Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of insulin-stimulated protein kinase cascade in normal and diabetic rats and the effects… Hei, Yong-jiang 1993

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

Item Metadata

Download

Media
831-ubc_1993_fall_phd_hei_yong-jiang.pdf [ 19.29MB ]
Metadata
JSON: 831-1.0098793.json
JSON-LD: 831-1.0098793-ld.json
RDF/XML (Pretty): 831-1.0098793-rdf.xml
RDF/JSON: 831-1.0098793-rdf.json
Turtle: 831-1.0098793-turtle.txt
N-Triples: 831-1.0098793-rdf-ntriples.txt
Original Record: 831-1.0098793-source.json
Full Text
831-1.0098793-fulltext.txt
Citation
831-1.0098793.ris

Full Text

CHARACTERIZATION OF INSULIN-STIMULATED PROTEINKINASE CASCADE IN NORMAL AND DIABETIC RATS ANDTHE EFFECTS OF VANADIUMbyYONG-JIANG HEIM.D., Shihezi Medical College, 1978M.Sc., West China University of Medical Sciences, 1983A thesis submitted in partial fulfillment ofthe requirement for the degree ofDoctor of PhilosophyinThe Faculty of Graduate StudiesDivision of Pharmacology and ToxicologyFaculty of Pharmaceutical SciencesWe accept this thesis as conformingto the required standardThe University of British ColumbiaAugust, 1993©Yong-jiang Hei, 1993In 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. I 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  PRA 1-P7.4^(24The University of British ColumbiaVancouver, CanadaDate^tc-irDE-6 (2/88)ABSTRACTMultiple protein kinases acting in a sequential manner have been identified toconstitute a kinase cascade that mediates growth factor or mitogen-activated signaltransduction. The cascade is triggered at the cell membrane by the activation ofreceptor tyrosine kinases, which pass on the signal in the order of p21 ras_.>p74raf3or mitogen-activated protein (MAP) kinase kinase kinase (MKKK)-MAP kinasekinase (MKK)-MAP kinase or extracellular-signal regulated kinase (ERK)->S6 kinase-3S6 protein phosphorylation. Insulin also activates such a cascade in cultured cellsor isolated tissues, but very few studies have investigated the kinase cascade in anintact animal model. In this study, MAP kinase, S6 kinase, p74ref/ and casein kinase-2 (CK-2) were characterized by chromatographic and immunoblotting techniquesusing both control and diabetic rats. The effects of insulin-mimetic vanadiumcompounds on these kinases were also examined.Intravenous injection of insulin activated the previously identified 42-kDa(p42erk2) and 44-kDa (p44erk 1) MAP kinases as well as other potential novel MAPkinases in rat skeletal muscle and spleen. p42erk2 appears to represent a majorportion of the insulin-stimulated MAP kinase activity. Also activated in rat skeletalmuscle following insulin injection were S6 kinases and p74ref/ . A 100-kDa S6 kinasewas tentatively identified as a member of the previously identified 90-kDa ribosomalS6 kinase (p9Orsk) family, while another novel S6 kinase activity was also identifiedand purified to near homogeneity as a potentially novel 32-kDa S6 kinase. Thiskinase appears to play a major role in the insulin-stimulated activation of S6 kinaseactivity. CK-2 was not significantly activated by insulin in skeletal muscle, but wasactivated significantly in rat spleen. These results suggest that the kinase cascademay be physiologically significant and that the intact rat may serve as a useful modelin characterizing these kinases.In streptozotocin (STZ)-induced diabetic rats, basal MAP kinase and S6 kinaseactivity were elevated in 2-month diabetic rats, but significantly reduced in 6-monthdiabetic rats. Furthermore, activation of MAP kinase and S6 kinase was defective inthe 2-month diabetic rats, indicating a disruption of the kinase cascade that mediatesthe action of insulin during diabetes. The molecular mechanism of these changesrequires further investigation. When the diabetic rats were treated with vanadylsulphate or an organic compound of vanadium, bis-maltolato-oxo-vanadium (IV)(BMOV), the changes in MAP kinase and S6 kinase did not occur, suggesting that theinsulin-like effects of vanadium may involve the effects of vanadium on the kinasesthat are critical components in the mitogen-activated kinase cascade.TABLE OF CONTENTS Content^ PageABSTRACT iiTABLE OF CONTENTS^ ivLIST OF FIGURES viiiLIST OF TABLES^ xiLIST OF