Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of the pim-1 oncogene-encoded protein kinase Palaty, Chrystal K. 1995

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

Item Metadata

Download

Media
831-ubc_1995-060365.pdf [ 7.03MB ]
Metadata
JSON: 831-1.0088373.json
JSON-LD: 831-1.0088373-ld.json
RDF/XML (Pretty): 831-1.0088373-rdf.xml
RDF/JSON: 831-1.0088373-rdf.json
Turtle: 831-1.0088373-turtle.txt
N-Triples: 831-1.0088373-rdf-ntriples.txt
Original Record: 831-1.0088373-source.json
Full Text
831-1.0088373-fulltext.txt
Citation
831-1.0088373.ris

Full Text

CHARACTERIZATION OF THEpini-1 ONCOGENE-ENCODED PROTEIN KINASEbyCHRYSTAL K. PALATYB.Sc., The University of Victoria, 1989A THESIS SUBMIYI’ED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of Experimental MedicineWe accept this thesis as conformingTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1995© Chrystal K. Palaty, 1995In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. 1 further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of vQ4”” II,PThe University of British ColumbiaVancouver, CanadaDate ce:pi;’: 2-3%SDE-6 (2/88)ABSTRACTPim- 1 is an oncogene-encoded serine-threonine kinase, expressed primarily inhematopoietic and germ cells. Previously identified only in mammalian systems, pim-1was cloned and sequenced from Xenopus laevis, the African clawed frog. The codingregion of X. laevis Pim- 1 encoded a protein of 324 amino acids, which exhibited 64%amino acid sequence identity with the full-length human protein. PCR was also used todemonstrate the existence of Pim- 1 in Pisaster ochraceus, the purple sea star. The highsequence conservation observed in the catalytic domain of Pim- 1 between divergent speciessupport a conserved and important function for this kinase.The full-length coding regions of both human and X. laevis Pim- 1 were expressed asrecombinant bacterial fusion proteins which exhibited phosphotransferase activity towardsexogenous substrates as well as serine, threonine and tyrosine autophosphorylationactivity. A kinase-inactive mutant was engineered to serve as a negative control. Thephosphorylation site consensus sequence for recognition by Pim- 1 was defined bystepwise alterations in the amino acid sequences of peptide substrate analogues todetermine which of the amino acid residues surrounding the substrate phosphorylation sitewere critical for kinase recognition. The optimal substrate peptide for Pim- 1 wasdetermined to be K/R - K/R - R - K/R - L - SIT - X, where X is an amino acid residue witha small side chain.Studies were undertaken to determine the autophosphorylation sites of the GST-Pim- 1kinase, using electron spray ionization mass spectroscopy (ESI-MS). Theautophosphorylation sites of the GST-Pim-1 were identified as Ser-4, Ser-190 and Thr205. These sites were conserved amongst all Pim-1 homologues. An additional site wasidentified on the GST protein, Thr-17. To assess the importance of the Ser-190 site onphosphotransferase activity, the Ser- 190 residue was changed to alanine and to glutamicacid using PCR site-directed mutagenesis. These mutants were expressed in bacteria asGST-fusion proteins, and their activites were compared to the wild-type.Together with Pim- 1-specific antibodies, the optimal Pim- 1 peptide substrate was usedto study endogenous Pim-1 during X. laevis and P. ochracheus oocyte maturation. Pim-1did not exhibit maturation-induced activation in sea star oocytes and the quantity of Pim- 1remained constant during the oocyte maturation process. This is the first study carried outto investigate this kinase in the oocyte system.UTABLE OF CONTENTSAbstract iiTable of contents iiiList of Tables XivList of Figures xvNomenclature and Abbreviations xviii1. Measurements, reagents and units xviii2. Amino acids xxii3. Nucleotides xxiiAcknowledgement xxiiiPART 1 - INTRODUCTION AND BACKGROUNDCHAPTER I. INTRODUCTION 1PJM-l EXPRESSION AND REGULATION1. Viral origins 1i. Discovery of the pim-] oncogene 1ii. Activation by proviral insertion 22. Cloning the pim-1 gene 33. Oncogenic potential of pim- 1 4i. Studies in murine leukemogenesis 4ii. No proof that pim-1 is a human oncogene 5iii. Expression of Pim- 1 in human cancers and cell lines 6a. pim-1 mRNA expression in human cancers and cell lines 6b. Pim- 1 protein expression in human cancers and cell lines 64. Pim-1 gene expression 8i. Selective mRNA expression patterns 8ii. Alternate mRNA transcripts & different half-lives 8iii. Regulation ofpim-1 gene expression 10a. Transcriptional attenuation by DNA secondary structure 12b. Pim-1 regulation by mRNA destabilization 12c. Pim-1 regulation by differing rates of transcription 13ifi5. Pim-1: the protein 14i. Expression in normal tissue 14ii. The product of the pim-1 gene is a kinase 14iii. Expression of the Pim- 1 protein product 16iv. Protein half-life of Pim-1 17STUDIES INTO THE FUNCTION OF PIM-16. Pim- 1 substrate specificity 187. Upstream regulators: growth factors and mitogens 20i. Mitogen and growth factor stimulation of Pim- 1 20ii. Effects of mitogen stimulation on Pim- 1 21iii. Pim-1 may play a role in T cell receptor signaling 21iv. Induction of Pim-1 expression by PKA and PKC 22v. Pim- 1 is induced in response to signaling through receptorsof the GM-CSF family 22a. Induction of Pim-1 by GM-CSF 23b. Background information on the GM-CSF/IL-311L-5 receptor family 23c. Pathways induced by the IL-3/GM-CSF receptors 24d. Potential upsteam activators ofpim-1 transcription 27e. Other signal transduction pathways possibly involving Pim-1 28f. Possible involvement in an apoptotic pathway? 288. Pim- 1 transgenic mice 29i. Transgenic animals confirm that pim-J is an oncogene 29ii. Proviral tagging to identify cooperating oncogenes 30iii. Studies with pim-1/myc double transgenic mice 31iv. Other studies with double transgenic mice 32v. pim-1 null mutants - embryonic stem cells 33vi. The lack of physiological effects in pim-1 null mice 34vii. Growth factor stimulation of cells from transgenic animals 34viii. Susceptibility ofpim-1 transgenic animals to chemical carcinogens 359. The role of other oncogene-encoded serine/threonine kinasesin oocyte maturation 3710. Summary of Pim-1 38ivCHAPTER II. RATIONALE, HYPOTHESIS, OBJECTIVES 40CHAPTER III. METHODS 431. Supplies and Sources 431. Chemical reagents and laboratory supplies 432. Photography supplies 493. Plasmids and bacterial strains 494. Antibody reagents 49i. Primary antibodies 49ii. Secondary antibodies 49iii. Unique antibody reagents 505. Sources of oligonucleotides 506. Peptide production/sources 507. Sources of cell lines/cell lysates 518. Additional reagents 512. Experimental Procedures - Molecular Biology 521. General molecular biology techniques 52i. Isolation of PCR bands from agarose gels 52ii. Restriction digests 52iii. Alkaline phosphatase treatment of linearized plasmids 53iv. Ligations 53v. Transformations 53vi. Small scale plasmid preparation 54vii. Large scale plasmid preparation 54viii. Construction of apim-1 probe: Labeling a DNA probe with [y—32PJ ATP 55ix. Prehybridization and hybridization of membranes with radiolabeled probe 552. Oocyte maturation 56i. Isolation of X. laevis oocytes 56ii. Progesterone maturation of X. laevis oocytes 57iii. P. ochraceus oocyte maturation 573. Isolation of total RNA from ooeytes 57i. Homogenization of oocytes for total RNA 57ii. Quantitation and purity assessment of RNA 58iii. Selection of poly(A)+ RNA 58iv. Northern blot analysis of oocyte RNA 59V4. Amplifying pim- 1 using polymerase chain reaction (PCR) 60i. Reverse transcriptase reaction 60ii. Cleaning the cDNA 61iii. Deprotection, cleavage and working up oligonucleotides 61iv. Quantitation of oligonucleotides 62v. Specific PCR reaction conditions for amplifying pim-1 62vi. Confirmation of identity of PCR clones 635. Screening a Zap cDNA library 64i. Screening the X. laevis cDNA library 64ii. Screening the P. ochraceus cDNA library 65iii. Southern blotting 65iv. Sequencing positive clones to confirm identity 666. Construction of Pim- 1 expression vectors 66i. PCR of aX. laevis cDNA clone 66ii. Construction of Pim-1 mutants using PCR site-directed mutagenesis 67iii. Construction of a H. sapiens Pim-1 expression vector 703. Experimental Procedures - Protein Biochemistry 731. General protein biochemistry techniques 73i. Protein quantitation 73ii. Column fractionations of protein extracts 73iii. SDS-PAGE gels 74iv. Staining SDS-PAGE gels 74a. Coomassie staining 74b. Silver staining 75c. Amido black staining 75v. Western blotting of SDS gels 75a. Alkaline phosphatase (ALP)-conjugated secondary antibody 76b. Horseradish peroxidase-conjugated secondary antibody (ECL) 76vi. Stripping and reprobing Western blots 76vii. Autoradiography and development of film 77viii. Immunoprecipitation 77ix. Kinase assays of immunoprecipitations 78x. Protein-protein interactions 78a. GST-Pim-1 protein affinity columns 78b. Far Western blotting procedure 79vi1. Labeled Pim-1 probe 792. Labeling of immobilized proteins with X. laevis GST-Pim- 1 80xi. Antibody production 80a. Pimi-Ill, Piml-NT, Piml-XI 80b. Anti-X. laevis GST-Pim-1 serum 81xii. Phosphatase treatment of proteins 81a. Acid phosphatase 81b. Alkaline phosphatase 82c. Phosphatase HPTPB 822. Production and purification of GST-fusion proteins in bacteria 82i. Expression and purification 82ii. Thrombin cleavage of fusion proteins 833. Assessment of endogenous kinase activity of expressed Pim-1 84i. Autophosphorylation of GST-Pim- 1 84ii. Determination of specific activity 84iii. Determining the stoichiometry of autophosphorylation of GST-Pim- 1 84iv. Phosphoamino acid analysis of in vitro autophosphorylated Pim-1 854. Assessment of exogenous kinase activity of expressed Pim- 1 85i. Phosphorylation of protein substrates 85ii. Phosphorylation of synthetic peptides 86iii. Determination of kinetic constants 86iv. Stability of GST-Pim-1 enzyme at 30°C 86v. Optimization of kinase reactions 87a. Metal ion optimizations of GST-Pim- 1 kinase reactions 87b. ATP optimizations of GST-Pim- 1 kinase reactions 87c. Time course of Pim- 1 activity 87vi. Antibody inhibition of GST-Pim-1 kinase activity 88vii. Peptide inhibition of endogenous GST-Pim- 1 activity 88viii. Peptide inhibition of exogenous GST-Pim- 1 activity 885. Identification of autophosphorylation sites of expressed Pim- 1 90i. In vitro autophosphorylation and tryptic digestion of fusionprotein for tryptic phosphopeptide analysis 90ii. Two-dimensional phosphopeptide mapping 90iii. Extraction of tryptic phosphopeptides from TLC plates 91iv. Phosphoamino acid analysis of tryptic phosphopeptides 91v. HPLC analysis of tryptic phosphopeptides 91viia. IMAC-HPLC-ESI-MS analysis of tryptic peptides from 2D mapping 91b. LCMS analysis of tryptic digests of X. laevis Pim- 1 926. Sources of endogenous Pim- 1 protein 94i. Analysis of Pim- 1 protein in X. laevis oocytes 94ii. Pim-1 protein in P. ochraceus oocytes 94a. Homogenization of oocytes for protein extraction 94b. P. ochraceus oocyte maturation time courses 94iii. Probing crude bovine spleen extract for activated Pim-1 protein 95iv. Human erythroblast cell line - K562 95v. Primary human lymphocytes 957. Computer search analysis 96PART 2 - RESULTS AND DISCUSSION: BACTERIALLY EXPRESSEDPIM-1CHAPTER IV. CLONING AND EXPRESSION OF XENOPUSLAEVIS GST-Pim-1 and COMPARISON TO HUMAN GST-PIM-1 971. Cloning pim-1 from an X. laevis oocyte eDNA library 972. Comparison of the X. laevis Pim-1 sequence with other proteins 102i. Nucleotide sequence searches 102ii. Protein sequence searches 102iii. Pim-1 homology tree 1083. Bacterial expression of Pim- 1 as a GST-fusion protein 114i. Expression of human Pim- 1 as a bacterial GST-fusion protein 114a. Attempted separation of different Pim- 1 fragmentsby column chromatography 119b. Thrombinization of human GST-Pim- 1 119ii. Expression of X. laevis Pim- 1 as a bacterial fusion protein 119iii. Expression of a kinase-dead mutant of X. laevis GST-Pim-1 1214. Stability of GST-Pim- 1 enzyme 1245. Autophosphorylation activity of GST-Pim- 1 1246. General characterization of bacterially expressed GST-Pim- 1 126i. Time course of activity 126vmii. Divalent cation requirement 126jjj. Km of ATP 128iv. Protein kinase inhibitors 1287. Summary of GST-Pim- 1 expression 129CHAPTER V. SUBSTRATE STUDIES: IDENTIFYINGIN VITRO TARGETS OF PIM-1 1301. Preliminary substrate studies 130i. In vitro phosphorylation of protein substrates by GST-Pim- 1 130ii. Phosphoamino acid analysis of phosphorylated substrates 133iii. Peptide substrate comparisons to published data 1332. Substrate analysis using peptide analogs 136i. Location of phosphoacceptor site 137ii. Influence of C-terminal residues 137iii. Amino acid specificity of the phosphoacceptor site 140iv. Influence of the -1 amino acid residue 140v. Influence of -2,-3 and -4 amino acid residues 143vi. Influence of the -5 and -6 amino acid residues 143vii. Determining the substrate consensus sequence 1463. Data bank search for potential substrates of Pim- 1 1474. Inhibition of Pim- 1 activity by inhibitor peptides 148i. Inhibition of phosphotransferase activity by pseudo-substrate peptides 148ii. Inhibition of autophosphorylation activityby substrate and pseudo-substrate peptides 1515. Conclusions 152CHAPTER VI. ANALYSIS OF AUTOPHOSPHORYLATION 1531. Detection of GST-Pim-1 with anti-phosphotyrosine antibodies 1532. Analysis and identification of GST-Pim- 1 autophosphorylation sites 156i. Two dimensional phosphopeptide analysis of GST-Pim- 1 156ii. Phosphoamino acid analysis of spots from 2D phosphopeptide mapping 158iii. Identification of tryptic phosphopeptides 159ixa. Analysis of isolated tryptic phosphopeptides by IMAC-HPLC-ESI-MS 159b. Analysis of trypsinized GST-Pim- 1 by LCMS 1663. Identifying other kinses with similar phosphorylatable residues 1684. Construction and bacterial expression ofSl9O mutants 171i. Phosphoamino analysis of mutants 171ii. Specific activity determinations 171iii. Exogenous kinase activity of mutants 1755. Construction of additional mutants 1756. Conclusions of Pim- 1 autophosphorylation 177xPART 3 - RESULTS AND CONCLUSIONS: PIM-1 IN BIOLOGICALSYSTEMSCHAPTER VIII. EXAMINING THE ACTIVATION OF ENDOGENOUSPIM-1 DURING XENOPUS LAEVIS OOCYTE MATURATION 1801. Detection of endogenous Pim- 1 in X. laevis oocyte extracts 1802. Analysis of activation of Pim- 1 during oocyte maturation 183i. Peptide substrates 183ii. Immunoreactivity of oocyte maturation time course 183iii. Summary of X. laevis oocyte maturation results 1893. Conclusions 191CHAPTER IX. EXAMINING PIM-1 ACTIVITY INMATURING P. OCHRACEUS OOCYTES 1921. Partial cloning of sea star Pim-1 by PCR, and comparison topublished sequences 1922. Northern blotting of endogenous sea star pim-1. 1973. Examining changes in Pim-1 during sea star oocyte maturation 199i. Western blotting of Pim- 1 from sea star oocyte extracts 199ii. In vitro peptide studies to examine Pim-1 activation duringtime courses of oocyte maturation 1994. Summary 205xiPART 4 - DISCUSSION/FUTURE DIRECTIONSDISCUSSION 206FUTURE DIRECTIONS 2131. Identification of proteins interacting/regulating/phosphorylated by Pim- 213i. The yeast two hybrid system. 213ii. Far Western blots and fusion protein affinity columns 213iii. Purification of Pim- 1 from endogenous sources 214iv. Identification of upstream activators of Pim- 1 using sea star oocyte extracts 2142. Examining in vivo effects of Pim-l in the oocyte system by microinjection 214i. Pim-l ablations 215ii. KD Pim-1 dominant-negative effects 215iii. Effects of WT and mutant Pim- 1 over-expression 2153. Mutation of other autophosphorylation sites of X. laevis Pim- 1 2154. Further studies of Pim- 1 in the sea star oocyte system 216i. Further examination of the oocyte-specific pim-1 RNA transcripts. 216ii. Cloning of the sea star cDNA 216iii. Fertilization studies of sea star oocytes 216iv. Localization of Pim- 1 activity 216BIBLIOGRAPHY 217APPENDIXESAPPENDIX I. Description of Pim-1 antibodies 229APPENDIX II. Antibody Characterizations1. Peptide selection and Western blotting 231i. H. sapiens Pim-1 antibodies - Piml-Ill and Piml-NT 231ii. X. laevis Pim-1 antibodies - GXP and Piml-XI 2342. Comparison of Pim- 1 antibodies from various sources 2373. Inimunoprecipitation 239i. Immunoprecipitation of Pim- 1 from the K562 (human erythroid) cell line 239ii. Immunoprecipitation of Pim-1 from P. ochraceus oocyte extracts 241iii. Immunoprecipitation of Pim- 1 from X. laevis oocyte extracts 241xu4. Antibody inhibition of kinase activity 2425. Antibody summary 244APPENDIX III. Description of oligonucleotides 245APPENDIX IV. Peptide substrates 249APPENDIX V. Calculation formulae for ATP usage 250APPENDIX VI. Calculation of kinetic constantsMichaelis Menten and Lineweaver Burke 251APPENDIX VII. Autophosphorylation/dephosphorylation and exogenouskinase activity 253i. Dephosphorylation of GST-Pim-1 253ii. Dephosphorylation curve of HPTPB 255iii. Activity of GST-Pim- 1 after autophosphorylation/dephosphorylation 255APPENDIX VIII. Preliminary experiments in oocyte systems1. Associated proteins and substrates of X. laevis Pim- 1 256i. Recombinant GST-Pim-1 fusion protein affinity columns 256a. X. laevis oocyte proteins bound to human GST-Pim-1 257b. X. laevis oocyte proteins bound to X. GST-Pim- 1 259ii. Far Western Blotting 2612. Preliminary data from P. ochraceus system 263i. GST-Pim- 1 affinity columns 263ii. Far Westerns of sea star oocyte maturation extracts 263iii. Upstream activators of Pim- 1 265iv. Initial in vitro fertilization studies 265XillLIST OF TABLESTable 1. Kinases sharing homology with X. laevis Pim-1. 104A. Pim-1 serine/threonine protein kinases. 104B. SNF- 1 reated serine/threonine protein kinases. 104C. Calcium/calmodulin-dependent protein kinases. 105D. Other serine/threonine protein kinases (poorly characterized). 106Table 2. Names and accession numbers of kinases listed in the Pim- 1relatedness tree. 113Table 3. Comparison of several peptides phosphorylated by GST-Pim-1. 135Table 4. Identification of serine residues phosphorylated by X. laevisand H. sapiens GST-Pim-1. 138Table 5. Influence of C-terminal residues on substrate phosphorylationby GST-Pim-1. 139Table 6. Amino acid specificity of the phosphoacceptor site. 141Table 7. Influence of the -1 amino acid residue. 142Table 8. Importance of -2, -3 and -4 arginine residues. 144Table 9. Influence of the -5 and -6 amino acid residues. 145Table 10. Phosphoamino acid analysis of peptides isolated from the2D phosphopeptide map. 158Table 11. Activity of Ser-190 site mutants toward P4 peptide. 176xivLIST OF FIGURESFigure 1. Model of GM-CSF receptor signal transduction as mediatedby two distinct domains of the Be subunit. 26Figure 2. Restriction map of X. laevis pim-1 in the pGEX-2T vector. 68Figure 3. Restriction map of H. sapiens pim-1 in the pGEX-2T vector. 72Figure 4. Southern blot of X. laevis cDNA clones probed withPim- 1 PCR fragment. 98Figure 5. Nucleotide and amino acid sequence of coding regionof X. laevis pim-1. 99Figure 6. Protein sequence alignment of Pim- 1 from Xenopuslaevis, mouse, rat and human. 101Figure 7. Protein sequence alignments of X. laevis Pim-1 withhomologous proteins. 110Figure 8. Homology tree of Pim- 1 with related kinases. 112Figure 9. Bacterially-expressed H. sapiens GST-Pim-1. 116Figure 10. Expression of human GST-Pim-1 in DH5a andUT5600 E. coli strains. 117Figure 11. Phosphoamino acid analysis of bacterially-expressed GST-Pim- 1. 118Figure 12. Bacterially-expressed X. laevis GST-Pim- 1. 120Figure 13. Comparison of WT and KD X. laevis GST-Pim- 1. 123Figure 14. Specific activity of autophosphorylation of H. sapiens andX. laevis GST-Pim- 1. 125Figure 15. Cations required for optimal GST-Pim-1 phbsphotransferaseactivity. 127Figure 16. Phosphorylation of various protein substrates by X. laevisGST-Pim- 1. 131Figure 17. Phosphorylation of various protein substrates by H. sapiensGST-Pim- 1. 132Figure 18. Competitive inhibition of exogenous phosphotransferaseactivity of H. sapiens GST-Pim-1 by P28 peptide. 149Figure 19. Inhibition of exogenous phosphotransferase activity ofH. sapiens GST-Pim-1 by P29 peptide. 150Figure 20. Tyrosine phosphorylation of expressed X. laevis GST-Pim- 1. 155Figure 21. Two-dimensional phosphopeptide map of X. laevis GST-Pim- 1. 157xvFigure 22. Total ion chromatogram of spot 7. 160Figure 23. Fragmentation of peptide bonds by ionization. 159Figure 24. Mass spectrum of aa 185-195, LIDFGSGALLK. 162Figure 25. Mass spectrum of aa 196-206, DTVYTDFDGTR. 163Figure 26. Mass spectrum of GST-Pim-1 peptide, GSMLLSK. 164Figure 27. Mass spectrum of peptide GST aa 10-20, IKGLVQPTRLL. 165Figure 28. Autophosphorylation sites of X. laevis Pim-1. 167Figure 29. Kinases with sites homologous to the Ser-190 in kinase.catalytic subdomain VII. 169Figure 30. Alignments of subdomain VIII of Pim-1 with several kinases. 170Figure 31. Kinases having a tyrosine site homologous to Tyr-133 site ofPim-1 in domain V. 170Figure 32. Nucleotide sequence of coding region of X. laevispim-1 showing expected location of mutations. 172Figure 33. Comparison of mutant and wild-type GST-Pim-1. 173Figure 34. Specific activity of autophosphorylation of S 190>A andSi 90>E mutants of Pim- 1. 174Figure 35. Immunodetection of Pim- 1 in homogenates of a X. laevisoocyte maturation time course. 182Figure 36. P4 peptide phosphorylation by fractionated immature andmature X. laevis oocyte extracts. 185Figure 37. P4 peptide phosphorylation by fractionated X. laevisoocyte maturation time course. 186Figure 38. Western blots of fractionated X. laevis oocyte extracts. 187Figure 39. Western blots of selected MonoQ fractions of X. laevisoocyte time course. 188Figure 40. Partial nucleotide sequence of sea star Pim-1. 194Figure 41. Comparison of the partial amino acid sequence of sea starPim-1 with other species. 196Figure 42. Northern blot of sea star RNA. 198Figure 43. Western blot of MonoQ-fractionated immature and maturesea star oocyte extract. 200Figure 44. Peptide phosphorylation profiles of fractionated sea star oocytematuration time course. 201Figure 45. Comparison of fractionated sea star oocyte maturation time points. 202Figure 46. Sea star oocyte time course: peak MonoQ fractions. 204xviFigure 47. Regions of Pim-1 proteins against which antibodies are directed. 232Figure 48. Detection of bacterially-expressed and endogenous Pim- 1with H. sapiens antibodies. 233Figure 49. Detection of bacterially-expressed and endogenous Pim- 1with X. laevis antibodies. 236Figure 50. Pim- 1 antibody comparison by immunodetection of Pim- 1on Western blots. 238Figure 51. Immunoprecipitation of Pim- 1 from K562 cells withvarious antibodies. 240Figure 52. Antibody inhibition of kinase activity. 243Figure 53. Dephosphorylation and subsequent phosphorylation ofX. laevis GST-Pim- 1. 254Figure 54. X. laevis oocyte extracts bound to H. sapiens GST-Pim-1fusion protein affinity columns. 258Figure 55. X. laevis oocyte extracts bound to X. laevis GST-Pim-1fusion protine affinity columns - silver stain. 260Figure 56. Far Western blot of fractionated X. laevis oocytesphosphorylated by expressed GST-Pim- 1. 262Figure 57. Sea star oocyte extracts bound to GST-Pim- 1 fusion proteinaffinity column. 264Figure 58. Phosphorylation of GST-Pim- 1 by a maturation activated kinase. 266Figure 59. Western blot of fractionated in vitro fertilized sea star oocytes. 267xviiNOMENCLATURE AND ABBREVIATIONS1. MEASUREMENTS. REAGENTS AND UNiTSaa amino acidALP alkaline phosphataseamp ampicillinamp’ ampicillin resistanceamu atomic mass unitAPS ammonium persulfateATP adenosine 5’-triphosphate disodium saltBCIP 5-bromo-1-chloro-3 indoyl phosphateBSA bovine serum albuminBSE bovine spleen extractsbp base pair°C degrees celsiusC-terminal carboxyl-terminalCaFASW calcium free artificial sea watercDNA copy DNA-product of reverse transcriptase reactionCi Currie, 2.22 x 1012 disintegrations per minuteConA Concanavalin AdH2O distilled waterddH2O autoclaved millipore waterDNA deoxyribonucleic acidDMF dimethylformamideDMSO dimethyl sulfoxidedNTP 2’-deoxynucleoside 5’-triphophateDTT dithiothreitolECL enhanced chemiluminescenceEDTA ethylene diamine tetraacetate disodium saltEGTA ethylene bis (oxyethylenenitrilo)] tetraacetic acidEpo erythropoetinESI-MS electrospray ionization mass spectrometryFCS fetal calf serumFeLV feline leukemia virusF-MuLV friend-helper leukemia virusg gramxviiix g times the force of gravityGST glutathione S-transferaseGVBD germinal vesicle breakdownh hourHEPES N-(2-hydroxyetheyl)piperazine-N’-(2-ethanesulphonic acid)HPLC high pressure liquid chromatographyhsp heat shock proteinIg immunoglobulinIL interleukinIFN-y interferon gammaIMAC immobilized metal affinity chromatographyIP immunoprecipitation, immunoprecipitateIPTG isopropyl 13-D-thiogalactopyranosidekDa kilodalton, a measure of molecular masskb kilobase (1000 bp)KOAc potassium acetate1 litreLC50 lethal concentration for 50% of animalsLCMS liquid chromatography/mass spectrometryLTR long terminal repeatmA miii Ampsml miflulitremg milligram (10-3)M moles/litremMol millimolmM millimolarmm minute2-ME 2-mercaptoethanolMES 2-[N-morpholino]ethanesulfonic acidMOPS 3-[N-morpholino]propanesulphonic acidMr molecular massmRNA messenger ribonucleic acidMuLV murine leukemia virusm/z mass to charge ratioN-terminus amino terminusNaOAc sodium acetatexixNBT nitro blue tetrazoliumNEM N-ethyl maleimideNLS nuclear localization signalnM nanomoles (1O- moles)nm nanometres (iO meters)NSW natural sea waterNP-40 Nonidet P-40nt nucleotides0/N overnightORF open reading framePAA phosphoamino acid analysisPBS phosphate-buffered salinePCR polymerase chain reactionpfu plaque forming unitPM cAMP-dependent protein kinase inhibitor peptidePLB phospholysis bufferPMSF phenyl methylsulphonyl fluoridePNP p-nitrophenyl phosphatepoly(A)+ polyadenylatedPVDF polyvinylidene difluoride membraneRNA ribonucleic acidrpm revolutions per minuteRT room temperatureSBTI soybean trypsin inhibitorSDS-PAGE sodium dodecyl sulphate polymerase gel electrophoresissec secondSLS sodium lauryl sarcosinateT112 half-lifeTBS Tris-buffered salineTBST Tris-buffered saline with tweentet tetracyclineTIC total ion chromatogramtRNA transfer RNAii. micron (106 meters)microgramp.1 microlitrexxUTR untranslated regionUV ultravioletvol volumeV volumew weightXOM Xenopus oocyte mediaxxi2. AMINO ACIDSName Three letter abbreviation One letter symbol CharacteristicsAlanine Ala A non-polarArginine Arg R basicAsparagine Asn N polarAspartate Asp D acidicCysteine Cys C polarGlutamate Glu E acidicGlutamine Gln Q polarGlycine Gly G polarHistidine His H basicIsoleucine Be I non-polarLeucine Leu L non-polarLysine Lys K basicMethionine Met M non-polarPhenylalanine Phe F non-polarProline Pro P non-polarSerine Ser S polarThreonine Thr T polarTryptophan Trp W basic (weak)Tyrosine Tyr Y polar (weak)Valine Val V non-polar3. NUCLEOTIDESAdenine ACytosine CGuanine GThymine TDeoxyadenosine 5’-triphosphate dATPDeoxycytosine 5’-triphosphate dACPDeoxyguanine 5’-triphosphate dAGPDeoxythymidine 5’-triphosphate dTTPxxiiACKNOWLEDGEMENTSThere are many people whom I would like to thank for their assistance in the past 6years. I would like to thank my supervisor Dr. Steven Pelech for everything, especiallyfor giving me the opportunity to do my graduate studies in his laboratory. Gratefulappreciation to my committee, Drs. Gerry Krystal, Michael Gold and Roger Brownsey forall their guidance and support. I would like to thank my “non-official graduatesupervisors” Drs. Gabe Kalmar and Jasbinder Sanghera, for all the advice andencouragement during the roughest times. I would like to express my gratitude to Mr.Roman Babicki for the generous financial support during the last two years, which allowedme to continue and complete the most interesting part of my research.I would like express my appreciation to my many colleagues and collaborators withwhom I have had the privilege and pleasure to work with during the better part of the lastdecade. In particular, I would like to thank Dr. Chris Siow for helping start me down theroad of pim- 1 bacterial expression and Dr. Allen Delaney for helping with sequencealignments and the relatedness tree construction and for patience and humor exhibited whiletrying to assist a computer illiterate like myself. Much appreciation to Harry Paddon, LabGuy Extraordinaire for the coffee, the inspiration and for trying unsuccessfully to instill the“work smart” attitude. Thanks also to Dr. Diana Lefebvre for all the great advice and thesea star fertilization blots. I would like to mention the contributions of my excellentcollaborators, Dr. Lawrence Amankwa, Dr. Michael Affolter, Dr. Reudi Aebersold, Dr.Ian Clark-Lewis and Mr. Peter Borowski.In addition, I would like to acknowledge all my coworkers and friends for theirimmeasurable contribution to this project. I would like to thank the “Pelech women”Donna Morrison and Lorin Chariton for the chicken wings and stuff, David Charest andRuth Lanius for the bad jokes and the fashion commentary, and Marie-Therese Little for allthe tea. Thanks to George Gill for all the great advice and for attaching my ski bindings.Gratitude to Dr. Martina Metzler for the proofreading, Naam natchos and belays. Anhonorable mention to Dr. Kevin Leslie (patron saint of graduate students everywhere) forthe molecular biology assistance and to Anita Kaiser for all the assistance and attitude-adjusting over the years. For technical assistance I would like to gratefully acknowledgeGeorgia Tai, Dan Leung, Helen Merkins, Jaspal Girn, and Krista Lemke. And to Dr.Carol Reinish (wherever she is now), a big thanks for kicking me in the butt.Finally, I would like to express my appreciation to my family for their support andpatience, in particular to Dr. Jan “the spouse” Palaty, my parents Anne McNabb, BillMcNabb and Vladimir and Marta Palaty. Thanks to all my friends (whom I dare notmention by name at the risk of leaving someone out) for making the last six years totallyworth it.xdiiCHAPTER I.INTRODUCTIONPIM-1 EXPRESSION AND REGULATIONThe transformation from normal to neoplastic cells involves a sequential series ofgenetic events, often resulting in the activation or deactivation of genes concerned withthe regulation of the cell cycle or the control of differentiation. The overexpression of thepim-1 gene is thought to be one of the steps involved in the tumorogenic process. Despitelarge amounts of research devoted to pim-1, this protein kinase has been poorlycharacterized and the physiological function of Pim- 1 remains unknown. Resolving themany unanswered questions about the regulation, activity and function of Pim-1 couldprovide insight into lymphomagenesis and could contribute greater understanding of themysteries of cell cycle regulation. This work attempts to characterize the kinase activityand expression of Pim- 1 using bacterially-expressed enzyme, with the aim of achieving afuller understanding of the physiological function of the kinase, especially as applied tothe oocyte maturation system.1. VIRAL ORIGiNSi. DISCOVERY OF THE PIM-1 ONCOGENEMany early oncogenes, such as src, abi and erbB, were first identified as alteredversions of cellular genes that had been extricated and incorporated into the viral genome,their altered function or expression lending a survival advantage to the virus [Hunter,1987; Ramakrishnan and Rosenberg, 1989; Graf and Beug, 1983]. In contrast, the pim-1oncogene was first identified in Murine Leukemia Virus (MuLV) induced T cell and Bcell lymphomas not as a pirated viral component but as a common site of proviralintegration [Cuypers et al., 1984; Berns et al., 1 988a, 1 988b]. Hence the nomenclature ofpim-1; the iroviral Integration site of MuLV. Other oncogenes activated by proviralinsertion include the c-myc, c-myb, mt-i, int-2 and bmi-1 [reviewed by van Lohuizen andBerns 1990]. The pim-1 is also a site of integration in the Friend helper leukemia virus(F-MuLV) -induced murine erythroleukemias and feline leukemia virus (FeLV) -inducedlymphomas [Dreyfus et al., 1990; Tsatsanis et al., 1994]. As the MuLV contains notransforming genes of its own, it was inferred that the neoplastic manifestations of theintegration must be attributable to the gene located at the site of integration. This1assumption proved correct, as proviral integration into the pim-1 gene caused a 10-50%increase of pim-1 mRNA transcripts as compared to lymphomas not having integrationsin the pim-1 site [Cuypers et at., 1984; Selton et at., 1985]. As the gene was found tocontain little homology to any of the known oncogenes, pim-1 was acknowledged as anew oncogene [Cuypers et al., 1984].ii. ACTIVATION BY PROVIRAL INSERTIONThe exact mechanics of proviral insertion into the pim-1 gene are still unclear. TheMuLV prefers to integrate at specific, conserved nucleotide positions in the pim-] gene;this integration region is not homologous with the integration sites commonly used byother viruses [Cuypers et al., 1984]. Integrations are almost always found outside of theopen reading frame (ORF) in the 3’ untranslated region (UTR), suggesting that forsuccessful transformation, a full-length pim-1 protein is required [Selton et at., 1985;Haupt et at., 1991]. One study of viral integration into the pim-1 locus found that ofthirty-six lymphomas with an activated pim-1 gene, twenty-five had integrations in the 3’non-coding region of the pim-] gene, eight had integrations downstream of pim-1 andthree had integration upsteam of the pim-1 transcription unit. In some lymphomas, thelevels of pim-1 mRNA were elevated even though no viral integration was detectedwithin the gene, possibly because the virus integrated elsewhere [Selton et a!., 1985].Another study by Mally et a!. [1985] failed to detect the expression ofpim-1 in murine Tcell lymphomas. However, the level of detection required may not have been achieved.Proviral integration significantly enhances the expression of pim-] mRNA, byincreasing the stability of the pim-] mRNA transcripts. This is achieved by removal ofthe AUUUA motif, a ‘destabilization signal’ targeting the mRNA transcript for rapidturnover [Berns et at., 1988a]. Indeed, almost all viral integrations located in the 3’untranslated region are located upstream of the AUUUA sequence, leading to a truncatedmRNA transcript without this AUUUA sequence. The size of the pim-] mRNA in theMMLV-induced leukemias varies from 2.0 to 2.6 kB and depends on the site of viralintegration [Selton et al., 1985; Nagarajan et at., 1986]. Alternatively, the activation ofthe gene may be a function of the provirus itself; all inserted proviruses examined fromthe pim-1 area have duplicate or triplicate enhancer regions in the long terminal repeat(LTR), possibly causing an increase in the rate of pim-1 gene transcription. However, nomRNA was produced and no transforming activity was observed when NIH/3T3 cellsand rat embryo fibroblasts were transfected with genomic pim-] “provirally activated” by2the MuLV LTR, thus disputing the importance of the LTR in pim-] activation [Berns etat., 1988a1.The resulting Pim- 1 protein remains unchanged by the proviral insertion, not onlywas the size of the Pim- 1 consistent with that of uninfected cells, but there were nomutations or amino acid changes observed, even when upstream insertion occurred[Berns et at., 1988a]. Therefore, Pim-l becomes an oncoprotein by virtue ofoverexpression alone; proviral insertion contributes to cellular transformation by causingan increase in the amount of active Pim- 1 protein in the cell produced as a result of pim-1mRNA stabilization.2. CLONING THE PIM-1 GENEThe genomic and cDNA sequences of the murine pim-] gene were first reported bySelton et at. [1986]. The murine pim-1 gene is located on chromosome 17, close to theMHC complex, localized to the a-crystallin and tw- 12 markers [Berns et at., 1 988a], andencodes a 2.8 kilobase (kb) mRNA transcript [Domen et at., 1987]. The size of somaticmurine and human pim-1 mRNA transcripts have been reported to be similar [Meeker etat., 1987a].The human genomic pim-1 clone was first obtained from the human B cell leukemia380 cell line [Nagarajan et at., 1986] and from a human genomic library [Zakut-Houri etat., 1987; Reeves et at., 1990]. The 2.9 kB human pim-] cDNA was cloned by severalgroups from the human chronic myelogenous leukemia tumor cell line K562, whichexpresses high levels of pim-1 [Domen et al., 1987; Meeker et a!., 1987a; Padma andNagarajan, 1991; Zakut-Houri et at., 1987]. A single non-coding polymorphism at site1003 distinguishes the pim-1 sequence from the K562 cell line from that of the 380genomic clone [Meeker et at., 1987b].The sequence of pim-1 is highly conserved and homologues have been identified bySouthern analysis in such diverse vertibrate species as cat, hamster, human, chicken andmouse, but not Drosophita hydei [Cuypers et at., 1986; Tsatsanis et at., 1994]. Thehuman genomic clone has an overall nucleotide homology to the murine pim-1 gene of53%. Both human and mouse cDNA ORF are 939 nucleotides, with a predicted proteinproduct of 313 amino acids [Berns eta!., 1988a; Domen et at., 1987; Reeves et at., 1990].3The primary sequence of the coding regions of the human and mouse genes are 88% to90% homologous, while the amino acid sequences are 94% homologous [Berns et al.,1988a; Reeves et a!., 1990]. The amino terminal 250 amino acids are 98% homologous,with most of the amino acid substitutions clustered at the carboxy terminus. Both humanand mouse contain a single sequence for N-linked glycosylation, Asn-Gly-Thr, at Asn-82[Domen eta!., 1987; Berns et at., 1988a].Until recently, there was believed to be only one pim- 1 locus in mammaliangenomes. A second related gene called pim-2, located on the X chromosome wasrecently cloned from pim-1-deficient mice using complementation tagging [van der Lugtet al., 1995]. This gene is expressed in hematopoietic cells and in the brain, and theprotein encoded by this gene has many characteristics similar to murine pim-i includingupstream translation from CTG sites and similar substrate specificity. The pim-i andpim-2 genes were demonstrated to be functionally redundant. A pim pseudogene wasalso identified on chromosome 8. With the discovery of this second pim family member,an unrelated M-MuLV proviral integration region in chromosome 17, called pim-2, wasrenamed Tic-i [Breuer eta!., 1989a; van Lohuizen and Berns, 1990; Haupt eta!., 1991;van Lohuizen and Berns, 1990].3. ONCOGENIC POTENTIAL OF PIM-1i. STUDifiS IN MURINE LEUKEMOGENESISStudies of proviral integration in murine lymphomagenesis have been extensive andhave implicated the involvement of pim-i in cellular transformation. Pim- 1 is a site ofproviral insertion not only in murine T cell lymphomas [Cuypers et at., 1984; Dreyfus etat., 1990], but in erythroleukemias and in B cell lymphomas as well [Muscenski et at.,1988; Dreyfus et at., 1990; Verbeek et at., 1991]. Genetic examination of the sites of FMuLV integration in erythroid, lymphoid and myeloid leukemias demonstrated that 50%of F-MuLV were integrated in pim-i, c-myc orpvt-i, and that this integration was oftenassociated with rearragements in p53 in the same tumor [Dreyfus et a!., 1990]. The pim-isite, along with the c-myc gene, is one of the most common MuLV integration sites inmurine lymphoid tumors, with at least 75% of early murine T-cell lymphomas showingproviral integration in pim-i or near c-myc [Cuypers et at., 1984]. When Baib C micewere infected with Moloney Murine leukemia virus (M-MuLV), 31 of the 66 mice thatdeveloped lymphomas within a six month period had a provirus integrated in the pim-i4region [Selton et at., 1985]. In another study, changes were observed in the pim-1 regionin 23 of 93 murine lymphomas screened and over 50% of the T-cell lymphomasexamined demonstrated integration in this region [Cuypers et at., 1984]. In a studyinvolving inbred mouse strains with a high incidence of spontaneous lymphomas, only24% of the lymphomas examined contained any type of observable geneticrearrangements, and rearragements of pim-] (as well as fis-1, mtvi-1, mtvi-2) wereobserved only in T cell lymphomas [Muscenski et at., 1988]. Finally, studies withtransgenic animals overexpressing pim-1 have confirmed that this oncogene is involvedin murine lymphomagenesis [van Lohuizen et at., 1989; Moroy et at., 1991].ii. NO PROOF THAT PIM-1 IS A HUMAN ONCOGENEDespite the strong link between pim-1 overexpression and mouse lymphomagenesis,there is very little evidence to prove that pim-1 is a human oncogene. Although pim-1mRNA and protein are overexpressed in many human cell lines and leukemias, there isno evidence that viral integration in humans occurs in or near the pim-1 gene or thatchromosomal translocations involve the pim-1 locus [Meeker et at., 1 987b; Amson et at.,1989]. It seems that the levels of the pim-1 gene products are elevated in many humanleukemias by methods other than translocation or amplification [Amson et al., 1989].The human pim-1 gene was mapped by somatic cell hybrid analysis and in situhybridization to site 6p2l [von Lindern et at., 1989], more specifically to the 6pter-q 12segment [Cuypers et at., 1986]. The gene is located quite close to the human HLA(human leukocyte antigen) complex, but it is not known if the two regions are linked[Cuypers et at., 1986]. Some chromosomal translocations associated with lymphomasand leukemia do occur at 6p2l, but it is doubtful if any of the observed genetic defectsactually involve pim-1 [Amson et at., 1989]. Genetic defects involving 6p2i include 6pdeletions associated with T cell lymphoma [Meeker et at., 1987b] as well as a reciprocaltranslocation in t(6:9)(p21:q33) as the sole chromosomal anomally in some humanchronic myeloid (CML), acute myeloblastic (AML) [Nagaraj an et at., 1986] and acutenonlymphocytic leukemias (ANLL) [von Lindern et at., 1989]. A detailed study of thebreakpoints in AML found that the translocation did not involve pim-1, but rather intronsof the can gene, icb-9, on chromosome 9 and of the dek gene, icb-6, on chromosome 6,which were spliced to produce a transcript yielding a 165 kDa DEK-CAN fusion protein[von Lindern et at., 1992]. ANLL patients were examined specifically for theinvolvement of the pim-1 (6p2l) and c-abl (9q34) genes in the reciprocal (t6;9)(p23;q34)translocation. Although the expression of pim-1 mRNA was elevated in 2/3 of the ANLL5patients examined, the transcript size (2.7 kB) was unaltered, so the pim-1 gene remainedon chromosome 6 during the translocation and no chomosomal breakpoints were detectedwithin 165 kb of the pim-1 locus [von Lindern et at., 1989]. In addition, tumors from 51patients revealed trisomies and diploidies but no gene rearrangements or genetranslocations in chromosome 6 [Amson et at., 1989]. Coincidently, the humanerythroleukemia cell line K562, which over-expresses pim-], has a 6p21 rearrangementnot involving the pim-] gene [Nagarajan et at., 1986].iii. EXPRESSION OF PIM-l TN HUMAN CANCERS AND CELL LNESa. pim-1 mRNA expression in human cancers and cell linesSeveral detailed studies have been conducted to examine the expression patterns ofpim-] mRNA in various human tumor cells and cell lines. Examination of 38 human celllines revealed that the pim-1 mRNA is detected mainly in B cell and myeloid cell lines.Of 19 B-cell lines examined, most were positive for pim-1 except for the very immatureand the very mature [Meeker et at., 1987b]. These results agree with those of Nagarajanet at., [1986] who also examined pim-1 mRNA expression in various humanhematopoietic cell lines and found that the pim-1 transcripts were expressed at variouslevels in the different cell types. Myeloid cell lines (K562 and KG-i) expressed thehighest amounts of pim-]; the high frequency of positive pim-1 cDNA clones (0.1%-0.01%) in the K562 library confirmed that pim-1 mRNA is highly expressed in thiserythroleukemia cell line [Nagarajan et at., 1986; Meeker et at., 1 987b1. Despite the factthat pim-1 is causal in viral-induced murine T cell leukemia, pim-1 mRNA was notdetected in any of the human T cell lines examined in one study [Meeker et at., 1987b],and detected in only one of seven T cell lines tested in a second study [Nagaraj an et at.,1986]. A recent study examining pim-1 mRNA levels in primary bovine lymphocytesconfirmed that pim-] is constitutively expressed in primary B cells, and is expressed atonly low levels in T cells [Wingette et at., 1995]. In summary, pim-] mRNA ispreferentially expressed in a some myeloid leukemias and B-cell lymphomas, includingBurkitts.b. Pim-1 protein expression in human cancers and cell linesThe expression of Pim- 1 protein was examined in 70 malignancies of human origin,ranging from fresh tumor tissues to cell lines representing various stages of differentiationincluding myeloid, myelomonocyte and cells of B and T cell origin. Pim- 1 protein wasdetected/overexpressed in 30% of the samples tested but expression was not correlated6with any particular cell type or stage of cell differentiation and was not caused byrearrangement of amplification of the gene [Amson et a!., 1989]. Pim-1 wasoverexpressed at various stages in 24 AML patients, in a stage IV B-lineage ALL and acompletely immature ALL. In these patients, the protein levels of c-Myc were alsoexamined; although 90% of tumors from 51 patients overexpressed c-Myc and 30%overexpressed Pim- 1, the two genes did not display cooperativity with each other [Amsonet a!., 1989]. This lack of cooperativity observed in the human tumors contrasts sharplyto the cooperativity observed in pim-ilmyc double transgenic mice [Moroy et a!., 1991;Verbeek eta!., 1991].In the human cell lines tested, expression of Pim- 1 was variable between the differentcell types and differentiation stages [Amson et a!., 1989]. The highest expression was inmyeloid cells (K562 and KCL cells) and in SUDHL-6 cells of histiocytic origin. Pim-1protein was detected in most myeloid leukemia cell lines examined including HL6O,PLB985, KG1, K562, U937, WEHI3B and Ml cell lines, but not in EM2 [Lilly, 1989].The lowest expression was found in T cells and myelomonocytic cell lines [Amson et a!.,1989], a result in agreement with other studies [Padma and Nagarajan, 1991]. Relativelevels of pim-] mRNA and Pim-1 protein detected in the Daudi, SB, HL6O, K562 celllines were in agreement [Amson et a!., 1989; Meeker et at., 1987b], suggesting thatmRNA and protein levels of pim-1 may be correlated.In summary, expression of pim-] in human neoplasms does not involve proviralinsertions, genetic rearrangements, translocations or truncations. Pim- 1 protein appearsto be overexpressed primarily in myeloid and B-cell tumors and cell lines and expressiondoes not seem cooperative with the expression of c-myc. Although pim-1 unquestionablyacts as an oncogene in the mouse, it is uncertain ifpim-1 has a similar role in humans.74. PIM-1 GENE EXPRESSIONi. SELECTIVE mRNA EXPRESSION PATTERNSPim-] mRNA has very selective patterns of expression, with levels varyingconsiderably between stages of development, tissues and cell lines [Sorrentino et at.,1988; Amson et at., 1989; Wingette et at., 1995]. Between some cell lines, the levels ofthe mRNA transcript vary from 0-0.2% of total mRNA [Meeker et al., 1987a; ZakutHouri et at., 1987; Nagarajan and Narayana, 1993]. Variation in pim-1 expression levelsand stability have been demonstrated between different tissue types in the same animal,possibly reflecting a difference in Pim- 1 function between the various tissues [Wingetteet at., 1991, 1995]. Although it is unusual for such a highly conserved gene to beexpressed differently in various species, the patterns of p im -1 expression varyconsiderably between mouse and human [Meeker et al., 1 987b]. It is unclear if variationsin the amounts of pim-1 mRNA detected are the result of different experimental methods,expression levels or a reflection of post-transcriptional regulation.Murine pim-] mRNA is detected in both B and T cells as well as in manyhematopoietic cell lines [Berns et at., 1988a]. During murine embryonic development,the expression of the gene follows migration of the hematopoietic cells, with pim-1mRNA expression resticted to liver, thymus and spleen. High levels of pim-1 mRNA arealso detected in 10-12 day-old placenta [Selton et at., 1985]. In normal newborn andadult mice, pim-1 mRNA is detected predominantly in the spleen and thymus, with minoramounts in the liver and none in the kidney, lung, heart, ovaries, testes or brain [Selton etal., 1985]. The levels of pim-] mRNA are higher in the embryonic tissues than in thecorresponding adult tissues [Berns et at., 1988a1. Maximum pim-1 mRNA levels weredetected in the thymus at birth, in the liver at 16-19 days of gestation, and pim-1 mRNAlevels increase up to 14 days after birth in the spleen [Selton et al., 1985].ii. ALTERNATE mRNA TRANSCRIPTS & DIFFERENT HALF-LIVESThe 2.8 kb pim-1 mRNA transcript detected in somatic cells is inducible and short-lived, and can be upregulated by mitogen and growth factor stimulation [Dautry et at.,1988; Wingette et al., 1991; Lilly et al., 1992; Wingette et at., 1995]. Growth factorsincrease the stability of the transcripts, contributing in part to increases in the levels ofpim-1 mRNA [Wingette et at., 1991; Yip-Schneider et al., 1995].8The stability of pim-1 mRNA transcripts varies substantially between different celllines, species and laboratories. The half-life (T112) of pim-] was measured in severalunstimulated human cell lines and ranged from 47 mm to over 3 hours [Meeker et a!.,1990; Wingette eta!., 1995]. Stimulation of bovine lymph node lymphocytes and bovinePBMC lymphocytes with Con A (a T cell mitogen) and phorbol ester (PMA) caused a3.5-fold increase in pim-1 mRNA in 4 hours (T112 > 80 mm), which had decreased to 50%of peak levels by 17 hours post-stimulation (T112 35 mm) [Wingette et at., 19911. Theinduction of pim-1 mRNA in primary lymphocytes stimulated with PMA and ionomycin,peaked in 4 hours (T112 40 mm) and dropped by 17 hours (T112 120 mm), indicatingthat more than just mRNA stablilization was involved in this increase in pim-1 mRNA.For unknown reasons, similar cells from different species yield different mRNAstabilities; rat lymphocyte pim-1 mRNA transcripts have a longer half life (T112 140mm) than those found in activated bovine lymphocytes (T112 80 mm) [Wingette et at.,19911.There are differences in the size and half-life of pim-1 mRNA transcripts in varioustissues from the same animal. Somatic tissues of rats and mice contain a 2.8 kb pim-1mRNA transcript, while a shorter 2.4 kb transcript has been observed in the testes[Sorrentino et a!., 1988; Wingette et a!., 1992]. This alternate 2.4 kb fragment is not aproduct of a related gene or a different isoform ofpim-1 generated by alternative splicing,but results from the removal of the A/U rich destabilizing regulatory region in the Yuntranslated region of the gene and polyadenylation at an alternate site, nucleotidenumber 1302 [Wingette et a!., 1992]. The removal of this same A/U rich region isthought to occur during proviral integration. Consequently, the 2.4 kb mRNA message ingerm cells is more stable (T112 > 3.5) than the longer 2.8 kb message produced inlymphoid cells (T112 -440 mm) [Wingette eta!., 1991, 1992].This 2.4 kB form is specifically expressed in the haploid postmeiotic early spermatidsof mature adult mouse testes, and seems to correlate with sexual maturation and thedevelopment of spermatids [Sorrentino et at., 1988]. This transcript was not observed inthe testes of newborn mice, ovaries of mature female mice, nor in spermatogonia,suggesting that the pim-1 plays a developmental role in male gametes [Sorrentino et at.,1988].One functional justification for the short, stable germ-cell specific transcript could bethat the increased stability of the transcript could allow it to survive through the9translational delay that occurs in post-meiotic germ cells as the early spermatids matureinto differentiated spermatozoa [Sorrentino et at., 1988]. There are several other genesthat have shorter testes-specific transcripts, including the mos [Sorrentino et at., 1988],abi, t-fer, calmodulin, cAMP-dependent protein kinase regulatory and catalytic subunits,possibly due to the existence of a unique germ-cell specific polyadenylation system[Sorrentino et at., 1988; Wingette et al., 1992]. With abt, a more stable mRNA transcriptoccurs in the mouse spermatids than in somatic cells [Wingette et al., 1992]. The PKAR1c was examined but unlike pim-1, no difference in stability was found between theshort, testes-specific transcript and the longer somatic cell-specific transcript. However,the A/U rich region of PKA-R1c is not as extensive as that of pim-1 [Wingette et al.,1992].iii. REGULATION OF PIM-1 GENE EXPRESSIONPim-1 mRNA expression is tightly regulated and occurs at many different levels, thespecifics of which remain hypothetical. The promoter region of pim-] was found to havemany characteristics of a constitutively expressed housekeeping promoter [Meeker et al.,1990; Nagarajan and Narayana, 1993], yet expression is selective and highly variablebetween different cell types and lines and during stimulation by growth factors [Dautry etat., 1988; Meeker et at., 1990; Wingette et al., 1991; Lilly et al., 1992; Wingette et at.,1995]. Although elements located outside of the promoter region are probablyresponsible for the differences in mRNA transcription, the pim-] promoter region hasundergone extensive examination for clues about how the gene may be physiologicallyregulated.Extensive examination of the 1.7 kb promoter region has revealed that pim-1 hasmany features of a constitutively expressed housekeeping gene including Spi bindingsites and a lack of TATA and CAAT boxes [Meeker et al., 1987b; Meeker et at., 1990;Reeves et at., 1990; Nagarajan and Narayana, 1993]. The human promoter region alsohas AP-1 sites and NF-A2 and NF-kB binding sites that may be involved intranscriptional regulation [Meeker et at., 1987b; Reeves et at., 1990]. The transcriptioninitiation site of human pim-] was identified by Si nuclease protection, from which itwas determined that only one transcription initiation site is viable [Meeker et at., 1 987b].The human pim-1 promoter region is more than 80% identical with the mouse pim-]promoter region, and the -ito -876 region is extremely G+C rich (71%) [Reeves et al.,1990; Meeker et at., 1990; Wingette et al., 1992].10Recently an interferon gamma (IFNy) responsive element was identified in the 5’region of Pim-1 [Yip-Schneider et at., 1995]. This region, called PMGAS, contained aStat (signal transducers and activators of transcription) binding site IICCCAGAA thatbound Stat lc, a 9lkDa subunit of the interferon-stimulated gene factor 3 (ISGF3). ThisStat binding site was functional and is capable of confering IFNy responsiveness ontoheterologous promoters. As many growth factors utilize the JAK-Stat pathway, thismight be the method by which this gene is stimulated in growth factor-activated cells.Two major functional regions of the human pim-1 promoter were defined by deletionmutations; a proximal element at -104 to -1 and a distal element at -429 to -336 [Meekeret at., 1990]. DNase I protection assays identified the specific binding sites for SP 1 andAP2 proteins within these elements, four of five Spi boxes [(G/T) (0/A) 0 G 0 C G(GIT) (0/A) (0/A) (C/T)] are conserved between mouse and human [Meeker et at.,1990].The pim-] promoter also contains a lymphoid-specific octamer motif begining at-248, similar to one that is found in the promoter or enhancer regions of IgG genes thatbinds lymphoid-specific transacting transcriptional regulating proteins [Selton et at.,1986; Meeker et at., 1987b; Bems et at., 1988a]. This region is normally found 70 kbupstream of the RNA cap site, but is located in a slightly different region in pim-1 [Seltonet at., 1986]. In the human, this motif (ATGCAGAT) is similar (7/8) to that of the IgGoctomer motif, while in the mouse gene, this octomer (ATGCAAAT) is identical. Therole of this octamer in pim-1 expression is unclear.Examination of the pim-1 promoter has revealed little information about theregulation of the oncogene or why there is such a tissue-specific expression of the pim-1.Several theories have emerged as to how pim-] is regulated including transcriptionalattenuation, transcription control by the rate of transcription and tissue-specifictranscriptional control elements and selective degradation of RNA by protein factors. Alltheories propose that unknown control factors, other than the sequence itself, are requiredin the regulation of pim-1 mRNA, and are discussed below.11a. Transcriptional attenuation by DNA secondary structureTranscriptional attenuation/repression is one theory to explain the differences in thesteady state levels of pim- 1 mRNA; formation of stem-loop structures or triple helixbinding in genomic DNA blocks transciptional machinery in the first exon-intron, leadingto a “transcriptional pause” [Nagarajan and Narayana, 1993; Svinarchuk et at., 1994].Dyad symmetry elements which could possibly form stem-loop structures were foundwithin the first 488 base pairs of the genomic coding sequence of pim-1, but neither thelocation nor the identity of these sequences was specified [Nagarajan and Narayana,1993]. This block might possibly be prevented or overcome by the presence of a mysteryfactor that could interact with the DNA directly, overcoming this “transcriptional pause”,allowing mRNA transcription and the subsequent protein translation.A novel potential method of eukaryotic gene regulation involving triple-helix bindinghas recently been suggested for pim-1 [Svinarchuk et at., 1994]. A polypurine oligonucleotide (GGGAGGGGGAGG) based murine pim-1 promoter residues -182 to -194,binds to a homopyrimidine duplex in the murine pim-1 promoter (-358 to -370 and -425to -437) forming a stable triple-helix complex. This complex is stable to 65°C, bindsirreversibly at 37°C and the interaction is dependent on the sequence in the pim-]promoter region; a one base pair substitution abrogates oligonucleotide binding andtriple-helix formation. A suggested method of gene regulation involves a regulatorymolecule interacting with the DNA on the basis of triple-helix recongnition [Svinarchuket at., 1994]. Interestingly, a yet unidentified transcription factor, PPF-348 (Pim-1promoter factor 348) binds to residues -348 to -374 [Meeker et at., 1990] which overlapswith the homopyrimidine domain. If regions of the pim-1 promoter are able to trimerizeinto a stable complex, this PPF-348 may be the transcription factor that interacts with thisregion and regulates gene expression.Recently another group has reported that elements within the pim-1 5’ untranslatedregion (UTR) may be responsible for translational repression, and that this repressionmay be relieved in vivo by a factor-dependent mechanism [Hoover, et at., 1995].b. Pim-1 regulation by mRNA destabilizationMessenger RNA stability is involved in the modulation and regulation of pim-1mRNA levels in normal lymphocytes and germ cells by protein synthesis-dependent posttranscriptional mRNA degradation [Wingette et at., 1991]. The 1.3 kb 3’ untranslatedregion of both the murine and human pim-] genes contain conserved UUAUUUAUU12motifs, which are believed to mediate mRNA instability [Lagnado et at., 1994; Zubiagaet al., 1995]. The removal of this region occurs during pim-1 proviral integration and inthe production of stable germ-cell specific pim-] transcripts [Domen et at., 1987; ZakutHouri et at., 1987; Wingette et at., 1991, 1992].The selective and rapid degradation of mRNA transcripts containing AU-richelements (AREs) are mediated by protein factors and is protein synthesis-dependent.Actinomycin D (an inhibitor of de novo mRNA synthesis) inhibits mRNA degradation,and cycloheximide increases the stability of pim-1 transcripts about three-fold [Wingetteet a!., 1991; Meeker et at., 1990]. ARE-containing mRNA sequences are stabilized bygrowth factors, phorbol esters, calcium ionophores, mitogenic antibodies, proteinsynthesis inhibitors and Con A, all of which may modulate the action of mRNAdegrading proteins [reviewed in Reeves and Magnuson, 1990].c. Pim-1 regulation by differing rates of transcriptionDifferences in the rates of pim- 1 transcription may be responsible for differences inthe levels of pim-1 mRNA observed. Levels of pim-] mRNA expression were studied