnk isoforms to c-Jun (Gupta et al, 1996). The high degree of substrate recognition was further demonstrated with JunB and JunD transcription factors. JunB binds effectively to Jun kinase but is a poor in vitro substrate because the transcription factor lacks the appropriate phosphorylation sites (Gupta et al, 1996; Kallunki et al, 1996). In contrast, JunD possesses the regulatory seryl residues but lacks the necessary docking sites to interact with Jun kinase and consequently is unable to be phosphorylated effectively (Gupta et al, 1996; Kallunki et al, 1996). Interestingly, heterodimeric complexes between c-Jun and JunB and JunD can promote Jnk to phosphorylate the JunD  transcription factor that lacks the docking site (Kallunki et al, 1996). These new complexes may create another level of specificity and diverstiy in the regulation of gene expression by different extracellular signals.  9.  REGULATION OF MKK4/MKK7 PROTEIN KINASES  9.1 Jun protein kinase activators.  The failure of Mekl and Mek2 to stimulate Jnk protein kinase activity indicated that a unique Jun kinase kinase regulatory enzyme was required to phosphorylate and activate this enzyme (Kyriakis et al, 1994). Previously, Yasharer al. (1993) isolated two MAP kinase kinase cDNAs from by PCR (XMek2 alias Mkk4, and XMek3). XMek2 and XMek3 were distinct from the previously identified MAP kinase kinase that was shown to be activated during Xenopus oocyte maturation and hence may define a new MAP kinase pathway in the frog. Interestingly, both XMek2 and XMek3 were similar to the yeast Hogl (High osmolarity glycerol) MAP kinase kinase activator Pbs2 (Polymyxin B sensitive) from S. cerevisiae such that both XMek2 and XMek3 lacked the prolyl-rich insert region located between the catalytic subdomains IX and X. Moreover, activation of Jnkl by osmotic shock in mammalian cells implied that the enzyme may be activated in a manner similar to Hogl (Galcheva-Gargova et al, 1994). Indeed, Jnkl expression in a yeast strain carrying a null mutation in Hogl (high osmolarity glycerol) complemented the defect under hyper-osmolar growth conditions, while Erk2 failed to rescue (GalchevaGargova et al, 1994). Jnk kinase (Jnkk and more commonly termed Mkk4) from human and murine Sekl (for Sapk/Erk kinase-1) share -40% amino acid identity with Mekl and Mek2 within the catalytic region (Sachez et al, 1994; Derijard et al, 1995; Lin et al, 1995). The ubiquitously expressed 43-kDa kinase is most abundantly expressed in mammalian brain and skeletal muscle (Sanchez et al, 1994; Derijard et al, 1995). This  may explain why this pathway is acted upon by a large number of stimulators including EGF in some cell types (Kyriakis et al, 1994). Activation of Jnkl by Mkk4 is achieved through the direct phosphorylation of Thr-183 and Tyr-185 regulatory sites (Lin et al, 1994; Derijard et al, 1995).  Targeted disruption of the Mkk4 gene in embryonic stem cells (ES cells) by homologous recombination impared the activation of Jnk following stimulation with known activators like anisomycin and heat shock (Nishina et al, 1997; Yang et al, 1997). Control experiments demonstrated that the Erk- and Hog-specific MAP kinase pathways remained unaffected by the impairment to Jnk signalling. This is the first genetic evidence in mammals that Mkk4 is required for activation of Jnk in vivo. Unexpectedly, however, U.V. irradiation treatment and changes in osmolarity retained the ability to fully activate the Jun kinases to normal levels (Nishina et al, 1997). These data support previous results that demonstrated the existence of several Jun kinase kinases in protein extracts (Morigushi et al, 1995; Meier et al, 1996). In fact, a new Jun kinase activator, termed Mkk7 (alias Jkk2), has been identified that may have compensated for the Mkk4 deletion in cells (Tournier et al, 1997; Wu et al., 1997; Moriguchi et al., 1997). Mkk7 is the mammalian homologue of Hep (Hemipterous), the activator for Drosophila Jnk (Glise et al, 1995; Sluss et al, 1996).  10.  ACTIVATION OF THE JNK PROTEIN KINASE MODULE  10.1 The Mek kinase (Mekk) node.  The first mammalian homologue of Stell (Sterile) and Byr2 (Bypass of Rasl deficiency) kinases, the cognate activators of the budding yeast Mekl kinases, Ste7 and Byr2, was isolated from a murine brain cDNA library (Lange-Carter et al, 1993). This  protein kinase was denoted Mekk (Mek kinase), since the enzyme activated Mek in an EGF-dependent manner in cells transiently transfected with the Mekkl cDNA (LangeCarter et al., 1993; Fanger et al., 1997). The evolutionary conservation of Mekkl was demonstrated when this enzyme was able to rescue the Stell homologue in the pheromone signalling pathway and Bckl (Bypass of C kinase) gene product cell-wall lysis defect in S. cerevisiae (Blumer et al., 1994). Further studies into the function of Mekkl revealed that expression of the truncated form of Mekkl in NIH 3T3 cells activated the Jnk protein kinase pathway (Yan et ah, 1994). This observation was corroborated in NGF- and EGF-treatment of PC 12 and Hela cells, respectively (Minden et al, 1994b). Jnk activation occurred through Mekkl phosphorylation of Mkk4 on Ser220 and Thr-224 residues; moreover, the location of these sites was equivalent to the regulatory sites (Ser-218 and Ser-222) identified in Mek (Yan et al, 1994). Recently, the full-length Mekkl from rat (94% homologous in the catalytic region) was shown to be a 196-kDa protein which possesses a larger than anticipated carboxy-terminal region (Xu et al, 1996). Three additional Mekk cDNAs were subsequently isolated from murine brain (Blank etal, 1996; Gerwins etal, 1997).  The Mekk2, Mekk3 and Mekk4 isoforms can stimulate the Jnk protein kinase pathway in transient transfection experiments (Blank et al, 1996; Gerwin et al, 1997). The kinase catalytic domains of the smaller 70-kDa Mekk2 and Mekk3 enzymes are 94% conserved within their carboxy-terminal kinase (Blank et al, 1996). The amino-terminal region, which are 65% homologous, may contain a modified bipartite nuclear localization signal (Blank et al, 1996; Fanger et al, 1997). No proximal activators have yet been identified for both Mekk2 and Mekk3 protein kinases. Although the larger 180-kDa Mekk4 is similar in size to Mekkl, both proteins share only a modest 55% conservation in their amino acid sequence (Gerwin et al, 1997). A similar level of homology is observed between Mekk4, Mek2 and Mekk3. Mekkl and Mekk4 contain a prolyl-rich  region and pleckstrin homology domain (PH) motifs located at the very amino terminus of the non-catalytic domain (Xu et al, 1996; Gerwin et al, 1997). A modified Cdc42/Racl interactive binding motif (CRIB motif consensus ISXPXXXXFXHXXHVG where X is any amino acid) also lies adjacent to the kinase catalytic domain in Mekk4, while no such motif was observed in the Mekkl isoform (Gerwin et al, 1997). However, both Mekkl and Mekk4 associate specifically with GTP-bound forms of Rac and Cdc42 (Gerwin et al, 1997). It is expected that these and other functional motifs would allow the Jnk protein kinase pathway to be regulated by a variety of unique regulatory inputs. Overexpression of the four Mekk isoforms in transient transfection assays in HEK293 cells stimulated the activation the Jnk signalling pathway (Blank et al, 1996; Gerwins et al, 1997). Recently, the human homologue of Mekk4, termed Mtkl (MAP three kinase 1), was shown to be a minor mediator of Jnkl activity in transient transfection assays using COS-7 and HeLa cell lines (Takekawa et al, 1997). In contrast, dominant negative Mtkl did not block Jnk activation in response to environmental stress (osmotic stress, UV irradiation, anisomycin) or exposure to the cytokine TNF-a (Takekawa et al, 1997). In support of the notion that all four Mekk protein kinase isoforms regulated specific as well as overlapping signalling pathways is the distinct intracellular distribution of these enzymes; Mekk2 and Mekk4 appear associated primarily with the Golgi, while Mekkl is both nuclear and cytoplasmic (Fanger et al, 1997).  10.2The mixed lineage kinase (Mlk) node.  The Mlk protein kinases were first isolated using degenerate oligonucleotide primers directed against conserved amino acid sequences from a large number of protein kinases (Dorow et al, 1993). A common feature of Mlk is that these enzymes possess amino acid residues that are homologous to both seryl/threonyl and tyrosyl families of protein kinases. In fact, the catalytic subdomains I through VII in Mlk are closely related  to sequences found in the Mekk family, whereas subdomains VIII through XI resemble sequences found in FGF and HER4 receptor tyrosyl kinases (Rana et al, 1996). To date, two members of the Mlk family that are closely related to the Mekk protein kinases are Mlk3 (alias Ptkl for protein tyrosine kinase 1 and Sprk for Src-homology 3 domaincontaining proline-rich kinase), and Dlk (dual-Ieucine-zipper kinase alias Muk for MAP upstream kinase and Zpk for leucine-zipper protein kinase (Ing et al., 1994; Ezoe et ah, 1994; Gallo et al, 1994; Holzman et al. 1994; Reddy and Pleasure, 1994; Hirai et al, 1994). The mixed lineage kinase family possess several structural motifs within their carboxy terminal regulatory domain that play an essential role in promoting specific protein interactions. Juxtaposed to the catalytic domain are two contiguous a-helical leucine/isoleucine zipper motifs followed by a basic region and a proline rich segment. Additionally, the Mlk3 and Dlk contain an amino-terminal SH3 domain and a CRIB domain with a modified consensus sequence (Burbelo et al, 1995). These proteinprotein interaction domains within the Mlk family of kinases add another level of complexity in that they may be activated by a variety of extracellular stimuli.  Until recently, the role these hybrid kinases played in signal transduction remained undefined at the molecular level. However, the high basal activity of these kinases when expressed in transfection experiments revealed that these Mekk-like proteins may be involved in the regulation of the c-Jun stress-activated protein kinase pathway. The dual leucine-bearing kinases Dlk and Muk caused the increase in Jnk phosphotransferase activity in COS and NIH3T3 cells (Hirai et al, 1996; Fan et al, 1996). Still, a catalytically compromised version of Dlk was only able to partially block Jnk activation by GTPase-deficient Cdc42 (Fan et al, 1996). Therefore, these results indicate that the Rho family of G-proteins can use more than one signalling pathway to activate Jnk protein kinase such as the Mekk route. The Mlk3 family also links the Cdc42/Racl GTP-binding proteins to the Jnk signalling cascade. Furthermore, there is  evidence that Mlk3 is able to bind to Cdc42 and Racl in vivo through the CRIB motif. However, the proteins appeared to interact with much lower affinity then observed with other Rho-binding proteins (Teramoto et al, 1996). Indeed, Mlk3 displayed less than a two-fold increase in its ability to stimulate Jnk phosphotransferase activity toward the substrate ATF2, indicating that other co-activators may be involved in the regulation of this enzyme. Whatever the mechanism of Mlk3 activation, signalling to Jnk was shown to be the result of direct phosphorylation of Mkk4 by this enzyme (Rana et al, 1996). Mutation of the two regulatory phosphorylation sites in Mkk4 (Ser-220 and Thr-224) completely abolished the ability of Mlk3 to activate Jnk in transfected COS cells.  10.3 Orphan MAPK kinase kinase nodes.  Tpl2, a protein kinase closely related to budding yeast Stell and mammalian Mekkl was demonstrated to activate the Jnk signalling module in addition to activating the prototypical Erk pathway (Troppmair et al, 1994; Patriotis et al, 1994; Salmeron et al, 1996; Ceci et al, 1997). Both full-length and carboxy-terminal truncated forms of Tpl2 phosphorylate recombinant Mkk4 on regulatory sites (Salmeron et al, 1996). Furthermore, transfection of intact and truncated forms of Tpl2 into COS-1 cells and Jurkat T cells caused the activation of the Jnk signal transduction cascade (Salmeron et al, 1996; Ceci et al, 1997). The extracellular signal that stimulates Tpl2 regulated activation of Jnk remains unknown. Askl (apoptosis signal-regulated kinase) a protein kinase distantly related to Rafl, Ksrl and Tpl2 protein kinases may act a a key intermediary in cytokine- and stress-induced apoptosis (Ichijo et al, 1997).  10.4The Rho monomeric G-protein node.  Morphological changes associated with rearrangements to the actin cytoskeleton depend on the coordinate activities of the monomeric Rho subfamily (Cdc42, Racl, Rac2, RhoA, RhoB and RhoC) of Ras-related proteins (Votjek and Cooper, 1995; Nobes and Hall; 1995). In mammalian cells, Rho induces stress fibers at cell adhesion sites; Cdc42 regulates a distinct signalling pathway that mediates the formation of actincontaining projections termed filopodia; while membrane ruffling or lamellopodia is controlled by Rac (Pelech and Charest, 1995). Although the function of the Rho proteins is associated with the regulation of events at the cell surface, these same small GTPbinding proteins can activate specific cellular kinases.  Besides coordinating the organization of the actin cytoskeleton, Rho family G proteins are involved in the regulation of specific cytoplasmic and nuclear targets (Votjek and Cooper, 1995; Nobes and Hall; 1995). Activation of the Jnk signalling pathway can be mediated by Rac, Cdc42 and Rho.  In transient transfection assays, Jnk  phosphotransferase activity toward c-Jun was stimulated by activated forms of Cdc42 and Racl expressed in Hela, NIH3T3, and COS-7 cell lines (Coso et al, 1995; Minden et al, 1995). In addition, the Cdc42 guanine nucleotide exchange factor protein (GEF) Dbl stimulated Jnk activity. The guanine nucleotide dissociation inhibitors (GDI) RhoGDI and RhoGAP overexpression blocked the activation effect of oncogenic Dbl on the Jnk enzyme activity (Coso et al, 1995). Furthermore, RhoA, B and C in addition to Cdc42 all induce the activation of Jnk in 293T human kidney epithelial cell line (Teramoto et al, 1996). Therefore, all three Rho G-proteins can participate in the activation of the Jnk stress-activated pathway.  10.5The Ste20-related kinase node  Ste20 is an essential gene that operates within the M A P kinase regulated pheromone response pathway in the budding yeast S. cerevisisae (Leberer et al, 1992). The monomeric (Cdc42) and heterotrimeric Gfty-subunits (Ste4 and Stel8, respectively) converge at the level of Ste20 located at the top of the Stell —> Ste7 —> Fus3/Kssl MAP kinase pathway.  Several Ste20-related mammalian homologues have been  identified by sequence analysis of tryptic peptides from Cdc42/Rac affinity purified proteins or by PCR using degenerate oligonucleotide primers (Manser et al, 1994; Katz et al, 1994; Manser et al, 1995; Martin et al, 1995; Teo et al, 1995; Bagrodia et al, 1995; Creasy and Chernoff, 1995a; Creasy and Chernoff, 1995b; Brown et al, 1996; Taylor et al, 1996; Pombo et al, 1996; Keifer et al, 1996; Su et al, 1997). These seryl/threonyl kinases form part of a rapidly expanding family of Ste20-related kinases that are linked to cognate MAP kinase signalling modules. The Ste20-related kinases display distinctive regulatory and structural features; and consequently these enzymes can be arranged into two classes based on enzyme topology: the p21-activated kinase family (Pak) and the germinal center kinase family (Gck).  10.5.1 Pak regulation of the Jnk protein kinase module.  The Pak kinase family consists of the original 68-kDa Pakl (aPak) and the most recently identified 62-kDa Pak2 (yPak) and the 65-kDa Pak3 (PPak) isoforms (Manser et al, 1994; Manser et al, 1995; Bagrodiaef al, 1995; Teo et al, 1995; Martin et al, 1995; Brown et al, 1996). These proteins share >95% identity within their catalytic domains and -70% with the kinase domain of yeast Ste 20 (Leberer et al, 1992). Furthermore, Pakl complements a Ste20 defect in S. cerevisiae, indicating that this enzyme may be functionally equivalent to the yeast counterpart (Brown et al, 1996). All three Pak  isoforms have their kinase subdomain located at the carboxy-terminal side of the protein. The Pak enzymes are targeted by the Rho subfamily of small G-proteins Cdc42 and Racl in a GTP-dependent manner, but not by RhoA (Bagrodia et al, 1995; Knaus et al, 1995). Interaction of the activated forms of Cdc42 or Racl with the conserved 8 amino acid CRIB binding located in the amino-terminal portion of the kinase induces autophosphorylation on seryl residues and activation of Pak phosphotransferase activity toward MBP. Although genetic epistasis experiments have positioned Ste20 as the immediate activator of Stell in budding yeast, there has been no direct evidence for Pak isoform directly regulating Mekk nor Mkk4 activity in mammalian cells (Su et al, 1997; Fanger et al, 1997). Therefore, it would appear that important accessory proteins may have been absent or expressed in such low abundance in the transient transfection assays that proper Pak/Mekk signalling was disrupted in these cells. Equally plausible is the possibility that the Pak subfamily of Ste-related kinases phosphorylate and activate distinct Mekk-like protein kinases. A potential candidate is the recently identified activator of the Jnk pathway, MAPKKK5 a seryl/threonyl kinase that differs from the four Mekk isoforms in that the catalytic region of these enzymes is centrally located between two presumptive amino- and carboxy-terminal regulatory domains (Wang et al, 1996). Despite the lack of evidence of a proximal kinase effector, Pakl and Pak3 were demonstrated to link Cdc42 and Racl G-proteins in the activtion of Jnk protein kinase in transfected COS cells (Bagrodia et al, 1995; Knauss et al, 1995; Brown et al, 1996). Similarly, constitutively activated forms of Pakl and Pak3 induced Jnk activation in these same cells. Activation of Jnk phosphotransferase activity by Pakl was also observed in cell-free extracts of Xenopus oocytes induced to undergo maturation with progesterone (Polverino et al, 1995). Coexpression of Mekkl and Pak3 in sf9 insect cells also lead to activation of Mekkl. However, Pak3 was unable to phosphorylate and activate Mekkl in vitro  (Siow et al, 1997). This result was not surprising since Mekk protein kinases also  display some phosphotransferase activity in transient transfection assays in the absence of extracellular stimulation (Lange-Carter et al, 1993; Fanger et al, 1997).  Most  importantly, however, was the observation that the Pak subfamily of Ste20-related enzymes stimulated the Jnk signalling module without effecting the parallel Erk protein kinase cascade. Further work will be required to determine the intermediary kinases involved in Jnk activation by Pak.  10.5.2  Gck regulation of the Jnk protein kinase module  Members of the Gck family of protein kinases, which have been cloned from several different mammalian sources are related to Spsl, a Ste20-like kinase that is expressed at specific stages during spore cell wall biosynthesis in S. cerevisiae (Krisak et al, 1991; Friesen et al, 1994). The topology of Gck differs from the Pak protein kinases in that the enzyme catalytic domain is positioned at the amino-terminal. Also in contrast to Pak enzymes, the Gck protein kinases lack the Cdc42/Racl CRIB motif. Two classes of Gck have been identified based on the carboxy-terminal region of the enzyme. Gck as well as the related kinases Nik (Nek interacting kinase), Hpkl (hematopoietic progenitor kinase 1), Khs (kinase homologous to Sps/Ste20) and Glk (Gck-like kinase), all range in size from 100- to 140-kDa, contain a large regulatory domain that possesses none of the known protein interaction motifs (Katz et al, 1994; Kiefer et al, 1996; Su et al, 1997; Tung and Blenis, 1997; Diener et al, 1997). The second class of Gck protein kinases including Mstl (mammalian sterile twenty-like alias Krsl for kinase responsive to stress) and Sokl (Ste-20/oxident stress kinase) have molecular masses in the range of 50- to 56kDa and conquently a smaller regulatory domain of unknown function (Creasy and Chernoff, 1995a; Taylor et al, 1996; Pombo et al, 1996).  Spsl regulates a specific MAP kinase, Smkl during the yeast sporulation process. It might be expected that the Gck family of mammalian Ste20 kinases may regulate unique MAP kinase modules. Indeed, transient transfection of Gck protein kinase in COS cells was shown to operate within the Jnk stress-activated kinase cascade (Pombo et al, 1995). In a manner similar to Jnk protein kinase, Gck was activated by TNF-cc in situ. Overexpression of the Gck isoform Hpkl and Khs in COS1 fibroblast cells also stimulated Jnk phosphotransferase activity toward c-Jun (Kiefer et al, 1996; Hu et al, 1996; Tung and Blenis, 1997). The kinase catalytic activity of Hpkl was required for the activation of the Jnk pathway, whereas the expression of carboxyterminal regulatory portion had no effect (Hu et al, 1996).  Mlk3 or a similar  intermediary kinase may link Hpkl to Jnk protein kinase. Hpkl was able to directly phosphorylate Mlk3 in vitro (Kiefer et al, 1996; Hu et al, 1996). Furthermore, coexpression studies demonstrated that the two enzymes could physically associate via the amino-terminal SH3 domain of Mlk3 and two of four prolyl-rich motifs in Hpkl. Khs and Glk also contain polyprolyl helix structures that may interact with SH3 binding sites on proteins that have yet to be identified experimentally (Tung and Blenis, 1997; Diener et al, 1997). The Jnk regulator Gck, which does not contain these prolyl-rich motifs was also capable of signalling to Jnk through Mlk3 (Pombo et al, 1995; Tibbies et al, 1996). It is possible that other sites in the carboxy-terminal domain may play essential roles in the regulation of cytoplasmic effectors. Mlk3 which has several protein interaction motifs co-immunoprecipitated with Mkk4 in co-transfection studies (Tibbies et al, 1996).  As mentioned previously, the mixed lineage kinase homologue, Sprk,  phosphorylated and activated Mkk4, the Jnk activator (Rana et al, 1996).  The yeast two-hybrid system was used to screen for proteins with prolyl-rich sequences that associated with SH3 domain of the adaptor protein Nek (Su et al, 1997). Nik (Nek interacting kinase) is a 140-kDa seryl/threonyl kinase homologous to  Ste20/Spsl. Like Gck and Hpkl, Nik activated the Mekk —> Mkk4 —> Jnk —> c-Jun module in transient overexpression assays. Although there was no evidence that Nik directly regulates Mekk in vivo, two pieces of indirect evidence support this notion. Catalytically-inactive mutants of Mekk inhibited stimulation of Jnk activity by Nik. Also, Mekk and Nik associate when co-expressed in 293 cells (Su et al., 1997). Interaction between the two enzymes was mediated through the amino-terminal of Mekk and carboxy-terminal of Nik. Su et al. (1997) observed a strong sequence conservation (>70%) between the carboxy-terminal domains of Nik (see below) and a C. elegans gene whose function is unknown. This 325 amino acid domain is present in Gck and Hpkl, and consequently may be critical for the regulation of dowstream effectors by the Gck family (Su et al, 1997). In contrast, the related Mstl/Sok class of Gck protein kinases which possess a smaller carboxy-terminal tail are missing this 325 amino acid region. This may explain the inability of Sokl, Mstl and Krsl kinases to activate the Jnk pathway (Creasy and Chernoff, 1995; Taylor et al., 1996; Pombo et al, 1996; Su et al, 1997). Therefore, these Sps 1-like protein kinases may operate within distinct MAP kinase modules. At present, the mechanism of how Nek couples Nik to receptor or nonreceptor tyrosyl activities that occur at the inner surface of the cell membrane are not well understood.  Recently, the SH3- SH2-adaptor protein Nek and Pak were shown to form a tight association in vitro and in vivo when co-expressed in L6 rat myoblast, Cos7 and 293T cell lines (Galesteo et al, 1996; Bokoch et al, 1996; Lu et al, 1997). The interaction is mediated between the first prolyl-rich SH3-binding domain of Pakl and the second SH3 domain of Nek (Galisteo et al, 1996; Bokoch et al, 1996). Furthermore, Pakl/Nck complexes are recruited to the inner face of the plasma membrane after cell stimulation with growth factors EGF and PDGF (Galisteo et al, 1996; Bokoch et al, 1996). Translocation to the membrane led to an increase in Pakl phosphotransferase activity  Figure 2: Diverse proximal inputs regulate the MkJc4/Mkk7-Jnk stress signalling module.  55  (Galisteao et al, 1995; Lu et al, 1997). Therefore, the targeting of Pak to receptor tyrosyl kinases may lead to the activation of one or more of the MAP kinase modules. Figure 2 summarizes the various regulatory imputs that modulate Jnk activity.  11.  HIGH OSMOLARITY MAP KINASE SIGNALLING PATHWAYS  11.1 The Hog MAP protein kinase pathway in Saccharomyces  cerevisiae  In budding yeast, changes in environmental osmolarity activates a stress-related MAP kinase to increase the expression of genes critical for survival of the cell (Herskowitz, 1995). 5. cerevisiae utilizes an osmosensor mechanism that is similar to the prokaryotic two-component system. This multistep phosphorelay system, Sln-Ypd-Sskl, is coupled to a MAP kinase module distinct from the pheromone pathway. Under homeostatic conditions, the transmembrane histdyl-kinase osmosensor, Slnl, maintains Sskl (suppressor of sensor kinase) in an inactive state (Maeda et al, 1994). Immediately following autophosphorylation of Slnl receptor on a histidyl residue, the high energy phosphate is relayed to an aspartyl site located in the receiver domain of the same protein. The phosphate is shunted from the aspartyl residue on Sin to the inactivating aspartyl site on Sskl via histidine on the Ypdl intermediary receiver protein. Elevated osmotic levels reduce Slnl activity and consequently promote Sskl stimulation through reduced phosphorylation of the protein. Sskl promotes the production of intracellular glycerol by recruiting a MAP kinase module composed of two redundant Mek kinases, Ssk2 and Ssk22, the Mek homologue, Pbs2 (polymyxin B sensitive) and the MAP kinase, Hogl.  11.2  Identification of Hog protein kinase in the regulation of stress signalling.  Engagement of the CD 14 glycosylphosphatidylinositol-anchored cell-surface glycoprotein with endotoxin (LPS for Hpopolysaccharide) triggers the rapid tyrosyl phosphorylation and activation of the 42- and 44-kDa MAP kinase isoforms Erk2 and Erkl (Weinstein et al., 1992). However, the appearance of an additional 38-kDa tyrosyl phosphorylated protein was also observed in RAW264.7 monocytic-like and pre-B 70z/3 cell lines (Weinstein et al, 1992; Han et al, 1993). Purification and molecular cloning revealed that p38 was related to the S. cerevisiae  gene, Hogl (Han et al, 1994).  Ultimately, LPS stimulates the production and release of proinflammatory cytokines IL-1 and TNF-cc from cells like monocytes and macrophages. LPS is an endotoxin associated with Gram-negative bacteria. Since this inflammatory stimulus can induce powerful immune responses which may lead to septic shock, compounds that can block this process could be useful therapeutically for many acute and chronic inflammatory diseases (Lee et al, 1994). Pyridinyl-imidazole compounds termed CSAID (cytokine-suppressive anti-inflammatory drug) were shown to be powerful inhibitors in the biosynthesis of these proinflammatory compounds (Lee et al, 1994 and references therein). Identification of the cellular targets of CSAIDs revealed that one of the CSAID binding proteins (Csbp) was a 38-kDa Hogl-related protein kinase (Lee et al, 1993). The bicyclic pyridinyl imidazole inhibitors were shown to prevent Hogoc activity by interfering with ATP access to the ATP binding pocket (Young et al, 1997). Selectivity most likely occurs through interaction of CSAID with non conserved regions close to the ATP binding site. One consequence of cell exposure to IL-1 and TNF-oc is the phosphorylation of small heat shock proteins, Hsp25 and Hsp27 (Kaur et al, 1989; Saklatvala et al, 1991). IL-1 and stress (arsenite and heat shock) stimulated a protein kinase pathway in human epithelial KB cells and Xenopus laevis oocytes involving a p38 homologue (p40 and Rk for Reactivating kinase, respectively) and MAPKAPK2 (mitogen-activated protein kinase-  activated protein kinase 2) that subsequently led to the phosphorylation of Hsp27 (Guesdon et al, 1993; Freshney et al, 1994; Rouse et al, 1994). Since p38 can complement a Hogl null mutant and become activated in response to changes in osmolarity, the mammalian isoform will be referred to as Hoga.  Hoga cDNAs have been isolated from mouse (p38), Xenopus (Mpk2), and human (Csbpl and Csbp2) (Han et al, 1994; Rouse et al, 1994; Lee et al, 1994). The two human Csbp isoforms were identical within the coding sequence except for a 75 nucleotide (25 amino acid polypeptide) mRNA splice variation in which there is a choice between two exons (Lee et al, 1994). An second alternatively spliced variant of Hoga, Mxi2 ( M ^ interactor 2), was isolated in a search for proteins that interact with the transcription factor Max in a two-hybrid screen (Zervos et al, 1995). Although catalytic subdomain XI is replaced with a shorter 17 amino acid sequence which leads to an abrupt truncation of Mxi2 carboxy-terminus, the enzyme retained the ability to phosphorylate Max in vitro. Typically, human Hoga is synthesized as a 360-amino acid polypeptide chain (Lee et al., 1994; Zervos et al, 1995). The predicted molecular mass of Hoga is -41-kDa; however, the activated form of the protein kinase appears to migrate faster under certain SDS-PAGE conditions.  To date, three new members of the Hog family of protein kinases have been isolated by using PCR amplification or by performing data base searches for cloned polypeptide sequences similar to Hoga in the expressed sequence tag (EST) database at the National Center for Biological Information. Once identified, the EST clones were then retrieved by DNA cloning. The largest of the Hog family is the 372-amino acid human P-isoform (p38(3) which is 74% identical with the Hoga (Jiang et al, 1996). In contrast, HogS (alias Sapk4) displays 58% and 59% identity over its 365-amino acid length when compared to the a- and (3-isoforms (Goedert et al, 1997). Finally, the 367-  amino acid human Hogy is 63%, 62% and 64% identical with the a-, (3- and 5- isoforms (Li et al, 1996; Lechner, et al, 1996; Mertens et al, 1996). Interestingly, Hogy tissue distribution is restricted to skeletal muscle whereas the other isoforms are expressed more ubiquitously, albeit at varying levels (Li et al., 1996). Like the Erk and the Jnk family of protein kinases, the mammalian and yeast Hog proteins share the signature threonyl and tyrosyl phosphorylation site motif located between catalytic subdomains VII and VIII. Also, the activation lip that harbours the TGY motif is six residues shorter than observed in MAP kinases Erkl and Erk2. Unexpectedly, only Hoga and P isoforms were inhibited by SB 202190 implying that there may be subtle differences at the amino acid level that enables certain isoforms of Hog protein to interact with pyridinyl imidazole compounds (Lee et al, 1994; Jiang et al, 1996)  11.3  Activation of Hog protein kinases  The four Hog isoforms become activated to varying degrees by the proinflammatory cytokines IL1 and TNF-oc (Freshney et al, 1994; Jiang et al, 1996; Cuenda et al, 1997; Goedert et al, 1997). In addition, these same enzymes are also regulated in response to other environmental stresses such as hyperosmotic conditions (Han et al, 1994; Jiang et al, 1996; Cuenda et al, 1997; Goedert et al, 1997), U.V. irradiation (Derijard et al, 1995; Lin et al, 1995), increases in temperature (Rouse et al, 1994), ribotoxic agents (Jiang et al, 1996; Cuenda et al, 1997; Goedert et al, 1997), thrombin or thromboxane activation of platelets through seven-transmembrane receptors linked to heterotrimeric G proteins (Kramer et al, 1995; Saklatvala et al, 1996). However, the Hogy isoform may be involved in tissue-specific functions because of the protein's unique expression pattern.  Indeed, C2C12 pre-muscle cell line stably  expressing Hogy (alias Erk6) were induced to undergo myoblast cell differentiation into myotubes (Lechner et al, 1996). Many of the prototypic growth factors that are known  to activate the Erk protein kinases such as EGF and IGF-1 fail to robustly stimulate Hog isoforms. One exception is fibroblast growth factor (FGF) which activates the Hog signalling pathway in the neuroblastoma SK-N-MC cell line (Tan et al, 1996).  11.4  11.4.1  Effectors of Hog protein kinase signalling.  Regulation of gene expression.  The amount of contribution made by the Hog isoforms in response to stress signalling is not well defined at present. To date, the amino terminal fragment of ATF2 has been shown to be a more useful in vitro substrate than MBP (Lee et al., 1994; Derijard et al, 1995; Raingeaud et al, 1995). It is unclear, however, what role ATF phosphorylation by Hog has in the regulation of gene expression (Bayaert et al, 1996; Hazzalin et al, 1996; Iordinov et al, 1997). To identify novel in vivo substrates for Hog, a human fetal brain cDNA library was screened by yeast two-hybrid analysis (Han et al, 1997). The transcription factor, myocyte-enhancer factor 2C (MEF2C), was recovered using an inactive mutant of Hoga as a bait. MEF2C was the preferred substrate of Hoga whereas MAP kinase family members Erk2 and Jnkl were less efficient (Han et al, 1997). Stimulation of MEF2C transcriptional activity in LPS-treated RAW264.7 cells was reversible by incubation with the p38-specific inhibitor SB202190 or expression of dominant-negative Hoga.  MEF2C transcriptional activity is regulated by Hoga  phosphorylation of Thr-293 and Thr-300 located within the transactivation domain and a third site, Ser-387, located outside this region (Han et al, 1997). Of critical importance is the presence of a MEF2C DNA-binding site in the c-Jun promoter region, which is LPS-inducible (Han et al, 1995). LPS triggers a defence response in the innate immune system by inducing changes in the de novo synthesis of cytokine factors like IL-1 and TNF-a in monocytic cells (Lee et al, 1994). Moreover, gene expression from many  cytokine promoter regions is regulated by the AP-1 transcription factor complex (Newell et al, 1994; Han et al, 1997). Hence, LPS-induced MEF2C phosphorylation may contribute to the regulation of the c-Jun expression and indirectly to cytokine production (Han etal, 1997).  11.4.2  Regulation of MAPKAPK2 activity  The mitogen-activated protein kinase-activated protein kinase 2 (MAPKAPK2) was first identified as an enzyme that was activated after phosphorylation by MAP kinase (Stokoe et al, 1992). Originally purified from rabbit skeletal muscle, MAPKAPK2 was demonstrated to exist as a 45- and 55-kDa protein by gel electrophoresis (Stokoe et al, 1992; Cano et al, 1994; Cano et al, 1996). The 55-kDa, 400 residue MAPKAPK2 protein contains a prolyl-rich amino-terminal domain followed by the catalytic domain and a putative nuclear translocation signal at the carboxy-terminal (Engel et al, 1993; Stokoe et al, 1993; Zu et al, 1994). A second related 382-amino acid protein with a predicted molecular mass of 42 kDa probably corresponds to the smaller isoform (Sithanandam et al, 1996; McLaughlin et al, 1996). MAPKAPK3 (alias 3pK for chromosome 3p kinase) is 75% identical at the amino acid level to MAPKAPK2 and is also structurally similar. Furthermore, both MAPKAP kinases lie within the Hog signalling pathway.  In addition to Erk protein kinases, Hoga was shown to be an activator of MAPKAPK2 in stress-induced cells (Rouse et al, 1994; Freshney et al, 1994). Pyridinyl imidazole compounds such as SB 203580 directly inhibit Hog in situ and prevent stress activation of MAPKAPK2 (Cuenda et al, 1995).  It was initially  demonstrated that MAPKAPK2 activation required phosphorylation on carboxy-terminal Thr-334 to become activated by Erkl and Erk2 (Stokoe et al, 1992). This implied that  the process of MAPKAPK2 activation is unique from that of other seryl/threonyl protein kinase enzymes that require phosphorylation within the catalytic domain (Johnson et al, 1996). Indeed, a second MAP kinases consensus site at Thr-222 was identified in the L12 activation loop (Engel et al, 1995; Ben-Levy et al, 1995). In fact, constitutive activation of MAPKAPK2 can be achieved by replacement of Thr-222 and Thr-334 with glutamic acid mimetic residues or by deleting the A-helix motif and consequently Thr334 in the carboxy-terminal region.  Therefore, M A P K A P K 2 may require two  phosphorylation events to become fully-active: 1) phosphorylation at Thr-222 to promote local conformational changes for peptide substrate binding and; 2) phosphorylation at Thr-334 located in the A-helix motif to relieve inhibition from the hydrophobic pocket between ATP- and substrate-binding catalytic lobes (Engel et al, 1995).  The consensus phosphorylation site sequence for MAPKAPK2 recognition is Hyd-X-Arg-X-X-Ser, where Hyd corresponds to a hydrophobic residue (Stokoe et al, 1993; Clifton et al, 1996). To date, only a few potential in vivo substrates have been identified for MAPKAPK2. Murine Hsp25 and human Hsp27 are phosphorylated in response to many of the stress stimuli that activate the Hog kinases (Pelech and Charest, 1995). Phosphorylation of heat shock proteins appears to initiate reconstruction of the actin microfilament network (Lavoie et al, 1995). Studies using SB 203580 supported a role for Hog in regulating MAPKAPK2 phosphorylation of Hsp25 and Hsp27 at physiologically relevent sites (Cuenda et al, 1995). Stress-induced stimulation of tyrosine hydroxylase (TH) has been noted in chromaffin cells (Sutherland et al, 1993). Inhibition of the Hog signalling pathway demonstrated that M A P K A P K 2 phosphorylation of Ser-19 modulates THs rate-limiting activity in the synthesis of catecholamine (Thomas et al, 1997).  11.5  Hog protein kinase activators.  The ability of Hog and Jnk protein kinases to complement a mutation in Hogl in budding yeast implied that a.kinase similar to the Hogl activator Pbsl might regulate both enzymes.  Indeed, the Jnk activator, Mkk4, was demonstrated in vitro to  phosphorylate Hogcc at the regulatory TGY motif (Derijard et al, 1995; Lin et al., 1995). However, cotransfection studies with constitutively activated Mekk, a known activator of Mkk4, did not effectively activate Hog a in COS cells (Lin et al., 1995). A separate protein kinase, namely Mkk3, appears to operate exclusively within the Hog signalling module. In fact, Mkk3 appears to activate only the Hoga isoform (Jiang et al., 1996; Cuenda et al., 1997; Goedert et al., 1997). The identities of several other upstream activators have been investigated in PC 12 and KB cells after treatment with cytokines or cellular stressors (Meier et al., 1996) Several research groups have identified other novel MAP kinase kinases (Han et al., 1996; Moriguchi et al, 1996; Raingeaud et al, 1996; Cuenda et al, 1996; Stein et al, 1996; Moriguchi et al, 1996). One such Hog protein kinase activator is Mkk6. The substrate specificity of Mkk6 (alias Mek6, Sapkk3) is similar to that of Mkk3 in that neither can enhance Erk or Jnk activity (Han et al, 1996; Raingeaud et al, 1996). However, one difference between Mkk3 and Mkk6 is that Mkk6 could phosphorylate the (3, y and 8 isoforms of Hog in coexpression experiments (Jiang et al, 1996; Cuenda et al, 1997; Goedert et al, 1997). These two closely related protein kinases (-80 identity between Mkk3 and Mkk6) are regulated by dual phosphorylation on conserved seryl and threonyl sites located in the activation loop.  