ABBREVIATIONS^ xiiLIST OF AMINO ACID CODES xviiACKNOWLEDGMENT^ xviiiDEDICATION^ xixINTRODUCTION 1I. Cell biology and physiology of insulin^ 11. Biosynthesis and secretion of insulin 12. Physiological function of insulin^ 4A) Glucose transporters 4B) Glucose and glycogen metabolism^ 8C) Metabolism of fatty acids^ 12D) Regulation of gene expression 14II. Insulin receptor (IR)^ 151. Insulin receptor structure and function^ 152. Processing of IR and modulation of IR function^23III. Mechanism of post-insulin receptor signal transduction 251. Substrate of insulin receptor tyrosine kinase (IRTK)^262. S6 phosphorylation and S6 kinases^ 283. MAP kinases^ 354. MAP kinase or ERK kinase kinase (MEK)^ 445. Raf and other MEK kinases^ 476. p21 ras in insulin signaling 51iv7. CK-2 and other insulin-activated kinases^ 568. G proteins in insulin signalling^ 619. Second messenger theory 6210. An integrated view^ 63IV. Insulin signalling in diabetes and insulin resistance^651. Overview of diabetes^ 652. Etiology and pathogenesis of diabetes^ 673. Cellular and molecular defects in diabetes 70and insulin resistanceV. Vanadium: insulin mimetic effects and mechanism of action^741. Biochemical properties and insulin-like effects of vanadium^752. Mechanism of action of vanadium^ 79VI. Rationale, hypotheses and objectives 831. Hypotheses^ 842. Objectives 84MATERIALS AND METHODS^ 86I. Materials^ 86II. Experimental procedures^ 871. Preparation of animals 872. Euglycemic clamp procedures^ 883. Determination of protein concentrations^ 894. Preparation of 40S ribosomes^ 905. Preparation of crude rat skeletal muscle extracts^916. SDS-polyacrylamide gel electrophoresis^ 947. Protein kinase assays^ 96V8. Fractionation of muscle extracts by column chromatography^979. Immunoblotting^ 9810. Purification of S6 kinase^ 10111. Phosphoamino acid analysis 10312. Statistical analysis^ 104RESULTS^ 105Part I. Insulin-stimulated protein kinases in normal rats^105I. Assay validationII. Characterization of insulin-stimulated Ser/Thr protein kinasesin rat skeletal muscle^ 1051. Time course of glucose concentration^ 1052. Insulin-induced tyrosine phosphorylationin rat skeletal muscle^ 1123. Effects of insulin injection on proteinkinase activities in rat skeletal muscle^ 1124. MonoQ column chromatography of MBP kinases^1205. Identification of MBP kinase isoforms byimmunoblotting and column chromatography 1206. Insulin-stimulated S6 kinases and its column profile^1337. Immunoblotting of MonoQ column fractionswith antibodies for S6 kinases^ 1388. Gel filtration of S6 kinases 1419. MonoQ chromatography and immunoblotting ofraf kinase in rat skeletal muscle^ 14410. MonoQ column profile and immunoblottingof CK-2 in rat skeletal muscle. 14711. Insulin-stimulated kinases in euglycemic clamp study12. Tissue distribution of kinases^ 150viPart II. Insulin-stimulated protein kinases in diabetic ratsand the effects of vanadium treatment^ 155I. General characteristics of the animals 1551. One-month study^ 1552. Two-month and six-month study^ 158II. Protein kinase activities in the crude muscle extracts^1641. One-month control and diabetic rats^ 1642. Two-month control and diabetic rats 1673. Six-month control and diabetic rats^ 167III. Chromatographic fractionation and immunoblotting^1721. Two-month control and diabetic rats^ 1722. Six-month control and diabetic rats 182Part III. Purification of a potentially novel S6kinase from rat skeletal muscle^ 1901. Purification of S6 kinase 1902. Properties of the purified kinase^ 1943. Substrate specificity of the purified kinase^ 1974. Modifiers of the purified kinase^ 2065. Regulation of the kinase by phosphorylation^210DISCUSSION^ 2161. Insulin-stimulated protein kinases in normal ratskeletal muscle^ 2162. Insulin-stimulated protein kinases in diabeticrats and the effects of vanadium^ 2343. Purification of a novel S6 kinase from rat skeletal muscle^2444. Future directions^ 248SUMMARY AND CONCLUSIONS^ 250BIBLIOGRAPHY^ 252viiLIST OF FIGURESPageFigure1. Structure of insulin receptor 172. Mitogen-stimulated protein kinase cascade 363. Isolation of 40S ribosome by sucrose gradient centrifugation 924. Calibration of Superose 12 and Superdex 200columns for gel filtration 995. Time course of protein kinase