inthree cell lines that exhibited levels of expression that were typical for unstimulated cellsof the myeloid (K562), B-cell (Daudi) and T-cell (Jurkat) lineages [Meeker et at. ,1990].pim-1 mRNA was most highly expressed in K562 cells, which had 20-fold greaterexpression than in Daudi and 50-fold greater expression than in Jurkat cells. Thevariations in the amount of mRNA were primarily due to differences in the rate oftranscription of the gene, as the rates of transcription were similar to the amounts ofmRNA. As the pim-] promoters were compared between the different cell lines and werefound to be identical, some tissue or cell specific mechanism must control the rate oftranscription in these cells [Meeker et at., 1990]. The half-life of the pim-1 mRNA variedbetween these different cell lines, with 47 minutes in the K562 cells, 71 minutes for theDaudi cells and 35 minutes for Jurkat. In this case, the half-life of the mRNA transcriptswere not related to the amounts of pim-1 mRNA detected in the cells [Meeker et at.,1990]. A second study showed that the amount of pim-1 mRNA transcript in a cell is notsoley related to the stability of the transcript, and suggested that other processes such asincreased transcription, mRNA processing or transport from the nucleus are involved[Wingette et a!., 1995]. Indeed, Yip-Schneider et at., [1995] demonstrated that anincrease in the rate of transcription of the pim-1 gene was partially responsible forelevations in the pim-1 levels in the cell in response to IFN’y13The whole question of regulation of pim-1 has been complicated by the recent claimthat the Pim-1 protein and pim-1 mRNA are not necessarily induced under the sameconditions [Hoover, et at., 1995], a finding that is in direct conflict with other groups[Yip-Schneider et al., 1995]. Although different levels of transcriptional and translationalcontrol over pim-1 expression exist, it is not known whether they operate independently.5. PIM-l: THE PROTEINi. EXPRESSION IN NORMAL TISSUEIn consideration of the wide disparity between levels of pim-] mRNA expression, it ishardly surprising that Pim- 1 protein levels vary between different tissues and cell types.Amson et at., measured Pim- 1 protein expression during fetal development and inhematopoietic malignancies [Amson et at., 1989]. Very high protein expression wasdetected in the liver and spleen during human fetal hematopoiesis, and low expressionwas detected in the kidney. In the fetal liver, expression was limited to the typical roundcells, which are hematopoietic progenitors, but not in surrounding tissue [Amson et at.,1989]. In human adults, the protein was expressed only slightly in circulatinggranulocytes and in bone marrow [Amson et at., 1989]. These results imply that Pim-lmay play a role during embryonal and fetal hematopoiesis, but does not reveal anythingfurther about the actual function of the kinase.The protein product of the pim-] gene is localized to the cytoplasm in pim-1 over-expressing (K562, 679thy) and low expressing (NIH3T3) cell lines and in primarylymphocytes of normal and Pim- 1 transgenic mice [Telerman et at., 1988; Saris et at.,1991]. The kinase does not have any hydrophobic stretches nor signal sequencescharacteristic of a membrane directed protein [Domen et at., 1987]. This cytoplasmiclocalization indicates that the kinase is potentially a component of a signal transductionpathway.ii. THE PRODUCT OF THE PIM-1 GENE IS A KJNASEWhen first sequenced, the protein products of both the human and murine pim-1genes were found to be structurally related to kinases, possessing all of the conservedresidues and domains of protein kinases [Hanks et at., 1989]. All studies examining theactivity of endogenous and expressed Pim-1 have confirmed that the protein has autokinase activity. The presence of a tyrosine at residue 198 corresponds to Tyr-416 in Src,14which is a conserved autophosphorylation site in all tyrosine kinases [Cooper andMacAuley, 1988; Telerman et al., 1988; Hanks et at., 1989], led to the initialclassification or Pim- 1 as a tyrosine kinase [Meeker et at., 1 987a; Telerman et at., 1988].An early study with endogenous Pim- 1 protein immunoprecipitated from K562 cells andtranslated in vitro, confirmed this belief [Telerman et at., 1988]. Not only did theimmunoprecipitated Pim-1 autophosphorylate on tyrosine residues, but this tyrosinespecific activity was recovered even after the immunoprecipitated Pim- 1 was boiled,electroluted and removed from an SDS-PAGE gel, suggesting that this activity was duespecifically to the Pim- 1 and not a contaminating kinase. Protein phosphotransferaseactivity towards serine and threonine with the Pim- 1 was not detected [Telerman et at.,1988].After extensive investigation by many different groups, Pim- 1 was reclassified as aserine/threonine protein kinase. Both the mouse and human pim-1 cDNA undergo invitro transcription-coupled translation in the rabbit reticulocyte system producing anactive kinase with serine-specific autophosphorylating activity [Saris et at., 1991; Padmaand Nagarajan, 1991]. Human and murine Pim-1 expressed in bacteria [Saris et at.,1991; Hoover et at., 1991; Friedmann et at., 1992] and in COS cells, andimmunoprecipitated from the 679thy and K562 cell lines [Saris et al., 1991], was foundto have autophosphorylating activity only on serine and/or threonine residues and haveonly serine/threonine phosphotransferase activity towards exogenous substrates.Phosphotyrosine-containing proteins were not detected on Western blots of E. coti cellsexpressing wild-type or mutated Pim- 1 [Saris et at., 19911. In contrast to the earlierstudy, Pim-1 was found to be thermolabile [Hoover et al., 1991]. It would seem fromthese studies that Pim-1 possesses serine/threonine but not tyrosine kinase activity.Among the groups claiming that Pim- 1 was not a tyrosine kinase, some conflict stillexisted; some groups claimed that the Pim-1 was strictly a serine kinase [Saris et at.,1991; Padma and Nagarajan, 1991], while others detected kinase activity towardsthreonine residues as well [Hoover et at., 1991; Friedmann et at., 1992]. The proteinproduct of the recently discovered pim-2 gene was also found to possess serineautocatalytic activity [van der Lugt et at., 1995].15iii. EXPRESSION OF THE PIM-1 PROTEIN PRODUCTExamination of the protein products of human and murine pim-1 genes has revealedthat several forms of Pim- 1 protein are produced in various systems as a result ofalternate initiation of translation. These multiple protein forms result from a singlemRNA transcript [Nagarajan et at., 1986].Human Pim-1 protein often appears as a tight 33-34 kDa doublet in Western blots andimmunoprecipitations of various cell lines [Telerman et at., 1988; Saris et at, 1991; Lilly,et at., 1992]. In vitro translation of the human Pim- 1 in the rabbit reticulocyte systemyields a doublet of 32 and 33 (34 kDa) [Domen et at., 1987; Padma and Nagarajan, 1991]in contrast to a predicted protein product of 35.6 kDa [Reeves et at., 1990]. In addition, asmear is often detected at about 50 kDa on SDS-PAGE gels [Domen et at., 1987].Although initially thought to be the results of degradation or phosphorylation, mutantstudies with murine Pim- 1 have found that the 34-35 doublet is a result of independentinitiation from an alternate initiation site, CUG, 4 codons upstream of the normal startsite [Saris et at., 1991]. Some groups also detected two minor, 28-29 kDa bands, in-frame products of the translation, which immunoprecipitate with the Pim- 1 C-terminalantibodies [Domen et at., 1987]. These proteins are speculated to arise from internalinitiation events, possibly at the Met-88 residue conserved in mouse, rat and human Pim1 [Domen et at., 1987; Wingette et al., 1992].Expression of murine Pim- 1 differs from the human in its production of an additionalprotein product. Murine Pim-l was expressed as a 33 to 34 kDa doublet [Domen et at.,1987] and a 41 kDa [van Lohuizen et at., 1989] (or 44 kDa [Saris et at, 1991]) protein inequimolar amounts in all murine tissues and cell lines tested [Saris et at, 19911. Theseproteins were also coexpressed in the in vitro rabbit reticulocyte translation system,bacterial and COS expression systems, and were detected by immunoprecipitation from amurine thymoma cell line, 679thy (from a transgenic overexpressing Ep-pim-1) [Saris etat, 1991]. The larger sized protein was due to an in-frame amino terminal extension,resulting from an alternate, albeit inefficient, translation event at an upstream CUG site[Saris et at, 19911. This upstream initiation is not conserved between mice and humans,but was observed during translation of the pim-2 gene product [van der Lugt et at., 1995].Overexpression of the Pim- 1 did not change the ratio or the cellular localization of thetwo proteins. However, the in vivo asssociation state of the cell-free synthesized kinasevaries, with the p34 existing in a monomeric state and the p44 forming either a dimer or acomplex with other proteins [Saris et at, 1991].16iv. PROTEIN HALF-LIFE OF PIM- 1In K562 cells, the Pim-l protein has a very short half-life implying that it istightly regulated. The half life of the Pim- 1 proteins immunoprecipitated from murine679thy cells were determined to be 1 hour for 44 kDa protein and 10 minutes for the 34kDa protein [Saris et al., 1991]. This is in agreement with another group that found thehalf-life of human Pim-1 to be 10 minutes [Amson et al., 1989].17STUDIES INTO THE FUNCTION OF PIM-1Pim- 1 is highly conserved between all species examined and is tightly regulated pretranscriptionally, post-transcriptionally, and post-translationally. The fact that pim-1 is aproto-oncogene implies that it serves an important function in the cell. Expressionpatterns of pim- 1 mRNA and protein implicate the kinase in hematopoiesis and malegerm cell development. However, differences in expression levels between differentsubtypes, developmental stages and species have made it difficult to assign a specific roleto the kinase based on expression patterns alone. Studies to elucidate the function of pim1 are divided into three major categories: substrate studies to identify downstream targetsof the kinase, studies with growth factors and receptors to understand if Pim- 1 isregulated in a specific signal transduction pathway, and studies with transgenic animals toanalyze the in vivo effects ofpim-1 overexpression or knockout.6. PIM-1 SUBSTRATE SPECIFICITYSubstrate specificity studies have been performed to a limited extent with endogenousPim-1 [Saris et al., 19911 and extensively with bacterially [Hoover et al., 1991;Friedmann et al., 1992] and in vitro expressed Pim-1 [Padma and Nagarajan, 1991].Studies have defined optimal conditions for kinase activity, and have tested a widevariety of proteins as exogenous substrates of Pim- 1. As will be described below,preliminary estimations of a substrate concensus sequence for Pim- 1 were determinedusing peptide and protein substrates.Studies using endogenous Pim- 1 immunoprecipitated from K562 cells found that theautophosphorylating activity of the kinase was constant in the range pH 5.5-7.5. Theoptimal divalent cations for Pim- 1 peptide assays, as defined by studies with bacteriallyexpressed Pim-1, were 10 mM MgCl2 or 5mM MnC12 [Friedmann et al., 1992].Autophosphorylation activity was inhibited by higher cation concentrations as well as bythe presence of zinc and sodium in concentrations over 50 mM [Saris et al. 1991;Friedmann et al., 1992]. Thrombin-cleaved Pim-1 and GST-Pim-1 phosphorylated thesame spectrum of substrate proteins and peptides in vitro but GST-Pim- 1 was moreefficient than thrombin-cleaved Pim-1 [Hoover et al., 1991].18Many proteins were tested and found to be substrates of Pim- 1 including histone Hi(on serine and threonine), enoiase (on serine and threonine) [Hoover et at., 1991;Friedmann et at., 1992], histone 2B (on serine), salmone protamine [Saris et at., 1991]which contains no tyrosine residues, and lactate dehydrogenase, which was only slightlyphosphorylated [Hoover et at., 1991]. Histone Hi was found to be the best substrate ofmammalian Pim-1, with a Km of 7 jiM under optimal conditions [Friedmann et at., 1992].All exogenous substrates were found to be phosphorylated on serine and/or threonineresidues only. Proteins tested but not phosphorylated by Pim-1 were casein, BSA,poly(GLU:TYR 4:1), GST [Hoover et at., 199 1], angiotensin I peptide (DRVYTHPFHL),Src peptide (RRLIEDAEYAARG) and c-Myc [Saris et at., 1991].V8 proteolysis of histone Hi phosphorylated by Pim- 1, followed by HPLC andsequencing revealed two sequences in histone Hi that are similar to the Kemptide:TAPAETAAPAP and AKPKA. One corresponded to the N-terminal fragment containinga cAMP-dependent protein kinase phosphorylation site, LRRASGP, which was referredto as “HiPEP” [Friedmann et at., 1992]. This H1PEP sequence is conserved in manyhistone Hi proteins and may be the site of Pim-1 phosphorylation [Friedmann et at.,1992].Peptide substrates have been used to define the substrate concensus sequence of Pim1. Pim-1 exhibited strong preference for peptide “HlPEP” (KRRASGP) six-fold overKemptide (LRRASLG), the optimal substrate of cAMP dependent protein kinase.H1PEP and Kemptide both exhibited Km values of 0.4 mM with Pim-1. Both N-terminalarginine residues (-2,-3) in the peptides were equally important for recognition, asreplacement of these residues with alanine reduced the efficiency of the peptides assubstrates to the same extent. Basic amino acids, in this case arginine, were essential onthe N-terminal side, while basic residues on the carboxy terminus of the phosphoacceptorsite were inhibitory, and interfered with substrate recognition [Friedmann et at., 1992].Saris suggested that in addition to a preference for serine surrounded by basic residues(arginine), the Pim- 1 also requires a proline residue [Saris et at., 19911. Under optimizedconditions, the consensus site was deduced to be (R/K)3-X-SIT-X’, where X’ cannot bearginine, lysine, or a large hydrophobic residue [Friedmann et at., 1992].Pim-1 showed a higher Km value for Kemptide than cAMP-dependent protein kinase,possibly due to the Pim- 1 having less hydrophobic residues at the sites that interactdirectly with the substrate [Friedmann et at., 1992]. Examination of the crystal structure19of the cAMP-dependent protein kinase revealed that the main residues interacting withthe substrate were Leu-198, Pro-202 and Leu-205 [Knighton et at., 1991]. In contrast,Pim- 1 has Phe-20 1, Arg-205 and Ser-209 at these same sites, suggesting that the lesshydrophobic residues at the + 1 site of the peptide are preferred for Pim- 1 recognition[Friedmann et at., 1992].The vague definition of a Pim- 1 substrate consensus sequence has had very fewfunctional implications. Although Pim-1 phosphorylates several proteins and peptideswith relatively high affinity in vitro, it is uncertain if these phosphorylations reveal anyphysiological relationship between the kinase and these substrates in vivo. Further workis required before the physiological substrates of Pim- 1 can be identified.7. UPSTREAM REGULATORS: GROWTH FACTORS AND MITOGENSi. MITOGEN AND GROWTH FACTOR STIMULATION OF PIM-1pim-] mRNA expresionis increased in response to growth factors and mitogens. Pim1 mRNA and protein expression is induced by cytokines associated with thehematopoietic (HP) receptor superfamily, such as GM-CSF, G-CSF, IL-2, IL-3, IL-5, IL6, IL-7 but is not limited to these types, as it is also stimulated with ConA, PMA, IFNy,Steel factor (SF) when costimulated with other mitogens and in response to TCR cross-linking [Dautry et at., 1988; Wingette et at. 1991; Lilly et at., 1992; Saito et at., 1992;Domen et at., 1993b, 1993c; Wingette et at. 1995; Yip-Schneider et at., 1995]. Pim-1expression is dependent on the mitogen used with effects being proportional to theproliferative effects of the cytokine [Wingette et at. 1991; Lilly et at., 1992]. Expressionof Pim- 1 in response to myeloid cytokines is varied, and depends on the nature of thegrowth factor and the response phenotype of the cell examined. However, it seems to begenerally associated with a proliferative response to GM-CSF and similar cytokines[Lilly, et at., 1992; Polotskaya et at., 1993]. Pim-1 does not seem to be induced inresponse to SF alone [Domen et at., 1993c], TNFc [Saito et at., 1992] or by receptortyrosine kinases [Polotskaya et at., 1993]. Many of the cytokines, such as IL-3, IL-S andGM-CSF, have similar functions in common target cells such as eosinophils, implyingthat Pim- 1 may have a similar function in response to each of these cytokines [Kinoshitaetat., 1995].20Many growth factors were found to synergistically upregulate the expression of Pim- 1when combined. For example, the induction of Pim- 1 mRNA and protein is muchstronger when stimulated with both IL-7 and SF, JFNy and SF, PMA and ConA, or PMAand ionomycin than when each of these growth factors is used separately [Domen et at.,1993b; Yip-Schneider et at., 1995; Wingette et at., 1995]ii. EFFECTS OF MITOGEN STIMULATION ON PIM-1In unstimulated, factor-dependent myeloid or lymphoid cells, there is normally a verysmall amount of pim-1 mRNA present and no protein. Mitogen stimulation causes arapid accumulation of pim-] mRNA within an hour, which is followed by an increase inprotein levels. For example, stimulation of two IL-2-dependent lymphoid (CTLL-2 andB6. 1) and one IL-3 dependent myeloid (FDC-P2) cell line, with ll.,-2 and ]L-3 caused andincreased accumulation of pim-] mRNA. All cell lines showed similar patterns of pim-]mRNA accumulation, with a small amount of pim- 1 mRNA present in serum-starvedcells which increased by 40 minutes, peaked between 2-8 hours and declined after 10hours post-stimulation [Dautry et at,, 1988]. Upregulation of pim-1 mRNA in responseto ll-2 and IL-3 occurred at the transcriptional level in the absence of protein synthesis[Dautry et at,, 1988]. pim-1 mRNA and protein levels are sustained for the duration ofgrowth factor exposure [Lilly, et at., 1992]. This time course of pim-1 mRNAaccumulation is similar to that of c-myc [Dautry et at., 19881.As discussed in a previous section, pim-1 mRNA is regulated post-transcriptionally inmitogen-stimulated lymphoid cells by a protein synthesis-dependent mRNA degradation[Wingette et at., 1991]. Changes in the levels ofpim-1 mRNA are controlled in part bychanges in pim-1 mRNA stability, with mitogens mediating the stabilization of themRNA transcript. mRNA stabilization leads to an accumulation of pim-1 transcript inthe cell which results in a greater amount of protein product translated. Other methods,such as an increase in the rate of transcription, contribute to an accumulation of mRNA inaddition to mRNA stabilization.iii. PIM-1 MAY PLAY A ROLE IN T CELL RECEPTOR SIGNALINGPim- 1 is likely to be active in T cells, as increased pim-1 expression is associated withmurine T cell lymphomas [Meeker et at., 1987b], the kinase is expressed in the thymusand also in ConA-stimulated spleen cells, a model system that is representative ofproliferating peripheral T cells [Mally et at., 1985]. Expression of pim-1 mRNA can beinduced in both ToJB (five-fold) and Ty/a cells (eight-fold) by PMA and ionomycin21[Wingette et at., 1995]. Pim-1 mRNA expression can also be induced by T cell receptor(TCR) cross-linking with anti-CD3 antibodies [Wingette and Magnuson, 1995],confirming that enhanced expression of the pim-1 gene is an early and transient event inactivation of normal lymphocytes [Wingette eta!., 1991].iv. INDUCTION OF PIM-l EXPRESSION BY PKA AND PKCAlthough IL-3 can produce an increase in Pim-l protein levels in 2-4 hours, the PKCactivator bryostatin, which can substitute for IL-3 as a mitogen in MAC-il or U937 cells,did not lead to any increase in Pim-l in these cells [Lilly et a!., 1992]. PMA did notcause prolonged changes in c-myc, pim-1 or cyclin D2 mRNA [Polotskaya et at., 1993],and failed to induce protein production in MO7E cells [M. Lilly, personalcommunication]. TPA treatment to stimulate PKC, and forskolin or dibutyryl-cAMPtreatment to stimulate PKA did not have any effect on pim-1 mRNA levels, indicatingthat PKC and PKA are not involved in pim-1 induction [Dautry et at., 1988]. Thisimplies that pim-1 is not stimulated through the PKC pathway, and even though bindingof GM-CSF to the receptor causes PKC translocation, this may not be how most of thesignal is transduced.By contrast, another group found that PMA synergized with ConA and withionomycin to induce a strong pim-1 mRNA induction in primary lymphocytes [Wingetteet a!., 1991, 1995]. In this system, pim-1 gene expression seemed to be upregulatedtranscriptionally and post-transcriptionally following activation of PKC. Not only didPMA induce rapid pim-1 expression, but PKC inhibitors H-7 and staurosporine blockedpim-1 expression. In stimulated T cells, pim-1 mRNA was not induced after elevation ofintracellular free Ca2 [Wingette and Magnuson, 1995] and was induced by PMA alone,suggesting that the activation of Pim- 1 involved a PKC signalling pathway [Wingette eta!., 1995]v. PIM-1 IS INDUCED IN RESPONSE TO SIGNALING THROUGHRECEPTORS OF THE GM-CSF FAMThYAlthough pim-1 expression is upregulated in response to a large number of growthfactors and mitogens, studies with the GM-CSF and IL-3 pathways have been mostproductive [Lilly, 1989; Polotskaya et a!., 19931. Recent studies with mutant GM-CSFBreceptor subunits have delineated specific signalling pathways and have identifiedpotential upstream regulators of Pim-1 [Polotskaya eta!., 1993; Sato eta!., 1993].22a) Induction of Pim- 1 by GM-CSFMany different cell types have been tested and found to exhibit induction of pim-1 inresponse to GM-CSF. The cell growth response to GM-CSF is correlated with prolongedincreases in cell levels of c-myc, pim-1 and cyclin D2 mRNA, but not with changes ineither immediate early genes or mitogen-activated protein kinase (MAPK)phosphorylation [Polotskaya et at., 1993]. The expression of Pim-1 protein is selectivelyinduced in the human factor-dependent myeloid leukemia cell line MO7E by IL-3 andGM-CSF [Lilly et at., 1992]. The promyelocytic cell line HL6O does not express pim-1with or without GM-CSF (and does not proliferate in reponse to the cytokine), but whenthe cell was made GM-CSF-dependent by DMSO treatment, pim-1 was induced. In themurine cell line MAC-TI, pim-1 is induced by IL-3 and GM-CSF in 2-4 hours, but not bybryostatin or M-CSF [Lilly et al., 1992]. The human line U937 (myeloid leukemia)expressed pim-1 in response to GM-CSF, G-CSF and IL-6 but not bryostatin, eventhough the cell line does not proliferate in response to GM-CSF [Lilly et a!., 1992]. Asadditional proof that pim-1 may mediate GM-CSF signalling, pim-1 anti-senseoligonucleotides reduced cell growth in GM-CSF cultures by 50-80% [Lilly, 1989].b) Background information on the GM-CSFJIL-311L-5 receptor familyInterleukin 3 (IL-3) and granulocyte-macrophage colony stimulating factor (GMCSF) are produced by activated T cells and mast cells, and serve as potent growth factorsfor immature multipotential hematopoietic progenitors. GM-CSF causes a dose-dependent proliferative response in immature myeloid cell lines, but acts on moredifferentiated cell types to cause activation as opposed to proliferation [Lilly et at., 1992].The receptors for human IL-3, IL-S and GM-CSF are heterodimeric and share a commonB subunit, but have distinct cytokine-specific a subunits. Ligand binding to the specifica subunit causes dimerization and activation of B subunits.The cytoplasmic portion of the 881 amino acid B-subunit has two distinct functionaldomains. The membrane distal domain between Leu-626 and Ser-763 is necessary formany of the functions of the GM-CSF receptor including the activation of tyrosinephosphorylation of cellular proteins including SHC, the activation of Ras, Raf-1, MAPK,p7OS6K and the induction of c-foslc-jun, yet it is dispensable for growth factor-inducedproliferation [Sato et at., 1993; Kinoshita et at., 1995]. The membrane proximal domainbetween Arg-455 and Glu-5 17 was essential for proliferation, JAK2 activation, and forthe induction of c-myc and pim-1 [Sato et a!., 1993; Quelle et a!., 1994; Kinoshita et at.,1995]. This proximal region is believed to stimulate a tyrosine kinase, as induction of c23myc and pim-1 by GM-CSF is sensitive to herbimycin A, a tyrosine kinase inhibitor, andreceptor deletion mutants containing only the proximal domain are also sensitive toherbimycin A [Sato et al., 1993; Queue et at., 1994].The 400 amino acid cytokine-specific a subunit functions not only in ligand bindingbut is necessary for signal transduction by protein phosphorylation and entry into the cellcycle [Polotskaya et at., 1993]. A short intracytoplasmic region of the a subunit (aa 346-382) is necessary for cell growth and is involved in translocation of PKC to the cellmembrane [Polotskaya et at., 1994]. Phosphorylation and activation of the B subunitrequires the entire intracytoplasmic domain of the a subunit. The functions of the a andB subunits are complementary and cooperative; not only is the B subunit necessary forhigh affinity ligand binding by the a subunit [Ronco et at., 1994] but the 54 amino acid asubunit cytoplasmic tail may mediate the specificity of the cellular reponse to cytokines,possibly by interacting with secondary signalling proteins {Polotskaya et at., 1993].c) Pathways induced by the IL-3/GM-CSF receptorsLong term cell proliferation of cultured cells requires two distinct pathways. Onepathway leads to the induction of DNA synthesis and is inhibited by both staurosporin (aprotein kinase C inhibitor that generally inhibits tyrosine and serine/threonine kinases)and genistein (an inactive ATP analogue that acts as a tyrosine kinase inhibitor). Theother route is an anti-apoptotic pathway that is inhibited only by staurosporin [Kinoshitaet at., 1995]. The IL-3 and GM-CSF activate the anti-apoptotic pathway even in thepresence of genistein by activating a signalling pathway distinct from the pathway thatcauses induction of DNA synthesis [Kinoshita et at., 1995]. Analysis of mutant IL3/GM-CSF receptor B subunits lacking one or more of the defined functional domains hasallowed definition of the downstream events and has permitted descrimination of the twosignal pathways involved in receptor signalling.Signals transduced through the membrane distal domain (Leu-626 and Ser-763) of theGM-CSF receptor are responsible for the prevention of apoptosis. The membrane distaldomain is necessary for phosphorylation and stimulation of SHC, ras, Raf-1 and MAPKand is not sensitive to inhibition by genistein. Removal of this domain causes the cells toundergo apoptosis, even in the presence of IL-3 [Kinoshita et at., 1995]. Expression of vRas in cells lacking this domain, rescues the cell from apoptosis but does not allow thecells to proliferate in long term culture [Kinoshita et at., 1995].24The membrane proximal domain (Arg-455 and Glu-5 17) of the GM-CSFIIL-3receptor B subunit is necessary for DNA synthesis and cell cycle progression, but not forthe prevention of apoptosis [Kinoshita et al., 1995]. This DNA synthetic pathway,repressed by both staurosporin and genistein, is responsible for JAK2 activation andcyclin E, c-myc and pim-1 mRNA induction. This domain does not seem to be necessaryfor the induction of cyclin D2, D3, CDK4 and bcl-2 [Kinoshita et al., 1995]. This datastrongly indicates that pim-1 induction in response to IL-3/GM-CSF is mediated throughthe membrane proximal domain and that Pim- 1 may be involved in a proliferativepathway [Fig. 1]. Pim-1 may occupy a similar position in pathways stimulated by 1L-3and erythropoietin (Epo), a glycoprotein produced in mammalian kidney and liver and isa member of the cytokine receptor superfamily.25Figure 1: Model of GM-CSF receptor signal transduction as mediated bytwo distinct domains of the B subunit. Protein tyrosine kinases (PYK)may be involved. Modified from Sato et al., 1993.GM-CSF receptor5176267%%k\i:EIshcSa 8c receptor subunitcell membrane455vav____Istati c-myc\\ pim-144PROLIFERATIONI c-fos/c-junlAPOPTOSIS26d) Potential upsteam activators ofpim-1 transcriptionOne candidate for the upsteam activator of pim-] is the JAK2 tyrosine kinase, amember of the family of receptor-associated soluble tyrosine kinases. JAK2 lacks SH2and SH3 domains and is activated and tyrosine-phosphorylated in response to GM-CSF,Epo, IL-3, IL-5, IL-6, ]FNy, prolactin as well as several other ligands [Schindler, 1995].JAK2 activation and pim-] expression are both mediated through the membrane proximalregions of the EpoR and the GM-CSF B chains, suggesting that JAK2 may be anintermediate factor upstream of Pim- 1 [Sato et at., 1993; Queue et at., 1994; Miura et at.,1994]. JAK2 associates with the proximal domain of the B chain independently of eitherGM-CSF or the GM-CSF x chain, implying that the GM-CSF B receptor subunit andJAK2 may be constitutively associated [Queue et at., 1994]. JAK2 phosphorylates a Statprotein in response growth factor stimulation, which then dimerizes, translocates to thenucleus and binds DNA, influencing the transciptional response of cytokine and growthfactor-inducible genes [Hill and Treisman, 1995; Schindler, 1995]. Recently, the 5’flanking region of pim-1 was found to contain a functional Statloc binding site [Yip-Schneider et at., 1995]. That Static is a specific target of JAK2, implies that Pim-1expression is upregulated through the JAK2-STAT pathway in contrast to acting as adirect substrate for the JAK2 kinase.Another protein that may function in the same signal transduction pathway as Pim-1is the proto-oncogene-encoded protein Vav. This 95-kDa protein contains one SH2 andtwo SH3 domains and has some of the structural features of a transcription factor. Vav isexpressed only in hematopoietic cells, and is tyrosine-phosphorylated in response tomany of the same growth factors that induce pim-1 expression. Vav is tyrosinephosphorylated in response to stimulation of the T-cell antigen receptor, cross-linking ofthe IgE or 1gM receptors by stimulation with Epo, and by p21OBCR/Ab1 expression inM07e cells [Matusuguchi et at., 1995; Miura et at., 1994]. In GM-CSF stimulated cells,Vav coprecipitates with JAK2, possibly by interaction of the SH2 domain of Vav withJAK2 [Matusuguchi et at., 1995]. Vav is tyrosine phosphorylated in response to Steelfactor (SF), but not GM-CSF or IL-3 [Alai et at., 1992], while pim-] mRNA is induced inresponse to GM-CSF and IL-3 but not SF alone. This implies that Pim-1 and Vav mayparticipate in a similar pathways, but in response to different stimuli.There are not many reports of other cytoplasmic mediators that have been tested asupstream activators of Pim- 1. Neither c-Fms tyrosine kinase [Lilly et at., 1992] nor cRaf-1 [Wingette and Magnuson, 1995] was found to influence pim-] expression.27e Other signal transduction pathways possibly involving Pim- 1Pim- 1 may play a role in growth factor signalling from the erythropoietin receptor(EpoR). Epo affects erythroid progenitors in hematopoietic organs and induced Pim-1expression in various 32DJEpoR cell lines. Stimulation of the EpoR caused receptorassociation with the JAK2 kinase, led to tyrosine phosphorylation of Vav and inducedexpression of the Pim- 1 protein [Miura et at., 1994]. Pim- 1 expression was stimulated invarious mutant EpoR-containing cell lines, except for those having an inactivated EpoRmutant; in these cell lines, only 1L-3 induced Pim- 1 expression [Miura et aL, 1994].Pim-1 expression is also stimulated in response to JFNy and SF in Mo7e cells [Yip-Schneider et al., 1995]. Stimulation of Mo7e cells with both growth factors has asynergistic effect on Pim- 1 levels in the cell. Stimulation with SF alone does not elicit aPim- 1 response, stimulation with IFNy induces both Pim- 1 mRNA and proteinaccumulation and stimulation with both growth factors causes a 2- to 3-fold increase ofboth Pim- 1 mRNA and protein levels. In this system, the upregulation of Pim- 1 occursboth transcriptionally and post-translationally: IFNy alone causes an increase in the rateof pim-] gene expression while the IFNy/SF stimulation increases the stability of themRNA transcript. STAT 1 x is thought to mediate the transcriptional effects of IFNy onpim-1 [Yip-Schneider et al., 1995].f) Possible involvement in an apoptotic pathway?Many of the growth factors that stimulate pim-], including IL-3, IL-5, GM-CSF andEpo, have an anti-apoptotic function [Koury and Bondurant, 1990; Kinoshita et at.,1995]. Apoptosis is a normal part of hematopoietic differentiation, with T cell apoptosistaking part during thymic selection and B cell apoptosis during the development of self-tolerance [Kinoshita et a!., 1995]. As tempting as it is to speculate about Pim- 1 being anapoptotic inhibitor, studies with mutant GM-CSF receptors have implicated Pim- 1 as acomponent of the cell proliferation pathway as opposed to the anti-apoptotic pathway[Sato et at., 1993; Kinoshita et a!., 1995]. Defining the function of Pim-1 in the cellcycle and in lymphopoiesis requires further study.The induction of pim-1 gene during the primary response to growth factors suggeststhat Pim- 1 functions as a cytoplasmic mediator in myeloid growth factor signal cascades.Although the mRNA and protein levels are upregulated, it is unclear if the activity of thekinase is modified in any way by phosphorylation. Indeed, changes in phosphorylationstate of Pim- 1 (as evidenced by band shifts on SDS-PAGE gels) in response to growth28factors have not yet been demonstrated. The recent identification of a STAT 1 a bindingsite in the pim-1 gene strongly imply that response of Pim- 1 to growth factor stimulationmay be a secondary effect, and that Pim-1 is not part of the primary signal transductionpathway from the cell membrane to the nucleus.8. PIM-1 TRANSGENIC MICEStudies with transgenic mice have contributed the most to our understanding of therole of pim-1 and have confirmed that it is, in fact, an oncogene. Provirus tagging withslow growing retroviruses and the creation of double transgenics have led to theidentification of other oncogenes with which pim-1 cooperates. The creation of pim-1null mice has permitted comparison of the growth factor reponse of cells from pim-1 null,wild type and overexpressing animals, allowing the role of pim-1 to be assessed inspecific signal transduction pathways.i. TRANSGENIC ANIMALS CONFIRM THAT PIM-1 IS AN ONCOGENEPim-1 transgenic mice were produced with an upstream immunoglobulin enhancer(Eji) in the promoter and a single MuLV long terminal repeat (LTR) inserted in the 3’UTR to boost expression of the transgene further [van Lohuizen et at., 1989]. The Eli.enhancer was used to achieve a high level of transcription, as fusion genes between pim-1and proviral sequences alone were not expressed in earlier studies. The transgene wasexpressed at similar high levels in in both B and T cells, and high levels of Eji-pim-1mRNA were detected in the thymus, bone marrow and spleen. No expression wasdetected in the testes and only low levels were detected in most other tissues, possibly asa result of the presence of circulating lymphocytes. Expression of the transgene did notinterfere with expression of the endogenous pim-1 mRNA; levels of endogenous Pim-1were comparable between tissues from transgenic and control mice [van Lohuizen et at.,1989]. Pim- 1 protein levels were increased in thymic and splenocytes of transgenicanimals. Histological examination and fluorescence-activated cell sorting (FACS)analysis of various tissues (liver, spleen, thymus, lymph node and bone marrow) in nondiseased Eji-pim- 1 transgenic mice revealed no abnormalities nor increase proliferationof hematopoietic cell populations, although a slight enlargment of the spleen wasobserved [van Lohuizen et a!., 1989; Verbeek et a!., 1991].29Overexpression of the Ei-pim- 1 transgene caused a rise in the incidence of tumordevelopment in transgenic animals. These tumor cells adapted to in vitro culture andcould cause tumors in syngenic hosts, confirming the identity of pim-1 as an oncogene[van Lohuizen et at., 1989]. The target cell population for transformation by Ej.t-pim-l isnot a reflection of the expression pattern of this oncogene; despite similar expressionlevels in both B and T cells, the Eji-pim-1 transgenic animals developed T celllymphomas exclusively. The tumors had a long, varied latency period, with tumorsappearing in 5-10% of the mice after 7 months [van Lohuizen et at., 1989]. This longlatency and the fact that lymphomas were monoclonal in nature implies that high pim-1expression is not sufficient, and that other genetic events are required to induce the fullymalignant phenotype [van Lohuizen et at., 1989].ii. PROVIRAL TAGGING TO IDENTIFY COOPERATING ONCOGENESStudies of proviral integration in murine lymphomagenesis indicate that usually morethan one gene is activated [Berns 1988b; 1991]. The method of proviral tagging byMuLV neonatal infection has been used to identify new oncogenes and to define sets ofcooperating oncogenes; taking advantage of the fact that slow transforming retrovirusescontain no oncogenes, genomic oncogenes are activated and ‘tagged’ by insertion of theprovirus. Oncogene-expressing transgenics can be infected with the virus and malignanttransformants analysed for the site of viral integration. Pim-] transgenic mice are verysucceptible to MuLV-infection, and thus form an ideal system to study provirus tagging.Proviral tagging identified N-myc and c-myc as being oncogenes that cooperate withpim-1 [van Lohuizen et at., 1989], despite the lack of cooperativity initially demonstratedbetween pim-1 and myc in tumors from human patients and in FeLV-induced murineleukemias [Amson et at., 1989; Tsatsanis et at., 1994]. Infection of Eji-pim- 1 transgenicmice with MuLV decreased the latency period of tumor formation to 7-8 weeks ascompared to 7 months in uninfected mice and in all lymphomas either c-myc (80%) or Nmyc (20%) had been activated by MuLV proviral integration [van Lohuizen et at., 1989].Myc/Pim- 1 cooperativity was confirmed after Mo-MuLV-infected Eji-c-myctransgenic mice also experienced acceleration of pre-B cell leukemia [Haupt et at., 1991;van Lohuizen et at., 1991]. Eji-c-myc is expressed in B cells, but not T cells or othersomatic tissues, and Ei-c-myc transgenic animals showed a predisposition to pre-B celllymphomas characterized by an enlarged pre-B cell compartment with some aberrantlyexpressed cell surface markers. MuLV infection decreased the tumor latency period in30EJi-myc transgenic mice from 150 to 50 days, and the pre-B lymphomas that developedhad high proportions of proviral integration in the pim-1, bmi-l, pim-2 and emi-l loci[Haupt et al., 1991; van Lohuizen et at., 1991].A second transgene cooperating with pim-1 identified by proviral tagging was v-abl.[Haupt et at., 1993]. Eji-v-abl transgenic animals develop plasmacytomas and afterinfection with MuLV, experience accelerated T cell tumor development. Insertions in cmyc, N-myc or pim-1 were observed in 42% of tumors and of tumors involving c-mycactivation, 14% also had pim-1 insertions, suggesting that all three oncogenes maycooperate in tumorigenesis.iii. STUDiES WITH PIM-JIMYC DOUBLE TRANSGENIC MICEThe creation of double transgenic animals strongly comfirmed the synergistic effectsof pim-1 and members of the myc family of oncogenes. All the myc family genes cansynergize with pim-1 to cause lymphoid tumors in vivo. The relative transformingefficiency (c-myc>N-myc>L-myc) was maintained when transgenics were crossed withEu-pim-1 mice, but the latency period was accelerated and the disease was moreextensive [Moroy et at., 19911.Overexpression of pim-1 and c-myc caused severe synergistic effects, as Eu-pim1/Ep.-c-myc double transgenic mice developed pre B cell leukemia in utero [Verbeek etat., 1991]. Analysis of 17-19 day fetuses revealed that although the gross morphology ofdouble transgenics was normal, the spleen was enlarged and there was a dramaticexpansion of pre-B lymphoid cells in the peripheral blood. Despite this strongcooperativity, additional genetic events were still needed for the development of a fullymalignant phenotype [Moroy et at., 1991; Verbeek et at., 1991].Transgenic mice were examined to determine if the N- and L- myc family membersmaintain lineage specific neoplasia when co-expressed with Eu-pim-1, and to assess therelative transforming activities of the myc genes when collaborating with otheroncogenes [Moroy et at., 1991]. The Eu-N-myc mice expressed the transgenepreferentially in B cells and were predisposed to B-cell neoplasia. The Eu-N-mycfEupim-1 double transgenic mice were smaller and more sickly than single transgenelittermates and experienced accelerated lymphomagenesis, with pre-B lymphomadeveloping in 36 days in contrast to 13-16 weeks with N-myc alone [Moroy et at., 1991].31The Eji-L-myc transgene had the longest latency and the lowest tumor incidencewhen compared to the other two myc transgenics, and was expressed preferentially in Tcells, leading to the development of thymic hyperplasia. Alone, Eu-L-myc and Eu-pim-itransgenics each had a transforming efficiency of about 5-10%. When expressedtogether, L-myc/pim-1 double transgenics developed tumors in the thymus with someinvolvement of spleen and lymph nodes; efficiency was increased to 82% with ashortened latency period of 94 days [Moroy et at., 1991].To summarize, all three myc genes cooperate with pim-i in vivo to generate lymphoidtumors but N- and L-myc synergized less efficiently than with c-myc. When expressedalone or with other oncogenes, the myc family members maintained descendingefficiency and lineage specificity but had more rapid development of tumors when coexpressed with pim-1. The myc/pim tumors were monoclonal, indicating that otherevents are required for full tumorigenesis [Moroy et at., 19911. Previous studies haveindicated two additional steps are required for tumorigenesis in pim-limyc mice [Berns,19911.iv. OTHER STUDIES WITH DOUBLE TRANSGENIC MICEThe bcl-2 gene was expressed in transgenic mice and was found to also cooperatewith pim-i [Acton et at., 1992]. The bcl-2 oncogene is an important inhibitor ofapoptosis and cooperates with myc [Mann et at., 1995]. Bct-2-Ig transgenic mice had alow tumor incidence and the expression of the transgene caused B cell malignancies bycellular immortalization. Coexpression of pim-i and bcl-2 affected several different celltypes and accelerated tumorigenesis with a long and variable latency. MuLV infection ofbct-2/pim-i transgenic mice led to increased tumorigenesis, with proviral integrations insites including N-myc, c-myc and pal-i. Surprisingly, when bcl-2 transgenics wereinfected with MuLV, there was no increase in the incidence of tumor formation than withMuLV-infected non-transgenics; bcl-2 is not important for MuLV-induced expression[Acton et al., 1992].Studies with Eu-pim-1 lpr/tpr double transgenic mice suggest a role for pim-i inapoptotic pathways [Moroy et al., 1993]. The lpr mutation in mice causes a structuralrearrangement of the FAS gene product, a 35 kDa surface receptor molecule expressed inlymphoid cells, structurally homologous to the TNF receptor and to CD4O. The FASgene product is involved in the thymic selection process by transducing an apoptoticsignal upon antibody binding at the extracellular domain. C57BL/6 mice homozygous32for the lpr mutation, developed a well described lymphoproliferative syndrome at 26-30weeks, characterized by the accumulation of abnormal T cells. Expression of the pim-1transgene rescued lpr lymph node cells from programmed cell death in vitro andprevented steroid-induced apoptosis in vivo. Expression of the pim-] transgene in thelpr/lpr mice led to a strong acceleration of lymphoproliferation and an increasedaccumulation of the non-malignant abnormal lpr T cells leading to a dramaticenlargement of lymph nodes in all areas of the body. Eu-pim- 1/lpr/lpr thymocytes wereprotected from steroid-induced apoptotic signals, as dexamethasone was unable to induceapotosis in these cells.Cross breeding bcl-2, lpr/lpr and myc transgenic mice with the Ej.t-pim-1 transgenicscaused them to maintain the lymphocytic lineage specificity of the original transgene, butresulted in rapid development and a high incidence of the lymphoid malignancies [Moroyet al., 1991; Verbeek et al., 1991; Acton et al., 1992; Moroy et al., 1993]. Pim-Joverexpression seems to enhance the effects of oncogenes promoting cell proliferationand cell survival by acting as an apoptotic inhibitor rather than a stimulator of cellproliferation [Moroy et al., 1993].Recently E.i-myc/Pim- 1 (-I-) double transgenic mice were used to identify the pim-2gene, a second pim family member [van der Lugt et al., 1995]. By complementationtagging, they managed to identify the pim-2 gene which became activated in the absenceofpim-1 expression, implying that the proteins encoded by the pim-1 and pim-2 genes arefunctionally redundant.v. PIM-1 NULL MUTANTS - EMBRYONIC STEM CELLSIn vitro studies of effects of pim-1 in early embryonic development were done byconsecutive inactivation of pim-] by homologous recombination in embryonic stem (ES)cells [te Riele et al., 1990]. A knockout was achieved in two steps by homologousrecombination, using G4 18 and hygromycin B for selection. The null allele was createdby deletion of the promoter, transcription and translation initiation sites and by removinga large segment of the coding region containing the conserved lysine residue of the ATPbinding site in subdomain II of the protein kinase catalytic domain. As the pim-1 genewas highly expressed in ES cells, it was thought that effects of pim-1 gene knockout maybe manifested during differentiation and during in vitro propagation. However, nodifferences in ES cell morphology were observed, with characteristic embryoid bodiesdeveloping in all cases, and with no selection against the double pim-1 knockout occuring33[te Riele et al., 1990]. This suggests that pim-1 is not required for normal ES cellproliferation and differentiation.vi. THE LACK OF PHYSIOLOGICAL EFFECTS IN PIM-1 NULL MICEPim-1 deficient or null transgenic mice were constructed and found to have asurprisingly lack of phenotypic abnormalities [Laird et at., 1993; Domen et al., 1993c].The pim-1 null mice displayed normal behavior, normal body weights, no morphologicaland histological abnormalities, no differences in tissue distributions were observed, andboth male and female were fertile [Laird et at., 1993]. The immature lymphoidcompartments were analysed by flow cytometry to reveal no difference between the pim1 null and normal mice. Splenocytes had stimulatory responses to ConA andlipopolysaccharide (LPS) similar to those of controls [Laird et al., 1993].The only phenotypic abnormality observed in pim-1 null mice was erythrocytemicrocytosis; the MCV (Mean Cell Volume) of erythrocytes was smaller in pim-1 nullmice compared to wild-type littermates [Laird et at., 1993]. The concentration of redblood cells was not elevated to compensate for the microcytosis, so hemoglobin levelswere reduced. Conversely, the erythrocytes in Eu-pim-1 transgenics overexpressing theprotein were larger than normal, and there was a compensatory decrease in erythrocyteconcentrations resulting in normal hematocrit and hemoglobin levels. The pim-1 genewas responsible for this abnormality, as introducing a pim-] transgene with its ownpromoter into the pim-1 null transgenics restored the low erythrocyte MCV to wild-typelevels [Laird et al., 1993]. It has not been determined if the lack of physiological effectsin pim-1 null mice is due to a compensatory effects of the pim-2 gene product [van derLugtetal., 1995].vii. GROWTH FACTOR STIMULATION OF CELLS FROM TRANSGENICANIMALSA significant functional anomaly observed in cells from the pim-1 null mice was animpaired, but not absent, proliferative response to IL-3 induction in bone marrow-derivedmast cells (BMMC) [Domen et al., 1993a, 1993c]. Conversely, expression of highamounts of pim-1 did not lead to growth factor independence, as withdrawal of IL-3 ledto apoptosis in BMMC cells from Ei-pim-1 mice [Domen et a!., 1993c]. Cellsheterozygous for pim-1 expression (pim(+I-)) showed dosage effects with IL-3 [Domen etal., 1 993a]. No difference in cell viability was observed between the wild-type, pim-1overexpressing and pim-1 null cells. Pim-1 null mice had a normal mast cell response to34nematode infection, despite impaired response to IL-3, possibly because of the multiplegrowth factors (i.e. IL-4) orchestrating an immune response in vivo.The effects of pim-1 levels on the size of the early B lymphoid compartments in bonemarrow were studied using cells from wild-type (WT) and from pim-1 null and pim-1overexpressing transgenic mice. The pim-] levels determined the number of SF and IL-7responsive early B-lymphoid colony forming cells (CFC) in the bone marrow [Domen etal., 1993b]. IL-7 induced pre-B colonies, while the combination of IL-7 and SF inducedmore primitive cells to differentiate into pre-B cells. The pim-] levels affected the size ofthe earliest B cell progenitor compartments in the bone marrow most profoundly; in thepim(+) CFCs there was an increased response to IL-7 and SF, as evidenced by anincreased number of pre-B colonies responsive to IL-7 and SF and a reduction in the sizein the mature B cell compartment by half as compared to the WT. In contrast, the pim(-)colonies had a reduced growth rates in response to IL-7 and SF, as reflected in thereduced numbers of early IL-7 and SF-responsive B lymphocytes. The lack of IL-7response in the pim(-) cells was rescued by a transgene that restored pim-1 expression toWT levels [Domen et al., 1993b]. Pim-l affects IL-7 response; pim-1 functioned in adose-dependent manner but while pim-1 overexpression did not cause IL-7 growth factorindependence, the pim(-) cells were still partly responsive to IL-7 and SF.The results indicate that Pim-1 seems to function in B-lymphopoiesis and is involvedin the response to growth factors acting on the early B-lymphocyte compartment bytransducing signals or halting differentiation. These response differences were notobserved when complex cell-cell interactions occurred between stomal and lymphoidcells in Whitlock-Witte cultures [Domen et al., 1 993b]. These studies have shown thatpim-] is somehow involved in the IL-3, IL-7 signal transduction pathways. However, thefunction of pim-1 may be redundant as effects of overexpression or knockout are maskedby complex cellular interactions.viii. SUSCEPTIBiLITY OF PIM-1 TRANSGENIC ANIMALS TO CHEMICALCARCINOGENSEi-pim- 1 transgenic mice are tumor-prone and display a high incidence of tumorformation when exposed to chemical carcinogens or viral infection [Breuer et al., 1989b,1991; Armstrong and Galloway, 1993]. Ej.t-pim-1 transgenics are approximately 25-foldmore susceptible to ENU (N-ethyl-N-nitrosourea)-induced lymphomagenesis [Breuer etal., 1989b, 1991]. H2K-pim-l transgenics were also tested, but with less impressive35results [Breuer eta!., 1989b1. When a simple low dose of ENU was administered, almostall Ei-pim-1 transgenic mice but only 15% of control mice developed T cell lymphomas,and c-myc levels were strongly elevated in most tumors supporting the concept of piml/c-myc cooperativity [Breuer et al., 1989b, 1991]. Approximately 10% of the tumorsalso had a ras mutation, but this was thought to be a later event, independent of the ENUinduction [Breuer eta!., 1991]. The levels of pim-1 in ENU-induced lymphomas of bothEu-pim- 1 and normal mice were highly variable, so tumor formation could not becorrelated to the levels of pim-1 expression [Breuer eta!., 1989b, 1991].Several groups have exploited this susceptibility of Eji-pim-1 transgenic mice to testthe effects of various carcinogens. Armstrong and Galloway [1993] found that 2-acetylaminofluorene (2-AAF) and benzene led to accelerated lymphogenesis in pim-1transgenics, but 1,2-diethylnitrosamine (DEN) and l,2-dichloroethane (1,2-DCE) did not.They used the formation of blood micronuclei (micronucleated erythrocytes) as ameasure of bone marrow genotoxicity, with the intention of developing the Ep-pim-1transgenic mice as a model for testing other oncogenic agents. A second study alsodemonstrated that both 2-AAF and N-nitrosodiethylamine (NDEA) led to a significantincrease in lymphomas in Eji-pim-1 transgenic mice [Storer et a!., 1995]. In contrast tothe previous study, 1,2-DCE was also shown to cause an increase in murinelymphomagenesis, while benzene did not produce any significant increases in lymphomainduction [Storer et a!., 1995]. The conflicting results from these two studies may resultfrom differing methods of chemical administration (oral versus injection) and fromdifferent methods of assessing toxicity (formation of micronuclei versus lymphomainduction). These studies demonstrate that while the Ei-pim-1 transgenic mice do havean increased sensitivity to chemical carcinogens, their sensitivity may not be sufficient tojustify their use as a tool to screen chemical carcinogens. Further work needs to be doneto standardize methods of chemical administration and assessment of effects before thistransgenic model can be used in carcinogenesis screening assays.369. THE ROLE OF OTHER ONCOGENE-ENCODED SERINE/THREONINEKINASES IN OOCYTE MATURATIONPim- 1 belongs to a small family of oncogenic serine/threonine kinases that includesTpl-2ICot, Mos, Raf-1, and AktfRac. Although the sequences and structures of thesekinases are not related, the fact that these are the only known serine/threonine kinasesencoded by oncogenes is indeed significant and implies that these kinase may havesimilar functions in the cells. Indeed, studies have indicated that some interestingfunctional similarities exist between these kinases.Tpl-2 (tumor progression locus 2) is the murine homologue of the human Cot (cancerosaka thyroid) protein [Miyoshi et al., 1991; Aoki et al., 1991; Makris et al., 1993]. Thegene was first identified as being involved in the progression of Moloney murineleukemia virus-induced leukemia in rats, with proviral insertion in the 3’ end of the geneleading to the production of truncated stabilized mRNA transcripts in a similar manner topim-] [Patriotis et al., 1994]. The protein is a close relative of MEK-1 and MEKK, actsdownstream of Ras and Raf- 1, and contributes to the activation of the MAPK cascade[Patriotis et al., 1994].Mos was one of the first kinases from this family to be studied. It first receivedrecognition as an active component of cytostatic factor (CSF) necessary for both thestablilization of maturation promoting factor (MPF) consisting of cyclin B and p34CdC2,and for activating MPF to promote the maturation of X. laevis oocytes [Sagata et aL,1989]. Mos is also necessary for the insulin and progesterone-induced maturation of X.laevis oocytes. Recently Mos has been demonstrated to activate maturation-activatedprotein kinase (MAPK) in X. laevis oocytes and to maintain the activity of MAPK duringmeiosis [Posada et a!., 1993; Nebreda and Hunt, 1993]. Mos phosphorylates MAPKK invitro, suggesting a method by which it may activate MAPK in vivo [Posada et al., 1993].Raf- 1 has also been shown to be a member of the MAPK signal transductionpathway, and is located upstream of MEK- 1 and downstream of Mos [Muslin et al.,1993]. Importantly, Raf- 1 plays a significant role in the regulation of progesterone-induced X. laevis oocyte maturation as well as in the early development of the X. laevisembryo [MacNicol et a!., 1993]. Raf-1 is a key signalling molecule in the developmentof the posterior stucture of the X. laevis embryo, mediating the cell differentiatingresponse to FGF in vertebrates [MacNicol et al., 1993].37The product of the AKT8 retrovirus, AktfRac, was first isolated form a rodent T celllymphoma. The cellular homologue is expressed in most tissues including testes, andexpression is especially high in the thymus [Bellacosa et al., 1991]. Unlike the otheroncogene encoded serine/threonine kinases, akt contains a protein interaction motif, aSH2-like domain called the Pleckstrin homology box in the N-terminal regulatory region.Although Mos, TPL-2 and Raf function in the MAPK cascade, there is no evidencethat Pim-1 has a role in this pathway. While Pim-l has been demonstrated to play a partin mammalian male germ cell development with the production of an alternative, 2.4 kbPim- 1 transcript in testicular tissue, no studies have examined the expression orregulation of Pim-1 in oocyte maturation [Sorrentino et al., 1988; Wingette et a!., 1992].Although pim-] mRNA transcripts were not initially detected in ovaries, this early studyalso did not detect pim-1 transcripts in testes, indicating that the sensitivity of detectionmay have been very low [Selton et at., 1985]. Because other oncogene-encodedserine/threonine kinase family members, Mos and Raf are important for X. laevis oocytematuration and because pim-1 has been shown to play a role during male germ celldevelopment, we hypothesize that Pim- 1 may function during oocyte maturation ordevelopment.10. SUMMARY OF PIM-1The pim-1 oncogene is highly conserved between species, suggesting that it plays animportant function in the cell. Studies with transgenic mice have confirmed that pim-1 isan oncogene, producing a low spontaneous rate of tumor incidence. The susceptibility ofPim- 1 overexpressing mice to chemical carcinogens and to MuLV infection has made theidentification of cooperating oncogenes possible. Pim- 1 cooperates strongly with c-myc,less strongly with N-myc, L-myc, bcl-2, lpr and possibly v-abl to contribute totransformation. In all cases, the coexpression of these oncogenes alone is not sufficient tocause malignant transformation; additional genetic events are required, supporting themodel of multistep tumorigenesis.Very little is known about the function of Pim-1, but expression patterns indicate thatit is involved in hematopoietic signal transduction and in the development of male germcells. Expression of pim-] mRNA and protein is very tightly regulated at different levels,38implying that is functionally potent. Pim-1 is induced by mitogens and growth factorsand may take part as a cytoplasmic mediator in a signal transduction pathway.Upregulation of Pim-l by signalling through the GM-CSF receptor family is dependenton the presence of the membrane proximal domain, and may involve the JAK2 tyrosinekinase. During GM-CSF stimulation, pim-1 is involved in the DNA synthetic as opposedto the apoptotic inhibition pathway.The only phenotypic anomally detected in pim-1 transgenic mice is an alteration inthe size of erythrocytes. The lack of physiological effects in transgenic mice bothoverexpressing and deficient in pim-1 implies that the oncogene is functionally redundantand any effects of the expression (or lack of) are masked by the effects of other growthfactors or pathways in vivo. The pim-1 null mice have an impaired response to severalgrowth factors, implying that the kinase functions as a signal transduction modulator,possibly in the pathways stimulated by IL-3 and Th-7. Pim-l may act to cause aninhibition of differentiation of early progenitors (especially B cell progenitors), or mayact to inhibit apoptosis.Although extensively studied in the hematopoietic system, the expression andregulation of pim- 1 has never been studied in the oocyte system. Recent studies of otheroncogene-encoded serine/threonine kinases (Mos, Rat) has implicated these kinases ashaving a vital role during oocyte maturation. It was the aim of this study to examine therole of this kinase in the maturing oocyte system.39CHAPTER II.HYPOTHESIS1. pim-1 is expressed in germ cells and is thought to play a role in male germ celldevelopment [Sorrentino et al., 1988, Wingette et al., 1992]. pim-1 may possibly play asimilar developmental role during oocyte maturation.2. As pim-1 is present and is highly conserved among all mammalian species examined,we expect the enzyme to also be expressed in Xenopus laevis. It is expected that Pim- 1from Xenopus laevis will behave similarly to mammalian Pim- 1.3. As a kinase with a high degree of autophosphorylation activity, autophosphorylationof Pim- 1 is likely to serve a functional purpose. Autophosphorylation may modify thekinase