To date only a couple of upstream activators have been demonstrated to communicate with the Mkk3/Hog signalling module. The murine homologue of Mekk4 was identified in functional complementation of osmosensitivity in budding yeast. Mtkl is structurally similar to Ssk2 and Ssk22 and shares 98% identity with human Mekk4  Figure 3: Hog-dependent signalling modules in osmosensing in Saccharomyces  cerevisiae  and stress signalling in mammals.  65  66  within the kinase catalytic domain (Takekawa et al, 1997). Furthermore, in coexpression studies, catalytically compromised Mtkl was still able to mediate Hogoc activation that was normally induced by changes in osmolarity (Takekawa et al, 1997). The ability of Mtkl to activate the Hog pathway depended on the activities of Mkk3 and Mkk6 enzymes. Another proposed activator of the Hog signal transduction pathway is the Rho family GTPases Cdc42 and Racl (Zhang et al, 1995). However, recent results by the same research group reported that overexpression of Cdc42/Racl does not lead to activation of Hog in a Mkk3/Mkk6-dependent-manner but instead uses Mkk4 in the process (Han et al, 1996). Figure 3 outlines the known players in the Hog stressactivated pathway in yeast and mammals.  12  12.1  ORPHAN MAP KINASE SIGNALLING MODULES  Erk3 signalling module  The 63-kDa Erk3 possesses many structural features that distinguish it from the other MAP kinase family members. In contrast to Erk, Jnk and Hog isoforms, the tyrosyl residue present in the signature Thr-Xaa-Tyr regulatory motif is absent in Erk3 (Boulton et al, 1991). Also, the regulatory threonine is replaced by serine at residue 189 in Erk3 (equivalent to Thr-202 in Erkl) (Boulton et al, 1991). The lack of tyrosine may explain the inability of Erk3 to phosphorylate any of the common MAP kinase substrates (Cheng et al, 1996a). Recently, the crystal structure of the active form of Erk2 revealed that Tyr-185 may play a critical role in recognizing the prolyl residue in the P + 1 site of the substrate (Canagarajah et al, 1997). In mammalian cells, Erk3 is a contitutively nuclear protein kinase which contrast the behavior of Erk, Jnk and Hog protein kinases which are located in the cytoplasm and the nucleus (Cheng et al, 1996a).  Rat Erk3 is most closely related to Erkl and Erk2 MAP kinase family isoforms (Boulton et al, 1991). The human homologue of Erk3 is almost identical (98%) to rat within the first two-thirds on the enzyme (Zhu et al, 1994). Alternative splicing at the carboxy-terminal region incorporates a unique extension of 178 amino acids that generates a 97-kDa protein (Zhu et al, 1994). Multiple Erk3-like genes have been reported to exist in mammalian cells (Boulton et al, 1991). An Erk3-related protein kinase that is 72% identical to the rat isoform has been identified for humans (Gonzalez etal,  1992).  An Erk3 protein kinase activator activity has been identified in rabbit muscle extracts and NGF-treated PC12 cells (Cheng et al, 1996b). The Erk3 kinase was shown to bind tightly to the Erk3 catalytic domain. Erk3 kinase phosphorylation site was determined experimentally to be the regulatory Ser-189 (Cheng et al, 1996b). The Erk3 kinase activator is present in both the cytosol and nuclear compartments of PC12 and 293 cells. In quiescent human fibroblasts, the HH1 kinase actvitiy of the 97-kDa Erk3 isoform is activated by serum and phorbol ester treatment (Zhu et al, 1994). Sauma and Friedman (1996) have reported that stable transfection of PKC [31 into colon cancer cells activates Erk3 MBP phosphotransferase activity. However, in both these instances the degree of phosphorylation appeared to be very low.  12.2  Erk5 signalling module  The fifth member of the MAP kinase family is Erk5 (also termed Bmkl for Big mitogen-activated kinase 1) (Zhou et al, 1995; Lee et al, 1995). Like Erkl and Erk2 isoforms, Erk5 contains the canonical TEY phosphorylation motif. However, Erk5 contains a 400-amino acid carboxy-terminal regulatory/localization domain. Erk5 is  Figure 4: M A P kinase specific signalling modules in mammalian cells.  69  70  uniquely paired with a specific MAP kinase activator, Mek5 (Zhou et al, 1995; English et al, 1995). Alternative splicing at the amino-terminal domain of Mek5 may confer specific subcellular localization (Zhou et al, 1995; English et al, 1995). Erk5 was activated in response to H2O2 and therefore may participate in a redox-sensitive pathway (Abe et al, 1996). Src, which also is stimulated by reactive oxygen species, appeared to be required for activation of Erk5 in mouse fibroblast cells (Abe et al, 1997). A summary of the five MAP kinase modules identified in mammals is depicted in Figure 4.  13  13.1  REGULATION OF MAP KINASE IN MATURING SEA STAR OOCYTES  Identification of MAP kinase in sea star oocytes.  Echinoderm oocytes are naturally arrested in prophase of meiosis I. Consequently, the sea star oocyte is a useful model system for studying the cell cycle since the immature egg can be induced to mature synchronically from a quiescent state. Furthermore, a large number of cells may be easily obtained for biochemical and biological analysis. Treatment of sea star oocytes with the hormone 1-methyladenine (1MeAde) stimulates ovulation in the female and resumption of oocyte maturation. Mature eggs from pisaster ochraceous, like other species of echinoderm, arrest at the Gi-phase of meiosis II (also considered the pronuclear stage). A number of protein kinases become activated during the maturation process that phosphorylate the exogenous substrates histone HI (HH1), MBP and S6 (Pelech and Krebs, 1987; Meijer et al, 1987; Pelech et al,  1988). This M-phase specific HH1 kinase which a component of maturation  promoting factor (MPF) was subsequently identified as a homologue of Cdc2 the cell cycle regulator of the fission yeast S. pombe (Arion et al, 1988; Labbe et al, 1988). MBP kinase was purified to homogeneity from 1-MeAde-treated oocytes (Sanghera et al, 1990). Microsequencing of tryptic peptide fragments revealed that the enzyme was a  member of the MAP kinase family (Posada et al., 1991). Contrary to other MAP kinases, Mpkl purified from sea star oocytes appeared not to require phosphorylation on threonine to become activated in maturing oocytes (Sanghera et al, 1991). However, recent cloning of the Mpkl cDNA revealed that the conserved threonyl and tyrosyl regulatory motifs are present in the sea star MAP kinase (Charest et al, unpublished data). In addition, recombinant Mpkl purified from bacteria was poorly activated in vitro by mouse Mekl (Charest et al, unpublished data). These results imply that sequence differences in Erkl and Mpkl protein kinases may confer substrate specificity for the upstream MAP kinase activators. Raf- or Mos-like kinases have yet to be unequivocally confirmed in echinoderms, although immunoblotting data supports the presence of Raf (Palaty and Pelech personal communication; Meijer, personal communication).  Although the identity of the 1-MeAde receptor at the cell surface of the sea star oocyte is unknown, it appears to activate heterotrimeric G proteins. Pretreatment of the oocytes with pertussis toxin, an inhibitor of G[ class of proteins, prevented the 1-MeAde induced maturation (Shilling et al, 1989; Chiba et al, 1992; Tademuda et al, 1992). Microinjection of the fty-subunit from bovine retina into immature sea star oocytes induced germinal vessel breakdown (GVBD) with the same time course as observed with 1-MeAde treatment (Jaffe et al, 1993). In contrast, microinjection of non-myristoylated G protein a subunit blocked 1-MeAde-induced maturation of sea star oocytes (Jaffe et al, 1993). It remains to be determined what effect heterotrimeric protein Gi Py- and asubunits have on activation of Mpkl in sea star oocytes.  HYPOTHESIS  1. MAP protein enzyme is activated in many cell types by a variety of stimuli (Pelech and Charest, (1995). Isolation of the cDNA will allow further characterization of the kinase at the protein level.  2. MAP protein kinase phosphorylates Rskl on regulatory sites that lead to its activation in vivo (Sturgill et al., 1988). As MAP kinase is regulated by reversible protein phosphorylation, it is expected that the enzyme will be regulated by an upstream MAP kinase activator.  3. As a kinase which is tightly regulated by phosphorylation, it will be important to examine the importance of the regulatory phosphorylation sites (Sturgill et al., 1988; Anderson et al., 1991). Characterization of the regulatory phosphorylation sites in MAP kinase will aid in understanding the specificity of the consensus site recognition sequence for its upstream activator.  OBJECTIVES  1. To clone human and sea star MAP kinase.  2. To express the cloned MAP kinase in prokaryotic cells.  3. To purify and characterize the direct activator of MAP kinase from sea star oocytes.  4. To clone the MAP kinase activator and express the protein in prokaryotic cells .  5. To determine the substrate specificity of the MAP kinase activator for MAP kinase by altering the regulatory phosphorylation sites by site-directed mutagenesis.  MATERIALS AND METHODS  1.  MATERIALS  A list of commonly used laboratory chemicals and their supplier(s) is summarized in Table 1. Although many of the chemicals are used for both molecular and biochemical technologies, only molecular biology grade reagents were used for the manipulation of DNA and RNA. This minimized degredation of the nucleic acids. As a second measure, all stock solutions used in molecular biology were sterilized by autoclaving for 20 min at 15 pound/square inch on liquid cycle. In those instances where a reagent could not be autoclaved the solutions were filter sterilized using a 0.2 um syringe filter or in the case of solutions over a 50 ml volume a self-contained 0.2 um membrane filtering unit. Also included in Table 1 are the miscellaneous reagents, consumables and photographic supplies required for many of the experimental protocols.  2.  CELL MANIPULATIONS  2.2.  Oocyte isolation and cell culture  2.2.1.  Oocyte preparation  2.2.1.1 Mechanical disruption  Gravid adult female sea stars {Pisaster ochraceus)  were gathered between the  months of March and July from the intertidal zone surrounding Vancouver, British Columbia. The animals were kept until needed in sea water tanks at the Department of Fisheries and Oceans located in West Vancouver, B. C. Immature sea star oocytes,  Table 1: Research materials and their commercial sources A. Chemicals Acetic acid (CH COOH) Acetonitrile Acrylamide Adenosine 5'-triphosphate disodium salt Agarose Agarose (low melting point) Ammonium bicarbonate (NH4HCO3) Ammonium hydroxide Ammonium persulphate Ammonium sulphate Bis-acrylamide N,N'-Methylene bis-acrylamide Boric acid Bovine serum albumin Brilliant Blue G 5-Bromo-4-chloro-3-indoyl phosphate (BCIP) 1-Butanol iso-butanol (3-glycerolphosphate P-methyl aspartic acid Calcium chloride (CaC^) Citric acid Chloroform Coomasie brilliant blue R Denatured alcohol 2'-Deoxynucleoside 5'-triphosphates (dNTP kit) Diethyl pyrocarbonate N,N-dimethyl formamide (DMF) Dimethyl sulfoxide (DMSO) Dithiothreitol (DTT) Ethanolamine Ethidium bromide Formaldehyde solution Gelatin Glutathione Glycerol Glycine Guanidine thiocyanate Hydrochloric acid N-(2-Hydroxyethyl)piperazine-N'-(2-ethanesulphonic (HEPES) N-Lauroyl sarcosine Liquid paraffin Lithium chloride anhydrous Magnesium acetate tetrahydrate Magnesium sulphate (MgSO -7H 0) Magnesium chloride (MgCl2-6H 0) 3  4  2  2  acid)  Fisher Scientific Applied Biosystems Fisher Scientific/ICN Sigma Gibco BRL BRL BDH Fisher Scientific Fisher Scientific Fisher Scientific Fisher Scientific Fisher Scientific Amersham Sigma Sigma Sigma Fisher Scientific Fisher Scientific ICN Sigma Sigma BDH Fisher Scientific E M Science Fisher Scientific Pharmacia Sigma Sigma/Fisher Scientific Fisher Scientific BDH Sigma Molecular Probes Inc. Fisher Scientific/BDH BioRad/Sigma Sigma Anachemia Sigma/Fisher Scientific ICN Fisher Scientific Sigma Sigma BDH BDH/Sigma/Fisher Scientific BDH Fisher Scientific Fisher Scientific  76  Table continued.. Maltose Manganous chloride (MnCl -4H 0) 2-Mercaptoethanol Methanol 1-Methyladenine DL-threo-(3-methylaspartic acid 2-[N-Morpholino]ethanesulfonic acid (MES) 3-[N-Morpholino]propanesulphonic acid (MOPS) Myelin basic protein (MBP) N-ethyl maleimide Ninhydrin Nitric Acid (HN0 ) Nitro blue tetrazolium (NBT) Nonidet P-40 Petroleum ether (60-80°) Phenyl phosphate disodium salt (phosphatase inhibitor) Phenolsulfonphthalein (Phenol red dye) Delbucco's phosphate-buffered saline (PBS) Phosphoric acid Potassium acetate (C2H3O2K) Potassium chlorate (KC1) Potassium dichromate (K Cr20 ) Potassium dihydrogen orthophosphate monobasic Potassium hydroxide (KOH) Potassium phosphate (dibasic) di-Potassium hydrogen orthophosphate 3-hydrate Potassium dihydrogen orthophosphate (KH2PO4) Ponceau S concentrate Propanol Pyridine Silver nitrate (AgN0 ) Sodium acetate (dibasic) Sodium azide Sodium bicarbonate (NaHC03) Sodium borate Sodium carbonate - anhydrous (NaC03) Sodium chloride tri-Sodium citrate Sodium deoxycholate Sodium dihydrogen orthophosphate (NaHP0 -H 0) Sodium dodecyl sulfate (SDS) Sodium fluoride di-Sodium hydrogen orthophosphate (Na HP0 ) Sodium hydroxide Sodium orthovanadate (Na3V04) TEMED (N,N,N',N'-Tetramethylethylenediamine) Tris hydroxylmethyl aminomethane hydrochloride (Tris-Cl) Tris hydroxylmethyl methyl ammonium chloride (Tris-Cl) 2  2  3  2  7  3  4  2  2  4  BDH BDH BioRad Fisher Scientific/BDH Sigma Sigma Sigma Sigma/ICN Kinetek/Sigma Sigma BDH Fisher Scientific Sigma BDH BDH ICN Sigma Gibco BDH Fisher Scientific BDH BDH BDH Sigma BDH BDH Sigma Fisher Scientific Fisher Scientific Fisher Scientific BDH Fisher Scientific Fisher Scientific Fisher Scientific BDH Fisher Scientific BDH BDH BDH Fisher Scientific BDH/Fisher Scientific BDH Fisher Scientific Fisher Scientific Fisher Scientific Fisher Scientific BDH  Table continued. Tris hydroxylmethyl methylamine Triton X-100 Tween-20 Urea Zinc Chloride  BDH/Fisher Scientific BDH/Fisher Scientific Fisher Scientific BioRad BDH  B. Miscellaneous reagents Ampicillin (D [-] -a-Aminobenzylpenicillin Bind silane Counting scintillant (biodegradable) DNA lkb ladder DNA herring sperm Geneclean kit Isopropyl (3-D-thiogalactopyranoside (IPTG) PKI - cAMP-dependent protein kinase peptide inhibitor Prestained SDS-PAGE standards Random Primers 5'pd(N)6 Repel silane RNase Guard RNasin Sephaglas Bandprep Kit  Sigma LKB Amersham Gibco BRL Boehringer Mannheim Bio 101 Fisher Scientific/Promeg Sigma Kinetek/BioRad Pharmacia Pharmacia Pharmacia Promega Pharmacia  C. Consumables Centricon tubes (10 and 30) 2070 Conical tubes (50 ml) 2059 Culture tubes (14 ml) disPo/ Culture Tubes (13 x 100 mm) GeneAmp Reaction Tubes Hybond-N hybridization membrane Immobilon P (PVDF) 3MM chromatography paper Membrane filter unit Micro centrifuge tubes (1.5 ml) Nitrocellulose transfer and immobilization membrane P81 phosphocellulose chromatography paper Pipet tips (1-200 ul) and (1-1000 ul) Syringe filter unit Tissue culture dishes (90-mm and 150-mm)  Amicon Falcon Falcon Baxtor Perkin Elmer Cetus Amersham Millipore/Dupont Whatman Nalgene Elkey/Eppendorf Schleicher and Schuell Whatman National Scientific Nalgene Corning  78  D. Photographic supplies Developer Fixer ISO 3000 Polaroid film 667 ISO 100 Polaroid film Reflection NEF-Autoradiography film X-OMAT AR Imaging film  Kodak Kodak Polaroid Polaroid DuPont Kodak  which were naturally arrested at the G / M border of meiosis I, were isolated and prepared 2  as previously described (Meijer et a l , 1984). Ovaries were removed from the sea stars by excision and were kept in cold natural sea water (4°C). The oocytes were released by gently rupturing the ovarian sacks with a sharp instrument and forceps. Free oocytes filtered through a sieve to remove large debris (eg. connective tissue and gut) were combined into large 500 ml bottles for centrifugation at 1,000 rpm (125 g) for 5 min. Normally a yellowish upper layer of broken oocytes and follicle cells appears above the pink layer of packed oocytes which is discarded with the supernatant. The oocytes were carefully washed by resuspending the pellet several times in cold natural sea water to remove the remaining contaminants. Following several washes, one litre of oocytes was added to 3,000 ml of natural sea water containing 10 uM 1-methyladenine.  After  incubation at 14°C for approximately 80 min post-hormone treatment, 80-90% of the oocytes underwent germinal vesical breakdown (GVBD), indicating that they had completed maturation. Cell cycle progression was monitored using a Leitz Labovert FS microscope at 25X magnification. The mature oocytes were concentrated by centrifuging 500 ml volume at 1,000 rpm (125 g) for 5 min. A 40% oocyte homogenate was prepared in chilled homogenization buffer A containing 20 mM MOPS [pH 7.2], 0.25 mM DTT, phosphatase inhibitors (50 mM (3-glycerolphosphate, 1 mM NaVC^ 5 u M |3-methyl aspartic acid), and several broad specificity protease inhibitors that are described in Table 2 (5 mM EGTA, 2 mM EDTA, 1.0 mM PMSF and 1.0 mM benzamidine) by applying two 30 sec bursts at 18,000 rpm using a Brinkmann Polytron PT300. Following cell disruption, cellular organelles (mitochondria, endoplasmic reticulum etc.) were removed by centrifuging for 10 min at 9,000 rpm (12,000 g) with a Beckman J2 HS centrifuge and its companion JA10 rotor. The combined post mitochondrial supernatant was pooled on ice and subsequently centrifuged at 40,000 rpm (100,000 g) for 30 min at 40°C using a Sorvall ultracentrifuge and T647.5 rotor to separate the cell membanes. The pooled  Table 2: Specificity of protease inhibitors. Aprotinin Benzamidine Ethylene bis (oxyethylenenitrilo) tetraacetic acid (EGTA) Ethylene diamine tetraacetic di sodium salt (EDTA) Leupeptin Phenyl methylsulphonyl fluoride Soybean trypsin inhibitor  Serine protease inhibitor Peptidase inhibitor Metalloprotease inhibitor (divalent cation-dependent) Metalloprotease inhibitor (divalent cation-dependent) Serine and thiol proteases Serine and thiol proteases Trypsin and Factor Xa  Sigma ICN Fisher Scientific Fisher Scientific Sigma/ICN Sigma Sigma  cytosolic supernatants were subsequently transferred to 50-ml conical tubes and frozen at -70°C.  2.2.1.2 1-Methyladenine injection  A second technique employed for isolating sea star oocytes involved inducing their maturation in vivo by injecting the hormone directly into the live animal. This approach proved extremely convenient since fewer volunteers were required for harvesting the oocytes and more environmentally friendly, because the sea stars were returned live to the waters in the Greater Vancouver area. Spawning in live gravid adult females was carried out according to the protocol described by Meijer et al. (1984). A minimum of 1 ml of 0.14 mM 1-MeAde diluted in sea water was injected into each sea star arm. Sea stars began shedding usually within 90 min after the first injection; if not, then a second injection of 1-MeAde was given. Once the sea stars began spawning they were inverted over 250- or 400-ml plastic beakers filled with 20 ml of natural sea water. The nearly 100% mature oocytes were collected, concentrated by gentle centrifugation at 1,000 rpm (125 g) for 5 min finally resuspended to 40% (vol/vol) with homogenization buffer. All subsequent steps were executed as described above.  2.2.2  Cell culture  The human hepatocellular carcinoma cell line Hep G2 and the epidermoid cell line A-431 were obtained from American Type Tissue Collection (Rockville, Maryland). The Hep G2 cells were maintained in Eagle's minimum essential medium (10% fetal bovine serum, Earle's salts, and 5 uM 2-mercaptoethanol) (Charest et al., 1993). Hep G2 g  cells were grown in monolayers to a confluency of 10 (eight to ten 150 mm plates) before treating them with 100 uM insulin for 5 min at 37°C. The cells were scraped from  the plate with a rubber policeman and collected by centrifugation 3,000 rpm (-800 x g) using a Heraeus Biofuge 15 microcentrifuge. The cells were resuspended in 2 ml of homogenization buffer A and lysed by sonication using a Vibra-cell sonicator at setting 45.  Using a Beckman TL100 Ultracentrifuge, the homogenate was clarified by  centrifugation at 70,000 rpm (175,000 x g) for 30 min and stored at -70°C until needed. The A431 cells were maintained in Dulbecco's modified Eagle medium (5% fetal bovine serum, 5% fetal calf serum and 5 uM 2-mercaptoethanol).  3.  BIOCHEMICAL TECHNIQUES  Table 3 outlines the kinases, phosphatases and proteases used during the course of this work. Several of the kinases and phosphatases were the generous gifts of many researchers.  3.1  Determination of protein concentrations  The protein concentration was determined by the Bradford (1976) method. Ten or twenty microlitres of crude cytosolic extract or column fraction were mixed with 2.5 ml of Bradford reagent (100 mg/1 Coomassie brilliant blue G-250) and was allowed to stand for 5 min before reading the colorimetric change at an optical density of 595 nm. Varying concentrations of BSA (1 mg/ml) were used as a standard. In the case of recombinant GST-fusion proteins, 20 ul of glutathione-GST beads were subjected to a 10% SDS-PAGE and visualized using Coomassie blue staining (see later). Varying concentrations of BSA electrophoresed side-by-side on the same gel were used as standards.  Table 3: Sources of kinases, phosphatases and proteases Acid phosphatase CD45 protein tyrosine phosphatase from human spleen a-Chymotrypsin Protein tyrosine phosphatase (3 from human (HPTP(3) Lck from baculovirus infected Sf9 cells Protein phosphatase 2A (PP2A) Lysozyme Trypsin sequencing grade Thrombin  Sigma Dr. Nicholas Tonks Sigma Dr. Ken Harder Dr. Ruedi Abersold Dr. David Brautigan Boehringher Mannheim Sigma Sigma  3.2  Column chromatography of sea star cytosolic extracts  3.2.1  Anion exchange chromatography  A list of the chromatography resins, their properties and their commercial sources is provided in Table 4. The most common analytical chromatography columns used during the course of this study are described below.  A protocol, used by all members of Dr. Steve Pelech's laboratory was created for routine fractionation and analysis of crude cytosolic extracts. This approach fulfilled two important purposes. First, the elution profiles of several different protein seryl/threonyl kinases have been well characterized after many years of biochemical analysis. Second, the activation of these kinases can be compared in such disparate model systems as echinoderms and mammalian cells. Approximately 1-2 mg of protein were applied at a flow rate of 0.8 ml/min to a Mono Q column (HR5/5) equilibrated with buffer B (20 mM MOPS [pH 7.2], 5 mM EGTA, 2.5 mM EDTA, 100 uM N a V 0 , 1 mM NaF, 25 mM 03  glycerophosphate, 5 uM |3-methyl aspartic acid and 2 mM DTT). The column was eluted at the same flow rate using a 15 ml linear NaCl gradient (0-0.8 mM) (Figure 5A). Two hundred and fifty microlitre fractions were collected and analyzed for enzyme activity and/or immunodetection with specific antibodies.  3.2.2  Cation exchange chromatography.  Samples containing 1-2 mg of protein were diluted with 10 volumes of buffer C (20 mM MES [pH 6.5], ImM EGTA, ImM EDTA, 100 uM NaV0 , 1 mM NaF, 25 mM 3  P-glycerolphosphate, 5 uM (3-methyl aspartic acid and 2 mM DTT) and subsequently loaded at a flow rate of 0.8 ml/min onto a Mono S (HR5/5) column equilibrated with the  Table 4: Chromatography resins used for protein analysis and purification. DEAE-cellulose Heparin-agarose Hydroxylapatite Mono Q (HR5/5) Mono S (HR5/5) Phosphocellulose P l l Polylysine-agarose S-Sepharose Superose 12  Fibrous anion exchanger Cation exchanger Ionic differential surface binding Strong anion exchanger Strong cation exchanger Fibrous cation exchanger Anion exchange Strong cation exchanger Gel filtration  Sigma Sigma BioRad Pharmacia Pharmacia Whatman Sigma Pharmacia Pharmacia  same buffer. The column was developed with a 15 ml linear NaCl gradient (0-0.5 mM) and the eluate was collected over a range of 50 fractions at 300 ul each (Figure 5B). A small sample from alternate fractions was tested for enzyme activity and/or Western blotting.  3.2.3  Gel filtration chromatography  The Superose 12 column (HR10/30) equilibrated with buffer B containing 150 mM NaCl was loaded with 200 ul of purified protein (approximately 1-2 ug) at a flow rate of 0.25 ml/min. Once the void volume of 27.5 ml had eluted from the column, 200 ul fractions were collected and assayed for enzyme activity. Protein molecular mass standards used for calibrating the column were: Blue Dextran, 443-kDa; alcohol dehydrogenase, 150 kDa; bovine serum albumin, 67 kDa; and P-lactoglobulin, 35 kDa.  Please note that all chromatography procedures described in this work were performed in a refrigeration unit at or near 4°C with a programmable Pharmacia Fast Protein Liquid Chromatography (FPLC) system.  3.4.  Purification of sea star MAP kinase activator-like protein  A schematic flow chart of the MAP kinase activator purification is outlined in Figure 6. Sea star cytosolic extract (100 ml; ~2 g of protein) prepared as described in Section 1 was thawed and diluted to 2,000 ml with buffer D (20 mM MOPS [pH 7.2], 12.5 mM (3-glycerolphosphate, 10 mM EGTA, 2 mM EDTA, 100 uM N a V 0 , 1 mM 3  NaF, 5 uM P-methylaspartic acid, and 1 mM DDT). The diluted homogenate was applied to a DEAE-cellulose column (5 cm x 7 cm) and hydroxylapatite column (2.5 cm x 5 cm) linked in series. Hence, the flow through material from the DEAE-cellulose  Figure 5: Mono Q and Mono S column elution profiles. Panel A. The Mono Q column was developed with a 15 ml linear NaCl (0-0.8M) at a flow rate of 0.8 ml/min into 50250 ul fractions. Panel B. The Mono S column was developed with a 15 ml linear NaCl (0-0.8M) at a constant flow rate of 0.8 ml/min into 50-300 ul fractions.  A  column was loaded directly onto the hydroxylapatite column. After washing the two columns with 300 ml of buffer D at 4 ml/min, the hydroxylapatite was uncoupled and developed separately with a 300 ml linear gradient of potassium phosphate (0-200 mM) at a flow rate of 1.5 ml/min. Four millilitre fractions were collected and tested for MAP kinase kinase activity using the M A P kinase activator assay (Section 3.5.2). The activator activity eluted between 25-60 mM NaCl.  The fractions that were able to increase recombinant GST-erkl activity toward MBP from the hydroxylapatite column were combined and diluted to 1:4 with buffer D prior to loading an S-Sepharose column (2.5 cm x 5 cm) at 1.5 ml/min. The column was washed, at the same flow rate, with 90 ml of buffer D before collecting 4 ml fractions using a 300 ml linear 0-500 mM NaCl gradient in buffer D. Maximal activator activity was detected at a concentration of 100-135 mM NaCl.  The fractions from the S-Sepharose column were pooled, diluted 1:1 with buffer D, and applied to a phosphocellulose column (2.5 cm x 3 cm) that had been equilibrated in buffer D. After washing the column with 50 ml of buffer D, the MAP kinase activator was collected in 3 ml fractions from the phophocellulose column that had been developed with a 195 ml linear 0-800 mM NaCl gradient in buffer D. The peak activity fractions eluted in the range of 225-325 mM NaCl.  The active MAP kinase activator fractions were combined, diluted to 1:3 with buffer D, and loaded at 1.5 ml/min onto a heparin-agarose column (2.5 cm x 3 cm). The column was washed with 50 ml of buffer D. The column was developed with a linear 0200 mM NaCl gradient in buffer D. Three millitre fractions were collected and the MAP kinase activator was eluted in the range of 125-155 mM NaCl.  Figure 6: Purification of sea star MAP kinase activator. The sea star MAP kinase activator was purified using a combination of ion exchange chromatographic resins. The concentration of salt required to elute the kinase activity is shown for each column.  Sea Star Cytosolic Extract 2 g diluted with 21 bufferD DEAE-celliiose Breakthrough material loaded directly Hydroxylapatite Acti vity eluted with 225-325 mM NaCl S-Sepharose Acti vity eluted with 100-135 mM NaCl Heparin-agarose Activity eluted with 125-155 mM NaCl  t MonoS  Activity eluted with 190 mM NaCl Biochemical and immunological analysis  The fractions containing activator activity were dialyzed against buffer E (20 mM MES [pH 6.5] and all the components of buffer D except the MOPS) for 2.5 h at 4°C. The dialysate was applied at 0.8 ml/min onto a Mono S column equilibrated with buffer E. Using the same flow rate, the column was washed with 3 ml of buffer E before application of a 15 ml linear 0-500 mM NaCl gradient in buffer E. The 300 ui fractions that were assayed for MAP kinase activator activity revealed that the enzyme eluted at a concentration of -190 mM NaCl.  3.5.  Protein kinase assays  3.5.1  Single-step seryl/threonyl protein kinase reaction  The MAP kinase phosphotransferase activity was measured using the exogenous substrate MBP or peptides patterned after the MBP Thr-97 phosphorylation site (ClarkLewis et al, 1991). Details of the filter paper assay have been published (Pelech et al, 1988). Briefly, the enzyme assays were carried out in a reaction volume of 25 ul: 5 ul column fraction or crude sea star extract, 15 ul phosphorylation buffer F (1 mg/ml MBP or 2 mM MBP peptides, 20 mM MOPS [pH 7.2], 30 mM (3-glycerophosphate, 20 mM M g C l , 5 mM EGTA, 2 mM EDTA, 0.5 mM N a V 0 , 1 mM DTT and 500 nM PKI). 2  3  The reaction was initiated by addition of 5 ul of 50 uM [y- P]ATP (-2,000 cpm/pmol) 32  and was incubated at 30°C for 5 min. The reaction was terminated by application of 20 2  ul of the reaction to a 2 cm P81 phosphocellulose paper. Once the unincorporated [y32  P]ATP was removed from the filter papers by repeated washing (10-20) using 1%  phosphoric acid, they were transferred to 6 ml scintillation vials containing 200 ul of scintillation fluid and counted in a Wallac 1410 Liquid Scintillation Counter.  3.5.2  MAP kinase activator assay  Figure 7 schematically outlines the two-step MAP kinase kinase assay. The principal of the assay was to assess MAPKK activity indirectly by phosphorylating and activating the fusion protein GST-Erkl immobilized on glutathione-agarose beads. In turn, the activated GST-Erkl then phosphorylated MBP as a substrate. The kinase reactions were performed according to the protocol described by Charest et al. (1993). Each 70 ul reaction consisted of 20 ul packed GST-Erkl glutathione beads (1-2 ug) combined with 40 ul of column fraction and 10 ul of phosphorylation buffer G (20 mM MOPS [pH 7.2], 10 mM M g C l , 10 mM MnCl , 2 mM NaF, 5 uM p-methyl aspartic 2  2  acid, 25 mM p-glycerolphosphate, 1 mM DTT, 10 (ig/ml aprotinin, 10 ug/ml SBTI, 10 Hg/ml leupeptin and 50 uM ATP). The kinase reactions commenced upon placing them at 30°C. At the completion of the 20 min incubation period, the supernatant was removed and the beads washed several times with chilled phosphorylation buffer G containing 0.1% Triton X-100 and 150 mM NaCl. The reaction was continued by addition of phosphorylation buffer H (20 mM MOPS [pH 7.2], 10 mM MgCl , 1 mM MnCl , 2 mM 2  2  NaF, 1 mM NaV03 P-methyl aspartic acid, 25 mM p-glycerolphosphate, 1 mg/ml MBP) and 50 mM [y- P]ATP (-2,000 cpm/pmol) in a final volume of 50 ul. The incubation 32  proceeded for 20 min at 30°C and was terminated by spotting 25 ul of the reaction mixture onto a P81 phosphocellulose paper and analysed as described for the MAP kinase assay.  Figure 7: MAP kinase activator assay. The bacterially expressed protein erkl linked to glutathione S-Sepharose beads fusion protein  (^^)—fj  ERK1  through its glutathione-S-transferase  |—was used to assay for MAP kinase activator activity.  The bound acti-  vator < A caused an increase in phosphorylation and consequentiy activation of Erkl kinase p  p  U ERK1 /  P  A  in vitro substrate MBP  The activated Erkl  P  _AZ.  1_  E  R  K  1  /  w a s  then combined with its  and [v -P] ATP . The phophorylated MBP 32  spotted onto a P81 chromatography paper and counted.  r±j was  Bacterial lysate  1 - Affinity Purification  2 - Erkl activation  ERKl^  1 r x x  glutathione S-Sepharose beads  P  Column Fraction  V .  50 UM ATP 10 mM M n 10 mM M g  c  m  P  ERK1 ^ A P  P  \y -  +  E  RK1<^ A  2 + 2  Incubate 20 min at 30°C  Decant Supernatant  3 - Washing  4 - MAP kinase assay P  P P  25 (ig MBP p p  cb  p  "  D-  "x,  P  \ /  [y-32p] ATP  P  \y  ERKl/ P  ERKl/  2000 cpm/pmol Lb  1=1  Incubate for 20 min at 30°C  P  f  A  P p  P  3 times with buffer B + 0.1% Triton X-100 and 150 mM NaCl  Lb  r_b  rzzi Apply 25 ul of the reaction to a P81 filter paper  <3  <3  96  3.5.3  Autophosphorylation  3.5.3.1 Recombinant proteins  Autophosphorylation experiments were performed with recombinant GST-fusion proteins linked to glutathione-agarose beads or thrombin-cleaved GST-fusion proteins (0.5-2.0 fig), 50 mM [y- P]ATP (9,000 cpm/pmol) and phosphorylation buffer I (25 32  mM sodium p-glycerolphosphate, 20 mM MOPS [pH 7.2], 10 mM M g C l and/or 10 mM 2  M n C l , 2 mM NaF, 1 mM DTT, 1 mM N a V 0 10 ug of aprotinin per ml, 10 ug of 2  3  soybean trypsin inhibitor (SBTI) per ml and 5 ug of leupeptin per ml) in a final reaction volume of 40 ul for 30 min. The reaction was terminated by addition of 30 ul of 5X concentration sodium dodecyl sulfate (SDS) sample buffer J (125 mM Tris-HCl [pH 6.8], 4% SDS, 0.01% bromophenol blue, 10 mM (3-mercaptoethanol, and 20% glycerol) and boiling for 5 min (Laemmli, 1970).  3.5.3.2 Purified sea star activator  Sea star M A P K K was tested for autophosphorylating activity by incubation of the most purified preparation of the activator (1-2 ug of protein) with phosphorylation buffer H without MnCl . The 40 ul reaction continued for 20 min at 30°C before addition of 2  sample buffer E.  3.6  3.6.1  Protein phosphatase assays  Protein tyrosyl phosphatase assays  Wild type GST-Erkl protein (100 ug) linked to glutathione-agarose beads was equilibrated by washing several times in dephosphorylation buffer K (25 mM Tris-HCl [pH 7.0] and 10 mM DTT) prior to being diluted 1:1 (vol/vol) in the same buffer containing 100 units of CD45 protein tyrosine phosphatase (see Table 3) The reaction was incubated for 60 min at 30°C. A time course of inactivation was performed by removing 40 ul of the bead slurry and combining it with 1 mM of the protein tyrosine phosphatase inhibitor sodium orthovanadate. The beads were washed with buffer G and half of the material was used to assay MBP phosphotransferase activity (see Section 3.5.1) while the remainder of the beads were used for Western blotting analysis using the anti-phosphotyrosine antibody 4G10 (see Section 3.9.3).  3.6.2  Protein seryl/threonyl phosphatase assays  The GST-Erkl beads were equilibrated with dephosphorylation buffer L (20 mM MOPS [pH 7.2], 1 mM DTT, 1 mM EDTA and 0.1 mg/ml) before being resuspended in the same buffer containing ~4 U of human protein phosphatase 2A (1 U releases 1 nmol of phosphate/min from 15 uM phosphorylase at 30°C). The time-course of inactivation proceeded for 60 min at 30°C. During this period 40 ul of slurry was removed from the reaction at 10 min intervals and combined with 1 uM okadaic acid to inhibit the phosphatase reaction. The beads from each measurement in the time-course were washed with buffer I before being assessed for MBP kinase activity (see Section 3.5.1) or immunoreactivity with 4G10 antibody (see Section 3.9.3).  Table 5: Polyclonal antibodies  Antibody MAP Kinase Rl MAP Kinase R2 MAP Kinase R3 MAP Kinase mpk-l MAP Kinase GEGA MAPKK-XI  Peptide Antigen Location 63-98 Rat 337-367 Rat 8-32 Rat unknown Sea star unknown Sea star 335-356  Subdomain Region  STE7-VIII  360-378 Rat 434-446  Reference  PFEHQTYCQRTLREIQIL LGFRHENVIGIRDILRAP PFTFDMELDDLPKERLK ELIFQETARQFPGAPEAP GGGGGEPRRTEGVGPG VPGEVEMVK N.A.  Boulton et al., 1990  GLAYIGEGAYGMV  Posada etal., 1991 Kosako et al., 1993  VIII  EFQDFVNKCLVKNPAER ADLKQ HSFIKQSELEEVDFAGW LC FVGTSTYMSPERIC  CT  CQDRPPSRELLKHPVFR  Lange-Carter et al., 1993 Rhodes et al., 1990  III CT NT complete protein I XI  Xenopus  MEK-CT  Amino Acid Sequence  CT  Boulton et al., 1990 Boulton et al., 1990 Sanghera et al., 1991  Kosako et al., 1993 Teague et al., 1986  S. cerevisiae  MEKK-CT STE11-II  667-684 Rat 464-474 S. cerevisiae  II  HTGELMAVKQVC  3.7  Antibody production  Kinetek Pharmaceuticals Inc. provided many of the anti-peptide antibodies used during the course of this study including a-MAPK, a - M A P K K a-Mekk and a-Raf (Table 5). The short polypeptides used to generate the antibodies were synthesized by Dr. Ian Clark-Lewis' research group at the Biomedical Research Centre (University of British Columbia), Darryl Hardie at the University of Victoria or supplied by Upstate Biotechnology Incorporated (Lake Placid).  3.7.1  Antigen preparation  Each synthetic peptide was conjugated to the carrier protein K L H (keyhole limpet hemocyanin) through their C- or N-terminal cysteine residues using the conjugating agent SMCC (succinimidyl 4-(N-maleimidomethyl) cyclohexane-l-_arboxylate). This usually created a peptide concentration of 5.3 mg/ml. Peptide dimerization through cysteine residues was reduced by first dissolving the lyophylized powder in PBS containing 4 M guanidine hydrochloride pH 7.5 and 5 mM DTT (see protocol described in Appendix 1 and 2). After removing the excess DTT by centrifuging through a Sephadex G10 spincolumn at 1,000 rpm (600 x g) for 3 min, the carrier/peptide reaction mixture was allowed to couple with gentle stirring overnight. Peptides crosslinked to K L H are then stored at -70°C. The initial rabbit immunization involved injecting equal volumes of complete Freunds adjuvent with 250 ug/ul of antigen diluted in PBS. However, incomplete Freunds adjuvent was used for subsequent injections of decreasing peptide antigen concentrations of 200 ug/ul and 100 ug/ul respectively.  A l l rabbit  immunizations were performed by trained animal technicians at the University of British Columbia's South Campus Animal Care Facility.  3.7.2  Antibody purification and quantitation  Appendix 3 provides a detailed description of the antibody purification proceedure. Approximately 30 ml of blood was collected from the rabbit's ear and was stored at room temperature to allow partial clotting. The clot was then removed by centrifugation at 2,500 rpm (-1300 x g) for 5 min using a Du Pont Sorvall RT 6000D bench-top centrifuge. The serum was centrifuged a second time to remove any remaining erythrocytes. Large impurities were removed by sequentially filtering the serum through a 0.45 um filter followed by a 0.22 um filter. The serum was added to 1-2 ml of thiollinked Sepharose peptide affinity beads. The complex formation between peptide and antibody proceeded overnight at 4°C with continuous mixing using a vertical rotating wheel. After the incubation period was complete, an affinity column was prepared from the bead/serum mixture. The column was washed with one bed volume of PBS and subsequently eluted with 5 ml of 100 mM glycine [pH 2.5]. The 1 ml eluates were immediately neutralized with saturated Tris-base. The antibody titre was determined by the enzyme-linked immunosorbent assay (ELISA) using a Bio-Tek Instruments EL 312e Bio-Kinetics Reader.  3.7.3  Other antibody sources  All the monoclonal and several of the polyclonal antibodies used during the course of this study were generous gifts from other research laboratories or were purchased from commercial sources.  