assays 1066. Protein concentration curve of protein kinase assays 1087. Time course of glucose levels after insulin injection 1108. Insulin-stimulated protein tyrosine phosphorylationin rat skeletal muscle 1139. Insulin-induced activation of protein kinasesin crude skeletal muscle extracts 11510. Immunoblotting of skeletal muscle extractswith antibodies against protein kinases 11811. MonoQ column chromatography of insulin-regulated MBP kinases and their timecourse of activation 12112. Immunoblotting of MonoQ-fractionated MBP kinases 12313. Sequential phenyl-Sepharose (PS) and MonoQfractionation of MBP kinases 12614. Immunoblotting of the MBP kinases resolvedsequentially on phenyl-Sepharose and MonoQ 12915. Phenyl-Superose chromatography andimmunoblotting of MBP kinases. 13116. MonoQ chromatography of S6 kinases 13417. MonoQ chromatography of S6 peptide kinases 13618. Immunoblotting of S6 kinases resolved on MonoQ 13919. Gel filtration of the MonoQ S6 kinase peaks 14220. MonoQ chromatography and immunoblotting of p74raf 1 14521. MonoQ column chromatography and immunoblottingviiiof casein kinase-2^ 14822. MonoQ chromatography of insulin-regulated MBP kinasesin skeletal muscles of glucose-clamped rats^ 15123. Tissue distribution of protein kinases 15324. MonoQ chromatography of spleen MBP kinase andcasein kinase-2^ 15625. Time course of glucose levels in control anddiabetic rats in an one-month study^ 15926. The dose of vanadyl sulphate and body weight ofcontrol and diabetic rats in a two-month study^ 16027. Time course of glucose levels in control anddiabetic rats in a two-month study^ 16228. Protein kinase activities in crude muscle extractsfrom 1-month control and diabetic rats 16529. Protein kinase activities in crude muscle extractsfrom 2-month control and diabetic rats^ 16830. Protein kinase activities in crude muscle extractsfrom 6-month control and diabetic rats 17031. MonoQ profile of MBP kinase in the 2-month control,diabetic, and vanadyl sulphate-treated diabetic rats^17332.^Immunoblotting of MAP kinase in the MonoQfraction of 2-month control, diabetic andvanadyl sulphate-treated diabetic rats^ 17533^MonoQ profile of S6 kinase in 2-month control,diabetic, and vanadyl sulphate-treated diabetic rats^17834. Immunoblotting of S6 kinase in the MonoQfraction of 2-month control, diabetic andvanadyl sulphate-treated diabetic rats^ 18035. MonoQ profile of MBP kinase and S6 kinase in 6-month rats^18336. Immunoblotting of MAP kinase in the MonoQ fractionsof the skeletal muscle extracts from 6-month control,diabetic and BMOV-treated diabetic rats^ 18537. Immunoblotting of S6 kinase in the MonoQ fractionsof the skeletal muscle extracts from 6-month control,diabetic and BMOV-treated diabetic rats^ 18838.^Chromatographic procedures in the purificationof a potentially novel S6 kinase^ 191ix39. SDS-PAGE and silver stain of S6 kinase peak fractionsfrom the Superose 12 column and separation of S6kinase from PKA by heparin-agarose 19540. Determination of kinetic parameters of purified S6 kinase 19841. Time course of S6 phosphorylation by purified S6 kinase 20042. Requirement for divalent metal ions of purified S6 kinase 20143. Phosphorylation of protein substrates by purified 20244. Phosphoamino acid analysis of S6 proteinphosphorylated by the purified S6 kinase 20445. Phosphorylation of peptide substrates by purified S6 kinase 20746. Phosphatase 2A inactivation of purified S6 kinase 21147. Effects of MAP kinase and CK-2 on the activityof purified S6 kinase 21248. Interaction of MAP kinase and CK-2 with purified S6 kinase 214LIST OF TABLESTables^ Page1. Stimulators of MAP kinases^ 172. Putative substrates of MAP kinases^ 363. Characteristics of Type 1 and Type 2 diabetes^664. S6 kinase purification table^ 1935. Modifiers of S6 kinase 209LST OF ABBREVIATIONSACC^acetyl CoA carboxylase;ASH^abundant Src homologyATP^adenosine triphosphateBB^BioBreedingBCIP^5-bromo-4-chloro-indoly1 phosphateBMOV^bis-maltolato-oxovanadium (IV)BSA^bovine serum albuminCAT^chloramphenicol acetyl transferaseCHL^chinese hamster lungCHO^chinese hamster ovaryCK^casein kinaseCSF-1^colony-stimulating factorDG^diacyl glycerolDMF^N,N-dimethylformamideDMSO^dimethylsulfoxidedrk^a homolog of Grb2 adaptor protein in