activity of the enzyme or may allow the protein to participate in interactions withother proteins. Identification and modification of the autophosphorylation site(s) willallow examination of changes to the autokinase activity and the kinase activity towardsexogenous substrates resulting from mutations introduced at these sites.4. As a highly conserved kinase, Pim-1 may phosphorylate important physiologicalsubstrates at specific sites that conform to distinct sequence motifs. Characterization ofthis consensus phosphorylation site sequence will aid in the identification of targets ofPim- 1.40RATIONALE1. X. laevis and P. ochraceus were selected as model systems, because these systemsprovide abundant sources of oocytes for biological characterization. The oocytes arearrested at the same stage of maturation and can be induced to mature by progesteronestimulation allowing the activity of the enzyme of interest to be studied at discrete stagesof maturation.2. The role of Pim- 1 in oocyte maturation and early development has not been explored.3. The X. laevis oocyte maturation system is well characterized with respect to otheroncogene-encocled serine/threonine protein kinases (Mos and Raf). As well, this systemoffers the opportunity to perform microinjection experiments to assess the activity of thekinase in vivo.4. Sequences important for the function and regulation of Pim- 1 are likely to beconserved between divergent species. Comparing the non-mammalian Pim-l sequenceswith the human sequence will help determine regions of the protein that are important forcatalytic function, substrate binding or regulation of the protein.5. Pim-1 has never been purified to homogeneity from an endogenous source, so verylittle enzymological characterization of Pim- 1 has been performed. Expression of Pim- 1as a recombinant protein will allow large amounts of the enzyme to be produced foranalysis.6. Many protein kinases autophosphorylate then become active towards other substrates.Despite the strong autophosphorylation activity of Pim- 1, there is no proof that Pim- 1activity is modulated by autophosphorylation. Novel regulatory sites may potentially beidentified in the Pim- 1 protein, which may then act as a model for regulation byautophosphorylation for other kinases containing homologous residues.7. There are no known or suspected physiological substrates of Pim-l. Understandingthe requirements for substrate recognition by Pim- 1 will help to identify potentialphysiological substrates and confirm the exact location of phosphorylation in a suspectedsubstrate.41OB.TECTIVES1. To clone pim-] from a non-mammalian species, in this case, from Xenopus laevis.2. To develop specific antibodies that can be used to detect Pim- 1 in the X. laevis oocytesystem.3. To express the cloned Pim-1 from frog and human as bacterial fusion proteins and touse these expressed proteins to determine the substrate recognition sequence using aseries of peptide substrates.4. To use the Pim- 1-specific antibody and peptide reagents to examine changes inexpression and activity of this kinase during oocyte maturation.5. To identify the sites of autophosphorylation in Pim-1, to mutate theseautophosphorylation sites using site-directed mutagenesis and to determine their potentialroles in the regulation of Pim- 1 activities.42CHAPTER IIIMETHODS1. SUPPLIES AND SOURCES1. CHEMICAL REAGENTS AND LABORATORY SUPPLIESBoehringer Mannheim = BM, Fisher Scientfic= FS, New England Biolabs= NEBAceteic Acid (CH3COOH) FSAcetonitrile Applied BiosystemsAcid phosphatase SigmaAcrylamide FS/ICNAdenosine 5’-triphosphate disodium salt SigmaAgarose Gibco BRLAgarose (low melting point) BRLAnildo black lOB ICNAlkaline phosphatase BMAmmonium bicarbonate (NH4HCO3) BDHAmmonium hydroxide FSAmmonium persulfate FSAmmonium sulphate ([NH4]SO) FSAmpicillin (D[-]-a-Aminobenzylpenicillin) SigmaAmpliTaq DNA polymerase Perkin-Elmer CetusAprotinin SigmaBind-Silane LKBBis-acrylamide FSN,N’- Methylene bis-acrylamide FSBovine serum albumin SigmaBrilliant Blue G Sigma5-Bromo-4-chloro-3 indoyl phosphate (BCIP) Sigma1-Butanol FSiso-Butanol FSBenzamidine ICN3-Glycerophosphate ICN43(3-Methyl aspartic acid SigmaCalcium chloride (CaC12) SigmaCentricon tubes (10 and 30) AmiconCitric acid BDHChloroform FSCollagenase SigmaConcanavalin A BMCoomassie Brilliant Blue R EM ScienceCounting scintillant Amershamc&Chymotrypsin Sigmao-casein, dephosphorylated SigmaDenatured alcohol FS2’-Deoxynucleoside 5’-Triphosphate (dNTP kit) PharmaciaDiethyl pyrocarbonate SigmaN,N-dimethyl formamide (DMF) Sigma/USDimethyl sulfoxide (DMSO) USDisodium pyrophosphate SigmaDispase BMDeoxyribonuclease SigmaDithiothreitol (DTT) BDHDNA 1 kb ladder Gibco BRLDNA -herring sperm BMDynabeads oligo (dT)25 DynalEnhanced chemiluminescence kit AmershamEnolase SigmaEthanolamine SigmaEthylene bis (oxyethylenenitrilo)J tetraacetic acid (EGTA) FSEthylene diamine tetraacetate disodium salt (EDTA) FSEthidium bromide Molecular Probes IncFinquel (methyl trisulphonate) ArgentFormaldehyde solution (HCOC) FS/BDHFormalin SigmaGelatin BioRad/SigmaGeneclean kit Bio 101Glacial acetic acid FS44Glutathione SigmaGlutathione cross-linked 4% beaded agarose SigmaGlycerol AnachemiaGlycine ICN/SigmaJFSGuanidine thiocyanate ICNHistone hA SigmaHistone ITS SigmaHistone ITT-S SigmaHistone Vu-S SigmaHybond-N hybridization membrane AmershamHydrochloric Acid FSIsopropyl 3-D-thiogalactopyranoside (IPTG) FisherBiotech/PromegaKemptide SigmaN-(2-Hydroxyetheyl)piperazine-N’-(2-ethanesulphonic acid)(HEPES) SigmaT4 DNA ligase NEB/BMN-Lauryl sarcosine SigmaLauryl sulfate (dodecyl lithium sulphate) SigmaLeupeptin SigmaJICNLiquid paraffin BDHLithium chloride anhydrous BDHISigmaJFSLysozyme SigmaIBMMagnesium acetate tetrahydrate BDHMagnesium sulphate (MgSO4-7H20) FSMagnesium chloride (MgC1-6H0) FSMaltose BDHManganous chloride (MnC12-4H0) BDH2-Mercaptoethanol BioRadMethanol FS/BDH1 -Methyladenine SigmaDL-Threo-B-methylaspartic acid SigmaMES (2-[N-Morpholino]ethanesulfonic acid) SigmaMonoS column PharmaciaMonoQ column PharmaicaMOPS 3-[N-Morpholino]propanesulphonic acid SigmaJICN45Myelin basic protein Kinetek/SigmaN-ethyl maleimide SigmaNinhydrin BDHNitric acid (HNO3) USNitro blue tetrazolium (NBT) SigmaNitrophenyl phosphate disodium salt SigmaNonidet P-40 BDHOligo(dT)-cellulose type 7 PharmaciaPetroleum ether (50-110°C) J.T. BakerPetroleum ether (60-80°C) BDHPhenyl phosphate disodium salt (P’tase inhibitor). ICNPhenol ICNPhenolsulfonphthalein (Phenol red dye) SigmaPhenyl methylsuiphonyl fluoride (PMSF) Sigmap-nitrophenyl phosphate Sigmap81 phosphocellulose filter paper Whatmannortho-Phosphoric acid (H3P04) FSPhosphatase, alkaline (high conc) BMPhosphate-buffered Saline (PBS) Gibco0-Phospho-L-serine Sigma0-Phospho-DL-threonine Sigma0-Phospho-L-tyrosine SigmaPhosphorylase B SigmaPhosvitin SigmaPlasmid kit QuiagenPotassium acetate (C2H30K) BDHPotassium chloride (KC1) FSPotassium dichromate (K2Cr07) BDHPotassium dihydrogen orthophosphate monobasic BDHPotassium hydroxide (KOH) BDHPotassium phosphate (dibasic) Sigmadi-Potassium hydrogen orthophosphate 3-hydrate BDHPotassium dihydrogen orthophosphate (KH2PO4) BDHPonceau S concentrate SigmaT7 polymerase NEB/PharmaciaDNA polymerase I large fragment (Kienow) NEB/Gibco-BRLIBM46Polynucleotide kinase NEB/BMPKI - cAMP-dependent protein kinase peptide inhibitor SigmaPrestained SDS-PAGE standards (Low Mr range) UBlIBioRadProgesterone CalbiochemPropanol FSProtamine chloride SigmaProtamine sulphate SigmaProtein A Sepharose CL-4B PharmaciaProteinase K PharmaciaPVDF membrane Millipore/DupontPyridine FSRandom primers (5’-pd(N)6) (100 pmol4tl) PharmaciaResource Q resin PharmaciaRestriction enzymes NEBIPromegaJBMRepel-Silane (dimethyldichiorosilane) PharmaciaReverse transcriptase (Superscript) BRLRibonuclease A PharmaciaRNA standards GIBCO BRLRNAsin PromegaRNAseGuard PharmaciaT7 Sequencing kit PharmaciaSephadex G50 PharmaciaSephacryl-HR 300 PharmaciaSephglas Bandprep Kit PharinaciaSilver nitrate FSSodium acetate (dibasic) BDHSodium azide FSSodium bicarbonate (NaHCO3) FSSodium borate FSSodium carbonate - anhydrous (Na2CO3) BDHSodium chloride (NaC1) FStn-Sodium citrate BDHSodium deoxycholate BDHSodium dihydrogen orthophosphate (NaHPO4-20) BDHSodium dodecyl sulphate (SDS) FSSodium fluoride (NaF) BDHIFS47di-Sodium hydrogen orthophosphate (Na2HPO4) BDHSodium hydroxide FSSodium orthovanadate (Na3V04) FSSoybean trypsin inhibitor (SBTI) SigmaSuperscript reverse transcriptase BRLSuiphadiazine (4-amino-N-2-pyrimidinyl-benzenosulfonate) SigmaTEMED (N,N,N’,N’-tetramethylethylenediamine) PSThrombin SigmaL-threonine, L-tyrosine, L-serine SigmaTrifluoroacetic acid (TFA) Applied biosystemsTris hydroxylmethyl aminomethane hydrochloride (Tris-Ci) FSTris (hydroxylmethyl) methylamrnonium chloride (Tris-Cl) BDHTris (hydroxylmethyl) methylamine BDHJFSTriton X-l00 BDHIFSTricaine methylsulfonate SyndelSequencing grade modified trypsin PromegaTween -20 FSUrea BioradIBDHVent DNA polymerase NEBZinc chloride BDH2059 tubes Becton Dickinson2070 tubes Becton Dickinson3mm filter paper Whatmann2. PHOTOGRAPHY SUPPLIESDeveloper and replenisher KodakFixer and replenisher KodakISO 3000 Polaroid film 667 PolaroidISO 100 Polaroid film PolaroidReflection NEF- autoradiography film DupontX-OMAT AR imaging film Kodak483. PLASMIDS AND BACTERIAL STRAINSBluescript II KS M13(+) plasmidpGEX-2T vectorR408 Interference resistant helper phageXL1 -Blues bacteriaJM1 10 bacteriaDH5o bacteriaDH5ct high competence cellsUT5600 protease-deficient bacteriaStratagenePharmaciaStratageneStratageneDr. G. Kalmar (SFU)BRLGIBCO-BRLNEB4. ANTIBODY REAGENTSi. Primary Antibody ReagentsAnti-glutathione S-transferase (GST) antibodyPY2O anti-phosphotyrosine antibody4G 10 anti-phosphotyrosine antibodyCRB anti-Pim- 1 antibodyPim-CT (C2)Pim-CT (DaniellA2)Tel, anti-Pim- 1 antibodyMolecular probesSanta CruzUBICambridge Research BiochemicalsDr. M. Lilly (Seattle VA Hospital)Dr. M. Lilly (Seattle VA Hospital)Drs. A. Telerman and R. Amson,(CEPH, Paris, France)ii. Secondary Antibody ReagentsBlotting grade affinity purified goat anti-rabbitIgG (H+L) alkaline phosphatase conjugateETA grade affinity purified goat anti-mouse IgG(H+L) alkaline phosphatase conjugateAffinity purified rabbit anti-sheep IgG (H+L)alkaline phosphatase conjugateGoat anti-rabbit IgG (H+L) horseradish peroxidase conjugateBio-RadCalbiochemBio-RadAmersham49iii. Unique Antibody ReagentsUnique antibodies created in this laboratory are listed below, and described in extensivedetail in Appendixes I and II.Piml-1 11 Anti-human and murine Pim-1 (Available from UBI)Piml-NT Anti-human and murine Pim-1 (Available from UBI)Piml -Xl Anti-X. laevis Pim- 1 peptide (Available from UBI)GXP Against X. laevis GST-Pim-l fusion protein5. SOURCES OF OLIGONUCLEOTIDESA detailed description of all oligonucleotides is located in Appendix III.Oligonucleotides Ki and K2 were synthesized on an Applied Biosystems 392 DNAsynthesizer by Georgia Tai at the Biomedical Research Centre, UBC. Oligonucleotides11 A, 1 2A, 1 3B, 14B, 9205, 9204, were synthesized by core technicians at the BiomedicalResearch Centre. Oligonucleotides were obtained bound to a column and requiredcleavage and deprotection.Oligonucleotides Pim-Y, Pim-5’, PM1, PM2, PM3 and PM4 were synthesized in thelaboratory of Dr. Gabe Kalmar (Simon Fraser University, Burnaby) on an AppliedBiosystems 392 DNA synthesizer. These oligonucleotides were obtained in adeprotected and lyophilized state, ready to hydrate and use.6. PEPTIDE PRODUCTION/SOURCESThe sequence of peptides Piml-1 11, Piml-NT and Pim-X1 are provided in AppendixI. The sequences of the substrate analog peptides are detailed in Appendix IV.Peptides Pim 1-111 and Pim 1-NT, used for antibody production and well as thoseused for kinase assays were synthesized on an A.B .1. 430A peptide synthesizer in thelaboratory of Ian Clark-Lewis (Biomedical Research Centre), and cleaved from the resinby hydrofluoric acid. Purity of the peptides was demonstrated by reverse phase HPLC,and identity was confirmed by ion spray mass spectrophotometry analysis (model APIIII). Due to the small size of the peptides and the high purity as confirmed by reversephase HPLC, the kinase substrate peptides did not require further purification after50lyophilization. Peptides Piml-1 11 and Piml-NT were further purified by HPLC andlyophilized. Peptides were carefully dried before weighing on an analytical balance. Allpeptides were readily soluble in assay dilution buffer or water.Peptide Pim-X1 was synthesized by contract through UBI.7. SOURCES OF CELL LNES/ CELL LYSATESPrimary human lymphocytes in culture were purified by Dr. Bill Sahi in ourlaboratory. K562 cells were kindly grown by Ms. Helen Merkins (Biomedical ResearchCentre).8. ADDITIONAL REAGENTSPhosphatase HPTPB was a kind gift from Mr. Ken Harder (Biomedical ResearchCentre). The 40S ribosomes were a gift from Dr. J. McNeil’s Lab (Dept. of Pharmacy,UBC). Lck was a generous gift from Dr. Julian Watts (Biomedical Research Centre).S6 kinase was a kind gift from Ms. Lorin Charlton (Pelech Lab), and GST-Raf-l was akind gift from Mr. Dan Leung (Pelech Lab).512. EXPERIMENTAL PROCEDURES - MOLECULAR BIOLOGY1. GENERAL MOLECULAR BIOLOGY TECHNIOUESi. Isolation of PCR bands from an agarose gelPCR products were separated on 0.75-1.5 % agarose gels in TAE buffer (10 mM Trisbase, 200 mlvi EDTA, glacial acetic acid (1.142 ml per litre)), were visualized byethidium bromide staining and photographed on a shortwave UV light. Sizedetermination was by comparison to 1 kb DNA ladder. DNA bands of interest wereexcised from the gel with a scalpel blade and the DNA was recovered from the agaroseby electrolution or by Sephglas or Geneclean systems.For electroelution, the DNA-containing agarose band was inserted into a piece ofdialysis tube with several ml of TAE buffer and clamped. The dialysis tube waselectroluted in the gel box at 100 mA for 30 mm. The current was reversed for 30 s, theTAE was carefully removed from the dialysis bag and the DNA precipitated with 3 vol ofice-cold ethanol.The Sephglas and Geneclean kits both relied on the affinity of DNA for a novel glassmatrix. The procedure was followed as recommended by the manufacturer, with thereagents supplied.ii. Restriction digestsFor restriction digests, 0.5 - 1 p.g of DNA were digested in a 20 p.1 reaction with 2 p.1of the appropriate buffer and 1-2 p.1 of enzyme. Each p.g of DNA should ideally bedigested in a 10 p.1 volume. Digests were performed for 1- 2 h (BamHl) or 0/N at 37°Cexcept for Smal digestions which were performed at 16°C. After the digests werecomplete, 2 p.1 of agarose gel loading buffer [Maniatis et al., 1989, Section 6.12. Type 2:0.25% bromophenol blue (wlv), 0.25% Xylene cyanol FF (wlv) and 30% glycerol (v/v)]were added to each reaction, the DNA was subjected to electrophoresis in an agarose gel,visualized by ethidium bromide staining and was photographed on a shortwave UV light.PCR reactions were assessed for the presence of DNA bands of the expected size and forthe amount of background amplification. The PCR band of interest was excised andpurified by electrolution, Geneclean or Sephglas system. The DNA fragment was thenused for ligation, probe construction or for further restriction digests.52The P. ochraceus and X. laevis pim-1 PCR products were digested with Smal tocreate blunt ends. Bluescript plasmid was likewise digested to create compatiblecohesive ends and treated with alkaline phosphatase.iii. Alkaline phosphatase treatment of linearized plasmidsLinearized plasmids were treated with alkaline phosphatase to remove the 5-phosphate group, preventing the plasmid from ligating to itself with the exclusion of thePCR fragment. Approximately 30 .tl of plasmid were incubated with 1.0 jil dilutedalkaline phosphatase (0.024 units/jil, 1/1000 dilution) in a total volume of 50 .tl of 1Xalkaline phosphatase buffer (provided by manufacturer) for 30 mm at 37°C. Oneadditional il of alkaline phosphatase was added and the reaction continued for anadditional 30 mm. The reaction was terminated by the addition of 0.5 mM EDTA (pH8.0) and incubated at 70°C for 20 mm. The volume was increased to 200 jil and theplasmid extracted sequentially with phenol:chloroform (1:1; v/v) and chloroform.iv. LigationsLigation reactions were performed in a final volume of 10 jil, containing 1 pJ ofligation buffer, 1 l of T4 DNA ligase, and varying concentrations of insert, plasmid(cleaved) and ddH2O water. For blunt-ended ligations, the amount of T4 DNA ligase wasincreased to 2-4 p.1 of enzyme per reaction. The ligations were performed for 16-48 h at16°C. Various controls were performed including insert alone (control for ligated orundigested plasmid contamination), plasmid alone with no ligase (control for spontaneousantibiotic-resistant bacterial revertants) and supercoiled plasmid (control for bacterialcompetence).v. TransformationsBacteria were made competent using a modified CaCl2 method of Maniatis et al.[1989, Section 1.82]. E. coli bacterial strains DH5c and UT5600 (protease deficient)were grown until an optical density of 0.4-0.6 (600 nm) was achieved. Aliquots (50 ml)were centrifuged and the pellets resuspended in 20 ml of ice cold 50 mM CaC12, andincubated on ice for 20 mm. The cells were pelleted and resuspended in 2.5 ml of icecold 50 mM CaCl2containing 20% glycerol, aliquoted and frozen immediately at -70°C.Transformations were carried out as described in Maniatis et al. [1989, Section 1.83].In brief, 0.5 p.1 of supercoiled plasmid or diluted ligation mixture were added to thecompetent bacterial solution and incubated on ice for 15 mm. For ligations, 90 p.1 of53freshly prepared 10-10-10 buffer (10 mM Tris-Ci, pH 7.5, 10 mM MgC12, 10 mM CaC12)were added. The solution was then gently agitated at 45°C for 90 s, then incubated on icefor 5 mm. Antibiotic-free media (2xYT) was added and the bacteria was incubated at37°C for 20 mm to induce ampR gene expression. The transformed bacteria was thenplated onto 2xYT(amp+) plates and grown at 37°C 0/N.Isolated colonies were selected and restreaked or used to inoculate an 0/N culture.For construction of the PCR probes, XL1-blue E.coli was used for all the transformationsand positive colonies were selected by blue/white selection on LB-amp plates. A whitecolony indicated that the 3-galactosidase gene was interrupted by the insertion of afragment of DNA, and was selected as potentially positive.vi. Small scale plasmid preparationThe small scale plasmid preparation protocol was that developed by He et a!., [19891.In brief, 0/N cultures of bacteria were grown in media with ampicillin (100 j.tg/ml).Bacteria were pelleted by centrifugation at maximum speed for 2 mm in an Eppendorfcentrifuge and resuspended in 200 111 of TELT (2.5 M LiC1, 50 mM Tris-HC1 (pH 8.0),4% Triton X-100 (v/v), 62.5 mM EDTA). The plasmid was extracted withphenol:chloroform (1:1; v/v) and precipitated with 500 il of 100% ethanol. The pelletwas washed with 70% ethanol, air dried for 10 mm and resuspended in 15-20 tl of TEbuffer (10 mlvi Tris (pH 8.0), 1 mlvi EDTA (pH 8.0)). Restriction digests were thenperformed using 5 tl of the DNA solution per digest.vii. Large scale plasmid preparationLarge scale preparations were initially done using a protocol obtained from Dr. F.Jirik (Biomedical Research Centre). Later (1994 and beyond), plasmid preparations wereperformed using the Quiagen Midi prep kit, using the protocol and reagents supplied.The procedure recommended with the kit is based on a modified alkaline lysis procedure.The alkaline lysis protocol obtained from Dr. F. Jirik is briefly detailed. A 500 mlculture of transformed bacteria was grown 0/N in LB media containing ampicillin (0.05mg/ml). The bacteria was pelleted by centrifugation at 4420 x g for 10 mm. Thesupernatant was discarded, the pellet was resuspended by intense vortexing and was kepton ice for the remainder of the procedure. Seven ml of glucose solution (25 mM TrisHC1, 10 mM EDTA, 50 mM glucose) at 4°C were added to the pellets and mixedthoroughly. In a 50 ml conical centrifuge tube, 14.0 ml of 0.2 N NaOH/1% SDS solution54were added, mixed by gentle inversion and incubated on ice for 10 mm. Fourteen mis ofKOAc solution (pH 5.7, 60 ml of 5M KOAc, 11.5 ml glacial acetic acid, 28.5 ml dH2O)were then added, mixed by gentle inversion and incubated on ice for 10 mm. Thepiasmid solution was centrifuged twice at 7250 x g for 10 mm, 4°C. The supernatantwas then extracted with an equal volume of phenol/chloroform (1:1; v/v), vortexed andcentrifuged for 10 mm at 5000 x g, at RT. The aqueous layer was extracted with an equalvolume of isopropyl alcohol and incubated on ice for at least 15 mm with occasionalinversions. The plasmid was pelleted by centrifugation at 12000 x g for 15 mm at 4°Cand air-dried for 20 mm. The plasmid pellet was resuspended in 200 il of TE buffer and5 j.ti of RNAse A (10 mg/mi) were added. After a 20 mm incubation at 37°C, the solutionwas extracted three times with phenol/chloroform (1:1; v/v) and centrifuged at 16000 x gfor 5 mm. The piasmid was precipitated for 20 mm at -20°C by the addition of 1 ml of100% ethanol and centrifuged for 30 mm at 16000 x g. After briefly drying in adessicator, the plasmid DNA was resuspended in 500-1000 j.tl of lx TE buffer andquantitated by measuring the 0D260 (0D260 x dilution x 42/50 = j.tg / ml).viii. Construction of a pim-] probe: labeling a DNA probe with[32P1 ATPAmplified pim-] PCR fragments were excised from pBiuescript with Smal andlabeled with y—32P, by a method obtained from Ms. Nicole Janzen (BRC). Theradiolabeled DNA fragment used as a probe to screen Northern blots, Southern blots andthe P. ochraceus and X. laevis cDNA libraries.For each probe, 100 ng of double stranded DNA in a volume of 9 jil were denaturedby boiling for 5 mm then cooled immediately on ice. The labeling reaction was carriedout for 30 mm at 37°C in a final volume of 20 j.il with 2.0 .tl of lOX Kienow buffer(supplied by GJBCO-BRL (400 mM Tris-Ci, pH 7.5, 66 mM MgC12, 10 mM 2-ME), 250.tM each of dATP, dTTP, and dGTP, 5.0 !IM of random hexamer, 1.0 j.U of DNApolymerase large fragment (Kienow) and 5.0 p.1 of [“y—32P]dCTP. Addition of 60 p.1 ofbuffer (10 mM Tris-HCL, pH 7.5, 10 mM EDTA, 0.5% SDS), 80 p.1 of 5M NH4OAc, 4p.1 tRNA (10 mg/mi), and 400 p.1 of 95% ethanol was followed by centrifugation for 15mm at RT. The probe was dissolved by boiling in 100 p.1 of dH2O for 5 mm, brieflychilled on ice, then added to hybridization solution.ix. Prehybridization and hybridization of membranes with radiolabeled probeMembranes were prehybridized for 30 mm at 55°C in hybridization buffer (350 mMNaPO4, 30% deionized formamide (v/v), 7% SDS (w/v) and 1% BSA (w/v)). The55labeled probe was hybridized to the membranes in a volume of 10 ml of hybridizationbuffer for 16 h at 55°C. The membranes were subjected to multiple washes with 150 mMNaPO4,0.1% SDS (wlv) for 10 mm at 55°C, until the radioactive counts emitted from themembranes was significantly reduced when measured with a hand-held Geiger counter.For more stringent washes, 50 mM NaP, 0.1% SDS (w/v) was used. When counts weresufficiently reduced, the membrane was air dried and autoradiographed.2. OOCYTE MATURATIONi. Isolation of Xenopus laevis oocytesMature female Xenopus laevis (African clawed frog) were immersed in a 3% Finquel(tricaine methanesulfonate in tap water) solution for approximately 10 mm, until noresponse was elicited by gentle pinching of feet and claws. As the LC50 of Finquel is 30mm in a 6.2% solution (wlv) [Argent Chemical Laboratories’ specification sheet], frogswere monitored carefully during the procedure to ensure that they were not overanaesthetized. Ovaries were surgically removed, and cardial puncture was perfomed.Ovaries were washed by gentle swirling in tissue culture plates with 1X X. laevis oocytemedia (XOM) containing 5.4 mM Tris-base, 86 mM NaC1, 0.8 mM KC1, 0.5 mM CaC12,0.6 mM MgSO4and 4.0 mM sulphadiazine (pH 7.6) [Zhang and Masui, 1992]. The sack-like covering of the ovary was gently pulled apart to expose eggs and ovary material tothe media and to wash away crushed eggs.The oocytes were isolated from the ovaries using a modified method [Belle et al.,1986] by digestion with dispase (0.04% (w/v) in XOM) for 4 h at RT with occasionalgentle swirling. The oocytes were washed with buffer, then digested in collagenase(0.1% (w/v) in XOM)) for 2 h at RT. Media was replaced when cloudy.Stage VI oocytes displaying definite bipolar pigmentation were selected by visualinspection, and placed in fresh media. Oocytes that were mottled or had an irregularshape were discarded. Isolated oocytes were stored 0/N in sulphadiazine-containingmedia at room temperature.ii. Proestèrone maturation of X. laevis oocytesOocytes were matured at ambient temperature with 100 j.tM of progesterone. Freshmedia was added to oocytes, and a 1/1000 dilution of progesterone in ethanol (100 mM56stock) was added. Germinal vesicle breakdown (GVBD) became visible between 4-9 h,as evidenced by the appearance of a symmetrical white spot in the middle of the darkhemisphere. Mature oocytes were selected and set aside for homogenization. Oocytesthat became cloudy or mottled were discarded.iii. P. ochraceus oocyte maturationOvaries were surgically removed from the arms of P. ochraceus (purple sea stars) andincubated in calcium free artificial sea water, CaFASW (475 mM NaC1, 10 mM KC1, 31mM MgC12.6H0, 18 mM MgSO4, 10 mM Tris) on ice. Ovaries were gently teased apartwith forceps to release the oocytes, and were strained through a large mesh to removeconnective tissue and residue. Oocytes were washed three times in cold CaFASW bypelleting the oocytes by centrifugation at 400 x g for 5 mm in the Beckman centrifuge.After the third wash, the oocytes were resuspended in natural sea water (NSW)containing 4 jiM 1-methyladenine at 14°C. Oocytes were allowed to mature by gentlestirring at 14°C for 70-100 mm.Maturation was achieved by the onset of GVBD, as evidenced by the disappearanceof the nucleus within the oocyte when viewed under high magnification. Mature oocyteswere harvested when GVBD occurred in over 80% of the oocytes, or 2 h after theinitiation of maturation. For the maturation time course, measured volumes of oocytesuspension were removed at discrete time points after the addition of 1 -methyladenine.3. ISOLATION OF TOTAL RNA FROM OOCYTESi. Homogenization of oocytes for total RNATotal RNA was isolated from immature and mature stage VI X. laevis and from P.ochraceus oocytes by the method of Maniatis et al. [1989, Section 7.16]. All glasswareincluding the homogenizer was washed with 0.1 M NaOH and rinsed several times withddH2O to remove RNAses. Approximately 1 ml of oocytes were homogenized withabout 10 volumes of homogenization buffer (50mM NaCl, 50 mM Tris-HC1 (pH 7.5), 5mM EDTA (pH 8.0), 0.5% SDS (w/v) and 200 jig/mi proteinase K). The homogenatewas incubated at 37°C for one h and then extracted with phenol:chloroform (1:1; vlv).Phases were separated by centrifugation at 1480 x g for 10 mm in the Beckmancentrifuge at RT, the upper phase was removed and re-extracted with phenol: chloroform(1:1; v/v) and a third extraction was performed with chloroform to remove any traces of57phenol. The aqueous phase was transferred to a fresh tube and 0.1 vol (1.1 ml) of 3 MNaOAc (pH 5.2) and 25 ml of ice cold ethanol (95%) were added, then the solution wasincubated on ice for 2 h.After centrifugation at 5000 x g for 15 mm at 4°C, the supematant was discarded andthe pellet was briefly air dried. Pellets were resuspended in 5 ml of ddH2O, 5 ml of 7.44M LiC1 were added and the pellets were stored at -20°C 0/N. The RNA was pelleted bycentrifugation in 2059 tubes at 9800 x g for 30 mm at 4°C, and washed with cold 70%ethanol. The pellet was briefly air dried and resuspended in 2 ml of ddH2O and 3 vol of100% ethanol, with the addition of 20 jil of RNAsin. The RNA was stored at -70°C untiluse.To recover the RNA, a volume of the RNAlethanol solution was removed, 0.1 vol of3 M NaOAc (pH 5.2) was added and mixed and the solution was centrifuged at 4°C for 5mm in an Eppendorf centrifuge. The supernatant was discarded, the RNA pellet washedonce with 70% ethanol, dried very briefly and resuspended in the buffer of choice.ii. Ouantitation and purity assessment of RNATo quantitate the amount of RNA in a purified sample, an aliquot of the ethanolsolution was withdrawn and recovered as described above. The pellet was briefly airdried and dissolved in ddH2O. The absorbance of the solution at 260 nm was measured,one optical density unit contained 40 jig of RNA per ml. The value 44.19 corresponds tothe extinction coefficient for RNA.[RNAJ jig/mi = A260 x 44.19 x dilution factorThe ratio of the optical density of 260/280 was obtained to determine the purity of theRNA, with a value of 2.0 being optimal.iii. Selection of polv(A)+ RNAPoly(A)+ RNA for reverse transcription reactions was selected using affinitychromatography on oligo(dT)-cellulose as described by Maniatis et at. [1989, Section7.26] with minor modifications. In brief, the RNA was dissolved in ddH2O and an equalamount of 2X column loading buffer (40 mM Tris-HC1 (pH 7.6), 1.0 M NaC1, 2 mMEDTA (pH 8.0), 0.2% SLS (w/v)) was added to the RNA. The RNA was not alwaysheated at 70°C to dissociate.58Oligo(dT)-cellulose (0.1 g) was hydrated with 0.1 N NaOH, and poured into a plasticcolumn that was prewashed with 0.1 M NaOH. The volume of the column varied from250- 500 j.iJ (each ml of resin bound 10 mg of RNA). The column was first washed with5 column volumes of ddH2O, then washed with 1X column loading buffer (20 mM TrisHC1 (pH 7.6), 0.5 M NaC1, 1 mM EDTA (pH 8.0), 0.1% sodium lauryl sulphate) until thepH of the effluent was less than 8.0. An alternate recipe for column loading buffer thatwas used was 20 mM Tris-HC1 (pH 7.5), 1.0 M LiC1 and 2 mM EDTA.The RNA containing solution was applied to the column, followed by one vol of 1Xcolumn loading buffer. The column flow-through was collected and reapplied to thecolumn. The column was washed with 10 vol of 1X column loading buffer and theremoval of nonpolyadenylated RNA was monitored by reading the absorbance of thecollected fractions at 260 nm.The poly(A)+ RNA was eluted from the column with 2-3 volumes of sterile RNAsefree elution buffer (10 mM Tris-HC1 (pH 7.6), 1 mM EDTA (pH 8.0) or alternately, 2mM EDTA (pH 7.5)). SDS was not added to the elution buffer as recommended byManiatis et al. [1989]. Fractions (0.3-0.5 vol) were collected, assessed by reading theabsorbance of the RNA at 260 nm, and the peak fractions were pooled. NaOAc (pH 5.2)was added to the poly(A)+ RNA to a concentration of 0.3 M, 2.5 vol ice-cold ethanolwere added, and then the RNA was stored at -70°C until use.To recover the mRNA, the RNA was centrifuged at 10000 x g for 15 mm at 4°C, andthe pellet was washed with 70% ethanol. The pellet was air-dried, and the RNAresuspended in a small volume of ddH2O and quantitated.Poly(A)+ RNA for Northern blots was isolated using Dynal Oligo(dT) beads with therecommended protocol. The protocol is similar to this except that the oligo(dT) isadsorbed to metal beads instead of cellulose, allowing very rapid separation of thepoly(A)+ RNA in a very small volume. All buffers used were provided in the kit.Approximately 75 jig of total RNA were purified with 200 jii (1.0 mg) of beads.iv. Northern blot analysis of oocyte RNAX. laevis and P. ochraceus RNA was analyzed by Northern blotting for the presenceof pim-1 mRNA. Before commencing, all equipment including the gel box and combwere rinsed with 0.1 M NaOH and ddH2O to minimize RNAse contamination. Total59RNA (20 tg) and poly(A)+ RNA (0.5-3 j.ig) were recovered from ethanol andresuspended in 4.7 jil of ddH2O. To each sample of RNA, 3.3 il of 37% formaldehyde(2.2 M final), 10 p1 of formamide and 2.0 p1 of lOX MOPS buffer (0.2 M MOPS (pH7.0), 50 mM NaOAc, 10 mM EDTA (pH 8.0)) were added and the sample was heated at55°C for 15 mm. RNA standards were treated the same as samples, with 3 p1 used in thecontrol lanes. Agarose gel loading buffer was added to the samples directly beforeapplication onto the gel. The RNA samples were applied onto a dry gel (1.2% agarose,1X MOPS, 0.66 M formaldehyde) and the gel was electroluted for 10 mm at 90V. Thegel was then flooded with 1X MOPS buffer and electroluted at 100V. After electrolution,the gel was soaked in 1X MOPS buffer to remove formaldehyde. The RNA was nickedby exposure to 320 nm UV light for 2 mm.The Northern blot was assembled as described in Maniatis et al. [1989, Section 7.46].Briefly, the gel was placed on a wick made from three pieces of Whatmann 3MM paper,and soaked in 20X SSC (333 mM NaCl, 930 mM NaCitrate, pH 7.0). After a briefhydration in ddH2O, a piece of nylon Hybond membrane was soaked in lox ssc for 5 to10 mm and placed on the gel without trapping air bubbles. Six sheets of Whatmannpaper soaked in 20X SSC were placed on top of the nitrocellulose, followed by a 6 cmstack of paper towels. A glass tray and a 500 g weight were placed on top and the gelwas left to blot 0/N in the fumehood. The RNA was crosslinked to the air-driedmembrane in the Stratalinker. The standards and the 18s and 28s bands were visualizedusing the hand-held crosslinker and the standards marked with a pencil.The Northern blot was prehybridized and hybridized with the same probes as detailedin the library screening section. The blots were stripped and reprobed several times,using 15 mM NaP and 1% SDS (w/v), for 30 mm at 70°C.4. AMPLIFYING PIM-1 USING POLYMERASE CHAIN REACTION (PCR)i. Reverse transcriptase reactionThe cDNA synthesis reaction was carried out using approximately 1-2 jig of X. laevisor P. ochraceus mRNA as a template with 0.5 p1 RNAsin (20 units). For each reaction,RNA in a volume of 10 jil (ddH2O) was denatured at 45-50°C for 2-3 mm, then incubatedon ice as the remaining reagents were added. The reverse transcriptase reaction with 1.0j.il Superscript reverse transcriptase, 100 jiM of each nucleotide triphosphate, 1 jig of60random primers, was carried out in a total volume of 20 p1 of polymerase chain reaction(PCR) buffer (50 mM KC1, 10 mM Tris-Ci (pH 8.3), 1.5 MgC12,0.01% gelatin (wlv)).The reaction was performed in a Perkin-Elmer Cetus thermal cycler at 23°C for 10 mm,42°C for 45 mm, 94°C for 3 mm and 4°C for 5 mm.For large scale PCR, the reverse transcriptase reaction was scaled up. For the scaledup reaction, 20 jig of mRNA in 100 jil of dH2O were denatured for 3 mm at 45°C thenbriefly incubated on ice. The reverse transcriptase reaction was carried out with 10 jil ofSuperscript reverse transcriptase, 100 jiM of each nucleotide triphosphate, 10 jig ofrandom primers, in a total volume of 200 p1 of PCR buffer (50 mM KC1, 10 mM Tris-Cl(pH 8.3), 1.5 MgCl2,0.0 1% gelatin (wlv)), at 23°C for 10 mm, 42°C for 45 mm, then at94°C for 3 mm and 4°C for 5 mm.In place of the random primers, oligo dTTT (ljiglj.tl) as well as oligo 14B (1.756j.tg/jil) were substituted as primers for the reverse transcription reaction. PCR usingcDNA prepared in this manner yielded a high background to product ratio, so this methodof cDNA preparation was discontinued.ii. Cleaning the cDNAProducts of large scale reverse transcription reactions were cleaned on a SephacrylHR 300 column to remove the random primers from the cDNA. Sephacryl-HR 300 wasused to fill a 5 ml syringe plugged with glass wool, and packed by centrifugation in aBeckman at 200 x g for 1 mm. The column was equilibrated with TE buffer under thesame conditions. The cDNA was loaded on the column and the column centrifuged in aBeckman at 200 x g for 1 mm. The eluate was collected and used as a template forpreparative PCR reactions.iii. Deprotection. cleavage and working up oligonucleotidesThis protocol was used only on oligonucleotides that were obtained bound to acolumn. The column was attached to a 1 ml syringe and to a G18 needle using a male-male luer connector. A small amount of fresh ammonia was drawn through the columnsuch that the level of the fluid barely entered the bottom of the syringe. The needle wasjabbed into rubber bungie and left to deprotect for 30 mm at RT, in the fume hood. After30 mm the crude oligonucleotide solution was expelled into a 1.5 ml screw top, 0-ringtube. Ammonia was again drawn through the column and left to deprotect once for 30mm, then twice for 15 mm. The oligonucleotide-containing tube was then capped, sealed61with parafilm and incubated in a 45-55°C water bath 0/N to cleave the oligonucleotide.The solution was briefly cooled at -20°C, for 15 mm, then the fluid dried under vacuum.The dried oligonucleotides were resuspended in 120 ul of 1X STE buffer (100 mM NaC1,20 mM Tris-Ci (pH 7.5), 10 mM EDTA (pH 7.5)).Oligonucleotides were purified on a G-50 spin column. G-50 beads were carefullyloaded into a 1 ml syringe plugged with glass wool and packed by centrifugation in a2059 tube in a bench top Clinical Centrifuge (International Equipment Company) atsetting “3” for 3 mm. The G-50 was topped up to a volume of 1 ml and the centrifugationrepeated. The fluid volume of the column was equilibrated by adding 120 p.1 of 1X STEto top of column and centrifuging as before. The STE was retrieved from the 2059 tubeand quantitated with a pipetteman to ensure that the entire volume was recovered. Ifnecessary, an additional 120 -300 p.1 of STE were added to top of column and centrifugedas before.The entire 120 p.1 of oligonucleotide solution were applied to the top of the column,and the column was centrifuged at setting “3” for 3 mm. The purified oligonucleotidewas collected at the bottom of the syringe column in a screw-top tube. A secondcentrifugation with an additional 120 p.1 of STE buffer was done to remove the remainderof the oligonucleotide from the column and was collected in a second screw-top tube.iv. Ouantitation of oligonucleotidesThe concentrations of oligonucleotides in the primary and secondary tubes weredetermined by measurement at 260 nm in a spectrophotometer. A small amount of theoligonucleotide solution was diluted in dH2O and measured at 260 nm. Each opticaldensity unit equaled a concentration of 1 jig/mI of oligonucleotide.[oligo]jig/mi = 0D260 x 20 x dilution factor (1000)The expected yield was 1-2 jig/p 1 for the primary tube and less for the secondary tube.Conversion of jig/jil to mol of oligonucleotides: [oligonucleotides (g/p.l)J x 1/325(number of bases)= mol. (10-’ 1 mol = pMolIp.l)v. Specific PCR reaction conditions for amplifying pim-1Pim-i PCR reactions were designated as A (oligos 1 1A to 13B), B (oligos 12A-13B),C (oligos 11 A-i 4B) and D (oligos 1 2A- 1 4B). Oligonucleotide sequences are listed inAppendix III and are based on the human and murine pim-1 sequences [Berns et al.,621988a]. The expected size of the PCR products were 669 bp for reaction A, 423 bp forreaction B, 798 bp for reaction C and 552 bp for reaction D.Approximately 50-100 pmols of each oligonucleotide were added to each reversetranscriptase reaction (20 p.1 of single stranded X. laevis or P. ochraceus cDNA template)in a total volume of 49 p.1 in 1X PCR buffer. The reaction was overlaid with 1-2 drops ofPCR oil, heated to 96°C for 1 mm, then cooled immediately on ice. Taq polymerase (1p.1) was added to the reaction and the tube was centrifuged briefly and the reactioninitiated. For negative controls, 20 p.1 of dH2O were used in the place of the template.Initial PCR reactions were performed at low stringency conditions, with initiation at96°C for 1 mm, annealing at 37°C for 90 s and elongation at 73°C for 3 mm. Thereaction was repeated through 29 cycles. To optimize the reactions, five different PCRreactions were carried out at annealing temperatures of 37°C, 4 1°C, 46°C, 50°C and55°C. The reaction times and the rest of the conditions were as above.Large scale preparative reactions for both P. ochraceus and X. laevis were performedfor reaction A only, using 20 p.1 of reverse transcriptase template, 3 p.1 of lOX PCRbuffer, 1.1 p.1 of oligo hA, 1.1 p.1 of oligo 13B, 23.8 p.1 of dH2O and 1 p.1 of Taqpolymerase. The reactions were carried out for 29 cycles with an initiation temperatureof 94°C for 35 s, annealing at 50°C for 90 s and elongation for 73°C for 90 s.vi. Confirmation of identity of PCR clonesThe PCR clones in pBluescript were first examined by restriction digest analysis. Intotal, five clones yielded bands of the expected size, and orientation was determined bythe restriction analysis. The two X. laevis pim-1 clones (in opposite orientation) weredesignated as 3A and 3B, and the three P. ochraceus pim- 1 clones were designated as 5A,SB, and SC. The identity of the PCR clones was confirmed by sequencing, using adouble template protocol as dictated by the Pharmacia kit. Details are listed in thesection following library screening. The identity of the clones was confirmed ascorresponding to pim-1.635. SCREENING A ZAP eDNA LIBRARYi. Screening the X. laevis eDNA libraryA X. laevis oocyte cDNA library in ZAP (EcoRl, Xhol) was kindly obtained fromDr. Leonard Zon (Childrens Hospital, Boston), at a titre of 4 x 10 pfu/pi. The librarywas retitred at 1.13 x 10 pfu/.tl. The library was screened using a protocol modifiedfrom Stratagene.Competent cells were prepared by inoculating a 50 ml culture of LB media containing0.2% maltose (w/v) and 10 mM MgSO4 with a single colony of E.coli XL1-blues andgrown at 37°C 0/N. The cells were pelleted by centrifugation at 2000 rpm for 10 mm,and resuspended in 15 ml of 10 mM cold MgSO4. The cells were stored at 4°C for up toone week.The competent XL 1-blue cells were transformed with an appropriate dilution ofphage library by incubation at 37°C for 20 mm with agitation and plated onto dried 2xYTplates at approximately 2.2500 - 2.5 x iO plaque forming units (pfu) per plate. Aftergrowing 0/N, duplicate plaque lifts were performed using Hybond membranes, whichwere oriented with respect to the plate by puncturing with an ink-soaked needle.Membranes were denatured in 1.5 M NaC1, 0.5 M NaOH for 2 mm. Membranes wereneutralized for 5 mm in 1.5 M NaC1, 0.5 M Tris-Ci (pH 8.0) and washed with 2X SSC(pH 7.0, 33.3 mM NaC1, 333 mM sodium citrate). After drying briefly, the DNA wascross-linked to the membrane with a Stratalinker UV linker (Stratagene).Membranes were prehybridized and hybridized with radiolabelled X. laevis pim-1nucleic acid probe by the method detailed in Section 2.1 .ix. Membranes were washedunder increasingly stringent conditions, then exposed to X-ray film 0/N to identifylabeled clones. Potential positives were selected by alignment of the autoradiographs ofduplicate membranes. The potential positives were picked and stored in 500 il of SMbuffer (100 mM NaC1, 17 mM MgSO4,50 mM Tris-HC1 (pH 7.5) and 0.01% gelatin(w/v)) with 20 p.1 of chloroform. As each plaque contained 106107 phages, the titre ofthe phage solution was determined to be between 2 x l0 - 2 x 10 pfu/p.l. Potentialpositive clones were amplified as per Stratagene protocol and screened several times withincreasingly stringent washes (10 - 60 mm at 65°C). For secondary and tertiary screens,the phage solution was diluted to yield approximately one hundred pfu per plate.64The protocol was followed for in vivo excision from the Stratagene InstructionalManual (#236211) for the Predigested Zap IT/EcoRl Cloning Kit. This procedureyielded the pBluescript double stranded phagemid with the cloned insert. ThepBluescript was amplified and the size of the inserts were examined by digestion withEcoRl.Potential positives were first analyzed by Southern blotting to confirm their identity,then sequenced by Dr. G. Kalmar (Simon Fraser University) to obtain the sequence of theentire coding region.ii. Screening the P. ochraceus cDNA libraryTwo P. ochraceus eDNA libraries were obtained from Dr. Michael Smith (SimonFraser University). Both libraries were in gt10; one was a cDNA library from P.ochraceus oocytes, the other from P. ochraceus testes. Both libraries were plated andscreened with the sea star pim-] probe, but did not yield any positive clones. Due to thequestionable quality of this library, the screening was discontinued.iii. Southern blottingThree positive pim-1 clones obtained from the X. laevis oocyte cDNA library werequantitated and 60 ng of each were digested by restriction enzymes. The X. laevis cloneswere digested with Accl/Ncol, Hind IIJJPvuII and an undigested sample was alsoprepared as a control. The digests were separated on a 1% agarose gel in TAE buffer, thegel was stained with ethidium bromide and photographed to visualize the DNA bands.The gel was soaked in denaturation solution (0.4 M NaOH) for 2 x 10 mm at RT, untilthe top dye (Xylene) turned green.The Southern blot was assembled as recommended by Maniatis et al. [1989, Section9.34]. Normally, Southern blots are left 0/N to blot; as we blotted digested plasmidDNA, 4 h was judged to be sufficient. After disassembly, the membrane was washed in2X SSC for 5 mm to rinse away debris, then air dried and crosslinked to the filter underUV light for 3 mm manually. The Southern blot was hybridized with a X. laevis pim-1nucleic acid probe as detailed in Sections 2.1 .viii-ix.65iv. Sequencing positive clones to confirm identityThe nucleotide sequence of X. laevis and P. ochraceus pim-1 PCR fragments weredetermined using the standard dideoxy chain termination method [Sanger et at., 1980].The protocol detailed in the Pharmacia sequencing kit manual, incorporating 35S-dCTPwas utilized. The sequence was visualized by autoradiography and read manually.Initially, oligonucleotides 11 A and 1 3B were tried as sequencing primers, but theydid not yield interpretable sequence. Universal primer (provided with the Pharmacia kit)and oligonucleotides T3 and T7 (see Appendix III) were successfully used forsequencing.The entire coding region of the X. laevis pim-1 cDNA clone was sequenced by Dr. G.Kalmar (Simon Fraser University) by automated fluorescent DNA sequencing using anABI 373A sequencing machine. Sequencing of pim-] mutants was perfomed using thestandard dideoxy chain termination method [Sanger et at., 1980] with custom primers.The sequence of X. laevis pim-1 was entered into the Genome Sequence Database,accession number L29495. A computer search for sequence similarity was performed atthe National Center for Biotechnology Information (NCBI) using the BLAST networkservice.6. CONSTRUCTION OF PIM-1 EXPRESSION VECTORSi. PCR of X. laevis cDNA cloneThe coding region of the X. laevispim-1 was amplified by PCR usingoligonucleotides modeled on the amino and carboxyl terminal regions of the open readingframe of the X. laevis pim-1 sequence. The oligonucleotides, pim5’, 5’-CGA TGG ATCCAT GCT TCT CTC TAA ATT CGG-3’, and pim3’, 5’-GAT CGA ATT CCA GAC TCTCGT TGC TTG A-3’, were designed to incorporate appropriate restriction sites to allowinsertion of the pim-1 PCR fragment in the correct reading frame into the EcoR 1 andBamHl restriction sites of the pGEX-2T vector. The polymerase chain reaction was usedto amplify an approximately 1000 base pair fragment using the X. laevis pim-1 cDNAclone 12.35 in the pBluescript plasmid as a template. Approximately 50 ng of BamHllinearized X. laevis pim-1 eDNA clone were used as a template in a PCR reaction, with75 pmol of each oligonucleotide, 0.2 jil of non-acylated BSA [10 mg/mi], 50 j.tM of eachnucleotide triphosphate and 1 pi of Vent polymerase in a total volume of 20 p.1 in PCR66buffer provided by NEB (10 mM KC1, 10 mM (NH4)2S0,10 mM Tris-HC1 (pH 8.8), 2mM MgSO4,0.1% Triton X-100). The initial PCR reaction was carried out for 25 cycleswith denaturation at 96°C for 45 s, annealing at 50°C for 2 mm and elongation at 73°Cfor 2 mm. Reaction conditions were optimized with respect to time and temperature andthe reaction was scaled up. Preparative PCR reactions were done using approximately 50ng of template, 75 pmol of each oligonucleotide, 0.5 j.il of non-acylated BSA [10 mg/mI],100 tM of nucleotide triphosphates and 2 j.tl of Vent polymerase in a total volume of 50jil in NEB PCR buffer. The PCR reaction was carried out in a Perkin-Elmer Cetusthermal cycler for 25 cycles with denaturation of 96°C for 45 s, annealing at 55°C for 60s and elongation at 73°C for 90 s. The PCR product was visualized on a 1% agarose gel,purified by Sephglas band prep kit. Subsequent restriction digests were done with bothBamHl and EcoRl, with a Sephglas purification between digestion reactions. pGEX-2Twas likewise digested to create compatible cohesive ends, and subjected to alkalinephosphatase treatment.The X. laevis pim-] PCR product and the pGEX-2T were ligated together,transformed into UT5600 E.coli, and plated onto 2xYT amp plates. Colonies wereamplified and digested with restriction enzymes Nco 1/Ava 1 to determine if the correctlysized fragment was inserted, and with BamHl to ensure that the 5’ restriction site hadbeen maintained. This was designated as “Clone 1” and is shown in Figure 2.The pGEX-2T vector containing X. laevis pim-1 was used to transform competentUT5600 bacteria as described in Section 2.1 .v. The fusion protein was expressed andpurified as detailed in the protein biochemistry Section 3.2.i.ii. Construction of pim-1 mutants using PCR site-directed mutagenesisNon-degenerate oligonucleotides were constructed to incorporate specific changes inthe coding region of the X. laevis pim-1 using PCR site directed mutagenesis. Adescription of oligonucleotides used and mutants constructed is detailed below.K69>AA kinase-inactive Pim- 1 mutant was constructed by changing Lys-69, essential forATP binding, to an alanine residue. Anti-sense primer Ki, 5’- CTC CTT AGC TACGTG GCG CAC AGC GAC CGG CTG -3’, corresponding to aa 64-74, contained thelysine codon, TCC (nt 205-207), changed to an alanine codon, GCG (underlined). Senseprimer K2 (aa 75-80) was constructed to allow amplification in the opposite direction.67PvuIIEc0RVFigure 2 Restriction map of X. laevis pim-1 in the pGEX-2T vectorThe map shows the the 969 base pair coding region of X. laevis pim-1 insertedinto the BamHl and EcoRl cloning sites (shown in bold type) of the pGEX-2Tplasmid. The Clal site used for construction of the ser-190 mutant is italicized.The unfilled arrow II] shows the location of the AmpT gene, the grey shadedarrow shows the location of the lac F! gene , the hatched box re resentsthe glutathione S-transferase coding region and the black box is thehuman pim-1 coding region. The line arrow —01 shows the direction of transcription of the fusion construct.BamHI (657)BgIII (733)(799)(1632)68These oligonucleotides were intended to be used with the existing 5’ and 3’ X. laevisoligonucleotides. Unfortunately, the oligos Kl and K2 were not phosphorylated so thePCR products would not ligate together. Instead, primer Ki was used with the 5’ senseprimer to first amplify a 232 base pair fragment from BamH 1-linearized X. laevis WTpim-1 in a PCR reaction containing approximately 50 ng of template, 100 pmol ofoligonucleotides, 0.5 p.1 of non-acylated BSA [10 mg/mi], 100 p.M of nucleotidetriphosphates and 2 p.1 of VENT polymerase in a total volume of 50 p.1 in PCR bufferprovided by NEB (10 mM KC1, 10 mM (NH4)2S0,10 mM Tris-HC1 (pH 8.8), 2 mMMgSO4,0.1% Triton X-l00). The PCR reaction was optimized with respect to time andtemperature and was carried out in a Perkin-Elmer Cetus thermal cycler for 20 cycles;96°C for 45 s, 50°C for 60 s and 73°C for 60 s. The PCR fragment containing theK69>A mutation was prepared in large amounts by PCR, visualized on an agarose geland purified by Sephglas. This PCR product was then used as an oligonucleotide toamplify the full-length coding region of pim-1 with the 3’ anti-sense oligonucleotide. ThePCR reaction was carried out in a total volume of 100 p.1 with approximately 100 ng oftemplate, 100 pmol of 3’ anti-sense primer, 5 p.1 of unquantitated 232 base-pair PCRproduct, 1.0 p.1 of non-acylated BSA [10 mg/mi], 100 p.M of nucleotide triphosphates and7 p.1 of Vent polymerase in NEB PCR buffer. The PCR reaction was optimized withrespect to time and temperature and was carried out in a Perkin-Elmer Cetus thermalcycler for 25 cycles: 96°C for 45 s, 56°C for 60 s and 73°C for 90 s. After purification,the approximately 1 kb fragment was digested with EcoRl and BamHl and ligated intothe corresponding sites of pGEX-2T (Pharmacia). The clone was sequenced by Dr. GabeKalmar to confirm that the mutation was as expected.S190>ASense primer PM1 (1.125 p.1), 5’- CTGATC GATTTT GGC GCC GGG GCG CTACTC -3’ was used with pim3, the 3’ anti-sense primer (4.5 p.1) to amplify a 415 base pairfragment that was used to replace the Clal/EcoRl fragment of WT X. laevis pim-1.Mutated base pairs are underlined. Conditions are listed below.S190>ESense primer PM2 (0.875 p.1), 5’ - CTG ATC GAT TTT GGC GAA GGG GCG CTA CTC-3’ was used with pim3, the 3’ anti-sense primer (4.5 p.1) to amplify a 415 base pairfragment that was used to replace the Clal/EcoRl fragment of WT X. laevis pim-1.Mutated base pairs are underlined. Conditions are listed below.69Y198>ASense primer PM3 (3.6 j.tl) 5’ - GGA TAC GGT GGA AAC GGA TTT TGA TGG -3’was used with pim3, the 3’ anti-sense primer (4.5 jil), to amplify an approximately 400base pair fragment containing the desired mutation (underlined). A second primer, PM4(2.7 jii), 5’ - TTG AGT AGC GCC CCG GAG CC -3’ was used with the 5’ sense primerpim5 (4.5 j.tl), to amplify a 560 base pair fragment. The two PCR products were ligatedtogether and inserted into the BamHl and EcoRl sites of pGEX-2T.The PCR reaction conditions used to create the S190 and Y198 mutants wereperformed as follows. Approximately 25 jig of BamHl linearized X. laevis pim-1 cDNAclone were used as a template, with the amounts of oligonucleotides specified, 1 111 ofBSA (10 mglml), 100 jiM dNTPs, and 5 j.tl VENT polymerase in a total volume of 100 p.1of NEB PCR buffer. The PCR reaction was optimized with respect to time andtemperature and was carried out for 29 cycles at 96°C for 45 s, 55°C for 120 s and 73°Cfor 120 s. The reactions continued for one cycle at 96°C for 45 s, 55°C for 120 s and73°C for 10 mm. The reaction was electophoresed through a 2% agarose gel, visualizedby ethidium bromide staining, and purified using Sephglas bandprep kit. PCR productswere digested and ligated into appropriate restriction sites of pGEX-2T. The ligationswere used to transform DH5x-high competence cells (GII3CO-BRL) and positive cloneswere selected on the basis of ampicillin resistance. The clones were sequenced by Dr. G.Kalmar to confirm that the mutations were as expected.iii. Construction of a H. sapiens Pim- 1 expression vectorThe human pim-1 cDNA clone, pCI, was obtained from Dr. T. Meeker via Dr. M.Lilly (Seattle VA hospital). This plasmid was a tetracycline-resistant derivative ofpBR322 with the human pim-] cDNA clone inserted into the Pstl site, abolishing theamp’ of the plasmid. Restriction digests were done to ensure that the restriction sites inthe plasmid were intact. The pCI plasmid was linearized by digestion with Smal to beused as a template for PCR.The coding region of the human pim-1 was amplified by PCR, using oligonucleotidesbased on the amino and carboxyl terminal regions of the open reading frame of thepublished human pim-1 sequence [Zakut-Houri et al., 1987]. Oligonucleotides 9402 and9405 (see Appendix III) were designed to incorporate appropriate restriction sites toallow insertion of the pim-1 PCR fragment in the correct reading frame into the EcoRland Smal restriction sites of the pGEX-2T vector.70For each PCR reaction, the amounts of template and oligonucleotides were varied inorder to optimize the reaction. Initial reactions were performed in a final volume of 20il, and scaled up to 50 jil for preparative purposes. Reactions contained 50 ng of pCItemplate, 50 pmol of oligo 9204, 50 pmol of oligo 9205, 100 jiM of nucleotidetriphosphates, 0.5 jil of BSA and 2.5 p1 of VENT in a total volume of 50 jil NEB PCRbuffer. The PCR reaction was optimized with respect to temperature and time of thevarious steps, denaturation was at 96°C for 45 s annealing at 55°C for 2 mm andelongation at 73°C for 2 mm for 25 cycles. The approximately 1 kb PCR product wassubjected to electrophoresis on a 1% agarose gel, visualized by ethidium bromidestaining, purified by electrolution for 90 mm at 50 mA, and then precipitated by theaddition of 0.1 volume of NaOAc (pH 4.8) and 2.5 volumes of ethanol.The pim-1 PCR product and the pGEX-2T plasmid were digested sequentially withEcoRl and Smal to create cohesive ends. In addition, the pGEX-2T fragment wasdephosphorylated with alkaline phosphatase. The DNA fragments were ligated for 16 hat 14°C, then for 6 days at RT and then used to transform XL1-blue E. coli. Positiveclones were selected with blue-white selection on LB-amp plates containing IPTG and Xgal and were restreaked for mini-preps. Clones were analyzed by restriction digestionwith enzymes Xhol, EcoRV, BamHl, EcoRl and Smal to ensure that the insert was theexpected size, in the correct orientation and the Smal site at the 5’ end of the PCRfragment was maintained (Figure 3). A large scale plasmid preparation was performedwith one of the positive clones.71Figure 3 Restriction map of H. sapiens pim-1 in the pGEX-2T vectorThe map shows the the 939 base pair coding region of H. sapiens pim-l insertedinto the Smal and EcoRl cloning sites (shown in bold type) of the pGEX-2Tplasmid. The unfilled arrow [I] shows the location of the Amp’ gene, thegrey shaded arrow shows the location of the lac Fl gene , the hatched boxrepresents the glutathione S-transferase coding region and the blackbox is the human pim-1 coding region. The line arrow 0’ showsthe direction of transcription of the fusion constructBamHI (657)(662)BalI (859)coRl (1610)PstI (2567)723. EXPERIMENTAL PROCEDURES - PROTEIN BIOCHEMISTRY1. GENERAL PROTEIN BIOCHEMISTRY TECHNIOUESi. Protein guantitationProteins were quantitated by the method of Bradford [1976]. A series of proteinstandards was prepared by adding 0-30 jig of BSA to tubes in 5 jIl increments, with 2 mlof Bradford reagent (100 mg Coomassie Blue G, 50 ml ethanol, 100 ml H3P04made upto 1:1 with dH2O and filtered) and mixed by gentle vortexing. The protein to bequantitated was diluted with dH2O so that the absorbance would fall in the linear range ofthe BSA standards (5-20 jig). The solutions were measured at 595 nm and theconcentrations were plotted by linear regression.For purified bacterially-expressed proteins, the amount of full-length fusion proteinwas quantitated visually by comparison to BSA protein standards on a SDS-PAGE gel.Serial dilutions of purified protein and BSA standard (0.5, 1, 2, 4, 8 jig) were subjected toelectrophoresis on an SDS-PAGE gel, then stained by Coomassie. The amounts of fusionprotein were compared to the standards and adjusted by the dilution factor and thevolume loaded. The percentage of the fusion protein in the sample was determined bycomparison of the fusion protein concentration to the total protein concentration.ii. Column fractionations of protein extractsCell extracts from various sources were fractionated by fast protein liquidchromatography (FPLC) on 2 ml MonoQ, MonoS or ResourceQ columns (Pharmacia).Columns were equilibrated before and after use with 2 ml of 2.0 M NaC1, and washedextensively with buffer A (0 M NaCl). All buffers were filtered (0.22 1’) before using.Samples were prepared as described and subjected to either high speed centrifugationin the Eppendorf centrifuge at 15000 rpm for 5 mm, or by filtration through a 0.22 jifilter. The sample was quantitated if necessary and the appropriate volume was diluted to2.2 ml with the appropriate buffer A (described below). A maximum of 5.0 mg of totalprotein were loaded onto the column (numbers of oocytes are indicated whereapplicable). The sample was loaded onto the 2.0 ml injection loop and applied directlyonto the column as part of the program. A standard elution program was followed: theprotein was applied onto the column in 2.5 ml of buffer A, eluted in a 10 ml 0-0.8 M73NaC1 gradient, and collected in 0.25 ml fractions. Column fractions were either assayedby peptide kinase assays or analyzed by SDS-PAGE.MonoQ and ResourceQ columns were equilibrated with Buffer A (10 mM MOPS (pH7.2), 25 mM B-glycerophosphate, 2 mM EDTA, 5 mM EGTA, 2 mM Na3VO4)andeluted with Buffer B (0.8 M NaC1, 10 mM MOPS (pH 7.2), 25 mM B-glycerophosphate,2 mM EDTA, 5 mM EGTA, 2 mM Na3V04).MonoS columns were equilibrated with MES buffer A, pH 6.5 (20 mM MES, 25 mMB-glycerophosphate, 1 mM EDTA, 1 mM EGTA, I mM DTT, 5 jiM B-methyl asparticacid, 1 mM NaF, 100 jiM Na3VO4)and eluted with MES buffer B, pH 6.5 (0.8 M NaC1,20 mM MES, 25 mM B-glycerophosphate, 1 mM EDTA, 1 mM EGTA, I mM DTT, 5.iM B-methyl aspartic acid, 1 mM NaF, 100 jiM Na3V04).iii. SDS-PAGE GelsProteins were separated using sodium dodecyl sulphate polyacrylamide gelelectrophoresis (SDS-PAGE) [Laemmli, 1970]. Proteins were subjected toelectrophoresis on 1.5 mm thick polyacrylamide gels, with a stacking gel containing 4%acrylamide and a separating gel containing 10% acrylamide. Before loading, the sampleswere boiled for 5 mm with 1 vol of 2X sample loading buffer (120 mM Tris-Cl (pH 6.8),4% SDS (wlv), 10% glycerol (v/v), 0.3 M 2-ME, 0.008% bromphenol blue (w/v)), thenbriefly centrifuged. The gels were subjected to electrophoresis in running buffer (25 mMTris, 192 mM glycine, 3.5 mM SDS) at 10 mA 0/N, until the dye front reached thebottom of the gel.iv. Staining SDS-PAGE gelsa. Coomassie stainingGels were immersed in Coomassie stain (0.1% Coomassie Brilliant Blue R150%methanol/lO % acetic acid (w/vlv)) for 10-30 mm, then soaked in 40% methanol/10%acetic acid (vlv) until the gel was sufficiently destained to allow appearance of proteinbands. The destaining solution was changed several times, and a sponge was added tosoak up the stain.74b. Silver stainingIn preparation for silver staining [Merril et al., 1981], gels were first soaked infixative 1 (40% methanol/10% acetic acid (vlvlv)) for 30 mm and fixative 2 (10%ethanol! 