3.8  One dimensional gel electrophoresis  The analytical electrophoresis of proteins by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) utilizes the discontinuous buffer system. The sample and stacking gel contain Tris-HCl [pH 6.8], upper and lower reservoirs contain Tris-glycine [pH 8.3], and resolving gel contains Tris-HCl [pH 8.8] described by Ornstein (1964) and Davis (1964) in combination with denaturing conditions (strongly anionic detergent SDS, reducing agent (3-mercaptoethanol and heat) devised by Laemmli (1970). A Hoeffer gel electrophoresis system was used to cast 1.5 mm thick separating gels (0.25 M Tris-HCl [pH 8.8], 0.1% SDS). By varying the acrylamide concentration (9-12%) (see Appendix 5 for details) it was possible to analyze medium (40 kDa) or large (80 kDa) molecular mass proteins. A 15 well 4 % acrylamide stacking gel (0.25 M Tris-HCl [pH 6.8], 0.1% SDS) was used to load the samples prepared from 80 ul of eluted column fraction or 20 ul GST-fusion proteins linked to glutathione-agarose beads after they were boiled in 5X SDS sample buffer J for 5 min. The proteins were electrophoresed for 16 h at 10 mA in Tris-glycine electrophoresis buffer K (0.25 mM Tris-base, 250 mM glycine [pH 8.3], and 0.1% SDS). For rapid protein analysis, the BioRad mini-gel system was used routinely (summarized in Appendix 5). A maximum sample volume of 80 ul was electrophoresed for 1.5 h at 400 V and 150 mA.  3.9  Protein visualization  3.9.1  Coomassie blue staining  Coomassie blue staining was used primarily for detecting and quantitating the expression levels of recombinant GST-fusion proteins purified from E. coli. At the  completion of the electrophoresis, the polypeptides separated in the SDS-polyacrylamide resolving gel were simultaneously fixed with methanol and acetic acid and stained with Coomassie blue (methanobglacial acetic acid:water [45:10:45 by volume] and 0.25% Coomassie Brilliant Blue R250 [wt./vol.]). The gels were immersed in 5 volumes of staining solution and placed on a rotating plateform to assure complete saturation of the gel. Typically the immersion lasted for 10-20 min at room temperature.  The protein  bands were revealed by removing excess stain from the gel with destain buffer (10 % MeOH and 10 % glacial acetic acid [vol/vol]). The gel was preserved by drying it between two sheets of cellophane using a Bio Rad gel slab drier.  3.9.2  Silver staining  Silver staining technique involves the differential reduction of silver ions bound to the polypeptide animo acid side chains. This procedure is 100- to 1000-fold more sensitive than staining with Coomassie Brilliant Blue R250. Due to its detection capabilities (minimum capacity in the 0.1-1.0 ng range) the silver staining method was used for assessing the purity of the protein preparations.  Appendix 6 outlines the  protocol described by Merril et al. (1981). The proteins were separated through an SDSpolyacrylamide gel electrophoresis as described in Section 9 and sequentially fixed in methanol:acetic acid:water (40:10:50 by volume) then ethanol:acetic acid:water (10:5:85 by volume). To improve protein affinity for the silver reagent, the gels were placed in an oxidizing solution containing 3.4 mM K^Ci^Oy and 3.2 mM nitric acid prior to being incubated with a silver nitrate solution (0.2 % AgN03 [wt/vol]). The fixed protein bands were revealed by reducing the bound AgNO-3 in a developer solution (2.5% sodium carbonate [wt/vol] and 0.017% formaldehyde [vol/vol]). After the desired contrast was obtained, the reaction was quenched with a 5% acetic acid solution. The gel was preserved by drying as outlined in Section 10, i.  3.9.3  Western immunodetection  Western blotting is a very sensitive technique for detecting a specific protein that is present at low concentration in a crude sample of polypeptides. Proteins separated electrophoretically are transferred from the polyacrylamide gel to a solid support and probed with an antibody specific for antigenic sequences displayed within the polypeptide. Generally, Western analysis was performed routinely on recombinant GSTErkl and GST-Mekl expressed in E. coli; MAP kinase activator and MBP/HH1 kinase purified from sea star oocytes. Appendix 7 summarizes the procedure used for protein gel transfer and Western immunoblotting. Briefly, the electrophoresed proteins were electroeluted onto nitrocellulose for 3 h at 400 mA in transfer buffer L (120 mM glycine, 20 mM Tris-base [pH 8.6], and 20% methanol by volume) using an LKB 2005 Transphor Power Supply. For total protein detection the blot was immersed in Ponceau S stain for 5 min and rinced several times with water to reduce the background. Immunoreactive protein bands were revealed by Western blotting with polyclonal anti-peptide or fulllength protein antibodies. The membrane was blocked for 2 h at ambient temperature in TBST (50 mM Tris-base, 150 mM NaCl, and 0.2% Tween-20) with 5% dry milk powder. After the blocking step, the membrane was washed twice with TBST for 10 min and incubated overnight with a 1/256 000 titre rabbit primary polyclonal anti-peptide antibody diluted 1:500 in TBST. The membrane was washed extensively in TBST then incubated with goat anti-rabbit alkaline phosphatase conjugated secondary antibody diluted 1:2500 in the same buffer. Trace amounts of secondary antibody was removed with successive washes with TBST followed by a final wash with TBS to remove the detergent. The immunoreactive protein bands were revealed by placing the membrane in alkaline phosphatase development solution (100 uM NaHC03 [pH 9.8], 1 mM MgCl2, 0.03% NBT and 0.015% BCIP). The development continued until the desired colour intensity was obtained after which the membrane was placed in deionized water.  The apparent molecular masses of protein from Western blotting and SDSpolyacrylamide gels were estimated by comparison with Bio Rad or Kinetek prestained protein standards: phosphorylase b 100-kDa; BSA 66-kDa; ovalbumin 49-kDa; GST (glutathione-S-transferase) 27-kDa; SBTI (soybean trypsin inhibitor) 21-kDa; and lysozyme 17-kDa.  3.10  Phosphoamino acid analysis  The P radiolabeled protein samples were electrophoresed on an SDS-polyacrylamide 3 2  gel as outlined in Section 3.8. The electrophoresed proteins were transferred to Immobilon P membrane in buffer L for 4 h at 500 mA. The membrane was autoradiographed and the appropriate bands attached to the solid support were excised for phosphoamino acid analysis. The radioactive membrane was minced into manageable pieces and placed into a reacti-vial (Pierce) for digestion with 6 N constant boiling HC1. Longer incubations were avoided due to the instability of the phosphate moiety on the phosphoserine and phosphothreonine (Posada and Cooper, 1992). The acid-hydrolyzed sample was transferred to an Eppendorf tube and lyophilized under vacuum using a Labconco Centrivap Concentrator. After successive washing and drying steps to remove the HC1, the free P-labelled phosphoamino acids were resuspended in electrophoresis 32  buffer containing pyridine-acetic acid-water (1:10:189 by volume) and 1 ug each of phosphoserine, phosphothreonine and phosphotyrosine standards. Samples spotted onto an thin layer chromatography (TLC) cellulose plate were electrophoresed with cooling at 1,000 V. Ninhyndrin (0.2% in ethanol) was used to reveal the migration positions of the standards while autoradiography revealed the phosphoamino acid content of the radiolabelled protein. A detailed protocol is outlined in Appendix 8.  3.11 Two-dimensional phosphopeptide mapping  Electrophesis of P-labelled proteins was performed as described in Section 3.8. 32  After identifying the bands by autoradiography, the radiolabeled proteins were excised from dried SDS polyacrylamide gels. The gel slice was reswollen in methanol:acetic acid:water (10:10:80 by volume) and washed in methanol:water (50:50 by volume) before being dessicated in a Labconco Centrivap Concentrator. The protein was digested in a 50 mM ammonium bicarbonate buffer [pH 8.0] containing 100 ug of TPCK-treated (tolysulfonyl phenylalanyl chloromethyl ketone-treated) trypsin. Following the digestion period, the supernatant was vacuum dessicated. The tryptic peptides were washed several f  times with decreasing volumes of water interspersed with drying in the vacuum concentrator and finally resuspended in first dimension eletrophoresis buffer (see below) containing 0.5% phenol red before spotting on a TLC cellulose plate. First dimension electrophoresis was executed in acetic acid:pyridine:water (10:1:89 by volume) at 750 V and second dimension chromatography was performed in a solution containing 1butanol:acetic acid:water:pyridine (10:3:12:15 by volume). The TLC cellulose plate was air dried before autoradiography. The 2-D phosphopeptide mapping protocol is detailed in Appendix 9.  4.  MOLECULAR BIOLOGY TECHNIQUES  All molecular biology enzymes were available from commercial sources (Table 6).  The Erkl and Mekl oligonucleotide primers used throughout this study were  synthesized  on a Applied  Biosystems D N A Synthesizer  model  300A.  Table 6: Sources of DNA and RNA modifying enzymes AmpliTaq DNA polymerase Phosphatase, alkaline DNA polymerase I large fragment (Klenow) Polynucleotide kinase Superscript reverse transcriptase Ribonuclease A RNAsin T4 DNA ligase T7 DNA polymerase T7 sequencing kit Vent DNA polymerase  Perkin-Elmer Cetus Boehringer Mannheim New England Biolabs Boehringer Mannheim BRL Pharmacia Promega New England Biolabs Pharmacia Pharmacia New England Biolabs  4.1  RNA Isolation  Total cellular RNA was isolated according to the protocol described by Chomczynski and Sacchi (1987). The guanidinium thiocyanate-phenol-chloroform extraction is a rapid method for isolating undegraded RNA in high yield by eliminating the need for ultracentrifugation.  A l l aqueous solution were treated with diethyl  pyrocarbonate and autoclaved. All glassware, polypropylene microcentrifuge tubes and pipette tips, etc. were autocaved. Individually wrapped sterile disposable plasticware was utilized whenever possible.  For RNA isolation, A431 cells were grown in monolayers to a confluency of 10 /per 150 mm plate. 8  After washing the cells with ice-cold IX PBS (lacking  magnesium and calcium) they were removed from the cell culture plate by mechanical disruption with a rubber policeman and collected in a Falcon 2059 polypropylene tube by centrifugation at 3,000 rpm (-800 x g) for 5 min at 4°C. The supernatant was aspirated before addition of 1 ml of guanidinium thiocyanate RNA extraction buffer M per 140mm plate. The RNA was extracted from the cell homogenate by adding the following reagents sequentially with thorough mixing by inversion between each addition: 100 ul of 2 M sodium acetate, 1.0 ml water-saturated phenol and 200 ul of chloroform:isoamyl alcohol (49:1 by volume). The liquid suspension was shaken vigorously for 10 sec before storage at 4°C for 15 min. The aqueous layer containing the RNA was separated from the organic layer by centrifuging at 15,000 rpm (10,000 x g). After transferring the aqueous to a fresh tube it was combined with one volume of isopropanol and stored at 20°C for a minimum of 1 h. The precipitated RNA was sedimented at 15,000 rpm (10,000 x g) for 20 min. The pellet was then resuspended in a 300 ul volume of buffer M , transferred to a microfuge tube and reprecipitated with 1 volume of isopropanol at 20°C for at least 1 h. The RNA suspension was sedimented at 15,000 rpm (10,000) for  10 min at 4°C. The pellet was washed with 75% ethanol, vacuum dessicated and dissolved in 10 mM Tris-Cl [pH 7.4], 5 mM EDTA and 0.1 % SDS. The RNA preparation was stored by adding 0.1 volume of 3 M sodium acetate [pH 5.2] and 2.2 volumes of ice-cold absolute ethanol. A detailed protocol is outlined in Appendix 10.  4.2  PCR amplification of a partial human Erkl cDNA  A single stranded cDNA template was synthesized from A431 cell RNA. In a final volume of 20 ul, approximately 2.5 ug of total cellular RNA were combined with 1 mM nucleotide triphosphates, 20 units RNasin, and 100 pmol six base oligonuceotide random primers in PCR buffer N (50 mM KC1; 10 mM Tris-Cl, [pH 8.3] at 20°C; 1.5 mM Mg Cl; 0.1% gelatin). The reaction was initiated by addition of 200 U of Superscript 2  reverse transcriptase (10 min at 23°C; 45 min at 42°C; and 1 min 95°C). The specific coding and complementary oligonucleotides (sense strand of subdomain II, GTG GCT/C ATC A A G A A G ATC AGC CCC TTC GAG CAT; antisense strand of subdomain VII, CTC AGG GCT AGC AAT CCG GGC A A G GCC G/AAA G/ATC; and antisense strand of subdomain IX, GCA GCC CAC AGA CCA GAT GTC A/GAT GGA T/CTT GGT GTA) were synthesized based on the sequence published for rodent Erkl (Boulton et al, 1991) and modified for human codon preference described by Lathe (1985). The DNA/RNA hybrid was separated by heating the reverse transcriptase reaction at 95°C for 5 min before diluting the mixture with 100 ul of IX PCR buffer N supplemented with 100 pmol of each paired coding and complementary oligonucleotides. The DNA was amplified using 10 U of ampliTaq DNA and an automated programmable Perkin-Elmer Cetus thermal cycler. The cDNA was amplified using a two-step approach: the first three cycles occurred at low primer-annealling temperatures to provide more efficient priming (95°C for 30 sec, 37°C for 60 sec and 72°C for 120 sec) followed by 35 cycles at higher primer-annealling temperatures to allow increased specificity (95°C for 30 sec,  55°C for 60 sec and 72°C for 120 sec). Synthesis of complete blunt-end termini was ensured by subjecting the PCR mixture to a final elongation cycle (72°C for 5 min). An aliquot of the amplified DNA sample was analyzed by horizontal agarose gel electrophoresis and visualized by ethidium bromide staining (Appendix 11).  Bands displaying the appropriate size were excised and purified using a Sephaglas Bandprep™ Kit as described by Pharmacia's specifications. The purified fragment was then blunt-end ligated into the Eco RV restriction site of Stratagene Bluescript™ vector (Appendix 12). The ligation reaction (Appendix 13) was allowed to proceed for 24 h at 16°C before transforming the cDNA-vector into the competent E. coli strain X L . l Blue (Appendix 14). Recombinant clones were selected using an a-complementation system. In this method, the deletion mutant E. coli host strain constitutively expresses the 3' distal portion of the |3-galactosidase {lacZ) gene while the plasmid vector expresses the 5' proximal portion of the same protein from an inducible operator region. Complementation occurs when the isogropylthio-P-D-galactoside (IPTG) inducible plasmid-encoded fragment associates with the host-encoded fragment to form an active lacZ enzyme complex that is detectable in the presence of the chromogenic substrate 5bromo-4-chloro-3-indolyl-P-D-galacoside (X-gal). As a consequence the lac bacteria +  form blue colonies. In the case of recombinant bacteria, a fragment of foreign DNA inserts into the mutiple cloning site which is located within the lacZ gene of the plasmid vector. Inevitably, bacteria carrying plasmids with amino-terminal fragment insertions, results in the disruption of the lacZ enzyme, are usually unable to complement and as a result form white colonies. Positive clones were grown in 3 ml of 2 x YT bacterial growth medium (Appendix 15) which was ideal for promoting the amplification of the high copy number Bluescript™ plasmid. A small sample of the plamid (2-4 ug DNA) was isolated by the boiling lysis procedure adapted from Holmes and Quigley (1981) (Appendix 16) and the DNA prepared for sequencing using the Pharmacia L K B Deaza  Sequencing Mixes  . The cDNAs templates were sequenced by the enzymatic  dideoxy-mediated chain-termination method developed by Sanger et al. (1977) and electrophoresed on 6% vertical buffer-gradient polyacrylamide gel (Appendix 17).  4.3  Cloning and sequencing a full-length human erkl cDNA  The 400 base pair partial erkl cDNA cloned by PCR was used to probe a human Hep G2 cDNA X Zap library commercially available from Stragene Cloning Systems (Short et al, 1988). The technique was modified from the work of Benton and Davis (1977).  4.3.1 Preparation of plating bacteria  A single bacterial colony of X L . l Blue strain of E. coli was inoculated into 50 ml of LB medium supplemented with 0.2% maltose in a sterile 250-ml Erlenmeyer flask. The culture was placed in a rotary shaker (220 cycles/min) preheated to 37°C. The addition of sugar to the medium improved the efficiency of A, bacteriophage infection, since expression of the X receptor gene lamB is controlled by the maltose operon. The bacteria were sedimented at 7000 rpm (-4,000 g) for 10 min at room temperature. The supernatant was discarded and the cell pellet resuspended in 20 ml of sterile 0.01 MgSC<4 solution to an OD6oo=2 or -1.6 x 10 cells/ml. The cells could be stored for up to 3 9  weeks at 4°C.  4.3.2  in situ hybridization of bacteriophage X plaques  Since first being introduced in the late seventies (Sim et al, 1979), in situ hybridization has been used widely in molecular biology to detect, isolate and purify  DNAs from complex cDNA and genomic DNA libraries consisting of hundreds of thousands recombinant bacteriophage.  4.3.2.1 Plating  The titre of the Strategene human Hep G2 cDNA X Zap library was determined by plating 1:10 serial dilutions onto plating bacteria LE. coli strain X L . 1 Blue). Infection was performed by combining 200 ul of plating bacteria (-1.6 x 10 cells/ml) with 1 ul of 9  each of the bacteriophage serial dilutions. The infection was allowed to proceed at 37°C for 20 min. During this period hardened NZY top agarose (Appendix 15) was liquified by heating the bottle in a microwave. The hot liquified top agarose was cooled to 55°C in a warm water bath. At the completion of the infection period, 4-5 ml of molten NZY top agarose were added and the mixture was immediately poured and spread onto pre-heated NZY plates (the NZY plates were warmed to 37°C for at least 2 h with their lids left open slightly to allow the water to evaporate). After the plates were sufficiently dried, they were incubated overnight in a 37°C bacterial incubator.  4.3.2.2 Immobilization of bacteriophage X  The plates were incubated until the plaques reached a suitable size and density (the plaques usually attained a diameter of 1.5-2.0 mm before almost coming in contact with one another). The plates were then placed at 4°C for a least 1 h to allow the top agarose to harden.  The plates were numbered with a ethanol-resistant felt pen. Two nitrocellulose filters/Hybond nylon membranes were numbered identically to their corresponding +  plates using a soft-lead pencil. In addition the filters/membranes were further identified  by using the letters 'A or B'. The filter/membrane carrying the same number was carefully placed on the surface of the top agarose. The first filter/membrane (A) was marked asymmetrically in at least three different locations by puncturing through the filter into the underlying agar with a 20G 1 1/2 gauge needle which was immersed in black India ink. The filter/membrane was removed after 30-60 sec with a pair of bluntend forceps. A replica was created by placing a second filter marked B onto the agarose gel. To assure sufficient transfer of bacteriophage, the filter remained in contact with the plate for 1-2 min. Note that the excess ink left from the previous puncture also transferred to the filter/membrane.  4.3.2.3 Fixation of bacteriophage DNA to nitrocellulose  The phage were lysed and the target bacteriophage DNA was denatured in situ by placing the filter/membrane plaque side up onto 2-3 Whatmann 3MM papers saturated with denaturation solution (0.5 M NaOH, 1.5 M NaCl) for 30-60 sec.  The  filter/membrane were then transferred to a 3MM paper saturated with neutralizing solution (0.5 M Tris-Cl [pH 7.0], 1.5 M NaCl) for 1-2 min. The filter/membrane were then placed on dry 3MM to dry at room temperature for 30-60 min. The DNA was fixed to nitrocellulose filters by baking for 30 min at 80°C or to Hybond" " nylon membranes by 1  cross-linking for 30 sec with U.V. light.  4.3.2.4 Hybridization of immobilized X DNA with a P-labeled probe. 32  The filter/membrane was rehydrated in prewashing solution (5X SSC see Appendix 18, 0.5% SDS and 1 mM EDTA [pH 8.0]) for 30 min with agitation using a rotating platform in a water bath maintained at 42°C. The solution was changed two times. The solution was decanted and the filters/membrane were washed with three  changes of 2X SSC for 15 min. The filters/membranes were transferred to heat-sealable bags containing a sufficient volume of prehybridization buffer [(50% formamide, 5X Denhardt's reagent (6X SSC, 0.5% SDS and 100 ug/ml denatured, fragmented salmon sperm DNA (see Appendix 18)] to cover their surface (230 ul per square centimeter of nitrocellulose filter or nylon membrane). The bag was sealed in a manner that removed any bubbles. The bag was submerged in water bath with a rotary platform for 1-2 h at 42°C. A P radiolabeled probe was prepared with a specific activity equal to 10 3 2  9  cpm/ug of DNA (Appendix 19) and denatured by heating for 5 min at 100°C. The hybridation was performed by addition of 1-2 ng denatured DNA/ml radiolabeled probe to the heat-sealable bag. The bag was submerged in a 42°C rotary water bath. At the completion of the incubation period the filters were tranferred to a plastic Tupperware container and and excess radiolabel removed by washing 5 x 100 ml with 5X SSC at room temperature. The washing conditions were made increasingly more stringent by placing the filters in 2X SSC and raising the the incubation temperatures in incremental stages (37°C, 42°C, 47°C and 52°C). The filters/membranes were dried on paper towel and immobilized, using adhesive tape, onto a piece of 3MM paper cut to the size of a film cassette. Using an S radioactive ink pen, several asymmetric symbols were inscribed 3 5  on the sheet of 3MM paper which facilitated the alignment of the X-ray film with the positive spots located on the filters/membranes.  The filters/membranes were  autoradiographed overnight. Autoradiographs from the two separate plate lifts were aligned and plugs from the potential double-positive Erkl clones were excised and stored in 500 ul of SM buffer and 20 ul of chloroform. Secondary and tertiary screens were performed as described above to eliminate false double-positives and isolate a single pure X bacteriophage recombinant plaque. The positive recombinant bacteriophage contains plasmid sequences that can be excised in vivo and converted to the Bluesript SK (Ml3-) plasmid vector (Short et al, 1988). The putative positive Erkl clones were rescued by the method described by Stratagene and sequenced using plasmid-based T3 and T7  114  bacteriophage-specific oligonucleotide primers (Appendix 17). After confirming that the inserts were bona fide Erkl clones, the complete cDNA sequence was obtained in a stepwise fashion by using custom-designed oligonucleotide primers (Table 7) based on previous sequence results and walking along the cDNA in both directions.  4.4  Sub-cloning human Erkl into the pGEX-2T vector  Expression of Erkl as a GST-fusion protein required that the cDNA be ligated into the polyclonal site situated downstream of the glutathione-S-transferase gene in the pGEX-2T vector (Figure 12). Toward this end, forward and reverse oligonucleotide primers were synthesized based on the 5' and 3' coding region of the original Erkl 26a(3-3 clone (Table 7). Several additional bases were included at the 5' ends of each of the primers. First, an additional two bases were placed immediately upstream of the ATG start codon of the Erkl cDNA to assure that the Erkl protein was in the correct reading frame. Second, Eco R l restriction sites were placed 5' to the start and stop codons of the forward and reverse primers. And third, five additional bases were placed 5' to the Eco R l restriction site to facilitate binding of the Eco R l enzyme to the PCR amplified DNA fragment. The 26a(3-3 clone was linearized with Eco R l (Appendix 12) before addition of 100-500 pg of it to the reaction cocktail containing specific coding and complementary oligonucleotides (Table 9) 1 mM nucleotide triphosphates, 20 units RNasin, PCR buffer J and 10 U of high-fidelity VentR DNA polymerase in a final volume of 50 ul. Twentyfive rounds of amplification (95°C for 30 sec, 55°C for 45 sec, and 72°C for 90 sec) were performed with a Perkin-Elmer Cetus thermo cycler.  The D N A sample was  electrophoresed on an horizontal agarose slab gel and visualized by staining with ethidium bromide (Appendix 11). The 1.2 kb. band was excised and purified using a Pharmacia Sephaglas Bandprep  Kit. The purified fragment was restriction digested  with Eco R l (Appendix 12) ligated in the pGEX-2T expression vector that was digested  Table 7: Erkl sequencing oligonucleotide primers Amino Acid Sequence  Oligonucleotide orientation  Oligonucleotide DNA Sequence  149-174  Reverse  GAG G TG GAG ATG GTG AAG  Glu-Val-Glu-Met-Val-Lys-  151-167  Forward  GAG GTG GAG ATG GTG AA  Glu-Val-Glu-Met-Val-Lys  214-231  Reverse  ATC GGC GAG GGC GCG TAC  Ile-Gly-Glu-Gly-Ala-Tyr  335-352  Forward  AGA TCC AGA TCC TGC TGC  Glu-Ile-Gln-Ile-Leu-Leu  443-459  Reverse  GTC CAG TCA GAG GTA GT  Leu-Met-Glu-Thr-Asp-Leu  483-500  Forward  GCT GAG CAA TGA CCA TAT  Ile-Ser-Asn-Asp-His-Ile  615-631  Forward  GAT TTG TGA TTT CGG CC  Ile-Cys-Asn-Phe-Gly  724-742  Reverse  ACT CCA AGG GCT ATA CCA  Gln-Ser-Lys-Gly-Tyr-Thr  724-742  Forward  ACT CCA AGG GCT ATA CCA  Gln-Ser-Lys-Gly-Tyr-Thr  916-934  Forward  CAG TCT CTG CCC TCC AAG  Gln-Ser-Leu-Pro-Ser-Lys  1086-1103  Reverse  TGA GCC AGT GGC CGA GGA  Glu-Pro-Val-Ala-Glu-Glu  1165-1182  Forward  TTC CAG GAG ACA GCA CGC  Phe- Gln-Glu-Thr-Ala-Arg  1205-1221  Reverse  CCC CCT AGC CCA GAC AG  Ala-Pro-stop-3'-untranslated  1396-1312  Forward  TTC TCC TCC CCA CCC GC  3'-untranslated region  1396-1313  Reverse  TCC TCC TCC CCA CCC GCC  3'-untranslated region  1476-1494  Reverse  CGC CCT TAC TCC CCC CAG  3'-untranslated region  1611-1629  Forward  CAT CTC ATT CAA ACC CCA  3'-untranslated region  1707-1725  Reverse  CCT GTC AAA GCT GTC ACT  3'-untranslated region  Sequence location  with the same enzyme (Appendix 13) and transformed into competent E. coli strain UT5600 (Appendix 14). A Sma I digest was performed to verify that the full-length Erk1 cDNA was subcloned in the proper orientation.  4.5  Chromosomal Assignment  The chromosomal localization of the human erkl gene was determined by probing nylon membrane panels of Hindlll digested hamster-human hybrid cell line purchased from BIOS Corporation. The DNA blots were placed in a heat sealable plastic bag and prehybridized with pre-warmed BIOS Speed-Hyb Solution (6X SSC, 5X Denhardt's, 10% dextran sulfate, 1% SDS and 100 |ig/ml denatured, fragmented salmon sperm DNA; Appendix 19) for 2 h in a rotator water bath set a 65°C. The partial human Erkl cDNA clone A431-400-8 was radiolabeled by the random oligonucleotide primer method (Feinberg and Vogelstein 1983,1984) to a specific activity of 10 -10 cpm/ug 8  9  DNA and added to the prehybridization solution. After hybridizing for 16 h at 65°C with agitation, the membranes were washed twice at low stringency (2X SSC, 0.5% SDS, room temperature) for 10 min, once and medium stringency (IX SSC, 1% SDS, 65°C) for 15 min and twice at high stringency (0.1% SSC.1% SDS, 65°C) for 15 min. To prevent the blots from drying, they were wrapped in plastic before placing them in an imaging cassette for autoradiography.  4.6.  Primer Extension  The primer extension technique was used to measure the size of the 5' terminal region of the human Erkl mRNA.  A complementary oligonucleotide primer was  patterned after a sequence located within 150 base pairs of the 5' terminus of the Erkl 26ap-3 clone (Table 8). In this procedure the bacteriophage T4 polynucleotide kinase  was used to catalyze the 5' phosphorylation of the primer with [y- P]ATP (Appendix 32  18). Once purified, 10 ul of labelled primer (1 ng /ul) was combined with 20 ul of total RNA from Hep G2 cells (50 ug), 6 ul of 3 M sodium acetate and sufficient DEPC water to increase the volume to 50 ul final. The nucleic acids were precipitated by addition of 2.5 volumes of 100% ethanol and followed by storage at -20°C for 30 min. The RNA was recovered by centrifugation at 13,000 rpm (~12,000x g) for 10 min in a microcentrifuge cooled to 4°C. The pellet was washed once with 70% ethanol, and recentrifuged. The supernatant was carefully removed by pipette and the residual ethanol allowed to evaporate from the pellet. Thirty microlitres of hybridization buffer O (40 mM PIPES [pH 6.4], 1 mM EDTA [pH 8.0], 0.4 NaCl and 80% formamide) were added to the precipitated nucleic acids and resuspended by repeatedly pipetting the liquid. To fully separate the nucleic acids, the hybridization mixture was heated to 85°C for 10 min. The hybridization reaction was immediately placed at the annealling temperature of 30°C for 16 h. The RNA/DNA hybrid was precipitated by addition of 170 ul of 3 M sodium acetate, 500 ul of 100% ethanol and storing at 4°C overnight. After centrifuging at 4°C for 15 min, the supernatant was removed and the pellet washed with 70% ethanol. The RNA/DNA hybrid was air dried before diluting the nucleic acids in 25 ul of reverse transcriptase reaction buffer (2.5 ul buffer N , 2 ul 100 mM each dNTPs, 2 ul 100 mM DTT, lul RNase inhibitor, 13 ul deionized water). The reaction was initiated by addition of 2 ul (400 U) of Superscript reverse transcriptase® At the completion of a 2 hour incubation period at 42°C, the reaction was quenched by adding 1 ul of 0.5 M EDTA. Ribonuclease A was added to the mixture to degrade the RNA template. The volume was increased with 150 ul of TE buffer containing 150 mM NaCl prior to extracting the DNA with 200 ul of phenol/chloroform solution. The upper aqueous phase was transferred to a fresh tube and the DNA precipitated with 500 ul of 100 % ethanol for 1 h at 0°C. The DNA was recovered by centrifuging at 4°C for 15 min at 12,700 rpm (12,000 x g), and washed with 70% ethanol. The pellet was air dried to remove any traces of residual  ethanol and then dissolved in 4 ul of TE buffer and 6 ul of Pharmacia loading buffer. The longest Erkl cDNA clone, 26a(3-3, was also subjected to sequence analysis using the very same primer extension oligonucleotide primer. Both reactions were electrophoresed on DNA sequencing gels as outlined in Appendix 17. The difference in length observed between the primer extended fragment obtained from the RNA template to that of the Erkl cDNA template was predicted to be the amount of 5' sequence missing from the 26a(3-3 clone.  4.7  PCR Amplification of a murine Mekl cDNAs  Total RNA from mouse liver was used to synthesize a single stranded cDNA template for PCR. One to two micrograms of total cellular RNA were combined with 20 units RNasin, 1 mM nucleotide triphosphates, 100 pmol mixture of six base random oligonucleotide primers and New England Biolabs PCR buffer N in a final reaction volume of twenty microliters. After addition of 200 U of Superscript reverse transcriptase, the mixture was subjected to annealling (10 min at 23°C), elongation (45 min at 42°C) and denaturation (3 min at 95°C) steps. Oligonucleotide primers specific for the coding and complementary DNA sequences (sense strand beginning with the ATG-initiation codon, ATG CCC A A G A A G A A G CCG ACG CCC ATC CAG CTG AAC; antisense strand beginning with the TGA stop codon, A A C CAG CCC AGC ACA CCA ACC CAC GCT GGC AGC ATC TGA) of the murine Mekl gene product were synthesized from the published results of Crews et al. (1992). At the completion of the denaturation step, the reaction mixture was diluted with 75 ul of I X PCR buffer N containing 100 pmol of the forward and reverse oligonucleotide primers. Amplification of the Mekl cDNA was performed by adding 10 U of high-fidelity VentR® DNA polymerase. The thermal cycler (Perkin-Elmer Cetus) was programmed for 30 rounds of amplification (95 °C for 45 sec, 50°C for 90 sec and 74°C for 120 sec). The PCR  reactions were analyzed by electrophoresis using an 0.8% agarose slab gel. The bands were visualized by low-intensity U.V. light after staining the gel in ethidium bromide (Appendix 11).  The appropriate size bands were excised and purified using Pharmacia's Sephaglas Band Prep Kit™. The purified DNA fragments were sequentially digested first with Sma I then Eco Rl restriction endonucleases (Appendix 12) and ligated into the identical cloning sites of the pGEX-2T prokaryotic expression plasmid (Appendix 13). Recombinant clones were identified by restriction analysis of small-scale preparations of plamid DNA (Appendix 16). The putative recombinant clones were sequenced by the Sanger D N A sequencing procedure (Appendix 17) using specific sequencing oligonucleotides (Table 8) to confirm the presence and fidelity of a Mekl cDNA.  4.8  Prokaryotic expression of recombinant GST-fusion proteins  The pGEX-2T vector, carrying a mutant of the lacZ inhibitor gene lacN that overproduces lac promoter repressor protein, was used for the expression of GST-Erkl and GST-Mekl fusion proteins used during the course of this study (Muller-Hill et al., 1968). The expression of recombinant proteins is more tightly controlled with laciv thereby minimizing the spurious transcription of the lac operon and the consequent synthesis of potentially toxic foreign polypeptides. Erkland Mekl fusions were expressed in the E. coli strain UT5600 generated by Elish et al. (1988). The unique feature of this strain for prokaryotic expression of recombinant proteins is that it has a mutation in the periplasmic protease. This reduced protein degradation during lysis and purification.  120  Table 8. Mekl sequencing oligonucleotide primers Sequence Location  Oligonucleotide Orientation  Oligonucleotide DNA Sequence  Amino Acid Sequence  197-214  Forward  TTT CTG ACG CAG AAG CAG  Phe-Leu-Thr-Gln-Lys-Gln  439-458  Reverse  TAC AGC GAC GGC GAG ATC  Tyr-Ser-Asp-Gly-Glu-Ile  575-592  Forward  TAT CTT CGG GAG AAG CAC  Tyr- Leu-Arg-Glu-Lys-His  663-682  Reverse  ATT TTG GGG TCA GCG GGC  Phe-Gly-Val-Ser-Gly-Gln  731-748  Forward  TCG CCT GAG AGA CTC CAG  Ser-Pro-Glu-Arg-Leu-Gln  879-898  Forward  GGA GAC GCA GCC GAA ACA  Gly-Asp-Ala-Ala-Glu-Trr  4.8.1  Protein expression in bacterial culture  The purification of glutathione-S-transferase  by  glutathione-affinity  chromatography was first described by Simons and Jagt (1977) and later adapted for the expression and purification of eukaryotic GST-fusion proteins in bacteria (Smith et al, 1988). The protocol used for the isolation of milligram quantities of fusion protein was modified from Guan and Dixon (1991). Before performing a purification of the fusion protein, the recombinant Erkl or Mekl clone was transformed into competent UT5600 bacteria (Appendix 14) to insure they grow and express the highest levels of fusion protein. A volume of 2-50 ml of 2 x YT medium containing 100 ug/ml ampicillin was inoculated with a single colony of E. coli containing the recombinant pGEX plasmid. The culture was incubated at 37°C overnight in a Lab Line shaker/incubator with vigorous shaking (220-230 rpm). The culture was diluted 1:10 (vol/vol) in fresh 2 x YT medium supplemented with 100 ug/ml ampicillin. To insure proper aeration, the volume of the media did not exceed 20-25% of the capacity of the flask. The bacteria were grown at 37°C with shaking until the cell density reached an absorbance of 1-2 at OD600The cells were then induced with 400 uM of IPTG for another 4-6 h.  4.8.2  Preparation of bacterial cytoplasmic sonicates  After inducing bacterial expression of the GST-fusion proteins for the required amount of time, the culture was transferred to centrifuge bottles or tubes and the cells sedimented at 8,000 rpm (-11,300 x g) for 10 min at 4°C with a Beckman J2 HS centrifuge. The supernatant was discarded and the residual media drained form the pellet. The bottles/tubes were placed on ice. The pellet was resuspended by adding 50 ul of ice-cold buffer P (50 mM Tris-HCl [pH 8.0], 1 mM EDTA [pH 8.0], and 150 mM NaCl) per millilitre of culture. The peptidylglycan structure of the outer cell membrane  was disrupted by addition of 100 ug/ml of lysozyme. An increase in the viscosity of the liquid usually observed within 20 min was indicative of cell disruption. Triton X-100 was added to a final concentration of 1% to facilitate solubilization of the protein. The genomic and plasmid DNA was sheered by sonicating with a Vibra-cell sonicator for 10 sec bursts at a setting 45-60. Using a DuPont Sorvall® ODT Combi ultracentrifuge and its companion T865 rotor, the homogenate was clarified at 30,000 rpm (-81,900 x g) for 15 min at 4°C and the supernatant transferred to a fresh container. A battery of broad spectrum protease inhibitors (10 ug/ml each of aprotinin, soybean trypsin inhibitor and leupeptin see Table 2 for details) was added to the cell extract. At this point, the GSTfusion was immediately purified or the homogenate was stored at -70°C until needed.  4.8.3  Purification of the fusion proteins  The glutathione-agarose beads were prepared by swelling them in buffer P containing 0.1% Triton X-100 and 1% bovine serum albumin. Prior to adding the beads to the bacterial homogenate, the beads were washed three times with buffer P. The supernatant was incubated with a volume of beads appropriate for the volume of cytoplasmic sonicate (1-2 ml bed volume per 50 ml of sonicate) for 30 min at 4°C on a rotating platform. The GST-fusion bound to the matrix was recovered by centrifugation at 1000 rpm (~200g) for 30-60 sec using a Sorvall RT600D centrifuge. After discarding the supernatant, the beads were washed with 10 bed volumes of buffer P. The beads were again recovered by centrifugation. The washing and centrifugation steps were repeated two more times. The protein concentration was determined by electrophoresing 10-20 ul of beads on a Bio-Rad SDS-PAGE mini-gel apparatus (Appendix 5) followed by staining with Coomassie blue (Section 10, ii). Small volumes of the beads were aliquotted into Eppendorf tubes and stored at -70°C.  4.8.4  Elution of glutathione bound GST-fusion proteins  The GST-fusion protein bound to the glutathione matrix was washed once with IX PBS (140 mM NaCl, 2.7 mM KC1, 10 mM N a H P 0 , 1.8 mM K H P 0 , pH 7.3). 2  4  2  4  Glutathione elution buffer (25 mM reduced glutatihone in I X PBS pH 7.0) was added to the equilibrated beads and the mixture incubated with rotation at 4°C for 1-2 h and sometimes overnight.  The eluted GST-fusion protein was recovered by gentle  centrifugation at 1,000 rpm (-200 x g) for 30-60 sec using a DuPont Sorvall RT 6000 bench-top centrifuge. The protein concentration was measured by the Bradford (1976) assay using bovine serum albumin as a protein standard (Section 3). Before performing any kinase reactions, the free-glutathione was removed by dialysis against I X PBS with Spectrapor® membrane tubing (M.W. cutoff 12,000-14,000) for 1 h at 37°C. Aliquots (100-250 (il volume) were stored at -70°C.  4.8.5  Thrombin cleavage of glutathione bound GST-fusion proteins  In preparation for thrombin cleavage, the GST-fusion protein matrix was first washed with 10 bed volumes of IX PBS. The reaction was performed by addition of 500 ul of thrombin solution (1 cleavage unit per microlitre of I X PBS) to an equivalent bed volume of bead matrix. The incubation was allowed to continue for 2-16 h at room temperature with rotation. At the completion of the reaction, the bead matrix was centrifuged at 1,000 rpm (-200 x g) for 2 min at 4°C with a Hereaus Biofuge 15. Protease inhibitors (10 ug/ml each leupeptin and SBTI see Table 2) were added to inhibit thrombin activity. Before storing the cleaved protein at -70°C, the concentration was determined by SDS-PAGE mini-gel electrophoresis (Appendix 5).  