DrosophilaDTT^dithiothreitolEDTA^ethylene diaminetetraacetic acidEGF^epidermal growth factorEGTA^ethylene glycol bis(b-aminoethyl ether-P-N,N,N'N'-tetraacetic acidERK^extracellular signal-regulated kinaseET^endothelinFBP^fructose-1,6-bisphosphatasefMLP^formayl-Met-Leu-PheFSAB^5'-p-fluorosulfonylbenzoyadenosinexiig^gramsG6P^glucose-6-phosphateG6Pase^glucose-6-phosphataseGAP^GTPase activating proteinGDI^GDP dissociation inhibitorGDP^guanine diphosphateGDS^GDP dissociation stimulatorGH^growth hormoneGip2^a mutant form of Gi proteinGLUT,^glucose transporterGP^glycogen phosphorylaseGPI^glycosyl-phosphatidylinositolGrb2^Growth factor receptor-bound protein 2GRF^guanine nucleotide releasing factorGS^glycogen synthaseGSK^glycogen synthase kinaseGTP^guanine triphosphateh^hourHLA^human leukocyte-associated antigen1-1^inhibitor-1IDDM^insulin-dependent diabetes mellitusIGF-1^insulin like growth factor-1IL^interleukinIP3^inosito1-1,4,5-trisphosphateIR^insulin receptorIRE^insulin response elementIRS-1^insulin receptor substrate-1IRTK^insulin receptor tyrosine kinaseLDL^low-density lipoproteinLPS^LipopolysaccharidesMAP^mitogen-activated proteinMAP-2^microtubule-associated protein-2MARCKS^myristoylated alanine-rich C kinase substrateMBP^myelin basic proteinMEK^MAP or ERK kinaseMEKK^MEK kinaseMHC^major histocompatibility complexmin^minuteMPF^Maturation promoting factorMr^molecular mass;NBT^nitroblue tetrazoliumNGF^nerve growth factor,NIDDM^non-insulin-dependent diabetes mellitusNOD^nonobesep42erk2^the 42-kDa MAP kinase encoded by ERK2 gene, also referred to asp42maPkp44erk1^^the 44-kDa MAP kinase encoded by ERK1 gene, also referred to asp44maPkp44mPk^the 44-kDa sea star MBP kinase, also a MAP kinase isoformp95vav^a mammalian GRF encoded by oncogene vavPAGE^polyacrylamide gel electrophoresisPAK-II^protease-activated kinase IIPAO^phenylarsenoxidePDGF^platelet-derived growth factorxivPDH^pyruvate dehydrogenasePEPCK^phosphoenopyruvate carboxyl kinasePFK^6-phosphofructose-1 kinasePhK^phosphorylase kinaseP13K^phosphatidylinositol-3 kinasePK^pyruvate kinase^•PKA^cyclic AMP-dependent protein kinasePKC^protein kinase CPKG^cyclic GMP-dependent protein kinasePKI^peptide inhibitor of PKAPLC^phospholipase CPMA^phorbol 12-myristate 13-acetatePMSF^phenylmethylsulfonyl fluoridePP^phosphatasepp15^the 15 kDa fatty acid binding protein 422(aP2)PP1 G^the G subunit of PP1PTF1^pancreatic transcription factor 1PTPase^protein tyrosine phosphatasePTX^pertussis toxinREKS^ras-dependent ERK-kinase stimulator;rsk^ribosomal S6 kinases^secondss6k^S6 kinaseSDS^sodium dodecyl sulfateSem5^a homolog of Grb2 adaptor protein in Caenorhabditis elegansSev^sevenlessSH2^src homology domain 2SH3^src homology domain 3smg^small molecular weight G proteinSOS^son of sevenlessSp-1^specificity protein-1Src^an oncogene isolated from chicken Rous sarcoma virusSRE^serum response elementSTZ^strecptozotocinTCA^tricarboxylic acidTCF^ternary complex factorTIG^human lung fibroblastsTNF^tumor necrosis factorWGA^wheat germ agglutininLIST OF AMINO ACID CODESName^Three letter code^One letter codeAlanine^Ala AArginine Arg^ RAsparagine^Asn NAspartic acid Asp^ DCysteine^Cys CGlycine Gly^ GGlutamic acid^Glu EGlutamine GlnHistidine^His^ Hlsoleucine lieLeucine^Leu^ LLysine Lys KMethionine^Met^MPhenylalanine^Phe FProline^Pro^ PSerine Ser SThreonine^Thr^ TTryptophan TrpTyrosine^Tyr^ YValine Val VxviiACKNOWLEDGMENTS I would like to express my sincere gratitude to my supervisors, Dr. Jack Diamondand Dr. John H. McNeill, for their continuous moral support and scientificguidance in my graduate studies. I deeply appreciate their understanding andconsideration at difficult times during the course of my Ph.D. training.My special thanks go to Dr. Steve Pelech, my unofficial supervisor, for hisscientific advise, inspiration, enthusiasm, and his assistance in my thesis work.I would like to thank my Ph.D. thesis committee members, Dr. Kath MacLeod,Dr. Sid Katz, Dr. Don Lyster, and Dr. Marc Levine for their assistance andsupport.I am appreciative of Dr. Jasbinder Sanghera and Mr. Harry Paddon in theBiomedical Research Center for their assistance at the beginning of my thesiswork.I am very grateful to Ms. Xunsheng Chen for her technical assistance in theexperiments for my thesis and for her