5% acetic acid (v/vlv)) for 2-15 mm periods. The gels were oxidized for 5 mmin oxidizer (3.4 mlviK2CrO7,3.2 mM nitric acid). The gels were washed three timeswith dH2O, then stained with 0.204% AgNO3 (w/v) for 20 mm. Gels were washedbriefly in deionized water and developed with Na2CO3 (0.28M) in a formaldehydesolution (0.166% (v/v)) and stopped with the addition of acetic acid (5% v/v).c. Amido black stainingMembranes or gels were immersed in 0.1% amido black!45% methanol!10% aceticacid (vlvlv/v) for 15-30 mm at RT with rotation. Destaining was carried out in a 5%methanol!10% acetic acid (v/v/v) solution until bands were visible and background wasreduced.v. Western blotting of SDS gelsAfter the dye front reached the bottom of the gel, the electrophoresis apparatus wasdisassembled and the bottom right corner of the gel cut for orientation. The gel wasequilibrated in transfer buffer (pH 8.6, 20 mM Tris, 120 mM glycine, 20% methanol(vlv)) to remove SDS. PVDF membrane was hydrated for 30 s in methanol, thenequilibrated for 5 mm in transfer buffer. If nitrocellulose membrane was used, it washydrated in transfer buffer for 60 s. The gel and membrane were placed between 6 piecesof 3 mm filter paper hydrated in transfer buffer and enclosed within the sponges andplastic sandwich apparatus. If more than one gel was transferred at a time, the gels werestacked with 4 pieces of filter paper and a piece of nitrocellulose between them. Carewas taken to prevent the trapping of any air bubbles between the gel and the membrane.The transfer sandwich was placed in a Hoefer transfer cell and transferred at 4°C for 3 hat 300 mA, or for 100 mA for 12-16 h.Membranes were fixed (40% methanol/10% acetic acid (v/v/v)) for 15 mm, thenwashed for 15 mm in dH2O. The membranes were then either Ponceau stained tovisualize protein or air dried and stored for later use.The wet membranes were blocked in 5% BSA in TBST (50 mM Tris base, 150 mMNaC1, 0.05% Tween-20, pH 7.5), for 2 h or 0/N, then rinsed briefly in TBST. Theprimary antibody was diluted to the optimum concentration (usually 1/500-1/1000) in75TBST and incubated with the membrane for 4 h at RT, shaking gently. The antibody wasthen removed and saved (0.1% azide added) and the membrane washed at for three 10mm washes with TBST. The secondary antibody was diluted to the optimumconcentration in TBST, and incubated with the membrane for 2 h at RT, with agitation.When using ECL, second antibody was diluted to 1:20000 and discarded after use.Membranes were subjected to at least three 10 mm washes with TBST and one wash withTBS.When using the 4G10 anti-phosphotyrosine antibody, blotting was carried out asabove, but with some modifications. Low salt TBS (20 mM Tris-base, 50 mM NaC1, pH7.5) and 0.05% NP-40 (v/v) instead of Tween-20 was used. The membrane wasincubated in the primary antibody for a maximum of 4 h and incubated with an ALP goatanti-mouse secondary antibody for a maximum of 2 h.a. Alkaline phosphatase (ALP)-conjugated secondary antibodyThe membrane was rinsed briefly with 1X AP buffer (0.1 M Tris-base, 0.1 M NaC1, 5mM MgCl2,pH 9.5), then developed in 50 ml AP buffer containing 340 jil of NBT (50mg/ml in 70% DMF) and 170 jil of BCIP (50 mg/mi in 100% DMF). The color reactionwas left to proceed for 30 sec to 1 h (depending on intensity of band) and was terminatedby rinsing membrane with dH2O before air drying.b. Horseradish peroxidase-conjugated secondary antibody (ECL)Membranes were blotted briefly on filter paper to remove excess TBS, then incubatedwith ECL reagent for exactly 60 s with gentle swirling. Membrane were blotted toremove excess reagent and wrapped carefully in Saran wrap, avoiding bubbles andwrinkles. The membranes were exposed to X-ray film from 10 sec to 30 mm, dependingon intensity of bands and amount of background.vi. Stripping and reprobing Western blotsFor ALP blots, membranes were stripped by washing in TBS (pH 2.5) for 10 mm,followed by two 5 mm washes with TBS (pH 7.5). Membranes were reprobedimmediately with primary antibody.For ECL blots, membrane was stripped with buffer (100 mM 2-ME, 2% SDS (wlv),62.5 mM Tris-HC1, pH 6.7) at 55°C for 20 mm with agitation. The membrane waswashed with TBST and then reblocked 0/N with 5% BSA in TBST.76vii. Autoradiography and development of filmRadioactive items including SDS-PAGE gels, Western blots, nylon membranes(library screen, Southern and Northern blots), sequencing gels, TLC plates and ECL blotswere wrapped in Saran wrap to protect film from moisture. The radioactive item wassecured with labeling tape onto cardboard and inserted into cassette. Film was exposed tothe radioactive signal for a specified amount of time and developed. Film was developedfor 20-30 sec (time dependent on exposure and background) in Kodak Developer withgentle swirling. Development was stopped by immersing the film in 3% acetic acid (vlv)for 60 sec then fixed by soaking in Kodak Fixer for 3-5 mm. The film was washed for 10mm in running water, then air-dried at RT.viii. ImmunoprecipitationThe protocol for immunoprecipitation was obtained from Dr. Chris Siow (KinetekBiotechnology Corp.) and was performed as described below. Protein A-SepharoseCL4B was swollen for 15 mm in 3% NEFT (3% NP-40, 100 mM NaC1, 5 mM EDTA,50 mM Tris-HC1, (pH 7.4), 50 mM NaF). The Protein A-beads were pelleted by spinningin a Sorvall RT 6000D lab top centrifuge at 3000 rpm for 1 mm, washed twice with 3%NEFT, and resuspended with an equal volume of 3% NEFT to yield a 1:1 suspension.Antibodies were bound to the Protein A-beads by preincubating the beads (40 ii ofthe 1:1 suspension) with an appropriate amount of purified antibody or sera. The titreand amount of antibody used is indicated in the results section. The antibody-Protein Asolution was mixed by rotation for 45 mm at 4°C, then pelleted by centrifugation in anEppendorf centrifuge at 5000 rpm for 5 mm, and washed once with 3% NEFT.Cell homogenate (approximately 100 j.il) or fusion protein (approximately 0.5-2 pjg)was clarified by centrifugation in an Eppendorf centifuge at 4°C for 5 mm at 15000 rpm.The solution was aliquoted and the volume increased to 100 j.ti with 0% NEFT buffer(100 mM NaC1, 5 mM EDTA, 50 mM Tris-HC1, (pH 7.4), 50 mM NaF), and SDS wasadded to a final concentration of 1%. To each sample were added 100 p.1 of 6% NEFT(6% NP-40, 100 mM NaC1, 5 mM EDTA, 50 mM Tris-HC1, (pH 7.4), 50 mM NaF). Thesamples were precleared with 10 j.tl of Protein A for 15 mm at 4°C, and centrifuged for 5mm at 15000 rpm. This precleared pellet was saved as a control for non-specific proteinbinding to the protein A-beads. The supernatant was added to the antibody-gel pellet,and rotated for 1 h at 4°C. The immunoprecipitate (IP) was centrifuged at 5000 rpm for 5mm, and the supernatant removed and saved as a control. The pellet was washed with771.25 ml of 6% NEFT, followed by a wash with 1.25 ml 0% NEFT. If the IP was not to beused for kinase or autophosphorylation assays, the pellet was then resuspended in 2Xsample buffer, boiled and loaded on a gel as usual.Alternatively, when the H-chain interfered with the detection of theimmunoprecipitated protein(s) on an SDS page gel, 1 mM N-ethyl maleimide (NEM) andmodified sample loading buffer (not containing 2-ME) were added to the IP and the entiresample loaded directly onto a gel without boiling. The remainder of the protocol wasfollowed as usual.ix. Kinase assays of immunoprecipitationsIf peptide assays or autophosphorylation assays were to be performed with IPs, somemodifications to the previous protocol were made. If kinase activity was expected to beretained after binding of antibody, the IP was washed several times with ADB. Forpeptide assays, the peptide cocktail as well as the assay ATP was added to each tube andthe assay initiated by gentle vortexing. The peptide assay was terminated by the carefulwithdrawal (avoiding the protein A-beads) and spotting of the supernatant on filter paper.Autophosphorylations were done in a similar manner, with the appropriate ADB and theATP added to the IP. The reaction was terminated by the addition of sample buffer andthe entire sample, beads and supematant, was loaded onto gels.In cases where the kinase activity was immobilized by antibody binding, the kinasewas dissociated from the antibody by the addition of 25 p.1 of 0.1 M glycine, pH 2.5. TheIP was mixed and incubated for 5 mm at RT, then the beads were quickly pelleted bycentrifugation and the supematant removed to a tube containing 20 p.1 of 1.0 M Tris-HC1(pH 8.8). Kinase assay constituents were added and the assay carried out as usual. Thereaction was terminated by either spotting on filter paper or by the addition of 2X samplebuffer.x. Protein-protein interactionsa. GST-Pim-1 protein affinity columnsMany combinations of cell lysates (immature and mature X. laevis and P. ochraceusoocytes) and affinity matrix (H. sapiens and X. laevis GST-Pim-1) were tested for proteinbinding. Cell lysates (approximately 4 mg of X. laevis oocyte lysates, 5 mg P. ochraceusoocyte lysates) were filtered with a 45 p. filter. The lysates were precleared to remove78any proteins non-specifically binding to the GST or glutathione by incubating with 250 tl(packed volume) of GST beads in the presence of protease inhibitors (10 pi each of SBTI,aprotinin, leupeptin) for 15 mm at 4°C with frequent inversion. The slurry was pouredinto a column, and the flow-through collected. The column was washed with 8-10 vol ofSTEC buffer.The flow through from the GST column was added to 250 jii (packed volume) ofGST-Pim-1 beads, and incubated for 10-30 mm at 4°C with frequent inversion. Theslurry was poured into a column and washed with 8-10 volumes of STEC buffer. Theflow-through was applied to the column a second time and the column washed again with8-10 volumes of STEC buffer.For in vitro phosphorylation reaction of bound proteins, the column was then washedwith ADB. A sample of 250 jil of GST-Pim-1 beads was introduced as a control. Thebeads were removed from the column and an in vitro phosphorylation reaction wascarried out in the presence of 10 mM MgC12,33.3 p.M [y_32P] ATP (10 cpmlpmol), 0.5j.LM PKI (cyclic AMP-dependent protein kinase inhibitor) in a final volume of 400 p.1ADB. The reaction was initiated with the addition of [y—32P]ATP and was allowed toproceed for 30 mm at 30°C. Beads were resuspended every 10 mm by withdrawing andexpelling supernatant. The reaction was terminated by the addition of 100 p.1 of 2Xsample buffer and the sample split in two before being subjected to SDS-PAGE. Half ofthe gel was silver stained to visualize protein and half was blotted onto nitrocellulose andautoradiographed.b. Far Western blotting procedureThe Far Western blotting protocol was a modified version of that described by Kaelinet al. [1992]. MonoQ fractions of immature and mature X. laevis oocyte extract wereseparated on an SDS-PAGE gel and Western blotted onto PVDF membrane. Membraneswere blocked 0/N in 3% BSA in low salt TBS, then briefly washed twice with NBST.Two probe methods were used, [y—32P] labeling of Pim- 1 probe and [y—32P] labeling ofimmobilized proteins with X. laevis GST-Pim- 1.1. Labeled Pim- 1 probePim- 1 protein was prepared as a probe, by radiolabeling X. laevis GST-Pim- 1 proteinbound to 250 p.1 of glutathione agarose beads. The beads were washed twice andresuspended in a slurry with 500 p.1 of ADB (10 mM MgC12,0.5 uM PKI). [y—32P] ATP79(25 jiM, 6 x io cpmlpmol) was added to the beads and incubated for 1 h at 30°C. Thereaction was terminated by the addition of 500 j.il of TBST, and the beads were washed3X with TBST to remove any non-incorporated [y—32P]ATP. The labeled fusion proteinwas removed from the beads by washing the beads three times with 500 j.il of 10 mMglutathione (15.4 mg/5 ml) in TBST. The eluted GST-Pim-1 was then added to themembrane in 10 ml of 3% BSA in TBST and incubated at RT 0/N with constant rotation.The membranes were washed with TBST, air dried and autoradiographed.2. Labeling of immobilized proteins with X. laevis GST-Pim-1After blocking of membrane 0/N with BSA, a negative control kinase assay wasperformed to ensure that none of the bound proteins would autophosphorylate. Assaydilution buffer (4 ml) was added to the blot, with 0.125 tiM PKJ and 10 jiM [‘y—32P]ATP(6 x i0 cpmlpmol). The kinase reaction was allowed to proceed 0/N at 30°C, withrotation. The blot was washed extensively with NTBS until unincorporated radioisotopewas no longer present. The blot was then briefly air-dried and exposed to X-ray film for16-24 h.The membrane was rehydrated and blocked for 1 h in 3% BSA (wlv) in NTBS.Assay dilution buffer (4 ml) and 0.125 jiM PM were added to the blot and 200 jil of1.75 mg/ml of GST-Pim-1 were added, with 10 jiM [y—32P]ATP (6 x iO cpmlpmol).The kinase reaction was allowed to proceed 0/N at 30°C, with rotation. The blot waswashed extensively with NTBS until unincorporated radioisotope was no longer present.The blot was then briefly air-dried and exposed to X-ray film for 16-24 h.xi. Antibody Productiona. Pimi-ifi. Piml-NT. Piml-XIThe affinity-purified antibodies Pim 1-ITT, Pim 1-NT, and Pim 1 -XI were prepared asdescribed [Sanghera, et al., 1992] and are commercially available from UpstateBiotechnology Inc. (Lake Placid). The peptides were suspended in PBS and Freund’sincomplete adjuvant and injected into rabbits. The rabbits were frequently boosted, andbled for titres. Antibodies were purified from sera using peptide affinity columns andtitred by ELISA.80b. Anti-X. laevis GST-Pim 1 serumRabbit polyclonal antisera was produced against the full-length X. laevis GST-Pim-1bacterial fusion protein. The X. laevis GST-Pim- 1 bacterial fusion protein was produced,purified and eluted from the glutathione beads as described in Section 3.2.i. The elutedfusion protein was concentrated in a Centricon 30 filter tube by centrifugation in aBeckman centrifuge at 4°C, 3000 rpm for 30-45 mm. After several washes with coldPBS, the volume was reduced by continued centrifugation. The protein concentrationwas quantitated as described in Section 3.1.i, aliquotted and frozen until use. The firstinjection required 250 jig of protein, 200 jig were used for the second and third injectionsand 100 jig were used for each successive injection.The appropriate aliquot of protein was thawed before injection, and mixed with PBSand Freund’s incomplete adjuvant (first injection only). Two rabbits (numbers 48, 49)were each injected and boosted according to the regular schedule. The rabbits were bledand the sera purified by protein G purification following the manufacturers protocol[Pharmacia]. The Protein G-purified sera was incubated with GST-beads to remove anyantibodies specific for the GST portion of the fusion protein. The sera (1.5 ml) wasincubated with 500 jil of GST bound to glutathione beads, rotating for 30 mm at 4°C.The sera was centrifuged briefly to pellet the beads and the supernatant was removed.The beads were washed once with 500 jil of PBS (4°C), the supernatants combined andaliquoted for future use. The GST-cleared sera was frozen and used for Western blotsand immunoprecipitations.xii. Phosphatase treatment of proteinsThe phosphatase activity of various enzymes was tested using p-nitrophenophosphate(PNP). The PNP was diluted to 10 mg/mi in the respective phosphatase buffer, and 2 j.tlof each phosphatase were added to 1000 jil of the PNP solution, and incubated at RT.Phosphatase activity was evidenced by a color change in the solution, from clear to paleyellow. Phosphatases were considered to be active if the color reaction occurred within30 s of phosphatase addition.a. Acid phosphataseAcid phosphatase was used to non-specifically dephosphorylate GST-Pim- 1. Acidphosphatase (10 jil) was added to 200 jil of a 1:1 suspension of GST-Pim-1 beads(prewashed in TBS, pH 5.0) and the reaction allowed to proceed for 30 mm at RT. To81terminate the phosphatase reaction, the beads were washed three times with isotonicassay dilution buffer.b. Alkaline phosphataseAlkaline phosphatase was used to dephosphorylate GST-Pim- 1 non-specifically.Alkaline phosphatase (5 jil) was added to 200 j.iJ of a 1:1 suspension of GST-Pim- 1 beads(prewashed in the buffer provided, containing MgC12,pH 9.0) and the reaction allowed toproceed for 30 mm at RT. To terminate that reaction, the beads were washed three timeswith isotonic assay dilution buffer.c. Phosphatase HPTPBThe HPTPB, was obtained from Mr. Ken Harder (Biomedical Research Centre) boundto glutathione beads and was thrombin cleaved prior to use. The HPTPB beads werewashed in thrombin buffer (0.05 M sodium citrate, 0.15 M NaC1, pH 6.5), and thesupernatant was removed. Thrombin was diluted in buffer to a final concentration of 20ngIi1, which was added in a 1:1 ratio (vlv) with the HPTPB beads. The cleavage reactionwas carried out at 22°C for 30 mm, with frequent agitation. After centrifugation, thesupernatant containing the cleaved HPTPB was carefully recovered and the beads werewashed twice with 100 j.tl of thrombin buffer. PMSF was added to inactivate thethrombin. The final concentration of the cleaved HPTPB was estimated to be 1 jiglj.il,which was confirmed by silver staining.Phosphatase reactions using HTPTB were carried out in TBS, pH 8.0 in the presenceof 5 mM 2-ME and SBTI. Ten jil of HPTPB were added to a reaction containing 200 tlof a 1:1 slurry of GST-Pim- 1 beads. Reactions proceeded at RT for 30 mm.2. PRODUCTION AND PURIFICATION OF GST-FUSION PROTEINS INBACTERIAi. Expression and purificationThe pGEX-2T vectors containing the various Pim-] inserts were used to transformUT5600 E. coli. Cultures were grown in 2xYT [Maniatis et al., 1989, Section A.3] mediawith 75 j.tg/ml amp 0/N at 37°C. After 16 h, the cultures were diluted 10-fold with 2xYTmedia with 75 pg/m1 amp and grown for 1 h at 37°C. The expression of the fusionprotein was induced with 200 IIM IPTG for 3-6 h at RT. The culture was centrifuged in a82Beckman centrifuge at 2000 x g for 15 mm, the pellet washed in 25 ml PBS containing0.2 mM PMSF, frozen and stored at -70°C until use. To extract the fusion protein, thepellet was resuspended in 10 ml of PBS containing 1 mM EDTA, 0.1 mM EGTA, 0.2mM PMSF, 1% Triton X-100 and protease inhibitors aprotinin, leupeptin and soybeantrypsin inhibitor (1 jig/mi each). The pH was adjusted to 7.5 and 200 jig/mi of lysozymewere added to each pellet. After a 20 mm incubation on ice, the final concentration ofMgC12 was adjusted to 2.5 mM and 30 j.il of DNAse (50 mg/mi) were added. Afterincubation on ice for 30 mm, the lysate was centrifuged (12000 x g for 10 mm at 4°C).The supernatant was added to glutathione-agarose beads and gently mixed on a rotator at4°C for 30 mm. The beads were then poured into a column 0.5 cm in diameter andwashed with 10 volumes of STE buffer (50 mM Tris-HC1 (pH 8.0), 4 mM EDTA, 150mM NaC1, 0.1% Triton X-100), followed by 10 volumes of STEC buffer (50 mM TrisHC1 (pH 8.0), 200 mM NaC1, 6 mM CaCl2). The fusion protein was eluted from thebeads with 10 mM glutathione in STEC buffer and immediately frozen at -70°C.ii. Thrombin cleavage of fusion proteinsThrombin digestion of GST-Pim-1 fusion protein was carried out on both eluted andimmobilized fusion protein. In both cases, the immobilized fusion protein was washedwith several volumes of thrombin buffer (0.05 M sodium citrate, 0.15 M NaC1, 5 mMEDTA, pH 6.5). Although the Pharmacia manual recommends thrombin treatment for 2-16 h at RT using 10 cleavage units of thrombin per mg of fusion protein, reactions wereperformed for 1 h at 30°C, for reasons of enzyme thermolability. For thrombinization ofeluted protein, the protein was eluted from the beads with 10 mM glutathione in thrombinbuffer and the eluate then subjected to thrombinization at 30°C for 1 h. Thethrombinization reaction was terminated by the addition of 5 mM EDTA. Alternatively,the thrombin was added directly to the immobilized GST-Pim- 1 to form a 1:1 slurry, andincubated at 30°C for 1 h. The supernatant containing the cleaved protein was thenremoved from the beads.In general, for 10 jig of protein, 20 ng of thrombin were used, or 10 cleavage units ofthrombin per mg of fusion protein. One thrombin cleavage unit was equal to 0.2 NIHunits, and could completely digest 100 jig of fusion protein in 16 h.833. ASSESSMENT OF ENDOGENOUS KINASE ACTIVITY OF EXPRESSED PIM-1i. Autophosphorviation of GST-Pim-1Phosphorylations were performed in a final volume of 50 p.1, with 5 p.1 of eluted GSTPim-1 in STEC buffer, 50 p.M [y—32P] ATP (6 x 10 cpmlpmol), 0.5 p.M PKI and ADBcontaining 10 mM MgC12 and 1.25 mM MnC12. Alternatively, for large scalepreparations, GST-Pim-1 immobilized to 50 p.1 of glutathione beads wasautophosphorylated in a total volume of 200 p.1 ADB containing 10 mM MgC12 and 1.25mM MnC12,0.5 p.M PM and 50 p.M [y—32P] ATP (6 x 10 cpmlpmol). Preincubationswere performed on ice and the kinase reaction at 30°C for 30 mm was started with theinclusion of the ATP and terminated by the addition of 50 p.1 of 2X SDS-sample buffer.The sample was boiled for 5 mm and clarified by centrifugation, the supernatant wassubjected to electrophoresis on a 10% SDS-PAGE gel then transferred onto PVDFmembrane for Western blotting. X-ray film was exposed to the PVDF membrane 0/N.ii. Determination of specific activityThe specific activity of autophosphorylation of the fusion proteins was determined bymeasuring the amount of [y—32P]ATP incorporated per mg of fusion protein at discretetime points. Varying quantities of fusion protein (5, 2.5 and 1 p.g) wereautophosphorylated in an assays containing 100 p.M [y_32P] ATP (6 x iO cpmlpmol), in25 p.1 of ADB. Assays were started by the addition of the ATP, allowed to proceed at30°C for various amounts of time and were terminated by the addition of 2X samplebuffer. Samples were boiled and centrifuged as usual and were subject to electrophoresison a SDS-PAGE gel. Gels were stained to ensure that the amount of protein loaded perlane was consistent, the gels were dried and radioactive bands were cut from the gel andcounted by scintillation counting. Values obtained from the scintillation counting wereadjusted for pmol [y—32P] incorporated per mg fusion protein and were plotted against thetime of assay in mm. The slope of the graph was calculated as the specific activity orautophosphorylation of the kinase in units of pmol.min-’.mg-’.iii. Determining the stoichiometrv of autophosphorvlation of GST-Pim- 1Stoichiometry of autophosphorylation was determined by measuring the moles ofATP incorporated per mol of GST-Pim- 1. The stoichiometry was determined byconverting the specific activity to pmol [y—32P]. pmol’ GST-Pim-1.84iv. Phosphoamino acid analysis of in vitro autophosphorviated Pim-1Phosphorylations were performed in a final volume of 200 jil, with GST-Pim- 1immobilized to 50 j.tl of glutathione beads, 25 jiM [y—32P] ATP (10 cpm/pmol), 0.5 jiMPKI and ADB containing 10 mlvi MgC12 and 1.25 mM MnC12. Preincubations wereperformed on ice and the kinase reaction at 30°C for 30 mm was started with theintroduction of the ATP. Reactions were terminated by the addition of 100 jil of 2x SDSsample buffer, boiled for 5 mm, clarified by centrifugation, separated on a 10% SDSPAGE gel and then transferred onto PVDF membrane by Western blotting.Radioactively labeled bands were visualized by autoradiography and excised from thePVDF membrane and chopped into 0.5 mm2 pieces. The membrane bound protein wasdigested in 300 111 of constant boiling HC1 at 105°C for 1.5 h. The acid was removed andthe membrane was washed briefly with dH2O to remove residual amino acids. Waterand acid were removed by evaporation in a vacuum centrifuge and amino acids weresequentially washed and vacuum-dried. Amino acids were redissolved in water/aceticacid/pyridine buffer (94.5/5/0.5 v/v/v) containing 1 mg/ml each of phospho-serine,phospho-threonine and phospho-tyrosine standards. Phenol red (0.5% in buffer (w/v))was spotted on the origin as a control. Approximately 2000 cpm were spotted 2 cm fromthe bottom of a cellulose sheet (Kodak chromagram) and subjected to electrophoresis for1.5 h at 750 volts, until the phenol red spot had migrated 7-8 cm from the origin. Theplate was air dried and sprayed with ninhydrin solution (0.25% in ethanol) and developedto visualize standards either by heating with a hairdryer or by baking in an oven at 90°Cfor 5 mm. X-ray film was exposed to the cellulose sheet for 18 h and subsequentlydeveloped.4. ASSESSMENT OF EXOGENOUS KINASE ACTIVITY OF EXPRESSED PIM-1i. Phosphorviation of protein substratesPhosphorylation reactions were carried out in ADB in a total volume of 50 jilcontaining 25 jig of protein substrates, 0.5 jiM of PKI, 50 jiM [‘y—32P]ATP (6 x iOcpmlpmol) and either 5 jig of X. laevis or 15 jig of H. sapiens GST-Pim- 1. For reactionscontaining 40S ribosome, 60 jig of ribosome was used. Control reactions containingeither substrate alone or enzyme alone were performed. Reactions were started with theaddition of the ATP, carried out at 30°C for 40 mm and terminated by the addition of 100jil of 2X sample buffer. Samples were subjected to electrophoresis on SDS-PAGE gels,transferred onto PVDF and autoradiographed for 2 h.85ii. Phosphorviation of synthetic peptidesPeptide phosphorylations were performed in a final volume of 25 j.il, with 5.0 p.1 ofdiluted peptide, 1.25 p.g of H. sapiens or 0.5 jig of X. laevis GST-Pim-l in STEC buffer,50 p.M [y-32P1 ATP (2000 cpmlpmol), 0.5 p.M PKI and ADB. All preincubations wereperformed on ice and the kinase reactions at 30°C for 10 mm were started with theaddition of the ATP and terminated by application of 20 p.1 of the reaction mixture onto a1.5 cm2 piece of Whatmann P81 phosphocellulose paper. Filter papers were washed inphosphoric acid (l%,v/v) and the radioactivity quantitated in an LKB Wallac 1410scintillation counter.iii. Determination of kinetic constantsKinetic constants Km and Vm were estimated from Lineweaver-Burke plots with atleast six different concentrations of a given peptide per determination. Assays wereperformed in triplicate, and the mean values of the Km and Vmax determinations areshown in Tables I-VT. The standard errors for all reported Km and Vmax were less than15%. The value of Vmax/Km was calculated for all peptides and inserted into the thirdcolumn of Tables I-VT. The VmIKm value was used to compare the relative efficienciesof various peptides as substrates of the GST-Pim- 1.iv. Stability of GST-Pim- 1 enzyme at 30°CThe stability of GST-Pim- 1 enzyme activity after incubation at 30°C was examined.H. sapiens and GST-Pim- 1 was incubated at 30°C and at discrete time points, aliquots ofthe enzyme solution were withdrawn and immediately added to either anautophosphorylation reaction or a peptide kinase reaction.For autophosphorylation reactions, a 1:1 slurry of GST-Pim- 1 beads in ADB wasincubated at 30°C with frequent resuspension. At intervals from 0-70 mm, 10 p.1 of theslurry were removed and immediately added to an autophosphorylation reactioncontaining 50 p.M [‘y-32P] ATP (2000 cpmlpmol), 20 mM MgCl2,0.5 p.M PM in 25 p.1 ofADB. Reactions were allowed to proceed for 5 mm at 30°C then terminated by theaddition of 25 p.1 of 2X SDS-PAGE sample buffer and subjected to electrophoresis on anSDS-PAGE gel. After electrophoresis, the gel was silver stained as a control for proteinloading and the radioactive bands were excised and quantitated by scintillation counting.86For peptide substrate assays, GST-Pim- 1 was incubated at 30°C. At discrete intervals(from 0-70 mm), 5 j.il of GST-Pim- 1 were removed and immediately added to a peptidesubstrate reaction containing 50 jiM [y-32P] ATP (2000 cpmlpmol), 0.5 p.M PKI in 25 p.1of ADB. Reactions were allowed to proceed for 5 mm at 30°C, then terminated byspotting 20 p.1 onto p81 filter papers. Filter papers were washed as usual and counted byscintillation counting.v. Optimization of kinase reactionsa. Metal ion optimizations of GST-Pim- 1 kinase reactionsAssays contained 3 p.1 GST-Pim-1 (eluted without calcium), 0.5 p.M PKJ, 250 p.M P4peptide (AKRRRLSA), 50 p.M [y-32P] ATP plus ions, in a final volume of 25 p.1 of ADBcontaining 1 mM MgCl2. Assay concentrations of CaCl2 ranged from 3.906 mM to 1 M;assay concentrations of NaC12,ZnC12,MnCl2,and MgC12ranged from 390.6 p.M to 200mM. Assays were performed in triplicate. The incubations were for 7 mm at 30°C, andterminated by spotting 20 p.1 onto p81 filter papers. Filters were washed and quantitatedas usual.A more detailed study was done to examine the effects of CaC12 on GST-Pim- 1activity. The same conditions were employed as above, except that a higherconcentration of ATP was used (90 p.M [y-32P] ATP (1100 cpmlpmol)).b. ATP optimizations of GST-Pim-1 kinase reactionsKinase assays were performed to ensure that the amount of ATP used in assays wasnot limiting. Assays contained 1.5 jig of X. laevis or 3 jig H. sapiens GST-Pim- 1, 250p.M P4 peptide, 0.5 p.M PKI in 20 p.1 of ADB (25 mM MgC12). Preincubations were overice and reactions started with the addition of [y-32P] ATP, from 0.5 p.1 to 10 p.1, inamounts varying from 4 j.LM to 83 p.M. Assays were carried out for 10 mm at 30°C andwere terminated by spotting 20 p.1 onto p81 filter papers. Filters were washed andquantitated as usual.c. Time course of Pim-1 activityA time course assay was performed ensure that the activity of Pim- 1 was linear forthe assay times selected. Kinase assays contained 1.5 jig of X. laevis or 3 jig H. sapiensGST-Pim-1, 250 p.M P4 peptide, 0.5 p.M PKI, 120 p.M [y-32P] ATP (1100 cpmlpmol) in29 p.1 of ADB (25 mM MgC12). Preincubations were over ice and reactions at 30°C werestarted with the addition of [y-32P] ATP. Reactions were terminated at discrete time87points by spotting 20 pJ of reaction onto p81 filter papers. Time points tested intriplicate, starting at 30 s then one mm intervals from 1-20 mm, two mm intervals from20-30 mm. Filter papers were washed and quantitated as usual.vi. Antibody inhibition of GST-Pim-1 kinase activityKinase reactions were carried out in a total volume of 25 p.1 of ADB containing 25mM MgCl2,0.5 jiM PM, 50 p.M [y-32P1 ATP (2000 cpmlpmol), 250 p.M P4 peptide, and0.5 jig of either H. sapiens or X. laevis GST-Pim-1 and serial dilutions (1:2.5 - 1:160) ofantibodies (Piml-III, Piml-NT, Tel, CRB-Pim-1, Piml-XI, GXP and a cocktail of A2and C2.). Assays were started with the addition of ATP and were allowed to proceed for10 mm at 30°C. Reactions were terminated by spotting 20 p.1 of reaction onto p81 filterpapers. Filter papers were washed and quantitated as usual.vii. Peptide inhibition of endogenous GST-Pim- 1 activityAutophosphorylations were carried out in 100 p.1 of ADB containing 10 mM MgCl2,25 p.M [“y-32P1 ATP (6 x 10 cpmlpmol), 25 jig of GST-Pim-l and peptides P4(AKRRRLSA), P7 (negative control peptide AKRRRLAA), and P15 (AKRRRLCA,constructed as an inhibitor) in concentrations varying from 0, 0.125 mlvi (0.5 Km), 0.25mM (1(m), 0.50 mM, 0.75 mM, 1.0 mM, 1.5 mM, 2.5 mM and 4.0 mM. For eachreaction, constituents were mixed on ice and started at 10 s intervals with the addition ofATP. Reactions were carried out at 30°C for 20 mm, and stopped by the addition of 30p.1 of 2X sample buffer. Samples were boiled for 5 mm and loaded onto gels, thentransferred onto PVDF. Radioactive bands corresponding to the full-length fusionprotein and smaller byproducts were cut out and quantitated by scintillation counting.viii. Peptide inhibition of exogenous GST-Pim-1 activityReactions were carried out in 25 p.1 of ADB containing 10 mlvi MgC12, 1.25 mMMnC12, 50 p.M [y-32P]ATP (2000 cpmlpmol), 1.5 jig of GST-Pim-1 and dilutions ofsubstrate peptide P4 (0, 1.953, 3.906, 7.812, 15.625, 31.25, 62.5, 125, 250, 500 and 1000jiM). Inhibitor peptide P7 was used at final concentrations of 100, 200, and 500 jiM, andP15 was used at final concentrations of 100 and 200 jiM. An additional assay with 200p.M of P15 was carried out under reducing conditions, with 284 mM 2-ME added toprevent dimerization of the P15 peptide during the assay. For each reaction, constituentswere mixed over ice and the reactions started at 10 s intervals with the addition of theATP. Reactions were carried out at 30°C for 10 mm, and stopped by spotting on to p8188filter papers. Filter papers were washed extensively and counted by scintillationcounting.895. IDENTIFICATION OF AUTOPHOSPHORYLATION SITES OF EXPRESSEDPIM-1i. In vitro autophosphorviation and trvptic digestion of fusion protein for trvpticphosphopeptide analysisIn vitro autophosphorylation of 25 jig of expressed fusion protein (immobilized onglutathione beads) was carried out in a total volume of 100 p1 ADB containing 160 p.M[‘y-32P]ATP (100 cpmlpmol), 1.25 p.M PKI, for 30 mm at 30°C. Reactions wereterminated by the addition of SDS-PAGE sample buffer, boiled for 5 mm and weresubjected to SDS-PAGE electrophoresis on a 10% acrylamide gel. The full-lengthradiolabelled fusion protein was detected by autoradiography and excised from the gelusing a scalpel blade. Gel particles were minced, washed with water to remove methanoland SDS and partially dried under vacuum. The gel pieces were resuspended in 50 mMammonium bicarbonate (pH 8-8.5), containing trypsin (10 jig/mi) and digested at 37°Cfor 24 h with constant agitation. The gel pieces were extracted several times with dH2Oand 20% acetonitrile to recover the peptides, washes were combined and dried downunder vacuum. The peptides were resuspended in decreasing volumes of dH2O andredried to ensure removal of all of the ammonium bicarbonate. The radioactivity of thedried peptides was quantified by scintillation counting; from this it was determined thatmore than 60% of the counts had been removed from the fragmented acrylamide gel.ii. Two-dimensional phosphopeptide mappingThe dried peptides were rehydrated in a minimal volume of electrophoresis buffer andapproximately 5-20 x iO counts (10-100 pmol of protein) were spotted onto the center ofa cellulose TLC (20x20) plate, 1.5 cm from the bottom edge. Phenol red (0.4 p1 of a0.5% solution in 5% acetic acid, 0.5% pyridine w/v/v) was spotted as a migration marker.Electrophoresis in the first dimension was performed in water/acetic acid/pyridine(89/10/1; v/v/v) at 750 V at RT, until the phenol red marker migrated 2 cm from thepositive electrode. The plates were air dried and the second dimension was developed byascending chromatography in water/butan- 1 -ol/pyridine/acetic acid (34/30/30/6; v/v/v/v)for approximately 1.5 h until the solvent front reached the top of the plate. After theplates were dried extensively, phosphopeptides were visualized by autoradiography 0/Nat -70°C.90iii. Extraction of trvptic phosphopeptides from TLC platePhosphopeptides contained within each visualized spot were eluted from the TLCplates by removing the cellulose matrix from the plastic support with a scalpel blade.The cellulose was extracted twice with 200 p.1 of 20% acetonitrile (0.1% TFA) and oncewith 200 p1 of 60% acetonitrile. The cellulose was vortexed vigorously after the additionof the acetonitrile solutions and the sample was sonicated in a water bath sonicator for 5mm to break up the cellulose. Following centrifugation, the phosphopeptide containingsupernatant was recovered and dried to a volume of about 5 p.1 in a vacuum centrifugeand quantitated in a scintillation counter. More than 90% of the radioactive counts werelocated in the supernatant. These counts were used to estimate the quantity of peptides ineach sample.iv. Phosphoamino acid analysis of tryptic phosphopeptidesIn vitro autophosphorylation of expressed fusion protein bound to glutathione beadswas carried out exactly as in Section 3.5.i. except that 100 p.M of [y-32P]ATP (10cpmlpmol) was used. The radiolabelled protein was subjected to electrophoresis on anSDS-PAGE gel, isolated, trypsinized, subjected to two dimensional chromatography andexcised from the TLC plate as described above.The phosphopeptides were extracted from the cellulose twice with 200 p.1 of 20%acetonitrile (0.1% TFA), and once with 200 p.1 of 60% acetonitrile. After centrifugation,the supernatant was recovered and dried under vacuum. The samples were digested with200 p.1 of boiling HC1 for 1.5 h at 105°C. The digested amino acids were sequentiallywashed with water and dried under vacuum, then subjected to phosphoamino acidanalysis as described in Section 3.3.iv.v. HPLC analysis of trvptic phosphopeptidesa. IMAC-HPLC-ESI-MS analysis of trvptic peptides from 2D mappingTryptic peptides were subjected to phosphoamino acid analysis or analyzed byimmobilized metal affinity chromatography (IMAC) high pressure liquidchromatography (HPLC) electrospray ionization mass spectrometry (ESI-MS).MicroIMAC-HPLC-ESI-MS was perfomed by Drs. Lawrence Amankwa and MichaelAffolter (Biomedical Research Centre) using instrumentation and protocols as detailedelsewhere [Watts et al., 1994]. In brief, the microlMAC column was assembled bymanually filling a Teflon tube (6 cm long x 250 p.M inner diameter, 1.59 mm outer91diameter) to about 3 cm with a 50% slurry of chelating Sepharose Fast flow (Pharmacia)in 20% ethanol, to yield a final column volume of approximately 1.5 tl. The column wasconnected to an injector and syringe pump and washed with water for 10 mm, at a flowrate of 5 pi/min. The matrix was activated with 30 mM FeC13, in 5 - 5 ul injections at 1mm intervals, then washed with H20 for 10 mm, followed by 0.1 M acetic acid for 10mm. Before the initial use, the column was washed with 5 jil of elution buffer (50 mMNa2HPO4in 0.1% ammonium acetate, pH 8). Samples were loaded onto the microlMACcolunm in volumes of less than 5 jil, and washed for 10 mm in 0.1 M acetic acid, pH 3, toremove unbound phosphopeptides. Bound phosphopeptides were eluted with 5 p.1 ofelution buffer, and applied directly to and purified by reversed phase-HPLC on a HypersilC 18 capillary column (0.32 x 50 mm). After an isocratic wash with buffer A (0.05%TFA, 2% acetonitrile in H20) the HPLC column was eluted with a 0-50% gradient ofbuffer B (0.045% TFA, 80% acetonitrile in H20) over 15 mm at a flow rate of 5 pi/min.Samples were subjected to ESI-MS analysis by Dr. Lawrence Amankwa on a PESciex (Thornhill, Ontario) APJJII triple quadruple mass spectrometer equipped with apneumatically assisted ESI source (ion spray). A total ion chromatogram was used todetect charged particles passing into the spectrometer. The mass to charge ratio (mlz)was scanned repetitively over the range of 300-2000.The output data from the mass spectrometer was analyzed with computer softwareprovided with the machine. Mass spectra were displayed for observed peaks of iondetection events and mass ranges from the data set could be extracted. Peptide masseswere calculated based on computer matching of observed signals with the predicted m/zvalues for the various possible charge states of the same peptide and the theoreticalfragmentation of the peptide sequence listing all possible fragments along with predictedcharged mass values.b. LCMS analysis of trvptic digests of X. laevis Pim- 1Two samples of expressed X. laevis GST-Pim- 1 were prepared, with an estimated 4.2mg of fusion protein bound to 5 mg of glutathione beads per sample. Two samples wereprepared: one was autophosphorylated in vitro with cold ATP, the other was assumed tobe autophosphorylated or phosphorylated in vivo. GST-Pim-1 was partially purified onthe glutathione beads and one sample was autophosphorylated in a total volume of 10 mlin ADB buffer with 25 mM MgC12and 120 p.M non-radioactive ATP, for 30 mm at 30°C.Both samples were washed extensively with thrombin buffer (Section 3.2.ii.) to change92the buffers and to remove unbound ATP. Both samples were digested with 10 jig ofthrombin for lh, RT. After thrombinization, the slurries were poured into columns andthe flow through and elute collected. The columns were washed with 2 volumes ofthrombin buffer to remove all cleaved fusion protein and the flow-through and eluteconcentrated in Centricon-lO tubes by sequential centrifugation and washes with 50 mMammonium bicarbonate buffer. Once the volume of the Pim-l was reduced to 100 j.tl, 10j.tl of trypsin were added, and the sample allowed to digest for 24 h at 37°C, with constantagitation.Both samples (in vitro and in vivo phosphorylated) analyzed by Dr. L. Amankwa(BRC) by LCMS. Half of each sample was dephosphorylated with HPTPB prior toLCMS analysis. Peaks that experienced a shift in retention time after phosphatasetreatment, indicating the loss of a phosphate group, were identified and analyzed.936. SOURCES OF ENDOGENOUS PllvI-1 PROTEINi. Analysis of Pim- 1 protein in X. laevis oocytesX. laevis oocytes were obtained as described in Section 2.2.i. Oocytes werehomogenized in a chilled glass homogenizer with one volume of 2x homogenizationbuffer (150 mM 8-glycerophosphate, 40 mM MOPS, 30 mM of EGTA, 4 mM EDTA, 2mM Na3VO4, 0.25 j.tM DTT, pH 7.2) containing SBTI. Lysates were centrifuged for 10mm at 4°C at 15000 rpm in TL-100, or for 15 mm at 15000 rpm in an Eppendorfmicrofuge. The clear supernatant was withdrawn, avoiding both the greyish pellet andthe yellow lipid layer. Additional protease inhibitors were added (aprotinin, leupeptin,SBTI (1 jig/ml each)), and the homogenate was quantitated, aliquoted and frozen untillater use.X. laevis oocyte extracts were subjected to immunoprecipitation, MonoQ columnchromatography, affinity chromatography and Western blotting. These procedures wereperformed as described earlier.ii. Pim- 1 protein in P. ochraceus oocytesa. Homogenization of oocytes for protein extractionP. ochraceus oocytes were obtained as described in section 2.2.iii. Oocytes werepelleted by centrifugation at 1500 rpm for 5 mm. The volume of the pellet wasdetermined and two volumes of homogenization buffer (50 mM 8-glycerophosphate, 20mM MOPS, 5 mM EGTA, 2 mM EDTA, 1 mM Na3V04 plus 0.25 jiM DTT, 5.0 jiM B-methyl aspartic acid, 1.0 mM PMSF, and 1.0 mM benzamidine) were added. The oocyteswere homogenized in a blender in 3-4 bursts of 15-20 s, then centrifuged at 9000 rpm for15 mm to remove particulate matter and organelles. The post-mitochondrial supernatantwas centrifuged at 42000 rpm in a Sorval ultracentrifuge for 25 mm. The supernatantwas immediately aliquoted and frozen at -70°C until further use.b. P. ochraceus oocyte maturation time coursesA time course of oocyte maturation was prepared, using a total of 60 p.1 of packedoocytes. Oocytes were washed three times with calcium-free artificial sea water(CaFSW) to remove follicle cells and the oocytes were diluted with artificial sea water toa final volume of 400 ml. At this time, the t=0 time point was removed. The maturationstimulant, 4 p.M methyladenine, was added and incubated at 15°C. Aliquots of oocytes94were removed at specific time points and pelleted by centrifugation for 5 mm at 177 x gat 4°C in 50 ml tubes. The supernatant was removed and the oocytes resuspended in 1:1homogenization buffer. The oocyte solution was sonicated for 30- 60 s, then clarified bycentrifugation for 30 mm at 42000 rpm in a Sorval ultracentrifuge at 4°C. Thesupernatant was aliquoted and frozen for future use.iii. Probing crude bovine spleen extract for activated Pim- 1 proteinCrude bovine spleen extracts (B SE) were fractionated and analyzed for the presenceof activated, endogenous Pim- 1 protein. Crude extracts were prepared by homogenizingchopped bovine spleen in a blender, (1:3, (w/v)) with homogenization buffer (50 mM Bglycerophosphate, 5 mlvi EGTA, 2 mlvi EDTA, 1 mM Na3V04plus 0.1 jiM DTT, 5.0 jiMB-methyl aspartic acid, 1.0 mM PMSF and 1.0 mM NaF). Homogenate was filteredthrough cheesecloth to remove particulate matter and clarified by centrifugation at 12 000x g for 15 mm. Supernatants were centrifuged for 40 mm at 42 000 rpm in a Sorvalultracentrifuge, then quantitated, aliquoted and frozen at -70°C until use.iv. Human ervthroblast cell line - K562Human erythroleukemia cells, K562, were cultured in RPMI 1640, 10% FCS and5x104 M 2-ME by Ms. Helen Merkins (Biomedical Research Centre). Cells werecentrifuged at 300 x g in a lab top centrifuge for 5 mm then quantitated using ahemocytometer. Cells were resuspended at a concentration of 3x 1 0 cells per ml in ice-cold phospholysis buffer (PLB) [Telerman et al., 1988] (1% Triton X-100, 0.5%deoxycholate, 0.1% SDS, 0.01 M NaH2PO4(pH 7.5), 0.1 M NaCl and 5 mM EDTA) plusprotease inhibitors, PMSF, aprotinin and leupeptin each at 10 jig/ml. Cells were lysed onice for 20 mm then sonicated at 40% for three 5 mm bursts. Cell homogenates werecentrifuged in a TL-100 for 10 mm at 10 000 rpm or in an Eppendorf microfuge for 10mm at 15 000 rpm. Supematants were aliquoted (2.26 - 3.0 x106 cells per aliquot) andfrozen at -70°C until use.v. Primary human lymphocytesHuman lymphocytes were produced by Dr. Bill Sahl (Pelech Laboratory) as abyproduct of platelet purifications using the method of Fotino et al. [19711. Purifiedlymphocytes were obtained and quantitated using a hemocytometer and were cultured at4 x 106 cells per ml in RPMI media. The lymphocytes were divided to provide controland stimulated samples and were stimulated with 0.666 mM Concanavalin A fromCanavalia ensfonnis which was obtained at 4 M (1.25 g/10 ml). After stimulation for 495h at 37°C, cells were pelleted by centrifugation at 300 x g for 10 mm at RT.Lymphocytes were washed once with PBS, then resuspended in 200 jil of PLB plusprotease inhibitors and incubated on ice for 15 mm. Cells were sonicated at 40% withtwo 2 s bursts, then frozen immediately at -70°C.7. COMPUTER SEARCH ANALYSISComputer analysis used a software package developed at the Biomedical ResearchCentre by Dr. Allen D. Delaney. Evolutionary trees were prepared using the DesoeteTree algorhythms with the PHYLIP program.Nucleotide and amino acid sequence searches were performed at the National Centerfor Biotechnology Information (NCBI) using the BLAST network service. These weredone by Dr. G. Kalmar.96PART 2 - RESULTS AND CONCLUSIONS: BACTERIALLY EXPRESSEDPIM-1CHAPTER IV.CLONING AND EXPRESSION OF XENOPUS LAEVIS GST-Pim-1 ANDCOMPARISON TO HUMAN GST-PIM-11. CLONING PIM- 1 FROM AN X. LAEVIS OOCYTE eDNA LIBRARYDue to sequence conservation between pim-1 genes from different species, degenerateoligonucleotides based on the human pim-1 sequence were successfully used to amplify partof the pim-1 coding region from X. laevis cDNA. This pim-1 PCR fragment was used as aprobe to screen a X. laevis oocyte eDNA library, and from 6.7 x105 plaques screened, threeidentical clones were obtained. The full-length X. laevis pim-] eDNA clones wereapproximately 2.7 kb in length, similar in size to that found in other species [Zakut-Houri etat., 1987; Domen et at., 1987; Meeker et at., 1987a; Padma and Nagarajan, 1991]. Theclones were initially analyzed by Southern blotting, using the X. laevis pim-1 PCR productas a probe (Fig. 4). The Southern blot demonstrated that the three clones (1.6/1.1, 6.22,12.35) were identical and that the probe hybridized even under conditions of highstringency.The 1348 nucleotides of the X. laevis pim-1 eDNA sequence include a predicted 969base pair open reading frame (ORF), which specifies a 323 amino acid protein with amolecular mass of 36970 daltons and a computed isoelectric point of 5.4998 (Fig. 5). ThePim- 1 protein sequence contains all the domains and conserved residues common to alleukaryotic protein kinases, including the critical lysine in subdomain II, important forphosphotransferase activity, and the sequence V-G-S-G-G-F-G-T-V (residues 46-54)conforming to the [L11/YJ-G-x-G-x-[ELY/M]-G-x-V protein kinase subdomain I region[Barioch and Claverie, 1988]. X. laevis Pim-1 also contains a serine/threonine kinasespecific signature pattern, V-V-H-R-D-I-K-D-E-N-L (residues 164-175) conforming to the[L[IJVJMIF/Y/C]-x-[jJLY]-x-D-[LIJLVIMJF/Yj-K-x-x-N4LLJJVJMIF/Y/C] motif. This differsfrom the tyrosine-specific kinase signature motif, [L/I/V/M/F/Y/C]-x-[H,Y]-x-D[L/TIVIM/FlY]- [R/S/T/AIC] -x-x-N-[L/I/V/MIF/Y/C]. Pim- 1 also features a consensuspattern for disulfide bond forming cysteines found in immunoglobulin (Ig) - related proteinsincluding the Ig constant chain domains and MHC extracellular domains [Beck and Barrel,97Figure 4. Southern blot of X. laevis eDNA clones probed with Pim-l PCR fragment.Clones isolated from the X. laevis cDNA library were digested with Acci and Ncol(lanes 2, 5, 8, 11, 14), HindIll and Pvull (lanes 3, 6, 9, 12, 15) or left undigested(lanes 1,4, 7, 10, 13). Clone 1.6, lanes 1-3; clone 1.1, lanes 4-6; clone 6.22, lanes 7-9;clone 12.35, lanes 10-12; human pim-1 in pGEX-2T (negative control), lanes 13-15.Panel A, an ethidium bromide stained agarose gel of digested plamids. The agarosegel was Southern blotted and probed with a radio-labelled X. laevis p/rn-i probe.Panel B is an autoradiograph of the Southern blot. Standards are shown on the left,estimated size of radioactive bands on the right. Clones 1.6 and 1.1 are duplicates.1 2 3 4 5 6 7 8 9 10 11 1213141598TAAATTTTTTTGTGACGGCCACGGATGTATGGTTACACGAGCTGAPTCTAPTATATACCTPGTCACGACGGATTCTCCCX3TGTCCAGARTGGGICACACACCACCCCGTTACTGGAGTCGACCTCAATPAAAGCGGCAACAAGGGAAGTACCGGCACAATACCCCAGAAGGGACCTUCPCAGTCAGACCCTAATTAGCCGCTAATGACATTACCGGAGCGTATrGAACM L L S K F G S L A H I C N P S N M E H L P V K 24ATGCTTCTCTCTAAATTCGGATCCCTGGCTCATATCTGCAACCCAAGCAACATGGACCATCTACCGCTTAAG 72I L Q P V K V D K E P F E K V Y Q V C S V V G S 48ATCTTACAGCCAGTGAAAGTGGACAAGGAGCCCTTCGAGAAGGTGTATCAGGTGGGCTCGGTTGTGGGCAGC 144ilk>G G F G T V Y S G S R I A D G Q P V A V K H V A 72GCTGGTTTCGGCACGGTGTATTCGGGCAGTCGGATTGCAGATGGACAGCCGGTCGCTGTGAAGNACGTAGCT 216K E R V T E W C T L N G V M V P L E I V L L K K 96AAGGAGACAGTCACAGAATCGGGCACTTTCAACGGCGTGATGGTCCCTTTGGAGATCGTCCTACTAAAGAAG 288V P T A F R G V I N L L D W Y E R P D A F L I V 120GTCCCCACCGCCTTCCGAGCCGTAATCAACCTACTCCACTGGTACGAGCGACCCGACGCCTTCCTGATCGTT 36012k>M E R P H P V K D L F D Y I T E K G P L D H D T 144ATGGAGAGACCGCAGCCGGTGA1GGATCTATTCGATTATATAACGGAAAAGGGGCCCCTGGACGAGGACACA 432A R G F F R Q V L E A V R H C Y N C C V V H R D 168GCCCGCGGTTTTTTCCGGCAGGTGCTGGAAGCGGTGCGACACTGCTATMCTGCGGGGTGGTGCATCGGGAC 504I K D E N L L V D T R N G E L K L I D F C S C A 192ATCAACGATGAGAACCTCCTCCTGCACACCAGCCCGCGAACTGAAACTGATCGATTTTGGCTCCGGGGCG 576L L K D T V Y T D F D G T R V Y S P P E W V R Y 216CTACTCAACGATACCGTGTACACGGATTTTGATCGAACCAGAGTCTATAGTCCACCAGAATGCGTCAGATAC 648H R Y 14 C R S A T V W S L G V L L Y D M V Y G D 240CACAGATACCATGGAAGATCAGCAACCGTGTGGTCTTTGGGTGTGCTTCTTTATGACATGGTTTACGGACAT 720I P F H Q 0 E H I V R V R L C F R R R I S T E C 264ATTCCCTTTGAGCAAGATGAAGAGATTTCGTGTCCCCTTGTGTTTCAGAAGAAGGATCTCTACAGAGTGC 792<13BQ Q L I K W C L S L R P 5 0 R P T L E Q I F D H 288CACCAACTTATCAAATGCTGCCTTTCCTTCAGGCCTTCTCATAGACCAACACTTCAGCAAATTTTTGACCAT 864P W M C K C 0 L V K S E 0 C 0 L R L R T I D N 0 312CCTTCGATGTGCAATGCGACCTTGTGAAATCTGAAGACTGTGATrrAAGACTAACCACAATTCACAATGAT 936<14BS S S T S S S N E S L 323TCATCAAGCACAACCTCAACCA7CGACACTCTG 969TAAAGTAAGCAATAVITCATATMTCCCATTACrAAGCACrCCTCCTCCCAACCTTACAAGGGGACCTCAAACATTTGCTGTTTTGGCTAAAACATTTTATACFigure 5. Nucleotide and amino acid sequence of coding region of X. laevis pim-l.The amino acid sequence of the coding region is shown above nucleotides, numberedin bold type. The nucleotides sequenced are numbered starting at “1” for the firstresidue of the start codon. Oligonucleotides used to amplify pim-1 from X. laeviscDNA were based on regions underlined, arrows indicate sense (>) and anti-sense(<)primers.9919881. Even though the sequence Y-N-C-G-V-V-H (residues 160-167) conforms to theconsensus pattern [F/]-x-C-x-]tA]-x-H, it is not clear if Pim-1 dimerizes or formsintramolecular disulfide bonds. X. laevis Pim- 1 also contains a coiled hydrophobic regionnear the 3’ end of the protein, from residues Leu282-Asp3Ol, that is conserved in Pim-lfrom all species. Other interesting regions that were identified in Pim- 1 include a potentialN-glycosylation site, (N-x-[S/TJ-x at residues 311-313) [Grand, 1989], a potential cAMP-dependent protein kinase site ([RIK]-[RIK]-x-[S)T] at residues 258-262) [Kennelly andKrebs, 1991], 4 potential protein kinase C phosphorylation sites ([S/T]-x-[RJK] at residues135, 144, 273 and 277) [Kennelly and Krebs, 1991], a potential casein kinase IIphosphorylation site ([S/T]-x-x-[D/E] starting at residue 200) [Pinna, 1990], a potentialtyrosine kinase phosphorylation site ([KIR]-x-x-x-[DIE]-x-x-Y at residues 33-4 1), and 5myristoylation sites (G-[EIDIRIK/HJPIF/Y/W]-x-x-[S/T/A/G/CIN]-P) starting at residues33, 43, 49, 52, 63 and 80). The physiological relevance of these sites is undetermined.X. laevis Pim- 1 shares a high degree of sequence similarity to the mammalian Pim- 1counterparts, especially within the catalytic region (Fig. 6). X. laevis Pim-1 amino acidsequence has 86% overall similarity and 65% overall identity with human Pim-1. While X.laevis Pim-1 N-terminus shares 69% similarity and 49% identity with human Pim-1, X.laevis Pim-1 C-terminus shares only 55% similarity and 15% identity with the humancognate. The C-terminus of the amphibian Pim-1 is eight amino acids longer than themammalian Pim-1 proteins and contains a high number of serine residues. The catalyticdomain of X. laevis pim-1 shares 71% nucleotide sequence similarity with the other pim-1sequences.100IXENOPUS MLLSKFGSLAHICNPSNMEHLPVKILQPVKVDKEPFEKVYQVGSVVGSGGFGTVYSGSRIMOUSE MLLSKINSLAHL -RARPCNDLHATKLAPGK-EKEPLESQYQVGPLLGSGGFGSVYSGIRVRAT MLJLSKINSLAHL -RAAPCNDLHANKLAPGK-EKEPLESQYQVGPLLGSGGFGSVYSGIRVHUMAN MLLSKINSLAHL -RAAFCNDLHATKLAPGK-- EKEPLESQYQVGPLLGSGGFGSVYSGIRV*****. ****. . . .* * * * .***.* ****. . .******.**** *.II III IVXENOPUS ADGQPVAVKHVAKERVTEWGTL-NGVMVPLEIVLLKKVPTAFRGVINLLDWYERPDAFL IMOUSE ADNLPVAIKHVEKDRI SDWGELPNGTRVPMEVVLLKKVSSDFSGVIRLLDWFERPDSFVLRAT ADNLPVAIKHVEKDRI SDWGELPNGTRVPMEVVLLKKVSSGFSGVIRLLDWFERPDSFVLHUMAN SDNLPVAIKHVEKDRI SDWGELPNGTRVPMEVVLLKKVSSGFSGVIRLLDWFERPDSFVL.* ***.*** *.*. . .** * ** **.*.******. . * *** ****.****.