4.9  Oligonucleotide-mediated mutagenesis  The in vitro site-directed mutagenesis technique is a powerful tool for analyzing the structural and functional properties of protein kinases for which the cDNA is available for study.  Short oligonucleotide primers are synthesized with base  substitutions, deletions or additions for one or more specific amino acid change(s) in the wild type protein. Using'the original cDNA as a template, paired mutant and wild type primers were subjected to several rounds of amplification by the polymerase chain reaction. After purification and restriction digestion, the wild type portion of the cDNA was substituted with the corresponding mutagenized fragment. Sequence analysis of the subcloned region confirms that only the desired mutations are generated by the PCR procedure. Two different approaches described below were used to generate the mutant cDNA fragments.  4.9.1  Mutagenesis by the megaprimer method  The two-step PCR megaprimer approach requires that the mutagenized fragment be synthesized in two separate PCR reactions (Sarkar and Sommer, 1990). In the first PCR reaction step, the mutant oligonucleotide primer in combination with a second template-based wild type primer are used to amplify a partial (300-500 base pair) mutagenized cDNA fragment. The purifed mutagenized cDNA fragment is then paired with a second template-based wild type primer to synthesize a full-length or nearly fulllength mutant fragment by PCR. A detailed protocol for in vitro mutagenesis by two-step PCR is outlined below. The sequence of the oligonucleotide primers is indicated in Table 9. The template was prepared by digesting 1-2 jxg of recombinant pGEX-2T vector containing the Erkl or Mekl cDNA insert with the appropriate restriction enzyme  Table 9: Erkl and Mekl mutant oligonucleotide primers A. Erkl mutant oligonucleotide primers Oligonucleotide DNA Sequence and base change(s)  Amino Acid Sequence  Amino Acid Location  Amino Acid Change  Oligonucleotide Orientation  Wild type  None  Forward  ATG GCG GCG GCG GCG GCT CAG GGG  Met-Ala-Ala-Ala-Ala-AlaGln-Gly  Wild type  None  Reverse  UUC CAG CCC GGA GAG CTG GAG GCC CCC  Phe-Gln-Pro-Gly-Val-LeuGlu-Ala-Pro  71  K->A  Forward  ACT CGC GTG GCC ATC GCG AAG ATC AGC CCC TTC  Thr-Arg-Val-Ala-Ile-AJaLys-Ile-Ser-Pro-Phe  *  Wild Type  Reverse  GAC CAC ACC GGC TTC CTG  Asp-His-Thr-Gly-Phe-Leu  *  T->S  Forward  TCT GAG TAT GTG GCT ACG  Ser-Glu-Tyr-Val-Ala-Thr  T->E  Forward  CAC ACC GCC TTC CTG GAG GAG TAT GTG GCT ACG  His-Thr-Gly-Phe-Leu Glu-Glu-Tyr-Val-Ala-Thr  *  Wild Type  Reverse  GCC TTC CTG ACG GAG  Gly-Phe-Leu-Thr-Glu  *  Y->T  Forward  ACT GTG GCT ACG CGC TGG  Thr-Val-Ala-Thr-Arg  Y->E  Forward  GGC TTC CTG ACG GAG GAG GTG GCT ACG CGC TGG  Gly-Phe-Leu-Thr-GluGJu-Val-Thr-Arg-Trp  *  T->Y  Reverse  CAC ACC GGC TTC CTG TAT  His-Thr-Gly-Phe-Leu-Tyr  *  Y->T  Forward  GAG ACT GTG GCT ACG CGC  Glu-Thr-Val-Ala-Thr-Arg  201  202  202  203  204  204  202  204  126  202/204  T->E Y->E  Reverse  GGC TTC CTG GAG GAG GAG GTG GCT ACG CGC TGG  Gly-Phe-Leu-GJu-GluGlu-Val-Ala-Thr-Arg-Trp  201/202*  L->E T->E  Reverse  CAC ACC GGC TTC GAG GAG  His-Thr-Gly-Phe-GJu-GJu  204/205*  Y->E V->E  Forward  GAG GAG GAG GCT ACG CGC TGG  Glu-Glu-Glu-Ala-ThrArg-Trp  B. Mek mutant oligonucleotide primers  Wild type  None  Forward  ATG CCC AAG AAG AAG CCG ACG CC  Met-Pro-Lys-Lys-Lys-ProThr-Pro  Wild type  None  Reverse  ACA CCA ACC CAC GCT GCC AGC ATC TGA  Thr-Pro-Thr-His-Ala-AlaSer-Ile-stop  97  K->A  Reverse  CTG GTT ATG GCT AGA GCG CTG ATC CAC CTG GAG  Leu-Val-Mat-Ala-ArgAla-Ile-His-Leu-Glu  218  S->E  Reverse  AGC GGG CAG CTA ATT GAC GAG ATG GCC AAC TCC TTC GTG  Ser-Gly-Gln-Leu-Ile-AspGlu-Met-Ala-Asn-SerPhe-Val  S->E  Forward  ATT GAC TCT ATG GCC AAC GAG TTC GTG GGC ACG AGA TCC  Ile-Asp-Ser-Met-Ala-GlnGJu-Phe-Val-Gly-ThrArg-Ser  222  Note: those letters that are underlined and bolded indicate the base alteration and the corresponding amino acid substitution.  127  (Appendix 12). Opening the plasmid enabled the oligonucleotide primers to have better access to their template. After heat-inactivating the enzyme at 90°C, the PCR cocktail was prepared in a volume of 50 ul by mixing 250-500 ng of the digested recombinant vector with 2mM nucleotide triphosphates, IX PCR buffer N , 100 pmol of each paired mutant and wild type primers (Table 9) and 10 U of high-fidelity VentR® DNA polymerase. The reaction mixture was subjected to thirty cycles of amplification (95°C for 45 sec, 55°C for 90 sec, and 74°C for 90 sec) using a Perkin-Elmer Cetus thermal cycler. The cDNA fragment was purified from an ethidium bromide stained horizontal agarose slab gel (Appendix 11) by a novel glass matrix system designed by Pharmacia. The second PCR amplification was executed by combining 200 ug of the purified mutant cDNA fragment with a second template-based primer using the same condition as described above except for the change in reaction conditions (95°C for 30 sec, 55°C for 30 sec and 74°C for 60 sec). The mutant cDNA fragment was gel purified (Appendix 11), restriction digested with the appropriate enzymes (Appendix 12), and ligated back into the wild type background (Appendix 13). The mutation(s) were confirmed by DNA sequencing (Appendix 17).  4.9.2 Mutagenesis by the double-primer method  Several Erkl and Mekl mutants were generated by the single-step site-directed mutagenesis method. In this procedure, two contiguous (non-overlapping) coding and complementary mutant oligonucleotide primers are synthesized with a phosphate group located at their 3' terminus to facilitate PCR fragment ligation. A second pair of coding and complementary wild type oligonucletide primers based on the 5' and 3' termini of the full-length cDNA are synthesized with vector specific restriction endonuclease sites for cloning purposes. Individual fragments of the mutant cDNA are created by performing two separate PCR reactions using paired combinations of 5' wild type coding and 3'  mutant complementary primers or 5' mutant coding and 3' wild type complementary primers. The complete Erkl and Mekl clones were reconstituted by digesting the plasmid vector and fragments with the specific restriction endonucleases (Appendix 12) and performing a three-way ligation (Appendix 13).  RESULTS  1.  ISOLATION OF A HUMAN ERK1 cDNA FROM HEP G2 CELLS  1.1  Amplification of a partial cDNA encoding a human MAP kinase  Erkl was originally purified as a 43-kDa protein from insulin-stimulated Rat 1 (HIRc B) cells that were transfected with the human insulin receptor (Boulton et al, 1991). Sanghara et al. (1992) identified two proteins of equivalent molecular mass, 42kDa (Erk2) and 44-kDa (Erkl), from human epidermoid A431 cell line that were immunoreactive with several M A P kinase antibodies.  The only M A P kinase  complementary DNA (cDNA) sequence that had been described in the literature was a partial Erkl clone obtained from rat (Boulton et al, 1990). Therefore, using the partial rat Erkl sequence data, a strategy was devised to isolate the full-length human Erkl homologue. The reverse transcriptase polymerase chain reaction (RT-PCR) technique and specific oligonucleotides were used to amplify a partial cDNA fragment from total A431 cellular RNA. Design of the primers was based on two criteria: (i) the sites for oligonucleotide design were chosen from within the kinase catalytic subdomains domain (i.e., II, VII and IX); and (ii) the codon selection of the base pairs adhered to the usage for homo sapiens (Hanks et al, 1988; Lathe, 1985). Figure 9 shows the location of the single 5' primer (subdomain II) and the two 3' primers (subdomains VII and IX) used for amplification. RT-PCR performed with paired forward and reverse primers yielded two partial cDNA fragments of 400 (A431-400-3) and 500 (A431-500-8) base pairs in size which fell within the range expected for a product of equivalent distance in rat cDNA (Boulton et al, 1990). Only two amino acid differences were observed between the predicted translational sequence of the PCR fragments and the published rat Erkl sequence (data not shown).  1.2  Detection of a MAP kinase in human Hep G2 cells  At the time this work was initiated, MAP kinase was demonstrated to be activated in a very small number of cell types (Pelech et al., 1990). Therefore, to successfully screen a cDNA library for a full-length MAP kinase, it was essential to determine which cell line expressed the enzyme. The exogenous substrate MBP was used to assay for the presence of MAP kinase(s) in the transformed human hepatocellular carcinoma line Hep G2. The cells were grown to confluency (~10 cells) before fasting overnight in medium 8  containing 0.5% calf serum. The quiescent (G ) cells were treated without or with 0.1 0  mM insulin for 5 min. Approximately 1.5 mg of the clarified control- and insulin-treated homogenates were applied to a Mono Q anion exchange column at a flow rate of 1 ml/min. In this system, MAP kinases Mpkl from mature sea star oocytes and Erk2 from Xenopus oocytes eluted as broad peaks between 330 to 470 mM NaCl (Sanghera et al, 1990; Posada et al, 1991). After elution of the bound proteins with a linear 0-0.8 M NaCl gradient, odd numbered fractions were assayed for MAP enzyme activity in a 5 min MBP kinase assay. As can be seen in Figure 8A, a rather sharp peak of MBP kinase activity eluted between 395 to 450 mM NaCl (fractions 27 through 31, closed circles). This marked a two-fold increase in kinase activity when compared to untreated Hep G2 cells over the same fractionation range (open circles). The elution profile of the MBP kinase activity was similar to the behavior of MAP kinases that had been described in other studies (Sanghera et al., 1990; Posada et al., 1991). Because MBP may be a substrate for other kinases a small volume of each assayed fraction was analyzed by Western immunoblotting. The MBP kinase activity peak coincided with a band that was immunoreactive with rat Erkl-CT anti-peptide antibody (Figure 8B). Therefore, both enzymatic and immunological data support a role for MAP kinase activation following insulin treatment in Hep G2 cells. Hence, a cDNA library of this cell line would be useful for screening human Erkl.  Figure 8: Detection of MAP kinase activity in Hep G2 cells. (A) The hepatocarcinoma cell line Hep G2 was stimulated for 10 min with 100 uM insulin (filled circles) or without insulin (open circles), lysed and 1.5 mg of the homogenate chromatographed on a Mono Q column as described in the Material and Methods. MAP kinase activity was assessed in a 25 ul reaction by incubating 5 ul of extract with Mg »[Y- P]ATP (2000 2+  32  pmol/cpm) and 1 mg/ml final of MBP substrate. After 5 min the reaction was terminated by applying 20 ml of the reaction onto a 2 cm X 2 cm Whatman filter paper. Activity is expressed in pmol/min/ml of the original column fraction. (B) The insulin-treated fractions from panel A were electrophoresed on a 10% SDS-PAGE, transferred to nitrocellulose and immunoblotted with Erkl-CT, a MAP kinase polyclonal antibody specific for the carboxy-terminal of the rat Erkl protein (Boulton et al., 1991).  A  0  10  20  30  40  50  Mono Q fraction number  B  50kDa cc-erkl-CT 33kDa •S 12 16 20 24 26 28 30 32 34 36 40 44 50 Mono Q fraction number  133  1.3  Cloning of a full-length cDNA encoding the human Erkl protein  The partial 500-bp A431-500-8 clone isolated by RT-PCR was labelled with P 3 2  and used to probe a XZAP bacteriophage cDNA library prepared from Hep G2 cell poly(A) enriched mRNA that were cloned into the Eco R l cloning site. Three sets of +  ten large plates containing a minimum of 10,000 plaques were 'lifted' twice with two sheets Hybond-N hybridization membranes. After incubating the membranes overnight at room temperature in the presence of the P-labelled partial human Erkl probe, the 32  first low-stringency screen (incremental washes at room temperature, 40°, 45°, 50° and 60° C using a solution of 0.1% SDS and 2 X SSC) of -300,000 plaques identified 26 potential MAP kinase recombinants. However, following more stringent conditions during the secondary (incremental washes at room temperature, 40° and 45° using a solution of 0.1% SDS and 2 X SSC) and tertiary screens (62.5°C using a solution of 0.1% SDS and 0.1 X SSC) only 4 double-positive clones were enriched following these plaque purification steps. Three of the clones were rescued as Bluescript plasmids from the Lambda ZAP® II vector using the R408 interference resistant Helper phage according to Stratagenes specifications. The cDNA inserts were then characterized by Eco R l restriction mapping. All three clones appeared to be identical, since they were isolated as single fragments ranging in size from 1.2-1.9 kbs after electorphoresis on an agarose gel. This was confirmed by limited sequence analysis from the Bluecscript plamid T3 and T7 sequencing primers. One of the clones encoded a partial Erkl sequence, while a second contained two poly(A) -tails and some partial Erkl coding sequence. The largest clone +  designated 26a [3-3 was sequenced completely since the 1.9 kb fragment was similar in size to the partial rat Erkl sequence (Boulton et al., 1990).  The complete sequence of 26a(3-3 clone yielded a cDNA insert that was 1,850 bp in length excluding the poly(A)+ region. Figure 9 shows the complete cDNA and  Figure 9: Nucleotide and predicted amino acid sequence of human Erkl cDNA from Hep G2 cells. The sequence is derived from the full-length cDNA clone identified using a partial Erkl PCR probe as outlined in Materials and Methods. The putative methionine translational initiation sites are indicated by the numbers 1 and 2 whereas the RNA polyadenylation signal (AUAAA) is italicized and underlined. The regulatory phosphorylation sites are designated with an asterisk. The catalytic subdomains (Roman numerals) and their conserved sequences (bolded and underlined) adhere to the nomenclature of Hanks et al. (1988). The horizontal arrows situated above the protein sequence correspond to the location of the oligonucleotides used for PCR amplification (Boulton et al., 1991). The vertical arrows located above the protein sequence correspond to potential thrombin cleavage sites.  © M A A A A A CGTTCCTCGGCGCCGCCGGGGCCCCAGAGGGCAGCGGCAGCAACAGCAGCAGCAGCAGCAGCGGGAGTGGAGATGGCGGCGGCGGCGGCT 15 30 45 60 75 90  © Q G G G G G E P R R T E G V G P G V P G E V E M V K G Q P F CAGGGGGGCGGGGGCGGGGAGCCCCGTAGAACCGAGGGGGTCGGCCCGGGGGTCCCGGGGGAGGTGGAGATGGTGAAGGGGCAGCCGTTC 105 120 135 150 165 180  I D V G P R Y T Q L Q Y I G E G A Y G M V S S A Y D H V R K T GACGTGGGCCCGCGCTACACGCAGTTGCAGTACATCGGCGAGGGCGCGTACGGCATGGTCAGCTCGGCCTATGACCACGTGCGCAAGACT 195 210 225 240 255 270  II  ^  III  R V A I K K I S P F E ^ H Q T Y C Q R T L R E I Q I L L R F R CGCGTGGCCATCAAGAAGATCAGCCCCTTCGAACATCAGACCTACTGCCAGCGCACGCTCCGGGAGATCCAGATCCTGCTGCGCTTCCGC 285 300 315 330 345 360  IV  V  H E N V I G I R D I L R A S T L E A M R D V Y I V Q D L M E CATGAGAATGTCATCGGCATCCGAGACATTCTGCGGGCGTCCACCCTGGAAGCCATGAGAGATGTCTACATTGTGCAGGACCTGATGGAG 375 . 390 405 420 435 450 T D L Y K L L K S Q Q L S N D H I C Y F L Y Q I L R G L K Y ACTGACCTGTACAAGTTGCTGAAAAGCCAGCAGCTGAGCAATGACCATATCTGCTACTTCCTCTACCAGATCCTGCGGGGCCTCAAGTAC 465 480 495 510 525 540  VI I H S A N V L H R D L K P S I J L L S N T T C D L K I C ATCCACTCCGCCAACGTGCTCCACCGAGATCTAAAGCCCTCCAACCTGCTCAGCAACACCACCTGCGACCTTAAGATTTGTGATTTCGGC 555 570 585 600 615 630  *  *  VIII  L A R I A D P E H D H T G F L T E Y V A T R W Y R A P E I M CTGGCCCGGATTGCCGATCCTGAGCATGACCACACCGGCTTCCTGACGGAGTATGTGGCTACGCGCTGGTACCGGGCCCCAGAGATCATG 645 660 675 690 705 720  <  —  L N S K G ^ T K S I D I W S V S C I L A E M L S N R P I F P CTGAACTCCAAGGGCTATACCAAGTCCATCGACATCTGGTCTGTGGGCTGCATTCTGGCTGAGATGCTCTCTAACCGGCCCATCTTCCCT 735 750 765 780 795 810  X  G K H Y L D Q L N H I L G I L G S P S Q E D L N C I I N M K GGCAAGCACTACCTGGATCAGCTCAACCACATTCTGGGCATCCTGGGCTCCCCATCCCAGGAGGACCTGAATTGTATCATCAACATGAAG 825 840 855 870 885 900 A R N Y L Q S L P S K T K V A W A K L F P K S D S K A L D L GCCCGAAACTACCTACAGTCTCTGCCCTCCAAGACCAAGGTGGCTTGGGCCAAGCTTTTCCCCAAGTCAGACTCCAAAGCCCTTGACCTG 915 930 945 960 975 990  XI L D R M L T F N P N K R I T V E E A L A H P Y L E Q Y Y D P CTGGACCGGATGTTAACCTTTAACCCCAATAAACGGATCACAGTGGAGGAAGCGCTGGCTCACCCCTACCTGGAGCAGTACTATGACCCG 1005 1020 1035 1050 i 1065 1080 T D E P V A E E P F T F A M E L D D L P K E R T L K E L I F Q ACGGATGAGCCAGTGGCCGAGGAGCCCTTCACCTTCGCCATGGAGCTGGATGACCTACCTAAGGAGCGGCTGAAGGAGCTCATCTTCCAG 1.095 1110 1125 1140 1155 1170 E T A R y F Q P G V L E A P STOP GAGACAGCACGCTTCCAGCCCGGAGTGCTGGAGGCCCCCTAGCCCAGACAGACATCTCTGCACCCTGGGGCCTGGACCTGCCTCCTGCCT 1185 1200 1215 1230 1245 1260 GCCCCTCTCCCGCCAGACTGTTAGAAAATGGACACTGTGCCCAGCCCGGACCTTGGCAGCCCAGGCCGGGGTGGAGCATGGGCCTGGCCA 1275 1290 1305 1320 1335 1350 CCTCTCTCCTTTGCTGAGGCCTCCAGCTTCAGGCAGGCCAAGGCCTTCTCCTCCCCACCCGCCCTCCCCACGGGGCCTCGGGAGCTCAGG 1365 1380 1395 1410 1425 1440 TGGCCCCAGTTCAATCTCCCGCTGCTGCTGCTGCTGCGCCCTTACCTTCCCCAGCGTCCCAGTCTCTGGCAGTTCTGGAATGGAAGGGTT 1455 1470 1485 1500 1515 1530 CTGGCTGCCCCAACCTGCTGAAGGGCAGAGGTGGAGGGTGGGGGGCGCTGAGTAGGGACTCAGGGCCATGCCTGCCCCCCTCATCTCATT 1545 1560 1575 1590 1605 1620 CAAACCCCACCCTAGTTTCCCTGAAGGAACATTCCTTAGTCTCAAGGGCTAGCATCCCTGAGGAGCCAGGCCGGGCCGAATCCCCTCCCT 1635 1650 1665 1680 1695 1710 GTCAAAGCTGTCACTTCGCGTGCCCTCGCTGCTTCTGTGTGTGGTGAGCAGAAGTGGAGCTGGGGGGCGTGGAGAGCCCGGCGCCCCTGC 1725 1740 1755 1770 1785 1800 CACCTCCCTGACCCGTCTAATATATAAATATAGAGATGTGTCTATGGCTGAAAAAAAAAAAAAAAA 1815 1830 1845 1860 1866  predicted amino acid sequence for human Erkl. A consensus polyadenylation signal (AUAAA) was located upstream of the poly(A) tail. A single translational termination +  codon U A G was located at nucleotide 1212 of the cDNA.  Two potential A U G  translational inititation codons, that satisfy the minimum consensus sequence first defined by Kozak (1987), were identified at nucleotides 73 and 160, respectively (Figure 9 and 10A).  Furthermore, the sequence surrounding both in-frame methionine start sites  (GUGGAGAUGG) is identical and thus both may serve to initiate translation.  The human Erkl, like the previously reported rat homologue, was published as a partial cDNA that lacked the initiating methionine (Boulton et al, 1991; Owaki et al, 1992; Gonzalez et al, 1992). Furthermore, no stop signal was identified 5' to either of the putative methione start sites in the human Erkl cDNA isolated from Hep G2 cells (Figure 10A). The GC-rich nature of Erkl 5' prime sequence may have hampered the first-strand cDNA synthesis and consequently more sequence may perhaps exist. To verify that the 26af3-3 clone contains the complete Erkl sequence, the size of the 5' region preceding the first methionine was analyzed by primer extension. A 26-base reverse primer patterned after a sequence (nucleotides 148 to 172) located 147 base pairs from the 5' terminus.  A complementary 185 base single-stranded cDNA fragment was  synthesized by this method (Figure 10B). The primer extended fragment generated from the mRNA template was approximately 13 bases longer than what was predicted from the Erkl cDNA clone. No heterogeneity was observed in the primer extended products.  If we assume that translation initiation begins at the first methionine (corresponding to nucleotide 73), then the full-length Erkl cDNA clone is expected to encode a 379-amino acid polypeptide with a predicted molecular mass of 43-kDa (Figure 9). The highly conserved invariant residues (bold type) that are found to be essential for catalytic function in all protein kinases are also conserved within the catalytic domain of  Figure 10: Primer extension analysis of the 5' terminal region of the human Erkl mRNA. (A) Nucleotide and predicted amino acid sequence of the 5' region of human Erkl clone 26a(3-3. The putative methionine initiation sites are numbered 1 and 2. (B) A complementary oligonucleotide primer was designed after a 5' sequence (amino acids 2732) of the human Erkl cDNA clone 26a(3-3. The primer was phosphorylated with [yP]ATP at its 5' terminus using T4 polynucleotide kinase. After elongation with reverse transcriptase, the RNA/DNA hybrid was treated with ribonuclease A followed by separation on a DNA sequencing gel (lane 2). A 1 kb DNA standard from Gibco BRL was end-labelled with [y- P]ATP and electrophoresed on the same gel (lane 1). 32  32  A.  R S S A P P D P N R A A A A T A CGT TCC TOG GCG CCG CCG GGG CCC CAG AGG GCA GCG GCA GCA A C A GCA  © A A A A A G V E M A A A A A Q G GCA GCA GCA GCA GCG GGA GTG GAG ATG GCG GCG GCG GCG GCT CAG GGG  G G G G E P R R T E G V G P G V GGC GGG GGC GGG GAG CCC CGT AGA ACC GAG GGG GTC GGC CCG GGG GTC  © P G E V E M V K G P CCG GGG GAG GTG GAG ATG GTG AAG GGG CAG  139  human Erkl (Hanks et al, 1988). Alignment of the full-length cDNA clones for rat and human Erkl revealed a high degree of identity (97%) throughout the entire protein sequence (Marquardt and Stabel, 1992; Charest et al., 1993). Rat Erkl is longer than its human homologue by a single alanine amino acid in the region of overlap. This residue is part of a poly-alanyl stretch located immediately following the initiating methionyl amino acid. MAP kinase isoform Erk2 from rat and human also features a string of six alanyl residues in its amino-terminal sequence (Boulton et al, 1991; Gonzalez et al, 1992). Interestingly, this poly-alanyl rich region of Erkl is followed by a repeat of five glycyl residues (Figure 10).  The minimal catalytic unit of a functional protein kinase has been defined as a 28kDa polypeptide chain comprising approximately 350 amino acid residues. This has been determined by truncation experiments and sequence alignment analysis of dozens of known enzymes (Hanks et al, 1988 and references therein). In figure 11, protein sequence alignments of the catalytic domains (corresponding to amino acids 39 to 338 in PKA as outlined by Hanks et al, 1988) were created for six human M A P kinase isoforms. Comparisons revealed that Erkl is more closely related to Erk2 (90%) than to any other MAP kinase family member. Erk5 (53%) and Hogl (49%) are the next MAP kinases most closely related to Erkl while Erk3 (42%) and Jnkl (41%) display the least homology. At first glance, the regions displaying the highest degree of amino acid conservation (subdomains I-II and VI-IX) are those regarded essential for enzyme catalysis.  These residues are involved in binding ATP and partiticipate in the  phosphotransfer reaction in all kinases (Hanks et al,  1988).  The regulatory  phosphorylation sites for Erkl, namely Thr-202 and Tyr-204, (equivalent to Thr-183 and Tyr-185 in Erk2) were mapped to an intermediate region between subdomains VII and VIII (Figure, 11), a location displaying little conservation among protein kinases (Hanks et al, 1988). MAP kinase isoforms Erk5, Jnkl and Hogl possess comparable Thr and  Figure 11: Sequence comparision of several MAP kinase isoforms. Sequences from human Erkl (Charest et al., 1993); Erk2 (Gonzalez et al., 1992); Erk3 (Gonzalez et al., 1992); Jnkl (Derjard et al., 1994); Hoga (Lee et al., 1994) and Erk5 (Zhou et al., 1995) were aligned using the BEST FIT program. To improve alignments spacing, dashed lines were introduced into the sequence. Residues that are identical or conserved in the MAP kinase isoforms are shown with an asterisk or a dot, respectively. Those residues in bold correspond to invariant or highly conserved residues and consequently define the eleven catalytic subdomains described for 62 of 65 unique kinase sequences by Hanks et al. (1988). Conserved amino acids are those residues with similar physical or structural properties: non-polar chain R goups (M, L, I, V, and C); neutral polarity R groups (A, G, S, T, and P); acidic and uncharged R groups (D, E, N, and Q); basic polar R groups (K, R, and H); and aromatic R groups (F, Y, and W).  ERK1 ERK2 ERK3 JNK1 H0G1  MA MA. MA. MSRSKRDNN MSQERP  AAAAQGGGGGEPRRTEGVGPGVPGEVEMVKGQPFDV GPRYTQLQYI 48 AAAAAGAGP EMVRGQVFDV GPRYTNLSYI 3 1 EKGDCIASVYG YDLGGRFVDFQPL 2 6 FYSVEIGDSTFTVLKRYQNLKPI 32 TFYRQELNKTIWEVPERYQNLSPV 30  ERK5  MAEPLKEEDGEDGSAEPPAREGRTPHRCLCAK  ERK1  GEGAYGMVSSAYDHVRKTRVAIKKIS-PFFJiQTYCQRTLREIQILLRFRHFJWIGIRDILRAST  LEAMRD  ERK2 ERK3 JNK1  GEGaYGMVCSAYDNVmTOVAIKKIS-PFEHQTYCQRTLREIKILLRFRHENIIGIIsm GFGVNGLVLSAVDSRACRKVAVKKI--ALSDARSMKHALREIKIIRRLDHDN^ GSGAQGIVCAAYDAILFIINVAIKKLSRPFQNQTHAK^  IEQMKD 100 100 LEEFQD 103  HOG1 ERK5  GSGAYGSVCAAFDTKTGLRVAVKF\XSRPFQS I I H A K R T Y R E L R L L K H M K H E N V I G L L D V F T P A R S GNGAYGWSSARRRLTGQQVAIKKIPNAFDVVTNAKRTLRELKILKHFKHDNIIAIKDILRPTVP  LEEFND YGEFKS  I  * *  N L A L L K A R S F D T F D V G D E Y E I I E T I D 58  II  * *  *  I I I  ** **  IV  **  * *  V  ERK1 ERK2  VI OTIVQDLMETDLYKLLKS-QQLSIsroHI VYIVQDIJffiTDLYKLLKT-QHLSNDH  101 129  V I I 190 173  ERK3  AYIVQEYMETDLARLLEGG-TLAEEHAKLFMYQLLRGLK^  JNK1  VYIVMELiMDANLCQVIQ—MELDHERMSYLLYQMLCGIKHLHSAGIIHRDLKPSNIVVKSDC-TLKILDFGLARTA  HOG1  VYLVTHMGADLNNIWC-QKLTDDWQFLIYQILRGL^  ERK5  VYVVLDLJffiSDLHQIIHSSQPLTLEHVRYFLYQLLRGL^ * * * * * ** * * * *** VIII  117  174  175 174  204 ****** *  *** *** **  IX  X  ERK1  ADPFJHDHTGFLTEWATRWYRAPEIMLNSKGYTKSIDOT  ERK2 ERK3 JNK1 HOG1 ERK5  ADPDHDHTGFLTEYVATOWYRAPEIMLNSKGYTKSIDIWSVGCILAElynijSMRP -DQHYSHKGYLSEGLVTKWYRSPRLLLSPNNYTK^^ GTS FMMTPYWTRYYRAPEVILG-MGYKFJWDLWSVGCIMGFJW^ DDE MTGWATRWYRAPEIMTJSMMHYNQTVDIWSVGCIMAELLTGRTLFPGTDHINQLC^ CTSPAEHQYFMTEYVATRWYRAPELMLSLHEYTQAIDLW  266  ERK1 ERK2 ERK3  EDLNCIINMKARNYLQSLPSKTKVAWAKLFP-KS EDLNCIINLKARNYLLSLPHKNKVPWMRLFP-NA —IPVIREEDKDELLRVMPSFVSSTWEVKRPLRKLLPE  JTMK1 HOG1 ERK5  EFMKKL-QPTWTYVENRPKYAGYSFEKLFP-DVLFPADSEHMKLKASQ YLINRMPSHEARNYIQSLTQMPKMNFANVFI-GA NPLAVDLLEKMLVLDSDKRITAAQALAHAY AVIQAVGAERVRAYIQSLPPRQPVPWETVYP-GA DRQALSLLGRMLRFEPSARISAAAALRHPF  249 243 245 243 280  XI DSKALDLLDRMLTFNPNKRITVEEALAHPY 329 DSKALDLLDKMLTFNPHKRIEVEQALAHPY 312 VNSEAIDFLEKILTFNPMDRLTAEMGLQHPY 310  *  *  *  ERK1  LEQYYDPTDEPVAEEPF-TFAMELDDLPKERLKELIFQETA  RFQPGV  ERK2  LEQYYDPSDEPIAEAPF -KFDMELDDLPKEKLKELIFEETA  RFQPGY  ERK3  MSPYSCPEDEPTSQHPFRIED-EIDDIV  RYPVSL  LMAANQSQLSNWDTCSS  *  319 306 343  * * 375 358 SSDLE  365  JNK1  INVWYDPSFJAEAPPPKIPDKQLDEREHTIEEWKELIYKEVMDLEER  365  HCG1 ERK5  FAQYHDPDDEPVAD-PY-DQSFESRDLLIDEWKSLTYDEVISF LAKYHDPDDEPDCAPPF-DFAFDRFALTRERIKEAIVAEIEDFHARREGIRO^^ *  347 418  ERK1  --  ERK2 ERK3  WRP-DRCQDASEVQRDPRGFGAL-AEDVQVD  JNK1  PRKDSHSSSERFLEQSHSSMERAFEADYGRS—CDYKVGSPS TKNGVIRGQ  434 374  HOG1  ERK5  PWAPSGDCAbffiSPPPAPPPCPGPAPDTIDLTLQPPPPVSEPAPPKKEGAISDNTKAALKAALLKSLRSRLRDGPSA  494  YLD  461  ERX1 ERK2 ERK3 jNKl  KLLWRD  NKPHHYSEPK  LILD  LSHW  HOG1 ERK5  PLEAPEPRKPWAQERQREREEKRRRRQERAKEREKRRQERERKERGAGASGGPSTDPLAGLVLSDNDRSL^  570  142  EKK1 ERK2 ERK3 jNKl H0G1 ERK5  KQ  AAGAPPTATGLADTGAREDEPASLFLEIAQWVKSTQGAQSTPARPPTTPS PSPLAQV  TRMARPAAPALTSVPAPAPAPTPTPTPVQPTSPPPGPVAQPTGPQPQSAGSTSGPVPQPACPPPGPAPHPTGPPG  ERK1 ERK2 ERK3  AACLPRPP  PPGPGGRRRQ  514 381 646  532  JNKI  HOG1 ERK5  VPPP 358 PIPVPAPPQIATSTSLI^QSLVPPPGLPGSSTPGVLPYFPPGLPPPDAGGAPQSSMSESPDVNLVTQQLSKSQV 722  ERK1 ERK2 ERK3 JNK1 HOG1 ERK5  EDPLPPWSGTPKGSGAGYGVGFDLEEFMQSFDMGVALXSPQIX^  ERK1 ERK2 ERK3 JNK1 HOG1 ERK5  LEAP RS -VHLPRPEALHQARGPAGQ QQ --S EIQMDSPMLLADLPDLQDP  PPVRPGR  539 LDQEEME  358 798  379 360 557 376 359 817  143  Tyr phosphorylation motifs located in similar positions in the sequence. Interestingly, two distinguishing features emerge from the sequence comparison within the activation domain for MAP kinases. First, the intervening glutamyl residue in TEY of Erkl, Erk2 and Erk5 is replaced by glycyl and prolyl residues in Hogl and Jnkl respectively. Second, the number of amino acid separating subdomain VII and VIII vary from six, eight and twelve residues each for Hogl, Jnkl and Erkl, respectively. Variations in the number and kinds of residues in this region may account for some of the specificity observed during activation of specific MAP kinase modules. Erk3 differs from other MAP kinases in that a seryl residue replaces the threonyl residue and the tyrosyl residue is completely absent. Erk3 and Erk5 have a unique structural feature that distinguishes them from other members of the MAP kinase family. Both kinases possess a large noncatalytic domain at their carboxy-terminal. It remains to be established whether this polypeptide extension plays a regulatory role similar to that observed with the aminoterminal of Rafl and protein kinase C (PKC) enzymes. Both of these enzymes possess cystein-rich motifs. In the case of PKC, interacation of the cystein-rich fingers with membrane-associated lipid derivatives stimulates the enzymes  phosphotransferase  activity. The function of these same motifs in Rafl activation may have a similar role in regulating Raf kinase activity..  144  2.  CHARACTERIZATION OF RECOMBINANT HUMAN ERK1 PROTEIN  2.1  Expression of human Erkl in E. coli as a recombinant GST-fusion protein  To express Erkl protein in sufficient quantity for biochemical analysis, we used the E. coli pGEX-2T plasmid. This system was shown to be ideal for the expression and purification of GST-fusion proteins in bacteria since the recombinant protein can be affinity purified from the contaminating proteins by binding the glutathione-S-transferase fusion to the glutathione substrate that is cross-linked to Sepharose beads (Guan and Dixon, 1991). Oligonuleotides with Eco R l sites were patterned after the 5' end and 3' end coding regions of the 26a(3-3 human Erkl clone. PCR was performed under very stringent conditions to minimize mutations. In addition, VENT polymerase which has proof-reading abilities, was used during the PCR reaction to maintain sequence fidelity. As a cautionary measure, the amplified product was verified for mutations by DNA sequencing. After amplification, the 1.2 kb fragment was digested with Eco Rland subcloned into the pGEX-2T vector (Figure 12). The proper orientation of the Erkl cDNA was determined by restriction analysis with Sma I.  Using the pGEX-2T  prokaryotic expression vector, production of recombinant fusion proteins is tightly regulated by the LacZ promotor. The highest expression levels of GST-Erkl were obtained by the addition of 80-100 uM of isopropyl-(3-D-thiogalactopyranoside .  We first determined the ability of human Erkl to be expressed as a recombinant fusion protein in E. coli by using a series of polyclonal antibodies directed against known human, rat Erkl and Mpkl peptide sequences as well as purified full-length sea star Mpkl. The glutathione-affinity purified GST-Erkl (4 ug), its Mono Q concentrated thrombin-cleaved Erkl (1 ug) and purified sea star Mpkl (0.5 ug) were electrophoresed  Figure 12: pGEX-2T Prokaryotic expression vector of the full-length human Erkl cDNA cloned in-frame with the glutathione-S-transferase (GST) protein. Construction of the vector is described in Materials and Methods. The 3' end of the glutathione-S-transferase protein was fused in frame with the the 5' human Erkl cDNA coding sequence by a two cytosine base linker (bold). As a result, a histidyl amino acid was exchanged for a prolyl residue (bold). This two base intervening linker was added to maintain the proper reading frame during protein translation. The thrombin cleavage site recognition sequence is indicated by a vertical arrow. The pBR322 origin of replication, ampicilin resistance gene, and the LacZ inhibitor protein gene, laclq, are also presented.  Thrombin cleavage site  Filler  Human Erkl  L V P R y G S P G I P M A A A A A Q G CGT GTT CCG CGT GGA TCC.CCG 5GA GGA ATT CCC ATG GCG GCG GCG GCG GCT CAG GGG  BamH  t  Smal  EcoRI  1632 Hind III 1892  NcoI2058 EcoRI 2079 AatII2359  PstI3036  Bpml 4639 AlwNI3756  and transferred to nitrocellulose for Western immunoblotting. These anti-peptide antibodies recognize unique sequences (see Table5 for details) near the amino terminal (anti-Erkl-NT [Figure 13A]) of human Erkl (Charest et al, 1993), the kinase subdomain III (anti-Erkl-III [Figure 13C]), and carboxy terminal (anti-Erkl-CT [Figure 13D]) sequences of rat Erkl (Boulton et al., 1990), as well as a peptide patterned after the ATPbinding site of subdomain I of Mpkl (anti-Mpk-I [Figure 13B]). The purified Mpkl (anti-Mpk-W [Figure 13E]) from sea star was also used to generate antibodies against the complete protein (Sanghera et al, 1990; Sanghera et al, 1992). A GST-fusion protein of approximately 70-kDa in size (molecular mass of Erkl, 44-kDa and of GST, 26-kDa) was recognized by all the antibodies from cell lysates of bacteria transformed with the pGEX-Erkl expression vector (Figure 13, lane 3). The 44-kDa thrombin-cleaved Erkl protein was also detected by the anti-Erkl-NT and -CT antibodies (Figure 13 lane 5). A ladder of smaller bands (Figure 13A, B and ) were also detected by anti-Erkl peptide antibodies. These are probably the result of aborted translation or proteolytic cleavage since a similar pattern was not observed with thrombin cleaved Erkl (Figure 13 compare lanes 3 and 5). No protein bands, however, were observed from control bacterial cell lysates expressing only GST (Figure 13, lane 2). Purified Mpkl from sea star protein was more easily revealed by whole protein Mpk antibody than with any of the peptide antibodies (Figure 13, lane 4). These data indicate that human Erkl can be expressed in E. coli  2.2  and subsequently purified in reasonably large quantities for experimentation.  Characterization of autophosphorylated Erkl  In vivo MAP kinases are activated in response to growth hormones by becoming phosphorylated on threonyl and tyrosyl residues (Ray and Sturgill, 1988; Anderson et al., 1990). The mechanism that regulated these phosphorylation events were not established at the time these enzymes were being cloned and sequenced.  As a first step in  Figure 13: Expression and immunoreactivity of recombinant Erkl protein from E. coli to M A P kinase antibodies. Isopropyl-(3-D-thiogalactopyranoside-induced bacterial lysates from pGEX control plasmid (lane 2), lysates from IPTG-induced pGEX-Erkl plasmid (lane 3), 0.5 mg of purified sea star Mpkl (lane 4) (Sanghara et al., 1990) and 1.0 mg of thrombin-cleaved glutathione affinity purified Erkl (lane 5) were electrophoresed on a 10% SDS-polyacrylamide gel. The MAP kinase proteins were revealed by Western immunoblotting the nitrocellulose filters with anti-peptide antibodies Erkl-NT (A), MpkI (B), Erkl-III (C), Erkl-CT (D), whole protein antibodies Mpkl-W (E). The Roman numeral designation specifies the catalytic subdomain location used to synthesize the peptides (Hanks et al., 1988). Lane 1 shows the prestained standard proteins: bovine serumn albumin, 87-kDa; ovalbumin, 50-kDa; carbonic anhydrase, 33-kDa soybean trypsin inhibitor, 29-kDa; and lysozyme, 21-kDa. The black marks indicate the positions of the markers.  1 2 3 4 5  1 2 3 4 1 2 3 4 B  A  GST-Erkl Mpkl -  C  <r  i  •m  •  kDa 106 80 50  -  28  a-Erkl-NT  a-Mpk-I  1 2 3 4 5 D  33  —  m  a-Erkl-III  1 2 3 4 1 2 3 4 E  GST-Erkl -  -a  Mpkl -  50 33  wmm  —  cc-Erkl-CT  kDa - 106 80  a-Mpk-W a-PY4G10  28  understanding the regulation of MAP kinases at a molecular level, the phosphorylation state of the human GST-Erkl preparation was analyzed with the monoclonal antiphosphotyrosine antibody 4G10. While MAP kinase purified from sea star oocytes was shown to be regulated by tyrosyl phosphorylation (Figure 13, lane 4) (Sanghara et al, 1994), we also detected the presence of phosphotyrosine in the non-activated form of Erkl purified from bacteria (Figure 13, lane 3). The detection of Erkl with tyrosylspecific 4G10 antibodies seemed to be comparable to the levels obsserved with antiErkl-specific antibodies (compare panels A, C, D, and F of Figure 13). This seemed unusual since there have been no reports of tyrosyl phosphorylation in prokaryotes and no reactivity with 4G10 was detected in control bacterial lysates (Figure 13, lane 2) (Schieven et ai, 1986). Therefore, the most likely explanation for the presence of phosphotyrosine in GST-Erkl was likely the result of an autophosphorylation reaction. As mentioned earlier, MAP kinases become fully activated by phosphorylation on threonyl and tyrosyl sites (Sturgill et al., 1988; Anderson et al., 1990). Because 4G10 is specific for phosphotyrosine, the data could not exclude the possibilty of a threonine autophosphorylation reaction.  A different approach to determing whether  autophosphorylation leads to increased phosphotheonine is to perform an autocatalytic reaction in the presence of [y- P]ATP followed by phosphoamino acid analysis of the 32  labelled protein.  We examined the possibility that autophosphorylation of Erkl might regulate its phosphotransferase activity. To do so, we linked GST-Erkl to glutathione-agarose beads and incubated the fusion protein with [y- P]ATP under conditions that promoted 32  autophosphorylation in vitro . A single radiolabelled protein band that was identical in size to the Erkl immunoreactive bands was identified as Erkl using MAP kinase-specific antibodies (Figure 14A and C).  Similar results were obtained using identical assay  conditions in which the thrombin-cleaved Erkl was incubated with radiolabel (Figure  Chapter 14: Autophosphorylation and phosphoamino acid analysis of the Erkl recombinant protein. GST-Erkl fusion (A and C) and thrombin-cleaved Erkl proteins, Erkl-U (U = upper band) and Erkl-L (L = lower band) (B and D) were incubated with Y "P[ATP] (9000 cpm/pmol) in the presence of either 10 mM M n C l (lanes 2) or MgCl (lanes 3) separately or combined (lanes 4). Autoradiograms are shown in panels A and B, and immunoblots in panels C and D. Phosphoamino acid analysis was performed on autophosphorylated bands of GST-Erkl (E) and cleaved-Erkl (F) proteins in the presence of M n or M g . The prestained protein standards (lanes 1) are displayed with radioactive ink (A and B) and with bars (C and D). The migration of the free-phosphate and phosphoamino acid standards are indicated in panels E and F. 32  2  2 +  2  2 +  152  1  c~ ca •— oh o  2 3 4  kDa  2 3 4  kDa  50  1-80  Erkl-U Erkl-L  GST-Erkl  —  OS iO  3  <  STD  A A S S  rs_  |  +  CN  2 3 4  1  kDa  2  3  4  kDa  50  80  *Erkl-U "Erkl-L  GST-Erkl  c  STD  <N  CM  <N  c  +  CN 00  3  4  -Phosphate-  f BO  -Phosphoserine-  9  -Phosphothreonine-  o 53  -Phosphotyrosine-  O  o CU  Mn  2 +  Mg  2 +  Mn  2 +  Mn  2 +  Mg  2 +  Mg  2 +  -J  GST-Erkl  _^  n  M  P3  153  14B and D). Thus, the presence or removal of the 26-kDa GST protein had no adverse effects on Erkl enzyme activity. Interestingly, antiphosphotyrosine immunoblotting of GST-free Erkl revealed two antibody-reactive proteins of approximately 43-kDa (ErklL) and 44-kDa (Erkl-U) while only a single band was observed with GST-Erkl (Figure 14C and D). When the autoradiogram in Figure 14B was underexposed both forms were shown to be radiolabeled with  [y- P]ATP. 32  Both species  autophosphorylation and were recognized by 4G10 antibodies.  underwent  The most likely  explanation for the doublet is the existence of a cryptic thrombin-cleavage site located at the carboxy-terminal end of the protein. Coincidently, the kinase-inactive form of Erkl (A71) also displayed a similar doublet after cleavage with thrombin. This result indicates that the observed doublet is the probably due to improper thrombin-cleavage of the GSTErkl and not a band shift that has been associated with activation of the Erkl protein kinase.  The divalent cation requirement for GST- and cleaved-Erkl during autophosphorylation reactions was examined (Figure 14A and B). M n C l and M g C l 2  2  were chosen since these two cations are required in most tyrosyl and seryl/threonyl phosphorylation reactions. The autophosphorylation activity was more efficiently stimulated in the presence of 10 mM M n  2 +  than with 10 mM M g  2 +  during the 30 min  incubation period. A similar level of radiolabel incorporation was achieved with the combination of divalent cations as with M n  2 +  alone.  