friendship. I am also grateful to mycolleagues and summer students who have made direct contributions to mythesis work. They are: Mr. Sep Farahbakhshian, Ms. Sibongile Bool, Mr. BrianWardrobe, Mr. !man Latifpour, Ms. Nancy Leung, and Mr. Jeff Moore.I appreciate very much the kindliness of Ms. Violet Yuen for her help in someexperiments and for providing me with the skeletal muscle samples from ratsthat were treated with vanadium for six-months.I would like to express my appreciation to our friendly lab manager Ms. MaryBattell for her consistent support in my work. Ms. Battell always manages toprovide me with help when it was needed. I thank Ms. Stephenie Lee for herassistance in some experiments and Ms. Sylvia Chan for her secretarialassistance.I am appreciative of all my friends for their moral support. In particular, I wouldlike to thank Ms. Karen MacDonald, Dr. Brian Rodrigues, Ms. Margaret Cam,and Mr. Ashwin Patel for their friendship. I also thank the support and friendshipof Dr. Margaret Prang, Dr. Maria Furstenwald, and Ms. Lynda Cowan for theirsupport and friendship.I would like to acknowledge the financial support from Stroke and HeartFoundation and B.C. and Yukon (1989-1990) and Medical Research Council ofCanada (1991-1993).Finally, I would like to thank all the faculty members, staffs and graduatestudents for making my Ph.D. training most enjoyable.xviiiDEDICATIONTo a great person, my friend and teacher, Mu Li.xixINTRODUCTIONInsulin is a polypeptide hormone which plays an essential role in themaintenance of some cellular functions. While being considered an endocrinehormone in higher animals, insulin was found to arise very early in evolution, existingnot only in mammals, but also in very primitive vertebrates and non-vertebrates. Aninsulin-related molecule was found even in unicellular eukaryotes like fungi andprokaryotes like protozoan Tetrahymena (LeRoith et aL, 1987; 1989). Theimportance of insulin has been well appreciated since its discovery in 1922 byCanadian scientist Banting and Best, and great advances and progress have beenmade in our understanding of almost every aspect of the physiology and biology ofinsulin action. I will summarize only some of the most important studies that arerelevant to this thesis.I. CELL BIOLOGY AND PHYSIOLOGY OF INSULIN1. Biosynthesis and Secretion of Insulin.The biologically active insulin molecule is a 6 kDa protein consisting of an Achain of 21 amino acids and a B chain of 30 amino acids. Disulfide bonds are foundto link A and B chains (A7-B7 and A20-B19) and to form a loop within the A chainbetween A6-A11 (Derewenda et al., 1990). Biosynthesis of insulin occurs mainly inpancreatic n-cells, which reside in the islet of Langerhans, although the nervoussystem may be an extrapancreatic site of insulin synthesis (LeRoith et a!., 1987;1989). The expression of the insulin gene results in an initial precursor of insulinknown as preproinsulin, a 12 kDa single-chain polypeptide consisting of a 9 kDaproinsulin and a 24 amino acid signal peptide at the N-terminus (Chan et al., 1976).1The signal peptide is apparently involved in directing the entrance of preproinsulin intothe rough endoplasmic reticulum (RER), where the signal peptide is immediatelycleaved, and proinsulin molecules fold to form disulfide bonds. Proinsulin istransported into the Golgi apparatus where it undergoes passage from the cis-cisternae and mid-cisternae to trans-cisternae, and is concentrated into theprosecretory granules which are pinched off from the clathrin-coated trans-cisternae(Steiner 1990). As the environment in the granules acidifies with maturation,cleavage enzymes remove the C-peptide to give rise to the formation of the finalproduct, insulin (Orci et al., 1988). It should be noted, however, that C-peptide is notrequired for folding and formation of disulfide bonds in insulin molecules, althoughcleavage of C-peptide occurs after the formation of disulfide bonds (Wang and Tsou,1991)The expression of the insulin gene is regulated in a cell-specific fashion due tothe presence of a cell-specific enhancer and a cell-specific promoter in the insulingene (Espinal, 1989). Glucose represents a major positive signal that exerts positiveregulation of the insulin gene