*.V VIXENOPUS VMERPEPVKDLFDYITEKGPLDEDTARGFFRQVLEAVRHCYNCGVVHRDIKDENLLVDTRMOUSE ILERPEPVQDLFDFITERGALQEDLARGFFWQVLEAVRHCHNCGVLHRDIKDENILIDLSRAT ILERPEPVQDLFDF ITERGALQEELARSFFWQVLEAVRHCHNCGVLHRDIKDENILIDLNHUMAN ILERPEPVQDLFDF ITERGALQEELARSFFWQVLEAVRHCHNCGVLHRDIKDENILIDLN****** • **** . *** . * . *. *• **. ** . ********* **** • ******** • * • *VII VIII IXXENOPUS NGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWVRYHRYHGRSATVWSLGVLLYDMVYGMOUSE RGEIKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRYHRYHGRSAAVWSLGILLYDMVCGRAT RGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRYHRYHGRSAAVWSLGILLYDMVCGHUMAN RGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRYHRYHGRSAAVWSLGI LLYDMVCG** • ****************************** • ********** • ***** • ****** *x XIXENOPUS DI PFEQDEEIVRVRLCFRRRISTECQQLIKWCLSLRPSDRPTLEQIFDHPWMCKCDLVKSMOUSE DIPFEHDEEIIKGQVFFRQTVSSECQHLIKWCLSLRPSDRPSFEEIRNHPWM-QGDLLPQRAT DI PFEHDEE IVKGQVYFRQRVS SECQHLIRWCLSLRPSDRPSFEEIQNHPWM-QDVLLPQHUMAN DI PFEHDEE I IRGQVFFRQRVSSECQHLIRWCLALRPSDRPTFEEIQNHPWM-QDVLLPQ*****•****• • • • **• •*•***•**•***•*******• •*•* •**** • *•XENOPUS EDCDLRLRTIDNDSS STSSSNESLMOUSE AASEIHLHSLSPGSSKRAT ATAEIHLHSLSPSPSKHUMAN ETAEIHLHSLSPGPSK* *Figure 6. Protein sequence alignment of Pim- 1 from X. laevis. mouse, rat and human.Stars (*) indicate residues identical and dots (.) indicate residues that are conserved betweenthe four species. Roman numerals indicate protein kinase subdomains. Residues that areidentical between all protein kinases are shown in bold type.1012. COMPARISON OF THE X. LAEVIS PIM-1 SEOUENCE WITH OTHERPROTEINSAs sequence similarities between proteins may indicate evolutionary relationships andprovide clues about domain function or protein-protein interactions, sequence searches wereperformed to identify other proteins sharing homology with X. laevis Pim- 1. The searchesmay also help to define important regulatory regions in Pim-1 based on strong conservation(e.g. phosphorylation sites). Most proteins that resulted from the search, including theother Pim-1 homologues, were serine/threonine protein kinases and displayed homology toPim- 1 in the catalytic domain. All proteins are listed with their accession number in boxedbrackets.i. Nucleotide sequence searchesNucleotide sequence comparisons were performed by NCBI using BLAST programs toidentify nucleic acid sequences sharing homology with the sequence of X. laevis pim-1. Allregions of homology (except with the other pim-] homologues) were located in the kinasecatalytic domain (nt 259-873). As expected, the nucleotide search identified other pim-1homologues including human pim-1 sequences [M16750, M27903, M54915], Rattusnorvegicus (rat) pim-1 [X63675] and Mus musculus (mouse) pim-1 [M13945]. X. laevispim-1 was 70-72% identical topim-1 from other species over a 615 nucleotide region in thecatalytic domain.Several sequences for calcium/calmodulin-dependent protein kinases were identified asbeing similar to pim-1 including rat calcium/calmodulin dependent kinase type II alphasubunit [J02942], murine calmodulin kinase type II [X14836], murine calcium/calmodulindependent kinase type II beta subunit [X63615], and Zea mays calcium-dependent proteinkinase [L27484]. These sequences were 6 1-69% identical to X. laevis pim-1 in kinasecatalytic subdomain domain VI. The rabbit gamma-subunit of phosphorylase kinase[Y00684] was identified as having 56% nucleotide sequence homology with X. laevis pim1 in kinase catalytic subdomains V-VT.ii. Protein sequence searchesProtein sequence comparisons were done by NCBI using BLAST programs to identifyprotein sequences that had some homology to the X. laevis Pim- 1. Most regions ofhomology (except with the other Pim- 1 homologues) were located within in the kinasecatalytic domain (aa 46-29 1). Apart from the other Pim-1 cognates, the sequence search did102not identify any kinases that were strongly related to X. laevis Pim- 1. The kinasesidentified as having the highest homology to Pim- 1 by the sequence search are summarizedin Table 1 along with the domains and identities.All proteins displaying homology to X. laevis Pim- 1 are protein-serine/threoninekinases. Most regions of homology were located in domains V-XI, with very few kinasessharing homology to Pim-1 in domains III and IV. Pim-1 has some homology to thecalcium/calmodulin-dependent kinases, SNF 1-related serine/threonine protein kinases(carbon catabolite derepressing protein kinase) as well as other signal translation kinasessuch as the 8-adrenergic receptor kinases and G protein-coupled receptor kinases. Theprotein homology search identified many members from these groups from a diverse rangeof species. As only the proteins with the highest homology were tabulated, the lesserrelated B-adrenergic receptor kinases and G protein-coupled kinases were not listed.Novel and interesting protein kinases identified by the search include p78, a humanputative serine/threonine protein kinase [P27448], murine tsk- 1 serine/threonine proteinkinase [S3 1333] [Bielke et a!., 1994], KKIALRE, human cdc-2-related serine/threonineprotein kinase [Q00532, S22745] [Meyerson eta!., 1992], and mouse ribosomal protein S6kinase II [P18653] [Alcorta et a!., 1989]. Pim-1 also had homology to the 35.1 kDaserine/threonine protein kinase from the African swine fever virus, a relationship reportedpreviously [Baylis et al., 1993].103TABLE 1. Kinases sharing homology with X. laevis Pim- 1.The name of the kinase (if known) as well as the species is listed. Identity reflects thepercentage of identical residues in the domains indicated and similarity reflects thepercentage of conserved substitutions plus identical residues in the domains indicated.Note that only a portion of a domain listed may be represented by the homology search.Accession Sub - Identity SimilarityKinase Species number domain (%) (%‘)A. Pim-1 serine/threonine protein kinasesH. sapiens [P11309] I-lI 75 83III-XI 75 89M. musculus [P06803] I-Il 75 83Ill-XI 75 89R. norvegicus [P26794] I-Il 75 83Ill-XI 75 89B. SNF1-related serine/threonine protein kinasesS. cerevisiae [P06782] I-lI 34 48V-IX 34 52A. thaliana [JC1446] 1-111 23 41V-DC 38 55RKIN 1 S. cereale (rye) [Q02723] V-DC 38 54BNASRKINB B. napus [L12394] V-VI 42 63BNASRKINA [L12393] VIII-X 46 55XI 32 47104TABLE 1. Kinases sharing homology with X. laevis Pim- 1.. .continued.C. Calcium/calmodulin-depenclent protein kinasesG. max [P28583] I-Il 35 56V-VT 42 61VII-X 38 61XI-N 27 45E. nidulans [JN0323] IV 36 68V-VT 44 62IX-X 31 53XI 34 56D. carota (carrot) [P28582] V-Vll 36 57IX-X 38 61XI 37 56type I S. cerevisiae [P27466] I-Il 35 50IV 31 57V-VT 37 62VIl-IX 40 70XI 35 59type II S. cerevisiae [P27466] I-lI 34 50lv 31 57V-VT 37 62IX 33 59XI 35 59type II S. cerevisiae [X65797] V-VT 40 57IX 41 66XI 40 62type II S. cerevisiae [P22517] I-il 28 541V 36 63V-VT 42 59IX 41 66XI 40 62Cam kinase II S. cerevisiae [B40896J I-lI 28 54lv 36 63V-VT 42 59IX 41 66XI 40 62Sch9 S. cerevisiae [P11792] V-X 32 48[X57629]105TABLE 1. Kinases sharing homology with X. laevis Pim-1...continued.D. Other serine/threonine protein kinases (most poorly characterized’)African swine fever virus kinase [S27892] II 50 71rn-VT 22 44VI-Vil 39 63VIII-XI 25 45KKTALRE (cdc-rel) H. sapiens [Q00532] I-IT 30 46[S22745] VI 46 69VII 47 64VIlI-IX 31 53XI 41 61niml+/cdrl S. pombe [A25958] V-VU 42 60[P07334] VIll-IX 37 56XI 35 54YALO17 serine/ S. cerevisiae [P31374] V-VT 47 72threonine protein kinase VII-X 38 60XI 34 59Abscisic acid-inducible T. aestivum [Q02066] V-VT 42 65serine/threonine VIII-X 40 50protein kinase XI-N 26 47tsk- 1 M. musculus [X70764] I-lI 28 48VIT-1X 35 52X-XI 41 56CDC5IPYK2 S. cerevisiae [P32562] V-VT 28 52VU-DC 32 47XI 46 67? M. crystallinum [Z26846] V-VT 42 65VIII-X 38 51XI 38 52? A. thaliana [L05562] V-VT 42 65VU-X 41 51XI 37 48? A. thaliana [M91548] V-VT 42 63VII-X 46 55XI 29 44A. thaliana [L05561] V-VT 40 63VII-X 41 51XI 32 47A. thaliana [S24586] V-VT 42 63VII-X 36 50XI 38 55106kern kinase M. musculus [S31333] I-Il 28 48V-Vll 35 52X-XI 41 56M. musculus [U01840] V-VT 34 57Vu-TX 39 62X-XI 26 51PK 1 G. max (soybean)[L01453] I-IT 27 45V-VT 44 67VIll-X 40 50XI 34 51PK 2 G. max (soybean) [L19360] I-lI 26 44V-VT 44 67VllI-X 40 50XI 34 48PK 3 G. max (soybean) [L19361] V-VT 42 63Vffl-X 36 51XI 35 52Hypothetical serine/ S. cerevisiae [P253 89] V-X 32 54threonine protein kinase XI 33 46p78 serine/threonine H. sapiens [P27448] I-Il 26 48protein kinase V-IX 37 53X-XI 37 54protein tyrosine S. cerevisiae [B28036] V-X 28 50kinases 1 &2 [A28036] XI 28 55BKIN12 H. vulgare [S24579] 1-11 33 51V-X 38 53XI 35 51BKIN12 H. vulgare [S24578] 1-11 30 48V-TX 38 53XI 35 54KIN1 S. pombe [P22987] I-Il 26 47[M64999] V-DC 33 50XT 36 63KIN2 S. cerevisiae [P13186] V-X 28 49XI 28 55Mkk2 S. cerevisiae [P3249 1] V-VT 28 52Vil-Vifi 35 51TX 50 68XI 33 55107iii. Pim-1 homology treeA Pim-1 homology tree was constructed using PHYLIP [Felsenstein, 1993] todemonstrate the relationship of the X. laevis Pim- 1 to related proteins. A protein sequencesearch was performed using genetic data environment (GDE), then X. laevis Pim-1 wasaligned to the 40 closest relatives using clustal. Because of limitations of the program, onlyshort regions of homology could be compared so the less related regions (the divergent N-and C-termini) had to be deleted to allow the alignments to be plotted on the tree. GDE usesthe Desoete Tree Fit algorithm to determine the relationship of the proteins to each other.The protein alignment is shown in Figure 7, and displays catalytic domains VI-XI of Pim-1(amino acids 15 1-294) aligned with the homologous regions of other proteins.PHYLIP was used to build a tree from the species alignments; species were assessed inthe order at which they appeared in the input file and as each additional species was added,the best tree was selected from all possible tree combinations. Local rearrangementsoccurred and if a better tree was made, the new rearrangement was accepted, guaranteeingthat the best possible tree was constructed. With this unrooted, additive tree model thegenetic distances are expected to equal the sums of vertical branch lengths between thespecies, with the tree “growing” from the left to the right.The Pim- 1 relatedness tree is shown in Figure 8, with a list of the names and accessionnumbers of the kinases shown in Table 2. The X. laevis Pim-1 is located near the far rightof the tree. Its closest relatives are, not surprisingly, the human, murine and rat Pim- 1sequences. The mouse and rat Pim-1 are most closely related to each other, then with thehuman Pim-1. The X. laevis Pim-1 is somewhat divergent from these other three species.The Pim-1 proteins do not have many close relatives, except the kinase encoded by cosmidc06E8 of C. elegans (celcO6e8). At this time the function is of this kinase is unknown.The next closely related protein to the Pim-1 family is a SNF1 homologue from wheat(a53467) followed by murine tsk- 1 testes-specific serine/threonine kinase (uO 1840), humanp78 serine/threonine kinase that is lost during chemically-induced pancreatic tumors(kp78_human), and calcium/calmodulin-dependent protein kinase from E. nidulans(jn0323). The murine, human and E. nidulans proteins were identified in the proteinsequence search in the previous section.The remainder of the proteins shown in Figure 8 all share a low amount of relatednessto the X. laevis Pim- 1 but high relatedness to each other. Many of these proteins are108Ca/calmodulin-dependent protein kinases including the rat IV B (S65840) and II a subunitsfrom cerebellum (kcc4_rat), mouse IV catalytic chain (kcc4_mouse), human (a53036) andrat lung Ca/calmodulin-dependent protein kinase (ratcampkaa) as well as thephosphorylase B kinase gamma catalytic chain from human (kpbh_human) and rat(kbph_rat). Members of a second group of calcium/calmodulin-dependent protein kinasesthat are related to each other include the type IV beta chains from mouse (kccb_mouse), rat(kccb_rat) and X. laevis (x1u06636), the type II alpha chain from Drosophila (ju0270),mouse (kcca_mouse) and rat (kcca_rat).The next group of kinases include uncharacterized protein kinases from barley (S24578,S24579) and homologues to SNF1, carbon catabolite derepressing protein kinase necessaryfor the release from glucose repression, from rye (rkil_secce), A. thaliana (jc1446),tobacco (tobpki) and rat (rnampapk). Two additional proteins shown are the yeast proteins,Ca-’-/calmodulin-dependent protein kinase (kcc4_yeast) and probable serine/threonine kinaseYKL1O1W (kkkl_yeast), both previously identified in the protein homology search.One difficulty with this program is that the limits of divergence in the homology regionsallow only the catalytic regions with the highest homology to be assessed. The program isdependent on the order that the species are presented into the input file, so to obtain the mostaccurate tree, a jumble’ option must be used to randomly input the species. Because thealignment was performed by computer using a mathematical program, the program does notalways select the most obvious residues to align. In addition, this program does not knowwhich changes have happened in the past (i.e. bases changing and then changing back) socannot predict how the sequences evolved or where they diverged from a common ancestor.This gives the results from this analysis limited statistical significance. Only a few proteinswere common to both protein sequence searches; both searches limited to the number ofproteins to be identified (100 for the previous search, 30 for this search) as there are manyproteins which share this low level of homology with Pim- 1.Most of the proteins on the homology tree are more closely related to each other than tothe Pim- 1 proteins. As all the proteins plotted on the tree are serine/threonine kinases eitherinvolved in cellular metabolism or cell cycle regulation, it is likely that Pim- 1 is involved inone of these processes.109Figure7.ProteinsequencealignmentsofX. laevis Pim-lwithhomologousproteins.Thisaugment wasusedtoconstructthehomologytree(Fig.8).Romannumeralsindicateproteinkinasecatalyticsubdomains.VIVIIVIIIkpim_xenQVLEAVRH-CYNCGVVHRDIKDENLLVDTR--NGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWVRYHRYHGkpim_ratQVLEAVRI-J-CHNCGVLI-IRDIKDENILIDLN--RGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRYHRYHGkpirn_mouQVLEAVRH-CHNCGVLF{RDIKDENILIDLS--RGEIKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRYHRYHGkpim_humQVLEAVRH-CI-INCGVLI-IRDIKDENILIDLN--RGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRYHRYHGcelcO6e8QVITTVFNNYSKHGLLI-{RDIKDENLIVNMN--TGEVKLVDFGATAYAEKATKKEFQGTRSYCPPEWFRDQLYLPtobpkiQIISGVEY-CHRNMVVI-{RDLKPENLLLDSKWN---VKIADFGLSN---IMRDGHFL---KTSCGSPNYAAPEVISGKLYAGjc1446QIISGVEY-CHRNNVVHRDLKPENLLLDSKCN---VKIADFGLSN--IMRDGHFL---KTSCGSPNYAAPEVISGKLYAGs24579QILAGVEY-CHRIMVVHRDLKPENLLLDSRYN---VKLADFGLSN--VMRDGHFL---KTSCGSLNYAAPEIISSKLYAGs24578QILAGVEY-CHRIMVVHRDLKPENLLLDSKYN---VKLADFGLSN--VMRDGHFL---KTSCGSLNYAAPEIISSKLYAGrkil_secQIISAVEY-CHRNKVVI-IRDLKPENLLLDSKYN----VKLADFGLSN--VM}-IDGHFL---KTSCGSLNYAAPEVISGKLYAGa53467HLISAVGF-CI-{TRGVFHRDLKPENLLVD---EAGNLKVSDFGLSAVAEPFQPEGLL--HTFCGTRAYVAPEVLARRGYEGu01840QLSSAIKY-CI-IDLDVV}-IRDLKSENLLLDKDFN---IKLSDFGFSKRCLRDDSGRLILSKTFCGSAAYAAPEVLQGIPYQPkcc4_mouQIL,EAVAY-LHENGIVHRDLKPENLLYATPAPDAPLKIADFGLSKIVE-HQV----LMKTVCGTPGYCAPEILRGCAYGx1u06636QILEAVLH-CHQMGVVHRDLKPENLLLASKCKGAAVKLADFGLAIEVQGEQQ----AWFGFAGTPGYLSPEVLRKEAYGs65840QILEAVAY-LI-IENGIVHRDLKPENLLYATPAPDAPLKIADFGLSKIVE-HQV----LMKTVCGTPGYCAPEILRGCAYGrnampapkQILSAVDY-CI-{RHMWHRDLKPENVLLDAQMN---AKIADFGLSN--MMSDGEFL---RTSCGSPNYAAPEVISGRLYAGratcampkQVLDAVKY—LHDLGIVHRDLKPENLLYYSLDEDSKIMISDFGLSKMED-PGS----VLSTACGTPGYVAPEVLAQKPYSkp78_humQIVSAVQY-CHQKRIVHRDLKAENLLLDADMN---IKIADFGFSN--EFTVGGKL---DTFCGSPPYAAPELFQGKKYDGkccb_ratQILEAVLH-CI-IQMGVVHRDLKPENLLLASKCKGAAVKLADFGLAIEVQGDQQ—---AWFGFAGTPGYLSPEVLRKEAYGkccb_mouQILEAVLH-CHQMGVVHRDLKPENLLLASKCKGAAVKLADFGLAIEVQGDQQ----AWFGFAGTPGYLSPEVLRKEAYGkcca_ratQILEAVLH-CHQMGVVHRDLKPENLLLASKLKGAAVKLADFGLAIEVEGEQQ----AWFGFAGTPGYLSPEVLRKDPYGkcca_mouQILEAVLH-CHQMGVVHRDLKPENLLLASKLKGAAVKLADFGLAIEVEGEQQ----RWFGFAGTPGYLSP4EVLRKDPYGkcc4yeaQIIIGISY-CHALGIVHRDLKPENLLLDSFYN---IKIADFGMAA--LQTDADLL---ETSCGSPHYAAPEIVSGLPYEGkcc4_ratQILEAVAY-LHENGIVHRDLKPENLLYATPAPDAPLKIADFGLSKIVE-I-{QV----LMKTVCGTPGYCAPEILRGCAYGa53036QILEAVAY-LHENGIV}-IRDLKPENLLYATPAPDAPLKIADFGLSKIVE-I-IQV----LMKTVCGTPGYCAPEILRGCAYGkpbh_hurnSLLEAVSF-LHANNIVHRDLKPENILLDD---NMQIRLSDFGFSCI-{LEPGEK----LR-ELCGTPGYLAPEILKCSMDETHPGYGkkkLyeaQIVEGVSY-CHSFNICHRDLKPENLLLDKKNR--RIKIADFGMAA--LELPNKLL---KTSCGSPHYASPEIVMGRPYHGkbphratSLLEAVNF-LHV1NIVHRDLKPENILLDD---NMQIRLSDFGFSCHLEPGEK----LR-ELCGTPGYLAPEILKCSMDETHPGYGju0270QILESVNH-CHQNGVVHRDLKPENLLLASKAKGAAVKLADFGLAIEVQGDHQ----AWFGFAGTPGYLSPEVLKKEPYGjn0323QVLDAVNY-LHQRNIVHRDLKPENLLYLTRDLDSQLVLADFGIAKMLDNPAE----VLTSMAGSFGYAAPEVMLKQGHGContinued...CFigure7....continued.IxxXIkpim_xenRSATVWSLGVLLYDMVYGDIPFEQDEEIVRVRLCFR----RRISTECQQLIKWCLSLRPSDRPTLEQIFDIIPWMC*kpim_ratRSAAVWSLGILLYDMVCGDIPFEHDEEIVKGQVYFR----QRVSSECQHLIRWCLSLRPSDRPSFEEIQNJ{PWMQ*kpim_mouRSAAVWSLGILLYDMVCGDIPFEI-IDEEIIKGQVFFR----QTVSSECQHLIKWCLSLRPSDRPSFEEIRNHPWI4Q*kpim_humRSAAVWSLGILLYDMVCGDIPFEHDEEIIRGQVFFR----QRVSSECQHLIRWCLALRFSDRPTFEEIQNHPNNQ*celcO6eSLEATSWSLGVLLFILLTGKLPFRNEIQICLGNVKFP----PDLSKEVCQLVKSCLTTSTSARASLAQIAAHPWME*tobpkiPEVDVWSCGVILYALLCGTLPFDDEN--IPNLFKKIKGGMISL--PSH--LSAGARDLIPRNLIVDPMKRNTIPEIRMHPWFQ*jc1446PEVDVWSCGVILYALLCGTLPFDDEN--IPNLFKKIKGGIYTL--PSH--LSPGARDLIPRMLVVDPMKRVTIPEIRQHPWFQ*s24579PEVDVWSCGVVLYALLCGSVPFDDDN--IPSLFRKIKGGTYIL--PSY--LSDSARDLIPKLLNIDPMKRITFHEIRV}lPWFK*s24578PEVDVWSCGVILYALLCGSVPFDDDN--IPSLFRKIKGGTYIL--PSY--LSDSARDLIPKLLNIDPMKRITIHEIRVI{PWFK*rkil_secPEIDVWSCGVILYALLCGAVPFDDDN--IPNLFKKIKGGTYIL--PIY--LSDLVRDLISRMLIVDPMKRITIGEIRK}ISWFQ*a53467AKADIWSCGVILFVLMAGYLPFHDQN--LMAMYRKFTRES---SMSRW--FSKDLTSLIMRFLDTNPSTRIT----_LPESW__*u01840KVYDIWSLGVILYIMVCGSMPYDD--SNIKKL-RIQKEHRVNF--PRSKHLTGECKDLIYRMLQPDVNRRLHIDEILNHCWVQ*kcc4_mouPEVDMWSVGIITYILLCGFEPFYDERGD-QFMFRRILNCEYYFISPWWDEVSLNAK]JLVKKLIVLDPKKRLTTFQALQJIPWVT*x1u06636KPVDIWACGVILYILLVGYPPFWDEDQH-K-LYQQIKAGAYDFPSPEWDTVTPEAKNLINQMLTINPAKRITAI-{EALK}IPWVC*s65840PEVDMWSVGIITYILLCGFEPFYDERGD-QFMFRRILNCEYYFISPWWDEVSLNAKDLVKKLIVLDPKKRLTTFQALQHPWVT*rnampapkPEVDIWSCGVILYALLCGTLPFDDEH--VPTLFKKIRGGVFYI--PEY--LNRSIATLLMHMLQVDPLKRATIKDIREHEWFK*ratcarnpkKAVDCWSIGVIAYILLCGYPPFYDENDA-K-LFEQILKAEYEFDSPYWDDISDSAKDFIRHLMEKDPEKRFTCEQALQHPWIA*kp78_humPEVDVWSLGVILYTLVSGSLPFDGQ--NLKELRERVLRGKYRI--PFY--MSTDCENLLKRFLVLNPIKRGTLEQIMKDRWIN*kccb_ratKFVDIWACGVILYILLVGYPPFWDEDQH-K-LYQQIKAGAYDFPSPEWDTVTPEAKNLINQMLTINPAKRITAHEALKI{PWVC*kccb_mouKPVDIWACGVILYILLVGYPPFWDEDQH-K-LYQQIKAGAYDFPSPEWDTVTPEAKNLINQMLTINPAKRITAI-IEALKHPWVC*kcca_rat.KPVDLWACGVILYILLVGYPPFWDEDQH-R-LYQQIKAGAYDFPSPEWDTVTPEAKDLINKMLTINPSKRITAAEALKHPWIS*kcca_mouKPVDLWACOVILYILLVGYPPFWDEDQH-R-LYQQIKARAYDFPSPEWDTVTPEAKDLINKMLTINPSKITAAEALK}IPWIS*kcc4yeaFASDVWSCGVILFALLTGRLPFDEENGNVRDLLLKVQKGQFEM--PNDTEISRDAQDLIGKILVVDPRQRIKIRDILSHPLLK*kcc4_ratPEVDMWSVGIITYILLCGFEFFYDERGD-QFMFRRILNCEYYFISPWWDEVSLNAKDLVKKLIVLDPKKRLTTFQALQHPNVT*a53036PEVDMWSVGIITYILLCGFEPFYDERGD-QFMFRRILNCEYYFISPWWDEVSLNAKDLVRKLIVLDPKKRLTTFQALQ}IPWVT*kpbh_humKEVDLWACGVILFTLLAGSPPFWH-RRQ-ILMLRI’4IMEGQYQFSSPEWDDRSSTVKDLISRLLQVDPEARLTAEQALQHP__*kkklyeaGPSDVWSCGIVLFALLTGHLPFNDDN--IKKLLLKVQSGKYQM--PSN--LSSEARDLISKILVIDPEKRITTQEILKF{PLIK*kbph_ratKEVDLWACGVILFTLLAGSPPFWH-RRQ-ILMLRMIMEGQYQFSSPEWDDRSNTVKDLIAKLLQVDPNARLTAEQALQHP___*ju0270KSVDIWACGVILYILLVGYPPFWDEDQH-R-LYSQIKAGAYDYPSPEWDTVTPEAKNLINQMLTVNPNKRIT*jn0323KAVDIWSLGVITYTLLCGYSPFRSENLT-D-LIEECRSGRVVFHERYWKDVSKDAKDFILSLLQVDPAQRPTSEEALKHPWLK*jn0323-kpim_mousekpim_rat________kirnxenIarceIcO6e8—1a53467kp78_humanuOl840kcc4_rats65840kcc4_mouse_____La53036_______—ratcampkaa________________________________kbphjatL_kpbh_humankccb_mousekccb_ratx1u06636_________—ju0270—kcca_mousekcca_rat_____s24578_{s24579kcc4yeastFigure8.HomologytreeofPim-1withrelatedkinases.X.laevisPim- 1islocatedtothetoprightofthetree, withtheotherPim-1sequencesincludingmouse,ratandhuman.Thefullnamesandaccessionnumbersof theotherkinasesshownarelistedinTable2.Table 2. Names and Accession numbers of kinases listed in the Pim- 1 relatedness tree.‘= proteins also identified in protein sequence search, Section 2.ii.Name of kinase Accession numberkpim_xenla frog Pim-1 L29496kpim_human human Pim-1 P11309 *kpim_mouse mouse Pim- 1 P06803 *kpim_rat rat Pim-1 P26794 *celcO6e8 C. elegans. cosmid c06E8 U00034a53467 wheat SNF1 homologue A53467kp78_human human p78 serine/threonine kinase P27448 *u01840 mouse tsk-1 testes-specific serine/threonine kinase u01840 *jn0323 E. nidulans Ca/calmodu1in-dependent protein kinase jn0323 *kcc4_rat rat Ca/calmodu1in-dependenttype II u. subunit P 13234S65840 rat Ca/calmodulin-dependent type 1V B subunit S65840kcc4_mouse mouse Ca/ca1modulin-dependent IV catalytic chain P084 14 *a53036 human Ca/ca1modulin-dependent PK homologues a53036ratcampkaa fetal rat lung Ca/calmodulin-dependent PK L26288kbph_rat rat phosphorylase 13 kinase gamma catalytic chain P31325 *kpbh_human human phosphorylase B kinase gamma catalytic chain P 15735 *kccb_mouse mouse CaVcalmodulin-dependent type IV beta chain P28652 *kccb_rat rat Ca/calmodulin-dependent type IV beta chain P08413 *x1u06636 frog Ca/caImodu1in-dependent type IV beta chain x1u06636ju0270 fly Ca/calmodu1in-dependent type II alpha chain ju0270 *kcca_mouse mouse Ca/calmodu1in-dependent type II alpha chain P11798kcca_rat rat Ca/calmodulin-dependent type II alpha chain P11275S24578 barley uncharacterized protein kinase S24578 *S24579 barley uncharacterized protein kinase S24579 *rkil_secce rye SNF1 homologue Q02723 *jc1446 A. thaliana SNF1 homologue jc1446 *tobpki tobacco SNF1 homologue D26602rnampapk rat SNF1 homologue Z29486kcc4yeast yeast Ca/calmodu1in-dependent protein kinase kcc4yeastkkkl....yeast yeast probable serine/threonine kinase P342441133. BACTERIAL EXPRESSION OF PJM-1 AS A GST-FUSION PROTEINThe pGEX plasmids allow recombinant genes to be expressed at high intracellular levelsin bacteria as fusion proteins with the Schistosoma japonicum glutathione S-transferase.The expression is chemically inducible; the fusion protein is under the control of the lacpromoter which is induced with the lactose analog isopropyl B-D-thiogalactoside (IPTG).The expressed fusion protein is easily purified from bacterial lysates under non-denaturingconditions by an affinity interaction with glutathione covalently attached to Sepharose 4Band can be eluted from the affinity matrix with glutathione. A thrombin digestion site isencoded between the GST and the fusion protein to allow cleavage of the purified proteinfrom the GST.i. Expression of human Pim-l as a bacterial GST-fusion proteinThe human pim-1 coding region was subcloned into the pGEX-2T expression vectorusing non-degenerate oligonucleotides based on the published nucleotide sequence ofhuman pim-1. The polymerase chain reaction was optimized to discourage the amplificationof unwanted byproducts, by employing stringent annealing conditions. VENT polymerasewas used in the PCR reaction because its high accuracy and proof-reading function wereessential to maintain the sequence integrity of the PCR product. The expression of thepredicted 62 kDa recombinant fusion-protein was confirmed by Western blotting withseveral different Pim-1 antibodies as well as with an anti-GST antibody (Fig. 9). Whilemost of the GST-Pim- 1 existed as a 62 kDa fusion protein, lower molecular mass speciesbetween 28 and 40 kDa were also detectable on Western blots with several of theantibodies, even in the presence of protease inhibitors. These bands may have arisen fromspecific proteolytic cleavage or aborted translation of the protein. While all antibodiesimmunodetected the full-length fusion protein, Piml-NT detected 30, 33 and 38 kDa bandswhile Pimi-ifi detected 38 and 40 kDa bands and the GST antibody detected a smear from28-32 and 38 kDa bands. The X. laevis antibodies did not detect the full-length humanfusion protein. Tel, Lilly cocktail and CRB antibodies strongly detected the full-lengthfusion protein and CRB and 4G10 also detected a 32 kDa band. Neither of the antiphosphotyrosine antibodies detected the full-length fusion protein. As all three antibodiesdirected against regions in the N-terminus detect lower Mr proteins, the lower molecularmass forms are possibly the result of aborted translation.The human GST-Pim-1 preparation was catalytically active and autophosphorylated invitro. In an autophosphorylation assay, the two lower molecular mass forms (32 and 38114kDa) were phosphorylated in addition to the 62 kDa band predicted to be the full-lengthGST-Pim-1 (Fig. 9, lane 12). These bands correspond with the three main immunoreactivebands detected by the Pim- 1 antibodies. It is uncertain if all the species are catalyticallyactive, or if the observed phosphorylation is a result of cross-phosphorylation by one activeprotein. Attempts to resolve these Pim-l species by various forms of columnchromatography (e.g. MonoQ and Superose) were unsuccessful.The expression of GST-Pim- 1 was compared between two different E. coli strains,DH5c and UT5600, to determine if a higher proportion of the fusion protein would beproduced in the UT5600 protease deficient strain (Fig. 10). Although protein expressionwas significantly higher in the protease-deficient UT5600 strain, the amounts of the full-length fusion protein relative to the smaller mass bands remained the same. The UT5600E. coli strain was used for the remainder of the expression experiments.Phosphoamino acid analysis of human GST-Pim- 1 was performed in order to determinethe nature of the residues that were phosphorylated in an in vitro autophosphorylationreaction (Fig. 11, Panel A). Although the majority of phosphorylation was on threonineresidues, there was also phosphorylation on serine and tyrosine residues as well. Thephosphorylation on tyrosine was unexpected, as antiphosphotyrosine antibodies did notdetect the full-length human GST-Pim-1 by Western blotting (Fig. 9, lanes 9, 10) andrecent studies have reported that the Pim- 1 autophosphorylated strictly on serine andthreonine residues [Saris et a!., 1991; Hoover et al. 1991; Padma and Nagarajan, 1991;Friedmann et a!., 1992]. This tyrosine autophosphorylation activity of GST-Pim-1 will beexamined further in Chapter VI.By comparing the relative amounts of the full-length fusion protein with BSA standardson an SDS-PAGE gel, the percentage of full-length human fusion protein in the glutathioneaffinity-purified Pim- 1 preparation relative to the total protein concentration was estimated tobe 1.8%. As this quantity was only an estimate, and because we were unable to determineif the lower species were catalytically active, the protein concentration of the entire samplewas used for all experiments unless otherwise indicated.115Western blot, silver stain and autoradiograph of H. sapiensGST-Pim-1. Western blots (lanes 1-10) of 0.75 ug of GSTPim-1 expressed in DH5a probed with various antibodies:lane 1, anti-GST; lane 2, Piml-NT; lane 3, Pimi-IlI; lane 4,GXP; lane 5, Piml-XI; lane 6, Tel; lane 7, Lilly; lane 8, CRB;lane 9, 4G10; lane 10, PY2O. Lane 11, silver stain of 0.75 ugof H. sapiens GST-Pim-1. Lane 12, autoradiogram of 0.75 ugof autophosphorylated GST-Pim- 1. Migrations of Mr standardsare shown at right, and arrow idicates the mobility of GST-Pim- 1.1 2 3 4 5 6 7 8 9 10Fianre 9. Bacteriallv-exnressed H. sapiens GST-Pim-l.116106 kDa—Figure 10. Expression ofhuman GST-Pim-lin DH5 and UT5600 E. coli strains.Approximately 5 i1 of affinity purified GST-Pim-1expressed m DH5cr (lanes 1 and 3) and UT5600(lanes 2 and 4) strains of E.coli were electrolutedand silver stained (lanes 1 and 2) or Western blottedwith Piml-NT antibody (lanes 3 and 4).80—GST-Pim-1 [49.5 —32.5 —27.5 —18.5—117Figure 11. Phosphoamino acid analysisof bacterially-expressed GST-Pim- 1.s H. sapiens and X. laevis GST-Pim- 1 were autophosphorylated in vitro and electroluted on anT SDS-PAGE gel. The radiolabelled band corresponding to the full-length fusion protein wasexcised and subjected to phosphoamino acidanalysis. Standards were visualized by ninhydrin.y The autoradiogram is shown at right with thelocation of the standard phosphoamino acidsshown.Human X. laevisGST-Pim- 1 GST-Pim- 1118a. Attempted separation of different Pim- 1 fragments by column chromatographyTo distinguish the activity of the full-length fusion protein from the smaller products,attempts were made to separate the different species by MonoQ column chromatography andby Superose gel exclusion. Uncleaved GST-Pim- 1 was loaded onto these columns andeluted by the standard protocols, and the fractions analyzed by peptide substrate assaysusing the P4 peptide (AKRRRLSA), in vitro autophosphorylation assays and by Westernblotting. These experiments indicated that the three main bands present in expressed humanGST-Pim- 1 preparations were not resolvable by the column chromatography methodsattempted.b. Thrombin cleavage of human GST-Pim-1Thrombin treatment of the GST-Pim- 1 fusion protein was done in order to isolate thePim-1 portion. Although there was a thrombin-cleavage site located at the GST-Pim-1junction, cleavage with this enzyme reduced the activity of the kinase. This is likely due tothermolability; GST-Pim- 1 is stable and retains most of its activity after a 25 mm incubationat 30°C, but begins to lose activity after 35 mm at 30°C. Reducing the time of incubationfrom 60 mm, reduced the efficiency of thrombinization. A large proportion of the cleavedproduct remained bound to beads, and the methods used to separate the thrombin from thecleaved product led to a further reduction in activity. Because of the inefficiency of thethrombinization reaction and the low activity of the end product, the uncleaved fusionprotein was utilized in subsequent studies to obtain maximal activity.ii. Expression of X. laevis Pim- 1 as a bacterial fusion proteinX. laevis Pim- 1 was subcloned into the pGEX-2T vector and expressed in UT5600protease-deficient bacteria. Analysis by Western blotting with X. laevis-specific Pim- 1antibodies as well as the anti-GST antibody demonstrated that the expected 549 amino acid,63.3 kDa fusion protein was produced (Fig. 12). Unlike the expression of human GSTPim- 1 fusion protein, expression of X. laevis GST-Pim- 1 yielded a relatively purepreparation with few degradation products. GXP antibody detects a plethora of proteinsthat may have also been present in the original immunizing preparation. Pim 1 -XI detectsonly the full-length protein, indicating that the lower mwt. bands observed in lane 2 areunlikely to be due to Pim-1 degradation or production of alternative Pim-1 products. TheX. laevis GST-Pim-l is immunodetected by both antiphosphotyrosine antibodies as well asby some of the human Pim-1 antibodies, most notably the Tel antibody (lane 6).119—96.5—65.54—46.8—29.2—20.211 12Figure 12. Bacterially-expressed X. laevis GST-Pim-1.Western blot, silver stain and autoradiograph of X. laevisGST-Pim-1. Western blots (lanes 1-10) of 0.75 ug of GSTPim- 1 expressed in DH5c cells.probed with variousantibodies: lane 1, anti-GST; lane 2, GXP; lane 3, Piml-XI;lane 4, Piml-NT; lane 5, Pimi-ifi; lane 6, Tel; lane 7, Lilly;lane 8, CRB; lane 9, 4G10; lane 10, PY2O. Lane 11, silverstain of 0.75 ug of X. laevis GST-Pim-1. Lane 12, autoradiogram of 0.75 ug of autophosphorylated GST-Pim- 1.Migrations of Mr standards are shown at right, arrow indicatesthe mobility of GST-Pim- 1.1 2 3 4 5 6 7 8 9 10120The full-length 62 kDa fusion protein was the major species radiolabelled in anautophosphorylation reaction of X. laevis GST-Pim-1 (Fig. 12, lane 12). A very faint 45kDa doublet was also visible, along with an even fainter band of about 15 kDa.Phosphoamino acid analysis revealed that like the human protein, the X. laevis GST-Pim-1autophosphorylates on serine and threonine as well as on tyrosine residues (Fig. 11).Interestingly, the proportion of radiolabel incorporated by each residue differs between thetwo enzymes.The X. laevis GST-Pim-1 protein was treated with thrombin to cleave the GST portionaway from the Pim- 1. The kinase activity of the thrombin-treated Pim- 1 was reduced to2.5% of the uncleaved control (data not shown). Staining of the glutathione-agaroseimmobilized sample revealed that although the thrombinization reaction was quite efficient,again most of the cleaved product adhered to the beads. Because of low activity and lowprotein recovery, the full-length GST-Pim-1 was used for all further experiments.iii. Expression of a kinase-dead mutant of X. laevis GST-Pim-1A kinase-dead (KD) mutant of X. laevis GST-Pim-1 was created as a negative control forpeptide substrate assays and to ensure that the GST portion of the fusion protein and othercontaminating or co-purifying proteins did not contribute to kinase activity. The KD Pim-1mutant was created by changing Lys-69, a residue required for ATP binding, to an alanineresidue (for another example of where this has been done, refer to Taylor, 1989). Althoughlysine has a significantly larger side chain, the substitution of alanine for lysine is unlikely tocause a major conformational change in the protein, nor should it interfere with therecognition sites for other substrates. The residue was altered by PCR site-directedmutagenesis using the wild-type (WT) X. laevis GST-Pim-1 as a template, and wasexpressed and purified in the same manner as for WT GST-Pim- 1. Western blot analysiswith X. laevis Pim- 1 antibodies revealed that an approximately 63 kDa protein was produced(Fig. 13).The KD mutant was not radiolabelled in a in vitro autophosphorylation reaction, eventhough other radiolabelled proteins were observed in the preparation (Fig. 13, lane 8). A 16hr expsoure is shown in Figure 13. However, no labeling of the KD GST-Pim-1 wasobserved even after the autoradiogram was exposed for 2 weeks (data not shown). Aradiolabelled 70 kDa protein was present in the KD preparation; as this protein was alsodetected by GXP antibody, it was probably part of the original immunizing preparation. Thisprotein was possibly the product of the bacterial dnaK gene, a close homologue of the121eukaryotic HSP family of proteins [Craig and Gross, 19911. This protein, sometimes calledchaperonin, is a major component of normally growing cells, binds to proteins and is thoughtto be important for protein motility within the cell [Craig and Gross, 1991; Leustek et al.,1992; Yu-Sherman and Goldberg, 19921. The expression of the 70 kDa protein increases inresponse to stress and may bind preferentially to abnormal (expressed) proteins along withother HSP (grpE, La) enhancing susceptibility to cellular proteases [Craig and Gross, 1991;Leustek et al., 1992; Yu-Sherman and Goldberg, 1992]. Interestingly, the dnaK proteinbinds ATP with high affinity and possesses a weak ATPase activity [Leustek et al., 1992; YuSherman and Goldberg, 19921. This could account for the radioactive labeling of the band inan autophosphorylating assay. Although the dnaK protein can be disassociated from thebacterial kinase by performing what is basically a kinase reaction [Yu-Sherman and Goldberg,19921, this procedure has been tested and found to be essentially ineffective [Dr. U. Dekkart,BRC]. Although this protein possessed autokinase activity, it did not phosphorylate the KDGST-Pim- 1.In summary, the KD GST-Pim-1 had no exogenous or auto kinase activity and althoughthe preparation was contaminated by a 70 kDa protein with autokinase activity, there were nokinases present that phosphorylated the KD mutant. Therefore, it can be concluded that theradio-labeling of the WT GST-Pim-1 was due solely to autophosphorylation. It is likely thatthe tyrosine phosphoryation of X. laevis GST-Pim-1 that was observed is the result of anautophosphorylation event. This suggests that Pim-l possesses tyrosineautophosphotransferase activity in addition to serine and threonine autophosphotransferaseactivity.122Figure 13. Comparison of WT and KD X. laevis GST-Pim- 1.Bacterially-expressed, purified and autophosphorylated GST-Pim-1WT (lanes 1, 3, 5, 7) and KD (lanes 2, 4, 6, 8) Western blotted withGXP (lanes 1 and 2) and 4G10 antibody (lanes 3 and 4). Amido blackstain of 2 ug of protein shown in lanes 5 and 6. Lanes 7 and 8 showautoradiograph of lanes 5 and 6. Full-length 63 kDa GST-Pim-1 fusionprotein indicated by solid arrow (4 ), 70 kDa DNAK protein indicatedby hollow arrow (<—). Migration of Mr standards are shown at right.1234 —12 34 5 67 84. STABILiTY OF GST-PIM-l ENZYMETime courses of autophosphorylation and substrate phosphorylation of both human andX. laevis GST-Pim-l were performed to determine how stable GST-Pim-1 was at 30°C.Preincubation at 30°C only slightly reduced the exogenous kinase activity of human and X.laevis GST-Pim- 1; the kinase activities were constant for the first 20-25 mm and began todrop off slowly at 30 mm (data not shown). The autophosphorylation activity of the twoPim-1 enzymes did not seem to vary much after preincubation at 30°C. The specific activitydeterminations demonstrated that the auto-phosphorylation activity of GST-Pim-1 at 30°Cwas linear for 25 mm, suggesting that preincubation at 30°C for 20 mm was unlikely toaffect activity.5. AUTOPHOSPHORYLATION ACTIViTY OF GST-PIM-lSpecific activity of autophosphorylation of the bacterially-expressed GST-Pim- 1preparations was quantitated. The specific activity of X. laevis GST-Pim-1autophosphorylation was 122 pmol.mg’min’, and the specific activity for theautophosphorylation of human GST-Pim-l was calculated to be 10 pmol.mg’min’ (Fig.14). This calculation takes into account the total concentration of the human GST-Pim-1preparation as it is unclear which Pim- 1 species in the preparation were catalytically-active.Reaction rates were linear for 20 mm, but started to drop by 25 mm for all concentrations ofsamples. As this rate reduction was observed even in the samples with the lowestconcentrations of protein, this reduction in reaction rate was unlikely due to saturation.Stoichiometry of autophosphorylation after 20 mm was determined to be 0.15 mmoles ATPincorporated per mole of X. laevis GST-Pim-1 fusion protein, and 0.013 mmoles ATPincorporated per mole of human GST-Pim- 1 fusion protein. These findings implied thatGST-Pim-l poorly autophosphorylated in vitro, or that it was substantiallyautophosphorylated in the E. coli prior to cell lysis.124Panel AaC.’E0EPanel B4003002001000 5 10 15 20Time of assay (mm)30002500200015001000500Time of assay (mm)Figure 14. Specific activity of autophosphorviation of H. sapiens and X. laevis GSTPim-1.The specific activity of autophosphorylation was determined for both H. sapiens and X.laevis GST-Pim-1. Panel A - specific activity of H. sapiens GST-Pim-1. Panel B -specific activity of X. laevis GST-Pim- 1. The pmol of 32P incorporated per mg of GSTPim-1 was calculated and plotted against the time of autophosphorylation assay. The slopeof the graph equals the specific activity and is shown on the graph.1256. GENERAL CHARACTERIZATION OF BACTERIALLY-EXPRESSED GST-PIM-1i. Time course of activityA time course study was performed to ensure that the exogenous activity of GST-Pim- 1towards a synthetic peptide substrate, AKRRRLSA (see later for criteria for selection of thispeptide), was linear for the time selected in most standard assays. These experimentsdemonstrated that the exogenous kinase assay of human GST-Pim- 1 was linear between 5and 10 mm, indicating that (a) the enzyme retained full activity during this time and (b) noneof the assay constituents were limiting. For this reason, assays with human GST-Pim- 1were performed for 5-10 mm. Preliminary results indicated that the X. laevis GST-Pim- 1reaction was linear for only about 3.5 mm, then decreased rapidly because of substratesaturation. To optimize the assays the amount of enzyme was reduced from .3.75 ug to 0.5ug per assay, and when the time course was repeated the reaction was linear with time for atleast 10 mm (data not shown).ii. Divalent cation requirement (Mg. Mn. Ca. Zn)The divalent metal cation requirement of GST-Pim- 1 for phosphotransferase activity wasexamined (Fig. iSa). The phosphotransferase activity of the human GST-Pim-1 towardsthe peptide substrate AKRRRLSA was optimally stimulated by 1.25 mM MnC12or 10 mMMgC12, so these metal ions were included at these concentrations in subsequentexperiments. These concentrations agree with the previously reported optimal values of 2mM MnC12 and 10 mM MgC12 for GST-Pim-l [Hoover et al., 1991], but differ slightlywith the corresponding value of 5 mM MnC12described by Friedmann et al. [1992]. Zincand calcium both inhibited GST-Pim- 1 kinase activity at concentrations higher than 0.6mM. All ions were tested to a final concentration of 250 mM, but as concentrations over100 mM had an effect comparable to 50 mlvi, values were only shown for up to 50 mM ofions.The phosphotransferase activity of the X. laevis GST-Pim- 1 towards the peptidesubstrate P4 was optimally stimulated by 1.5 mM MnCl2 or 25 mM MgC12, so these metalions were included at these concentrations in subsequent experiments (Fig. 1 5b). Theseconcentrations differ slightly from the values obtained for human GST-Pim-1, possiblybecause of enzyme quantities in the reaction. There was a slight stimulation of Pim- 1activity with 10 uM calcium, but at higher concentrations zinc and calcium both inhibitedGST-Pim-1 kinase activity. Sodium chloride was inhibitory at concentrations over 250mM.126AEIIIBI0C.)Cation concentration (mM)Figure 15. Cations required for optimal GST-Pim-1 phosphotransferase activity.Peptide assays using the P4 peptide as a substrate were performed in the presenceof varying cation concentrations. Cations tested: O Ca2, “Mg2+Na, - 0 Mn, - -- Zn. Panel A: results with X. laevis GST-Pim- 1,panel B: results with H. sapiens GST-Pim- 1.Cation concentration (mM)0 5 10 15 20 25 30 35 40 45 50127jjj. Km of ATPThe amount of ATP used per reaction was assessed using the equation described inAppendix V. The apparent Km of Pim- 1 for ATP was calculated using Lineweaver andBurke plots, as described in Appendix VI. The Km for ATP of human GST-Pim-1 wasdetermined to be 14 jiM ATP in the reaction. As enzyme reactions should optimally containconcentrations of substrate at least three-fold higher than their Km values, the standardhuman Pim- 1 assays normally contained 50 jiM of ATP. The Km of X. laevis GST-Pim- 1for ATP was determined to be 154 jiM, so a final concentration of 450 jiM was used for allX. laevis GST-Pim-1 assays.iv. Protein kinase inhibitorsThe sensitivity of Pim- 1 to a peptide inhibitor (PKI) of cAMP-dependent protein kinasewas tested, because of the high amount of sequence homology between Pim-1 and secondmessenger-dependent protein kinases [Cheng et at., 1985]. The activity of human and frogGST-Pim- 1 was independent of the amount of PKI in the reaction, indicating that PKJcould be added to assays to eliminate the possibility that the phoshotransferase activityobserved was due to cAMP-dependent protein kinase. This was especially important, inretrospect, when the Pim- 1 substrate peptides were found to fulfill the consensus sequencerequirements of cAMP-dependent protein kinases.1287. SUMMARY OF GST-PIM-l EXPRESSIONAn amphibian pim-] homologue was cloned from a X. laevis cDNA library. The kinasecontained all the conserved residues in catalytic subdomains common to serine/threonineprotein kinases, and featured a consensus pattern for a disulfide bond-forming cysteine, andconsensus phosphorylation site motifs for cAMP-dependent kinases, PKC, CKII, and atyrosine kinase. Nucleotide and amino acid searches revealed that Pim-1 had homology to thecatalytic domain of serine/threonine kinases, especiallyCa2/calmodulin-dependent proteinkinases and SNF-1 homologues. The Pim-l homology tree indicated that the X. laeviskinase was most homologous to the Pim- 1 from mammalian species.The human and X. laevis Pim- 1 were subcloned into the pGEX-2T bacterial expressionvectors and expressed as GST-Pim-l fusion proteins. Full-length proteins were produced inboth cases, and several smaller human Pim- 1 byproducts were produced, possibly a result ofalternate or aborted translation. The fusion protein preparations were active and had a lowspecific activity. They displayed autophosphotransferase activity primarily towards serine,threonine and tyrosine residues. The enzymes were stable at 30°C for about 20 mm, and Pim1 kinase assay conditions were optimized with respect to the nucleotide and ion requirements.129CHAPTER V.SUBSTRATE STUDIES: IDENTIFYING IN VITRO TARGETS OF PIM-1Although the exact function of Pim- 1 in the cell has not yet been defined, its nature as acytoplasmic protein kinase is consistent with a role within a signal transduction pathway.Upstream activators (GM-CSF, Epo, IL-3) [Wingette et at., 1991; Lilly et at., 1992; Sato etat., 19931 and proteins existing in similar signal transduction pathways (i.e. JAK, p95VaV)[Matsuguchi et at., 1995; Miura et at., 1994; Queue et at., 1994] are being intensivelyinvestigated by other groups, but little has been done to identify direct targets of Pim- 1. Togain insight into the function of the kinase, various proteins and peptides were tested as invitro substrates of Pim- 1. It was intended that the definition of substrate requirements ofPim- 1 would lead to the identification of physiological substrates of this kinase.1. PRELIMINARY SUBSTRATE STUDIESi. In vitro phosphorviation of protein substrates by GST-Pim- 1A large number of proteins were tested and found to be in vitro substrates of both H.sapiens and X. laevis GST-Pim-1. X. taevis GST-Pim-1 strongly phosphorylated MBP(21 kDa), and all the histones, including histone hA (14-17 kDa), histone hIS (14-17 kDa),histone Ill-S (31 kDa) and histone Vu-S (14-17 kDa) (Fig. 16). X. taevis GST-Pim-1moderately phosphorylated the 40 S ribosomal protein (31 kDa), casein (40 kDa), GSTRaf-1 (95 kDa), enolase (40,42 kDa) and phosphorylase b (94 kDa). This GST-Pim-1 didnot phosphorylate GST, phosvitin, protamine sulfate or protamine chloride.The protein spectrum phosphorylated by human GST-Pim- 1 was similar to thatphosphorylated by X. taevis GST-Pim-1 (Fig. 17). Human GST-Pim-1 stronglyphosphorylated MBP (21 kDa) and all the histones, including histone hA (14-17 kDa),histone ITS (14-17 kDa), histone Ill-S (31 kDa) and histone Vu-S (14-17 kDa). HumanGST-Pim-1 moderately phosphorylated the 40 S ribosomal protein (31 kDa), GST-Raf-1(120 kDa) and phosphorylase B (94 kDa). Human GST-Pim-1 did not phosphorylateGST, phosvitin, protamine sulfate and protamine chloride and it was difficult to determineif it phosphorylated casein (40 kDa) or enolase (40, 42 kDa) as these proteins comigratedwith autophosphorylated byproducts in the GST-Pim-1 preparation.130140kDa86.8GST-Pim-1Figure16.Phosphorviationof variousproteinsubstrates byX.laevisGST-Pim- 1.Phosphorylationreactionswerecarriedoutinthepresence(+)orabsence(-)ofGST-Pim- 1.GST-Pim-1controlonleft.Reactionscontained5ugeachofprotaminesulphate(A),protaminechloride(B),casein(C),phosvitin(D),phosphorylaseB(E),enolase(F), GST-Raf-1(G),GST(H),histoneHA(I), histone11S(J), histoneVHS(K),histoneifiS(L),40Sribosome(M)andmyelinbasicprotein(N).MigrationsofMrstandardsareshownontheleft.TheautophosphorylatedGST-Pim-1isindicatedbyarrow(—).GST—Pim—1+1+11+1+—11+11+—11+—1+—1+—11+—11+—I1+—ABCDEFGHIJKLMN140kDa—tJGST-Pim-1-*47.8—Figure17.PhosphorviationofvariousproteinsubstratesbyH.sapiensGST-Pim-1.Phosphorylationreactionswerecarriedoutinthepresence(+)orabsence(-)ofGST-Pim-1.GST-Pim-1controlonleft.Reactionscontained5ugeachofprotaminesulphate(A),protaminechloride(B),casein(C),phosvitin(D),phosphorylaseB(E),enolase(F),GST-Raf- 1(G),GST(H),histonehA(I),histone11S(J),histoneVhS(K),histoneifiS(L),40Sribosome(M)andmyelinbasicprotein(N).MigrationsofMrstandardsareshownontheleft.TheautophosphorylatedGST-Pim-1fragmentsareindicatedbyarrows(-*).33.3—28.6—GST—Pim—1+1+—11+—11+—11+—11+—11+—1+—1+1+—•+—11+—,+—ABCDEFGHIJKLMNii. Phosphoamino acid analysis of phosphorviated substratesPhosphoamino acid analysis of GST-Pim-i phosphorylated substrates was performedto establish the nature of phosphorylation catalyzed by GST-Pim- 1. Radiolabelled proteinsfrom Figures 16 and 17 were excised and subjected to acid hydrolysis. Both H. sapiensand X. laevis GST-Pim- 1 phosphorylated proteins on serine and threonine, but not tyrosineresidues. Both human and X. laevis GST-Pim-i phosphorylated the 40 S ribosomalprotein, histone hA, histone hIS and histone ITT-S on both serine and threonine residues,while MBP and histone VhS were phosphorylated on serine only. Because of lowstoichiometries of phosphorylation, the phosphoamino acid analysis was not performed onenolase, casein, phosphorylase b and GST-Raf1.In agreement with our results, previously published data indicated that Pim-1phosphorylated histone Hi on both serine and threonine residues, enolase on serine andthreonine residues [Hoover et al., 1991; Friedmann et al., 1992], histone 2B on serine[Saris et al., 1991] and did not phosphorylate GST [Hoover et al., 1991]. In contrast toour results, GST-Pim-i phosphorylated salmon protamine [Saris et al., 1991] but did notphosphorylate casein [Hoover et al., 19911. The reason for these conflicts is most likelydue to the fact that phosphorylation of these substrates by this kinase was very low anddifferent preparations of these substrates are already phosphorylated to variable extents.The fact that the GST-Pim- 1 phosphorylated so many of these proteins in vitro is ofminimal physiological significance. This in vitro situation involves very large amounts oftwo highly purified proteins: an active kinase and a potential substrate. When presentedwith such high concentrations of a potential substrate without interference by regulatoryproteins or competing substrates, the GST-Pim-1 is likely to seem more promiscuous in itssubstrate preference. Therefore, the in vitro protein phosphorylation data should be usedonly as a guideline of substrate preference and not an ultimate determination ofphysiological activity.iii. Peptide substrate comparisons to published dataIn an earlier study by Hoover et al. [1991], histone Hi and Kemptide (LRRRASLG, apeptide modeled after the cAMP-dependent protein kinase phosphorylation site in pyruvatekinase) were used as substrates for GST-Pim- 1. These substrates were used in the presentstudy to test the exogenous phosphotransferase activity of bacterially-expressed human andX. laevis GST-Pim-1. Both substrates were found to be phosphorylated by GST-Pim-i;histone Hi was phosphorylated by the X. laevis GST-Pim-i with a Vmax of 825 pmol.miw1331.mg,with an apparent Km of 2 .tM, while Kemptide was phosphorylated with a Vm ofless than 7.5 pmol.min’.mg’, with an apparent Km of greater than 1200 pM (Table 3). Apeptide based on the C-terminus of ribosomal S6 protein routinely used in the lab, wasphosphorylated by X. laevis Pim-1 with a Vmax of 178 pmol.min’.mg’ with an apparentKm of 70 p.M. Human GST-Pim- 1 phosphorylated Kemptide with a Vmax of 0.6 pmol.miir1.mg, and an apparent Km of greater than 800 p.M. This same preparation of humanGST-Pim-1 phosphorylated the S6-CT peptide with a Vmax of ii pmol.miir1. gwith anapparent Km of 16 p.M.Friedmann et al. [19921, found that Pim-l phosphorylated histone Hi in vitro with anapparent Km of 7 jiM. They found that the histone Hi was a 6-fold better substrate thanKemptide and suggested that histone might actually be a physiological substrate. HistoneHi was used early in our study to initially test and optimize the activity of the expressedfusion proteins and to analyze Pim-1 activity during early colunm purifications. However,as histone Hi is a substrate for many other kinases, we decided to explore the activity ofPim-l using a more specific and convenient substrate. As a panel of S6 peptide analogswas available in the lab, substrate studies were initiated using peptides modeled on the S6.134TABLE3.COMPARISONOFSEVERALPEPTIDESPHOSPHORYLATEDBYGST-PIM-1.Vmaxisexpressedaspmol.min1.mg.Kmisexpressedasi.LM.HUMANXENOPUSVmaxKmVmaxVmaxKmVmaxKmKmHistoneHinotperformed8252412KemptideLRRASLG0.6>800<0.0017.5>1200<0.0061S6-CTAKRRRLSSLRASTSKSESSQK11.0160.69178702.52. SUBSTRATE ANALYSIS USING PEPTIDE ANALOGSTo delineate the consensus phosphorylation site recognition sequence for GST-Pim- 1,we intended to test the affinity of GST-Pim- 1 for a series of peptide analogs. Since the S6protein in the 40 S ribosome served as a substrate for both human and X. laevis GST-Pim1, a peptide that encompassed the major phosphorylation sites located at the C-terminus ofthe human S6 protein (residues 229 to 249) [Heinze et al., 1988] was tested as a substrate(Table 3). This peptide (S6-CT, AKRRRLSSLRASTSKSESSQK) was at least a 20-foldbetter substrate for GST-Pim-l than Kemptide (LRRASLG). A panel of shorter peptideanalogs of S6-CT was constructed to identify the critical residues surrounding thephosphoacceptor site that were needed for substrate recognition and phosphorylation byGST-Pim- 1.Peptides were synthesized in the laboratory of Dr. Ian Clark-Lewis (BiomedicalResearch Centre). Peptides were initially purified by HPLC, but because of the small sizeand relative purity of the peptides, this step was eliminated. Comparison of the GST-Pim-1phosphorylation of the HPLC-purified with the non-purified peptides (P1, P2, P3),demonstrated that the activity towards the HPLC treated peptides averaged 15% higher thatwith the non-purified peptides (data not shown). Since the peptide selectivity wasmaintained despite the differences in the rate of the reaction, non-HPLC treated peptideswere used for the subsequent kinetic determinations. Determinations were conducted intriplicate and each experiment repeated at least three times with highly reproducible results.Values were plotted and kinetic constants determined using Michaelis-Menten andLineweaver and Burke plots as detailed in Appendix VI. There was some difficulty in theestimation of the Vmax values, because substrate inhibition occurred with higherconcentrations of certain peptides. The mean values of the apparent Km and Vmaxdeterminations as well as the Vmax/Km value for all peptides are shown in Tables 4-9. TheVmax/Km value was used to compare the relative efficiencies of various peptides assubstrates of the GST-Pim- 1.To ensure that the phosphorylation was due to the phosphotransferase activity of Pim- 1and not to a contaminating/copurifying kinase, the KD mutant was incubated with the samepeptides in control reactions. No phosphotransferase activity was detected towards anypeptide which confirmed that activity was specifically due to Pim-1 (data not shown).136i. Location of phosphoacceptor siteTo determine which serine was phosphorylated by the kinase, the rates ofphosphorylation of S6-CT (AKRRRLSSLRASTSKSESSQK), P1 (AKRRRLSSLRA), P2(AKRRRLSALRA), and P3 (AKRRRLASLRA) by GST-Pim-1 were compared (Table 4).To identify which of the two serine residues in Pt was targeted for phosphorylation, theserine residues were selectively replaced with alanine residues in P2 and P3. Alanineresidues were selected as the closest non-hydroxylated analogs for serine. The P1 and P2peptides were similarly effective as substrates and were phosphorylated to a much greaterextent than P3. These results demonstrated that GST-Pim-1 exhibited strong preference forthe first serine residue in P1. Consequently, all other peptides were constructed with onlythe first serine residue; for reference, the position of the phosphorylatable residue wasdesignated as “0”.ii. Influence of C-terminal residuesTo examine the influence of neighboring C-terminal residues on serine phosphorylationby GST-Pim-1, peptides P4 (ALRRRLSA), P5 (ALRRRLS-amide) and P6 (ALRRRLSacid) were tested as substrates (Table 5). The P4 peptide was more stronglyphosphorylated than P2, implying that the C-terminal residues, “ALRA” of P2 exerted aninhibitory effect on phosphorylation. P5 was phosphorylated more efficiently than P4,confirming the inhibitory effect of residues C-terminal to 0. However, it remains possiblethat different C-terminal residues might contribute to improved substrate recognition in aphysiological substrate of Pim- 1.To evaluate the importance of a peptide bond on the C-terminal side of thephosphoacceptor site, P6 was constructed with a C-terminal free acid instead of an amide.P6 was poorly phosphorylated by GST-Pim- 1, suggesting that the negative charge at the Cterminus acted as a negative determinant and physiological substrates of Pim- 1 are unlikelyto be phosphorylated by this kinase at the C-terminal residue. As P4 was morerepresentative of a natural substrate than PS or P6 in having a peptide bond after thephosphoacceptor residue, P4 was used as the prototype for the design of additional analogpeptides for the remainder of this study and a standard for comparison with other peptides.137TABLE4.IDENTIFICATIONOFSERINERESIDUESPHOSPHORYLAThDBYX.LAEVISANDH.SAPIENSGST-PIM-1.VmaxisexpressedaspmoLmin1.mgKmisexpressedasriM.HUMANXENOPUSVmaxKmVmaxVmaxKmVmaxKmKmKemptideLRRASLG0.6>800<0.0017.5>1200<0.006S6-CTAKRRRLSSLRASTSKSESSQK11.0160.69178702.5P1AKRRRLSSLRA9.31 30.7295156.3P2AKRRRLSALRA9.3130.72101185.6P3AKRRRLASLRA3.070.4338132.9GoTABLE5.INFLUENCEOFC-TERMINALRESIDUESONSUBSTRATEPHOSPHORYLATIONBYGST-P1M-1.Vmaxisexpressedaspmo1.min.mg1KmisexpressedastiM.HUMANXENOPUSVmaxKmVmaxVmaxKmVmaxKmKmP2AKRRRLSALRA9.3130.72101185.6P4AKRRRLSA15.3151.02124353.5P5AKRRRLS-amide15.362.5120264.6P6AKRRRLS-freeacidnotmeasurable92180.04iii. Amino acid specificity of the phosphoacceptor siteThe specificity of the phosphoacceptor site for recognition by GST-Pim- 1 wasexplored with P4 (AKRRRLS.A), P7 (AKRRRLIA) and P8 (AKRRRLYA) (Table 6).Peptides P4 and P7, which featured serine and threonine respectively, at thephosphoacceptor site (position 0) were recognized comparably as substrates.Phosphorylation of the peptide P12, which contained a tyrosine residue at position 0 wasnot detected.iv. Influence of the -1 amino acid residueTo examine the influence of the residue directly before the phosphoacceptor site,phosphorylations of peptides P4, P9, PlO, P11, P12, P13, and P14 by GST-Pim-l werecompared (Table 7). There were marked differences in the selectivity of human and X.laevis GST-Pim- 1 within this set. PlO (AKRRRRSA) with a basic arginine residue at the-1 position was poorly phosphorylated peptide by both human and X. laevis GST-Pim- 1.Although P8 (AKRRRKSA) with a basic lysine residue at the -1 position was not welltolerated by the human enzyme, P18 was phosphorylated to a comparable if not greaterextent than the P4 standard by X. laevis GST-Pim- 1. An acidic residue at the -1 site, as inP11 (AKRRRSA), yielded only low phosphorylation by both human and X. laevis GSTPim-1. Other polar residues such as glutamine in P12 (AKRRRQSA) at the -1 location alsoproduced only low phosphorylation by human GST-Pim-l, but were phosphorylated to thesame extent as the P4 standard by X. laevis GST-Pim- 1. Non-polar residues such asalanine in P13 (AKRRRASA) and isoleucine in P14 (AKRRRISA) at the -1 position weretolerated by both enzymes, yielding intermediate results. Thus, most amino acids, exceptfor acidic residues, were acceptable at the -1 site. Basic residues in this location were notaccommodated by the human GST-Pim- 1, while the X. laevis GST-Pim- 1 tolerated a lysinebut not an arginine residue at this site, an effect that may be sterically related. Although anynon-polar residue was tolerated by the X. laevis GST-Pim- 1, the optimal amino acid for thehuman GST-Pim-l for the -l position was found to be leucine, as in P4 (AKRRRLSA).For this reason, leucine was used for the -1 site for the construction of all further peptideanalogs. The fact that the human and the amphibian GST-Pim-1 results did not agree for allthe peptides tested indicate that either the sites examined were not particularly critical forsubstrate recognition by Pim- 1 or that minor differences in the primary structure of theproteins contribute to substrate recognition. To our knowledge, this is the first example ofa change in substrate specificity for a protien kinase from different species.140TABLE6.AMINOAC1I)SPECIFICITYOFTHEPHOSPHOACCEPTORSITE.Vmaxisexpressedaspmol.min1.mgKmisexpressedasIIM.HUMANXENOPUSVmaxKmVmaxVmaxKmVmaxKmKmP4AKRRRLSA15.3151.02124353.5P7AKRRRLTA18.9190.9980352.3P8AKRRRLYAnotmeasurablenotmeasurableTABLE7.INFLUENCEOFTHE-1AMINOACIDRESIDUE.VmaxisexpressedaspmoLmin1.mg1-.KmisexpressedasjiM.HUMANXENOPUSVmaxKmVmaxVmaxKmVmaxKmKmP4AKRRRLSA15.3151.02124353.5P9AKRRRKSA2.9230.13146265.6PlOAKRRRRSA0.63400.0023593.9P11AKRRRESA2.2260.0877352.2P12AKRRRQSA4.2330.1298313.2P13AKRRRASA6.3240.2686165.4P14AKRRRISA5.7230.2599313.2v. Influence of -2,-3 and -4 amino acid residuesThe importance of the arginine residues at the -2, -3, and -4 positions for GST-Pim-1recognition of a substrate was evaluated with peptides P4, P15, P16, P17, P18, P19,P20 and P21 (Table 8). Peptide P15 (AKRRALSA) demonstrated that substitution of analanine for the arginine at the -2 position dramatically reduced the phosphorylation of thepeptide by GST-Pim-1. A lysine residue at the -2 position in P16 (AKRRKLSA)restored phosphorylation by human GST-Pim-1 and partially restored phosphorylation byX. laevis GST-Pim- 1 and indicated a strong preference for a basic residue at this location.When an alanine residue was substituted for arginine at the -3 position in P17(AKRARLSA), phosphorylation by GST-Pim- 1 was markedly reduced. Conservativesubstitution of a lysine for an arginine at the -3 site in peptide P18 (AKRJRLSA) did notrestore phosphorylation to the original levels, which emphasizes the absolute requirementfor an arginine residue at this location. Substitution of an alanine for the arginine at the -4position in peptide P19 (AKARRLSA) reduced the affinity of the peptide slightly, but itdid not affect the rate of phosphorylation by human GST-Pim- 1. However, the alaninesubstitution at the -4 position in peptide 19 reduced the rate of phosphorylation by X.laevis GST-Pim- 1. Peptide P20 (AKKRRLSA) demonstrated that a substitution of lysinefor the arginine at the -4 position did not affect phosphorylation, and indicated that that thesubstrate requirement is fulfilled by any basic residue at the -4 position. A doublesubstitution of alanine at the -2 and -4 positions, further decreased the K and Vmax ofphosphorylation of peptide P21 (AKARALSA), as compared to -2 and -4 singlesubstituted peptides, P15 and P19, respectively, and emphasized the need for a strongbasic environment upstream of the Pim-1 phosphorylation site.This data indicated that all three arginine residues are optimal for recognition by GSTPim- 1, and that other basic residues like lysine, cannot completely replace the arginines.The most important arginine is located at the -3 position, followed by the -2 and the -4arginines. The results from both the H. sapiens and the X. laevis Pim- 1 were consistantin this regard.vi. Influence of the -5 and -6 amino acid residuesThe influence of the residues at the -5 and -6 positions for GST-Pim-1phosphorylation was tested with peptides P4, P22, P23, P24, P25, P26 and P27 (Table9). The results with P22 (KRRRLSA) demonstrated that the presence of a specificresidue at the -6 position was not essential for phosphorylation by GST-Pim-1. Thereduction in phosphorylation by human GST-Pim-1 that was observed with P22 could143TABLE8.IMPORTANCEOF-2.