The stoichiometry of GST-Erkl or thrombin-cleaved Erkl phosphorylation as assessed by  3 2  P incorporation was approximately 0.01 mol of phosphate per mol of  enzyme (data not shown). Since MAP kinases apparently require threonyl and tyrosyl phosphorylation to become activated, the low stoichiometry implied that Erkl poorly autophosphorylated in vitro. Another possible explanation for these findings is that Erkl  was already heavily autophosphorylated in E. coli during the course of protein expression (Figure 14C and D). The amount of phosphate incorporated in a half hour reaction was quite low when compared to the phosphotyrosyl content that occurs in situ in bacteria. This is best illustrated when comparing the differences in P labelling of Erkl-U and 3 2  Erkl-L proteins incubated in the presence of M n  2 +  or M g  2 +  divalent metal cations  (Figure 14B, compare lanes 2 and 3). Although the autophosphorylation reaction was easily observed by radiolabelling with [y- P]ATP, this increase in Erkl phosphotyrosine 32  did not result in greater immunoreactivity in Western blots with 4G10 antiphosphotyrosine antibody (Figure 14D, compare lanes 2 and 3). Perhaps the higher phosphotyrosine content of GST-Erkl was attributable to the lengthy IPTG induction period (12 h) that was used to maximize GST-Erkl protein synthesis. Since prokaryotes are unlikely to express endogenous tyrosyl phosphatases, Erkl that autophosphorylated on tyrosyl residues could accumulate over a longer duration.  Another possible  explanation for the absense of any observable differences may be the lack of sensitivity in detecting subtle changes by the Western immunoblotting technique. MAP kinases have been reported to require threonyl in addition to tyrosyl phosphorylation to become activated in vivo (Anderson et al, 1990; Payne et al, 1991). Therefore, the in vitro autophosphorylated proteins in Figure 14A and B were further characterized by phosphoamino acid analysis. In the presence of either divalent cation, GST-Erkl (Figure 14E lanes 1 and 2) and cleaved Erkl (Figure 14F Erkl-U, lanes 1 and 3; Erkl-L, lanes 2 and 4) were phosphorylated on tyrosyl. Additionally, GST-Erkl and Erkl-U were partially seryl phosphorylated in presence of M n  2 +  (Figure 14E and F lane 1). However,  almost no phosphothreonine was detected for any of the phosphorylated protein bands.  The GST-Erkl and thrombin-cleaved Erkl autophosphorylation sites were determined by two-dimensional tryptic phosphopeptide mapping. The phosphopeptide pattern of seven distinct spots was identical for both forms of Erkl (Figure 15A and B).  Figure 15: Two-dimensional tryptic phosphopeptide mapping and phosphoamino acid analysis of autophosphorylated recombinant Erkl. GST-Erkl fusion (A) and thrombincleaved Erkl (B) proteins were incubated with [y- P]ATP (10,000 cpm/pmol) and separated on 10% SDS-polyacrylamide gels before excising and digesting the bands with trypsin as described in Materials and Methods. The open circle marks the origin. The direction of the electrophoresis (positive and negative poles) and thin layer chromatography steps are indicated by arrows. The numbered spots in panel B were removed and further analyzed by phosphoamino acid analysis (C). 32  Cleaved Erkl protein  GST-Erkl protein  Thin layer chromatography  Phosphate  Phosphoserine Phosphothreonine  ^>  Phosphotyrosine  • Origin  *  Erkl tryptic spot number  A number was assigned to each phosphopeptide for comparison purposes. Spots 1 and 2 appear to be the major phosphorylated peptides. The phosphopeptides were further subjected to phosphoamino acid analysis (Figure 15C).  The majority of the  phosphorylated spots (1, 2, 3, 5, and 7) were tyrosyl phosphorylated while the remaining two spots (4 and 6) were seryl phosphorylated. phosphorylation was present.  Interestingly, no threonyl  Therefore, these results indicate that Erkl  autophosphorylates at multiple tyrosyl and seryl residues in vitro.  Although the the  protein was digested for close to 24 hours, it may be possible that an incomplete tryptic digest generated multiple phosphopeptides.  To determine which tryptic phosphopeptides in Figure 15 corresponded to potential in vivo regulatory sites (i.e., Tyr-204) murine Lck purified from baculovirus was used to phosphorylate recombinant thrombin-cleaved Erkl. A similar approach was used to tyrosyl phosphorylate and activate sea star Mpkl (Ettehadieh et al, 1992). Erkl was initially preincubated with unlabelled ATP as a measure to reduce autophosphorylation prior to addition of [y- P]ATP and Lck. A similar tryptic phosphopeptide mapping 32  pattern was achieved during the in vitro phosphorylation of Erkl by Lck and some of this may be the result of autophosphorylation (Figure 16B).  The major tyrosyl  phosphopeptide sites 1 and 2 were differentially phosphorylated. The only novel Lck site was phosphopeptide 8. To confirm which tryptic phosphopeptides, if any, contained the Thr-202 and Tyr-204 (TEY peptide IADPEHDHTGFLTEYVATR) identified by Payne et al. (1991), a synthetic peptide patterned after this sequence was phosphorylated with Lck and subjected to the same trypsin treatment as performed for the Erkl proteins (Figure 16C). Tryptic phosphopeptides sites 1 and 2 were generated with the TEY phosphorylation peptide and therefore Tyr-204 represents the major tyrosyl autophosphorylation sites in Erkl (Figure 16C). The GST-Erkl and TEY peptide both generated spots 1 and 2 after trypsin digest and electrophoresis. This result was  Figure 16: Two-dimensional tryptic phosphopeptide mapping and phosphoamino acid analysis of p56 phosphorylated recombinant Erkl. Panel A, Thrombin-cleaved Erkl was incubated with [y- P]ATP under autophosphorylating conditions for 20 min; Panel B, the same thrombin-cleaved Erkl protein was pretreated with unlabeled ATP as 'in panel A before incubation with [y- P]ATP in the presence of Lck; Panel C, the Erkl regulatory phosphorylation site T E Y peptide (IADPEHDHTGFLTEYVATR) was phosphorylated by Lck and purified by high-pressure liquid chromatography before trypsin treatment; Panel D, thrombin-cleaved Erkl-E204 allele phosphorylated by Lck as in panel B. Variable migration was observed with spot 6. lck  32  32  Erkl  Erkl + Lck  °  "  A  ,  3^  »  5  6  o  4 2  2  •n^^HP^^^^^^ 1"  TEY peptide + Lck  Erk 1 (E204) + Lck  o 3 2  '  5  ot 4  'TP) u 0  & O u  4—1  o _o U  O Thin layer chromatography  160  unexpected, since the sequence of the synthetic peptide contained just one tyrosyl residue. A possible explanation for this anomally may be that the trypsin preparation was contaminated with another protease that cleaved the TEY phosphopeptide. A second experimental approach was used to confirm that Tyr-204 was the principal autophosphorylation and Lck phosphorylation site in vitro. A Tyr-204 phosphorylation site allele was created by converting it to a negatively charged glutamyl residue by polymerase chain reaction site-directed mutagenesis. In phosphorylation experiments with Lck, Glu-204 Erkl substrate was phosphorylated specifically at phosphopeptide sites 5 and 8 (Figure 16D). No autophosphorylation was observed with the catalytically compromised Glu-204 Erkl. Furthermore, Lck was also demonstrated to phosphorylate the regulatory T E Y peptide in a parallel experiment which is apparent from the appearance of spots 1 and 2 in Figure 16C. Hence, Lck is capable of phosphorylating Erkl protein kinase on the Tyr-204 in vitro.  2.3  MBP phosphotransferase activity of GST-Erkl  A small amount of MBP phosphotransferase activity was detectable in recombinant GST-Erkl and cleaved Erkl purified from bacteria.  M B P was  phosphorylated at a rate of ~ 2 nmol/min/mg of enzyme, a rate that was substantially lower than the 324 nmol/min/mg obtained from the 42-kDa MAP kinase purified from Xenopus  oocytes using the same substrate (Gotoh et al, 1991). The importance of  threonyl and tyrosyl phosphorylation in Erkl activity was examined with phosphatases that selectively removed these residues in vitro. Tyrosine-specific protein phosphatase CD45 incubation with GST-Erkl led to a 50% decrease in MBP phosphotransferase activity in 2 min and complete inactivation of the enzyme after 60 min of incubation (Figure 17A). This was associated with a comparable decrease in the amount of detectable tyrosyl phosphorylation in GST-Erkl by Western analysis (Figure 17C).  Figure 17: Phosphatase treatment of recombinant Erkl. GST-Erkl bound to glutathioneagarose beads was incubated with 0.5 ug of purified protein tyrosine phosphatase CD45 (A and B) or ~4 U of human protein phosphatase 2A (C and D) for the specified times before quenching the reaction with 1 mM sodium orthovanadate or 1 uM okadaic acid respectively. After washing the beads several times with Buffer G, the material was divided in half and subjected to M B P kinase assays (A and B) or Western immunoblotting analysis with the 4G10 antiphophotyrosine antibody (C and D). The data are expressed relative to the non-phosphatase treated control GST-Erkl beads. For protein phosphatase 2A, 1 U releases 1 nmol of phosphate per min from 15 u M phosphorylase at 30°C.  A  C  I r3 |  e 12  CL  CL  100 80  h  Control PP2A  60  CQ  CQ  2 w  uy •—  c  a CL  a  u  CL  40  20  T—•—r  0  10  20  30  40  50  60  0  10  20  Time (min)  >\  % 50KDa-  40  50  60  Time (min)  B 2 80KDaO •*  30  D  11:111 0  2  5 10 20 30 60  Time (min)  LGST-Erkl  2 80KDaO  GST-Erkl  % 50KDa0  2  5 10 20 30 60 Time (min)  163  However, treatment with the seryl/threonyl specific protein phosphatase 2A did not substantially reduce the MBP phosphotransferase activity of GST-Erkl (Figure 17B). The phosphotyrosyl levels for GST-Erkl as revealed by the antiphosphotyrosine antibodies remained constant throughout the incubation with PP2A (Figure 17D). Staining of the protein bands with Ponceau S solution revealed that the amount of Erkl enzyme assayed was identical for each time point (data not shown). As a control, protein phosphatase 2A (PP2A) was shown to be active through its ability to dephosphorylate a synthetic phosphopeptide with the sequence Lys-Arg-Thr(P)-Ile-Arg-Arg (data not shown). Hence, autophosphorylation on Tyr-204 was necessary to minimally activate Erkl phosphotransferase activity.  2.4  Activation of GST-Erkl by a MAP kinase kinase  Lck phosphorylated Erkl in vitro to a stoichiometry of 0.1 mol of P per mol on a number of different tyrosyl sites including the TEY containing sites 1 and 2 (Figure 17) However, no increase in Erkl MBP phosphotransferase activity by Lck was noted during these kinase reactions even though it was reported that sea star Mpk was activated by Lck under similar circumstances (Ettehadieh et al, 1992) (data not shown). Because recombinant expressed bacterial Erkl is already substantially phosphorylated on tyrosyl, further phosphorylation by Lck may be insufficient to stimulate further its MBP kinase activity. It is also possible that Lck phosphorylates a tyrosyl site that inactivates the phosphotransferase activity of Erkl. The unique tryptic phosphopeptide site 8 which was only detected when Erkl was incubated with Lck is a good candidate. One candidate tyrosyl site is located in the ATP-binding region (IGEGAYGMV) of subdomain I which is found in some M A P kinases. Phosphorylation of a homologous site in cyclindependent kinase 2 (CDK2) by the dual-specificity kinase Weel leads to the diminution  of its histone HI phosphotransferase activity in vivo (Parker and Piwnica-Worms, 1992; McGowan and Russell, 1993; Parker et al, 1995).  MAP kinase activators that have been identified in different PMA-treated mammalian cells and apparently stimulate MBP phosphotransferase activity by phosphorylating the regulatory threonyl and tyrosyl phosphorylation sites of Erks (Wu et al, 1991; Alessandrini et al, 1992; Rossomando et al, 1992). Since Lck was unable to activate Erkl directly in vitro, a different approach using a GST-Erkl two-step activation assay was used to identify possible regulatory kinases. This work was undertaken as part of a separate study to investigate the role MAP kinase pathway plays in the regulation of platelet aggregation (Samiei et al, 1993).  The experimental treatment used to detect MAP and S6 protein kinases in sheep platelets was also used to investigate the presence of an upstream activator. Purified platelets were exposed to 200 nM PMA and homogenized after 5 min incubation period. The crude extract was applied to a DEAE-cellulose resin at neutral pH. The break through material was recovered and diluted in 10 volumes of MES buffer pH 6.5 before fractionation on a Mono S cation exchange column. Activation of the partially purified preparation of MAP kinase kinase (MAPKK) was determined by assaying for GST-Erkl phosphotransferase activity in the presence of [y- P]ATP. A three-fold increase in 32  phosphorylation was observed when compared with autophosphorylated Erkl (Figure 18A, lanes 1 and 3). Analysis of the phosphoamino acid content of Erkl phosphorylated with M A P K K showed an elevation in tyrosyl phosphorylation while the presence of seryl and threonyl was almost undetectable (Figure 18C). MAPKK purified from unstimulated platelets displayed almost no GST-Erkl phosphorylation under the same experimental conditions (Figure 18A, lane 2). The increase in GST-Erkl phosphorylation by PMAactivated sheep platelet MAPKK stimulated GST-Erkl MBP phosphotransferase activity  Figure 18: MAP kinase kinase phosphorylation and activation of recombinant Erkl. Partially purified MAP kinase kinase from sheep platelets stimulated for 5 min with 200 mM PMA (P-MKK) and from unstimulated sheep platelets (U-MKK) was used to phosphorylate GST-Erkl linked to glutathione agarose beads (A) and promote its phosphotransferase activity (B) by the technique outlined in Materials and Methods. Approximately 1 ug each of Erkl wild type (K71) and catalytically compromised (A71) in which the Lys that is essential for the phosphotransferase reaction is changed to an Ala (Kamps et al., 1994). SDS-PAGE loading buffer was added at the completion of the reaction. After electrophoresis, the proteins were transferred to nitrocellulose then autoradiographed for 15 min. The incorporated P was quantitated by liquid scintillation counting and is described as fold change above wild type with no extract. The top and bottom edges of each gel define the positions of the prestained standards: bovine serum albumin, 80-kDa; and ovalbumin, 50-kDa in panel A and soybean trypsin inhibitor, 28-kDa; and lysozyme, 19-kDa; in panel B. Lane 1, wild type Erkl TEY with no extract; lane 2, wild type Erkl TEY with untreated extract; lane 3, wild type Erkl TEY with PMA-treated extract; lane 4, inactive Erkl TEY* without extract; lane 5, inactive Erkl TEY* with PMA-treated extract. The P-MKK phosphorylated GST-Erkl (K71) in lane 3 of panel A was excised and subjected to phosphoamino acid analysis (C). Migrations of the free phosphate and phosphoamino acid standards are indicated. 3 2  by five-fold while the same kinase purified from control platelets had only marginal effects (Figure 18B lanes 1-3).  There was some speculation that the MAP kinase upstream regulator was a protein devoid of kinase catalytic activity that acted as an allosteric activator, since MAP kinases are capable of limited activation through autophosphorylation (Wu et al, 1991; Seger et al, 1991; Ahn et al, 1991; Robbins and Cobb, 1992). To address this controversy, Lys-71 was mutated to an Ala in human Erkl thereby compromising catalytic activity of GST-Erkl. A homologous Lys residue located in subdomain II of other protein kinases has been shown to be essential for the catalytic activity of protein kinases (Kamps et al, 1984).  The Ala71-GST-Erkl allele displayed almost no  autophosphorylating activity (6%) or MBP phosphotransferase activity (7%) in the absence of activated M A P K K from sheep platelets (Figure 18A and B, lane 4). However, Ala71-GST-Erkl was phosphorylated to 80% of what was observed in the same experiment with wild type Lys-GST-Erkl when treated with PMA-activated M A P K K and 38-fold greater than its autophosphorylation (Figure 18A compare lanes 3, 4 and 5). These results confirmed that the MAP kinase activator from PMA-treated sheep platelets was a bona fide protein kinase. Interestingly, M A P K K phosphorylation of catalytically compromised Ala71-GST-Erkl unexpectedly stimulated its MBP phosphotransferase activity 16-fold above what was achieved with control autophosphorylation (Figure 18 lanes 4 and 5). This limited Ala-71-GST-Erkl stimulation may be attributable to a second lysyl residue (Lys-72) that may compensate in a limited fashion for the mutation at Lys-71 (Figure 10).  PURIFICATION AND CHARACTERIZATION OF A SEA STAR MAP  3.  KINASE ACTIVATOR  3.1  Detection of MAP kinase activator during oocyte maturation  Germinal-vesicle breakdown (GVBD) indicates meiotic maturation in sea star oocytes. This event, which occurs within -80 min of ooctyes being exposed to the hormone 1-methyladenine, is characterized by disintegration of the nuclear envelope and chromosomal condensation in a manner similar to meiotic and mitotic cell divisions in other model systems. This makes the sea star oocyte system suitable for studying cell cycle events, since large numbers of synchronized cells arrested in prophase I of meiosis I can be isolated for cell and biochemical studies. Unlike sheep platelets which yield an insufficient quantity of protein for purification, each sea star can yield millions of quiescent cells to provide gram amounts of protein for this same purpose.  Activation of several distinct protein kinases have been detected in maturing oocytes (Pelech et al, 1988; 1991). One of these kinases, meiosis-activated MBP kinase (Mpk), was first identified because of its ability to phosphorylate the exogenous substrate MBP, was purified to homogeneity (Pelech et al., 1988; Sanghera et al, 1990). Peptide sequences derived from purified Mpk displayed a high degree of homology to rat and Xenopus  MAP kinases (Posada et al, 1990). Furthermore, a partial cDNA sequence  obtained from RT-PCR using sea star mRNA yielded two contiguous fragments that displayed a large degree of similarity with human Erkl in the regions of overlap (Figure 19). Of the seven conserved regions defined by Hanks et al. (1988), the amino acid sequence located between subdomains X and X I reveals the highest amino acid divergence (42% identical) between the partial sea star cDNA and the full-length human Erkl clone. However, a high degree of identity was also observed in the central region of  Figure 19: Sequence comparison between Erkl cDNA from human and an Erk-like cDNA from sea star oocytes. Full-length Erkl cDNA obtained from the liver Hep G2 cell liver line was aligned with a partial sequence from several overlapping cDNA fragments identified by PCR amplification from sea star oocytes. The MAP kinase signature regulatory phosphorylation sites (TXY) are indicated by asterisks. Amino acid residues that are conserved in all kinases are highlighted in boldface type. The consensus sequence is indicated while non-matching residues are represented by a hyphen.  Human E r k l  IQILLFEFtfEIWIG:^  Sea s t a r n p k  IKILTRFRHENIINIQDIIHAOT  consensus  I-IL-RFRHEN-I-I-DI—A-T  M-LWYTA7Q-Ijy[ET-LYKLIi<;-Q-LSN  VI Human  Erkl  VII  DHICYFLYQITJ^GIJC^^  Sea s t a r n p k  rjHTSYFT  consensus  DHI-YT^YOII^GLKYIHSAIWIHRK^  , Y Q T T  * Human  Erkl  *  ,Pf?T ^ Y T H C I A M T r HRDT ,TCP.qKTT ,T ,T iNTTTTlT .KTCETOIARIAD  VIII  IX  PEHDHTCFLTEY\/ATE^^  Sea s t a r n p k  PVHDHIC^TEYAZATRWYRA^  consensus  P-EDHIGT^TEYmTHATYT^  DIWSVGCILAEML  PI  Human E r k l  FPGKHYJIX)TJNmLGimSPSQ^  Sea s t a r n p k  FPGKHYIJDQTJSHILNITJGSPSCE^  consensus  FPGKHYIiy^IieiL-imSPS-EDL-CI-N-KA--Y-QSLP-K--V-W--L-  XI Human E r k l Sea s t a r n p k consensus  PKSDSKATX1IJ1DRMLTEIXEENKR GAADPKS  I^IJLDRILTFNPDKR  EKS  L-LLDR-LTENP-KR  171  the catalytic portion of partial Mpk protein (Figure 19). Therefore, there appeared to be sufficient conservation between human Erkl and the partial Mpk clone to justify using the bacterial expressed recombinant human GST-Erkl fusion protein to identify and purify to homogeneity a MAP kinase activator from sea star oocytes.  To determine the existence of a MAP kinase activator in sea star oocytes, 2 mg of cytosolic extract from immature oocytes and cells treated with 1-methyladenine were applied separately to a Mono Q anion exchange column. The protein was eluted with a 15 ml linear 0.8 M NaCl gradient and assayed for kinase activities toward a battery of substrates. As shown in Figure 20D, the MBP phosphorylating activity eluted as a broad peak at a concentration between 200-400 mM NaCl. The protein kinase displayed an identical chromatographic profile to the previously purified sea star MAP kinase, Mpk (Sanghera et al, 1990). When these same fractions were assessed for MAP kinase kinase activity using the kinase-inactive GST-Erkl fusion protein as a substrate, three major peaks of activity were observed in fractions 21, 25 and 33, respectively (Figure 20A). Additionally, three minor peaks of kinase activity were detected in fractions 7, 15 and 47 from the same Mono Q column. A similar kinase activity profile was observed using the TEY peptide as a substrate (compare Figure 20A and B). The 19 amino acid TEY peptide (IADPEHDHTGFLJEYVATR)  was patterned  after  the  regulatory  phosphorylation sites Thr-202 and Tyr-204 and surrounding sequence in the activation loop of human Erkl (Charest et al, 1993). These phosphorylation sites were first uncovered at a similar position in the MAP kinase isoform Erk2 (Thr-183 and Tyr-185) by mass spectrometry (Payne et al, 1991). Since phosphorylation of the full-length Erkl and TEY peptide uncovered at least six distinct kinase activities that were capable of phosphorylating one or more of the threonyl or tyrosyl residues present in this peptide region, it was necessary to determine which one these phosphorylation events could regulate Erkl phosphotransferase activity. A two-step Erkl phosphorylation and  Figure 20: Detection of Erkl activator activity in 1-methyladenine-treated sea star oocytes. Sea star extracts (2 mg) from immature (open circles) and mature (closed circles) were applied to a Mono Q anion exchange column and developed with a linear 0.8 M NaCl salt gradient and collected in 25 ul fractions. A 10 ul aliquot from every odd numbered fraction was tested for phosphotransferase activtity toward (A) full-length recombinant GST-Erkl protein; (B) a peptide termed 'TEY' which is patterned after the MAP kinase regulatory phosphorylation sites; (C) Erkl phosphotransferase activity toward MBP following preincubation with column fractions (see Materials and Methods); and (D) myelin basic protein.  0.5 i  NaCl concentration (M) 0 0.4 0.8 1  1  Fraction number  1  activation reaction was performed by first pre-incubating GST-Erkl immobilized on glutathione beads with column fraction in the presence of 50 mM non-radiolabelled ATP. After 20 min, the beads were washed three times in buffer and the Erkl assessed for MBP phosphotransferase activity. A maturation-stimulated MBP phosphorylating activity eluted in the 100 mM NaCl range just after the void volume from the Mono Q column. This indicated that the MAP kinase activator binds with low affinity to anion exchange resins equilibrated at pH 7.2. The MBP phosphotransferase activity for Erkl was stimulated 7-fold by a MAP kinase activator that became stimulated when the oocytes were treated with hormone.  However, the peak GST-Erkl M B P  phoshotransferase activity ( Figure 20C; fractions 11 through 15) was stimulated by the MAP kinase activator that eluted earlier from the Mono Q column. If fact this GST-Erkl activation peak did not coincide with the GST-Erkl nor the TEY phosphorylation peaks (Figure 20A and B; fractions 19 through 39) when these were used as in vitro substrates. In fact, Western blotting with anti-peptide antibodies revealed that peak IV was Casein kinase II. Therefore, the enzyme peaks that were capable of phosphorylating GST-Erkl and TEY peptide did not overlap with the activation peak that stimulated Erkl MBP phosphotransferase activity. This contradicted the result decribed earlier in sheep platelets (Figure 18 compare lanes 3 and 5). However, at that time, several research groups speculated that the upstream MAP kinase regulator may be a factor that could activate MAP kinase by an allosteric reaction (Wu et al., 1991; Seger et al., 1991; Ahn et al., 1991; Robbins and Cobb, 1992). Therefore, it was critical to determine if the sea star activating factor that regulated GST-Erkl activity was distinct from the MAP kinase kinase detected in sheep platelets.  3.2  Purification of MAP kinase activator  The major peak of MAP kinase activator activity was easily separated by anion exchange chromatography from the sea star MBP kinase Mpk (Figure 20 compare panels C and D). A purification strategy was devised for the isolation and characterization of this activating factor. Extracts prepared from the cytosolic portion of mature oocytes were applied to a DEAE-cellulose and the breakthrough material directly loaded onto a hydoxylapatite column linked in series. Assessment of the eluted DEAE-cellulose fractions for kinase activity revealed that none of the MAP kinase activator activity bound to this weaker anion exchange column (Figure 21A). However, the activator activity did adsorb to hydroxylapatite. It was released from the hydroxylapatite column as a broad peak between 20-100 mM phosphate, in a fractionation region separate from the bulk of the contaminating proteins (Figure 21 A).  The fractions containing the highest activator activity were combined, and after dilution in buffer, adsorbed onto an S-Sepharose column. The highest Erkl activator activity eluted as a narrow peak between 100-135 mM NaCl, with a large amount of contaminating proteins eluting in the later fractions (Figure 2IB). The most active fractions were pooled and the relative salt concentration reduced by dilution in column buffer before application onto a phosphocellulose column. A sharp peak of activator activity was released from the phosphocellulose resin with -270 mM-NaCl just prior to the elution of bulk of the contaminating proteins (Figure 21C). The combined fractions were diluted and loaded onto a heparin-agarose column. After development of the heparin-agarose with 200 mM linear NaCl gradient, the activator activity was eluted as a single peak with 150 mM salt (Figure 21D). Only minor amounts of contaminating protein was present in these fractions.  Figure 21: Purification of Erkl activator from mature sea star oocytes. The sea star Erkl activator was purified using six different chromatographic resins; DEAE-cellulose (reverse column); (A) hydroxylapatite; (B) S-Sepharose; (C) phosphocellulose; (D) heparin-agarose; and (E) Mono S. All columns were developed with a linear NaCl gradient except the hydroxylapatite which required potassium phosphate. Every odd numbered fraction was assayed for the presence activator indirectly by assaying for increased Erkl MBP phosphotransferase activity (closed circles) as described in Material and Methods. The protein concentration (open circles) was determined at 595 nm for each fraction assayed.  177  Phosphate concentration (M)  Hydroxylapatite fraction number NaCl concentration (M)  0  10 20 30 40 50 S-Sepharose fraction number NaCl concentration (M)  0  10  20  30  40  50  60  Phosphocellulose fraction number 178  NaCl concentration (M)  Heparin-agarose fraction number NaCl concentration (M) 0.015 ^4 \-<  0.012  a  0.009  m ON m -4—•  (3  w O  co  0.006  a 2 2  o  <  0.003  CL  1/1  (D o C  •e o  1/3  o 0 MonoS fraction number  179  For the remaining purification step, the most active fractions from the heparinagarose column were combined and the pH reduced by dialyzing against MES buffer [pH 6.5].  The dialysate was applied to a Mono S column (Figure 21E). With this  concentration step, the volume of the MAP kinase activator was reduced to one tenth the original volume. The most active fractions were detected at a concentration of 175 mM NaCl. A summary of the MAP kinase activator purification is outlined in Table 10. After six column chromatography steps, the MAP kinase activator was purified 536-fold for an overall recovery of 0.2%. Approximately 8 ug of protein was obtained from 2 g of crude oocyte cytosolic extract. The hydroxylapatite column eliminated nearly 75% of the contaminating proteins, while both hydroxylapatite (8-fold) and heparin-agarose (7-fold) resins yielded the best overall purification of the activator.  This MAP kinase activator was a highly labile enzyme. Often the stimulatory activity of the activating factor was lost after two or three column purification steps for no apparent reason. Attempts to stabilize the protein by increasing the concentration of phosphatase inhibitors or by addition of protease inhibitors (soybean trypsin inhibitor, aprotinin, and protamine) had minimal effect on maintaining the enzyme activity during the purification. The purified enzyme was also inactivated following storage at 4°C for more than 3 h or after a freeze-thaw cycle. Therefore, it would appear that the sea star activator was highly susceptible to denaturation at lower protein concentrations. Consequently, the purification of the sea star MAP kinase activator was performed on a continuous basis over a 24 h period. immediately following its purification.  Characterization of the enzyme was performed  o OO ON un T—i un  i—i  ro un  OH  CN O  O 00 o r-  o  cn  vb  cn cn  o  oo ON oo  ^t-  y -— i »  o  o  ^t- ON OO >—I c n un c n CN c n  O  ^  Cu 00 c3  cd  oo r oo CN cm c n  o o o un oo , o o o r - TJc  un Tt IcN ON  O  H vn  o  I-l  O O  O O  o  un  oo  o  ON ,-)-  CN  o  O  o  CN  OH  52  o o  O O ^ O  H  O N  O O o vo CN es.  > 00 O  •83  5§  O  on  c 2 s  3.3  Identification of the purified MAP kinase activator  To determine the purity of the sea star MAP kinase activator after the final Mono S column, fractions were electrophoresed on SDS-PAGE and the proteins revealed by the silver staining technique (Figure 22A). Both high and low molecular mass contaminating proteins were observed. However, at least three silver-stained bands of 82-90-kDa, and 44-kDa displayed an enrichment that correlated with the Mono S fraction possessing the greatest ability to increase Erkl MBP phosphotransferase activity in the two-step MAP kinase activator assay.  To clarify which band might be the active enzyme, the most purified preparation of MAP kinase activator was tested for its ability to undergo autophosphorylation in the presence of high specific activity y P[ATP]. Although autophosphorylation in vitro is 32  known to occur with many kinases, the purified MAP kinase activator failed to undergo any self-phosphorylation. Therefore, a different approach was necessary to confirm which of the silver-stained polypeptide bands represents the M A P kinase activator. Western blot analysis was performed on the most purified preparation. The Mono S fractions displaying the highest sea star Erkl activator activity were combined before electrophoresis on SDS polyacrylamide gel. After protein transfer, the nitrocellulose was immunoblotted with 3 distinct M A P kinase kinase polyclonal antibodies.  These  antibodies were raised against peptides patterned after the carboxy-terminus and subdomain XI of mouse Mekl and subdomain VIII of budding yeast cerevisiae  Saccharomyces  (Teague et al, 1986; Crews et al, 1992). All three anti-peptide antibodies  detected a 44-kDa band (Figure 22B, lane 1). No other bands were consistently detected with all three antibodies. As a control, partially purified sheep platelet MAP kinase kinase and glutathione-agarose affinity purified mouse recombinant GST-Mekl were also  Figure 22: Analysis of Erkl activator purification (A) The purified Erkl activator fractions from the final Mono S column step were electrophoresed on a 10% SDS-PAGE. The separated proteins were revealed by silver staining (see Material and Methods). The fraction containing the peak Erkl activator activity is denoted by an asterisk. An arrow indicates the position of the putative sea star Erkl activator. (B) The peak Erkl activity fraction and the adjacent fractions were amalgamated for Western analysis. The combined material was separated on a 10% SDS-PAGE and transferred to nitrocellulose. Purified sea star Erkl activator (lane 1); partially purified sheep platelet Mekl (lane 2); and recombinant mouse GST-Mekl (lane 3) were probed with polyclonal antibodies directed against peptides specific for mouse MAPKK-XI and MAPKK-CT as well as yeast Saccharomyces cerevisiae Ste7-VIII (where the Roman numerals refer to the kinase subdomains of these proteins). The molecular masses of prestained standards are indicated: bovine serum albumin, 87-kDa; ovalbumin, 50-kDa; carbonic anhydrase, 34kDa; soybean trypsin inhibitor, 29-kDa; and lysozyme, 21-kDa.  B.  GST-Mekl mouse  Mekl sheep platelets Erkl activator sea star  1  2  3  1  2  3  1  2  a-MAPKK-XI a-MAPKK-CT a-STE7-VIII 184  separated on the same gel (Figure 22B, lanes 2 and 3). Both the mouse and sheep isoforms were recognized by these same antibodies. In summary, it appears that the sea star MAP kinase activator is homologous to mammalian Mek.  This is supported by the  fact that the echinoderm isoform was detectable with three antibodies that were derived from three different kinase subdomain regions in Mek isoforms from two disparate model species.  To verify the subunit composition of the MAP kinase activator, a small sample of the most purified material was subjected to gel filtration analysis on a Superose 12 column (Figure 23). The activator activity chromatographed as a double peak. The smaller peak eluted from the Superose column with an apparent molecular mass of 150 kDa. The larger peak was estimated to have a molecular mass of 40 to 44 kDa which was in the same size range as the enriched 44kDa band identified from the silver-stained gel (compare Figure 22A and B). These data cannot exclude the possibility that the smaller 44 kDa band was generated from the larger 150 kDa species observed from the Superose 12 column. It is feasible that the larger 150 kDa species from the Superose 12 column was a dimer composed of the smaller 44 kDa Mek-like protein and one of the enriched 80 to 90 kDa proteins that copurified. These results do not exclude the possibility that the larger species is a homotrimer of Erkl activators  3.4  Mechanism of human Erkl activation by the sea star MAP kinase activator in vitro  A number of research groups have reported that recombinant Erks expressed in bacteria were phosphorylated on tyrosyl in vivo. This phosphorylation was shown to be important for Erk kinase activity since treatment with the tyrosyl phosphatase CD45 caused its inactivation, while the seryl/threonyl protein phosphatase 2A had no effect on  Figure 23: Gel filtration chromatography of the Erkl activator. A small volume (200 ul) of activator protein purified from the final Mono S step was chromatographed on a Superose 12 gel filtration column in buffer B containing 150 mM NaCl. The fractions were analyzed for the presence of activator by assaying for recombinant Erkl activation. Protein molecular mass standands used for calibrating the column are indicated: Blue Dextran, 443-kDa; alcohol dehydrogenase, 150-kDa; and p-lactoglobulin, 35-kDa.  Superose 12 fraction number  the enzyme (Seger et al, 1991; Crews et al, 1991; Wu et al, 1991; Charest et al, 1993). It was also shown that the Erks were capable of undergoing a slow intramolecular autophosphorylation on the regulatory Tyr-204 (Tyr-185 in Erk2) in vitro and that this was accompanied by a small increase in its MBP phosphotransferase (Rossomondo et al., 1992; Robbins et al, 1993; Charest et al, 1993). These observations lead several researchers to speculate that MAP kinases may be regulated by an allosteric mechanism in which a non-enzymatic factor may be responsible for inducing conformational changes and thus increase its kinase activity (Seger et al, 1991; Wu et al, 1991).  To examine the mechanism of MAP kinase activation in vitro, affinity-purified recombinant Erkl proteins immobilized on glutathione-agarose beads were incubated with the most purified preparation of of sea star MAP kinase activator. It is evident that phosphorylation of Erkl did increase in the presence of the sea star activator when compared to Erkl alone (Figure 24A, compare lanes 2 and 4). Furthermore, the sea star activator was capable of augmenting phosphorylation on tyrosyl- threonyl- and serylresidues on wild type Erkl when assessed with its autophosphorylated counterpart (Figure 24C). It is noteworthy that the increased phosphorylation of wild type Erkl resulted in a 3-fold increase in its phosphotransferase activity (Figure 24B, compare lanes 2 and 4). This is expected if a kinase is to be activated in a robust fashion after sea star oocyte exposure to the hormone 1-MeAde. However, the same sea star activator was unable to phosphorylate the catalytically-compromised Erkl (TEY*) in a similar assay to levels above what was obtained with Erkl (TEY*) alone (Figure 24A lanes 1 and 3). As expected, the kinase-inactive Erkl (TEY*) was unable to phosphorylate the MBP substrate (Figure 24B lanes 1 and 3). These data indicate that the purified sea star MAP kinase activator was capable of activating Erkl by a phosphotransferase-independent mechanism perhaps by inducing a conformation change that opened the active site of the Erkl kinase and promoted its activation by autophosphorylation. It is also possible that  the sea star activator interacts weakly with human Erkl and that a mutation in Lys-72 may lead to small comformational changes that could further destabilize interactions between the two proteins. In fact, comparison between human Erkl and sea star Mpkl MAP kinase isoforms that was cloned recently for our laboratory revealed that the two enzymes possess lower sequence homology in the amino- and carboxy-terminal regions of the enzymes (data not shown).  Consequently, lower amino acid conservation  combined with small conformational changes due to the Lys-72 substitution may be sufficient to prevent phosphorylation of the catalytically compromised Erkl by the sea star MAP kinase activator.  Figure 24: Sea star Erkl activator phosphorylation and activation of GST-Erkl. Recombinant kinase inactive (TEY*) and wild type (TEY) GST-Erkl were combined without (-) or with (+) purified sea star activator and y- P[ATP] (9000 cpm/pmol) and assayed for Erkl phosphorylation (A) or activation of MBP phosphotransferase activity (B) during a 20 min time period. The reaction was quenched with 5X gel loading. After separating the proteins by 10% SDS-PAGE, they were transferred to nitrocellulose and the phosphate incorporation into Erkl and M B P assessed by autoradiography. Quantitation was obtained by excising the phosphorylated bands for liquid scintillation counting. In a separate experiment, the wild type GST-Erkl band in (A) was excised and hydrolyzed with constant-boiling HC1 and the free amino acid assessed by phosphoamino acid analysis (C) as outlined in Materials and Methods. 32  TEY*  TEY  TEY*  TEY  0.01  1.5  0.05  2.4  Sea star Activator  -  -  +  +  Lane  1  2  3  4  TEY*  TEY  1  GST-Erkl pmol/min/ml  TEY*  TEY  pmol/min/ml  0.1  46  0.2  142  Sea star Activator  -  -  +  +  3  4  B MBP  Lane  1  2  Phosphate  Phosphoserine Phosphothreonine  Phosphotyrosine  Origin Sea star Activator 191  4.  