expression, at the level of both transcription (Walker1990) and translation (Steiner, 1990), probably via Ca 2+ (Efrat et al., 1991). Specificsequence elements in the 5'-flanking region of the insulin gene have been defined forthe positive or negative regulation of gene expression (e.g., the glucose responsiveelement and repressor protein binding element), whereas no definitive "transcriptionfactors" have been identified for the regulation of the insulin gene. In cells that do notproduce insulin, repressor proteins may be present to inhibit the expression of theinsulin gene (Espinal, 1989), which is mediated by multiple negative-acting controlelements identified in the promoter of the insulin gene (Cordle et al., 1991).2The secretion of insulin from pancreatic 13-cells is subject to fine regulation bythe glucose concentration in the interstitial fluid and by neurohumoral factors (Howelland Bird, 1989). Two major hypotheses have been proposed to explain themechanism of glucose-induced insulin secretion, i.e., the presence of a so-called"glucoreceptor' in the membrane of the 13-cells, and the generation of intracellularsignals by the uptake and utilization of glucose in p-cells (Espinal, 1989). Theproposed "glucoreceptor", however, has never been identified biochemically, whilerecent studies have provided strong support for the latter hypothesis. Usingtransgenic techniques, Epstein et al. (1992) demonstrated that the serum insulin levelwas increased in transgenic mice overexpressing p-cell hexokinase in spite of a 20-50% lower glucose concentration. On the other hand, mutations of the glucokinasegene leading to a lower glucokinase activity have been associated with diabetescharacterized by defective insulin secretion (Gidh-Jain et al., 1993). The pancreatichexokinase is of Type IV, also known as glucokinase, and catalyzes phosphorylationof glucose, a rate-limiting step in glucose utilization (Magnums, 1990). Thus, the (3-cell hexokinase (glucokinase) acts as a glucose sensor, and metabolism of glucoseresulted in insulin secretion (Mueckler, 1993).The proximal signals generated by glucose metabolism for insulin secretionhave not been identified (MacDonald, 1990). It has been suggested that ATPproduction by glycolysis changes the ratio of ATP/ADP, which closes the ATP-sensitive K+ channels causing hyperpolarization of cell membrane, resulting in Ca 2+influx and insulin secretion (Espinal, 1989; Malaisse, 1990; Rajan et al., 1990; Holzand Habener, 1992). This prevailing hypothesis, however, has been challenged byobservations that the ratio of ATP/ADP did not change significantly over an increasein glucose levels that induced a 10-fold increase in insulin release (Ghosh et al.,1991). This hypothesis must therefore be viewed with caution. However, well-known3second messengers such as cAMP, inosito1-1,4,5-trisphosphate (IP3) and diacylglycerol (DG), and protein kinase C (PKC) are probably involved in the secretion ofinsulin triggered by neuronal and humoral stimulators (Rasmussen et al., 1990;Malaisse, 1990), although the role of these signalling molecules in glucose-stimulatedinsulin secretion is still controversial (Regazzi et al., 1990 Wang et al., 1993).2. Physiological Function of InsulinThe main function of insulin is to regulate substrate metabolism. This is bestexemplified in the homeostasis of glucose, which is maintained by a fine balancebetween the production and disposal of blood glucose. The major target tissues forinsulin are liver, skeletal muscle, and adipose tissues. In addition to glucose, insulinhas profound effects on the metabolism of fat and protein. The dynamic integration ofthe metabolism of the three major sources of energy in the body is achieved by theconcerted efforts of insulin and other neurohumoral mechanisms.A) Glucose Transporters. The first step in the stimulation of glucosemetabolism by insulin is to increase glucose uptake into the cell. This is mediated byfacilitative carriers in the cell membrane known as glucose transporters (Simpson andCushman, 1985). Studies in the past few years have established that glucosetransporters are a group of structurally-related but genetically distinct proteins, someof which are subject to rigorous regulation by