-3AND-4ARGININERESIDUES.Vmaxisexpressedaspmo1.min1.mg.Kmisexpressedasp.M.HUMANXENOPUSVmaxKmVmaxVmaxKmVmaxKmKmP4AKRRRLSA15.3151.02124353.5P15AKRRALSA11.41110.121201171.0P16AKRRKLSA16.5200.82146881.7P17AKRARLSA1.31100.014290.1P18AKRKRLSA3.7890.0432291.1P19AKARRLSA19.8520.3890701.3P20AKKRRLSA17.5151.17140642.2P21AKARALSA1.92500.008141630.09TABLE9.INFLUENCEOFTHE-5AND-6AMINOACIDRESIDUES.Vmaxisexpressedaspmol.min1.mgKmisexpressedaspM.HUMANXENOPUSVmaxKmVmaxVmaxKmVmaxKmKmP4AKRRRLSA15.3151.02124353.5P22KRRRLSA3.280.4095185.3P23RRRLSA0.64440.00111180.6P24AARRRLSA8.2300.2748232.1P25ARRRRLSA9.3110.8594551.7P26ALRRRLSA6.7300.2237231.6P27AERRRLSA3.2600.05201240.2(I’possibly result from the removal of the peptide bond. In contrast, removal of this alaninecaused an increase in phosphorylation by the X. laevis GST-Pim- 1. The influence ofthis site on substrate phosphorylation was probably minimal, accounting for the conflictin results between the two enzymes. The influence of this residue was not examinedfurther.Peptide P23 (RRRLSA) demonstrated that a residue at the -5 position was essentialfor recognition by both H. sapiens and X. laevis GST-Pim- 1. Replacement of the lysineat the -5 location with hydrophobic residues in P24 (AA.RRRLSA) and P26(ALRRRLSA) reduced phosphorylation of the peptide. The restoration ofphosphorylation that occurred when a basic arginine was inserted at the -5 position as inP25 (ARRRRLSA) showed that human GST-Pim- 1 exhibited strong preference for abasic residue at this site, while X. laevis GST-Pim- 1 was ambivalent. Substitution of anacidic glutamic acid residue at this position as in P29 (ARRRLSA) caused very lowphosphorylation, confirming the preference for a basic residue at this site.vii. Determining the substrate consensus sequenceThese results demonstrate that H. sapiens and X. laevis GST-Pim-1 have very similarsubstrate specificities. Residues at the +1, 0, -2, -3, -4 and -5 sites had similar selectivitywith both the human and the X. laevis GST-Pim-1 enzymes, indicating that these residuesare important determinants in the substrate consensus sequence. Residues at the -1 and -6sites were not in agreement between the two enzymes, implying that these two sites are ofminimal importance for substrate recognition by GST-Pim-1. Although the Km and theVmax differ, a similar selectivity was maintained for most peptides.The deduced consensus sequence for substrate recognition by GST-Pim- 1 based on thesubstrate peptide studies was [K/RI - [K/RI - R - [K/RI - L - [SIT] - X, where X isoptimally a residue with a shorter side chain. It is possible that additional residues locatedN- and C-terminal to this sequence also influence substrate recognition. This sequence isnot present in Pim-l from human, mouse [Breuer et at., 1989] or rat [Wingette et at.,1992]. However, a sequence similar to this, F-R-R-R-I-S-T (amino acid residues 256-263)is located in the primary structure of the X. taevis Pim- 1 near catalytic subdomain XI.Perhaps this might be an auto- or cross-phosphorylation site in X. taevis Pim- 1, withphosphorylation at this site leading to altered phosphotransferase activity towardsexogenous substrates.1463. DATA BANK SEARCH FOR POTENTIAL SUBSTRATES OF PIM-1Using deduced consensus phosphorylation site motifs to scan protein sequence databases is potentially a useful approach to identify putative kinase substrates. However, insome cases, secondary or tertiary structure may deny access to a potential phosphoacceptorsite through steric hindrance [Kemp and Pearson, 1990; Kennelly and Krebs, 1991]. Acomputer search with the consensus sequence, [KIR]-[KIR]-R-[KIR]-L- [SIT] identifiedmany proteins that featured this sequence. In addition to S6, other potential substrates forPim-l include: human heterogeneous ribonucleoprotein [KKRRLS]; human cAMPdependent protein kinase type 1-13 regulatory chain [RRRRLS]; Schizosaccharomycescerevisiae DNA polymerase delta large chain [KRRRLS]; Saccharomyces cerevisiaeCTP:cholinephosphate cytidylyltransferase [KRRRLT]; human epidermal growth factorprecursor [KRRRLT]; human Ski oncoprotein [RKRKLT]; the mouse CDC-25 proteinhomologue [RRRKLS]; and the rat guanine-nucleotide releasing protein [RRRKLSJ.1474. INHiBITION OF PIM-1 ACTIVITY BY INHII3ITOR PEPTIDESi. Inhibition of phosphotransfer activity by pseudo-substrate peptidesPeptides P28 (AKRRRLAA) and P29 (AKRRRLA) were constructed as analogs ofP4 that lacked phosphorylatable amino acid residues. Neither peptide was detectablyphosphorylated by GST-Pim- 1, which confirmed that all other peptides were radiolabelledwith [y32P]-ATP from direct phosphorylation and not due to of radioactive ATP beingtrapped onto the phosphocellulose filters indirectly via the basic residues of the peptide (datanot shown).When P28 and P29 were tested as inhibitors of human GST-Pim-1 phosphorylation ofP4 substrate, P28 was found to inhibit the phosphorylation in a competitive fashion (Fig.18); increasing amounts of P4 competed with P28 for the substrate binding domain of thekinase. This is not surprising as the sequences of the peptides differed only between theserine and alanine residue and these two amino acids are both very small. Peptide P29inhibited phosphorylation of P4 more potently, acting competitively at low concentrationsand non-competitively at higher concentrations (Fig. 19), indicating that a more complexinteraction existed between P29 and GST-Pim-l. The inhibition of GST-Pim-lphosphorylation by P29 was repeated in the presence of 100 mM 2-mercaptoethanol withsimilar results. This reducing agent was added to inhibit dimerization of the cysteinecontaining P29 peptide and disulfide bond formation with GST-Pim- 1. The GST-Pim- 1did not rely on intact disulfide bonds to be functional and a disulfide bond did not appear toform between GST-Pim-1 and peptide P29.Pseudosubstrate sequences are located in regulatory domains of certain protein kinasesand regulate phosphotransferase activity [Kennelly and Krebs, 19911. They correspond tosequences resembling protein kinase phosphorylation site motifs, except that they usuallycontain an alanine residue substitution for a serine or threonine residue [Kennelly andKrebs, 1991]. Peptide P28, which contained an alanine residue in place of thephosphorylatable serine and P29, which contained a cysteine instead of the serine, weremodeled after the consensus phosphorylation site sequence of Pim- 1 and tested as inhibitorsof Pim- 1 phosphotransferase activity. The results with these peptides indicated that theactivity of Pim- 1 is not likely to be regulated by the presence of a pseudosubstrate sitewithin the kinase.148Figure 18. Competitive inhibition of exogenous phosphotransferase activity ofH. sapiens GST-Pim- 1 by P28 peptide.The concentrations of P28 peptide (AKRRRLAA) indicated in the legend were usedto inhibit the phosphorylation of P4 substrate peptide (AKRRRLSA) by H. sapiensGST-Pim- 1. The inverse incorporation of 32p is indicated on the vertical axis, theinverse concentration of P4 peptide is indicated on the horizontal axis. Origin isindicated by vertical dotted line.1 .000e-39.000e-40C.)EC.)8.000e-47.000e-46.000e-4’5. 000e-44.000e-43.000e-42.000e-41 .000e-4O.000e+0• ControlD 100uMP28A 200uMP28500 uM P28•-0.05 0.00 0.05 0.10 0.15 0.20[uM P4 peptide) -1149Figure 19. Inhibition of exogenous phosphotransferase activity of H. sapiensGST-Pim- 1 by P29 peptide.Concentrations of P29 peptide (AKRRRLCA) were used to inhibit the phosphorylationof P4 substrate peptide (AKRRRLSA) by H. sapiens GST-Pim- 1. The inverseincorporation of 32p is indicated on the vertical axis, the inverse concentration ofP4 peptide is indicated on the horizontal axis. Panel A shows a close-up of panel B.The origin is indicated by the vertical dotted line.A 4.000e-33.500e-33.000e-3- I‘ 2.500e-3 I• Control2.000e-3 I El 100uMP29I El A 200uMP29j 1.500e-3 I 200uMlP29 reduced1.000e-3 I5.000e-40.000e+0 .. I • • • •-0.02 -0.01 -0.00 0.01 0.02 0.03 0.04 0.05[uM P4 Peptide] -1B 0.10-0.08 I- I Q0.06-I • ControlI C C 100uMP29I A 200uMP290.04- I 200uM1P29-reduced0.02- A0.00--0.10 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30[uMol P4 Peptide] -1150ii. Inhibition of autophosphorylation activity by substrate and pseudo-substratepeptidesPeptides P4, P28 and P29 were tested as a potential substrate and pseudosubstrates ofhuman GST-Pim- 1 that might specifically block the active site of the kinase and stericallyinhibit autophosphorylation. Peptides Pim 1-Ill and Piml-NT, used to immunize rabbits forPim- 1 antibodies, were used as controls with peptide concentrations ranging from 0.125 -4.0 mM per assay.The autophosphorylation of GST-Pim- 1 was inhibited strongly by substrate P4 atconcentrations as low as 0.125 mM (data not shown). Peptide P29 slightly inhibited theautophosphorylation at concentrations above 0.5 mM and moderately inhibitedautophosphorylation at concentrations greater than 1.0 mM, while P28 caused only veryminor inhibition of autophosphorylation at concentrations that exceeded 0.75 mM. Theseresults indicated that GST-Pim- 1 preferentially phosphorylated exogenous substrate ratherthan itself. Peptides used to inject rabbits for antibody production (Pimi-ifi and Piml-NT)did not inhibit the autophosphorylation activity whatsoever, indicating that the inhibition ofautophosphorylation was selective and did not result from high amounts of non-specificpeptide in the reaction.1515. CONCLUSIONS OF SUBSTRATE STUDIESThe substrate consensus recognition sequence for Pim- 1 was deduced to be [K/RI -[K/RI - R- [K/RI - L -[SIT] - X, where X cannot be a residue with a large side chain. Therewere slight differences in peptide substrate selectivity between H. sapiens and X. laevisGST-Pim- 1; the H. sapiens enzyme displayed a clear preference for a leucine residue at the-1 site, while the X. laevis enzyme tolerated most residues tested at this site. Friedmann etat. [1992] published a similar consensus phosphorylation site motif for Pim-1 while thiswork was in progress, i.e. (R/K)3-X-S T-X’, where K cannot be arginine, lysine or alarge hydrophobic residue. They defined this sequence using a series of only 6-8 peptides,some of which did not resemble this consensus sequence, and the influence of each of thesites near the phosphorylatable residue was not tested.Potential physiological substrates containing this motif are not yet obvious. Many of theproteins that featured the Pim- 1 phosphorylation site consensus sequence are enzymes orstructural proteins, many of which are functional in the cytoplasm. This is in agreementwith several recent publications implicating Pim- 1 upregulation as a cytoplasmic event inresponse to growth factor receptor stimulation.152CHAPTER VI.ANALYSIS OF AUTOPHOSPHORYLATIONAs demonstrated in Chapter IV, expressed GST-Pim- 1 was able to autophosphorylate invitro on serine, threonine and tyrosine residues. Whether Pim- 1 autophosphorylates ontyrosine and/or serine/threonine has been a matter of controversy in the literature [Meeker etal., 1987a; Telerman et at., 1988; Saris et at., 1991; Padma and Nagarajan, 1991;Friedmann et at., 1992]. Although the nature of this autophosphorylation has beenextensively investigated, the sites of autophosphorylation have not yet been identified. Thegoal of this study was to identify the major autophosphorylation site of X. taevis GSTPim-1, to change this site by PCR mutagenesis, and to assess the functional consequence ofthis perturbation in the expressed mutant in comparison to WT GST-Pim- 1. The X. laevisGST-Pim-1 was used for all experiments as the full-length fusion protein stronglyautophosphorylated and the results were easier to assess because of the lack of degradationproducts as associated with human GST-Pim-1.1. DEThCTION OF GST-PIM- 1 WITH ANTI-PHOSPHOTYROSINE ANTIBODIESIn Chapter IV we demonstrated immunodetection of bacterially-expressed X. laevisGST-Pim-1 protein by anti-phosphotyrosine antibodies (Fig. 12). We wanted to determineif this imrnunodetection was specific for phosphotyrosine or if it was due to cross reactivitywith phosphoserine or/and phosphothreonine residues. We also wanted to assess if in vitroautophosphorylation would cause an increase in the amount of detectable phosphotyrosineassociated with Pim- 1, and we wanted to confirm that the tyrosine phosphorylation of Pim1 was due to auto-kinase activity and not due to contamination by bacterial kinases in theGST-Pim- 1 preparation.Western blots of KD, WT Pim-1 and WT Pim-1 autophosphorylated with [y32P]-ATPwere probed with a X. laevis-specific antibody, Piml-XI, and an anti-phosphotyrosineantibody, PY2O (Fig. 20, panel A). Probing with Piml-XI antibody demonstrated that theamount of protein loaded in each lane was similar (lanes 1-3). Although WT GST-Pim-1(lanes 2) did not experience a size shift with in vitro autophosphorylation (lane 3), bothunphosphorylated and in vitro autophosphorylated WT GST-Pim-1 samples appeared to beslightly retarded on a Western blot, compared to the KD- GST-Pim- 1. The difference in153size between the KD and WT GST-Pim-1 may be due to the incorporation of phosphateresidues, as a result of in vivo autophosphorylation of the WT GST-Pim- 1.The anti-phosphotyrosine antibody detected OST-Pim- 1 both before and after in vitroautophosphorylation and indicated that the GST-Pim- 1 was already substantiallyphosphorylated on tyrosine prior to lysis of the bacteria. The KD mutant was not detectedby the anti-phosphotyrosine antibody, indicating that the tyrosine phosphorylation was notthe result of a contaminating kinase (lanes 4-6). To ensure that this effect was not due tocross-reactivity with phosphoserine and phosphothreonine residues, free phosphoaminoacids were used to compete with the PY2O anti-phosphotyrosine antibody. Phosphoserine(lanes 7-9) and phosphothreonine (lanes 10-12) did not compete with the antibody, butphosphotyrosine competed and eliminated detection with the PY2O antibody (lanes 13-15).This experiment was repeated using a second anti-phosphotyrosine antibody (4G10) withsimilar results (data not shown).From this experiment we concluded that the X. laevis GST-Pim-l canautophosphorylate on tyrosine residues in addition to serine and threonine. While sometyrosine phosphorylation occurred during an in vitro autophosphorylation reaction, it seemsthat the majority of tyrosine phosphorylation happened during expression in the bacteria.Experiments were performed to assess if the kinase activity of GST-Pim- 1 could bechanged by autophosphorylation or by dephosphorylation. The results of these experimentswere inconclusive and are described with the results in Appendix VIII.154Piml-XI PY2OFigure 20. Tyrosine phosphorviation of expressed X. laevis GST-Pim- 1.Panel A - Western blots of expressed X. laevis GST-Pim- 1 probed with antiphosphotyrosine antibodies. Lanes 1, 4, 7, 10, 13 each contain 0.75 ug ofK69-A kinase-inactive GST-Pim- 1. Lanes 2, 5, 8, 11, 14 each contain 0.75 ugof wild-type GST-Pim-1. Lanes 3,6,9, 12, 15 each contain 0.75 ug if wild-typeGST-Pim- 1 that has undergone an in vitro kinase reaction with 32P-ATP. Theantibodies used to probe the Western blots are indicated. Antibody binding wasinhibited by the addition of phosphoamino acids; lanes 7-9 with 100 ug/mI ofphospho-serine, lanes 10-12 with 100 ug/mi of phospho-threonine, and lanes13-15 with 100 ug/mI of phospho-tyrosine. Panel B - Autoradiograms of panel A.1 2 3 4 5 6 7 8 9 10 11 12 13 14 151 2 3 4 5 6 7 8 9 10 11 12 13 14 151552. ANALYSIS AND IDENTIFICATION OF GST-PIM-1 AUTOPHOSPHORYLATIONSITESIdentification of the autophosphorylation sites of GST-Pim-1 was attempted. Theintention was to alter the identified autophosphorylation sites by site-directed mutageneisand to assess the effect on Pim- 1 catalytic activity. As well, a comparison of the Pim- 1site(s) to autophosphorylation sites of other kinases could potentially help determine a modeof activation of Pim- 1.i. Two dimensional phosphopeptide analysis of GST-Pim-1Two dimensional phosphopeptide analysis of in vitro autophosphorylated X. laevisGST-Pim-1 yielded 2 strongly phosphorylated peptides (spots 1 & 2), as well as a numberof moderate and weakly phosphorylated peptides (Fig. 21a). Each spot was numbered(Fig. 2 ib) for future reference. Addition of increasing amounts of trypsin caused therelative strength of some spots to change, indicating that some of these peptides were theresult of incomplete tryptic cleavage. None of the minor spots disappeared completelyeven when large amounts of trypsin were added. Dephosphorylation of GST-Pim-1before in vitro autophosphorylation did not result in any significant changes in the relativeintensities of the spots.156AB11013 Q100Figure 21. Two-dimensional phosphopepetide map of X. laevis GST-Pim- 1.Panel A : Autoradiogram of two-dimensional phosphopepetide map of autophosphorylated and trypsonized X. laevis GST-Pim- 1. Panel B: labeling of thespots from the 2D map. Origin is indicated by small circle, arrow indicatesdirection of electrophoresis in first dimension.0123Q4 0QQ57608157ii. Phosphoamino acid analysis of spots from 2D phosphopeptide mappingAfter extraction from the TLC plate, peptides identified by 2D phosphopeptide analysiswere subject to phosphoaniino acid analysis. The results of this experiment are summarizedin Table 10. Sample numbers correspond to the spot numbers in Figure 21b. Somesamples did not yield enough radioactivity to warrant further analysis by IMAC-HPLCESI-MS.Samples # Phosphoamino acid residue1. 52. T3. S4. S (T)5. S (Y)6. S (Y)7. T8. S9. T,S (Y)10. 511. S12. T13. STable 10. Phosphoamino acid analysis of peptides isolated from the 2D phosphopeptidemap.Radiolabelled peptides appearing as spots on the 2D phosphopeptide map (Fig. 21) wereextracted from the cellulose matrix and subjected to phosphoamino acid analysis. Residuesin brackets indicate that the radioactive labelling of that amino acid was difficult to discern.Samples 6 and 8 did not have enough radioactivity to allow analysis by this method.158iii. Identification of Trvptic PhosphopeptidesA collaboration was undertaken with Drs. Lawrence Amankwa and Michael Affolter ofRuedi Aebersold’s laboratory at the Biomedical Research Centre to analyze and identify thepeptides isolated by 2D tryptic phosphopeptide mapping, utilizing new methods developedin their laboratory.a. Analysis of isolated tptic phosphopeptides by IMAC-HPLC-ESI-MSImmobilized metal affinity chromatography (IMAC) - high pressure liquidchromatography (HPLC) - electron spray ionization mass spectroscopy (ESI-MS) is asequential series of methods that allow isolation and characterization of trypticphosphopeptides. IMAC exploits the fact that phosphate adheres to charged ferric cationsnon-covalently, so FeC13 was used to charge a chelating Sepharose Fast Flow matrix.Phosphorylated peptides were selectively retained by the IMAC column and eluted fromthe matrix with sodium phosphate buffer and were subjected to subsequent analysis byHPLC-ESI-MS.Charged particles entering the spectrometer were detected by total ion chromatography.A total ion chromatogram (TIC) for sample 7 is shown in Figure 22, which demonstratesthe presence of two major charged species. As the amount of information from a TIC isvery limited, the chromatograms for the rest of the samples analyzed are not shown.The peaks observed in the TIC from each sample were subjected to mass spectroscopy.Peaks that corresponded to tryptic peptides of GST-Pim- 1 were sequenced by the sequentialremoval of residues during the partial fragmentation by mass spectrometry. Peptide bondswere broken by electron bombardment during the ionization process (Fig. 23) and speciesgenerated by this process are indicated in the mass spectra. Samples not yieldinginterpretable results were not described.Figure 23. Fragmentation of peptide bonds by ionization.N-terminal fragments are represented by A, B and C species, C-terminal fragments arerepesented by X, Y and Z.159jUJy2O/spot#7(IMAC)/EractjuIy2O/spot#7QMAC) -7/20/94-4:49PM320umC18, 4u1/min, 7mm0%B,0-50%B/l5min,g6sec(Th I500/2000(I) c a, a, > a,5.70 I. I-C0.0 1835.010.016515.024532625.040730.0489Time(miri)/ScanUsing IMAC-HPLC-ESI-MS, two definite sites of Pim-1 autophosphorylation wereidentified, and one additional autophosphorylation site was suggested. A tryptic peptidecorresponding to amino acids 185-195 (LIDFGSGALLK), containing the phosphorylatedresidue Ser-190, was identified in sample 1 (Fig. 24). The mass spectrum shows thesingly charged species [M+Hi, the sodium adduct [M+Na], the doubly charged species[M+2H]2+, as well as the fragmentation products y, Y8 y and yio• A tryptic peptidecorresponding to amino acids 196-206 (DTVYTDFDGTR) was identified in sample 7(Fig. 25). Although this peptide contained four phosphorylatable residues, sequencing ofC-terminal fragments indicated that the only residue phosphorylated was the Thr-205.A third site of Pim- 1 autophosphorylation was suggested, Ser-4, in a tryptic peptidecorresponding to the last two residues of GST, and the first 5 residues of Pim-1(GSMLLSK) (Fig. 26). This sample also contained a peptide and peptide fragmentscorresponding to amino acids 56-60, of GST, PYYID. Although this second peptide wasnot phosphorylated, it may have been retained on the IMAC column by interaction with theother bound peptides.An additional autophosphorylation site was identified in the GST portion of the fusionprotein, Thr-17, corresponding to GST residues 10-20 (IKGLVQPTRLL) (Fig. 27). TheGST phosphorylation site was unexpected, as earlier experiments (Figures 16 and 17)demonstrated that GST was not phosphorylated by either H. sapiens or X. laevis GSTPim- 1. It is possible that this phosphorylation resulted either from the in vivo activity of abacterial kinase or results from a strictly intramolecular reaction.CH- -- a-iI IRnAn Bn CnXn Yn Zn161—.(P03H2)100[M+2Hj2LIDFGSGALLKCD a C7500>,*I[M÷HJ‘I50Y94-’ C—G)25[M÷Na]Y3Yiü04006008001000120014001600M/Z(P03H2)DTVYTDFDGTRCD C)100[M+HJ[M÷2HI‘075y7C>C’.4-’Y8C w[M÷Na]4-,Y6-j25y4Y50L I.?iii.iillII.JL.1iL.1I____________________4006008001000120014001600M/Z— C’(P03H2)c:jGSMLLSKCt,[M+H]÷6O{M+Na]rj45>.CD3O0CDy415a___L°Iy50I0Ik11k1..,.i,•1400500600700800900M/Zq(P03H2)IKGLVQPTRLLcl[M+2H12+100[M÷H]+CD 0757 CD>50.[M÷Na]÷C) C’, -3CI) 4-,Y8.25ICy5y6b80Lhr1i...1LLJiLII[Ai.LJALJ1&L4l4006008001000M/Zy10/bb9L’—I.1Il.nvFj,III’LLLIJ IJ?’120014001600b. Analysis of trvpsinized GST-Pim- 1 by LCMSTo determine if the autophosphorylation of GST-Pim- 1 detected was the result of an invivo or an in vitro event, Pim- 1 was analyzed after extraction and purification from thebacteria and prior to autophosphorylation. GST-Pim-1 was thrombin-treated to remove theGST, trypsinized, then the entire sample loaded onto the IMAC column and analyzed byliquid phase mass spectrometry (LCMS). Phosphopeptides containing the phophorylatedSer- 190 site (mlz 1213) as well as the dephosphorylated form (mlz 1133) were identified.This implies that the Pim-1 is autophosphorylated in vivo, in the bacteria. The exactamount of phosphorylation by the time of purification was not established as the LCMStreatment causes removal of the phosphate group, but over 50% of GST-Pim-1 wasestimated phosphorylated on Ser-190 (data not shown).A modification of this method was attempted to specifically identify tyrosinephosphorylation sites, using HPTPI3 to selectively dephosphorylate phosphotyrosinecontaining peptides which were then analysed by LCMS. This exploited the fact thatdephosphorylation would cause a reduction in the atomic mass of a peptide, and thuswould appear shifted in an LCMS chromatogram as compared to non-dephosphorylatedpeptide. Although several peaks that underwent a shift with HPTPB treatment wereidentified in both Pim-1 samples (those that had undergone in vitro autophosphorylation aswell as those which had not) the peaks did not relate to tryptic peptides of Pim- 1.However, all peptides contained a common Tyr-133 site, which may be a potentialautophosphorylation site. No other data were obtained to confirm this result.All the identified autophosphorylation sites of Pim-1 are shown in Figure 28.1664 IXENOPUS MLL(FGSLAHICNPSNMEHLPVKILQPVKVDKEPFEKVYQVGSWGSGGFGTVYSGSRIMOUSE MLL(INSLAHL-RARPCNDLHATKLAPGK-EKEPLESQYQVGPLLGSGGFGSVYSGIRVRAT MLL(INSLAHL-RAAPCNDLHANKLAPGK- EKEPLESQYQVGPLLGSGGFGSVYSGIRVHUMAN MLLCINSLAHL-RAAPCNDLHATKLAPGK-EKEPLESQYQVGPLLGSGGFGSVYSGIRVII III IVXENOPUS ADGQPVAVKHVAKERVTEWGTL-NGVMVPLEIVLLKKVPTAFRGVINLLDWYERPDAFL IMOUSE ADNLPVAIKHVEKDRISDWGELPNGTRVPMEVVLLKKVSSDFSGVIRIJLDWFERPDSFVLRAT ADNLPVAIKHVEKDRISDWGELPNGTRVPMEVVLLKKVSSGFSGVIRLLDWFERPDSFVLHUMAN SDNLPVAIKHVEKDRI SDWGELPNGTRVPMEVVLLKKVSSGFSGVIRLLDWFERPDSFVLV 1133 VIXENOPUS VMERPEPVKDLFEY1ITEKGPLDEDTARGFFRQVLEAVRHCYNCGVVHRDIKDENLLVDTRMOUSE ILERPEPVQDLFDFITERGALQEDLARGFFWQVLEAVRHCHNCGVLHRDIKDENILIDLSRAT ILERPEPVQDLFDFITERGALQEEL,ARSFFWQVLEAVRHCHNCGVLHRDIKDENILIDLNHUMAN ILERPEPVQDLFDFITERGALQEELARSFFWQVLEAVRHCHNCGVLHRDIKDEWIL IDLNVII 190 205 VIII IXXENOPUS NGELKLIDFGI3ALLKDTVYTDFDGfI1RVYSPPEWVRYHRYHGRSATVWSLGVLLYDMVYGMOUSE RGEIKLIDFG3ALLKDTVYTDFDGWkVYSPPEWIRYHRYHGRSAAVWSLGILLYDMVCGRAT RGELKLIDFG3ALLKDTVYTDFDGIRVYSPPEWIRYHRYHGRSAAVWSLGILLYDMVCGHUMAN RGELKLIDFGJbALLKDTVYTDFDGVYS PPEWIRYHRYHGRSAAVWSLGILLYDMVCGx XIXENOPUS DI PFEQDEE IVRVRLCFRRRISTECQQLIKWCLSLRPSDRPTLEQIFDHPWMCKCDLVKSMOUSE DIPFEHDEEI IKGQVFFRQTVSSECQHLIKWCLSLRPSDRPSFEEIRNHPWM-QGDLLPQRAT DI PFEHDEEIVKGQVYFRQRVSSECQHLIRWCLSLRPSDRPSFEEIQNHPWM-QDVLLPQHUMAN DIPFEHDEEI IRGQVFFRQRVSSECQHLIRWCLALRPSDRPTFEEIQNHPWM-QDVLLPQXENOPUS EDCDLRLRTIDNDS S STSSSNESLMOUSE AASEIHLHSLSPGSSKRAT ATAEIHLHSLSPSPSKHUMAN ETAEIHLHSLSPGPSKFigure 28. Autophosphorviation sites of X. laevis Pim-1.Protein sequence alignments of Pim- 1 from X. laevis, mouse, rat and human showing autophosphorylation sites identified by JMAC-HPLC-ESI-MS. Autophosphorylation sites Ser-4,Ser-190 and Thr-205 are boxed, the number of the residue is indicated above the site. Thesuggested Tyr-133 site is also shown. Roman numerals indicate protein kinase subdomains.Residues that are common to all protein kinases are shown in bold type.1673. IDENTIFYING OTHER KINASES WITH SIMILAR PHOSPHORYLATABLERESIDUESProtein sequences were examined to identify kinases with phosphorylatable residues inthe same locations as the Pim- 1 autophosphorylation sites, as these enzymes may alsoautophosphorylate at these sites and may possess similar modes of regulation as the Pim- 1[Hanks, 1993]. Kinase catalytic subdomains containing homologous phosphorylatableresidues are shown with surounding residues. The names of the kinases are shown in thetext below the figures.The Ser- 190 site was very easy to identify in other kinases, as it immediately follows theconserved DFG region in domain VII (Fig. 29). Sequence comparisons of catalytic domainVII indicate that most other kinases contain hydrophobic methionine or leucine residues atthis site. The Thr-205 site was also easy to examine, as it is located seven residuesupstream of the conserved glutamic acid residue in domain VIII. This site is conserved aseither a serine or a threonine residue in most kinases, so only a few kinases are listed inFigure 30. This site undergoes autophosphorylation in other protein kinses such as Rsk.The suspected Tyr-133 autophosphorylation site in domain V of GST-Pim-1 is shown withseveral other kinases having similar tyrosine residues in Figure 31.Many kinases containing homologous phosphorylatable residues to the Pim-l arehomologues of glycogen synthase 3 including the rat GSK-3x/13, S. cerevisiae MCK1,Drosophila sgg/zw3,Arabidopsis ASK-&y’, and the product of the S. cerevisiae MDSIgene. Kinases homologous to cdc/CDC28 family include the KNSI gene product and thehuman Cik protein kinase. Of interest is the fact that many of these kinases, like Pim- 1,have been reported to possess both tyrosine and serine/threonine autophosphotransferaseactivity including the ASKx/y, Clk, and MCKJ/YPK1 gene products. Several of thekinases identified as having homologous phosphorylatable residues were also previouslyidentified as having a high overall sequence homology to the Pim-1 including tsk-1, ASFVand p78.168Pim-1 (Xen) ELKLIDFGSGALLKASFV IIKVIESAVRLNDM HIRLASCLKLRGSK-3a VLKLCQSAKQLVGSK-3i VLKLCSAKQLVsgg/zw3 VLKLCSAKQLLMCK1/YPK1 V L K I C S A K K L EMDS1 SLKLCESAKQLKASK-a QVKLCESAKVLVASK-g QVKLCESAKVLVPSK-G1 AARVVESATFDHPSK-H2 GIKVIESSCYEHDoa DVRLIFGSATFDHKNS1 EIKIISAIFHYYAK1 ELKIISSCEEAELM-i VAKLSSCIFTPPstki NAKLTDFGSARNINCik DIKVVSATYDDFigure 29. Kinases with sites homologous to the Ser-190 in kinase catalyticsubdomain VII.Alignment of amino acids 182-195 of X. laevis Pim-l with corresponding residues ofother kinases. The phosphorylation site is in bold, the conserved kinase motif isunderlined. Homologous kinases include: ASFV, kinase from African Swine fever virus[Baylis et al., 1993]; DM, myotonic dytrophy kinase [Brook et al., 1992]; GSK-3a/B, ratglycogen synthase kinase-3 [Woodgett, 1990]; sgg/zw3,kinase encoded by Drosophilasegment polarity genes [Siegfried et al., 1990]; MCK1, a S. cerevisiae kinase encoded bya meiotic induction gene [Shero and Hieter, 1991]; MDS1 a S. cerevisiae GSK homologueand suppreser of mckl mutants [Puziss et al., 1994]; A. thaliana ASK-a/g, GSK-3homologues, [Bianchi et al., 1994], PSK-Gi and PSK-Hi, putative kinases from HeLacells [Hanks, S .K.]; Doa, the darkener of apricot locus [Yun et al., submitted]; KNS 1, anon essential protein kinase homologue [Padmanabha et al., 1991]; YAK1, S. cerevisiae akinase downstream/parallel to the Ras/cAMP pathway [Garret and Broach, 1989]; PSTKi, a distant relative of c-mos [Lohia and Samuelson, accession L05668], Elm-i, a S.cerevisiae kinase involved in a differentiation pathway induced by nitrogen starvation[Blacketer et al., 1993]; Clk, a human kinase with homology to cdc2-CDC28 [Johnsonand Smith, 1991].169Pim-1 VYTDFDGTRVYSPPEWVASFV PQYNMFGTWEYVCPEFYp78 K L D T F C G S P P Y A A P L Ftsk-1 LSKTFCGSAAYAAPEVLGsk-3cL PNVSYICSRYYRAPELIKIN1 QLHTFCGSLYFAAPELLCik HHSTLVSTRHYRAPEVIMCK1/YPK1 P S I S Y I C S R F Y R A P L IFigure 30. Alignments of subdomain VIII of Pim- 1 with several other kinases.Alignment of amino acids 198-214 of X. laevis Pim-1 with corresponding residues ofother kinases. The phosphorylation site is in bold, the conserved glutamic acid residue isunderlined. Homologous kinases include: ASFV, kinase from African Swine fever virus[Baylis eta!., 1993]; the p78 kinase lost in chemically induced human pancreatic tumors[Maheshwar et al., P27448]; tsk-l, murine testes-specific kinase [Bielke et a!., 1994];GSK-3x rat glycogen synthase kinase-3 [Woodgett, 1990]; KIN1, a S. cerevisiae kinaseinvolved in cell polarity [Levin eta!., 1987]; Clk, a human kinase with homology to cdc2-CDC28 [Johnson and Smith, 1991]; YPK1 a S. cerevisiae putative dual specificity kinase[Dailey et at., 1990], MCK1, a S. cerevisiae kinase encoded by a meiotic induction gene[Shero and Hieter, 1991].Pim-1 (XEN) VKDLFDYITEKASFV SIDLLHYHYFKp78 GGKVFDYLVAHran+ NGDLFTYITEKAKIN1O SGELFDYIVEKSNF1 GNELFDYIVAHKIN1/KIN2 GGQLLDYIIQHc-Abl YGNLLDYLRECARG YGNLLDYLRECPDGFR-f YGDLVDYLHRNFigure 31. Kinases having a tyrosine residue homologous to Tyr-133 site of Pim-1 insubdomain V.Alignment of amino acids 127-137 of X. !aevis Pim-1 with corresponding residues of otherkinases. The phosphorylation site is in bold. Homologous kinases include: ASFV, kinasefrom African Swine fever virus [Baylis et at., 1993]; the p78 kinase lost in chemicallyinduced human pancreatic tumors [Maheshwar et a!., P27448]; ran+, a gene required fornormal meiotic division [Mcleod and Beach, 1986]; AKIN 10, an A. thaliana kinase relatedto SNF1 [Le Guen eta!., 1992]; SNF1, aS. cerevisiae glycogen repression release protein[Celenza and Carlson, 1986]; KIN 1/KIN2, S. cerevisiae genes involved in cell polarity[Levin eta!., 1987]; c-Abl tyrosine kinase [Reddy et at., 1983]; ARG, an abl related gene[Kruh eta!., 1990] and the PDGF-R [Gronwald eta!., 1988].1704. CONSTRUCTION AND BACTERIAL EXPRESSION OF S 190 MUTANTSThe S 190 site was selected for further study, as its identity was most definitive, andrather unusual in its placement as compared to the Thr-205. PCR-mutagenesis was used toconstruct two Pim-1 mutants by changing the Ser-190 residue to alanine (S190>A) and toglutamic acid (S 190>E). Alanine, a non-phosphorylatable residue, mimicked theunphosphorylated state of the Pim- 1, and glutamic acid, a charged acidic residue, was usedto partially mimic phosphorylation at this site. The mutant Pim- 1 species were expressed asbacterial fusion proteins. The the sites of mutation are shown in Figure 32.i. Phosphoamino acid analysis of mutantsThe mutants were expressed as bacterial fusion proteins, and products of the expectedsize were immunodetected with X. laevis Pim- 1 antibodies by Western blotting analysis(Fig. 33). The autophosphorylation of both mutants was significantly reduced as comparedto the WT Pim- 1, which confirmed that the S190 site is an autophosphorylation site of Pim1. Although phosphoamino acid analysis was repeatedly attempted to determine if theproportion of phosphoserine was reduced, autophosphorylation of the mutants did notallow enough isotope to be incorporated into the protein to allow analysis by this method.ii. Specific activity determinationsThe specific activity of autophosphorylation of the mutants was analyzed and found to bereduced in comparison to that of the wild-type X. laevis Pim-1 (107 pmol.miwmg’). Thespecific activity of S 190>A was determined to be 4.9 pmol.min’mg’ (Fig. 34a) and thespecific activity of S190>E was determined to be 5.2 pmol.miir’mg1(Fig. 34b).In summary, the specific activities of autophosphorylation of the S190 mutants werereduced to about 5% of the activity of the WT GST-Pim-1. The differences in the specificactivity determinations between the S190>A and S190>E mutants were not significant.These findings confirm that S190 was indeed one of the major sites of autophosphorylation,since autophosphorylation was reduced in these mutants.171Figure 32. Nucleotide sequence of coding region of X. laevispim-1 showing expected locationof mutations. Amino acid sequence is shown above nucleotides, residues changed by site directedmutagenesis shown in bold type. The * indicates the Clal site where the two PCR fragments wereligated together.ML L S K F G S LA H IC N P SN ME H L P V KATGCTTCTCTCTAAATTCGGATCGCTGGCTCATATCTGCAACCCAAGCAACATGGAGCATCTACCGGTTAAG 72IL Q P V K V D K E P FE K V Y Q V G S V V G SATCTTACAGCCAGTGAAAGTGGACAAGGAGCCCTTCGAGAAGGTGTATCAGGTGGGCTCGGTTGTGGGCAGC 144G G F G TV Y S G SRI AD G Q P V A V K H V AGGTGGTTTCGGCACGGTGTATTCGGGCAGTCGGATTGCAGATGGACAGCCGGTCGCTGTGAAGCACGTAGCT 216K ER VT E W G T L N G VMV P L El V L L K KAAGGAGAGAGTCACAGAATGGGGCACTTTGAACGGCGTGATGGTCCCTTTGGAGATCGTCCTACTAAAGAAG 288V P TA FR G V IN L L D WY ER PD A F LIVGTGCCCACCGCCTTCCGAGGGGTAATCAACCTACTGGACTGGTACGAGCGACCCGACGCCTTCCTGATCGTT 360ME R PEP V K DL F DY IT E KG P L D EDTATGGAGAGACCGGAGCCGGTGAAGGATCTATTCGATTATATAACGGAAAAGGGGCCCCTGGACGAGGACACA 432AR G F FR Q V LEA V RH C Y N C G VV H RDGCCCGCGGTTTTTTCCGGCAGGTGCTGGAAGCGGTGCGACACTGCTATAACTGCGGGGTGGTGCATCGGGAC 5041K DEN L LV D T RN GEL K LI D F G S GAATCAAGGATGAGAACCTGCTGGTGGACACGAGGAACGGGGAACTGAAACTGAT*CGATTTTGGCTCCGGGGCG 576S to A mutant: GCCS to E mutant: GAAL L K D TV Y T D F D G T R V Y 5 PP E WV R YCTACTCAAGGATACGGTGTACACGGATTTTGATGGAACCAGAGTCTATAGTCCACCAGAATGGGTCAGATAC 648Y to E mutant: GAAHR YR G R SAT VW S L G V L L Y D MV Y GDCACAGATACCATGGAAGATCAGCAACCGTGTGGTCTTTGGGTGTGCTTCTTTATGACATGGTTTACGGAGAT 720I P F EQ DEE IV R V R L CF R R RI STE CATTCCCTTTGAGCAAGATGAAGAGATTGTTCGTGTCCGCTTGTGTTTCAGAAGAAGGATCTCTACAGAGTGC 792Q Q LI K W CL S L R PS DR PT L EQ IF D HCAGCAACTTATCAAATGGTGCCTTTCCTTGAGGCCTTCTGATAGACCAACACTTGAGCAAATTTTTGACCAT 864P W MC K CDLV K SE DC DL R L R TI D NDCCTTGGATGTGCAAATGCGACCTTGTGAAATCTGAAGACTGTGATTTAAGACTAAGGACAATTGACAATGAT 936SS ST S S SN ES LTCATCAAGCACAAGCTCAAGCAACGAGAGTCTG 969172A95 —6949—26 —20——.Figure 33. Comparison of mutant and wild-type GST-Pim-l.X. laevis GST-Pim-l mutants S19OA (lane 1), S19OE (lane 2) andwild-type (lane 3) were subjected to an in vitro autophosphorylationreaction then analysed by SDS-PAGE. Panel A is a Western blotprobed with Piml-XI antibody, panel B shows an autoradiogramof Panel A, exposed for 64 h. Panel C shows a silver stain of themutants. Approximately 0.33 ug of protein were used for eachreaction. Migrations of Mr standards are shown on the left.1 2 31734Panel BTime of assay (mm)20Figure 34.Pim- 1.Soecific activity of autoohosohorviation of S190>A and S190>Emutants ofThe pmol of [y32P] incorporated per mg of GST-Pim-1 was plotted against the time of theassay. The slope, as calculated by the equation, is the specific activity. Panel A is thespecific activity determination of the S 190>A mutant, panel B is the specific activitydetermination of the Si 90>E mutant.Panel A120100804 6040200 5 10 15 20Time of assay (mm)110 5 10 15174iii. Exogenous kinase activity of mutantsTo assess whether the exogenous phosphotransferase activities of the Ser-190 mutantswere different from the WT, peptide substrate assays were done using P4 peptide as asubstrate. The Km and Vmax values of the mutants for the P4 peptide are shown in Tableii. The apparent Km values were similar between Si 90>E and the WT, indicating that thereplacement of a serine to a glutamic acid did not change the affinity of the enzyme for thesubstrate peptide. The Km value for the S 190>A mutant was much lower than with theWT, indicating that the mutant has a higher affinity for the substrate peptide. The Vmaxvalue of both Si 90>A and Si 90>E are similar and much lower than the Vm value of theWT, indicating that mutations of the Ser- 190 site caused a reduction in activity as comparedtotheWT.5. CONSTRUCTION OF ADDITIONAL MUTANTSThe Tyr- 199 site has been suggested as an autophosphorylation site of Pirn- 1, based onits homology to the Src autophosphorylation site, Tyr-4i6 [Cooper and MacAuley, 1988].A Yi99>E mutant was attempted as described in the Methods. Sequencing of thecompleted mutant revealed that gene duplication had occurred during the PCR mutagenesis.As advanced analysis of Pim-1 by Dr. Lawrence Amankwa failed to identify the Tyr-199site as being phosphorylated, we elected not to continue construction of this mutant.175TABLE 11. ACTIVITY OF SER-190 SITE MUTANTS TOWARD P4 PEPTIDE.Vmax Km Vmax(pmol.miwl.mg-l) (.tM) KmWild type 124 35 3.5S190>A 17 5 3.4S190>E 16 42 0.41766. CONCLUSIONS OF PIM-l AUTOPHOSPHORYLATIONPhosphoamino acid analysis of the human and amphibian bacterially-expressed GSTPim-i (Chapter IV) revealed that the proteins autophosphorylated on serine, threonine andtyrosine residues. Construction of a kinase dead mutant confirmed that this activity wasdue to autophosphorylation and not due to the activity of contaminating kinases. Neitherautophosphorylation nor dephosphorylation seemed to affect the exogenousphosphotransferase activity of GST-Pim- 1. Although there did appear to be a slightreduction in autophosphorylation activity after phosphatase treatment, this could have beendue to residual phosphatases in the preparation.The majority of samples analyzed by IMAC-HPLC-ESI-MS did not contain peptidesthat corresponded to expected tryptic peptides of GST-Pim- 1. This was not surprising, asphosphopeptides may have been purified in amounts too small to be detected or elseunderwent massive degradation during the analysis procedure. Although the resolution ofphosphopeptides by 2D mapping was quite effective, TIC revealed that most samplesanalyzed contained several peptides that were retained on the IMAC column. Nonphosphorylated pepides might have been non-selectively retained on the IMAC due to thepresence of highly acidic residues or by interactions with the selectively retainedphosphopeptides. Partial fragmentation bombardment of the peptide during the ionizationprocess in the ion source during mass spectrometry allowed sequencing through thepeptide, confirmed the identify of the peptide and allowed identification of the exact site ofphosphorylation.Many of the samples analyzed contained peptides with the Ser-190 site, which indicatedthat the large numbers of spots on the 2D map resulted from incomplete tryptic digestionand did not represent a large number of different autophosphorylation sites. Peptidescontaining the S190 site were also identified by LCMS analysis serving to confirm that themain site of phosphorylation of Pim- 1 was Si 90. This site is conserved in all other Pim- 1homologues as well as in several other kinases and may be a physiological site ofautophosphorylation.The importance of the Ser- 190 site was investigated using mutants generated by PCR-sitedirected mutagenesis and expressed in bacteria. Reduced autophosphorylation of themutants indicate that the Ser-190 was a major autophosphorylation site. It was notdetermined in the reduction of autophosphorylation was due to the removal of the177phosphorylatable residue or due to a reduction in the activity caused by a conformationalchange. The Si 90>A mutant had a higher affinity (apparent Km) for the P4 peptidesubstrate that the wild-type or the S 190>E mutant, indicating that replacement of the serinefor an alanine induced some kind of change in the peptide binding site. Despite theincreased affinity for the peptide substrate, the S 190>A mutant had a 7-fold reduction incatalytic activity compared to wild type Pim- 1. This implies that autophosphorylation at thissite may be activating. The Si 90>E mutant displayed a similar affinity for the substratepeptide as with the WT mutant, but also had an 8-fold reduction in activity. Presumably theglutamic acid residue was only able to partially mimic a phosphorylation event.The other autophosphorylation sites of Pim-i identified using IMAC-HPLC-ESI-MSanalysis of tryptic phosphopeptides were Ser-4 and Thr-205. The Ser-4 phosphorylationsite is also conserved in all other Pim- 1 homologues, but as it is outside of the catalyticdomain, it was difficult to compare to other kinases. Phosphorylation sites are often locatednear the N- and C- termini of proteins as these regions are generally located at the surface ofprotiens and are often more accessible. It is not known if phosphorylation of this site hasany physiological importance. The functional or physiological importance of the Thr-205site was not assessed by mutational analysis. In contrast to X. laevis GST-Pim-l themajority of autophosphorylation of H. sapiens GST-Pim-1 is located on threonine residues(Fig. 20), so it is possible that this Thr-205 site may represent the physiologically mostimportant autophosphorylation site, at least in the human enzyme.Although the PAA indicates that GST-Pim- i autophosphorylated on tyrosine residues,we were unable to unequivocally identify a tyrosine phosphorylation site by IMAC-HPLCESI-MS. The presence of tyrosine autophosphorylation sites were examined using thenovel approach of identifying peaks that were shifted in mass and retention time aftertreatment with the tyrosine-specific phosphatase HPTPB. Although we were able toobserve peptides that shifted in LCMS retention time after HPTPB treatment, these peptideswere difficult to relate to the Pim-l. A tentative site, Tyr-133, was suggested by thismethod. However, we were unable to confirm this result. This site was not conserved inthe other pim- 1 cognates and therefore was unlikely to be a physiologically important sitefor the modulation of Pim-1 activity. As well, we have shown that human GST-Pim-l isnot immunodetected by anti-phosphotyrosine residues, which confirms that this tyrosineautophosphorylation may be species-specific, unlikely in such a conserved kinase, or mayeven be an artifact of bacterial expression. For this reason, this site was not exploredfurther by mutational analysis.178Results of the mutant studies combined with the fact that over 50% of GST-Pim- 1 isphosphorylated prior to purification from bacteria would suggest that autophosphorylationof the Ser- 190 site is an activating event. This does not preclude the requirement of otherprotein subunits for the regulation or localization of Pim- 1 activity. This will be consideredin greater detail in the discussion.179PART 3 - RESULTS AND CONCLUSIONS: PIM-1 IN BIOLOGICALSYSTEMSCHAPTER VIII.EXAMINING THE ACTIVATION OF ENDOGENOUS PIM-1 DURINGXENOPUS LAEVIS OOCYTE MATURATIONOne of the original aims of this research project was to study the changes in Pim- 1during Xenopus laevis oocyte maturation. The regulation and role of Pim- 1 in cells of thegerm line has been relatively neglected in the scientific literature, despite the high level ofPim- 1 expression in spermatozoa. The work described in Chapters IV-VI was initiated inpart to develop reagents to assist in the assessment of Pim- 1 in the oocyte system. Notonly were new antibodies developed to specifically recognize X. laevis Pim- 1, but peptideswere also constructed as substrates to measure the activity of this kinase.The studies involving bacterially-expressed GST-Pim- 1 were very productive.However, it was unclear how these findings applied to the endogenous kinase. It was ourintention to exploit the X. laevis oocyte system not only to study the Pim- 1 duringmaturation, but to hopefully confirm the earlier results obtained with the bacterially-expressed kinase.1. DETECTION OF ENDOGENOUS PIM-1 IN XENOPUS LAEVIS OOCYTEEXTRACTSAs described in Appendix II, X. laevis-specific Pim-1 antibodies, GXP and Piml-XI,were developed. Both of these antibodies recognized the bacterially expressed X. laevisGST-Pim-1 and detected proteins in crude X. laevis oocyte homogenates on Western blots.Unfortunately, not only did these antibodies fail to successfully immunoprecipitate theendogenous Pim-1, but they recognized different proteins on Western blots.Homogenates from a X. laevis oocyte maturation time course were screened with bothGXP and Piml-XI antibodies (Fig. 35). Although the gels were obviously overloaded, avery strong 42 kDa band was detected in all samples with the GXP antibody. Because ofthe large amount of protein present, it was difficult to assess if the quantity or size of theprotein changed during the maturation. The immunoblot probed with the Piml-XI180antibody fared slightly worse, with a high degree of non-specific binding of the antibody tothe blot. In this case it is impossible to assess which band corresponded to Pim- 1.Inhibition of antibody binding with expressed GST-Pim- 1 was attempted to confirm thespecificity of binding of the antibodies to Pim-1. Inhibition by GST-Pim-1 profoundlychanged the pattern of immunoreactivity of the blot, so it was difficult to determine whichof the proteins was specifically Pim- 1. The most notable changes involved the competititveinhibition of a 42 kDa Piml-XI immunoreactive band and a 44-46 kDa GXPimmunoreactive doublet (data not shown).181Time(h)01Piml-XIFigure35.ImmunodetectionofPim-1inhomogenatesofaX.laevisoocytematurationtimecourse.X.laevisoocytetimecoursehomogenateswereWesternblottedandprobedwithPiml-XIandGXPantibodies.Thenumbersbelowtheblotsshowthetimecourseofmaturationinhoursfrom0(immatureoocytes)to7hourspost-progesterone.GVBDoccuredat3.5hours.Theantibodiesusedareindicatedaboveblots,andthemigrationsof Mrstandandsareshownontheleft.Thearrowindicatesanapproximately42kDaproteinthatmaycorrespondtoPim-1.(W23456701234567002. ANALYZING ACTIVATION OF PIM-1 DURING OOCYTE MATURATIONi. Peptide substratesThe activation profile of Pim- 1 during oocyte maturation was examined using peptidesubstrates. Immature and mature X. laevis oocytes were subjected to by MonoQ columnchromatography and the fractions were analyzed by kinase assays. Initially, histone Hiwas used as a substrate, until a series of Pim- 1 -specific peptide substrates were developed.The S6, P3 and P5 peptides were used to distinguish between the activities of S6 kinaseand Pim- 1. No difference was observed in the activation patterns as detected using thesepeptides, so for later experiments only P4, the optimal Pim- 1 peptide substrate peptide,was used.Preliminary studies detected a large peak of peptide P4 (AKRRRLSA) phosphorylatingactivity that eluted with 0.3 M NaCl (fractions 24-25) and a second minor peak was elutedwith 0.35 M NaC1 (fraction 29) (Fig. 36). Maturation induced-activation of the first peakwas 1.3 X that of the control, and maturation induced-activation ofthe second peak was4.3 X over the control.Due to the promising results obtained between immature and mature X.laevis oocytes,a time course study of P4 phosphorylating activity during X. laevis oocyte maturation wasundertaken (Fig. 37). A significant increase in total P4 phosphotransferase activity did notoccur with oocyte maturation, the peak of P4 phosphorylating activity shifted radicallybetween time points and there were not trends in activation discernable. For this reason,this approach was discontinued.ii. Immunoreactivitv of oocyte maturation time courseWestern blots to analyze the Pim- 1 protein at each time point of the time course wereprobed with the Piml-XI antibodies (Fig. 38). ECL was used to visualize the blots, aswehoped that a more sensitive detection system would allow distinction ofthe antibody bandfrom the background. In the blots for the time points shown, there was a strong 45 kDaimmunoreactive band in fractions 24-26, and a slightly higher band infractions 26 to 30.The bands were shifted slightly between one fraction and anotherin the different timepoints; this is consistent with the elution of the peak fractions as observed in Figure 37.Reprobing the blots with GXP antibody detected a single 40 kDa band in lanes 22-24. Theproteins detected by the two antibodies eluted in different fractions (data not shown).183To try to better assess changes between the different time points, fractions 25, 27 and30 from each time point were analysed by Western blotting with various Pim- 1 antibodies(Fig. 39). The Piml-XI antibody detected a series of proteins with Mr from 42-48 kDa inall fractions. The differences in these proteins between time points were too profound toestablish actual changes in protein quantity and size occurring during maturation and wereprobably a consequence of the peak shifts in the Mono Q elution. Reprobing the Westernblots with the GXP demonstrated that some proteins immunoblotted with both antibodies.However, it was not possible to clearly establish whether changes occurred in theseproteins during the maturation process.1849080- 7.70• Ohours! ---- 5hours< 60 I’I ‘VI;’n 50 Io I40 I; Io IC.) I • 4.S 30 I ‘ / .4I ! V20Ii \10-*___•i•i•i•i.i•i•i•i•i•t•i•i•i•20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38MonoQ fraction numberFigure 36. P4 peptide phosphorviation by fractionated immature and matureX. laevis oocyte extracts.Immature and mature (5 hours) X. laevis oocyte extracts were fractionated by MonoQcolumn chromatography and used to phosphorylate peptide P4 in the presenceof [y-32P]ATP. The enzyme activity of of the mature extracts in fractions 22-32as measured by the activity under the graph, was 1 .8X that of the immature extracts.GVBD occured at 3.5 hours.185CEC)20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38MonoQ fraction numberFigure 37. P4 peptide phosphorviation by fractionated X. laevis oocytematuration time course.X. laevis oocyte extracts were fractionated by MonoQ column chromatographyand used to phosphorylate peptide P4 in the presence of[y-32P]ATP. The timepoints used are shown in the legend. GVDB occured at 3.5 hours.10090807060—0--—DA---A---.-4.-..O hours0.5 hours1 hours1.5 hours2 hours2.5 hours4 hours5 hours6 hours30186Figure 38. Western blots of fractionated X. laevis oocyte extracts.X. laevis oocyte extracts from various time points were fractionated by MonoQcolumn chromatography. Western blots of selected fractions (fraction numbersshown below blots) were probed with Piml-XI antibody. GST-Pim-1 was usedas a positive control (on right). Migration of Mr standards are shown on the left.Although the time course was performed from 0-6 hours, only selected time pointsare shown. Fractions highlighted by stars (25, 27, 30) were used for a comparisonof time points (Fig. 39). GVBD occurred at 3.5 h.16 18 2022232425 26 2728 293031 32 3436 - Pim* * *187Piml-XIGXP45600.51.01.522.53456Time(h)Figure39.WesternblotsofselectedMonoOfractionsofX. laevisoocytetimecourse.TimepointsfromaX.laevisoocytematurationtimecoursewerefractionatedbyMonoQcolumnchomatography.SelectedfractionswereWesternblottedandprobedwiththePiml-XI andtheGXPantibodies(labelled abovepanels). Thetimepointsareshownbelowblotsandareexpressedinhours.Note:Timepoint3containedalmost noprotein.GVBDoccurredat3.5hours.00.51.01.522.53Go Goiii. Sunimarv of X. laevis maturation time course resultsIt was hoped that X. laevis oocyte maturation would be an ideal system in which tostudy the activation of Pim- 1 protein, but we experienced several problems with thissystem and the methods of analysis. First of all, there was ambiguity about which of theimmunodetected proteins actually corresponded to Pim- 1, as both antibodies detecteddifferent bands in the X. laevis oocyte extracts. By fractionating the extracts, we hoped toestablish which band was Pim- 1. Unfortunately, our antibodies detected proteins of theappropriate size range in slightly different fractions. Some of the same proteins weredetected by the different antibodies, but it was still a “pick and choose” situation and theuncertainty as to the assignment of which of these was Pim-1 was not resolved.Secondly, the assessment of Pim- 1 activation was difficult because the P4 peptide usedto analyze the different fractions was also a substrate for other protein kinases beside Pim1. As we did not know if the P4 peptide phosphotransferase activity observed arose fromPim-1 or from an other kinase, we hoped that immunodetection by Pim-1-specificantibodies would facilitate identification of which fraction contained Pim- 1. Unfortunately,there were immunoreactive bands in the expected size range present in most of the fractionsanalyzed. Shifting of the peak P4 phosphorylating fractions from one MonoQ elution toanother also made analysis of results difficult. Minor shifts (1-3 fractions) in proteinretention might be expected to sometimes occur between immature and mature extracts, asproteins that become phosphorylated during the course of activation sometimes experienceslight differences in adhesion to the MonoQ matrix. Consistently cleaning and washing thecolumn between uses did not prevent the minor shifting from occurring. As varyingcharacteristics of different Mono Q columns could profoundly influence the outcome of theexperiment, all samples in a set were analyzed on the same day with the same Mono Qcolumn.The X. laevis oocyte extracts displayed some biological inconsistencies, possibly dueto the frogs themselves (age, health, diet, seasonal). The number of oocytes from a givenfrog varied immensely, as did the proportion of eggs in the different developmental stages,which was reflected by the differences in time for GVBD between the eggs from differentfrogs. In some sets of oocytes, GVBD would first begin to appear after 3 hours ofprogesterone treatment and by 6 hours, a large percentage (65%) of stage VI oocytes wouldmature. Sometimes only 10% of oocytes would experience GVBD. In other sets ofoocytes, GVBD would first occur at 8 hours and by 16 hours, 50% of the oocytesunderwent GVBD. Yet in other experiments, the oocytes would just mottle and GVBD189would not occur at all. As the treatment of the oocytes was consistent as possible for morethan 20 different frogs was studied, these differences may be due to the genetic or seasonaldifferences of the X. laevis population.1903. CONCLUSIONSThe determination of the regulation of endogenous Pim- 1 during X. laevis oocytematuration was not very successful due to inconsistencies of immunodetection with the twoantibodies and to difficulties with the X. laevis oocyte system. The two different X. laevisbased Pim- 1 antibodies immunodetected different proteins in fractionated extracts causingsome concern about antibody specificity. Although there seemed to be a slight activation inPim- 1 activity as measured by P4 peptide phosphorylation, it is not entirely certain if thisactivity was due to Pim-l or to another kinase.Preliminary experiments were performed to identify proteins in the X. laevis extractsthat interact with Pim-1. These experiments, Fusion protein affinity columns and FarWestern blotting, are detailed in Appendix VIII. Unfortunately, these methods were notvery fruitful in this study. Although proteins of 25 and 40 kDa from X. laevis oocyteextracts bound to and were phosphorylated by the human GST-Pim- 1 matrix, they werenot observed when X. laevis oocyte extracts were used to bind to X. laevis GST-Pim- 1.These techniques might be applied in the future when the identity of a downstream target issuspected.In conclusion, the X. laevis oocyte system was not a good model for the study of Pim1 activation. The variable results that were consistently obtained with this system are incontrast to the reproducible results obtained with the sea star oocyte system. The peptidesubstrates and reagents developed for use in this system seem to work well in otherspecies. As the results from the different batches of X. laevis oocytes varied considerably,and because intensive amounts of work was required to obtain small numbers of oocytes,this system was impractical and not recommended for further study.191CHAPTER IX.EXAMINING PIM-1 ACTIVITY IN MATURING PISASTER OCHRACEUSOOCYTESThe Pisaster ochraceus (the purple sea star) was selected as a second model system inwhich to study the role of Pim- 1 in oocyte maturation. An advantage of this system is thatthe immature oocytes are uniformly arrested at the begining of prophase of the cell cycle,and can be induced to resume meiotic maturation by the addition of 1-methyladenine.Germinal vessicle breakdown (GVBD) typically occurs about 70 minutes following theaddition of 1-methyladenine to the P. ochraceus oocytes. Unlike the X. laevis oocytesystem, ample quantities of sea star oocyte homogenates can be obtained for purificationpurposes. Additionally, the echinoderm oocytes can easily be fertilized in vitro, and therelatively large size of the oocyte is ideal for microinjection.1. PARTIAL CLONING OF SEA STAR PIM-1 BY PCR. AND COMPARISON TOPUBLISHED SEOUENCESTo validate the study of Pim- 1 in sea star, we initially exploited PCR techniques toconfirm the presence of this kinase in this primitive organism. Oligonucleotides based onthe nucleotide sequence of mammalian pim-1 were successfully used to amplify a part ofthe pim-1 coding region from P. ochraceus eDNA using PCR. The fragment wassubcloned into the XL 1-blue plasmid and sequenced. The partial nucleotide sequence isshown with the translated amino acid sequence in all three reading frames (Fig. 40).Sequencing of this fragment was carried out until the identity of the fragment wasconfirmed as being a homologue of pim-1. The alignment of translated regions of the seastar pim-1 with the mammalian and X. laevis amino acid sequences is shown in Figure 41.This sea star pim-1 PCR fragment was used as a probe to screen three P. ochraceuslibraries that were in the lab. One library was labelled simply “P. ochraceus”, while thetwo libraries obtained from Dr. Michael Smith (Simon Fraser University) were P.ochraceus oocyte (2 x 108 pfulul) and testes (3x106pfu/ul) cDNA libraries in ?gt10 (3x106pfu/ul). The oocyte library was retitred at 3.5 x 106 pfu/ul and the testes library wasretitred at 5.23 x iO pfu/ul. The libraries were plated and screened several times, withlittle success; the fact that both libraries had been reduced to less than 2% of the originaltitre indicates that the libraries had probably both undergone degradation due to improper192storage or DNAse contamination. Screening was discontinued due to the questionablequality of the libraries.The success in amplifying a fragment of pim-1 from sea star cDNA was due to a highlevel of pim-1 conservation that extends to species as diverse as sea star and X. laevis. Incontrast, we were unsuccessful at amplifying raf- 1, another oncogene-encodedserine/threonine kinase, from sea star cDNA, also using oligonucleotides based on themammalian raf-1 sequence. Apparently, pim-1 is more highly conserved between sea starand mammals than raf- 1 is.193Figure 40. Partial nucleotide and amino acid sequence of sea star Pim- 1.Partial nucleotide sequence of sea star Pim- 1 showing translated regions in three readingframes. Nucleotides in brackets represent uncertain residues. The symbol “@“ represents astop codon.CCCCCCTGACCCGGGCTCGAGGCGCCGGGGAAGGAGAA (A/G) GAGCCXTTC (C) (G) GAAAAPP @ PG LEAP G K E K E P F R KP P D P G S R R R G R R K/R S ? S G KP L T R AR GAGE GE G/R A? P E KGACGTATGCAATCGGTTCACTGCTTGGGAGCGGAGGTTTTGGGACCGTTTACTCGGGCACTCGR RN Q S V H CL GA E V L G P FT R ALDV C N R FT AWE AR F W DR L L G H ST Y Al G S L L G S G G F G TV Y S G T RAATCAGAGACAATTTGCCGGTTGCTATCAAGCTTGTGAC (G/ C) AAAGAAAAAGTGAACGATTE S E T I C R L L S S L P/R K K K T IN Q R Q PA G C Y Q A CD E/Q R K SE R LF RD N L P V A 1K LV T K E K V ND.GGAACATGATTAATGGACAGAAAGTTCCTCTAGA (A) GT (T) CATCTCCTGAAGAAGGTCGACG T @ L N DR K FL @ K F IS @ R R STE H Y @ W TESS SR S S S PEE G RW N MI N GO K VP L E V H L L K K V DCACATACCGATGTATAAAGATGCTGGATATCTATGATAGGGCAGACAATTTTATCATCGT YR C 1KM L DI Y DR AD N F IIPH T DV @ R C WI SM I G Q TI L SSH 1PM Y K D AG Y L @ G R Q F F HR(A! T) CCATGGAACGGCCCCGAACCTGCAAAGGACTTATTTGATTTCATCACCGAGAGTGGGC(D/V) H G T A P N L Q R T Y L I S S P R V GS/T M E R P R T C K G L I @ F H I-I R E W AP W N G PEP A K DL F D FIT ES G194Fig. 40. (Continued).CCGC (A/ T) GGGAGAGGAGACAGACGAAAGTTTTTCCACCAAGTTGTAGAGACGACTTCGGCTP Q/L GEE T DES F ST K L @ R R L R LR R/W ERR Q T K V F P P SC RD D F GPA G R GD R R K F F HO VV E T T S AACGGTGCCACGAGGCTGGTGTTCTCCACAGAGACCTCAAAGACGAGAACATTCCGGTCGATCTR C HE AG V L H RD L K DEN I P V DLY GA T R LV F STE T SK T R T FR SMTV PR G W CS P Q R P Q R RE H S G R SCTCAAAACTGGAGACCT (T / C) AAACTCATCGACTTTGGATCAGGCGCTATTCTCAAAGATACS K LET L/S N SS T L D Q AL F SKIS Q N W R P Q/@ T H R L W I R R Y S Q R YL K T GD L K LID F G S GAIL K D TCGTCTACAAAGATTTCGATGGTACTCGTGTGTACAGCCCACCAGAGTGGATTCGATCCCATCGPS T K IS MV LV C TA H Q S G F DPIR L Q R TRW Y SC V Q PT R V D SIPSV Y K D F D G T R V Y S PP E WI R 5 HRTTACCACGGTCGTCCTGCCACAGTCTGGTCGCTGGGCATCCTCCTGTATGACATGGCCTGTGGVT TV V L P Q S G R WAS SCM T W P VL PR S SC H S L VA G H PP V ‘ HG LWY HG R PAT VW S L GILL Y DMA C GAGATATACCCTTCGAACACGACGAGGAAATCTCGAGCCCGGGTCAGE I Y P 5 N T T R K S R A R V S/RR Y T L R T R R G N L E PG SDIP FE H DEE IS S PG 0.195Figure 41. Comparison of the partial amino acid sequence of sea star Pim- 1 with other species.Alignment of the sea star Pim-1 amino acid sequence (amino acids 37-24 1) with the other knownhomologues. The sea star open reading frame was taken from Figure 40, and frame shifts havenot been indicated. Unknown residues are indicated by question marks (?), stars (*) indicateresidues identical between all Pim-1 proteins, and dots (.) indicate residues conserved between allPim-1 proteins. Roman numerals indicate kinase catalytic subdomains. Residues in bold typecorrespond to Pim-1 autophosphorylation sites.I II IIIPIM1 SEASTAR EKTYAIGSLLGSGGFGTVYSGTRFRDNLPVAIKLVTKEEVNDWNMI -NGQKVPLEVHLLKPIM1cXENLA EKVYQVGSVVGSGGFGTVYSGSRIADGQPVAVKHVAKERVTEWGTL—NGVMVPLEIVLLKPIM1cMOUSE ESQYQVGPLLGSGGFGSVYSGIRVADNLPVAIKHVEKDRI SDWGELPNGTRVPMEWLLKPIM1 cRAT ESQYQVGPLLGSGGFGSVYSGIRVADNLPVAIKHVEKDRI SDWGELPNGTRVPMEVVLLKPIM1 cHUMAN ESQYQVGPLLGSGGFGSVYSGIRVSDNLPVAIKHVEKDRI SDWGELPNGTRVPMEVVLLK* * *. . .