MOLECULAR ANALYSIS OF THE REGULATORY PHOSPHORYLATION SITES (TEY) WITHIN THE L12 ACTIVATION LIP OF ERK1  4.1  Phosphorylation and activation of Erkl by activated Mekl (EE)  To conduct Erkl phosphorylation site studies, a constitutively active mouse Mekl mutant was constructed that could be expressed and purified from bacterial lysates as a GST-fusion protein (see Appendix 20).  Site-directed mutagenesis using specific  oligonucleotides and PCR was employed to alter the Rafl phosphorylation sites on Mekl (Zheng and Guan, 1994). Both Ser-218 and Ser-222 amino acids were exchanged with glutamic acid mimetics of phosphoresidues.  The recombinant double glutamic acid  allele, Glu-218/Glu-222-GST-Mekl (termed Mekl (EE)), activated Erkl M B P phosphotransferase activity 900-fold above what the dephosphorylated wild type Mekl could activate this same enzyme under identical experimental conditions (Figure 42, Appendix 20).  Erk2 has been shown previously to be activated by phosphorylation on neighbouring Thr-183 and Tyr-185 residues in vivo (Thr-202 and Tyr-204 in Erkl) (Payne et al, 1991). Subsequently, phosphorylation on these same sites in MAP kinase isoforms Erkl and Erk2 in vitro was demonstrated by addition of partially purified preparations of Mek from mammalian tissue or with marginally activated forms of recombinant Mek from E. coli (Haystead et al, 1992; Robbins et al, 1993; Charest et al, 1993; Butch and Guan, 1996). However, the results described in these publications were not conclusive since phosphoamino acid analysis revealed that Mekl phosphorylated Erk protein principally on tyrosyl residues and to a lesser degree on threonyl residues. This contrasted in vivo labelling studies that showed that activation of Erk MBP  192  phosphotransferase activity required phosphorylation on neighbouring threonyl and tyrosyl residues (Payne et al., 1991).  To confirm these previous reports , a time course of Erkl activation by Mekl (EE) was compared with the appearance of phosphate incorporation into the three hydroxyl containing amino acids, serine, threonine and tyrosine of this same enzyme. The experiment was initiated by addition of [y -P] ATP (10,000 cpm/pmol) to an 32  Eppendorf tube containing the Erkl and Mekl (EE) enzymes as well as the required cofactors. Thirty seconds prior to the completion of each time point, an aliquot was removed from the reaction mixture and added to a second tube containing the Erkl exogenous substrate MBP. After incubation for one half minute, the reaction was terminated by addition of gel loading buffer and the proteins separated by SDS-PAGE. The amount of radioactive P incorporated into Erkl attained a maximum of ~ 0.08 mol 3 2  phosphate/ mol enzyme after the first thirty minutes of incubation with activated Mekl (EE) (Figure 25A). As expected, Erkl MBP phosphotransferase activity increased in parallel with Mekl phosphorylation albeit at a slightly delayed rate (Figure 25B). The specific enzyme activity of Mekl (EE) activated Erkl was approximately 15 Hmol/min/mg. A steady state level of Erkl phosphotransferase activity was reached within forty minutes post ATP addition. Mekl (EE) phosphorylated Erkl below stoichiometric levels (2 mol phosphate/mol enzyme). This resulted in a poor activation of Erkl MBP phosphotransferase activity.  The Erkl bands in Figure 25A were subjected to phosphoamino acid analysis. Separation of the hydrolyzed phosphoamino acids by thin layer chromatography revealed that incorporation of phosphate by Mekl (EE) occurred in an ordered manner (Figure 25B). As will be discussed later, similar results were obtained with a more active Mekl  Figure 25: Time course of Mekl (EE) phosphorylation, activation and phosphoamino acid analysis of wild type Erkl. Recombinant thrombin-cleaved Erkl protein (~1 u,g) was incubated with eluted constitutively active GST-Mekl (EE) (-53 ng) in the presence of [y- P] ATP (10,000 cpm/pmol) for the indicated times. At thirty seconds prior to 32  completion of the reaction, a 5 ul aliquot of MBP (5 mg/ml) was added to the mixture. The reaction was incubated for a further 30 sec before termination of the reaction by addition of 5 ul of 4X gel loading buffer. The proteins were electrophoresed on a 10% Bio-Rad mini gel, transferred to PVDF and visualized by autoradiography. The radioactive bands were excised from the PVDF membrane. (A) The phosphorylated MBP bands were counted by liquid scintillation (dotted line, closed circles) while the phosphorylated Erkl bands were counted by Cerenkov (solid line, open circles). (B) The recombinant Erkl proteins were subjected to phosphoamino acid analysis. The hydrolyzed amino acids were separated by thin layer chromatography followed by autoradiography. The migration of the phosphoamino acid standards are indicated to the right of the panel  195  ireonine  /rosine  o  CL  CL  c sz  OO  CO  jrine o O J3  OH  O  -C  OH  H in in o in  m  m  CL  •5b  c  §  JC  CL  •  • %  o  -1-  •  • •  •  If  m •  •  m  •  O ro  «n O ri  •• * w  •  •  *  *  *  m w  •  •  196  (AN3EE).  In this experiment, the appearance of phosphotyrosine preceded that of  phosphothreonine which, in turn, preceded phosphoserine (Figure 25B). However, all three phosphoresidues were present at their highest levels when Erkl reached maximal activity. In fact, phosphoserine only became visible when Erkl reached maximal MBP kinase activity. There have been no reports of Mekl phosphorylating Erkl on residued other than Thr-202 and Tyr-204 despite the fact that seryl autophosphorylation has been detected in MAP kinases phosphorylated in vitro (Robbins et al, 1993; Charest et al, 1993). In addition, Mekl (EE) phosphorylation of Erkl predominantly on tyrosine and to a lesser extent on threonine may explain the low specific enzyme activity achieved by Erkl in these assays.  4.2  Phosphorylation and activation of Erkl by Mekl (AN3EE)  A second approach was used to create a more constitutively active Mekl for stimulating Erkl M B P phosphotransferase activity and mapping the sites of phosphorylation. To obtain a more active Mekl protein kinase Mansour et al. (1994) identified an oc-helix structural motif (A-helix) in the amino-terminal region located outside the kinase catalytic domain that is implicated in the regulation of Mekl protein kinase activity. Deletion of the hydrophilic region between residues 32 to 51 resulted in a mutant that was several fold more active than the basal activity of the recombinant dephosphorylated form of Mekl. Furthermore, the combination of deletion mutant and phosphorylation site substitutions with Glu-218 and Glu-222 caused an even greater Mekl phosphotransferase activity toward Erkl in vitro (Mansour et al, 1994). This Mekl hyperactive mutant induced a very dramatic increase in Erkl MBP kinase activity when compared to the wild type protein.  To examine the pattern of phosphate incorporation into Erkl by hyperactive Mek 1 (AN3EE), a deletion mutant using the mouse Mekl (EE) cDNA clone was prepared as reported by Mansour et al. (1994). Identical reaction conditions and Mekl protein concentration described in the previous Mekl (EE) time course were used for the Erkl phosphorylation and activation experiments with Mekl (AN3EE) hyperactive mutant (see Figure 25A). Mekl (AN3EE) phosphorylation of recombinant wild type Erkl attained a maximum P incorporation of -0.5 mol phosphate/ mol of enzyme within ten minutes of 3 2  initiating the reaction (Figure 26A). The amount of phosphate incorporated into Erkl was 6-fold greater for the AN3EE allele than what was observed with the EE allele. This steady state level was maintained for most of the experiment, except for a minor decline near the completion of the time course. Again the Erkl enzyme was not phosphorylated to stoichiometric levels in this experiment. Mekl (AN3EE) phosphorylated Erkl reached a maximum enzyme activity of -50 umol/min/mg within 10 min of initiating the reaction. In fact, Mekl (AN3EE) activated Erkl three-fold higher in one third less reaction time it took Mekl (EE) to phosphorylate Erkl to its maximal level (compare Figure 25A with Figure 26A). However, in the later time points Erkl MBP phosphotransferase activity declined to near basal levels.  To determine the amino acids phosphorylated by Mekl (AN3EE), the Erkl protein was hydrolyzed with acid and subsequently separated in one-dimension on a thin layer chromatography plate. After one minute, Erkl was predominantly phosphorylated on tyrosyl (Figure 26B). As the reaction progressed phosphothreonyl became more apparent. Finally, all three phosphoamino acids became manifest when Erkl reached full enzymatic activity at ten minutes. In fact, the appearance of phosphoserine with Erkl peak activation was reminiscent of what occurred with Mekl (EE) phosphorylation and activation of Erkl in that the appearance of phosphoserine occurred when Erkl reached maximal activity (Figure 26B). However, seryl phosphorylation in the presence of Mekl  Figure 26:  Time course of Mekl (AN3EE) phosphorylation, activation and  phosphoamino acid analysis of wild type Erkl. Recombinant thrombin-cleaved Erkl protein (~1 |ig) was incubated with eluted constitutively active GST-Mekl (AN3EE) (-53 ng) in the presence of [y- P] ATP (10,000 cpm/pmol) for the indicated times. At 32  thirty seconds prior to completion of the reaction, a 5 ul aliquot of MBP (5 mg/ml) was added to the mixture. The reaction was incubated for a further 30 sec before termination of the reaction by addition of 5 ul of 4X gel loading buffer.  The proteins were  electrophoresed on a 10% Bio-Rad mini gel, transferred to PVDF and visualized by autoradiography. The radioactive bands were excised from the PVDF membrane. (A) The phosphorylated MBP bands were counted by liquid scintillation (dotted line, closed circles) while the phosphorylated Erkl bands were counted by Cerenkov (solid line, open circles). (B) The recombinant Erkl proteins were subjected to phosphoamino acid analysis. The hydrolyzed amino acids were separated by thin layer chromatography followed by autoradiography. The migration of the free-phosphate and phosphoamino acid standards are indicated to the right of the panel.  A.  200  in  m o  in o -t  m m 9  cn in cN o CN in  CN  ca  201  (AN3EE) was more robust. These results revealed that Erkl became phosphorylated on all three hydroxylamino acids in a Mekl-dependent manner. Although Mekl has been demonstrated to phosphorylate catalytically-inactivated Erkl on threonyl and tyrosyl residues both in vivo and in vitro, these data are unable to negate the possibility that Mekl could phosphorylate sites distinct from the TEY regulatory motif. Alternatively, full activation of Erkl by Mekl (AN3EE) may promote a more pronounced Erkl autophosphorylation on serine.  4.3  Mutational analysis of Erkl regulatory phosphorylation sites  With the exception of Erk3 isoforms, all MAP kinases examined to date possess the canonical T X Y phosphorylation sequence (Figure 27A). Erk3 appears to be regulated by phosphorylation on a seryl residue located at the same position as Thr-183 in Erk2 (Thr-202 in Erkl), while the homologous tyrosyl residue is substituted with a glycyl residue (Figure 27A). This implies that Erk3 may be regulated by a mechanism distinct from the other M A P kinase family members (Pelech and Charest, 1995). X-ray crystallographic studies revealed that Thr-183 and Tyr-185 phosphorylation residues in Erk2 protein are contained within a loop structure (L12) known as the activation loop that lies within the cleft of the active site (Figure 27B and see Zhang et al, 1994). In the inactive form of Erk2, Tyr-185 (Tyr-204 in Erkl) is buried within the active site of the kinase, whereas Thr-183 (Thr-202 in Erkl) is exposed on the outer surface of the protein. It is predicted that phosphorylation of both these regulatory sites would cause a dramatic reorganization of enzyme structure and assist in stabilization of MAP kinase in the active conformation. Therefore, it is expected that the amino acid residues that comprise the TXY phosphorylation motif play a central role in recognition by the cognate upstream MAP kinase kinase activator.  202  Figure 27: Sequence alignments of MAP kinase isoforms. A. Sequence comparisons of the activation loop for several MAP kinase isoforms human (Charest et al., 1993); Erk2 (Gonzalez et al., 1992); Erk3 (Gonzalez et al., 1992); Jnkl (Derijard et al., 1994); Hogl (Lee et al., 1994) and sea star Mpkl (Posada et al., 1991). The two regulatory phosphorylation sites (TXY) are each designated with an asterisk and the intervening amino acid X is outlined with a box. B. Schematic representation for human Erkl mutation sites (Lys-71, Thr-202, Glu-203 and Tyr-204). The site of ATP-binding, within the kinase catalytic domain is located in subdomains I and II; a region which is conserved in all kinases studied to date (Hanks et al., 1988). The regulatory phophorylation sites for many kinase are located at within the enzyme activation loop between subdomains VII and VIII (Marshall, 1993).  ON  X X HH  I-H I-H I-H  N cn  J  o o o  >  CN CN f-1 I-H I-H >  > H  hH  J J J 0 0 h  0  Q P P a  fl  u id  w  5 "5 H CN ro H H H X u w  fl rji A! o a  SO  C  b  c I  h PQ  204  Mekl displays a very narrow substrate specificity for Erk enzymes, since it is unable to phosphorylate non-native forms of the Erkl or short peptides containing the TEY sequence (Seger et al, 1992a). Furthermore, Mekl also phosphorylates other exogenous substrates in vitro with reduced efficacy relative to Erkl phosphorylation. To elucidate Mekl substrate specificity, seven site-directed mutations were generated in Thr202 and Tyr-204 regulatory phosphorylation sites of Erkl (Table 9). Mekl activation sites Thr-202 and Tyr-204 were exchanged with conserved phosphorylatable amino acids serine and threonine, respectively, to determine the effect these mutations have on Mekl recognition. Also, a double mutant, in which the TEY phosphorylation site was inverted, was constructed to assess the effect orientation has on Mekl phosphorylation. The nonpolar glycyl amino acid that is present in Erk3 in place of tyrosyl found in other members of the MAP kinase family was used to substitute Thr-202 and Tyr-204 in Erkl, while the phosphorylation mimetic, glutamic acid, replaced these same regulatory sites.  4.4  Analysis of autophosphorylation and basal MBP phosphotransferase activities of Erkl regulatory phosphorylation site alleles  The different Erkl alleles were expressed under identical growth conditions. Each allele was expressed as a GST-fusion protein and purified by adsorption onto glutathione-agarose resin. The Erkl enzyme was released from the GST-beads by thrombin cleavage, resolved by SDS-PAGE and subsequently visualized by staining with Coomassie Blue dye or transferred to nitrocellulose and immunoblotted with Erkl-CT antipeptide antibody (Figure 28A and B). Erkl migrated as a 42- and 44-kDa doublet (Figure 28A). The enzyme concentration was adjusted to 1 ug of protein for all kinase reactions (Figure 28A). MAP kinases Erkl and Erk2 are known to autophosphorylate slowly on tyrosyl when expressed as recombinant proteins in bacteria (Seger et al, 1991; Crews et al, 1991; Wu et al, 1991, Rossomondo et al, 1992; Robbins et al, 1993;  205  Figure 28: Expression, immunodetection and quantitation of recombinant Erkl proteins from bacteria. The regulatory phosphorylation site mutant allele constructs were transformed into E. coli strain UT 5600 and subsequently grown in 2 x YT medium. The collected cells were disrupted in homogenization buffer P and the recombinant proteins purified by glutathione affinity chromatography. Thrombin digestion liberated the Erkl protein from its GST fusion. Approximately 1 ug of protein was subjected to 10% SDSPAGE Bio-Rad mini-gels. The proteins were quantitated by Coomassie Blue staining (A) or transferred to nitrocellulose for Western blotting (separate gels) with Erkl-CT polyclonal antibody (B) or phosphotyrosine specific 4G10 monoclonal antibody (C).  o 3 o3 U oq  PM I  H H H  rn  as  CM  1  w  rt  rt  4  u  -t  Charest et al, 1993).  As expected, Erkl wild type T E Y displayed tyrosyl  phosphorylation after Western blotting with the antiphosphotyrosine 4G10 monoclonal antibody (Figure 4C). The catalytically-inactive Erkl TEY*, however, displayed no immunoreactivity with 4G10 antibody since mutation of Lys-71, an amino acid critical for phosphotransferase activity, compromised the capacity of the enzyme to autophosphorylate on tyrosyl residue (Figure 28C). Substitution of Thr-202 with seryl, glutamyl or glycyl residues seemed to have little effect on basal autophosphorylation, while similar mutations at Tyr-204 abolished the phosphotyrosyl signal (Figure 28C). These results indicate that Tyr-204 is the major site of tyrosyl autophosphorylation when Erkl is expressed in bacteria. The dual phosphorylation site was reversed so that Thr202 was replaced by tyrosine and Tyr-204 by threonine. Surprisingly, Erkl YET displayed comparable immunoreactivity with the 4G10 antibody as seen with wild type TEY autophosphorylation. These prelimary data imply that there may be some flexibility with regards to the orientation and amino acid composition at these two phosphorylatable sites. To verify whether the narrow substrate specificity of Mekl is determined by the class of amino acid present at positions 202 and 204, both constitutively active Mekl mutants were used to phosphorylate and activate the Erkl regulatory phosphorylation site mutants.  First it was important to test the ability of each Erkl allele to further autophosphorylate in vitro was assessed in a ten and thirty minute assay reactions. These time points were selected based on the ability of Mekl (EE) and (AN3EE) to maximally activate wild type Erkl M B P phosphotransferase  activity.  The pattern of  autophosphorylation supports the observations made with the antiphosphotyrosine Western blot analysis in that those enzymes with an intact Tyr-204 present were able to undergo autophosphorylation (compare Figure 28C, 29A and 31 A). The three Erkl Thr202 mutant alleles, SEY, GEY, EEY and the inverted YET allele autophosphorylated to  Figure 29: Mekl (EE) phosphorylation of Erkl regulatory phosphorylation site mutant alleles.  Equal amounts of Erkl recombinant proteins (-lug) were  incubated without (A) or with (B) constitutively active recombinant Mekl (EE) (-53 ng) for 30 min in the presence of [y- P] ATP (1250 cpm/pmol). The 32  reaction was terminated by addition of 4X SDS-PAGE sample buffer. The proteins were separated on a 10% Bio-Rad mini gel, transfered to nitrocellulose and visualized by autoradiography. The radioactive bands were cut from Ponceau stained nitrocellulose and counted by liquid scintillation. (C), Erkl phosphorylation without (light grey) or with Mekl (EE) (dark grey) was normalized to Erkl wild type autophosphorylation activity which was given a value of 100; mean ± s.d., n=3 (Erkl autophosphorylating activity varied between 5200-8300 cpm/30 min reaction). The data were analyzed by the ANOVA (analysis of variance) test. Ranked sample means: TEY* T E E T E G  TEG  T E T  TEE  G E Y  TET  SEY Y E T E E Y  G E Y YET T  E Y  TEY* E E Y SEY TEY  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutants incubated in the absence (bold face) or presence of Mekl (EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  209  1  2  3  4  5  6  7  8  9  TEY TEY* SEY GEY EEY YET TET TEG TEE  TEY  TEY* SEY GEY EEY YET TET TEG TEE Erkl mutant alleles  40-75% of wild type T E Y (Figure 29A and 31 A).  In contrast, the Tyr-204  phosphorylation site mutants TET, TEG, and TEE mustered on average less than one quarter of the self-phosphorylation of the Erkl wild type (Figure 29A and 31 A). No appreciable kinase activity was observed for any of the recombinant dephosphorylated forms of Erkl in a 30 sec MBP phosphotransferase reaction (Figure 30 and 32). These data indicate alteration of either regulatory phosphorylation site causes a reduction of erkl autophosphorylation, however, a more pronounced effect was observed at Thr-202 than with Tyr-204 mutants alleles. Overall, neither Erkl wild type nor the mutant alleles displayed a strong M B P phosphotransferase  activity in the absence of Mekl  phosphorylation.  To understand what effect the Erkl phosphorylation site mutations may have on Mekl substrate recognition and phosphorylation, the various Erkl alleles were combined with constitutively active isoforms of Mekl (AN3EE) and (EE) for ten and thirty minute kinase reactions, respectively. Previously, Erkl was demonstrated to be phosphorylated and activated maximally at these two time points (Figure 26 and 27).  4.5  Mekl (EE) phosphorylation and activation of Erkl regulatory phosphorylation site alleles  In a thirty minute assay, Mekl (EE) phosphorylated Erk TEY, TEY*, SEY, and EEY alleles 2.5- to 3.0-fold above the level of autophosphorylation of these same enzymes (Figure 29). Moreover, the analysis of variance test showed that the amount of radioactive  3 2  P label incorporated into TEY*, SEY and EEY was not significantly  different from that of Erkl wild type TEY; however only the wild type form of the Erkl was effectively activated by Mekl (EE). This constitutively active Mekl (EE) caused a 25-fold stimulation of Erkl wild type MBP phosphotransferase activity above that of  Figure 30: Mekl (EE) activation of Erkl regulatory phosphorylation site mutant alleles. The Erkl activation was performed as a two-step assay in the presence of [y-32p]  ATP (1250 cmp/pmol). Erkl protein (~1 ug) was incubated without (light  grey) and with (dark grey) constitutively active Mekl (EE) (-53 ng) before performing the MBP kinase assay. At the completion of the 29 min preincubation described in Figure 25A, a 5 ul aliquot of the exogenous substrate MBP (5 mg/ml) was added to the mixture. The reaction was incubated for a further 1 min before addition of 5 ul of 4X gel loading buffer. The Mekl activation activity is normalized to Erkl wild type basal autoactivation activity which is given a value of 100; mean ± s.d., n=3 (Erkl wild type MBP phosphotransferase activity varied between 750-1550 cpm/1 min reaction). The data were analyzed by the ANOVA (analysis of variance) test. Ranked sample means: T E E T E Y * YET YET T E T G E Y T E G G E Y E E Y TEE TET E E Y T E Y TEY* S E Y Y E T SEY T E Y  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutant alleles incubated in the absence (bold face) and presence of Mekl (EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the kinase-inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  -Mekl (EE) I +Mekl (EE)  TEY TEY* SEY GEY EEY YET Erkl mutant alleles  TET  TEG  TEE  Erkl alone (Figure 30). In contrast, the Erkl SEY allele, a mutant that possesses a conserved threonyl to seryl substitution, achieved only 20% of wild type activity even though both alleles were phosphorylated to identical levels with Mekl (EE) (Figure 29). Analysis of variance revealed that the activation of SEY was significantly different from the remainder of the mutant alleles.  4.6  Mekl (AN3EE) phosphorylation and activation of Erkl regulatory phosphorylation site alleles  Under the identical reaction conditions described above for the double mutant (EE), the Mekl (AN3EE) was demonstrated to stimulate maximally Erkl wild type enzymatic activity within 10 min (Figure 26). Erkl wild type TEY and kinase-inactive TEY* were robustly phosphorylated 12- to 15-fold above Erkl autophosphorylation, respectively (Figure 31C). In addition, a pronounced band shift was detected for phosphorylated Erkl which has been previously reported to be indicative of the stimulation of enzyme activity. Almost no change in the electrophoretic mobility was observed with the catalytically-inactive Erkl (Figure 3 IB compare lanes 1 and 2). The threonyl phosphorylation site mutants SEY, GEY and EEY in the presence of activated Mekl (AN3EE) were phosphorylated 2- to 4-fold less than the non-phosphorylation site mutants TEY and TEY* (Figure 31). Comparable mutations at the tyrosyl site limited the phosphotransferase activity of Mekl (AN3EE) toward Erkl at 1- to 3-fold above autophosphorylated wild type Erkl. However, activation of the MBP phosphotransferase activity of these regulatory site mutant alleles was incongruent with amount each of these enzymes was phosphorylated by Mekl (AN3EE) during the same reaction. Analysis of variance revealed that only wild type Erkl TEY and the conserved SEY mutant displayed significant activation by Mekl (AN3EE) (Figure 32). Erkl TEY was activated 230-fold above the endogenous kinase activity and twice the level observed for Erkl SEY allele.  214  Figure 31: Mekl (AN3EE) phosphorylation of Erkl regulatory phosphorylation site mutant alleles. Equal amounts of Erkl recombinant proteins (~1 ug) were incubated without (A) or with (B) constitutively active recombinant Mekl (AN3EE) (-53 ng) for 10 min in the presence of  [Y- P] 3 2  ATP (1250 cpm/pmol).  The reaction was terminated by addition of 4X SDS-PAGE sample buffer. The proteins were separated on a 10% Bio-Rad mini gel, transfered to nitrocellulose and visualized by autoradiography. The radioactive bands were cut from Ponceau stained nitrocellulose and counted by liquid scintillation. (C), Erkl phosphorylation without (light grey) or with Mekl (AN3EE) (dark grey) was normalized to Erkl wild type autophosphorylation activity which was given a value of 100; mean ± s.d., n=3 (Erkl autophosphorylating activity varied between 5200-8300 cpm/10 min reaction). The data were analyzed by the ANOVA (analysis of variance) test. Ranked sample means: T E Y * T E E G E Y T E G T E T S E Y Y E T Y E T T E G E E Y T E Y TEE G E Y TET E E Y SEY T E Y T E Y *  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutants incubated in the absence (bold face) or presence of Mekl (AN3EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  1  TEY  2  3  TEY*  4  5  6  SEY GEY EEY YET TET  7  8  9  TEG TEE  ^WliiMlP 1  TEY TEY* SEY GEY EEY YET TET TEG TEE Erkl mutant alleles  No significant enzyme activation was observed for the remainder of the Erkl alleles. Furthermore, comparisons between Mekl (EE) and (AN3EE) phosphorylation and activation of Erkl proteins revealed that the triple mutant allele (AN3EE) was able to stimulate both Erkl TEY and SEY 10 to 20 times above the levels observed with the double mutant (EE) during a shorter incubation period (compare Figures 30 and 32).  Hay stead et al. (1992) demonstrated that Erk protein kinases undergo phosphorylation in an ordered fashion. Initially, Mekl phosphorylates Erkl on tyrosyl followed by threonyl residues. Phosphoamino acid analysis of Erkl phosphorylated by Mekl (EE) revealed that the phosphotyrosyl content was 3- to 4-fold greater than phosphothreonine (Figure 25B). In comparison, the more active Mekl (AN3EE) phosphorylated wild type Erkl equally on tyrosine and threonine (Figure 26B). The difference in the level of threonine phosphorylation by the two activated Mekl alleles resulted in a ten-fold difference in the activation of Erkl by each. Therefore, the phosphothreonine content of Erkl correlates with the level of MBP phosphotransferase activity of this enzyme. With the exeption of the Erkl SEY allele and to a lesser degree the EEY allele, alteration of Thr-202 or Tyr-204 has a significant impact on the ability of Mekl to phosphorylate Erkl. More importantly, only Erkl SEY diplays any significant MBP phosphorylating activity. The more robust Mekl (AN3EE) which has the deleted amino-terminal a-helix appears to have less constraint in recognizing tyrosine when serine is present in position 202 of Erkl. The enzyme phosphorylated SEY 3-fold higher than Mekl (EE) (compare Figure 29C and 31C) which resulted in 30-fold difference in Erkl SEY kinase activity (compare figure 30 and 31). Further studies will be required to determine the degree to which Mekl (AN3EE) is able to phosphorylate Erkl SEY.  Figure 32: Mekl (AN3EE) activation of Erkl regulatory phosphorylation site mutant alleles. The Erkl activation was performed as a two-step assay in the presence of [y- P] ATP (1250 cmp/pmol). The Erkl proteins (~1 ug) were 32  incubated without (light grey) and with (dark grey) constitutively active Mekl (AN3EE) (-53 ng) before performing the MBP kinase assay. At the completion of the 29 min preincubation described in Figure 26A, a 5 ul aliquot of the exogenous substrate MBP (5 mg/ml) was added to the mixture. The reaction was incubated for a further 1 min before addition of 5 ul of 4X gel loading buffer.  The Mekl  activating activity is normalized to Erkl wild type basal autoactivation activity which is given a value of 100; mean ± s.d., n=3 (Erkl wild type MBP phosphotransferase activity varied between 750-1550 cpm/1 min reaction). The data were analyzed by the ANOVA (analysis of variance) test. Ranked sample means: TEE TEG TET YET TEY* GEY YET TEY EEY TEG G E Y TEE SEY E E Y TEY* TET SEY TEY  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutant alleles incubated in the absence (bold face) and presence of Mekl (AN3EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the kinase-inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  218  IOOOO H  TEY  TEY* SEY GEY  EEY YET  Erkl mutant alleles  TET  TEG  TEE  4.7  Mutational analysis of the intervening glutamic acid residue in TEY of Erkl  The M A P kinase family of protein kinases are defined by the dual phosphorylation on the threonyl and tyrosyl site motif (Ferrel, 1996). However, MAP kinases have been further classified into subfamilies; based on the length of the activation loop that extends between kinase catalytic subdomains VII and VIII and the intervening amino acid that separates the regulatory threonyl and tyrosyl residues (Figure 27A). The Erk subfamily possesses the longest activation loop sequence followed by the Jnk and Hog subfamilies, which individually are four and six amino acids shorter (Figure 27A). Furthermore, the threonyl and tyrosyl phosphorylation sites in the Erkl, Erk2, and Erk5 subfamilies are separated by the charged amino acid glutamic acid; in contrast, prolyl and glycyl residues divide these two same phosphorylatable residues in Jnkl and Hogl, respectively. The combination of the activation loop length and the type of intervening amino acid is expected to have a dramatic effect on the regulation of these kinases by their specific upstream activators. To gain insight into the role the intervening amino acid has on the recognition and phosphorylation of MAP kinase by Mekl, the glutamyl residue in Erkl was exchanged with the prolyl or glycyl residues normally present in the Jnk and Hog MAP kinase family members.  4.8  Analysis of autophosphorylation and basal MBP phosphotransferase activities of Erkl amino acid 203 mutant alleles  The MAP kinase alleles were affinity purifed by binding to glutathione agarose beads and subsequently separated from the GST-fusion by thrombin cleavage. A small sample (~1 jxg) was electrophoresed on an acrylamide gel and visualized by Coomassie Blue dye or transferred to nitrocellulose and probed with an antibody specific for the carboxy-terminal region of Erkl (Figure 33A). As expected, the faster migrating 42-kDa  220  Figure 33: Expression, immunodetection and quantitation of recombinant Erkl protein from bacteria. The constructs were transformed in E. coli strain UT 5600 and subsequently grown in 2 x Y T medium. The collected cells were disrupted in homogenization buffer G. and the recombinant proteins purified by glutathione affinity chromatography. Thrombin digestion liberated the Erkl protein from the GST-fusion. Approximately 1 ug of protein was subjected to 10% SDS-PAGE Bio-Rad mini-gels. The proteins were quantitated by Coomassie Blue staining (A), or transferred to nitocellulose and Western blotted (seperate gels) with Erkl-CT polyclonal antibody. (B), or phosphotyrosine-specific 4G10 monoclonal antibody (C).  221  A. 44-kDa  Coomassie Blue  44-kDa -I  a-Erkl-CT  44-kDa -I  CC-PY-4G10  B.  c.  TEY  TEY*  TPY  TGY  222  band was present in higher concentrations than the slower 44-kDa protein band. Wild type Erkl TEY displayed some autophosphorylation on tyrosyl following purification of the protein from E. coli (Figure 33C). The TGY mutant protein also cross-reacted with the anti-phosphotyrosine monoclonal antibody 4G10 in the same Western blots. However, no tyrosyl signal was observed with Erkl TPY or the catalytically-inactivated TEY* alleles (Figure 33C).  To evaluate how changes to the glutamyl residue may effect catalyic activity, Erkl was assayed for basal autophosphorylation and MBP phosphotransferase activities. After a ten and thirty minute incubation period, only Erkl T E Y and TGY showed detectable levels of autophosphorylation activity (Figure 34A and 36A). The TPY allele, however, retained approximately 25% of wild type activity while T G Y autophosphorylating activity remained the same as wild type after 30 minutes (Figure 34C and 36C). Since no phosphotyrosyl signal was detected with 4G10, Erkl TPY may undergo seryl/threonyl autophosphorylation.  Neither the prolyl nor the glycyl  substitutions were able to relieve any conformational constraints imposed by Erkl phosphorylation lip to activate the mutant alleles above wild type TEY basal activity in an autophosphorylation reaction (Figure 34 and 36).  4.9  Mekl (EE) phosphorylation and activation of Erkl amino acid 203 mutant alleles  The Erkl TEY and TEY* recombinant proteins were phosphorylated by Mekl (EE) to an average of 4.0- to 4.5-fold, above the threshold autophosphorylation levels (Figure 36A and C). In contrast, the prolyl and glycyl substitutions were phosphorylated only 50% above the level observed for autophosphorylated wild type Erkl assayed under the same reaction conditions (Figure 36A and C). Analysis of variance demonstrated that Mekl (AN3EE) phosphorylation of Erkl TPY and TGY alleles were not significantly  Figure 34: Mekl (EE) phosphorylation of Erkl intervening amino acid 203 mutant alleles.  Equal amounts of Erkl recombinant proteins (-lug) were  incubated without (A) or with (B) constitutively active recombinant Mekl (EE) (-53 ng) for 30 min in the presence of [y- P] ATP (1250 cpm/pmol). The 32  reaction was terminated by addition of 4X SDS-PAGE sample buffer. The proteins were separated on a 10% Bio-Rad mini gel, transfered to nitrocellulose and visualized by autoradiography. The radioactive bands were cut from Ponceau stained nitrocellulose and counted by liquid scintillation. (C), Erkl phosphorylation without (light grey) or with Mekl (EE) (dark grey) was normalized to Erkl wild type autophosphorylation activity which was given a value of 100 mean ± s.d., n=3 (Erkl autophosphorylating activity varied between 350-1050 cpm/30 min reaction). The data were analyzed by the ANOVA (analysis of variance) test. Ranked sample means: TEY* TPY TGY TEY TPY T G Y TEY TEY*  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutants incubated in the absence (bold face) or presence of Mekl (EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  1  2  3  4  TEY*  TPY  TGY  - ,„, ,„,.,. Si  lsm  TEY  TEY  TEY*  TPY  Erkl mutant alleles  TGY  Figure 35: Mekl (EE) activation of Erkl intervening amino acid 203 mutant alleles. The Erkl activation was performed as a two-step assay in the presence of [y-32p] ATP (1250 cmp/pmol). The Erkl proteins (~1 ug) were incubated without (light grey) and with (dark grey) constitutively active Mekl (EE) (-53 ng) before performing the MBP kinase assay. At the completion of the 29 min preincubation described in Figure 25A, a 5 ul aliquot of the exogenous substrate MBP (5 mg/ml) was added to the mixture. The reaction was incubated for a further 1 min before addition of 5 ul of 4X gel loading buffer. The Mekl activating activity is normalized to Erkl wild type basal autoactivation activity which is given a value of 100 mean ± s.d., n=3 (Erkl wild type MBP phosphotransferase activity varied between 100-200 cpm/30 min reaction). The data were analyzed by the ANOVA ("analysis of variance) test. Ranked sample means: TEY* TPY TGY TEY TGY TPY TEY* T E Y  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutant alleles incubated in the absence (bold face) and presence of Mekl (EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the kinase-inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  226  TEY  TEY  TPY  Erkl mutant alleles  TGY  different than autophosphorylated wild type Erkl. Mutations at the Glu-203 caused an adverse affect on MBP kinase activity, since Mekl (EE) was able to only activate Erkl TEY (Figure 35). Therefore, Mekl phosphorylation of Erkl is acutely sensitive to the specific R-group present on the amino acid side chain at position 203.  4.10  Mekl (AN3EE) phosphorylation and activation of Erkl amino acid 203 mutant alleles  The more active Mekl (AN3EE) displayed the same specificity toward Erkl TPY and TGY mutant alleles as the less active Mekl (EE) (Figure 34C and 36C). Although Mekl (AN3EE) phosphorylated TPY and TGY proteins to one third the level of what was observed for the wild type enzyme and 10-fold higher than autophosphorylated wild type Erkl, no appreciable stimulation of the kinase activity was observed for these enzymes above the level observed for the kinase-inactive Erkl TEY* (Figure 34C and 35). In the presence of Mekl (AN3EE), wild type Erkl TEY was activated 370-fold above Erkl alone and 4-fold more than with Mekl (EE) (compare Figure 35 and Figure 37). As noted above with Mekl (EE), even the more active Mekl (AN3EE) recognition of Erkl mutant alleles is effected by substitutions in Glu-203.  To date, Erkl in the only known substrate for Mekl (Seger et al., 1992a). Part of Mekl substrate specificity for Erkl lies in the three amino acid TEY motif. Substitution of Thr-202 and Tyr-204 in Erkl with either hydroxyl or carboxyl group amino acids dramatically changed phosphorylation by Mekl. Even serine, which contains a hydorgen atom in place of a methyl group in its amino acid side chain, effects the ability of Mekl to recognize Erkl SEY. Another determinant in Mekl recognition of Erkl is the glutamic acid positioned between the regulatory threonyl and tyrosyl. Substitution of Glu-203 with glycyl or prolyl residues decreases the ability of Mekl to properly  Figure 36: Mekl (AN3EE) phosphorylation of Erkl intervening amino acid 203 mutant alleles.  Equal amounts of Erkl recombinant proteins (-lug) were  incubated without (A) or with (B) constitutively active recombinant Mekl (AN3EE) (-53 ng) for 10 min in the presence of [y- P] ATP (1250 cpm/pmol). 32  The reaction was terminated by addition of 4X SDS-PAGE sample buffer. The proteins were separated on a 10% Bio-Rad mini gel, transfered to nitrocellulose and visualized by autoradiography. The radioactive bands were cut from Ponceau stained nitrocellulose and counted by liquid scintillation. (C), Erkl phosphorylation without (light grey) or with Mekl (AN3EE) (dark grey) was normalized to Erkl wild type autophosphorylation activity which was given a value of 100 mean ± s.d., n=3 (Erkl autophosphorylating activity varied between 90-275 cpm/10 min reaction). The data were analyzed by the ANOVA (analysis of variance) test. Ranked sample means: TEY* TPY TEY TGY TGY TPY TEY TEY*  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutants incubated in the absence (bold face) or presence of Mekl (AN3EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  c o  —  c o  X ft ca  o  ft c  B.  3 <  TEY  TEY^  TPY  TGY  W co  C.  c -  7000  4  6000  L  5000  U  4000  U  o  -H 4-1  •~  o Xi y. O  X CU —  M  •—  > *->  3000 2000  CD  1000  L  TEY  TEY*  TPY  TGY  Erkl mutant alleles  230  Figure 37: Mekl (AN3EE) activation of Erkl intervening amino acid 203 mutant alleles. The Erkl activation was performed as a two-step assay in the presence of [y-32pj ATP (1250 cmp/pmol). The Erkl proteins (~1 ug) were incubated without (light grey) and with (dark grey) constitutively active Mekl (AN3EE) (-53 ng) before performing the MBP kinase assay.  