insulin, nutritional state of the cells, anda variety of other factors (reviewed in Kahn and Flier, 1990; Silverman, 1991; Pessinand Bell, 1992). Five major isoforms of glucose transporters (GLUT1-GLUT5) thatexhibit tissue specific expression have been identified. GLUT1 is primarily expressedin erythrocytes, GLUT2 in liver and pancreas, GLUTS in brain, GLUT4 in muscle andfat, and GLUT5 in small intestine (reviewed in Kahn and Flier, 1990; Silverman, 1991;4Pessin and Bell, 1992). The cDNAs for these transporters have been cloned andcharacterized, which reveals a 39-60% identity and 50-76% similarity between theamino acid sequence of the various isoforms of glucose transporters, although theyarise from different genes located on different chromosomes (Mueckler, 1990;Silverman, 1991). The major type of glucose transporter that is regulated by insulin isGLUT4, which translocates from intracellular vesicles to plasma membrane inresponse to insulin. Over 90% of glucose transporters in rat adipose tissues areGLUT4, 95% of which are intracellular under basal conditions. In response to insulin,there is as much as a 20-fold increase in glucose transport activity in adipocytes,which correlates well with a 13-fold increase in the number of plasma membraneGLUT4 transporters, and is concomitant with 40% decrease in the intracellular pool ofGLUT4 transporters (Smith et al., 1991; Silverman, 1991). Similar observations havebeen made in rat and human skeletal muscle (Fukumoto et al., 1989; Friedman et al.,1991; Rodnick et al., 1992). In contrast, GLUT1 represents 3-5% of the total glucose.transporters in adipocytes, the amount of which increases by only about 5-fold in theplasma membrane in response to insulin treatment, suggesting that GLUT4 isresponsible for most of the increase in the insulin-stimulated glucose transport activity(Simpson and Cushman, 1986).In tissues such as blood cells and brain cells, which are not very sensitive toinsulin stimulation, the predominant forms of glucose transporters are GLUT1 andGLUT3. These transporters have a high affinity for glucose (K m=6.9 mM), thereforethe amount of the transporters in the cell membrane determines the rate of glucosetransport. In contrast, GLUT2 has a high capacity but low affinity (K m=13.2 mM) forglucose, and the expression of GLUT2 is limited to liver, pancreas, kidney andintestines (Kahn and Flier 1990). Therefore, the glucose transport activity in these5tissues largely depends on the glucose concentration, a feature well suited for theglucose sensoring function of liver and pancreas (Pessin and Bell, 1992).The function and expression of glucose transporters are regulated by a varietyof factors. Certain oncogene (e.g., ras or src)-induced cell transformation wasassociated with increased transcription of GLUT1, and stimulation of cells with growthfactors such as platelet-derived growth factor (PDGF) and insulin also resulted in aincreased level of GLUT1 mRNA in fibroblasts (Rollins et al., 1988; Flier andMatsouka 1989; , Kahn and Flier, 1990). Treatment of mature 3T3-F442A adipocytesinduced a dose-dependent decrease in the level of GLUT1 mRNA without affectingthe level of GLUT4 mRNA (Tai et aL, 1990). Chronic insulin treatment for 4 days inthe same cell line also failed to change the level of GLUT4 mRNA (Hainque et al.,1991), while the expression of GLUT4 was down-regulated by insulin in 3T3-L1 pre-adipocytes (Kozka et al., 1991; Flores-Riveros et al., 1993) and L6 muscle cells(Koivisto et al., 1991). These observations suggest that the expression of glucosetransporters may be regulated in a cell- and tissue-specific manner, and that celldifferentiation may play a role in regulating the expression of glucose transporters.This notion is supported by an 8-10-fold increase in insulin-stimulated glucosetransport activity with the differentiation of murine fibroblast 3T3-L1 cells to adipocytelike 3T3-L1 cells (Weiland et al., 1990). Additionally, a striking difference in thelocalization of glucose transporter transcripts was also observed in early post-implantation mouse embryos (Smith and Gridley, 1992).Glucose was shown to be another important regulator of glucose transportergene expression. The level of GLUT1 was increased by glucose starvation