******.**** *. .* ***.* * *. . . .* ** **.*. ***IV VPIM1 SEASTAR KVSTTYRC-IKMLDIYDRADNFIIVMERPEPAKDLFDFITESGPAGRGDRRKFFHQVVE?PIM1 cXENLA KVPTAFRGVINLLDWYERPDAFL IVMERPEPVKDLFDYITEKGPLDEDTARGFFRQVLEAPIM1cMQUSE KVSSDFSGVIRLLDWFERPDSFVLILERPEPVQDLJFDFITERGALQEDLARGFFWQVLEAPIM1 cRAT KVS SGFSGVIRLLDWFERPDSFVLILERPEPVQDLFDFITERGALQEELARSFFWQVLEAPIM1 cHUMAN KVSSGFSGVIRLLDWFERPDSFVLILERPEPVQDLFDFITERGALQEELARSFFWQVLEA**. . * ****. * * * *****. .**** *** * *.** **.*.VI VII VIIIPIM1 SEASTAR ? ? ?CHEAGVLHRDLKDENIPVDLS?GDLKLIDFGSGAILKDTVYKDFDGTRVYSPPEWIRPIM1cXENLA VRHCYNCGVVHRDIKDENLLVDTRNGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWVRPIM1cMOUSE VRHCHNCGVLHRDIKDENILIDLSRGEIKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRP IM1cRAT VRHCHNCGVLHRDIKDENILI DLNRGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIRPIM1cHUMAN VRHCHNCGVLHRDIKDENILIDLNRGELKLIDFGSGALLKDTVYTDFDGTRVYSPPEWIR**.***.****. .* *. .*********.****** *************.*IxPIM1 SEASTAR SHRYHGRPATVWSLGILLYDMACGDIPIM1 cXENLA YHRYHGRSATVWSLGVLLYDMVYGDIPIM1 cMOUSE YHRYHGRSAAVWSLGILLYDMVCGDIPIM1cRAT YHRYHGRSAAVWSLGILLYDMVCGDIPIM1cHUMAN YHRYHGRSAAVWSLGILLYDMVCGDI****** *.*****.***** ***1962. NORTHERN BLOHING OF ENDOGENOUS SEA STAR PIM-1To determine if p/rn-i mRNA was expressed in sea star oocytes, the PCR fragment ofthe P. ochraceus p/rn-i coding region was used to probe Northern blots of immature seastar oocyte mRNA. The total RNA (Fig. 42, lane 1) contained 1.7, 2.5 and 2.7 kbfragments. The poly(A)+ sample (lane 2) contained only the 2.7 kb species whichindicated that only this species was polyadenylated. This 2.7 kb mRNA species wassimilar in size to that found in other species [Domen et al., 1987; Meeker et al., 1987a].Although the 2.7 kb fragment in the poly(A)+ lane seemed to be slightly higher than in thepoly(A)- lanes, this was probably due to interference by the 2.8 kb tRNA in the poly(A)samples. The 2.7 kb band present in lane 2 may be poly(A)+ RNA that was notsuccessfully separated from the rest of the RNA. The presence of the 2.6 kb and 1.7 kbbands in the poly(A)- mRNA other bands was unexpected. These shorter transcripts mayhave resulted from other genes that were highly homologous to pirn-i. However, p/rn-ihas no close relatives and these bands were visible even after stringent washes of the blot.This indicates that these 1.7 and 2.6 kb bands were probably shorter non-polyadenylatedversions of the p/rn-i transcript, possibly arising as a result of aborted transcription oralternate mRNA splicing. Although no descriptions of p/rn-i oocyte mRNA have beenpublished, several other groups have reported finding a shorter, 2.4 kb p/rn-i transcript inmurine and rat testes [Sorrentino et al, 1988; Wingette et a!., 19911. These blots were notprobed with a glycerol-3-phosphate dehydrogenase (GAP-DH) probe to control for theamount of RNA loaded.197Figure 42. Northern blot of sea star RNA.Nothern blot of total RNA (lane 1) and poly(A)+RNA (lane 2) purified on an oligo(dT) column.RNA was isolated from immature P. ochraceusoocytes. Location of 18 and 28s rRNA fragmentsare indicated by the arrows, migrations of RNAstandards are shown on right.1983. EXAMINING CHANGES IN PIM-l DURING SEA STAR OOCYTEMATURATIONi. Western blotting of Pim- 1 from sea star oocyte extractsAs was demonstrated in Appendix II, Pimi-Ill and Piml-NT antibodiesimmunodetected a 42 kDa protein from crude sea star extracts. The antibodies were used totest Mono Q-fractionated sea star extracts, at various time points during oocyte maturation.Initially we probed the Western blots of Mono Q-fractionated immature and mature oocyteextracts with both antibodies, and found a strong, approximately 42 kDa band detected infractions 25-27 (0.32-0.38 M NaC1) with Piml-NT, and in fractions 24-26 (0.28-0.32 MNaC1) with Pimi -III (Fig. 43). From this blot, it was difficult to determine if the bandsimmunodetected were the same size. However, the fact that they were eluted in slightlydifferent fractions strongly implies that the antibodies detected different proteins. Thebands were shifted by one fraction from the immature to the mature suggesting that theproteins experience a change during maturation (e.g. phosphorylation) which causes themto be retained on the colunm. These bands were detected in repeated experiments but theredid not seem to be quantitative or qualitative changes in the Pim- 1 protein with maturationas evidenced by band shifts.ii. In vitro peptide studies to examine Pim- 1 activation during time courses of oocytematurationChanges in endogenous Pim-l protein during sea star oocyte maturation were assessedby peptide substrate assays of MonoQ-fractionated oocyte extracts. There was someconcern that other maturation activated kinases, in particular, S6 kinase, would alsophosphorylate the P4 peptide. For this reason, we screened our Mono Q fractions withtwo different peptides to try to distinguish the activity of Pim- 1 from that of S6 kinase. Wescreened with P3 peptide which was phosphorylated strongly by purified S6 kinase andweakly by GST-Pim- 1 and P4 peptide which was phosphorylated efficiently by GST-Pim1 but poorly by S6 kinase. There was a large maturation-activated P3 activity peak infractions 27 (0.38 M NaCl), with shoulder peaks at fractions 31(0.4 M NaC1) and 34(0.46 M NaCl) (Fig. 44). P4 showed a large maturation-activated peak in fractions 26-27(0.38 M NaC1) and a shoulder peak in fraction 36(0.5 M NaC1).The peak fractions from each time point were Western blotted and probed with bothPimi-ifi and Piml-NT antibodies (Fig. 45). The two antibodies detected slightly differentbands; Piml-NT detected a 40 kDa band that eluted in a number of fractions, from 22-26,199Pim1-NTPimi-ifiI M49.5-32.5-i49.5-I32.5-Figure43.WesternblotofMonoO-fractionateclimmatureandmatureseastaroocyteextract.WesternblotsofimmatureandmatureMonoQ-fractionatedseastaroocytecytosolicextractsprobedwithPiml-NTantibody,thenstrippedandreprobedwithPimi-HIantibodies.TheMonoQfractionnumbersareshownbelowblots.Immature(I)extractsareonthetop,mature(M)extractsonthebottom.Migrationsof Mrstandardsareshownontheleft.101620222425262728293031323334353637101620222425262728293031323334353637CA P3 peptide phosphorylation profile140 -130- • 0mm120---0” i5min110-• 30mm2021 2223 2425 2627 28 29 30 31 32 33 3435 36 37 38 39MonoQ fraction numberB P4 peptide phosphorylation profile140130.120.C—• 100.90.. 80.70. \60. 150. 140. I30.20.10. 0- 2--_______02021 2223________2728 29 30 31 3233 3435 3637 38 39MonoQ fraction numberFigure 44. Peptide phosphorviation profiles of fractionated sea star oocytematuration time course. Sea star oocyte maturation time course extracts werefractionated by MonoQ column chromatography and used to phosphorylate substratepeptides P3 (AKRRRASLRA), Panel A or peptide P4 (AKRRRLSA), Panel B, in thepresence of [y-32P]ATP. The bar (------ ) indicates fractions containing immunoreactive protein. Time points tested are indicated in legend in panel A. GVBDoccurred at 90 minutes.201Piml-NT Pimi-Ill32.5 kDa —Figure 45. Comparison of fractionated sea star oocyte maturation timepoints.Various time points from an oocyte maturation time course were MonoQ-fractionated and electroluted on an SDS-PAGE gel. Western blots were probed withPim 1-NT and Pimi-ilI antibody. Panel A: immature oocytes; panel B, 15 mm;panel C, 30 mm; panel D, 45 mm; panel E, 60 mm; panel F, 90 mm; panel G, 120mm. MonoQ fraction numbers are indicated on the bottom of diagram. Migrationsof Mr standards are shown on the left.49.5 kDa-&22 23 24 25 26 27 28 22 23 24 25 26 27 28202and Pim i-TI seemed to detect a slightly higher molecular mass band of about 42 kDaprimarily in fractions 24 and 25. Even though the antibodies detected different proteins,the immunoreactivity of both antibodies corresponded with the minor, non-stimulated P4kinase activity peak in Mono Q fractions 24-25 (0.28-0.32 M NaC1).The peak immunoreactive fractions (24-26) from each time point were pooled andWestern blotted with Pim- 1 antibodies to determine if there are changes in the proteinamount or migration during the oocyte maturation (Fig. 46). The Pim-l antibodies detectedslightly different bands; Piml-NT detected a 40 kDa band strongly and a 42 kDa bandweakly, while Pim 1-ITT detected a 40 kDa band weakly and a 42 kDa band strongly. TheLilly antibody visualized only the 42 kDa Pim- 1 band. Although there is no easyexplanation for why these antibodies have different immunoreactivities, it is possible thatthere might be two related Pim protein products expressed in the oocyte. linportantly, therewas no increase in the amount or size of the immunoreactive proteins in the different timepoints of maturation.These results indicate that sea star Pim- 1 is present and apparently active in oocyteextracts, but the amount and the size of the protein does not change during the maturationprocess. There was no significant increase in kinase activity as indicated by the P4 peptidephosphorylation assay. Therefore, Pim-l does not appear to be regulated during meioticmaturation.20366kDa—lFigure46.Seastaroocytetimecourse:peakMonoOfractions.Westernblotsofpooledfractions(24-26) ofMonoQ-fractionatedseastaroocytematurationtimecourse.WesternblotsprobedwithPim-1antibodiesindicatedbelowblotsinboldtypeanddevelopedwithECL.Maturationtimepoints(mm)areshownbelowblots.MigrationsofMr standardsareshownonright.o15304560901200153045609012001530456090120Piml-NTPimi-IllLilly4. SUMIvIARYThe sea star oocyte system has potential for future studies with Pim- 1, because of highreproducibility of results and large quantities of biological starting materials. Although notenough of the sea star pim-1 sequence was obtained to warrant a thorough comparison tothe mammalian and amphibian sequences, it was enough to confirm that pim-1 mRNA wasexpressed in this system. A transcript of the expected size was detected on Northern blotsalong with two smaller, poly(A)- transcripts of 1.7 and 2.6 kb, possibly resulting fromaborted translation or alternative splicing of the mRNA transcript.Analysis of the sea star oocyte maturation time course with Pim- 1 antibodies andpeptides indicated that Pim-l eluted in fractions 24-25 (0.28-0.3 M NaC1), before the mainpeak of P4 kinase activity. The P4 phosphorylating activity of Pim- 1 did not changeduring maturation. The Pim- 1 antibodies immunodetected slightly different proteins from40-42 kDa, however none of these proteins underwent a change in quantity or size duringmaturation. The Pim-l protein seemed to disappear at 6 and 24 hours post-fertilization,reappear at 2 days and disappear again at 4 days post-fertilization. The fertilization studiesneed to be repeated to confirm these results.Preliminary work to identify Pim-l interacting proteins was promising, and has openedsome new areas for future study (Appendix VIII). GST-Pim-1 binding proteins in the seastar extracts were not identified by GST-Pim-l affinity columns and Far Western blotting.Pim-1 kinase (P1K) activity was detected in MonoQ fractions 28-32 (0.35-0.45 M NaC1).This P1K is completely uncharacterized, except for the fact that it undergoes an activationduring maturation, unlike Pim- 1.205PART 4 - DISCUSSION AND FUTURE DIRECTIONSDISCUSSIONThis study utilized a wide variety of recombinant DNA and biochemistry techniques toexamine and characterize the activity of the mammalian, amphibian and echinoderm Pim- 1serine/threonine protein kinases in vitro. Although the findings of this study were obtainedfrom work with expressed enzyme in vitro, these results can be extrapolated to endogenousPim- 1 and can assist in constructing a model of Pim- 1 activity in vivo.The Pim- 1 cDNA was cloned from X. laevis, and is the first non-mammalian pim-1homologue characterized. The eDNA open reading frame encoded a 35-kDa proteincontaining all the conserved serine/threonine kinase subdomains. The X. laevis Pim-1 hasstrong sequence homology to the mammalian Pim-1 cognates, implying that the functionsof these kinases are highly conserved in all species. The catalytic domain of the sea starPim- 1 was also found to show high homology to Pim- 1 of other species, in keeping withthis hypothesis.Although the Pim-1 family members are closely related to each other, they have fewimmediate relatives. Homology searches identified several proteins sharing some similarityto Pim- 1 in the catalytic domain, all of which were serine/threonine kinases. The pim-1nucleotide sequence is homologous to calcium/calmodulin-dependent protein kinases andphosphorylase b in kinase subdomains V and VI. Pim-l protein displays homology in thecatalytic subdomains with cAMP-dependent protein kinases, calcium/calmodulin-dependentprotein kinases, phosphorylase kinases, yeast kinases involved in metabolic control as wellas many yet uncharacterized kinases including human p78, the ASFV and tsk-l proteins.The structural homology between Pim- 1 and these other kinases was high in thecatalytic domain, implying that all these kinases may have functional homology. Indeed,the consensus phosphorylation site sequence that we deduced for Pim- 1 (K/R-K/R-RK/RXS*/T*X’) resembles the phosphorylation site motifs found for many of thesecond messenger-dependent protein kinases that have been analyzed [Kennelly andKrebs, 1991]. These include the consensus sequences of calcium/calmodulin dependentprotein kinase II (CaM1I) (RXXS*/T*), cAMP-dependent protein kinase (R-RIK-XS*IT*) and cGMP dependent protein kinase (RJK2..3XS*/T*). Similarly, thephosphorylation sites recognized by protein kinase C, phosphorylase kinase and the S6206kinase Rsk also feature basic residues on the N-terminal side of the phosphoacceptorresidue [Kennelly and Krebs, 1991]. Nonetheless, while Kemptide (LRRASLG) is anexcellent substrate for cAMP-dependent protein kinase and Rsk, but it was a very poorsubstrate for GST-Pim-1. Thus, while there is partial overlap in the substraterequirements of these kinases, they are also distinct.To study the activity of Pim- 1, a concerted effort was made to immunoprecipitate theenzyme from various cell types. However, this did not yield protein with sufficientactivity, purity or quantity for kinetic studies. As well, the native Pim- 1 has not yet beenpurified to homogeneity. Both Xenopus laevis and human pim-1 were expressed asbacterial fusion proteins with glutathione S-transferase. Both species were constitutivelyactive and both proteins had auto-phosphotransferase activity and phosphorylated a wideselection of exogenous substrates. Because the substrate specificity of eukaryoticserine/threonine-specific protein kinases is largely determined by the -250 amino acidcatalytic domain [Hanks et al., 1988], the GST protein fused to the amino terminus of thePim- 1 should not have influenced the substrate specificity of the kinase. There is nopublished evidence for the substrate specificity of a kinase being detenuined by regionsoutside of this catalytic domain. However, the catalytic efficiency of protein kinases isoften regulated by regions outside of the catalytic domain.The phosphorylation site consensus sequence for Pim- 1 varied slightly between themammalian and amphibian species. The substrate binding domain is thought to be locatedin the 60 carboxy-terminal amino acids of the catalytic domain, in kinase subdomains (IVVI) [Kemp and Pearson, 1990]. As single residues can influence the substrate specificities[Knighton et al., 1991], differences of one or more amino acids within these domainsbetween species could account for minor changes in the substrate consensus sequence for akinase. This variation in the substrate specificity between species may allow a difference infunction of Pim- 1 between the two species. Alternatively, the differences observedbetween X. laevis and H. sapiens Pim- 1 may merely indicate that some of the residuessurrounding the phosphoacceptor site are not critical for substrate specificity.Potential physiological substrates containing the K/R - K/R - R - K/R - X - S*/T* - X’motif for phosphorylation by Pim- 1 were not immediately obvious. Many of the proteinsthat featured the Pim- 1 phosphorylation site consensus sequence are enzymes or structuralproteins, some of which are functional in the cytoplasm. Nuclear localization signals,(NLS) used as targeting signals for nuclear proteins, consist of stretches of basic residues207[Srinivasan et al., 1994]. As well, stretches of basic residues form part of DNA bindingdomains [Hill and Treisman, 1995]. Regulation of transcription factors is achieved byphosphorylation, which seems to encourage dimerization, nuclear localization and modulateDNA binding activity [Schindler, 1995]. Some of these NLS or DNA binding domains arefollowed by serine residues, and conform to the Pim- 1 phosphorylation sequence. Forexample, the Calmodulin kinase II a8 subunit from rat heart, contains an 11 amino acidinsert between the catalytic/regulatory and association domains (VKKRKSSSS), whichconforms to the Pim-1 phosphorylation consensus site [Srinivasan et a!., 1994].Phosphorylation of this sequence is believed to modify nuclear targeting. Similarly, theSV4O T antigen is phosphorylated near the NLS in order to enter the nucleus [Rihs et a!.,1991]. In addition, it has been shown that serine phosphorylation of Stat3 by a yetuncharacterized kinase, is necessary for Stat3-Stat3 dimerization and optimal DNA binding[Zhang et al., 1995]. Perhaps Pim- 1 functions “late” in signal transduction, regulatingtranscription factors or other nuclear proteins by the serine phosphorylation of NLS,dimerization or DNA binding domains, to mediate the entry into the nucleus, to promotedimerization or to prevent DNA binding.Substrate mixing experiments identified many proteins that either phosphorylated Pim-1or were phosphorylated by Pim-1. Although it is intriguing to speculate that theseinteractions may have some physiological significance, these reactions involving with mMamounts of proteins may not resemble the in vivo situation.The main site of autophosphorylation of Pim- 1 was identified using IMAC-HPLC-ESIMS, in collaboration with Drs. Lawrence Amanakwa and Michael Affolter, as Ser-190.The Ser-190 as well as the Ser-4 and Thr-205 residues are conserved in all Pim-1homologues, suggesting that they may represent physiological sites ofautophosphorylation. Many other kinases were identified that contained conservedresidues in these locations, including many of yeast kinases involved in cellularmetabolism. Most of the cyclin-dependent kinases (CDKs) and cdc homologues contain athreonine residue in catalytic subdomain VIII, which is in the T loop, phosphorylation ofwhich is essential for kinase activity [Morgan, 1995; Pines 1995]. Pim-1 Thr-205 islocated in domain VIII, and although it does not seem to align exactly with the Thrl6OIThr-161 residues of CDK2/cdc-2, it may also be have a similar function, to stabilizethe structure of the enzyme. Other kinases having activating phosphorylation sites in thisdomain include MAP kinase (Thr-183, Tyr-185), src (Tyr-416) and cAMP-dependentprotein kinase (Thr-197) [Veron et a!., 1994].208The Mos serine/threonine protein kinase also contains a major phosphorylation siteroughly homologous to the Pim-l Ser-4 site in the N-terminus, at Ser-3 [Nishizawa eta!.,1992; Freeman et at., 1992]. The importance of this site is disputed; one group hasdetermined that phosphorylation of this site is necessary for metabolic stability and for fullphysiological activity during the cell cycle [Nishizawa et at., 1992]. The phosphorylationis the result of an autokinase event and stabilizes the kinase by preventing ubiquitination ofthe kinase. In contrast, a second group has determined that the Ser-3 site isphosphorylated by an unidentified protein kinase, and that phosphorylation of this site isnot essential for CSF activity or to induce oocyte maturation [Freeman et at., 1992]. Thereasons for these discrepancies are not known, but when resolved may provide a modelfor Pim- 1 regulation by phosphorylation/autophosphorylation.In the search for tyrosine phosphorylation sites, the site corresponding to the Src Tyr416 autophosphorylation site, i.e. Tyr-199 [Cooper and MacAuley, 1988], was notidentified as a site of tyrosine phosphorylation. That we did not detectautophosphorylation of this site does not preclude that it may be an important location ofphosphorylation by other kinases. It just means that by our methods, we could notidentify it as a phosphorylated residue.Although the phosphoamino acid analysis indicated that GST-Pim-1 autophosphorylatedon tyrosine residues, we were unable to unequivocally identify a tyrosine phosphorylationsite by IMAC-HPLC-ESI-MS. The presence of tyrosine autophosphorylation sites wereexamined using the novel approach of identifying peaks that were shifted in mass andretention time after treatment with the tyrosine-specific phosphatase HPTPB. Although wewere able to observe peptides that shifted in LCMS retention time after HPTPB treatment,these peptides were difficult to relate to the Pim-1. A tentative site, Tyr-133, wassuggested by this method, but we were unable to confirm this result. This site wasconserved only between X. laevis Pim- 1 and murine Pim-2 proteins, and therefore wasunlikely to be a physiologically important site for the modulation of Pim- 1 activity. Forthis reason, this site was not explored further by mutational analysis. Interestingly, someof the kinases with homologous residues to the Ser-190 site were found to also exhibit dualkinase activity including Cik, ASK-a/y, and MSDI [Johnson and Smith, 1991; Bianchi etat., 1994; Puziss et at., 1994]. It is not known if the apparent dual specificity of thesekinases is physiological or is an effect of the particular expression systems used.209The importance of the Ser-190 site was investigated using mutants generated by PCRsite directed mutagenesis and expressed in bacteria. Reduced autophosphorylation of themutants confirm that the Ser- 190 was the main autophosphorylation site. Both the S 190>Aand S 190>E mutants had a lower phosphotransferase activity (Vmax) towards exogenoussubstrates than the WT GST-Pim- 1, suggesting that a change in the structure of the proteinhas occurred, inhibiting the activity of the kinase. The S 190>A mutant has a much higheraffinity (Km) for the peptide substrate than the WT GST-Pim- 1 implying that a nonphosphorylated residue at this site allows tighter binding of the peptide to the enzyme. Thecharged glutamic acid of the Si 90>E mutant partially mimics a phosphorylation event, andthe affinity of the Si 90>E mutant and the WT GST-Pim- 1 are identical. This suggests thatautophosphorylation of the GST-Pim-1 on Ser-190 is activating. That the Ser-190 site was50-iOO% phosphorylated by the time that active GST-Pim-i was purified from the bacteria,supports this hypothesis. Additional methods are probably also used to control the activityof the kinase posttranlationally.Despite the changes in kinase activity obtained with the Si90 mutants, the activity ofthe expressed WT GST-Pim-i did not appear to be modulated by in vitroautophosphorylation, dephosphorylation or by phosphorylation by other purified kinases.It is very likely that autophosphorylation is not the sole method of posttranscriptionalregulation of Pim- 1 activity. Examination of the regulation of other small serine/threoninekinases may provide some insight into the methods by which the activity of Pim- 1 ismodulated. Many other kinases in the cell are part of dynamic, multiprotein complexes, theconstituents changing as the regulatory or localization requirements of the kinase changes.For example, the activity of CDKs are controlled by four highly conserved biochemicalmechanisms: activation by cyclin binding, activation by phosphorylation of the Thr-160residue, deactivation by phosphorylation of residues Thr-14 and Tyr-15 within kinasesubdomain I, and deactivation by the binding of cyclin kinase inhibitory domains [Pines,1995; Morgan, 1995]. The cAMP-dependent protein kinases have a cAMP-bindingregulatory domain that binds and represses the catalytic domain. Liberation of the activecatalytic domain occurs when two cAMP molecules bind and induce a steric change,releasing the regulatory domain. In addition to regulatory domains, catalytic domains alsohave targeting domains which direct and localize the activity of the kinase catalytic domain[Hubbard and Cohen, 1993]. An example of a protein employing the target signalhypothesis method of regulation is CaM1I, where each 54-60 kDa isoform contains a Nterminal catalytic domain, a regulatory domain containing a kinase autoinhibitory segment2i0and an association domain [Srinivasan, 1994]. The subunits, associate to from large 500-600 kDa multimeric complexes composed of 6-12 subunits.It is possible that Pim- 1 is also a component of a larger protein complex. Unlike theCaM11 isoforms, Pim- 1 consists of a catalytic domain without characterized autoregulatoryor association domains to allow interaction with other kinases. Our results with inhibitorpeptides indicate that Pim-1 was not inhibited by sequences within the kinase. However,the activity of Pim- 1 might be inhibited by pseudosubstrates located on associatedregulatory proteins. Pim-1 must associate with other kinases by virtue of stilluncharacterized binding domains in the N- or C-terminus. Although Pim-1 does not haveany of the known protein modules essential for protein-protein interaction such as the Srchomology domains (SH2 and SH3), the Pleckstrin (PH) domain, PTB, LIM Armadillo andthe Notch/ankyrin repeat, this does not preclude the possibility that other proteinscontaining these domains may bind to Pim-1 [Pawson, 1995; Cohen et al. 1995; van derGeer and Pawson, 1995]. An imperfect amphipathic a-helix motif in the C-terminus ofPim-1, similar to that of cAMP-dependent protein kinase [Veron et al., 1994], as of yethas no defined function, and may be involved in an interaction with associated proteins.Pim-1 activity may be regulated in a manner similar to that of second-messenger kinases,with binding to regulatory proteins causing steric alteration or blocking of active sites.Although there is little to suggest that Pim-1 shares any structural or functionalsimilarity with other oncogene-encoded serine/threonine kinases such as Mos and Raf,Pim- 1 shares homology and may be part of the same family as some of the newly identifiedprotein kinases. Some of the more interesting serine/threonine kinases with homology toPim-1 include the human p78 protein kinase that is lost during chemically-inducedpancreatic tumors, the product of the recently characterized African swine fever virus(ASFV) and tsk-1, a testes-specific murine kinase. The product of the ASFV is similar insize to the Pim-1 (299 residues) and contains residues homologous to theautophosphorylation sites of Pim-1 [Baylis et al., 1993]. The ASFV kinasephosphorylated histones but not BSA, casein, phosvitin or protamine, similar to the Pim-1.The murine testes-specific kinase tsk- 1 is also similar in size to Pim- 1 (364 residues) andhas a conserved serine residue at the location homologous to T205 [Bielke et a!., 1994].Both of these kinases are similar to Pim- 1 in that they are comprised of a catalytic domainwithout a known targeting domain or an obvious autoregulatory domain. It is likely thatthese kinases, along with Pim- 1 may form a new family of protein kinases that interact with211other regulatory or targeting proteins. Many of the yeast kinases identified in homologysearches are also of similar sizes and may be part of a homologous protein complexes.Pim- 1 is thought to participate in signal transduction pathways induced by WN- Epo,IL-3, IL-5 and GM-CSF, and requires an intact receptor membrane proximal domain forupregulation [Polotskaya et a!., 1993]. Also stimulated by the same growth factors areJAK2 and the STAT proteins [Hill and Treisman, 1995; Schindler, 1995]. The STATproteins are dimeric DNA binding proteins that respond to cytokines and growth factors toinduce cytokine and growth factor inducible genes. They exist in the cytoplasm asunphosphorylated monomers, but after phosphorylation, dimerize and translocate to thenucleus [Cohen et a!., 1995; Schindler, 1995]. Although it has been recently beendemonstrated that the pim-1 gene contains a STAT binding site, it is possible that Pim-1functions to in turn phosphorylate STAT or other STAT-like transcription proteins in thecell.Most work in other laboratories has focused on characterizing the expression of thePim-1 mRNA transcript or the upregulation of the Pim-1 protein in cytokine-stimulatedhemopoietic cells. This study explored the expression and activity of the Pim- 1 protein inthe maturing oocyte system. Although the meiotic cell cycle of the maturing oocyte doesnot have an S phase and has shortened Gl and G2 phases, many of the same kinases thatare active during the mitotic cell cycle become active during the meiotic cycle. It was hopedthat by defining the activity and expression patterns of Pim- 1 in the oocyte, that the resultscould be applied to the activity and expression of Pim- 1 during the cell cycle in general.Despite many promising results, the studies of Pim- 1 in the oocyte model systems weresomewhat inconclusive. Immunoreactive Pim- 1 displayed phosphotransferase activitytowards P4 peptide in sea star oocytes, but it did not become activated, and the quantity didnot change during oocyte maturation. This method for examining the activity ofendogenous Pim- 1, involving the analysis of fractionated extracts, is extremely limitedwhen the previous discussion is taken into account. Although Pim-1 is present duringoocyte maturation, the enzyme may be sequestered in a non-active state in multi-proteincomplexes with changes in the tightly regulated activity too insignificant or transient to bedetected by these methods. Although there were changes observed in the quantity of Pim- 1protein after oocyte fertilization, these changes were not thoroughly investigated.212FUTURE DIRECTIONSThis study has advanced the knowledge of the in vitro activity of expressed GST-Pim- 1and in the process has developed many tools and reagents for the further investigation ofthis kinase in endogenous systems. Many unanswered questions were raised, which form anew platform for future study. The most important questions remaining to be answeredare: (1) What are the physiological substrates of Pim-l? (2) What are theregulatory/targeting proteins which associate with Pim- 1 in vivo? (3) What is the functionof Pim- 1? and (4) What is the importance of the other autophosphorylation sites? Toanswer these remaining questions, some of the following directions could be considered.1. Identification of proteins interacting/regulating/phosphorvlated by Pim-1The small size and the constitutively active state of Pim- 1 suggest that other proteinsinteract with the kinase to regulate and target the activity of the kinase. Several approachescan be utilized to identify these associated proteins:i. The yeast two hybrid systemAs Pim- 1 is a relatively small kinase composed of a catalytic domain with very short N-and C- terminal regions, it is highly likely that other proteins interact with Pim- 1 to regulateor modulate the activity of the kinase. Preliminary work to screen a HeLa library with LexA human pim-1 was initiated by Mr. Michael Kyba (Biotech Lab, UBC). Initial resultsfrom his laboratory and others indicate that Pim- 1 non-specifically causes indiscriminantactivation of transcription, and removal of acidic domains of the Pim- 1 did not change thisactivity. This implies that Pim- 1 either acts as or interacts with a transcription factor toshort-circuit the 2-hybrid selection system, or that the Pim-1 phosphorylates a protein thatmay activate transcription of the markers in this system. Screening the library with a LexAKD pim-] would determine if this effect is dependent on kinase activity of the Pim- 1 or isstrictly due to a protein/protein interaction.ii. Far Western blots and fusion protein affinity columnsPreliminary work to test new methods of identification of protein-protein interactionswere promising. Both the Far Western and fusion protein affinity column methodsdemonstrated some binding of X. laevis proteins to GST-Pim-1 and some phosphorylationof X. laevis proteins by GST-Pim- 1. The identities of these associated proteins need to beinvestigated further. These methods would be best applied for confirmation when aninteraction between Pim- 1 and another protein is suspected.213iii. Purification of Pim- 1 from endogenous sources.Purification of Pim-1 from endogenous sources would not only provide material withwhich to confirm the results of the in vitro activity characterizations, but might allowcopurification and identification of associated proteins. Sea star oocytes would be a goodsource of endogenous Pim- 1 for purification, using the peptide reagents and antibodiesdeveloped in this study. Although it does not experience further stimulation duringmaturation, Pim- 1 appeared to be active in the sea star oocyte, based on P4 peptidephosphorylation.Initial characterization of bovine spleen extracts indicates that Pim- 1 is active andpresent in ample quantities to justify purification. Unlike many of the other Pim- 1expressing cell lines, large volumes of starting material can be inexpensively and easilyobtained. One disadvantage of using BSE is that the extracts contain many different celltypes including various different spleen cells and circulating hemopoietic cells. Not allthese different cell types will contain the active Pim- 1 protein; analysis of column fractionswith the P4 peptide should select for activated Pim- 1 species.iv. Identification of upstream activators of Pim- 1 using sea star oocvte extractsThe kinase-dead X. laevis GST-Pim-1 could be used to screen fractions for Pim-1kinase (P1K) activity, allowing potential upstream activators of Pim- 1 to be purified andidentified. This could be performed as in Appendix Vifi, or the expressed KD GST-Pim-1could be used as a substrate in an in gel assay.2. Examining in vivo effects of Pim-1 in the oocvte system by microinjectionMicroinjection studies using WT and mutant Pim-1 could be performed to determine thefunction of Pim- 1 in vivo. Although the putative Pim- 1 activity does not seem to increaseduring sea star oocyte maturation, the fact that some activity was observed implies that thiskinase has some role in the oocyte. Microinjection studies could be performed with bothX. laevis and sea star oocytes. However, as sea star oocytes are easier to obtain, aretransparent and yield more consistent results, the sea star oocyte system is recommended.Microinjections of many reagents, including expressed protein, expressed mutant proteins,antibodies and oligonucleotides, could be performed to examine effects in the oocyte.214i. Pim-1 ablationsMicroinjection of antibodies directed against a specific protein has been shown toinactivate or neutralize the activity of the protein, i.e. Mos. As the GXP antibodyrecognizes and inhibits the kinase activity of Pim-1, microinjection of this antibody intooocytes at different stages of maturation or fertilization may allow assessment of Pim- 1ablation in immature oocytes, during maturation and at different stages post-fertilization.An advantage of using embryos, is that just one cell in a two cell blastocyst can be injected,and the effects compared to the non-injected cell. As a negative control, some of the serumthat had been depleted of anti-Pim-1 antibodies could be microinjected.Pim-1 ablation could also be achieved through the microinjection of specific pim-1 anti-sense oligonucleotides. These oligonucleotides could bind the pim-1 mRNA, blockingprotein translation. This approach may be more successful over longer term experiments(during embryo development) where new Pim-l protein would be translated.ii. KD Pim-1 dominant-negative effectsMicroinjection of KD Pim- 1 into oocytes in situations identical to that of the GXPantibody will allow determination of whether or not the KD pim-1 can exert a dominant-negative effect on oocyte maturation or embryonic development. The results of theseexperiments would be expected to be similar to that of the Pim- 1 ablation by antibodyinjection.iii. Effects of WT and mutant Pim- 1 over-expressionSimilar to the experiments described above, expressed active or mutated GST-Pim- 1could be microinjected into oocytes to determine the effects of Pim-1 over-expression inimmature, maturing and post-fertilized oocytes and developing embryos. Variousautophosphorylation site mutants (i.e. S 190>A) could also be microinjected to assess theeffects of these mutants in the in vivo situation.3. Mutation of other autophosphorviation sites of X. laevis Pim- 1The Thr-205 and Tyr-133 phosphorylation sites could be altered by PCR-mutagenesis,and the activity of the expressed mutants assessed to determine the importance of these sitesfor regulation of the kinase. Pim-1 clones could be constructed to have mutations at 2 ormore of these autophosphorylation sites (i.e. both Ser-190 and Thr-205) and the combinedeffects on expressed Pim- 1 activity assessed.215To determine if the autophosphorylation sites mediate protein-protein interactions, theaforementioned approaches (yeast 2-hybrid or affinity columns) could be used with themutant Pim- 1. Results could be compared to those obtained with the wild-type Pim- 1, toassess the effects of the mutation on protein-protein interaction.4. Further studies of Pim- 1 in the sea star oocyte systemThe studies undertaken with Pim-l from oocytes were productive, but not conclusive.Additional experiments need to be performed in order to more fully examine the activity ofthe endogenous kinase in the oocyte system.i. Further examination of the oocyte-specific pim-1 RNA transcripts.These alternately sized RNA transcripts (2.6 and 1.7 kb) need to be more completelycharacterized to determine if they are pim-1, and if so, what they represent (i.e. are they aresult of alternative mRNA splicing?). What does the presence of these transcripts indicateabout the regulation and expression ofpim-1 in the oocyte system?ii. Cloning of the sea star eDNAThe sea star-specific cDNA probe could be used to screen a sea star cDNA libraryallowing the pim-1 to be cloned from this system. Experiments such as GST-Pim-1affinity chromatography, Far Western blotting and microinjection could then be performedusing the kinase from sea star.iii. Fertilization studies of sea star oocytesInitial results indicate that the amount of Pim- 1 changes after oocyte fertilization.Further studies could be done in fertilized oocytes to determine if the activity of the kinasechanges during oocyte fertilization and embryo development. Of interest, the activity ofanother oncogene-encoded serine-threonine kinase, Raf, is essential for mesodermformation in developing embryo. Pim- 1 may have a similar role.iv. Localization of Pim- 1 activityIf more specific Pim- 1 antibodies were developed, immunofluorescent studies could beundertaken to determine the exact location of the Pim- 1 in the oocyte. Although Pim- 1 hasbeen localized to the cytoplasm in the cell types examined, it would be of particular interestto examine if the Pim-1 was located at the endoplasmic reticulum to phosphorylate newproteins, or existed as part of a nuclear transport complex, modulating the transport ofproteins to the nucleus.216BIBLIOGRAPHYActon, D., Domen, J., Jacobs, H., Vlaar, M., Korsmeyer, S. and Berns, A. (1992)Current Topics in Microbiology and Immunology 182:293-298. Collaboration of PllvI-1 andBCL-2 in lymphomagenesis.Alcorta, D.A., Crews, C.M., Sweet, L.J., Bankston, L., Jones, S.W. and Erikson, R.L.(1989) Mol. Cell. Biol. 9:3850-3859. Sequence and expression of chicken and mouse rsk:homologues of Xenopus laevis ribosomal S6 kinase.Amson, R., Sigaux, F., Przedborski, S., Flandrin, G., Givol, D. and Telerman, A. (1989)Proc. Nati. Acad. Sci. U.S.A. 86: 8857-8861. The human protooncogene product p33P1mis expressed during fetal hematopoiesis and in diverse leukemias.Aoki, M., Akiyama, T., Miyoshi, J. and Toyoshima K. (1991) Oncogene 6:1515-1519.Identification and characterization of protein products of the cot oncogene with serinekinase activity.Armstrong, M.J. and Galloway, S.M. (1993) Mutation Research 302:61-70. Micronucleiinduced in peripheral blood of Ep-PIM- 1 trangenic mice by chronic oral treatment with 2-acetylaminofluorene or benzene but not with diethyl-nitrosamine or 1 ,2-dichloroethane.Bairoch, A. and Claverie, J.-M. (1988) Nature 331:22. Sequence patterns in proteinkinases.Bianchi, M.W., Guivarc’h, D., Thomas, M., Woodgett, J.R. and Kreis, M. (1994) Mol.Gen. Genet. 242:337-345. Arabidopsis homologs of the shaggy and GSK-3 proteinkinases: miecular cloning and functional expressionin Escherichia coli.Baylis, S.A., Banham, A.H., Vydelingum, S., Dixon, L.K. and Smith, G.L. (1993) J.Virology 67:4549-4556. African swine fever virus encodes a serine protein kinase which ispackaged into virions.Beck, S. and Barrell, B.G. (1988) Nature 331:269-272. Human cytomegalovirus encodesa glycoprotein homologous to MHC class-I antigen.Bellacosa, A., Testa, J.R., Staal, S.P. and Tsichlis, P.N. (1991) Science 254:274-277. Aretroviral oncogene, akt, encoding a serine-threonine kinase containing an SH2-like region.Belle, R., Mulner-Lorillon, 0., Marot, J. and Ozon, R. (1986) Cell Differentiation 19:253-261. A possible role for Mg2 ions in the induction of meiotic maturation of Xenopusoocytes.Ben-David, Y., Letwin, K., Tannock, L., Bernstein, A. and Pawson, T. (1991) EMBO J.10:317-325. A mammalian protein kinase with potential for serine/threonine and tyrosinephosphorylation is related to cell cycle regulators.Berns, A. (1991) J. Cell. Biochem. 47:130-135. Tumorigenesis in transgenic mice:identification and characterization of synergizing oncogenes.Bems, A., Selten, 0., Cuypers, H.T. and Domen, J. The pim-1 oncogene, pp 121-134,The Oncogene Handbook (1988a) Elsevier Science Publishers. Eds: E.P. Reddy, A.M.Skalka and T. Curran.217Berns, A. (1988b) Arch Virol 102:1-18. Provirus tagging as an instrument to identifyoncogenes and to establish synergism between oncogenes.Bielke, W., Blaschke, R.J., Miescher, G.C., Zurcher, G., Andres, A.-C. and Ziemiecki,A. (1994) Gene 139:235-239. Characterization of a novel murine testes-specificserine/threonine kinase.Blacketer, M.J., Koehler, C.M., Coats, S.G., Myers, A.M. and Madaule, P. (1993) Mol.Cell. Biol. 13:5567-5581. Regulation of Dimorphism in Saccharomyces cerevisiae:involvement of the novel protein kinase homolog Elmip and protein phosphatase 2A.Bradford, M.M. (1976) Anal. Biochem. 72:248-254. A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing the principle of protein-dye binding.Breuer, M., Wientjens, E., Verbeek, S., Siebos, R. and Berns, A. (1991) Cancer Research5 1:958-963. Carcinogen-induced lymphomagenesis in pim-1 transgenic mice: dosedependence and involvement of myc and ras.Breuer, M.,Cuypers, H.T. and Berns, A. (1989a) EMBO J. 8:743-747. Evidence for theinvolvement ofpim-2, a new common proviral insertion site, in progression of lymphomas.Breuer, M., Siebos, R., Verbeek, S., van Lohuizen, M., Wientjens, B. and Berns, A.(1989b) Nature 340:61-63. Very high frequency of lymphoma induction by a chemicalcarcinogen in pim-1 transgenic mice.Brook, J.D., McCurrach, M.E., Harley, H.G., Buckler, A.J., Church, D., Aburatani, H.,Hunter, K., Stanton, V.P., Thirion, J.-P., Hudon, T., Sohn, R., Zemelman, B., Snell,R.G., Rundle, S.A., Crow, S., Davies, J., Shelbourne, P., Buxton, J., Jones, C.,Juvonen, V., Johnson, K., Harper, P.S., Shaw, D.J. and Housman, D.E. (1992) Cell68:799-808. Molecular basis of myotonic dystrophy: expansion of a trinucleotide (CTG)repeat at the 3’ end of a transcript encoding a protein kinase family member.Celenza, J.L. and Carlson, M. (1986) Science 233: 1175-1180. A yeast gene that isessential for release from glucose respression encodes a protein kinase.Cheng, H-C., Van Patten, S.M., Smith, A.J. and Walsh, D.A. (1985) Biochem J.23 1:655-661. An active twenty-amino-acid-residue peptide derived from the inhibitorprotein of the cyclic AMP-dependent protein kinase.Cohen, G.B., Ren, R. and Baltimore, D. (1995) Cell 80:237-248. Modular bindingdomains in signal transduction proteins.Cooper, J.A. and MacAuley, A. (1988) Proc. Natl. Acad. Sci. USA 85:4232-4236.Potential positive and negative autoregulation of p60CSrC by intermolecularautophosphorylation.Cooper, J.A., Esch, F.S., Taylor, S.S. and Hunter, T. (1984) J. Biol. Chem. 259: 7835-7841. Phosphorylation sites in enolase and lactate dehydrogenase utilized by tyrosineprotein kinases in vivo and in vitro.Craig, E.S. and Gross, C.A. (1991) TiES 16:135-140. Is hsp70 the cellular thermometer?218Cuypers, H.T., Selten, G., Quint, W., Zijistra, M., Robanus-Maandag, E., Boelens, W.,van Wezenbeek, P., Melief, C. and Berns, A. (1984) Cell 37:141-150. Murine leukemiavirus-induced T-cell lymphomagenesis: integration of proviruses in a distinct chromosomalregion.Cuypers, H.T., Selten, G. and Geurts van Kessel, A.H.M. (1986) Hum Genet. 72: 262-265. Assignment of the human homologue of Pim-1, a mouse gene implicated inleukemogenesis, to the pter-q12 region of chromosome 6.Dailey, D., Schieven, G. L., Lim, M.Y., Marguardt, H., Gilmore, T., Thorner, J. andMartin, G.S. (1990) Mo!. Cell. Biol. 10:6244-6256. Novel yeast protein kinase (YPKJgene product) is a 40-kilodalton phosphotyrosy! protein associated with protein-tyrosinekinase activity.Dautry, F., Weil, D., Yu, J. and Dautry-Varsat, A. (1988) J. Biol. Chem. 263: 17615-17620. Regulation ofpim and myb mRNA accumulation by interleukin-2 and interleukin-3in murine hematopoietic cell lines.Domen, J., van der Lugt, N.M.T., Laird, P.W., Saris, C.J.M., Clarke, A.R., Hooper,M.L. and Berns, A. (1993a) Blood 82: 1445-1452. Impaired interleukin-3 response inPim-1 deficient bone marrow-derived mast cells.Domen, J., van der Lugt, N.M.T., Acton, D., Laird, P.W., Linders, K. and Berns, A.(1993b) 3. Exp. Med 178:1665-1673. Pim-] levels determine the size or early B lymphoidcompartments in bone marrow.Domen, 3., van der Lugt, N.M.T., Laird, P.W., Saris, C.J.M. and Berns, A. (1993c)Leukemia 7, Suppl 2: S108-S1 12. Analysis of Pim-1 function in mutant mice.Domen, J., Von Lindern, M., Hermans, A., Breuer, M., Grosveld, G. and Berns, A.(1987) Oncogene Research 1:103-112. Comparison of the human and mouse pim- 1 cDNAs:Nucleotide sequence and immunological identification of the in vitro synthesized pim-1protein.Dreyfus, F., Sola, B., Fichelson, S., Varlet, P., Charon, M., Tambourin, P., Wendling,F. and Gisselbrecht, S. (1990) Leukemia 4:590-594. Rearrangements of the Pim-1, c-mycand p53 genes in Friend helper virus-induced mouse erythroleukemias.Felsenstein, J. (1993) PHYLIP (Phylogeny Inference Package) version 3.5c. Distributedby the author. Department of Genetics, University of Washington, Seattle.Fotino, M., Merson, E.J. and Allen, F.H. (1971) Ann. Clin. Lab. Sci. 1:131-133.Micromethod for rapid separation of lymphocytes from peripheral blood.Freeman, R.S., Meyer, A.N., Li, J. and Donoghue, D.J. (1992) J. Cell. Biol. 116:725-735. Phosphorylation of conserved serine residues does not regualte the ability of mosprotein kinase to induce oocyte maturation of function as cytosolic factor.Friedmann, M., Nissen, M.S., Hoover, D.S., Reeves, R. and Magnuson, N.S. (1992)Arch. Biochem. Biophys. 298:594-601. Characterization of the proto-oncogene Pim-1:kinase activity and substrate recognition sequence.219Garret, S. and Broach, J. (1989) Genes and Development 3:1336-1348. Loss of Rasactivity in Saccharomyces cerevisiase is suppressed by disruptions of a new kinase gene,YAK 1, whose product may act downstream of the cAMP-dependent protein kinase.Grand, R.J.A. (1989) Biochem. J. 258:625-638. Acylation of viral and eukaryoticproteins.Gronwald, R.G.K., Grant, F.J., Haldeman, B.A., Hart, C.E., O’Hara, P.J., Hagen,F.S., Ross, R., Bowen-Pope, D.F. and Murray, M.J. (1988) Proc. Natl. Acad. Sci. USA85:3435-3439. Cloning and expression of a eDNA coding for the human platelet-derivedgrowth factor receptor: evidiense for more than one receptor class.Graf, T. and Beug, H. (1983) Cell. 34:7-9. Role of the v-erbA and v-erbB oncogenes ofAvian Erythroblastosis Virus in erythroid cell transformation.Ranks, S.K., Quinn, A.M. and Hunter, T. (1988) Science 241:42-52. The protein kinasefamily; conserved features and deduced phylogeny of the catalytic domains.Harder, K.W., Owen, P., Wong, L.K.H., Aebersold, R., Clark-Lewis, I. and Jirik, F.R.(1994) Biochem J. 298:395-401. Characterization and kinetic analysis of the intracellulardomain of human protein tyrosine phosphatase B (HPTPB) using syntheticphosphopeptides.Raupt, Y., Harris, A.W. and Adams, J.M. (1993) mt. j. Cancer 55:623-629. MoloneyVirus induction of T-cell lymphomas in a plasmacytomagenic strain of Et-V-Abl transgenicmice.Haupt, Y., Alexander, W.S., Barn, G., Klinken, S.P. and Adams, J.M. (1991) Cell65:753-763. Novel zinc finger gene implicated as myc collaborator by retrovirallyaccelerated lymphomagenesis in Ej.t-myc transgenic mice.He, M., Wilde, A. and Kaderbhai, M.A. (1989) Nucleic Acids Res. 18: 1660. A simplesingle-step procedure for small scale preparation of Escherichia coli plasmids.Heinze, H., Arnold, H.H., Fischer, D. and Kruppa, J. (1988) J. Biol. Chem. 263:4139-4144. The primary structure of the human ribosomal protein S6 derived from a clonedeDNA.Hill, C.S. and Treisman, R. (1995) Cell 80:199-211. Transcriptional regulation byextracellular signals: mechanisms and specificity.Hoover, D.S., Wingette, D.G., Reeves, R. and Magnuson, N.S. (1995) J. Cell. Biochem.S19A:A1-232. Translational repression: a strategy for cellular control of Pim-1 proteinlevels. (Keystone symposia on Molecular and Cellular Biology).Hoover, D., Friedmann, M., Reeves, R. and Magnuson, N.S. (1991) J. Biol. Chem. 266:14018-14023. Recombinant human Pim- 1 protein exhibits senine/threonine kinase activity.Hubbard, M.J. and Cohen, P. (1993) Trends Biochem. Sci. 18:172-177. On target with anew mechanism for the regulation of protein phosphorylation.Hunter, T. (1987) Cell 49:1-4. A tail of two src’s: mutatis mutandis.220Hunter, T. and Karin, M. (1992) Cell 70:375-387. The regulation of transcription byphosphorylation.James, P., Vorherr, T. and Carafoli, E. (1995) Trends. Biochem. Sci. 20:38-42.Calmodulin-binding domains: just two faced or multi-faceted?Johnson, K.W. and Smith, K.A. (1991) J. Biol. Chem. 266:3402-3407. Molecularcloning of a novel human cdc2/CDC28-Iike protein kinase.Kaelin, W.G., Krek, W., Sellers, W.R., DeCaprio, J.A., Ajchenbaum, F., Fuchs, C.S.,Chittenden, T., Li, Y., Farnham, P.J., Blanar, M.A., Livingston, D.M. and Flemington,E.K. (1992) Cell 70: 351-364. Expression cloning of a cDNA encoding a retinoblastomabinding protein with E2F-like properties.Kemp, B.E. and Pearson, R.B. (1990) Trends Biochem. Sci. 342-346. Protein kinaserecognition sequence motifs.Kennelly, P.J. and Krebs, E.G. (1991) J. Biol. Chem. 266: 15555-15558. Consensussequences as substrate specificity determinants for protein kinases and proteinphosphatases.Kinoshita, T., Yokota, T., Arai, K. and Miyajima, A. (1995) EMBO J. 14: 2266-275.Suppression of apoptotic death in hematopoietic cells by signalling through the IL-3/GM-CSF receptors.Knighton, D.R., Zheng, J., Ten Eyck, L.F., Xuong, N., Taylor, S.S. and Sowdadski,J.M. (1991) Science. 253:414-420. Structure of a peptide inhibitor bound to the catalyticsubunit of cyclic adenosine monophosphate-dependent protein kinase.Koury, M.J. and Bondurant, M.C. (1990) Science 248:378-38 1. Erythropoietin retardsDNA breakdown and prevents programmed death in erythroid progenitor cells.Kruh, G.D., Perego, R., Miki, T. and Aaronson, S.A. (1990) Proc. Natl. Acad. Sci. USA87:5802-5806. The complete coding sequence of arg defines the Abelson subfamily ofcytoplasmic tyrosine kinases.Laemmli, U.K. (1970) Nature 227:680-685. Cleavage of structural proteins during theassembly of the head of bacteriophage T4.Laird, P.W., van der Lugt, N.M.T., Clarke, A., Domen, J., Linders, K., McWhir, J.,Berns, A. and Hooper, M. (1993) Nucleic Acids Res. 2 1:4750-4755. In vivo analysis ofPim-1 deficiency.Lagnado, C.A., Brown, C.Y. and Goodall, G.J. (1994) Mol. Cell. Biol. 14:7984-7995.AUUUA is ot sufficient to promote poly(A) shortening and degradation of an mRNA: thefunctional sequence within AU-rich elements may by UUAUUUA(U/A)(U/A).Le Guen, L., Thomas, M., Bianchi, M., Halford, N.G. and Kreis, M. (1992) Gene120:249-254. Structure and expression of a gene from Aradopsis thaliana encoding aprotein related to SNF 1 protein kinase.Levin, D.E., Hammond, C.I., Ralston, R.O. and Bishop, J.M. (1987) Proc. Nati. Acad.Sci. USA 84:6035-6039. Two yeast genes that encode unusual protein kinases.221Lilly, M. (1989) Blood 74:245a. The proto-oncogene pim is an important intermediate inthe mitogenic response to myeloid growth factors. (ABSTRACT)Lilly, M., Le, T., Holland, P. and Hendrickson, S.L. (1992) Oncogene 7:727-732.Sustained expression of the pim- 1 kinase is specifically induced in myeloid cells bycytokines whose receptors are structurally related.Leustek, T., Amir-Shapira, D., Toledo, H., Brot, N. and Weissbach, H. (1992) Cell. Mol.Biol. 38:001-0 10. Autophosphorylation of 70 kDa heat shock proteins.MacNicol, A.M., Muslin, A.J. and Williams, L.T. (1993) Cell 73:571-583. Raf-1 kinaseis essential for early Xenopus development and mediates the induction of mesoderm byFGF.Makris, A., Patriotis, C., Bear, S.E. and Tsichlis, P.N. (1993) J. Virol. 67:4283-4289.Genomic organization and expression of Tpl-2 in normal cells and Moloney murineleukemia virus-induced rat T-cell lymphomas: activation by provirus insertion.Mally, M. I., Vogt, M., Swift, S. E. and Haas, M. (1985) Virology 144:115-126.Oncogene expression in murine splenic T cells and in murine T-cell neoplasms.Maniatis, T., Fritsch, E.F. and Sambrook, J. (1989) Molecular Cloning: A laboratorymanual. Cold Spring Harbour Press. Cold Spring Harbor Laboratory, NY.Mann, M.C., Hsu, B., Stephens, L.C., Brisbay, S. and McDonnell, T.J. (1995) Exp.Cell. Res. 217:240-247. The functional basis of c-myc and bcl-2 complementatio duringmultistep lymphomagenesis in Vivo.Matsuguchi, T., Inhorn, R.C., Carlesso, N., Xu, G., Druker, B. and Griffin, J.D. (1995)EMBO J. 14:257-265. Tyrosine phosphorylation of p95”’ in myeloid cells is regulated byGM-CSF, IL-3 and Steel factor and is constitutively increased by p210BC11ABL.Mcleod, M. and Beach, D. (1986) EMBO J. 5:3665-3671. Homology between the rankgene of fission yeast and protein kinases.Meeker, T.C., Nagarajan, L., ar-Rushdi, A. and Croce, C.M. (1987a) J. Cell. Biochem.35:105-112. Cloning and characterization of the human PIM-1 gene: a putative oncogenerelated to the protein kinases.Meeker, T.C., Nagarajan, L., Ar-Rushdi, A., Rovera, G., Huebner, K. and Croce, C.M.(1987b) Oncogene Res. 1:87-101. Characterization of the human PIM-1 gene: A putativeproto-oncogene coding for a tissue specific member of the protein kinase family.Meeker, T.C., Loeb, J., Ayres, M. and Sellers, W. (1990) Mol. Cell. Biol. 10:1680-1688.The human pim-1 gene is selectively transcribed in different hemato-lymphoid cell lines inspite of a G+C-nich housekeeping promoter.Melchior, F. and Gerace, L. (1995) Curr. Opin. Cell. Biol. 7:310-318. Mechanisms ofnuclear protein import.Merril, C.R., Goldman, D., Sedman, S.A. and Ebert, M.H. (1981) Science 211:1437.Ultrasensitive stain for proteins in polyacrylamide gels show regional variation incerebrospinal fluid proteins.222Meyerson, M., Enders, G.H., Wu, C.-L., Su, L.-K., Gorka, C., Nelson, C., Harlow, E.and Tsai, L.-H. (1992) EMBO J. 11:2909-2917. A family of human cdc2-related proteinkinases.Miura, 0., Miura, Y., Nakamura, N., Queue, F.W., Witthuhn, B.A., Ihie, J.N. and Aoki,N. (1994) Blood 84:4135-4141. Induction of tyrosine phosphorylation of Vav andexpression of pim- 1 correlates with Jak2-mediated growth signaling from the erythropoietinreceptor.Miyoshi, J., Higashi, T., Mukai, H., Ohuchi, T. and Kakunaga, T. (1991) Mol. Cel. Biol.11:4088-4096. Structure and transforming potential of hte human cot oncogene encoding aputative protein kinase.Morgan, D.O. (1995) Nature 374:131-134. Principles of CDK regulation.Moroy, T., Verbeek, S., Ma, A., Achacoso, P., Berns, A. and Alt, F. (1991) Oncogene11:1941-1948. Eji. N- and Ej.i. L-myc cooperate with Eji. pim- 1 to generate lymphoid tumorsat high frequency in double transgenic mice.Moroy, T., Grzeschiczek, A., Petzold, S. and Hartmann, K-U. (1993) Proc. Nati. Acad.Sci. U.S.A. 90:10734-10738. Expression of a Pim-1 transgene accelerateslymphoproliferation and inhibits apoptosis in lprflpr mice.Muscenski, M.L., Bedigian, H.G., Shull, M.M., Copeland, N.G. and Jenkins, N.A.(1988) J. Virology, 62:839-849. Comparative molecular genetic analysis of lymphomasfrom six inbred mouse strains.Muslin, A.J., MacNicol, A.M. and WIlliams, L.T. (1993) Mol.Cell. Biol. 13:4197-4202.Raf- 1 protein kinase is important for progesterone-induced Xenopus oocyte maturation andacts downstream of mos.Nagarajan, L. and Narayana, L. (1993) Biochem. Biophys. Res. Comm. 190:435-439.Transcriptional attenuation of Pim- 1 gene.Nagarajan, L., Louie, E., Tsujimoto, Y., Ar-Rushdi, A., Heubner, K. and Croce, C.M.(1986) Proc. Natl. Acad. Sci. U.S.A. 83:2556-2560. Localization of the human pimoncogene (PIM) to a region of chromosome 6 involved in translocations in acute leukemias.Nebreda, A.R. and Hunt, T. (1993) EMBO J. 12:1979-1986. The c-mos proto-oncogenekinase turns on and maintains the activity of MAP kinase, but not MPF, in cell-free extractsof Xenopus oocytes and eggs.Nishizawa, M., Okazaki, K., Furuno, N., Watanabe, N. and Sagata, N. (1992) EMBO J.11:2433-2446. The “second condon rule” and autophosphorylation govern the stability andactivity of Mos during the meiotic cell cycle in Xenopus oocytes.Padma, R. and Nagarajan, L. (1991) Cancer Res. 51:2486-2489. The human PIM-1 geneproduct is a protein serine kinase.Padmanabha, R., Gehrung, S. and Snyder, M. (1991) Mo! Gen Genet. 