At the completion of the 9 min  preincubation described in Figure 26A, a 5ul aliquot of the exogenous substrate MBP (5 mg/ml) was added to the mixture. The reaction was incubated for a further 1 min before addition of 5 ul of 4X gel loading buffer. The Mekl activating activity is normalized to Erkl wild type basal autoactivation activity which is given a value of 100 mean ± s.d., n=3 (Erkl wild type MBP phosphotransferase activity varied between 85-101 cpm/1 min reaction). The data were analyzed by the ANOVA (analysis of variance) test. Ranked sample means: TEY* TPY TEY TPY TGY TEY* TGY TEY  This is a graphical representation of the Tukey multiple comparisons test. At the 0.05 significance level, the means of any two groups underscored by the same line are not significantly different. The Erkl mutant alleles incubated in the absence (bold face) and presence of Mekl (AN3EE) (normal) are indicated by their tripeptide sequence. The wild type enzyme is denoted by the three letter abbreviation TEY and the kinase-inactive Erkl allele by TEY*. All other alleles are indicated with their amino acid change.  60000 A  -Mekl (AN3EE) +Mekl (AN3EE)  TEY  TEY*  TPY  Erkl mutant alleles  TGY  phosphorylate Erkl. It appears that the charge, structure and size of the amino acid side chain of residue 203 plays a role in Mekl recognition.  233  DISCUSSION  MAP kinase was originally identified in a number of model systems following growth factor or hormone stimulation (Sturgill and Ray, 1986; Ray and Sturgill, 1987; Cicirelli et al, 1988; Pelech et al, 1988; Hoshi et al, 1988). In the work presented here, a systematic approach was used to uncover the players involved in regulating the MAP kinase pathway in both sea stars and mammals. To isolate MAP kinase from human, a partial cDNA was obtained from A-431 epidermoid cell line by applying RT-PCR with paired forward and reverse oligonucleotide primers that were patterned after the partial rat Erkl (Boulton et al, 1990). Two partial overlapping cDNAs of the appropriate size were obtained using this approach. Sequence analysis revealed that these four and five hundred base pair fragments were -98% identical within the same interval to the rat Erkl (data not shown).  Before isolating the full-length human Erkl from a Hep G2 cDNA library, the same hepatocellular carcinoma cell was verified for the presence of Erkl. Indeed, insulin stimulation of Hep G2 cells resulted in a two-fold activation of an MBP kinase activity relative to control cells. Immunoblot analysis of the fractionated extracts with an antipeptide antibody directed toward the carboxy-terminal region of Erkl showed that the MBP kinase activity and the Erkl enzyme co-eluted. Therefore, MAP kinase isoform Erkl protein is expressed in the Hep G2 cell line.  The human Erkl clone was obtained in a screen of a Hep G2 cDNA library. The 1850 base pair fragment excluding the poly(A)+ track corresponded in length with the published sequence from rat. However, an additional stretch of approximately one hundred base pairs were present in human Erkl that were missing from the rat isoform (Charest et al, 1993). Isolation of truncated forms of Erkl from a number of model  systems may be due to the GC-rich content of this region that probably interferred with the synthesis of a full-length cDNA (Boulton et al, 1990; Gonzalez et al, 1992; Owaki et al, 1992; Tanner and Mueckler, 1993). In contrast to previously reported rat Erkl sequences, two potential translational initiation codons were identified at the 5' end of the cDNA. The two start codons located at nucleotides 73 and 160 conformed to the minimal consensus sequence defined by Kozak (1987). In fact, the DNA sequence surrounding the ATG start site was identical with respect to both methionines. To further complicate matters, no signal translational termination codon was located between the two initiation sites nor within the putative 72 base pair 5' untranslated region.  To define the 5' boundary of the Erkl message, primer extension analysis of this region revealed that only 13 bases were missing from the message.  The possible  existence of a third A U G translation initiation site located so close to the 5' cap site was remote since ribosomes normally require longer stretches to initiate translation (Kozak, 1987). Therefore, this result supports the presence of a fixed site for translation initiation in Erkl from Hep G2 cells. The assignment of cDNA nucleotide 73 as the translation initiation site for human Erkl is supported by protein sequence analysis of different MAP kinase isoforms. Although there is a perfect alignment between the second in-frame methionine in human and those of MAP kinases Fus3 and Kssl from S.cerevisiae,  a  valinyl residue is present at the exact position in rat Erkl (Courchene et al, 1989; Elion et al, 1990; Boulton et al, 1990; Charest et al, 1993). In addition, the first initating methionyl site followed by a string of alanyl residues (MAAAAA) is reminiscent of the predicted amino-terminal sequence of Erk2 from human. These results in combination with the full-length sequence from rat indicate that the first A U G is the most likely recognition site for translation from the human Erkl mRNA (Marquardt and Stabel, 1992; Charest etal, 1993).  235  Although Erkl was more effectively autophosphorylated in the presence of the divalent metal cation M n . The reason for this requirement remains unclear since 2 +  purified sea star Mpkl MBP phosphotransferase activity was sensitive to elevated M n  2 +  concentrations (Sanghera et al, 1990a). Perhaps recombinant Erkl expressed in bacteria has a slightly different conformation than the native form and consequently alters the enzymes divalent cation usage. Another possible explanation for this difference is that autophosphorylation on Tyr-204 in Erkl requires the divalent cation M n . 2 +  Recombinant Erkl purified from bacteria also displayed a minor amount of MBP phosphotransferase activity in vitro.  This low level activity was likely due to the  accumulation of phosphotyrosine within the bacteria, since treatment of the recombinant enzyme with the tyrosyl phosphatase CD45 inactivated the enzyme. The maximum level of Erkl autophosphorylation following 20 min incubation was 0.01 mol of P per mol distributed at multiple sites. These data are similar to the results reported for the low activity structure of pTyr-185 in Erk2 (Seger et al, 1991; Robbins et al, 1993). Recombinant Erkl autophosphorylation in vitro occurred at a very low stoichiometry considering the enzyme has two regulatory phosphorylation sites (Payne et al, 1991). The explanation for the low level activation of Erkl by autophosphorylation of Tyr-204 may be explained by examinining the location of the residue within the context of the protein kinase topology. In Erk2, Thr-183 is exposed on the surface of the L12 activation loop, whereas Tyr-185 is buried in a hydrophobic pocket facing toward the active site (Zhang et al, 1994). To become fully active, however, Erk protein kinases require dual phosphorylation on neighbouring threonyl and tyrosyl residues (Anderson et al, 1990; Payne et al, 1991; Canagarajah et al, 1997).  This slow autophosphorylation on tyrosyl residues raised the possibility that a cellular 'enhancing factor' distinct from a protein kinase may stimulate this activity (Wu  et al. 1991; Seger et al, 1991; Ahn et al, 1991; Robbins and Cobb, 1992). However, the Src-related tyrosyl kinase Lck could directly phosphorylate the activating tyrosyl site in purified sea star Mpkl and stimulate its MBP phosphotransferase activity (Ettehadiah et al, 1992). In addition, Tyr-204 in GST-Erkl was also shown to be the major site of phosphorylation by Lck. Unfortunately, this did not stimulate increased activation of human Erkl MBP kinase activity. The reason for the discrepancy between Lck activation of purified sea star Mpkl and recombinant human Erkl is difficult to reconcile since both proteins display close to 80% identity at the amino acid level. It is quite possible that a contaminating protein in the Mpkl preparation was activated in the presence of Lck. Since Lck was shown not to be an Erkl activator, activated sheep platelets were used to assay for a MAP kinase activator. Indeed, a factor was detected in PMA-treated sheep platelets that stimulated GST-Erkl phosphotransferase activity by nearly three-fold. To verify the mechanism of MAP kinase activation, a catalytically compromised Lys-71 mutant of Erkl was used as a substrate. In a manner similar to wild type Erkl, the kinase-inactive version was phosphorylated to nearly identical levels, indicating that the activator possessed intrinsic phosphorylation activity. Moreover, the phosphorylation of GST-Erkl by sheep platelet fractionated extract occurred predominantly on tyrosine and to a lesser extent on threonine.  Evidence using catalytically compromised Erk2  microinjected into Xenopus oocytes revealed that a 'MAP kinase kinase' was responsible for phosphorylating the physiologically relevant sites (Posada and Cooper, 1992). Several groups reported identical observations in PMA- or growth factor- induced MAP kinase activation in mammalian cells (Alessandrini et al, 1992; L'Allemain et al, 1992; Shirakabe et al, 1992). In an analogous approach, MAP kinase kinase partially purified from EGF-activated 3T3 cells displayed an identical amino acid specificity (L'Allemain etal,  1992).  237  The high degree of amino acid sequence conservation between the partial sea star Mpkl and the full-length human Erkl cDNA implied that the mammalian isoform could be a useful reagent for purifying the sea star MAP kinase kinase. A M A P kinase activator, that was capable of increasing Erkl MBP phosphotransferase activity above basal levels, was detected in sea star oocytes induced to mature with the hormone 1MeAde. Like mammalian Mekl, the sea star isoform adsorbed very weakly on a Mono Q anion exchange resin at neutral pH (Gomez and Cohen, 1991; Ahn et al, 1991; Shirakabe et al, 1992; LAllemain et al, 1992). Major and minor peaks of Erkl activator activity eluted at 100 and 200 mM NaCl, respectively. The detection of two activator peaks has been noted in NGF-, EGF- and phorbol ester-treated mammalian cells (Gomez and Cohen, 1991; Ahn et al, 1991; Alessandrini et al, 1992). However, the peak of Erkl activation did not coincide with any of the 6 peaks of Erkl protein phosphorylation. Previously, heat-denatured forms of MAP kinase as well as peptides harbouring the MAP kinase phosphorylation site were shown to be poorly recognized by the upstream activator (Ahn et al, 1991). In control experiments, the TEY peptide phosphorylation profile superimposed that of the full-length Erkl protein. These results indicate that the low level of Erkl phosphorylation by the sea star activator coincides with a small activation of the Erkl phosphotransferase activity. However, these results do not rule out the possibility of an allosteric activation of Erkl by a factor unique to the sea star oocyte system.  The MAP kinase activator from sea star oocytes was purified using a protocol that incorporated six chromatographic resins including DEAE-cellulose (reverse column), hydroxylapatite, S-Sepharose, phosphocellulose, heparin-agarose and Mono S at acidic pH. Two or more of these column steps were also employed in the purification strategies of MAP kinase kinase from mammalian cells, rabbit skeletal muscle and Xenopus oocytes (Nakielny et al, 1992; Seger et al, 1992b; Matsuda et al., 1992; Wu et al, 1992;  Crews and Erikson, 1992). The overall purification of MAP kinase activator from 1MeAde-treated oocytes was about 500-fold. This recovery was much lower than the observed 5,000- to 40,000-fold purification of MAP kinase kinase obtained from other model systems.  Although the purified preparation of sea star MAP kinase activator was not homogeneous, an enriched polypeptide band of 42- to 44-kDa was identified on a SDSPAGE silver stained gel. This band directly correlated with the peak activity of the MAP kinase activator in the final Mono S purification step. Furthermore, gel filtration chromatography indicated that the protein existed as a monomeric protein with an approximate molecular mass of 40- to 42-kDa. On occasion, a second peak of MAP kinase activator activity was eluted earlier from the Superose 12 column fractionation. This may be a complex of proteins which has as one of its components the sea star Erkl activator. Purified fractions from the Mono S and Superose 12 column step were analyzed by Western blotting with amino- and carboxy-terminal antipeptide antibodies directed against a number of different mammalian, Xenopus, and yeast kinases including Stell, Rafl, Mos and Src proteins. No immunoreactive bands were detected in these experiments (data not shown). However, these results cannot eliminate the possibility that the contaminating protein bands identified on the silver-stained gel are protein kinases that were unable to immunoreact with these antibodies due to species-specific sequence divergence at the amino acid level. Mek and Erk proteins are know to form complexes in vitro and in vivo and Erk2 has been shown to exist as protein dimers in crystal structures (Fukuda et al, 1997; Canagarajah et al, 1997). The size of the polypeptide band obtained after six distinct purification steps correlated to within 2-kDa of the molecular mass of the purified MAP kinase kinase from other sources (Nakielny et al, 1992; Seger et al, 1992; Matsuda et al, 1992; Wu et al, 1992; Crews and Erikson, 1992). Several attempts were made to demonstrate that the MAP kinase activator was a  239  bona fide  kinase by monitoring the activity of the kinase under reaction conditions that  favored autophosphorylation. However, despite repeated attempts with different purified preparations of the protein, no autokinase activity was observed for the sea star MAP kinase activator. This contrasted the evidence presented from other MAP kinase kinase purifications. (Seger et al, 1992; Nakielny et al, 1992; Matsuda et al, 1992).  Sequence analysis of tryptic peptides derived from a number of different purified MAP kinase kinase preparations revealed that the isolated Mek proteins were closely related to the yeast proteins Byrl (bypass of Ras) and Ste7 (Nadim-Davis and Nasim, 1988; Teage et al, 1986). Byrl from the fission yeast S. pombe and its cognate, Ste7, from the budding yeast S. cerevisiae, have been implicated in cell cycle control during yeast mating. These two enzymes act as intermediary proteins that transduce signals emanating from receptors at the plasma membrane to specific MAP kinase effectors. The 45-kDa protein product of the Spkl gene in S pombe which has sequence similarity to S. cerevisiae  Fus3 and Kssl are highly related MAP kinases that regulate the conjugation  (Teage et al, 1986; Nadin-Davis and Nasim, 1990; Gartner et al, 1992). In light of the difficulty in demonstrating that the single silver-stained -42 to 44-kDa band observed by SDS-PAGE was the desired polypeptide, the combined peak Erkl activator activity fractions were subjected to Western immunoblotting analysis using antibody probes specific for mouse Mekl and budding yeast Ste7 MAP kinase kinases. Two antibodies directed against peptides patterned after amino acid sequences from highly conserved catalytic subdomain VIII in yeast Ste7 and subdomain XI in mouse Mekl immunoreacted with a 42-kDa band from the most purified preparation of sea star Erkl kinase activator. In contrast, the mouse Mekl carboxy-terminal antibody displayed the least immunoreactivity with the purified protein. All three protein antibodies detected to varying degrees the bacterially expressed recombinant GST-Mekl and partially purified Mek enzyme from TPA-treated sheep platelets. In terms of the size, chromatographic  behavior and immunoreactivity to Mek antibodies derived from two disparate species, the 42-kDa protein purified from maturing sea star oocytes appeared to be the sea star isoform of MAP kinase kinase. The most important difference, however, remains the relatively low degree of activation of Erkl MBP phosphotransferase activity induced by the sea star Mek enzyme relative to purified isoforms from other model systems.  The mechanism of MAP kinase activation has been subject to much speculation given that Erk protein kinases were capable of being activated by a slow, linear autophosphorylation reaction that resulted in a detectable increase in the enzymes MBP phosphotransferase activity (Wu etal,  1991, Seger et al, 1991; Crews et al, 1991;  Rossomondo et al, 1992b; Robbins et al, 1993; Charest et al, 1993). Many originally believed that Erkl autophosphorylation may play a significant role in Erkl regulation (Knighton et al, 1991; Seger et al, 1991; Crew et al, 1991; Wu et al, 1991; Robbins and Cobb, 1992). To determine the mechanism of Erkl activation in the sea star oocytes, both the wild type and its catalytically compromised variant were analyzed in the presence of purified sea star Mek. As expected, addition of the sea star Mek promoted the Erkl phosphorylation and a promoted a 4-fold activation of recombinant GST-Erkl MBP phosphotransferase activity in vitro while at the same time failing to phosphorylate kinase-inactive Erkl. Analysis of the phosphoamino acid content of the sea star Mek 'phosphorylation' of Erkl revealed that all three hydroxyl residues displayed a marked increase in phosphorylation. Similar results were observed with different sea star Mek preparations.  There are several possible scenarios that could explain the low level activation observed for human Erkl by the sea star MAP kinase activator. It is plausible that purified sea star Mek may function as an allosteric activator of Erkl, perhaps via stimulation of MAP kinase autophosphorylation at the regulatory sites. Allosteric  activation has not yet been described between pairs of protein kinases. However, the transcription factor E l k l , which is a MAP kinase substrate, is able to increase the autophosphorylation of Erks and Mpkl in vitro (Rao and Reddy, 1993; Rao and Reddy, 1994). Specific allosteric interactions between kinases may explain Mek's narrow specifity for its physiological substrates Erkl and Erk2 (Ahn et al, 1992; Charest et al, unpublished data). In fact, Zhang et al. (1994) observed that because Tyr-185 in Erk2 is buried in a hydrophobic pocket in the low activity state, the activation loop must assume a completely different conformation when complexed with Mek. The mechanism of Erkl activation by mammalian and sea star Mek is presently under investigation. We are in the process of analyzing how Mek-Erk association may stimulate Erkl autophosphorylation and lead to the activation of MBP phosphotransferase activity. To do so we have constructed unique Mekl mutant alleles that combine an Ala-97 mutation in the essential catalytic lysyl residue with Glu-218 and Glu-222 mutations at the regulatory seryl sites. In so doing, the catalytically-inactive but conformationally active Mek mutant can be tested as an allosteric activator of the Erkl enzyme.  The low specific activity observed for the purified MAP kinase activator from sea star oocytes may result from the presence of a non-kinase 'activation factor' purified during the course of the purification proceedure. This 'activation factor' may stimulate the in vitro autophosphorylation activity of Erkl and consequently stimulate MBP phosphotransferase activity. Evidence of an 'autokinase-enhancing factor' which stimulates the intrinsic autokinase activity of Erk2 has been observed in Swiss 3T3 fibroblast cells (L'Allemain et al, 1992). However, more detailed analysis of the 'enhancing factor' by Sturgill's research group revealed that this novel protein stimulated Mek phosphotransferase activity toward MAP kinase in vitro (L'Allemain et al., 1992). Although the identity of the mammalian 'enhancing factor' remains unknown, it is similar to the 1-MeAde-stimulated sea star Mek activator in that the protein factor activity is  242  sometimes induced in mitogen-stimulated cells (L'Allemain et al, 1992). A second Erkl 'enhancing factor' that activated recombinant Erkl isolated from bacteria was detected in mammalian cells. The purified Erkl activator was shown to be a 16 kDa thiol transferase enzyme. Apparently, Erkl expressed in bacteria forms inactive kinase complexes through disulfide linkages between cystyl residues on the surface of the protein. Exposure to the thiol transferase causes the cleavage of the disulfide bonds and subsequent liberation of the active enzyme (Haystead, personal communication).  Many of the signalling components within each of the M A P kinase modules display a high degree of selectivity based on specific protein-protein interactions. In fact, the formation of large signalling complexes may be the modus operandi within the cell. In mammalian cells, individual kinases Rafl, Mekl/Mek2, Erkl/Erk2 and Rskl/Mnk2 form protein contacts specifically with their immediate upstream activators and their downstream effectors. Discreet interactions between signalling kinases may be important for maintaining fidelity within the signal transduction pathway by limiting interaction with other M A P kinase modules or by regulating its activation/attenuation. The analogous situation exists in the S. cerevisiae, with the exception that the non-protein kinase, Ste5, provides specific docking sites for Stell, Ste7 and Fus3/Kssl kinases (Choi et al, 1994; Marcus et al, 1994; Printen and Sprague, 1994). The specific interactions that occur between protein kinase pairs may explain the absence of a robust phosphorylation and activation of human Erkl by sea star Mek activator. In support of this notion, Ste7 is unable to bind with other yeast (Mpkl and Spkl) or mammalian (Erk2) MAP kinase homologues (Bardwell et al, 1996). To further investigate the role of Mek/MAPK complexes in oocytes, we have isolated the full-length sea star Mpkl cDNA by the RACE (rapid amplification of cDNA ends) method. Sequence comparsison revealed that sea star Mpkl and human Erkl are identical in size (379-amino acid residues), in addition to both enzymes containing the signature TEY activation loop (data  243  not shown). However, both kinases display a great deal of variability within the aminoand carboxy-terminal regions (data not shown). Because purified sea star Mek poorly activates human E r k l , we are currently assessing whether constitutively active Mekl(AN3EE) is capable of stimulating the MBP phosphotransferase activity of sea star Mpkl. Futhermore, we are presently constructing protein chimerics of these enzymes by interchanging the amino- and carboxy-terminal regions between sea star Mpkl and human Erkl to determine whether these regions may regulate Mek binding to MAP kinase. In addition, we want to attempt to redirect mitogen signals in mammalian cells through expression of Erkl/Mpkl chimeric proteins.  Since several high and low molecular mass proteins were present in the purified sea star Mek preparations, the data cannot exclude the possibility that a contaminating factor may regulate Erkl activity. It is possible that one or more of the contaminating silver-stained bands observed on SDS-PAGE  may  promote the renaturation of  misfolded Erkl protein purified from bacteria. As a result, the relative amount of active enzyme that could undergo autophosphorylation would increase during the incubation period. This seems unlikely in light of the fact that the Erkl activator appears to be stimulated during the course of oocyte maturation. Unexpectedly, however, we have observed a factor that is present in both 1-MeAde-stimulated and unstimulated sea star oocytes that is capable of inhibiting Erkl MBP phosphotransferase activity in a concentration-dependent manner. Like sea star Mek, this inhibitory factor does not adsorb to anion exchange columns at pH 7.0. It may be possible that one of the contaminating protein bands in the purified Erkl activator preparation may be a contaminating inhibitory factor. Investigations are presently underway to determined the identity of this heat-stable MAP kinase inhibitory factor.  244  MAP kinases are regulated by dual phosphorylation on neighbouring threonyl and tyrosyl residues in the L12 activation loop of catalytic subdomain VIII (Payne et al, 1992; Canagrarajah et al, 1997). Although the presence of seryl phosphorylation has been observed by a number of investigators, it is assumed to play no role in the regulation of protein kinase activity (Robbins and Cobb, 1992). To understand the mechanism of Erkl activation, a time course of phosphorylation and activation of Erkl by constitutively activated forms of Mekl was performed in vitro with recombinant proteins. The recombinant GST-Mekl (EE) fusion protein which has the regulatory residues (Ser-218 and Ser-222) substituted with glutamic acid mimetics phosphorylated thrombin cleavedErkl to 0.08 mol/mol. In fact, Erkl was phosphorylated 60 to 80 percent more on tyrosyl than on the adjacent threonyl site during the first 30 min. No further changes were observed during the remainder of the reaction. The inability to achieve a stoichiometry of phosphorylation of 2 mol/mol is consistent with previous in vitro results (Robbin et al, 1993; Scott et al, 1995; Robinson  al, 1996). Others have demonstrated that catalytic  amounts of Mek resulted in the accumulation of the Tyr-185 form of Erk2 in vitro (Haystead et al., 1992). In fact, a 40-fold excess of Mek was critical to achieve identical phosphotyrosyl and phosphothreonyl content in Erk (Scott et al, 1995). As mentioned above, Haystead found that a substantial percentage of bacterial expressed Erkl exists in a inactive complex and therefore may not accessible for activation by Mek.  However, proper signal transmission from Mek to Erk in vivo may require interactions with other proteins in a signalling complex. Ste5 in yeast has been proposed to act as a tether in the pheromone signalling pathway in budding yeast to assist in the sequential activation of the members in MAP kinase module. In mammalian cells, factors have been detected that stimulated the rate of Erkl autophosphorylation and increase the rate at which this same enzyme is phosphoryated by Mek in vitro (L'Allemain et al, 1992; Scott et al, 1995). In support of this notion, removal of the  inhibitory amino-terminal NES sequence in Mekl (EE) leads to a dramatic increase in the constitutive phosphorylation of Erk enzyme and transformation of mammalian cells (Mansour et al, 1994). It will be interesting to determine whether the NES region in Mekl becomes displaced by a non-kinase factor.  This factor may induce a  conformational change in the kinase and thereby increase Mekl phosphorylation of Erkl. Mekl (AN3EE) phosphorylation of Erkl reached a maximum between 5 and 10 min under identical conditions used for the Mekl (EE) experiment. Furthermore, this 5-fold higher stoichiometry of phosphorylation (-0.5 mol/mol) translated into a similar increase in Erkl MBP phosphotransferase activity. Although Erkl phosphorylation was not stoichiometric, phosphoamino acid analysis revealed that Mekl (AN3EE) phosphorylated the enzyme equally on threonyl and tyrsosyl residues.  In addition to phosphorylation on threonine and tyrosine, a very robust increase in serine phosphorylation of Erkl was observed in the presence of Mekl(AN3EE). Following serine phosphorylation, Erkl enzyme activity slowly declined to near basal levels during the remainder of the experiment. At the same time phosphoserine also coincided with the appearance a slower migrating Erkl band on SDS-PAGE (data not shown). Several groups have demonstrated that overexpression of MAP kinase dualspecificity phosphatases can lead to the inactivation of Erkl in mammalian cells (Campbell et al, 1995). However, there are no reports of Erk down regulation by Mek phosphorylation or Erk autophosphorylation. Analysis of the primary sequence of Erkl did not reveal any related SEY-like phosphorylation sites. In fact, the data indicate that serine phosphorylation may depend on the activity of Erkl, since phosphoserine appeared at the point where Erkl attained full enzymatic activity. Interestingly, two canonical MAP kinase phosporylation sites are present at Ser-74 and Ser-263 in Erkl. Ser-74 is positioned three amino acid carboxy-terminal to Lys-71 in the ATP-binding site. The Ser263 site is located in the M A P kinase insert region. We are currently taking the  246  necessary steps to identify the site of serine phosphorylation in Erkl.  The crystal  structure has been solved for the active form of Erk2 (Canagarajah et al, 1997). Conformational changes associated with phosphorylated Erk2 causes enzyme dimerization. Apparently, three residues at the carboxy-terminus (Tyr-356 to Ser-358) bind in the active site of the neighbouring molecule (Canagarajah et al, 1997). Therefore, the association of activated Erkl molecules in solution may lead to the slow inactivation of the kinase activity observed in these experiments.  It will also be  interesting to determine if phosphorylation on serine effects the dimeric interaction between individual Erk molecules  At present five distinct MAP kinase modules have been identified in mammalian cells. The individual pathways respond to specific stimuli or environmental perturbations by regulating a unique subset of effector molecules. In many instances, activation of one MAP kinase cascade occurs without stimulating parallel pathways.  Some of the  specificity appears to occur at the level of the Mek/MAP kinase nodal point. Primary sequence comparisons and crystal structure analysis of the members of the MAP kinase family revealed structural motifs that may be essential in regulating kinase function (Zhang et al, 1994). Specifically, the dual phosphorylation motif, Thr-Xaa-Tyr, in the LI2 linker of MAP kinase has been proposed to be an important determinant in the recognition and phosphorylation by upstream activators (Han et al, 1994; Butch and Guan, 1996). Therefore, site-directed mutagenesis was used to examine the role of the intervening sequence Thr-Xaa-Tyr and the individual phosphorylation sites has on the mechanism of Erkl activation.  To test the importance of Erkl threonyl and the tyrosyl regulatory phosphorylation sites in Mekl specificity, these residues were substituted with the small amino acid glycine, the hydroxy amino acid serine or the glutamic acid mimetic. The  SEY and EEY mutants were phosphorylated on tyrosine to the same extent as wild type Erkl. As discussed earlier, this limited phosphorylation on tyrosine leads to minimal activation of Erkl. However, in the presence of the more active Mekl(AN3EE) isoform, wild type Erkl is phosphorylated 2-fold more than the SEY mutant indicating that serine is a poor substitute for threonine in this position. It has been shown previously that a serine substitution at position 183 in Erk2 and 202 in Erkl was a poor phosphoacceptor site (Butch and Guan, 1996; Robinson et al, 1996). It is possible that the methyl moiety present in the threonine R-group is critical for Mekl recognition. To further analyze the role each phosphorylatable residue plays for Mek specificity, the orientation of the TEY motif was inverted in Erkl.  Like the wild type enzyme, recombinant Erkl YET  contained a minor amount of phosphotyrosine after purification from bacteria. Further, YET displayed the same level of autophosphorylation in vitro as did the wild type enzyme. However, there was no observable increase YET phosphorylation in the presence of either activated forms of Erkl. These data demonstrate that at postions 202 and 204 in Erkl, Mekl has a high degree of specificity for the amino acids threonine and tyrosine.  With the exception of the SEY allele, none of the mutants diplayed any appreciable MBP phosphotransferase activity. This probably reflects the specific interactions Thr-202 and Tyr-204 must make with specific residues in the Erkl molecule after phosphorylation by the Mekl. It has been hypothesized that the inactive form of MAP kinase would require both local and global conformational changes after phosphorylation that would be controlled by Erkl phosphorylation (Zhang et al, 1994). Indeed, phosphothreonine and phosphotyrosine interaction with conserved arginine residues induces structural changes that allow the enzyme to assume an active conformation (Canagarajah etal, 1997).  248  The intervening residue between the regulatory threonine and tyrosyl phosphorylation sites differs among the MAP kinase family members. Each subfamily possesses a unique conserved residue (Ferrell, 1996).  In the Erk isoforms, the  intervening Xaa residue is occupied by glutamic acid whereas Jnk and Hog MAP kinase subfamilies encode proline and glycine, respectively, at this position. To examine what role the intervening amino acid plays in directing Mek substrate specificity, the glutamic acid residue in TEY was replaced with the cognate residue present in the two stressactivated M A P kinases.  The TPY and T G Y alleles were phosphorylated to  approximately 30% of the original Erkl wild type enzyme by Mekl. Phosphorylation of TPY and TGY did not stimulate any increase in MBP phosphotransferase activity of the mutant alleles. Previously, it has been suggested that the intervening sequence in Erk2 has no effect on Mekl substrate specificity (Robinson et al, 1996). Similarly, it has been reported that replacement of the glycine residue in Hog had no effect on the specificity of activation by Mkk3 or Mkk6 in transient transfection assays (Jiang et al, 1997). The present work shows that replacement of an acidic amino acid with a residue containing a bulky or small side chain does indeed have a profound effect on Mekl substrate recognition. The unusually long Mek/Erk2 preincubation period, transfection time (48 h) or absence of certain controls may account for why this was not apparent in the earlier studies (Robinson et al, 1996; Jiang et al, 1997).  These data demonstrate that Mekl phosphorylates Erkl sequentially on tyrosyl and threonyl residues that are positioned in the correct orientation.  Further, the TEY  motif plays a pivotal role in Mek substrate specificity. It seems likely that there are unique Mek determinants in the Erk protein kinase backbone that promote the specific association between these two enzymes. A MAP kinase-binding domain has already been mapped to the amino-terminal regions of Mek (Fukuda et al, 1997). Using HogErk chimera, Brunet and Pouyssegur (1996) have mapped a M A P kinase kinase  specificity determinant site to a region between catalytic subdomains III and V in MAP kinase. However, these same authors acknowledge that the chimeric molecules are substantially less active than the intact Hog and Erk proteins. We have recently identified a sea star isoform that contains the same number of amino acids as mammalian Erkl (data not shown). Although the proteins share more than 80% identity at the amino acid level, mammalian Mekl was unable to activate the sea star Erkl enzyme to the same magnitude as the human isoform (data not shown). Since most of the sequence differences between the two enzymes map to regions outside the catalytic domain, it may be possible to create chimeric molecules that will uncover specific Mekl determinant without compromising enzyme activity.  BIBLIOGRAPHY Abe, J., Kusuhara, M . , Ulevitch, R.J., Berk, B.C. and Lee, J.-D. (1996) Big mitogenactivated protein kinase 1 (BMK1) is a redox-sensitive kinase. J. Biol. Chem. 271: 16586-16590. Abe, J., Takahashi, M , Ishida, M . , Lee, J.-D. and Berk, B.C. (1997) c-Src is required for oxidative stress-mediated activation of big mitogen-activated protein kinase 1 (BMK1). J. Biol. Chem. 272: 20389-20394. Adam, L.P. and Hathaway, D.R. (1993) Identification of mitogen-activated protein kinase phosphorylation sequences in mammalian h-caldesmon.. FEBS Lett. 322: 5660. Adams, P.D. and Parker, P.J. (1991) TPA-induced activation of MAP kinase. FEBS Lett. 290: 77-82. Adler, V., Franklin, C.C. and Kraft, A.S. (1992a) Phorbol esters stimulate the phosphrylation of c-Jun but not v-Jun: Regulation of the N-terminal domain. Proc. Natl. Acad. Sci. U. S. A. 89: 5341-5345. Adler, V., Poloskaya, A., Wagner, F. and Kraft, A.S (1992b) Affinity-purified c-Jun amino-terminal protein kinase requires serine/threonine phosphorylation for activity. J. Biol. Chem. 267: 17001-17005. t  Adler, V., Unlap, T. and Kraft, A.S. (1994) A peptide encoding the c-Jun 5 domain inhibits the activity of a c-Jun amino-terminal protein kinase. J. Biol. Chem. 269: 11186-11191. Ahn, N.G. and Krebs, E.G. (1990) Evidence for an epidermal growth factorstimulated protein kinase cascade in Swiss 3T3 cells. Activation of serine peptide kinase activity by myelin basic protein kinases in vitro. J. Biol. Chem. 265: 1149511501. Ahn, N.G., Seger, R., Bratlien, R.L., Diltz, C D . , Tonks, N.K. and Krebs, E.G. (1991) Multiple components in an epidermal growth factor-stimulated protein kinase cascade. J. Biol. Chem. 266: 4220-4227. Alessandrini, A., Crews, C M . and Erikson, R.L. (1992) Phorbol ester stimulates a protein-tyrosine/threonine kinase that phosphorylates and activates the E r k l gene product. Proc. Natl. Acad. Sci. U.S.A. 89: 8200-8204.  Alessi, D.R., Saito, Y., Campbell, D.G., Cohen, P., Sithanandam, G., Rapp, U . , Ashworth, A., Marshall, C.J. and Cowley, S. (1994) Identification of the sites in M A P kinase kinase-1 phosphorylated by p74 f- . EMBO J. 13: 1610-1619. Ra  1  Alvarez, E., Northwook, L C , Gonzalez, F.A., Latour, D.A., Seth, A., Abate, C , Curran, T. and Davis, R.J. (1991) Pro-Leu-Ser/Thr-Pro is a consensus primary sequence for substrate protein phophorylation. J. Biol. Chem. 266: 15277-15285. Amano, M., Mukai, FL, Ono, Y., Chihara, K., Matsui, T., Hamajima, Y., Okawa, K., Iwamatsu, A. and Kaibuchi, K. (1996) Identification of a putative target for Rho as the serine-threonine kinase protein kinase N. Science 271: 648-650. Ammerer, G. (1994) Sex, stress and integrity: the importance of M A P kinase in yeast. Curr. Opin. Genet. Dev. 4: 90-95. Anderson, N.G., Mailer, J.L., Tonks, N.K. and Sturgill, T.W. (1990) Requirement for integration of signals from two distinct phosphorylation pathways for activation of M A P kinase. Nature (London) 343: 651-653.  Anderson, N.G., Li, P., Marsden, L.A., Williams, N., Roberts, T.M. and Sturgill, T.W. (1991) R a f l is a potential substrate for mitogen-activated protein kinase in vivo. Biochem J. 277: 573-576. Angel, P. and Karin, M . (1991) The role of Jun, Fos and the AP-1 complexes in cell proliferation and transformation.  Biochem. Biophys. Acta 1072: 129-157.  Arion, D., Meijer, L., Brizuela, L. and Beach, D. (1988) cdc2 is a component of the M H I kinase: Evidence for identity with M P F . Cell 55: 371-378.  phase specific histone  Ashworth, A., Nakielny, S., Cohen, P. and Marshall, C. (1992) The amino acid sequence of a mammalian M A P kinase kinase. Oncogene 7: 2555-2556. Bagrodia, S., Derijard, B., Davis, R.J. and Cerione, R.A. (1995) Cdc42 and P A K mediated signaling leads to Jun kinase and p38 mitogen-activated protein kinase activation. J. Biol. Chem. 270: 27995-27998.  Baker, S.J., Kerppola, T.K., Luk, D., Vandenberg, M.T., Marshak, D.R., Curran, T., Abate, C (1992) Jun is phosphorylated by several protein kinases at the same sites that are modified in serum-stimulated fibroblasts. Mol. Cell. Biol. 12: 4694-4705. Ballou, L.M., Luther, H. and Thomas, G. (1991) M A P 2 kinase and 70K S6 kinase lie on distinct signalling pathways. Nature 349: 348-350.  Bannerjee, P., Ahmed, M.F., Grove, J.R., Kozlosky, C., Price, D.J. and Avruch, J. (1990) Molecular structure of a major insulin/mitogen-activated 70-kDa S6 kinase from Swiss mouse 3T3 cells. J. Biol. Chem. 263: 1188-1194. Bardwell, L., Cook, J.G., Chang, E.C., Cairns, B.R. and Thorner, J. (1996) Signaling in the yeast pheromone response pathway: Specific and high-affinity interaction of the mitogen-activated protein (MAP) kinases Kssl and Fus3 with the upstream MAP kinase kinase Ste7. Mol. Cell. Biol. 16: 3637-3650. Barrett, C.B., Erikson, E. and Mailer, J.L. (1992) A purified S6 kinase kinase from Xenopus eggs activates S6 kinase II and autophosphorylates on serine, threonine, and tyrosine residues. J. Biol. Chem. 267: 4408-4415. Belka, C , Wiegmann, K., Adam, D., Holland, R., Heuloh, M . , Hermann, F., Kronke, M . and Brach, M A . (1995) Tumor necrosis factor (TNF)-alpha activates c-raf-1 kinase via the p55 TNF receptor engaging neutral sphingomyelinase. EMBO J. 14: 11561165. Bender, A.D. and Sprague Jr., G.F. (1986) Yeast peptide pheromones, a-factor and ccfactor, activate a common response mechanism in their target cells. Cell 47: 929937. Benito, M . , Porras, A., Nebreda, A.R. and Santos, E. (1991) Differentiation of 3T3-L1 fibroblasts to adipocytes induced by transfection of ras oncogenes. Science 253: 565568. Ben-Levy, R., Leighton, LA., Doza, Y.N., Attwood, P., Morrice, N . , Marshall, C.J. and Cohen, P. (1995) Identification of novel phosphorylation sites required for activation of MAPKAP kinase-2. EMBO J. 14: 5920-5930.  Benton, W.D. and Davis, R.W. (1977) Screening A.gt recombinant clones by hybridization to single plaques in situ. Science 196: 180-182. Beyaert, R., Cuenda, A., Vanden Berghe, W., Plaisance, S., Lee, J.C., Haegeman, G., Cohen, P. and Fiers, W. (1996) The p38/RK mitogen-activated protein kinase pathway regulates interleukin-6 synthesis response to tumor necrosis factor. EMBO J. 15: 1914-1923. Birnboim, H.C. and Daly, J. (1979) A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acid Res. 7: 1513-1523 Birchmeier, C , Broek, D. and Wigler, M . (1985) ras proteins can induce meiosis in Xenopus oocytes. Cell 43: 615-621.  Bishop, J.M. (1991) Molecular themes in oncogenesis. Cell 64: 235-248. Blank, J.L., Gerwins, P., Elliott, E.M., Sather, S. and Johnson, G.L. (1996) Molecular cloning of mitogen-activated protein/Erk kinase kinase (Mekk) 2 and 3. J. Biol. Chem. 271: 5361-5368. Blenis, J. and Erickson, R.L. (1986) Regulation of protien kinase activities in PC12 pheochromocytoma cells. EMBO J. 5: 3441-3447. Blenis, J., Kuo, C.J. and Erikson, R.L. (1987) Identification of a ribosomal protein S6 kinase regulated by transformation and growth-promoting stimuli. J. Biol. Chem. 262: 14373-14376. Blumer, K.J. and Thorner, J. (1990) (3 and y subunits of a yeast guanine nucleotidebinding protein are not essential for membrane association of the a subunit but are required for receptor coupling. Proc. Natl. Acad. Sci. U.S.A. 87: 4363-4367.  Blumer, K.J., Johnson, G.L. and Lange-Carter, C A . (1994) Mammalian mitogenactivated protein kinase kinase (Mekk) can function in a yeast mitogen-activated protein kinase pathway downstream of protein kinase C . Proc. Natl. Acad. Sci.  U.S.A. 91: 4925-4929.  Boguslawski, G. and Polazzi, J.O. (1987) Complete nucleotide sequence of a gene conferring polymyxin B resistance on yeast: similarity of the polypeptide to protein kinases. Proc. Natl. Acad. Sci. U.S.A. 84: 5848-5852.  Bohmann, D., Ellis, M . C , Staszewski, L.M. and Mlodzik, M . (1994) Drosophila jun mediates Ras-dependent photoreceptor determination. Cell 78: 973-986. Bokoch, G.M., Wang, Y., Bohl, B.P., Sells, M.A., Quillium, L.A. and Knaus, U.G. (1996) Interaction of the Nek adapter protein with p21-activated kinase (PAK1). J. Biol. Chem. 271: 25746-25749. Boulton, T.G., Yancoupoulous, G.D., Gregory, J.S., Slaughter, C , Moomaw, C , Hsu, J. and Cobb, M.H. (1990) A n insulin-stimulated protein kinase similar to yeast kinase involved in cell cycle control. Science 249: 64-67 Boulton, T.G., Gregory, J.S. and Cobb, M.H. (1991) Purification and properties of extracellular signal kinase 1, an insulin-stimulated microtubule-associated protein 2 kinase. Biochemistry 30: 278-286.  Boulton, T.G., Nye, S.H., Roggins, F.J., Ip, N.Y., Radziejewska, E., Morgenbesser, S.D., DePinho, R.A., Panayotatos, N., Cobb, M.H. and Yancopoulos, G.D. (1991) E R K s : A family of protein-serine/threonine kinases that are activated and tyrosine phosphorylated in response to insulin and NGF. Cell 65: 663-675.  Boulton, T.G. and Cobb, M.H. (1991) Identification of multiple extracellular signalregulated kinases (ERKs) with antipeptide antibodies. Cell. Regul. 2:357-371. Bowtell, D.D.L., Fu, P., Simon, M . and Senior, P. (1992) Identification of murine homologues of the Drosophila Son of sevenless gene: potential activators of ras.  Proc. Natl. Acad. Sci. U.S.A. 89: 6511-6515.  Bradford, M . M . (1976) A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-254. Brown, J.L., Stowers, L., Baer, M . , Trejo, J., Coughlin, S. and Chant, J (1996) Human Ste20 homologue h P A K l links GTPases to the J N K M A P kinase pathway. Curr. Biol. 6: 598-605. Brvta, T. R., Drugan, J. K., Sujoy, G., Terrell, R. S., Campbell-Burk, S., Bell, R. M . and Der, C. J. (1995) Two distinct Raf domains mediate interaction with Ras. J. Biol. Chem. 270: 9809-9812. Brunet, A., Pages, G. and Pouyssegur, J. (1994) Constitutively active mutants of M A P kinase kinase ( M E K 1 ) induce growth factor-relation and oncogenicity when expressed in fibroblasts. Oncogene 9: 3379-3387.  Buday, L., Warne, P.H. and Downward, J. (1995) Downregulation of the ras activation pathway by M A P kinase phosphorylation of sos. Oncogene 11: 1327-1331. Bunone, G., Briand, P.-A., Miksicek, R.J. and Picard, D. (1996) Activation of the unliganded estrogen receptor by E G F involves the M A P kinase pathway and direct phosphorylation. EMBO J. 15: 2174-2183.  Burbelo, P.D., Drechsel, D. and Hall, A . (1995) A conserved binding motif define numerous candidate target proteins for both Cdc42 and Rac GTPases. J. Biol. Chem. 270: 29071- 29074. Burgering, B.M.T., Medema, R.H., Maassen, J.A., van de Wetering, M.L., van der Eb A.J. and McCormick, F. (1991) Insulin stimulation of gene expression mediated by p21ras activation. EMBO J. 10: 1103-1109  Burgering, B.M.T., Pronk, G., van Weeren, P.C., Chardin, P. and Bos, J.L. (1993) c A M P antagonizes p21ras-directed activation of exracellular signal-regulated kinase 2 and phosphorylation of mSos nucleotide exchange factor. EMBO J. 12: 4211-4220.  Burkholder, A.C. and Hartwell, L.H. (1985) The yeast a-factor receptor: Structural properties deduced from the sequence of the STE2 gene. Nucleic Acids Res. 13: 8463-8475. h  Butch, E R . and Guan, K.-L. (1996) Characterization of E R K 1 activation site mutants and the effect on recognition by M E K 1 and M E K 2 . J. Biol. Chem. 277: 4230-4235. Cai, H., Szeberenyi, J. and Cooper, G.M (1990) Effect of a dominant Ha-ras mutation on mitogenic signal transduction in NIH3T3 cells. Mol. Cell. Biol. 10: 5314-5323. Campbell, J.S., Seger, R., Graves, J.D., Graves, L.M., Jensen, M . and Krebs, E.G. (1995) The M A P kinase cascade. Recent Progress in Hormone Research Vol. 50 pp. 131-159. Canagarajah, B.J., Khokhlatchev, A., Cobb, M . H . and Goldsmith, E.J. (1997) Activation mechanism of the M A P kinase E R K 2 by dual phosphorylation. Cell 90: 859-896. Cano, E., Hazzalin, A . and Mahadevan, L.C. (1994) Anisomycin-activated protein kinases p45 and p55 but not mitogen-activated protein kinase E R K - 1 and -2 are implicated in the induction of c-fos and c-jun. Mol. Cell. Biol. 14: 7352-7362.  Cano, E., Doza, Y . N . , Ben-Levy, R., Cohen, P. and Mahadevan, L.C. (1996) Identification of anisomycin-activated kinases p45 and p55 in murine cells as  M A P K A P kinase-2. Oncogene 12: 805-812.  Cantley, L . C , Auger, K.R., Carpenter, C , Duckworth, B., Graziani, A., Kapeller, R. and Soltoff, S. (1991) Oncogenes and signal transduction. Cell 64: 281-302. Cavegelli, M . , Dolfi, F., Fabrizio, D., Claret, F.-X., Karin, M . (1995) Induction of c-fos expression through JNK-mediated T C F / E l k l phosphorylation. EMBO J. 14: 59575964. Ceci, J.D., Patriotis, C P . , Tsatsanis, C , Makris, A . M . , Kovatch, R., Swing, D.A., Jenkins, N.A., Tsichlis, P.N. and Copeland, N.G. (1997) Tpl-2 is an oncogenic kinase that is activated by carboxy-terminal truncation. Genes Dev. 11: 688-700. Chant, J. and Stowers, L. (1995) GTPase cascades choreograghing cellular behavior: movement, morphogenesis, and more. Cell Si: 1-4.  Chardin, P., Camonis, J.H., Gale, N.W., Van Aelst, L., Schlessinger, J., Wigler, M.H. and Bar-Sagi, D. (1993) Human Sosl: A guanine nucleotide exchange factor for Ras that binds to Gbr2. 260: 1338-1343. Charest, D.L., Mordret, G., Harder, K.W., Jirik, F. and Pelech, S.L. (1993) Molecular cloning, expression, and characterization of the human mitogen-activated protein kinase p44 . erkl  Mol. Cell. Biol. 13: 4679-4690.  Chen, R.H., Sarnecki, C. and Blenis, J. (1992) Nuclear localization and regulation of erk- and rsk-encoded protein kinases. Mol. Cell. Biol. 12: 915-927. Cheng, M., Boulton, T.G. and Cobb, M.H. (1996a) E R K 3 is a constitutively nuclear J. Biol. Chem. 271: 8951-8958.  protein kinase.  Cheng, M., Zhen, E., Robinson, M J . , Ebert, D., Goldsmith, E. and Cobb, M.H. (1996b) Characterization of a protein kinase that phosphorylates serine 189 of the mitogenactivated protein kinase homolog E R K 3 . J. Biol. Chem. 271: 12057-12062.  Cheng, J.-T., Cobb, M.H. and Baer, R. (1993) Phosphorylation of the T A L I oncoprotein by the extracellular signal-regulated protein kinase E R K L Mol. Cell. Biol. 13: 801-808.  Cherniack, A.D., Klarlund, J.K., Conway, B.R. and Czech, M.P. (1995) Disassembly of sos-of-sevenless proteins form Grb2 during p21ras desensitization by insulin.  Biol. Chem. 270: 1485-1488.  J.  Chiba, K., Tadenuma, H., Matsumoto, M., Takahashi, K., Katada, T. and Hoshi, M . (1992) The primary structure of the a subunit of a starfish guanosine-nucleotidebinding regulatory protein involved in 1-methyladenine-induced oocyte maturation.  Eur. J. Biochem. 207: 833-838. Childs, T.J., Watson, M.H., Sanghera, J.S., Campbell, D.L., Pelech, S.L. and Mak, A.S. (1992) Phosphorylation of smooth muscle caldesmon by mitogen-activated protein (MAP) kinase and expression of M A P kinase in differentiated smooth muscle cells.  J. Biol. Chem. 267: 1868-1874.  Choi, K.-Y., Satterberg, B., Lyons, D.M. and Elion, E.A. (1994) Ste5 tethers multiple protein kinases in the M A P kinase cascade required for mating in S. cerevisiae. Cell 79: 131-141.  Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162: 156159.  Chuang, E., Barnard, D., Hettich, L., Zhang, X. F., Avruch, J. and Marshall, M.S. (1994) Critical binding and regulatory interactions between Ras and Raf occur through a small stable N-terminal domain of Raf and specific Ras effector residues. Mol Cell.  Biol. 14: 5318-5325.  Chung, J., Chen, R.-H. and Blenis, J. (1991a) Coordinate regulation of pp9W and a sk  distinct protein-serine/threonine kinase activity that phosphorylates recombinant pp9QT in sk  vitro.  Mol. Cell. Biol. 11: 1868-1874.  Chung, J., Pelech, S.L. and Blenis, J. (1991b) Mitogen-activated Swiss mouse 3T3 R S K kinases I and II are related to pp44 "/'* from sea star oocytes and participate in the regulation of pp90™* activity. Proc. Natl. Acad. Sci. U.S.A. 88: 4981-4985. /  Chung, J., Kuo, C.J., Crabtree, G.R. and Blenis, J. (1992) Rapamycin-FKBP specifically blocks growth-dependent activation of and signaling by the 70 kd S6 protein kinase. Cell 69: 1227-1236.  Cicirelli, M.F., Pelech, S.L. and Krebs, E.G. (1988) Activation of multiple protein kinases during the burst of protein phosphorylation that precedes the first meiotic cell division in Xenopus oocytes. J. Biol. Chem. 263: 2009-2019.  Clark-Lewis, I., Sanghera, J.S. and Pelech, S.L. (1991) Definition of a consensus sequence for peptide substrate recognition by p44 M p k , the meiosis-activated myelin basic protein kinase. J. Biol. Chem. 266: 15180-15184.  Clifton, A.D., Young, P.R. and Cohen, P. (1996) A comparison of the substrate specificity of M A P K A P kinase-2 and M A P K A P kinase-3 and their activation by cytokines and cellular stress. FEBS Lett. 392: 209-214.  Cobb, M.H. (1986) A n insulin-stimulated ribosomal protein S6 kinase. J. Biol. Chem. 261: 12994-12999. Cobb, M.H., Boulton, T.G. and Robbins, D.J. (1991) Extracellular signal-regulated kinases: E R K s in progress. Cell Regulation 2: 965-978. Cohen, P. (1983) Protein phosphorylation and the control of glycogen metabolism in skeletal muscle. Phil. Trans. R. Soc. Lond. B302: 13-25. Cook, S.J. and McCormick, F. (1993) Inhibition of c A M P of Ras-dependent activation of Raf. Science 262: 1069-1072.  Cooper, J.A., Sefton, B . M . and Hunter, T (1983) Detection and quantification of  phosphotyrosine in proteins. Methods Enzymol. 99: 387-402.  Corbalan-Garcia, S., Yang, S.-S., Degenhardt, K.R. and Bar-Sagi, D. (1996) Identification of the Mitogen-Activated Protein kinase phosphorylation sites on human Sosl that regulate interaction with Grb2. Mol. Cell. Biol. 16: 5674-5682.  Coso, O.A., Chiariello, M . , Yu, J.-C, Teramoto, H., Crespo, P., Xu, N., Miki, T. and Gutkind, S. (1995) The small GTP-binding proteins R a c l and Cdc42 regulate the activity of the Jnk/Sapk signaling pathway. Cell 81: 1137-1146. Courchesne, W.E., Kunisawa, R. and Thorner, J. (1989) A putative protein kinase overcomes pheromone-induced arrest of cell cycling in S. cerevisiae. Cell 58: 11071119.  Cowley, S., Paterson, H., Kemp, P. and Marshall, C J . (1994) Activation of M A P kinase kinase is necessary and sufficient for P C 12 differentiation and for transformation of NIH 3T3 cells. Cell 77: 841-852.  Creasy, C L . and Chernoff, J. (1995a) Cloning and characterization of a human protein kinase with homology to Ste20. J. Biol. Chem. 270: 21695-21700. Creasy, C L . and Chernoff, J. (1995b) Cloning and characterization of a member of the M S T subfamily of Ste20-like kinase. Gene 767: 303-306. Crews, C M . , Alessandrini, A. and Erikson, R.L. (1991) Mouse E r k l gene product is a serine/threonine protein kinase that has the potential to phosphorylate tyrosine.  Proc. Natl. Acad. Sci. U.S.A. 88: 8845-8849.  Crews, C M . and Erikson, R.L. (1992) Purification of a murine proteintyrosine/threonine kinase that phosphorylates and activates the Erk-1 gene product: Relationship to the fission yeast byrl gene product. Proc. Natl. Acad. Sci. U.S.A. 89:  8205-8209.  Crews, C M . , Alessandro, A. and Erikson, R.L. (1992) The primary structure of Mek, a protein kinase that phosphorylates the E r k gene product. Science 258: 478-480. Cuenda, A., Rouse, J., Doza, Y.N., Meier, R., Cohen, P., Gallagher, T.F., Young, P.R. and Lee, J.C. (1995) SB 203580 is a specific inhibitor of a M A P kinase homologue which is stimulated by cellular stresses and interleukin-1. FEBS Lett. 364: 229-233.  Cuenda, A., Alonso, G., Mortice, N., Jones, M., Meier, R., Cohen, P. and Nebreda, A. (1996) Purification and cDNA cloning of S A P K K 3 , the major activator of RK/p38 in stress- and cytokine- stimulated monocytes and epithelial cells. EMBO J. 16: 41564164. Cuenda, A . , Cohen, P., Buee-Sherrer, V. and Goedert, M . (1997) Activation of stressactivated protein kinase-3 (SAPK3) by cytokines and cellular stresses is mediated via S A P K K 3 ( M K K 6 ) ; comparison of the specificities of S A P K 3 and S A P K 2 (RK/p38). EMBO J. 16: 295-305.  Davis, B.J. (1964) Disc electorphoresis II. Method and application to human serum proteins. Ann. N.Y. Acad. Sci. 121:404Dent, P., Haser, W., Haystead, T.A.J., Vincent, L . A . , Roberts, T . M . and Sturgill, T . M . (1992) Activation of mitogen-activated protein kinase by v-raf in NIH3T3 cells and in vitro. Science 257: 1404-1407. Dent, P., Jelinek, T., Morrison, D.K., Weber, M.J. and Sturgill, T.W. (1995) Reversal of Raf-1 activation by purified and membrane-associated protein phosphatases.  Sciences 268: 1902-1906.  de Vries-Smits, A.M.M., Burgering, B.M.T., Leevers, S.J., Marshall, C.J. and Bos, J.L. (1992) Involvement of p21ras in activation of extracellular signal-regulated kinase 2. Nature (London) 357: 602-604. Derijard, B., Hibi, M., Wu, I-H., Barrett, T., Su, B., Deng, T., Karin, M . and Davis, R.J. (1994) J N K 1 : A protein kinase stimulated by U V light and Ha-ras that binds and phosphorylates the c-Jun activation domain. Cell 76: 1025-1037. Derjard, B., Raingeaud, J., Barrett, T., Wu, I.-H., Han, J., Ulevitch, R.J. and David R.J. (1995) Independent human M A P kinase signal transduction pathways defined by M E K and M K K isoforms. Science 267: 682-685. Diener, K., Wang, X.S., Chen, C , Meyer, C.F., Keesler, G., Zukowski, M., Tan, T.-H. and Yao, Z. (1997) Activation of the c-Jun N-terminal kinase pathway by a novel protein kinase related to human germinal center kinase. Proc. Natl. Acad. Sci. U.S.A. 94: 9687-9692. Dbring, F., Drewes, G., Berling, B. and Mandelkow, E.-M. (1993) Cloning and sequencing of cDNA encoding rat brain mitogen-activated protein (MAP) kinase activator. Gene 131: 303-304.  Dorow, D.S., Devereux, L., Deitzsh, E. and De Kretser, T. (1993) Identification of a new family of human epithelial protein kinases containing two leucine/isoleucine-  zipper domains. Eur. J. Biochem. 213: 701-710.  Downward, J., Graves, J.D., Warne, P.H., Rayter, S. and Cantrell, D.A. (1990) Stimulation of p21 upon T-cell activation. Nature (London) 346: 719-723. Dressier, K.A., Mathias, S. and Kolesnick, R.N. (1992) Tumor necrosis factor-a activates the sphingomyelin signal transduction pathway in a cell-free system.  Science 255: 1715-1718.  Drewes, G., Lichtenberg-Kraag, B., Doring, R, Mandelkow, E.M., Biernat, J., Goris, J., Doree, M . and Mandelkow, E. (1992) Mitogen-activated protein (MAP) kinase transforms tau protein into an Alzheimer-like state. EMBO J. 11: 2131-2138. Egan, S.E. and Weinberg, R.A. (1993) The pathway to signal achievement. Nature (London) 365: 781-783. Elion, E.A., Brill, J.A. and Fink, G.R. (1990) FUS3 encodes a cdc2/cdc28-related kinase required for the transition from mitosis to conjugation. Cell 60: 649-664. Elish, M.E., Pierce, J.R. and Earhar, C F . (1988) Biochemical analysis of spontaneous fepA mutants of Escherichia coli. J. Gen. Microbiol. 134: 1355-1364. Engel, K., Plath, K. and Gaestel, M . (1993) The M A P kinase-activated protein kinase 2 contains a proline-rich SH3-binding domain. FEBS Lett. 336: 143-147. Engel, K., Schultz, H., Martin, F., Kotlyarov, A., Plath, K., Hahn, M., Heinemann, TJ. and Gaestel, M . (1995) Contitutive activation of mitogen-activated protein kinaseactivated protein kinase 2 by mutation of phosphorylation sites and an A-helix motif. J. Biol. Chem. 270: 27213-27221.  English, J.M., Vanderbilt, C.A., Xu, S., Marcus, S. and Cobb, M.H. (1995) Isolation of M E K 5 and differential expression of alternatively spliced forms. J. Biol. Chem. 270: 28897-28902. Erhardt, P. Troppmair, J., Rapp, U. R. and Cooper, G. M . (1995) Differential regulation of Raf-l and B-Raf and Ras-dependent activation of mitogen-activated protein kinase by cyclic A M P in PC12 cells. Mol. Cell. Biol. 75: 5524-5530.  Erikson, E. and Mailer, J.L. (1986) Purification and characterization of a protein kinase from Xenopus eggs highly specific for ribosomal protein S6. J. Biol. Chem. 261: 350-355. Erickson, A.K., Payne, D.M., Martino, P., Rossomondo, A.J., Shabanowitz, J., Wever, M.J., Hunt, D.F. and Sturgill, T.W. (1990) Identification by mass spectrometry of Thr97 in bovine myelin basic protein as a phosphorylation site for M A P (mitogenactivated protein) kinase. J. Biol. Chem. 265: 19728-19735.  Errede, B. and Ammerer, G. (1989) STE12, a protein involved in cell-type specific transcription and signal transduction in yeast, is part of protein-DNA complexes.  Genes Dev. 3:1349-1361.  Ettehadieh, E., Sanghera, J.S., Pelech, S.L., Hess-Bienz, D., Watts, J., Shastri, N. and Aebersold, R. (1992) Tyrosyl phosphorylation and activation of M A P kinases by p56Lck. Science 255: 853-855. Ezoe, K., Lee, S.T., Strunk, K.M. and Spritz, R.A. (1994) PTK1, a novel protein kinase required for proliferation of human melanocytes. Oncogene 9: 935-938.  Fan, G., Merritt, S.E., Kortenjann, M., Shaw, P.E. and Holzman, L.B. (1996) Dual leucine zipper-bearing kinase (DLK) activates p46SAPK and p38MAPK but not ERK2. J. Biol. Chem. 271: 24788-24793. Fanger, G.R., Johnson, N.L. and Johnson, G.L. (1997) Mek kinases are regulated by E G F and selectively interact with Rac/Cdc42. EMBO J. 16: 4961-4972. Faramisco, J.R., Gross, M., Kamata, T., Rosenberg, M.D. and Sweet, R.W. (1984) Microinjection of the oncogene form of the human H-ras (T-24) protein results in rapid proliferation of quiescent cells. Cell 38: 109-117. Feinberg, A.P. and Vogelstein, B. (1983) A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132: 613. Feinberg, A.P. and Volgelstein, B. (1984) A technique for radiolabeling D N A restriction endonuclease fragments to high specific activity (Addendum).  Biochem. 137: 266-267.  Anal.  Ferrell, J.E., Jr. (1996) M A P kinases in mitogenesis and development. In Current Topics in Developmental Biology. Edited by Pederson, R.A. and Schatten, G.P. London: Academic Press In.; 33: 1-60  Freshney, N.W., Rawlinson, L., Guesdon, F., Jones, E., Cowley, S., Hsuan, J. and Saklatvala, J. (1994) Interleukin-1 activates a novel protein kinase cascade that results in the phosphorylation of Hsp27. Cell 78: 1039-1049. Friesen, H., Lunz, R., Doyle, S. and Segall, J. (1994) Mutation of the SPSl-encoded protein kinase of Saccharomyces cerevisiae leads to defects in transcription and morphology during spore formation. Genes Dev. 8: 2162-2175.  Frodin, M . , Peraldi, H.J. and Van Obberghen, E. (1994) Cyclic A M P activates the mitogen-activated protein kinase cascade in PC12 cells. J. Biol. Chem. 269: 62076214. Fukuda, M . , Gotoh, Y. and Nishida, E. (1997) Interaction of M A P kinase with M A P kinase kinase: its possible role in the control of nucleocytoplasmic transport of M A P kinase. EMBO J. 16: 1901-1908.  Fukunaga, R. and Hunter, T. (1997) M N K 1 , a new M A P kinase-activated protein kinase, isolated by a novel expression screening method for identifying protein kinase substrates. EMBO J. 16: 1921-1933.  Galcheva-Gargova, Z., Derijard, B., Wu, I-H. and Davis, R.J. (1994) A n osmosensing signal transduction pathway in mammalian cells. Science 265: 806-808. Galisteo, M.L., Chernoff, J., Su, Y . - C , Skolnik, E.Y. and Schlessinger, J., (1996) The adaptor protein Nek linked receptor tyrosine kinases with the serine-threonine kinase P a k l . J. Biol. Chem. 271: 20097-21000.  Gallo, K.A., Mark, M.R., Scadden, D.T., Wang, Z., Gu, Q. and Godowski, P.J. (1994) Identification and characterization of S P R K , a novel src-homology 3 domaincontaining proline-rich kinase with serine/threonine kinase activity. J. Biol. Chem.  269: 15092-15100.  Gartner, A . , Nasmyth, K . and Ammerer G. (1992) Signal transduction in Saccharomyces cerevisiae requires tyrosine and threonine phosphorylation of FUS2 and K S S l . Genes & Dev. 6: 1280-1292. Gerwin, P., Blank, J.L. and Johnson, G.L. (1997) Cloning of a novel mitogen-activated protein kinase kinase kinase, Mekk4, that selectively regulates the c-Jun amino terminal kinase pathway. J. Biol. Chem. 272: 8288-8295.  Gibbs, J.B., Marshall, M.S., Scolnick, E.M., Dixon, R.A.F. and Vogel, U.S. (1990) Modulation of guanine nucleotides bound to Ras in HIH3T3 cells by oncogenes, growth factors, and the GTPase activating protein (GAP). J. Biol. Chem. 265:  20437-20442.  Gille, Ff., Sharrocks, A . D . and Shaw, P.E. (1992) Phosphorylation of transcription factor p 6 2 T C F by M A P kinase stimulates ternary complex formation at c-fos promoter. Nature (London) 355:414-417.  Glise, B., Bourbon, H. and Noselli, S. (1995) hemipterous encodes a novel Drosophila M A P kinase kinase, required for epithelial cell sheet movement. Cell 83: 451-461. Goedert, M . , Cohen, E.S., Jakes, R. and Cohen, P. (1992) p42 M A P kinase phosphorylation sites in microtubule-associated protein tau are dephosphorylated by protein phosphatase 2A. FEBS Lett. 312: 95-99.  Goedert, M., Cuenda, A., Craxton, M., Jakes, R. and Cohen, P. (1997) Activation of the novel stress-activated protein kinase S A P K 4 by cytokines and cellular stresses is mediated by S K K 3 ( M K K 6 ) ; comparison of its substrate specificity with that of other SAP kinases. EMBO J. 16: 3563-3571.  Gomez, N. and Cohen, P. (1991) Dissection of the protein kinase cascade by which nerve growth fetor activates M A P kinases. Nature (London) 353: 170-173. Gonzalez, F.A., Raden, D.L., Rigby, M.R. and Davis, R.J. (1992) Heterogeneous expression of four M A P kinase isoforms in human tissues. FEBS Lett. 304: 17-178. Gosh, S., Strum, J.C, Sciorra, V.A., Danier, L. and Bell, R.M. (1996) Raf-1 kinase possesses distinct binding domains for phosphatidylserine and phosphatide acid. J. Biol. Chem. 271: 8472-8480.  Gotoh, Y., Nishida, E., Yamashita, T., Hoshi, M . , Kawakami, M . and Sakai, H. (1990) Microtubule-associated-protein (MAP) kinase activated by nerve growth factor and  epidermal growth factor in PC12 cells. Eur. J. Biochem. 193: 661-669.  Gotoh, Y., Moriyama, K., Matsuda, S., Okumura, E., Kishimoto, T., Kawasaki, H., Suzuki, K., Yahara, I., Sakai, H. and Nishida, E. (1991) Xenopus M phase M A P kinase: isolation of its cDNA and activation by M P F . EMBO J. 10: 2661-2668. Gotoh, Y., Matsuda, S., Takenaka, K., Hattori, S, Iwamatsu, A., Ishikawa, M., Kosado, H., Nishida, E. (1994) Characterization of recombinant Xenopus M A P kinase kinase mutated at potential phosphrylation sites. Oncogene 9: 1891-1898.  Graves, L.M., Bornfeldt, K.E., Raines, E.W., Potts, B.C., Macdonald, S.G., Ross, R. and Krebs, E.C. (1993) Protein kinase A antagonizes platelet-derived growth factorinduce signaling by mitogen-activated protein kinase in human arterial smooth muscle cells. Proc. Natl. Acad. Sci. U.S.A. 90: 10300-10304.  Gregory, J.S., Boulton, T.G. and Sang, B.-C. (1989) A n insulin-stimulated ribosomal protein S6 kinase from rabbit liver. J. Biol. Chem. 264: 18397-18401. Gressner, A . M . and Wool, I.G. (1974) The stimulation of the phosphorylation of ribosomal protein S6 by cyclohexamide and puromycin. J. Biol. Chem. 249: 69176925. Guan, K.-L. and Dixon, J.E. (1991) Eukaryotic proteins expressed in Escherichia coli: an improved thrombin cleavage and purification procedure of fusion proteins with glutathione -S-transferase. Anal. Biochem. 192: 262-267.  Guesdon, F., Freshney, N., Waller, R.J., Rawlinson, L . and Saklatvala, J. (1993) Interleukin 1 and tumor necrosis factor stimulate two novel protein kinases that phosphorylate the heat shock protein hsp 27 and (3-casein. J. Biol. Chem. 268: 4236-  4243. Gupta, S., Campbell, D., Derijard, B. and Davis, R.J. (1995) Transcription factor A T F 2 regulation by the J N K signal transduction pathway. Science 267: 389-393. Gupta, S., Barrett, T., Whitmarsh, A.J., Cavanagh, J., Sluss, H.K., Derijard, B. and Davis, R. J. (1996) Selective interaction of J N K protein kinase isoforms with transcription factors. EMBO J. 15: 2760-2770. Hafner, S., Adler, H.S., Mischak, H., Janosch, P., Ffeidecker, G., Wolfman, A., Pippig, S., Lohse, M., Ueffing, M . and Kolch, W. (1994) Mechanism of inhibition of Raf-1 by protein kinase A. Mol. Cell. Biol. 14: 6696-6703. Hagag, N., Halegoua, S. and Viola, M . (1986) Inhibition of growth factor-induced differentiation of PC12 cells by microinjection of antibody to ras p21.  (London) 319: 680-682.  Nature  Hagen, D.C., McCaffrey, G.and Sprague, G.F., Jr. (1991) Pheromone response elements are necessary and sufficient for basal and pheromone-induced transcription of the FUS1 gene of Saccaromyces cerevisiae. Mol. Cell. Biol. 11:  2952-2961.  Hall, A. (1994) A biochemical function for Ras - At last. Science 264: 1413-1414.  Han, J., Lee J.-D., Tobias, P.S., Ulevitch, R.J. (1993) Endotoxin induces Rapid protein tyrosine phosphorylation in 70Z/3 cells expressing CD14. J. Biol. Chem. 268: 2500925014. Han, J., Lee, J.-D., Bibbs, L . and Ulevitch, R.J. (1994) A M A P kinase targeted by endotoxin and hyperosmolarity in mammalian cells. Science 265: 808-811. Han, J., Lee, J.-D., L i , Z., Feng, L. and Ulevitch, R.J. (1996) Characterization of the structure and function of a novel M A P kinase kinase (MKK6). J. Biol. Chem. 271: 2886-2891. Han, J., Jiang, Y., L i , Z., Kravchenko, V.V. and Ulevitch, R.J. (1997) Activation of the transcription factor M E F 2 C by the M A P kinase p38 in inflammation. Nature (London) 386: 296-299. Hanks, S.K., Quinn, A . M . and Hunter, T. (1988) The protein kinase family: Conserved features and deduced phylogeny of the catalytic domains. Science 241: 42-52. Haselbacher, G.K., Humbel, R.E. and Thomas, G. (1979) Insulin-like growth factor: Insulin or serum increase phosphorylation of ribosomal protein S6 during transition of stationary chick embryo fibroblasts into early G l phase of the cell cycle. FEBS  Lett. 100: 185-190.  Hattori, S., Fukuda, M., Yamashita, T., Nakamura, S., Gotoh, Y. and Nishida, E. (1992) Activation of mitogen-activated protein kinase and its activator by ras in intact cells and in a cell-free system. J. Biol. Chem. 267: 20346-20351.  Haycock, J.W., Ahn, N.G., Cobb, M.H. and Krebs, E.G. (1992) E R K 1 and E R K 2 , two microtubule-associated protein 2 kinases, mediate the phosphorylation of tyrosine hydroxylase at serine-31 in situ. Proc. Natl. Acad. Sci. U.S.A. 89: 2365-2369.  Haystead, T.A., Dent, P., Wu, J., Haystead, C.M.M. and Sturgill, T.W. (1992) Ordered phosphorylation of p42 P by M A P kinase kinase. FEBS Lett. 306: 17-22. raa  k  Hazzalin, C.A., Cano, E., Cuenda, A., Barratt, M.J., Cohen, P. and Mahadevan, L.C. (1996) p38/RK is essential for stress-induced nuclear responses: J N K / S A P K s and cJun/ATF2 phosphorylation are insufficient. Curr. Biol. 6: 1028-1031. He, M., Wilde, A. and Kaderbhai, M.A. (1989) A simple-step procedure for small-scale preparation of Escheria coli plasmids. Nucleic Acids Res. 18: 1660.  Heider, H., Hug, C. and Lucocq, J.M. (1994) A 40-kDa myelin basic protein kinase, distinct from erkl and erk2, is activated in mitotic HeLa cells. Eur. J. Biochem. 219: 513-520. Heldin, C.-H. (1995) Dimerization of cell surface receptors in signal transduction. Cell 80: 213-223. Her, J.-H., Lakhani, S., Zu, K., Vila, J., Dent, P., Sturgill, T.W. and Weber, M J . (1993) Dual phosphorylation and autophosphorylation in mitogen-activated protein (MAP) kinase activation. Biochem. J. 296: 25-31/  Herskowitz, I. (1995) M A P kinase pathways in yeast: for mating and more. Cell 80: 187-197. Hibi, M . , Lin, A., Smeal, T., Minden, A. and Karin, M . (1993) Identification of an oncoprotein-and UV-responsive protein kinase that binds and potentiates the c-Jun activation domain. Genes Dev. 7: 2135-2148.  Hirai, S., Izawa, M . , Osada, S., Spyrou, G. and Ohno, S. (1996) Activation of the J N K pathway by distantly related protein kinases, M E K K and M U K . Oncogene 12: 641650. Holmes, K. and Quigley, M . (1981) A rapid boiling method for the preparation of bacterial plasmids. Anal. Biochem 114: 193-197. Holzman, L.B., Merritt, S.E. and Fan, G. (1994) Identification, molecular cloning, and characterization of dual leucine zipper bearing kinase. A novel serine/threonine protein kinase that defines a second subfamily of mixed lineage kinases. J. Biol.  Chem. 269: 30808-30817.  Hordjik, P.L., Verlaan, I., Jalink, K., van Corven, E J . and Moolenaar, W.H. (1994) c A M P abrogates the p21ras-mitogen-activated fibroblasts. J. Biol. Chem. 269: 3534-3538  protein kinase pathway  in  Hoshi, M . , Hishida, E. and Sakai, H. (1988) Activation of a Ca2+-inhibitable protein kinase that phosphorylates microtubule-associated protein 2 in vitro by growth factors, phorbol ester, and serum in quiescent cultured human fibroblasts. J. Biol.  Chem. 263: 5396-5401.  Howe, L.R., Leevers, S J., Gomez, N., Nakielny, S., Cohen, P. and Marshall, C J . (1992) Activation of the M A P kinase pathway by the protein kinase raf. Cell 71: 335-342.  267  Hu, C.-D., Kariya, K., Tamada, M . , Akasaka, K., Shirouzu, M . , Yokoyama, S. and Kataoka, T. (1995) Cystein-rich region of Raf-1 interacts with activator domain of post-translationally modified Ha-Ras. J. Biol. Chem. 270: 30274-30277. Hu., M.C.-T., Qui, W.R., Wang, X., Meyer, C F . and Tan, T.-H. (1996) Human HPK1, a novel hematopoietic progenitor kinase that activates the J N K / S A P K kinase cascade.  Genes Dev. 10: 2251-2264.  Hubbard, S.T., Wei, L., Ellis, L. and Hendrickson, W.A. (1994) Crystal structure of the tyrosin kinase domain of the human insulin receptor. Nature (London) 372: 746-754. Huleihel, M . , Goldsborough, M . , Gunnell, M . , Cleveland, J.L., Bonner, T.I. and Rapp, U.R. (1986) Characterization of murine A-raf, a new oncogene related to the v-raf oncogene. Mol. Cell. Biol. 6: 2655-2662. Hunter, T. and Karin, M . (1992) The regulation of transcription by phosphorylation. Cell 70: 375-387. Ichijo, H., Nishida, E., Irie, K., ten Dijke, P., Saitoh, M . , Morigushi, T., Takagi, M . , Matsumoto, K., Miyazono, K. and Gotoh, Y. (1997) Induction of apoptosis by ASK1, a mammalian M A P K K K that activates SAPK7JNK and p38 signaling pathways.  Science 275: 90-94.  Ing, Y.L., Leung, I.W., Heng, H.H., Tsui, L.C. and Lassam, N.J. (1994) M L K - 3 : identification of a widely-expressed protein kinase bearing an SH3 domain and a leucine-zipper basic region domain. Oncogene 10: 1447-1451.  Iordanov, M.S., Pribnow, D., Magun, J.L., Dinh, T.-H., Pearson, J.A., Chen, S.L.-Y. and Magun, B.E. (1997) Ribotoxic stress response: Activation of the stress-activated protein kinase J N K 1 by inhibitors of the peptidyl transferase reaction and by sequence-specific R N A damage to the a-sarcin/ricin loop in the 28S r R N A . Mol.  Cell. Biol. 17: 3373-3381. Irie, K., Takase, M . , Lee, K.S., Levin, D.E., Araki, H., Matsumoto, K. and Oshima, Y. (1993) M K K 1 and M K K 2 , which encode Saccharomyces cerevisiae mitogenactviated protein kinase kinase homologs, function in the pathway mediated by protein kinase C . Mol. Cell. Biol. 13: 3076-3083.  Ish-Horovicz, D. and Burke, J.F. (1979) Rapid and efficient cosmid vector cloning. Nucleic Acids Res. 9:2989-2998.  Ishizakik, T., Maekawa, M . , Fujisawa, K., Okawa, K., Iwamatsu, A., Fujita, A., Watanabe, N., Saito, Y., Kakizuka, A., Morii, N. and Narumiya, S. (1996) The small GTP-binding protein Rho binds to and activates a 160 kDa Ser/Thr protein kinase homologous to myotonic dystrophy kinase. EMBO J. 15: 1885-1893.  Itoh, T., Kaibuchi, K., Masuda, T., Yamamoto, T., Matsuura, Y., Maeda, A., Shimizu, K. and Takai, Y. (1993) A protein factor for ras p21-dependent activation of mitogenactivated protein (MAP) kinase through M A P kinase kinase. Proc. Natl. Acad. Sci. U.S.A. 90: 975-979. Jaffe, L.A., Gallo, C.J., Lee, R.H., Ho, Y . - K . and Jones, T.L.Z. (1993) Oocyte maturation in starfish in mediated by the Py-subunit complex of a G-protein. J. Cell. Biol. 121: 775-783. Jahng, K., Ferguson, J. and Reed, S.I. (1988) Mutations in a gene encoding the a subunit of a Saccharomyces cerevisiae G protein indicate a role in mating pheromone signaling. Mol. Cell. Biol. 8: 309-320.  Jaiswal., R., Moodie, S., Wolfman, A. and Landreth, G. (1994) The mitogen-activated protein kinase cascade is activated by B-Raf in response to nerve growth factor through interaction with p 2 1 . Mol. Cell. Biol. 14: 6944-6953. r a s  Jaiswal, R., Weissinger, E., Kolch, W. and Landreth, G.E. (1996) Nerve growth factormediated activation of the mitogen-activated protein (MAP) kinase cascade involves a signaling complex containing B-Raf and HSP90. J. Biol. Chem. 271: 23626-23629.  Jamal, S. and Ziff, E. (1990) Transactivation of c-fos and (3-actin genes by raf as a step in early response to transmembrane signals. Nature (London) 344:463-466. Janknecht, R., Ernst, W.H., Pingoud, V. and Nordeim, A. (1993) Activation of ternary complex factor Elk-1 by M A P kinases. EMBO J. 12: 5097-5104. Jelinek, T., Dent, P., Sturgill, T.W. and Weber, M.J. (1996) Ras-induced activation of Raf-1 is dependent on tyrosine phosphorylation. Mol. Cell. Biol. 16: 1027-1034. Jenness, D.D., Burkholder, A.C. and Hartwell, L.H. (1983) Binding of a-factor pheromone to yeast a cells: Chemical and genetic evidence for an a-factor receptor.  Cell 35: 521-529.  Johnson, L.N., Noble, M.E.M. and Owen, D.J. (1996) Active and inactive protein kinases: structural basis for regulation. Cell 85: 149-158. Jones, S., Erikson, E., Blenis, J., Mailer, J. and Erikson, R.L. (1988) A Xenopus ribosomal protein S6 kinase has two apparent kinase domains that each are similar to distinct protein kinases. Proc. Natl. Acad. Sci. U.S.A. 85: 3377-3381.  Joneson, T., McDonough, M., Bar-Sagi, D. and Van Aelst, L. (1996) R A C regulation of actin polymeraization and proliferation by a pathway distinct from Jun kinase.  Science 271: 1374-1376.  Joseph, C.K., Byun, H,-S., Bittman, R. and Kolesnick, R.N. (1993) Substrate recognition by ceramide-actived protein kinase. J. Biol. Chem. 268: 20002-20006. Kallunki, T., Su, B., Tsigelny, I., Sluss, H.K., Derijard, B., Moore, G., Davis, R. and Karin, M . (1994) JNK2 contains a specific-determining region responsible for efficient c-Jun binding and phosphorylation. Genes Dev. 8: 2996-3007. Kallunki, T., Deng, T., Hibi, M . and Karin, M . (1996) c-Jun can recruit J N K to phosphorylate dimerization partners via specific docking interactions. Cell 87: 929939. Kamps, M.P., Taylor, S.S. and Sefton, B . M . (1994) Direct evidence that oncogenenic tyrosine kinases and cyclic AMP-dependent protein kinse have homologous A T P binding sites. Nature (London) 310: 589-592.  Karin, M . and Smeal, T. (1992) Control of transcription factors by signal transduction patyways: The beginning of the end. Trend. Biochem. Sci. 17: 418-422. Kato, S., Endoh, H., Masuhiro, Y., Kitamoto, T., Uchiyama, S., Sasaki, H., Masushiga, S., Gotoh, Y., Nishida, E., Kawashima, H.., Metzger, D. and Chambon, P. (1995) Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 270: 1491-1494.  Katz, P., Whalen, G. and Kehr, J.H. (1994) Differential expression of a novel protein kinase in human B lymphocytes. J. Biol. Chem. 269: 16802-16809. Kaur, P., Welch, W.J. and Saklatvala, J. (1989) Interleukin  1 and tumour necrosis factor increase phosphorylation of the small heat shock protein. FEBS Lett. 258:  269-273.  Keifer, F., Tibbie, L.A., Anafi, M., Janssen, A., Zanke, B.W., Lassam, N., Pawson, T., Woodgett, J.R. and Iscove, N.N. (1996) H P K 1 , a hematopoietic protein kinase activating the S A P K / J N K pathway. EMBO J. 15: 7031-7025. Kharbanda, S., Ren, R., Pandey, P., Shafman, T.D., Feller, S.M., Weichselbaum, R.R. and Kufe, D.W. (1995) Activation of the c-Abl tyrosine kinase in the stress response to DNA-damaging agents. Nature (London) 376: 758-788. Klarland, J.K., Cherniack, A.D., McMahon, M . and Czech, M.P. (1996) Role of the Raf/Mitogen-activated protein kinase pathway in p21ras desensitization. J. Biol. Chem. 271: 16674-16677. Knaus, U.G., Morris, S., Dong, H.J., Chernoff, J. and Bokoch, G.M. (1995) Regulation of human leukocyte p21-activated kinases through G protein-coupled receptors.  Science 269: 221-223.  Knighton, D.R., Zheng, J., Ten Eyck, L.F., Ashford, V.A., Xuong, N.-H., Taylor, S.S. and Sadowski, J.M. (1991) C