229:1-9. TheKNSJ gene of Saccharomyces cerevisiae encodes a nonessential protein kinase homologuethat is distantly related to members of the CDC28/cdc2 gene family.223Patriotis, C., Makris, A., Chernoff, J. and Tsichlis, P.N. (1994) Proc. Nati. Acad. Sci.91:9755-9759. Tpl-2 acts in concert with Ras and Raf-1 to activate mitogen-activatedprotein kinase.Pawson, T. (1995) Nature 373:573-580. Protein modules and signalling networks.Pines, J. (1995) Biochem J. 308:697-7 11. Cyclins and cyclin-dependent kinases: abiochemical view.Pinna, L.A. (1990) Bioch. Biophys. Acta. 1054: 267-284. Casein kinase 2: an ‘eminencegrise’ in cellular regulation?Polotskaya, A., Zhao, Y., Lilly, M.L. and Kraft, A.S. (1993) Cell Growth Different.4:523-532. A critical role for the cytoplasmic domain of the granulocyte-macrophagecolony-stimulating factor x receptor in mediating cell growth.Polotskaya, A., Zhao, Y., Lilly, M.L. and Kraft, A.S. (1994) J. Biol. Chem. 20:14607-146 13. Mapping the intracytoplasmic regions of the x granulocyte-macrophage colony-stimulating factor receptor necessary for cell growth regulation.Posada, H., Yew, N. Ahn, N.G., Vande Woude, G. and Cooper, J.A. (1993)Mol.Cel.Biol. 13:2546-2553. Mos stimulates MAP kinase in Xenopus oocytes andactivates a MAP kinase kinase in vitro.Puziss, J.W., Hardy, T.A., Johnson, R.B., Roach, P.J. and Hieter, P. (1994) Mol. Cell.Biol. 14:831-839. MDSJ, a dosage suppressor of an mckl mutant, encodes a putativeyeast homolog of glycogen synthase kinase 3.Queue, F.W., Sato, N., Witthuhn, B.A., Inhorn, R.C., Eder, M., Miyajima, A., Griffin,J.D. and Ihie, H.N. (1994) Mol. Cell. Biol. 14:4335-4341. JAK2 associates with the Bchain of the receptor for granulocyte-macrophage colony-stimulating factor, and itsactivation requires the membrane-proximal region.Ramakrishnan, L. and Rosenberg, N. (1989) Biochem. Biophys. Acta. 989:209-224. abigenes.Reddy, E.P., Smith, M.J. and Srinivasan, A. (1983) Proc. Natl. Acad. Sci. USA 80:3623-3627. Nucleotide sequence of Abelson murine leukemia virus genome: structural similarityof its transforming gene product to other onc gene products with tyrosine-specific kinaseactivity.Reeves, R. and Magnuson, N.S. (1990) Progress in Nucleic Acid Research and MolecularBiology 38: 24 1-282. Mechanisms regulating transient expression of mammalian cytokinegenes and cellular oncogenes.Reeves, R., Spies, G.A., Kiefer, M., Barr, P.J. and Power, M. (1990) Gene 90:303-307.Primary structure of the putative human oncogene, pim-1.Rihs, H-P., Jans, D.A., Fan, H. and Peters, R. (1991) EMBO J. 10:633-639. The rate ofnuclear cytoplasmic protein transport is determined by the casein kinase II site flanking thenuclear localization sequence of the 5V40 T-antigen.224Ronco, L.V., Silverman, S.L., Wong, S.G., Slamon, D.J., Park, L.S. and Gasson, J.C.(1994) J. Biol. Chem. 269:277-283. Identification of conserved amino acids in the humangranulocyte-macrophage colony-stimilating factor receptor a subunit critical for function.Rostas, J.A.P. and Dunkley, P.R. (1992) J. Neurochem. 59:1191-1202. Multiple formsand distribution of calcium/calmodulin-stimulated protein kinase II in brain.Sagata, N., Watanabe, N., Vande Woude, G.F. and Ikawa, Y. (1989) Nature 342:512-518. The c-Mos proto-oncogene product is a cytostatic factor responsible for meiotic arrestin vertebrate eggs.Saito, Y., Tada, H., Nazarea, M. and Honjo, T. (1992) Growth Factors 7:297-303.Interleukin 2 and tumor necrosis factor a are complementary for proliferation of thehematopoietic stem cell line LyD9.Sanger, F., Coulson, A.R., Barrell, B.G., Smith, A.J.H. and Roe, B.A. (1980) J. Mol.Biol. 143:161-178. Cloning in single stranded bacteriophage as an aid to rapid DNAsequencing.Sato, N., Sakamaki, K., Terada, N., Arai, K. and Miyajima, A. (1993) EMBO J. 12:4181-4 189. Signal transduction by the high-affinity GM-CSF receptor: two distinct cytoplasmicregions of the common B subunit responsible for different signalling.Saris, C.J.M., Domen, J. and Berns, A. (1991) EMBO J. 10: 655-664. The pim-1oncogene encodes two related protein-serine/threonine kinases by alternative initiation atAUG and CUG.Schindler, C. (1995) Annu. Rev. Biochem. 64:621-651. Transcriptional responses topolypeptide ligands: the JAK-STAT pathway.Selton, G., Cuypers, H.T., Boelens, W., Robanus-Maandag, E., Verbeek, J., Domen, J.,van Beveren, C. and Berns, A. (1986) Cell 46:603-611. The primary structure of theputative oncogene pim-1 shows extensive homology to protein kinases.Selton, G., Cuypers, H.T. and Berns, A. (1985) EMBO J. 4:1793-1798. Proviralactivation of the putative oncogene Pim- 1 in MIJLV induced T-cell lymphomas.Shackleford, G.M. and Varmus, H.E. (1987) Cell 50:89-95. Expression of the protooncogene mt-i is restricted to protemeiotic male germ cells and the neural tube of midgestational embryos.Shero, J.H. and Hieter, p. (1991) Genes & Develop. 5:549-560. A suppressor of acentromere DNA mutation encodes a putative protein kinase (MCK1).Siegfried, E., Perkins, L.A., Capaci, T.M. and Perrimon, N. (1990) Nature 345:825-828.Putative protein kinase product of the Drosophila segment-polarity gene zeste-white3.Soderling, T.R. (1990) J. Biol. Chem. 265:1823-1826. Protein kinases. Regulation byautoinhibitory domains.225Sorrentino, V., Mckinney, M.D., Giorgi, M., Geremia, R. and Fleissner, E. (1988) Proc.Nati. Acad. Sci. U.S.A. 85:2191-2195. Expression of cellular proto-oncogenes in themouse male germ line: a distinctive 2.4-kilobase pim-1 transcript is expressed in haploidpostmeiotic cells.Srinivasan, M., Edman, C.F. and Schulman, H. (1994) J. Cell. Biol. 126:839-852.Alternative splicing introduces a nuclear localization signal that targets multifunctional CaMkinase to the nucleus.Storer, R.D., Cartwright, M.E., Cook, W.O., Soper, K.A. and Nichols, W.W. (1995)Carcinogenesis 16(2): 285-293. Short-term carcinogenesis bioassay of genotoxicprocarcinogens in PIM transgenic mice.Svinarchuk, F., Bertrand, J.R. and Malvy, C. (1994) Nucleic Acids Res. 22:3742-3747. Ashort purine oligonucleotide forms a highly stable triple helix with the promoter of themurine c-pim-1 proto-oncogene.Taylor, S.S. (1989) J. Biol. Chem. 264:8443-8446. cAMP-dependent protein kinase.Model for an enzyme family.Telerman, A., Amson, R., Zakut-Houri, R. and Givol, D. (1988) Mol. Cell. Biol.8(4):1498-1503. Identification of the Human pim-1 gene product is a 33-kilodaltoncytoplasmic protein with tyrosine kinase activity.te Riele, H., Maandag, E.R., Clark, A., Hooper, M. and Berns, A. (1990) Nature348:649-651. Consecutive inactivation of both alleles of the pim-1 proto-oncogene byhomologous recombination in embryonic stem cells.Tsatsanis, C., Fulton, R., Nishigake, K., Tsujimoto, H., Levy, L., Terry, A., Spandidos,D., Onions, D. and Neil, J.C. (1994) J. Virology 68(12):8296-8303. Genetic detenuinantsof feline leukemia virus-induced lymphoid tumors: patterns of proviral insertion and generearrangement.van der Geer, P and Pawson,T. (1995) Trends Bioc. Sci. 20:257-293. The PTB domain: anew protein module implicated in signal transduction.van der Lugt, M.T.V., Domen, J., Verhoeven, E., Linders, K., van der Gulden, H., Allen,J. and Berns A. (1995) EMBO J. 14:2536-2544. Proviral tagging in Ej..t-myc transgenicmice lacking the Pim-1 proto-oncogene leads to compensatory activation of Pim-2.van Lohuizen, M., Verbeek, S., Scheijen, B., Wientjens, E., van der Gulden, H. andBerns, A. (1991) Cell 65:737-752. Identification of cooperating oncogenes in Eji-myctransgenic mice by proviral tagging.van Lohuizen, M. and Berns, A. (1990). Biochim. Biophys. Acta. 1032:213-235.Tumorigenesis by slow-transforming retroviruses-an update.226van Lohuizen, M., Verbeek, S., Krimpenfort, P., Domen, J., Saris, C., Radaszkiewicz, T.and Berns, A. (1989) Cell 56:673-682. Predisposition to lymphomagenesis in pim- 1transgenic mice: cooperation with c-myc and N-myc in murine leukemia virus-inducedtumors.Verbeek, S., van Lohuizen, M., van der Valk, M., Domen, J., Kraal, G. and Berns, A.(1991) Mol. Cell. Biol. 11:1176-1179. Mice bearing the Ej.t-myc and Eji-pim-1 transgenesdevelop pre-B-cell leukemia prenatally.Veron, M., Radzio-Andzelm, E., Tsigelny, I. and Taylor, S. (1994) Cell. Mol. Biol.(France) 40:587-596. Protein kinases share a common structural motif outside theconserved catalytic domain.von Lindern, M., Fornerod, M., van Baa!, S., Jaegle, M., de Wit, T., Buijs, A. andGrosveld, G. (1992) Mol. Cell. Biol. 12:1687-1697. The translocation (6;9), associatedwith a specific subtype of acute myeloid leukemia, results in the fusion of two genes, dekand can, and the expressionof a chimeric, leukemia-specific dek-can mRNA.von Lindern, M., van Agthoven, T., Hagemeijer, A., Adriaansen, H. and Grosveld, G.(1989) Oncogene 4:75-59. The human pim-1 gene is not directly activated by thetranslocation (6;9) in acute nonlymphocytic leukemia.Watts, J.D., Affolter, M., Krebs, D.L., Wange, R.L., Samelson, L.E. and Aebersold, R.(1994) J. Biol. Chem. 269: 29520-29529. Identification by electrospray ionization massspectrometry of the sites of tyrosine phsophorylation induced in activated Jurkat T cells onthe protein tyrosine kinase ZAP-70.Wingette, D., Stone, D., Davis, W.C. and Magnuson, N.S. (1995) Cell.Immunol.162:123- 130 Expression of the pim-1 protooncogene: differential inducibility betweena/B- and ‘yla-T cells and B cells.Wingette, D. and Magnuson, N.S. (1995) J. Cell. Biochem. S19A:A1-271. Induction ofpim-1 proto-oncogene expression in T lymphocytes by anti-CD3 monoclonal antibody isdependent on protein kinase C and involves both transcriptional and post-transcriptionalmechanisms (Keystone symposia on Molecular and Cellular Biology).Wingett, D., Reeves, R. and Magnuson, N.S. (1992) Nucleic Acids Res. 20:3183-3189.Characterization of the testes-specific pim- 1 transcript in rat.Wingett, D., Reeves, R., and Magnuson, N.S. (1991) J. Immunol. 147:3653-3659.Stability changes in pim- 1 proto-oncogene mRNA after mitogen stimulation of normallymphocytes.Woodgett, J.R. (1990) Embo J. 9:2431-2438. Molecular cloning and expression ofglycogen synthase kinase-3fFactor A.Yip-Schneider, M.T., Hone, M. and Broxmeyer, H.E. (1995) Blood 12:3494-3502.Transcriptional inductionof pim-1 protein kinase gene expression by interferon y andposttranscniptional effects on costimulation with steel factor.Yu-Sherman, M. and Goldberg, A.L. (1992) Embo J. 11:71-77. Involvement of thechaperonin dnaK in the rapid degradation of a mutant protein in Escherichia coli.227Zakut-Houri, R., Hazum, S., Givol, D. and Telerman, A. (1987) Gene 54:105-111. ThecDNA sequence and gene analysis of the human pim oncogene.Zhang, S.C. and Masui, Y. (1992) J. Exp. Zoology 262:317-329. Activation of Xenopuslaevis eggs in the absence of intracellular Ca activity by the protein phosphorylationinhibitor, 6-dimethylaminopurine (6-DMAP).Zhang, X., Blenis, J., Li, H.-C., Schindler, C. and Chen-Kiang, S. (1995) Science267:1990-1994. Requirement of serine phosphorylation for formation of STAT-promotercomplexes.Zubaiga, A.M., Belasco, J.G. and Greenberg, M.E. (1995) Mol. Cell. Biol. 15:2219-2230. The nonamer UTJAUUUAUU is the key AU-rich sequence motif that mediatesmRNA degradation.228APPENDIX I.DESCRIPTION OF PRIMARY ANTIBODIES USEDUsual DilutionName Animal used (Western Blots) Source & DescriptionCRBPim-1 Sheep antisera 1/500 Cambridge Research BiochemicalsDirected against the peptide corresponding to human and murine amino acids 26-38,ATKLAPGKEKEPLESQY.A2/Daniel Rabbit antisera 1/500 Dr. Michael LillySeattle VA HospitalDirected against a peptide cooresponding to the 18 carboxy-terminal amino acids of humanPim- 1, residues 294-313, PQETAEIHLHSLSPGPSK.C 2 Rabbit antisera 1/500 Dr. Michael LillySeattle VA HospitalDirected against the 12 carboxy-terminal amino acids of murine Pim- 1, amino acid residues302-313, IHLHSLSPGSSK.Tel Mouse monoclonal 1/500 Drs. A. Telerman and R. Amson(CEPH, France)Directed against a TrpE-fusion protein with 69(128-196) residues of human Pim- 1.Pimi-ill Rabbit antisera 1/1000 Pelech Lab, UBIAntipeptide antibody directed against the kinase subdomain ifi region of the human Pim- 1,amino acid residues 70-91,VEKDRISDWGELPNGTRVPMEV-GGC.Piml-NT Rabbit antisera 1/1000 Pelech Lab, UBIAntipeptide antibody directed against the amino terminus of human Pim- 1, amino acidresidues 1-37, MLLSK1NSLAHLRARACNDLHATKLAPGKEKEPLES-GC.Piml-X1 Rabbit antisera 1/1000 Pelech Lab, UBIAntipeptide antibody against kinase domain XI of Xenopus Pim- 1 amino acids 275-292,RPSDRPTLEQLFDHPWMC.229GXP Rabbit antisera 1/500 Pelech LabSera produced against the bacterially expressed full-length GST-Xenopus Pim- 1 fusionprotein. Sera was purified with Protein G and GST-beads.230APPENDIX II.ANTIBODY CHARACTERIZATIONS1. PEPTIDE SELECTION AND WESTERN BLOTTINGi. H. sapiens Pim-1 antibodies - Pim 1-Ill and Piml-NTUnique antibodies were created in order to study the expression patterns of Pim-1protein during the maturation of Pisaster ochraceus (purple sea star) and Xenopus laevis(African clawed frog) oocytes. Initially, only the Pim- 1 sequences of Homo sapiens(human) and Mus musculus (mouse) Pim-1 were known. As these mammalian Pim-1protein sequences were very highly conserved (94%) [Domen et at., 19871, wehypothesized that the Pim- 1 sequence would be similar in other species as well. Therefore,the first antibodies were produced against sequences of mammalian Pim- 1. The rabbitpolyclonal antibodies Pim 1-NT and Pim 1-Ill were directed against peptides based onsequences in the amino-terminus and catalytic subdomain III region of human Pim-l,respectively. As demonstrated in Figure 47, these regions are highly conserved betweenmammalian species. These peptide sequences contain several proline residues whichintroduce ‘kinks’ into the secondary structure of a protein and are often located at thesurface of the macromolecule. By immunizing rabbits with these peptides, we hoped tocreate stronger epitopes for antibody recognition and potentially, immunoprecipitation.The Piml-NT and Pimi-Ill antibodies immunodetected Pim-1 protein from varioussources on Western blots. Figure 48 demonstrates the specificity of Piml-NT and PimiIII towards bacterially-expressed H. sapiens and X. laevis glutathione S-transferase(GST)-Pim-l fusion protein and endogenous Pim-1 from the K562 (human erythroid) cellline. Both antibodies strongly detected the 62 kDa full-length human GST-Pim-1 fusionprotein. Piml-NT detected a strong 34 kDa doublet in K562 lysates. This is in closeagreement with previous reports of proteins of 33-35 kDa in K562 extracts [Telerman etat., 1988; Saris eta!., 1991]. Pimi-IlI antibody detected a faint doublet in K562 lysates;low immunoreactivity in K562 extracts is a result observed by other groups as well [M.Lilly, personal communication].231CRB Pim-] IXENOPUS MLLS KFGSLAHICNPSNMEHLPVKILQPVKVDKEPFEKVYQVGSWGSGGFGTVYSGSRIMOUSE MLLSKINSLAHL- RARPCNDLHLAPGK- EKEPLESQYbVGPLLGSGGFGSVYSGIRVRAT MLLSKINSLAF{L-RAAPCNDLHANKLAPGK-EKEPLESQYQVGPLLGSGGFGSVYSGIRVHUMAN MLLSKINSLAHL - RAAPCNDLHATKLAPGK—EKEPLESbYQVGPLLGSGGFGSVYSGIRVPimi -NTII ifi IVXENOPUS ADGQPVAVKHVAKERVTEWGTL -NGVMVPLEIVLLKKVPTAFRGVINLLDWYERPDAFLIMOUSE ADNLPVAIKHJVEKDRI SDWGELPNGTRVPME\VLLKKVSSDFSGVIRLLDWFERPDSFVLRAT ADNLPVAIKHVEK1JRI SDWGELPNGTRVPMEXItVLLKKVSSGFSGVIRLLDWFERPDSFVLHUMAN SDNLPVAIKHIyEKDRI SDWGELPNGTRVPME\jVLLKKVS SGFSGVIRLLDWFERPDSFVLPimi-IIIV VIXENOPUS VMERPEPVKDLFDYITEKGPLDEDTARGFFRQVLEAVRHCYNCGVVHRDIKDENLLVDTRMOUSE I LERPEPVQDLFDFITERGALQEDLARGFFWQVLEAVRHCHNCGVLHRDIKDENILIDLSRAT I LERPEPVQDLFDFITERGALQEELARSFFWQVLEAVRHCHNCGVLHRDIKDENILIDLNHUMAN ILERPEPVDLFDFITERGALQEELARSFFWQVLEAVRHCHNCGVLHRDIKDENIL IDLNITel Pim-] AbVII VIII IXXENOPUS NGELKL IDFGSGALLKDTVYTDFDGTRVYS PPEWVRYHRYHGRSATVWSLGVLLYDMVYGMOUSE RGE IKL IDFGSGALLKDTVYTDFDGTRVYSPPEWIRYHRYHGRSAAVWSLGILLYDMVCGRAT RGELKLIDFGSGALLKDTVYTDFDGTRVYS PPEWIRYHRYHGRSAAVWSLGILLYDMVCGHUMAN ROELKLIDFGSGALLKDTJYTDFDGTRVYSPPEWIRYHRYHGRSAAVWSLGILLYDMVCGX XI Piml-XIXENOPUS DI PFEQDEE IVRVRLCFRRRI STECQQLIKWCLSLPSDRPTLEQIFDHPWMC(CDLVKSMOUSE DIPFEHDEEI IKGQVFFRQTVSSECQHLIKWCLSLRPSDRPSFEEIRNHPWM—QGDLLPQRAT DI PFEHDEE IVKGQVYFRQRVSSECQHLIRWCLSLRPSDRPSFEEIQNHPWM-QDVLLPQHUMAN DIPFEHDEEIIRGQVFFRQRVSSECQHLIRWCLALRPSDRPTFEEIQNHPWM-QDVLI]Lilly’s C2 AntibodyXENOPUS EDCDLRLRTIDNDSSSTSSSNESLMOUSE AASEIIHLHSLSPGSSK-I-RAT ATAEIHLHSLSPSPSKHUMAN ETAEIHLHSLSPGPSKI-Lilly’s “Daniel”Figure 47. Regions of Pim-l proteins against which antibodies are directed.Protein sequence alignments of Pim- 1 from X. laevis, mouse, rat and human; openboxes designate peptides/regions against which antibodies were made. The namesof the antibodies are italicized above or below the sequence. Roman numeralsindicate protein kinase subdomains. Specific details are listed in the text and inAppendix I.232Piml-NTPimi-ifiFigure48.Detectionofbacterially-expressedandendogenousPim-1withH.sapiensantibodies.Pimi -NT(leftpanel)andPimi-III(right panel) antibodieswereusedtoprobeWesternblots ofcellextracts.Lanes1and2contained0.75ugofbacterially-expressedH.sapiensandX.laevisGST-Pim-1, respectively.Lane3containedK562cellextracts(7.5 x10cellsperlane).Lanes4and5containedimmatureandmatureseastaroocyteextracts, respectively(approx.350ugperlane)andlanes6and7containedimmatureandmatureX.laevisoocyteextracts, respectively(approx.300ugperlane).Solidarrows(-*)showlocationofexpressedGST-Pim- 1.Hollowarrows(-‘)showlocationofPim-1doublet inK562cellextracts.Migrationsof Mrstandardsareshownonright.3456749.734567—49.7—26.4—26.4L)The Piml-NT and Pimi-Ill antibodies were also used to screen for Pim-1 expression inP. ochraceus and X. laevis oocytes. Piml-NT and Pimi-Ill both detected a strong seriesof proteins of about 44 kDa proteins in immature and mature sea star oocytes (Fig. 48,lanes 4 and 5). Both antibodies recognized the 62 kDa X. laevis GST-Pim-1 fusionprotein very poorly, and while Piml-NT detected proteins of 31 and 42 kDa in X. laevisoocyte extracts, Pimi-Ill did not detect any proteins of the appropriate size in theseextracts. This is hardly surprising, for when the sequence of X. laevis Pim- 1 is comparedwith that of the mammalian sequences in the regions against which the antibodies weredirected, the sequences are dissimilar in these regions. Additional proteins of inappropriatesizes were also immunodetected on Western blots with these antibodies. As the antibodieswere polyclonal, they contained a mixture of antibodies directed against different epitopesin immunizing peptides. Some of these antibodies might have cross-reacted with unrelatedproteins sharing epitopes with Pim- 1.ii. X. laevis Pim-1 antibodies: GXP and Piml-XIThe lack of immunoreactivity of the human-directed Pim-1 antibodies to endogenousand bacterially-expressed X. laevis Pim- 1 necessitated the production of frog-specificantibodies. The Piml-XI rabbit polyclonal antibody was directed against the kinasesubdomain XI of X. laevis Pim- 1, which showed considerable conservation between allspecies (Fig. 47). The GXP rabbit polyclonal antibody was directed against the full-lengthbacterially-expressed X. laevis GST-Pim-1 fusion protein. Although it would have beenpreferable to immunize the rabbits with the Pim- 1 portion of the fusion protein alone, onlysmall quantities of the thrombin-cleaved Pim- 1 were obtainable, as will be discussed inChapter IV, Section 3.i.b. Instead, the GXP serum was passed through a GST-boundglutathione agarose column to remove most of the anti-GST antibody. Although there wasstill some cross-reactivity, immunoreactivity towards expressed GST was significantlyreduced.Both antibodies were tested for immunodetection with recombinant X. laevis GSTPim- 1, H. sapiens GST-Pim- 1 and native Pim- 1 in extracts from X. laevis and sea staroocytes and from K562 cells (Fig. 49). Both GXP and Piml-XI strongly detectedrecombinant X. laevis GST-Pim- 1, GXP faintly detected the human recombinant andnative Pim-1, and Piml-XI did not detect human Pim-1 at all (Figure 49, lanes 1 - 3).Both antibodies detected strong 44 kDa proteins in the immature and mature sea starextracts and the GXP antibody detected an additional 33 kDa protein in these lanes (Figure49, lanes 4 and 5). GXP detected a strong 31 and 42 kDa proteins similar to those234observed with Pim 1-NT in the X. laevis oocyte extracts, but the Piml-XI antibody onlydetected a 38 kDa band.The frog Pim- 1 antibodies were designed to cross-react with Pim- 1 from other species;the Pim 1 -XI antibody was directed against a region in the catalytic domain of Pim- 1 that ishighly homologous between species, and the GXP antibody was directed against the fulllength X. laevis Pim- 1 which contains many regions very similar to the mammalian Pim- 1.The GXP antibody displayed some cross-reactivity with sea star and human Pim-l, whilePim 1 -XI primarily detected sea star and frog Pim- 1. That these antibodies cross-reactedwith Pim-1 from several different species was desirable, as they allowed examination ofPim- 1 expression in other systems (i.e. sea star).235GXPPiml-XIFigure49.Detectionofbacterially-expressedandendogenousPim-1withX.laevisantibodies.GXP(leftpanel)andPiml-XI(rightpanel) antibodieswereusedtoprobeWesternblotsofcellextracts.Lanes1and2contained0.75ugofbacterially-expressedH.sapiensandX.laevisGST-Pim-1, respectively.Lane3containedK562cellextracts(7.5x10cellsperlane).Lanes4and5containedimmatureandmatureseastaroocyteextracts, respectively(approx.350ugperlane)andlanes6and7containedimmatureandmatureX.laevisoocyteextracts, respectively(approx.300ugperlane).Solidarrows(—.)showlocationofexpressedGST-Pim-l.Hollowarrows(-)showlocationofPim-ldoubletinK562cellextracts.Migrationsof Mrstandardsareshownonright.-49.7—26.42. COMPARISON OF PIM- 1 ANTIBODIES FROM VARIOUS SOURCESFigure 47 shows regions of the Pim- 1 sequence against which the antibodies fromother sources (detailed in Appendix 1) were directed. The Pim-1 antibodies obtained fromother researchers were compared with those produced in our laboratory (Fig. 50).Although titres of our antibodies were available, those from other sources were not. Wedid not attempt to titre these antibodies as only limited quantities for confirmatory workwere obtained. Because both of the antibodies kindly donated by Dr. Michael Lilly weredirected against C-terminal residues, a 1:1 cocktail of A2 and C2 antibodies was usedunless otherwise indicated. This cocktail was designated as “Lilly”.K562 cell lysates were used as a positive control for Pim- 1 detection, as they have beenreported to express high amounts of endogenous Pim-1 [Telerman eta!., 1988; Saris et al.,1991]. Approximately 7.5 x 1O cells were electrophoresed on each lane of a Western blot.A 32-35 kDa doublet was observed with most antibodies tested. A 34 kDa doublet/smearwas detected strongly with Tel, CRB, Piml-NT and Lilly, detected faintly with GXP andPimi-ifi, but not detected with the Pimi-IX antibody.Most antibodies detected the 62-kDa bacterially-expressed H. sapiens and X. laevisGST-Pim-1 fusion proteins by Western blotting (Fig. 50, panels B, C). Human GSTPim-1 protein was detected very strongly by the Pim-NT, Pim 1-Ill, Tel, Lilly, and CRBantibodies (panel B). The protein was detected only faintly by GXP and was not detectedat all by Piml-XI, 4G10 and PY2O. Several lower Mr proteins were observed in the PimiIll, Piml-NT, GXP and CRB lanes. These bands are likely to be the products of abortedtranslation, as these antibodies are directed towards regions in the amino-terminus of theprotein. The anti-GST, GXP, Piml-XI, Tel and 4G10 antibodies strongly detected frogGST-Pim-1, Pimi-Ill and PY2O faintly visualized frog GST-Pim-1, and Piml-NT, Lillyand CRB negligibly detected frog GST-Pim-1 (panel C).237AS.z’.2 3 4 5Figure 50. Pim-l antibody comparison by irmnunodetection of Pim-1 on Western blots.Panel A: K562 cell extracts (7.5 x l0 cells). Panel B: bacterially-expressed H. sapiensGST-Pim-l (0.75 ug per lane). Panel C: bacterially-expressed X. laevis GST-Pim-1(0.75 ug per lane). Antibodies used: lane 1, anti-GST; lane 2, GXP; lane 3, Piml-XI;lane 4, Piml-NT; lane 5, Pimi-Ill; lane 6, Tel; lane 7, Lilly; lane 8, CRB. Open arrow(<— ) indicates endogenous Pim- 1 protein, filled arrow (4—) indicates bacterially-expressed GST-Pim- 1. Migrations of Mr standards are shown at right.49.7- 26.412345678123456782383. IMMUNOPRECIP1TATIONThe antibodies were tested for the ability to immunoprecipitate Pim- 1 from varioussources including preparations of expressed fusion proteins, sea star and X. laevis oocytesand from K562 cell lysates. Immunoprecipitation allows rapid and selective resolution ofthe target protein from other proteins in a cell homogenate, yielding a relatively pure proteinpreparation for examination on a Western blot or for enzyme activity analysis (i.e.autophosphorylation or kinase assays). For the purposes of this study, it was hoped thatactive Pim- 1 could be immunoprecipitated from oocyte preparations in order to examineboth changes in the protein levels, apparent molecular mass or kinase activity of theendogenous enzyme during different stages of oocyte maturation.Initially, bacterially-expressed X. laevis and H. sapiens GST-Pim- 1 were used aspositive controls for immunoprecipitation experiments. Unfortunately, GST-Pim- 1adhered non-specifically to Protein-A Sepharose, so large quantities of the fusion proteinwere removed during the Protein-A pre-clearing step. Secondly, because of the highconcentrations of fusion protein and the high affinity of the antibodies for the fusionprotein, we were never certain if the fusion protein that was visible in theimmunoprecipitates was immunoprecipitated or was present in trace amounts as acontaminant. Increased washes did not reduce the amount of fusion protein in theimmunoprecipitates. Antibody cocktails with a mixture of dissimilar antibodies were oftenused to improve the prospects of immunoprecipitation.i. Immunoprecipitation of Pim- 1 from the K562 (human ervthroid) cell linePim-1 was immunoprecipitated from K562 cells by Tel, GXP, Lilly and weakly byPiml-XI antibody (Fig. 51). The inefficiency of Piml-XI for immunoprecipitating Pim-1from K562 cells was not surprising, as this antibody was directed against a X. laevisspecific peptide and did not recognize the human GST-Pim- 1. The other X. laevisantibody, GXP, was able to immunoprecipitate human Pim- 1, probably because itrecognized epitopes common to both proteins. The human antibody, Pim 1-NT did notimmunoprecipitate Pim- 1, possibly because it was directed against the amino-terminus ofthe kinase, which may be folded into the interior and may therefore be unaccessible. Thefact that the Pim 1-HI antibody was able to immunoprecipitate Pim-1 from K562 cells wassurprising, as it only faintly detected Pim- 1 protein in this cell line on Western blots (Fig.48, lane 3). The anti-phosphotyrosine mouse monoclonal antibodies 4G10 and PY2O didnot immunoprecipitate Pim-1 from K562 cells.239TelGXPLillyPiml-XI4G10PY2OhuNTPimi-ifiPiml-NTIIIIIIIIIIIIIiiIIFigure51.Immunoprecipitationof Pim-1fromK562cellswithvariousantibodies.K562cellextracts(+) andPBScontrols(-)wereimmunoprecipitatedwiththevariousPim-1andantiphosphotyrosineantibodiesshownonthetopoftheblot.Migrationof Mrstandardsareshownontheright.The33kDaPim-1bandisindicatedbyanarrow(>).Eachimmunoprecipitationwasfromapprox.3x106cells.Cii. Immunoprecipitation of Pim- 1 from P. ochraceus oocyte extractsA cocktail of Pimi-IJI and Piml-NT antibodies was used to immunoprecipitate Pim-1from sea star oocyte time course homogenates (data not shown). A 40 kDa band thought tobe Pim-l was observed in both the supernatants and in the whole oocyte homogenates, butnot in the immunoprecipitates (data not shown). Similar results were observed insubsequent experiments and with the Lilly cocktail (data not shown), leading to theconclusion that Pimi-ilI and Pim 1-NT antibodies did not immunoprecipitate Pim-1 fromsea star oocytes, even though they detected the presence of Pim-1 in these lysates.iii. Immunoprecipitation of Pim-1 from X. laevis oocyte extractsGXP, Piml-XI and the Lilly antibody cocktail were used to immunoprecipitate Pim-1from immature and mature X. laevis oocytes. There were no unique proteins present in theimmunoprecipitates that were not already detected in the controls and proteins at the sizeexpected for the Pim-1 were detected in the supernatants with the GXP and Piml-XIantibodies, indicating that Pim- 1 was not immunoprecipitated. Both antibodies detectdifferent bands in the supematants; it is uncertain which band is actually Pim-1. The Lillycocktail did not detect Pim- 1 protein in the controls or immunoprecipitates (data notshown).It was hoped that GXP and Pim 1-NT would be the most efficient antibodies forimmunoprecipitation of endogenous X. laevis Pim- 1, as both are directed against epitopesthat should be exposed on the surface of the native protein. GXP antibodyimmunoprecipitated Pim- 1 from K562 cells, but was unsuccessful at immunoprecipitatingthe kinase from X. laevis oocyte extracts possibly due to differences in homogenizationbuffers. While X. laevis oocyte extracts were homogenized under non-denaturingconditions, K562 cells were lysed in buffer containing 1% Triton X-lOO, 0.5%deoxycholate and 0.1% SDS. The presence of these detergents may have caused proteinunfolding and denaturation allowing antibodies better access to the epitopes. As weultimately wanted to obtain active enzyme from the oocyte preparations, non-denaturingconditions were required and the use of high detergent buffers was unacceptable.2414. ANTIBODY INHIBiTION OF KINASE ACTIVITYAll antibodies were tested for the ability to inhibit the phosphotransferase activity ofexpressed GST-Pim- 1 toward a peptide substrate in order to determine if any of theepitopes to which the antibodies were directed were important for kinase-substrate bindingor catalysis. The kinase activity of 8 nmoles of human or X. laevis GST-Pim-l wasassessed in the presence of 0.184 to 32 nmoles of antibody per assay. The activity towardsP4 substrate peptide was plotted against the molar ratio of antibody to enzyme in order toassess the ability of the antibodies to inhibit the kinase activity (Fig. 52). Antibodies thatreduced the kinase activity more than 25% in a molar ratio of 2 or less were judged to beinhibitory. X. laevis GST-Pim-l kinase activity was strongly inhibited by GXP and Tel,and weakly inhibited by Pim 1-NT and Pimi-Ill. H. sapiens GST-Pim-l was also stronglyinhibited by GXP and Tel antibodies, and weakly by the Pimi-ifi antibody.The results of antibody inhibition study can be interpreted when the relationshipbetween the antibodies and the Pim-1 protein is considered. Antibodies such as Tel andGXP, which are directed against large portions of the protein are most likely to inhibit theactivity of the kinase, as they can recognize epitopes that are on the “outside” of the foldedprotein. Antibody binding may inhibit or induce a conformational change in Pim- 1necessary for phosphotransferase activity or may sterically block sites important forsubstrate binding. The Pim 1-Ill antibody, directed at similar regions in the catalyticdomain caused a very slight inhibition of Pim- 1 activity. As the N-terminus of Pim- 1 wasfused to GST, normal protein folding of the Pim- 1 was likely be hindered, so the inhibitionby Pimi-ifi may not accurately reflect the situation with the endogenous, native protein.The results using the Lilly cocktail were not shown as the antibody did not inhibit the GSTPim-1 activity even at very high amounts. Of note but of unclear significance, is the factthat low concentrations of the antibodies caused a slight increase in the amount of 32Pincorporated into the P4 peptide by Pim- 1.Although it was hoped that immunoprecipitation would be a suitable method forobtaining active, purified endogenous Pim-1 from oocytes, these results demonstrated thatthe Pim-1 antibodies inhibited the kinase activity of Pim-1 if they were able toimmunoprecipitate the kinase. Therefore, the kinase activity of immunoprecipitatedendogenous Pim-1 could not be accurately assessed in our study.242A—0——— Pimi-lIlPim-NT--4-- TelCRB------ Piml-XlGXP—0--— Piml-lllPim-NT——4—— TelCRB—‘——0—-— Piml-XIGXP0C)3.0Molar ratio (antibody:GST-Pim-l)Figure 52. Antibody inhibition of kinase activity.The activity of 8 nmol of X. laevis (panel A) and H. sapiens (panel B) GST-Pim-ltowards P4 peptide was inhibited with increasing concentrations Pim-1 antibodies.The amount of 32p -incorporation by P4 peptide is shown on the vertical axis and themolar ratio of antibody to GST-Pim-l protein is shown on the horizontal axis. The1:1 molar ratio is shown by a dotted vertical line.0.0 0.5 1.0 1.5 2.0 2.5 3.0Molar ratio (antibody:GST-Pim-l)B0.0 0.5 1.0 1.5 2.0 2.52435. ANTIBODY SUMIvIARYWe produced four distinct Pim-1 antibody preparations that detected recombinant andnative forms of amphibian and mammalian Pim- 1. The antibodies displayed some speciesspecificity; Piml-NT and Pimi-HI recognized mammalian Pim-l, while Piml-XI and GXPrecognized X. laevis GST-Pim- 1. Comparison studies with Pim- 1 antibodies fromvarious sources confirmed that the antibodies constructed for this study immunodetectedPim-l protein from bacterially-expressed and endogenous sources. Although the GXP andto a lesser extent, Piml-XI immunoprecipitated Pim-l from K562 cells, Pimi-Ill andPiml-NT did not immunoprecipitate Pim-1 from sea star oocyte extracts, and GXP andPimi -XI did not immunoprecipitate Pim- 1 from X. laevis oocyte extracts. AntibodiesGXP and Tel inhibited the activity of expressed GST-Pim- 1 possibly by blocking substratebinding sites or by inducing or inhibiting a conformational change in the protein thatinterfered with catalytic activity. These antibodies were used to examine the proteinexpression of Pim- 1 in the bacterial expression system (Chapter IV), in maturing X. laevisoocytes (Chapter VII) and in maturing P. ochraceus oocytes (Chapter VIII).244APPENDIX III.DES CRIPTION OF OLIGONUCLEOTIDEShA 5’-CTG ACC CGG GCT CGA GGC ICC IGG IAA (G/A)GA (G/A)AA (G/A)GA(G/A)CC-3’Based on the human and mouse sequence APGKEKEP, aa 26-33. 39 bp.Degenerate sense primer for the amplification of a pim-1 probe from X. laevis and P.ochraceus cDNA. CTGA clamp, CCCGGG Smal site, CTCGAG Xhol site. (0.886ug/ul, 0.56 ul per PCR reaction)12A 5’-CTG ACC CGG GCT CGA GGA (C/T)TG OTT (C/T)GA (G/A)AG ICC ICA3,Based on human and mouse sequence DWFERPD, aa 108-114. 36 bp.Degenerate sense primer for the amplification of a pim-1 probe from X. laevis and P.ochraceus eDNA. CTGA clamp, CCCGGG Smal site, CTCGAG Xhol site. (0.7 14ug/ul, 0.7 ul per PCR reaction)13B 5’-CTG ACC CGG OCT CGA GAT (C/T)TC (C/T)TC (G/A)TC (G/A)TG TTC(G/A)AA 100-3’Based on human and mouse sequence PFEHDEEI, aa 241-248. 39 bp.Degenerate anti-sense primer for the amplification of a pim-1 probe from X. laevis and P.ochraceus eDNA. CTGA clamp, CCCGGG Smal site, CTCGAG Xhol site. (0.938ug/ul, 0.53 ul per PCR reaction)14B 5’-CTGACC CGGGCT CGA GC(C/T) TGC ATC CAl GG(G/A) TG(G/A)TT-3’Based on human and mouse sequence NHPWMQ, aa 286-29 1. 35 bp.Degenerate anti-sense primer for the amplification of a pim-1 probe from X. laevis and P.ochraceus cDNA. CTGA clamp, CCCGGG Smal site, CTCGAG Xhol site. (1.756ug/ul, 0.285 ul per PCR reaction)9402 5-COT ACC CCG GGC ATG CTC TTG TCC AAA ATC-3’Based on human sequence MLLSKI, aa 1-6. 30 bp.Sense primer for the amplification of the human pim-1 coding region for cloning into thepGEX-2T expression vector. CGTAC clamp, CCCGGG Smal site. (0.244 ug/ul, 25.02pmollul)2459405 5’-GCG GAA TCC TAT TTG CTG GGC CCC GGC GAC AGG-3’Based on human sequence LSPGPSK, aa 307-3 13. 30 bp.Antisense primer for the amplification of the human Pim- 1 coding region for cloning intothe pGEX-2T expression vector. GCG clamp, GAATTC EcoRl site. (0.200 ug/ul, 18.64pmollul)pim5 5’ - CGA TGG ATC CAT GCT TCT CTC TAA ATT CGG -3’Based on X. laevis sequence MLLSKF, aa 1-6. 30 bp.Sense primer for the amplification of the X. laevis Pim- 1 coding region for cloning into thepGEX-2T expression vector. CGAT clamp, GAATTC BamHl site. (28.2 pmollul)pim3 5’- GAT CGA AlT CCA GAC TCT CGT TGC TTG A -3’Based on X. laevis sequence SSNESL, aa 3 18-323. 33 bp.Antisense primer for the amplification of the X. laevis Pim- 1 coding region for cloning intothe pGEX-2T expression vector. GATC clamp, GAAITC EcoRl site. (27.9 pmol/ul)K 1 5’- CTC CTT AGC TAC GTG GCG CAC AGC GAC CGG CTG -3’Based on X. laevis sequence QPVAVAHVAKE, aa 64-74. 33 bp.Antisense primer for site directed mutagenesis (K69-A) of X. laevis Pim- 1. The lysinecondon, TCC, (nt 205-207) is changed to GCG (alanine). This oligonucleotide was usedwith pim5 to amplify a 230 base pair PCR fragment which was designated as oligo 5K1.K2 5’- AGA GTC ACA GAA TGG GGC -3’Based on X. laevis sequence RVTEWG, aa 75-80. 18 bp.Sense primer for site directed mutagenesis (K69-A) of X. laevis Pim- 1. Thisoligonucleotide was meant to be used with pim3 to create a PCR fragment to ligate with thepim5 - Ki fragment (oligo 5K1).5K1232 bp, corresponding to the X. laevis amino acid sequence, aa 1-74.Sense primer containing the K69-A mutation at nt residue 207. This oligonucleotide wasamplified by a PCR reaction with pim5 and Ki, and used with pim3 to amplify the entirecoding region of the pim-1.246PM1 5’ - CFGATC GATTT GGC GCC GGG GCG CTA CTC -3’Based on X. laevis sequence, aa 184-194, LIDFCGAKK. Sense primer for site directedmutagenesis (S190-A) of X. laevis Pim-1. This oligonucleotide was meant to be used withpim3 to create a PCR fragment to ligate into the Clal/EcoRl sites of the X. laevis pim-] inpGEX-2T. Italics indicate Clal site, bold underline denotes mutation. [10 ug/ul, 102.562pmollul]PM2 5’ - CTG ATC GAT TTT GGC GAA GGG GCG CTA CTC -3’Based on X. laevis sequence, aa 184-194, LIDFCGAKK. Sense primer for site directedmutagenesis (S 190-E) of X. laevis Pim- 1. This oligonucleotide was meant to be used withpim3 to create a PCR fragment to ligate into the Clal/EcoRl sites of the X. laevis pim-1 inpGEX-2T. Italics indicate Clal site, bold underline denotes mutation. [13.78 ug/ul, 141.3pmollul]PM3 5’ - GGA TAC GGT GGA AAC GGA TTT TGA TGG -3’Based on X. laevis sequence, aa 195-204, DTVXTDFDG. Sense primer for site directedmutagenesis (Y199-E) of X. laevis Pim-1. This oligonucleotide was meant to be usedeither with pim3 to create a large oligo to use with pim5 to amplify the full-length codingregion of pim- 1, or to be used with oligo pim3 to create a PCR product to ligate with thePCR product of PM4 and pim5’. Bold underline denotes mutation. No restriction sites.[1.8 ug/ul, 20.51 pmollulJPM4 5’ - TTG AGT AGC GCC CCG GAG CC -3’Antisense primer based on X. laevissequence, aa 189-195, KLLAGSG. To be used withpim5 to create a fragment to ligate with the PCR product of PM4 and pim5’. No mutation,no restriction sites. [3.0 ug/ul, 46.153 pmol/ul]T3 5’-AATTAACCCTCAcTAAAGGG-3’Primer for sequencing reactions. (0.126 ug/ul)T7 5’-GTAATACGACTCACTATAGGGC-3’Primer for sequencing reactions. (0.277 ug/ul)2475H6Pim 5’- (P) ATG GAA GAG GAA GAG GAA GAG CTT CTC TCT AAA TTCGGA TCG -3’ MHHHHHHLLSKFGSPrimer used to amplify the coding region of WT X. laevis pim-1 and mutants to put ahistidine tag on Pim- 1 to insert into Sma 1 site of PEF 1. Used by Dr. Gabe Kalmar atS.F.U.pim3’ 5’- CAA AGC TTT ACA GAC TCT CGT TGC HG AGCPrimer used to amplify the coding region of WT pim- 1 and mutants and to insert into Sma1 site of PEF1. Used by Dr. Gabe Kalmar at S.F.U.248APPENDIX IV.PEPTIDE SUBSTRATESName Sequence Molecular MassKemptide LRRASLG 937.1 g/molS6 10 AKRRRLSSLRASTSKSESSQK 2360 g/molP1 AKRRRLSSLRA 1312 g/molP2 AKRRRLSALRA 1296 g/molP3 AKRRRLASLRA 1296 g/molP4 AKRRRLSA 956 g/molP5 AKRRRLS - amide 885 g/molP6 AKRRRLS - free acid 885 g/molP7 AKRRRLTA 970 g/molP8 AKRRRLYA 1032 g/molP9 AKRRRKSA 971 g/molPlO AKRRRRSA 999 glmolP11 AKRRRESA 972 glmolP12 AKRRRQSA 971 glmolP13 AKRRRASA 914 g/molP14 AKRRRISA 956 g/molP15 AKRRALSA 871 glmolP16 AKRRKLSA 928 g/molP17 AKRARLSA 871 g/molP18 AKRKRLSA 928 g/molP19 AKARRLSA 871 g/molP20 AKKRRLSA 928 g/molP21 AKARALSA 786 g/molP22 KRRRLSA 885 g/molP23 RRRLSA 757 g/molP24 AARRRLSA 899 g/molP25 ARRRRLSA 984 g/molP26 ALRRRLSA 941 glmolP27 AERRRLSA 957 g/molP28 AKRRRLAA 940 g/molP29 AKRRRLCA 972 g/mol249APPENDIX VFORMULAE FOR CALCULATIONS1. Calculation of ATP concentration and cpm/pmolConcentration of [y-32P1-ATP stock: 2.2 uM in 100 ul volume (220 pmoles). Diluted in 4-5 ml of 250 uM ATP to a final concentration of 250 uM, designated as ‘Assay ATP’.x counts (cpm in 5 ul of assay ATP)/1250 (pmol in 5 ul ) = cpmlpmol=D2. Calculation of activity of enzyme in filter paper assayx (cpm of assay) x total volume of assay = pmollmin.mgD x time (mm) x [enzyme (mg) vol spotted on filter250APPENDIX VICalculation of Kinetic Constants -Michaelin Menten and Lineweaver Burke PlotsFor the calculation of kinetic constants, two methods were employed, theMichaelis-Menten and the Lineweaver Burke equations, which can be expressedas graphs. Representatives of both methods are shown below for the measurementof the activity of expressed wild-type X. laevis GST-Pim- 1 towards the P4 substratepeptide (AKRRRLSA).The Michaelis-Menten plot is derived from plotting the enzyme activityagainst peptide concentration. The best-fit curve is plotted, and the maxima isdetermined as the Vmax value. Half of the Vm is plotted and used to determinethe Km value of the enzyme for the peptide.Michaelis- Menten Plot for GST-Pim-1 activity toward P4 peptideVmax = 88 nmol.min1. gKm 42jiMECE1007060i 90max80501/2 Vm403020100 25 50 75 100 125 150 175 200 225 250P4 peptide concentration (ILM)251A more accurate method of kinetic constant determination is theLineweaver-Burke plot. The inverse of the kinase activity is plotted againstthe inverse of the peptide concentration. The Vmax and Km values arederived from the Y and X-intercepts of the slope equation.Lineweaver-Burke plot of GST-Pim-1 activity toward P4 peptideSlope equation y = 1 .5092e-2 + 0.70528x RA2 = 0.99 1Vmax= 66.3 nmol.min’.mg’.Km = 471J.MInverse peptide concentration (1/[jiM])0.1200.1100.1000.090o b 0.080. E 0.0700.060) 0.0500.0400.030— —— 0.0200.0100.000\\1/Km0.100 0.125 0.150252APPENDIX VIIAUTOPHOSPHORYLATION/DEPHOSPHORYLATION AND EXOGENOUSKINASE ACTIVITYi. Dephosphorviation of GST-Pim-1As GST-Pim- 1 was already autophosphorylated by the time it was purified from thebacteria, we wanted to determine if dephosphorylation would cause a shift in molecularmass of GST-Pim-1 or would alter the activity of the kinase. The phosphatases that weretested for Pim-1 dephosphorylation included broad specificity acid and alkalinephosphatases and a tyrosine-specific phosphatase corresponding to the intracellular domainof HPTPB. The phosphatases were removed from the glutathione-bound GST-Pim-1 afterphosphatase treatment, so that the in vitro autophosphorylation of the dephosphorylatedGST-Pim- 1 could be subsequently examined.Dephosphorylation by the various phosphatases did not cause a shift in apparentmolecular weight of GST-Pim-1 (Fig. 53, panel A). The slightly undulating band patternsobserved in lanes 2, 4, 5 and 7 were due to a slight problem with the stacking gel and arenot due to band shifting. Panel B demonstrates that the dephosphorylated Pim- 1 sampleswere active and able to autophosphorylate. The acid phosphatase treated sample retained58% of activity, the alkaline phosphatase-treated sample retained 55% of activity and theHPTPB-treated sample retained 65% of activity as compared to the control. Although it istempting to speculate that the reduction in activity observed was due to dephosphorylation,it likely reflected residual phosphatase contamination. The control sample was incubated inparallel with the other samples, so any effects of protease contamination would have beenconsistent between samples; as degradation of the samples was not observed in panel A,the reduction in activity was unlikely due to proteolytic degradation. Panel C demonstratesthat both the acid and alkaline phosphatases caused a very minor reduction in the amount oftyrosine phosphorylation as detected by the anti-phosphotyrosine antibody (lanes 3 and 5)and autophosphorylation did not change this amount (lanes 4 and 6). HPTPB treatmentreduced tyrosine phosphorylation by an estimated 80%, as detected by the antiphosphotyrosine antibody (lane 7), and autophosphorylation seemed to cause a 5-10%recovery of this signal (lane 8).253ABCFigure 53. Dephosphorviation and subsequent phosphorylation ofX. laevis GST-Pim-l. Control and phosphatase-treated GST-Pim-1were loaded directly onto an SDS-PAGE gel (lanes 1, 3, 5, 7) or weresubjected to autophosphorylation reactions (2,4, 6, 8). Panel A,silver stain of SDS-PAGE gel; panel B, autoradiograph of Westernblot; panel C, Western blot probed with 4G10 antibody. Control(lanes l&2), acid phosphatase-treated (lanes 3&4), alkaline phos.-treated (lanes 5&6) and HPTP3-treated (lanes 7&8) GST-Pim- 1.Boundaries of panels are defined by the 97.2 and 50.0 kDa Mr standards.12345678254ii. Dephosphorviation curve of HPTPBA dephosphorylation curve was performed using serial dilutions of HPTPB todetermine if complete dephosphorylation of GST-Pim- 1 by HPTPB could be achieved. Ateven very high concentrations of HPTPB, the tyrosine phosphorylation of GST-Pim- 1 wasnot completely reduced as detected by antophosphotyrosine antibodies (data not shown).As Figure 20 demonstrated that recognition of GST-Pim-1 by the antiphosphotyrosineantibody was specific, it is possible that Pim- 1 has several tyrosine phosphorylation sitesand only certain ones are targeted by the HPTPB. Indeed, kinetic analysis of theintracellular domain of HPTPB has revealed that the phosphatase displays substratespecificity [Harder et al, 1994].iii. Activity of GST-Pim- 1 after autophosphorvlationldephosphorvlationTo determine if differences in phosphotransferase activity toward exogenoussubstrates existed between the control and phosphatase-treated GST-Pim-1 and to examineif the exogenous phosphotransferase activity of the control and phosphatase-treated GSTPim-1 changed after autophosphorylation, a large scale in vitro autophosphorylationexperiment was performed. At discrete time points from 1-20 mm, aliquots of theautophosphorylation reaction were removed and added to P4 peptide substrate assays.Results indicated that the activity of all samples towards exogenous substrates was similarand no trends were apparent as a consequence of autophosphorylation (data not shown).255APPENDIX VIIIPRELIMINARY EXPERIMENTS IN OOCYTE SYSTEMS1. ASSOCIATED PROTEINS AND SUBSTRATES OF X. LAEVIS PIM- 1Many cytoplasmic proteins associate in large multimeric complexes in cells. Thesecomplexes may function to bring regulatory subunits into close proximity to catalyticproteins, or may allow enzyme-substrate combinations to interact with greater efficiency.In many cases, these proteins have been discovered by cross-linking orimmunoprecipitation experiments.Pim-1 is a relatively small kinase with very short N- and C- terminal regions outside ofthe catalytic domain. It is highly likely that other proteins interact with the Pim- 1 tomodulate its activity. These putative factors might function in a cell-cycle dependentmanner, similar to the interaction of other cell cycle modulators with CDKs and cyclins.Attempts were made to identify proteins that associated with or acted as substrates for thekinase. We used two novel approaches for this study, fusion protein affinity columns andFar Western blotting.i. Recombinant GST-Pim-l fusion protein affinity columnsBecause of the initial lack of availability of a Pim- 1-specific immunoprecipitatingantibody, a novel approach was used to isolate regulator or substrate proteins that associatewith Pim-1 in the cell. Recombinant GST-Pim-1 bound to glutathione beads by the GSTglutathione affinity interaction were mixed with frog oocyte extracts to allow binding to thePim- 1 portion of the fusion protein. The slurry was poured into a syringe column, thebeads were washed extensively, removed from the column and subjected to analysis bySDS-PAGE electrophoresis. The affinity column extracts were examined for proteinbinding by several different methods. Silver staining was used to detect large amounts ofbound proteins. Immunodetection with various antibodies was used to screen for thepresence of specific proteins. Phosphorylation assays were done to determine if any of theproteins retained by the columns were substrates for GST-Pim- 1. As a control, cellularextracts were first applied to GST columns to remove any proteins that would bind nonspecifically to either GST or to the glutathione beads. To ensure that any of the proteinspresent were not co-purifying bacterial proteins, identical experiments were performed withPBS or STE buffer instead of cell extracts. Many combinations of cell extracts and affinitymatrix were examined with inconsistent results.256a. X. laevis oocyte extracts bound to human GST-Pim-1Before the X. laevis Pim-1 was cloned and expressed, immobilized human GST-Pim-1was used as an affinity matrix to bind proteins in the X. laevis oocyte extracts. Silverstaining detected a number of bands in all GST-Pim- 1 lanes including controls (Fig. 54,panel A). Unique bands of 25 kDa, 40 kDa, 55/60 doublet, 80 kDa and a 80/110 kDadoublet were observed in lanes corresponding to the immature and mature X. laevisextracts bound to the GST-Pim-1, but not in control lanes.The blot was probed with various antibodies directed against Raf (Raf1-CT, UBI) andcyclins A, D and E (from Dr. F. Hall, University of California, L.A.). Raf1-CT and cyclinD antibodies did not detect any bands, cyclin A antibody detected a 65-68 kDa singlet and a80 kDa triplet in all the X. laevis lanes, and the cyclin E antibody detected a 27.5 kDa bandin all lanes including controls [data not shown]. When this experiment was repeated,similar results with the cyclin A and E antibodies were not obtained.Phosphorylation assays revealed strongly autophosphorylated bands of 60 kDa, 30 and35 kDa in each lane corresponding to the full-length and degraded Pim-1 fusion protein, aswell as a faint band at 50 kDa (Fig. 54, panel B). In the immature and mature frog oocytesamples there were phosphorylated proteins of 25, 40 and 45 kDa. It is possible that the24 and 40 kDa phosphorylated proteins were the same ones as visualized by silver staining.257Figure 54. X. laevis oocyte extracts bound to H. sapiens GST-Pim-1fusion protein affinity columns.Control (lanes 1, 4, 7), inmiature X. laevis oocyte extracts (lanes 2, 5, 8)and mature X. laevis oocyte extracts (lanes 3, 6, 9) were precleared onGST (lanes 1-3) then bound to human GST-Pim-1 columns (lanes 4-9).Panel A shows a silver stain of the GST control and GST-Pim- 1 affinitymatrix, and panel B shows an autoradiograph of the autophosphorylatedGST-Pim-1 affinity matrix from lanes 4-6. Sharp arrows (—*)indicate X. laevis proteins bound to the GST-Pim-1, open arrows (<—)indicate phoshorylated GST-Pim-1-bound X. laevis proteins. Migrationsof Mr standards are shown on the left.258b. X. laevis oocyte extracts bound to X. laevis GST-Pim-1When the X. laevis Pim-1 fusion protein became available, it was used to bind proteinsin X. laevis oocyte extracts, with the rationale that this would be more representative ofphysiological protein-protein interaction. Silver staining of the extracts revealed thatseveral proteins bound to the GST controls including those of 30, 40 and 45 kDa (Fig. 55).An 85 - 90 kDa protein bound to the X. laevis GST-Pim- 1 but not to the control. In an invitro phosphorylation assay of proteins bound to beads, only the fusion protein and fusionprotein byproducts were phosphorylated; the 85-90 kDa protein was not. As well, noproteins were phosphorylated when the GST-Pim-1 was used to phosphorylate themembrane in a modified Far Western blot.The affinity column method initially seemed like a very promising approach to identifyproteins that associate with Pim- 1. Although this method worked well to confirm theinteraction of known proteins, it was not very suitable for detecting the interaction withunknown proteins. Screening bound proteins with a selection of antibodies proved not tobe very efficient unless a specific protein to which an antibody was available was expected.The phosphorylation assay detected Pim-1 substrates, but may not have detected regulatorsthat associate and modulate the activity of Pim- 1 independent of phosphorylation.259Figure 55. X. laevis oocyte extracts bound to X. laevis GST-Pim-lfusion protein affinity columns - silver stain.Control (lanes 1, 4), immature X. laevis oocyte extracts (lanes 2, 5)and mature X. laevis oocyte extracts (lanes 3, 6) were precleared onGST-glutathione agarose (lanes 1-3) then bound to X. laevis GSTPim-1 columns (lanes 4-6). Arrows (—) show X. laevis proteinsbound to the GST-Pim- 1 glutathione agarose. Migration of Mr standardsare shown on the left.260ii. Far Western BlottingA second method to detect potential physiological substrates of Pim- 1 was Far Westernblotting. This method has been successfully used by other groups in the past to identifyproteins that specifically interact [Kaelin et al., 1992]. This method relied on the ability ofexpressed kinase to recognize and phosphorylate denatured proteins on a membrane.Extracts of Mono Q-fractionated X. laevis oocytes were subjected to SDS-PAGE and weretransferred onto PVDF membrane. The membranes were then subjected to an “in vitro”autophosphorylation assay and autoradiography to detect any proteins that wouldautophosphorylate while bound to the PVDF. None were detected. The membrane wasthen incubated in the presence of radiolabelled ATP with large amounts of X. laevis GSTPim-1, extensively washed and then autoradiography was performed (Fig. 56). The arrowindicates the GST-Pim-1 used as a control.The strongest protein bands phosphorylated by GST-Pim- 1 were in the MonoQ washthrough fractions 8-13. These proteins were 33-35, 38, 53, 55, 66, and 90 kDa. Infractions 22/24 and faintly in 26/28 was a phosphorylated 42 kDa band that might be theendogenous Pim-1 protein, implying that Pim-1 can cross-phosphorylate. There was alsoa strongly phosphorylated group of proteins of 20-25 kDa of an unknown identity infractions 33-40 (0.46-0.65 M NaC1).An elegant aspect of the Far Western blotting procedure is that the phosphorylated blotcan be subsequently probed with antibodies to identify potential substrates. Unfortunately,as with the affinity column binding approach, this is method is most useful if a specificprotein is already suspected of being a potential substrate and does not work well forgeneral screening.A second variation of Far Western blotting was tested with less successful results.Radiolabelled, inactivated fusion protein was used to screen renatured Western blots whichwere then washed under conditions of low stringency and autoradiographed. Alternatively,antibodies specific for the fusion proteins could be used to detect the bound fusion protein.This method allows identification of protein-protein interactions between proteins that arenot necessarily substrates, an approach that could be especially useful for finding upstreamregulators of a protein. This method was attempted, but no signal was detected on the blotsand futher experiments along these lines were not performed.261—106 kDaFigure 56. Far Western blot of fractionated X. laevis oocytesphosphorviated by expressed GST-Pim- 1.OST-Pim- 1 positive control in far left lane, unfractionated X. laevisoocyte extract in far right lane. Numbers below indicate MonoQfractions loaded in each lane, in most cases several fractions werepooled. Arrow indicates location of GST-Pim-1 control. Migrationsof Mr standards is shown on the right. The autoradiogram shown wasexposed for 21 h.?‘ 00 © 00 f) t. — ,—— c— (• C”I •- .JC%o6 — — C’ m •‘.0I .‘ .‘0’0.02622. PRELIMINARY DATA FROM THE P. OCHRACEUS SYSTEMDuring the course of examination of Pim- 1 in sea star extracts, several other approacheswere employed. Although these preliminary studies were not completed, these approachesrevealed potentially productive areas of future Pim-1 research.i. GST-Pim-1 affinity columnsGST-fusion protein affinity columns can be utilized to help identify protein-proteininteractions. Using human GST-Pim-1, immature and mature sea star oocyte extracts wereexamined for proteins that bind to recombinant Pim- 1. A 48 kDa protein as well as severalhigher molecular weight species bound to the GST control lanes (Fig. 57). There were noproteins detectable by silver staining that selectively bound to the the GST-Pim- 1 column,despite the fact that very large amounts of sea star extracts (10 mg protein) were used.Although no GST-Pim- 1 binding proteins were detected, it is possible that silver stainingwas not sensitive enough to detect this interaction if the binding protein was rarelyexpressed and present in only low quantities.ii. Far Western blots of sea star oocyte maturationFar Western blotting was done with sea star oocyte extracts in an attempt to identifypotential substrates of Pim- 1. Cytosol from immature and mature sea star oocytes werefractionated on a Mono Q and Western-blotted. The blots were first incubated with ADBand [y32PJATP, washed and autoradiographed as a control. No bands were observed inthe control. Bacterially expressed human GST-Pim-1 was used to radiolabel the blots inthe presence of ADB and [‘y-32PJATP and -the blots were washed and autoradiographed.The control lane (GST-Pim-1) was phosphorylated, but unlike with the X. laevis oocytes,the sea star blots did not feature many radioactively labelled bands.263Figure 57. Sea star oocyte extracts bound toGST-Pim-1 fusion protein affinity column.PBS control (lanes 1 & 4), 10 mg of immature(lanes 2 & 5) and 10 mg mature (lanes 3 &6)sea star oocyte extract protein was preclearedon GST - glutatione beads (lanes 1-3) thenapplied to GST-Pim-1-glutatione beads(lanes 4-6). After washing, the entire beadsupernatant slurry was electrophoresed on aSDS-PAGE gel and silver stained. Migrationsof Mr standards are shown at right.1 23456GST GST-Pim-1‘—106 kDa—80264iii. Upstream activators of Pim- 1To determine whether there were any maturation-activated kinases in the sea star extractthat could phosphorylate Pim- 1, Mono Q-fractionated sea star extracts were used tophosphorylate deactivated H. sapiens GST-Pim- 1 bound to glutathione beads. The GSTPim- 1 was incubated with MonoQ-fractionated immature and mature sea star oocyteextracts in the presence of [“y32P]-ATP, washed extensively and Western blotted with Pim1 antibodies (Fig. 58). Panel A demonstrated that the amount of protein in each lane wasfairly constant. Panel B showed that mature sea star oocytes contained Pim-1phosphorylating kinases in fractions 28-32 (0.35-0.45 M NaC1). Immature oocytefractions contained no Pim-1 phosphorylating activity (data not shown). As sea star Pim-1was eluted primarily in fractions 23-26, this phosphotransferase activity was unlikely toresult from endogenous sea star Pim-1 phosphorylating the GST-Pim-1, but may be due tothe action of a unique maturation-activated Pim- 1 kinase (P1K).iv. In vitro fertilization studiesPreliminary studies were carried out by Dr. Diana Lefebvre in our laboratory toexamine Pim- 1 protein following in vitro fertilization of 1 -methyladenine-treated sea staroocytes. In brief, immature and mature oocyte samples as well as those corresponding to 6and 24 hours post-fertilization, and 2 and 4 day embryos were fractionated by MonoQcolumn chromatography. Western blots of the fractions were probed with Pim 1-NTantibody (Fig. 59). A 45 kDa band was observed in fractions 13/14 - 15/16, and wasdetected very strongly in fractions 23/24 -25/26 in the immature oocytes. This 45 kDaband was also observed in the mature extract, fractions 23/24 -25/26. The 45 kDa bandswere hardly detected in the 6 and 24 hour post-fertilized samples, were strongly evident inthe 2 day embryo, but not seen in the 4 day embryo. This implies that the Pim-1 proteinwas present in the immature and in the unfertilized, mature oocyte, but that the protein wasvirtually non-existent after fertilization and reappeared again in the 2 day embryo. ThePim-1 was then gone again in the 4 day embryo. These initial studies are quite promising,as the fluctuating Pim- 1 protein levels of indicate that the kinase may have a changing roleat different stages of early development.265Figure 58. Phosphorviation of GST-Pim-1 by a maturation-activated kinase.Panel A- Western blot of X. laevis GST-Pim- 1 phosphorylatedby MonoQ-fractionated mature sea star oocyte cytosolic extracts.Membrane was probed with 4G10 anti-phosphotyrosine antibody.Panel B- Autoradiogram of membrane from panel A. Westernblot and autoradiogram of GST-Pim- 1 phosphorylated by immaturesea star extracts was not shown. Migrations of Mr standards areshown on right.266Figure 59. Western blot of fractionated in vitro fertilized seastar oocytes.Immature (panel A), mature (panel B) and fertilized oocytesat 6 h (panel C), 24 h (panel D), 48 h (panel E) and 96 h (panel F)post-fertilization were fractionated on a MonoQ column andelectroluted on a 10% SDS-PAGE gel. MonoQ fractions areshown on the top of the blot. Blots probed with Piml-NT antibody.The 46 kDa Mr marker is shown on right. Western blots doneby Diana Lefebvre.c.’ c’+ +—C+D 00— ,- —++ +A\D 00 C’ ‘.0cs c”I c-+ + + + + +cn -00 C+ +N267

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share