in 3T3-L1adipocytes (Kozka et al., 1991), pancreatic islet 13-cells (Tal et al., 1992), as well as inL6 muscle cells (Koivisto et al., 1991), while GLUT4 was increased only in the 3T3-L16cells (Kozka et al., 1991). Conversely, insulin-stimulated glucose transporter activitywas reduced by high glucose concentrations (Garvey et aL, 1987). It is not knownwhich of the isoforms of the glucose transporters were affected in this situation,although levels of the GLUT1 mRNA appears unchanged (Kahn and Flier, 1990).Regulation of gene transcription would apparently involve promoters of thegenes of interest. While potential binding sites for known transcription factors wereidentified in the 2.0 kb promoter region of the GLUT4 gene, transfection of GLUT4promoter constructs with chloramphenicol acetyl transferase (CAT) reporter gene intoadipocytes failed to reveal strong promoter activity (Buse et al., 1992). In transgenicmice carrying a 2.4 kb sequence of the 5'-flanking region of the GLUT4 gene,however, the tissue-specific expression of GLUT4 and certain responses to hormonaland metabolic regulation were observed, suggesting that this region contains thesequence necessary for the regulation of GLUT4 expression (Liu et al., 1992). Theexact mechanism for the regulation of gene expression of glucose transporters stillremains unclear.In addition to transcriptional and translational regulation, the glucosetransporters are regulated at the post-translational level as well. The proteinsynthesis inhibitor cycloheximide was shown to inhibit the translocation of glucosetransporters to plasma membrane but the insulin-stimulated glucose transport activitywas not affected by the treatment (Matthaei et al., 1988). Further studies revealedthat the intrinsic catalytic activity of GLUT4 in 3T3L-1 cells was increased bycycloheximide treatment (Clancy et al., 1991). Interestingly, the intrinsic activity ofglucose transporters may also be regulated by phosphorylation, although this is stillhighly controversial (Czech et al., 1992). Isoproterenol was shown to inhibit glucosetransport without altering the level of phosphorylation or the concentration of glucose7transporters in the plasma membrane, neither was insulin stimulated glucosetransport associated with a change in the level of phosphorylation of glucosetransporters (Joost et al., 1987; 1988; Clancy and Czech, 1991). Conversely, phorbolesters led to phosphorylation of glucose transporters, without altering theirtranslocation or transport activity (Joost et al., 1988). However, evidence indicatingexactly the opposite is also available. Cyclic AMP analogs were shown to inhibit theintrinsic activity of GLUT4 in adipocytes, presumably via the cAMP-dependent proteinkinase (PKA) (Bonen et al., 1992; Kelada et al., 1992). The use of okadaic acid, aninhibitor of phosphatase-1 (PP1) and phosphatase-2A (PP2A), also yieldedcontroversial results on the relationship between phosphorylation, intrinsic activity,and translocation of glucose transporters. While treatment of adipocytes with okadaicacid alone stimulated the translocation of GLUT4, insulin-induced GLUT4translocation was blocked by okadaic acid (Lawrence et al., 1990; Corvera et aL,1991). The intrinsic activity of GLUT4 did not appear changed by the treatment ineither of the studies. Thus, considerable confusion exists with regard to the effect ofphosphorylation on glucose transporters. However, recent studies provide evidencesupporting an inverse relationship between PKA- or Ca 2+-dependent GLUT4phosphorylation and its intrinsic activity (Reusch et al., 1993; Begum et al., 1993).B) Glucose and Glycogen Metabolism. Once inside the cell, glucose can bedisposed of via one of several metabolic pathways depending on the tissue and celltype. In muscle and brain cells, the main route of metabolism for glucose is oxidationfor energy production, while in fat cells glucose is used to synthesize triglycerides forenergy storage. In the liver, the fate of glucose depends on the physiologicalconditions and the level of glucose in the circulation, but the synthesis anddegradation of glycogen and gluconeogenesis occur mainly in the liver, althoughmuscle cells also participate in the synthesis of glycogen as an energy store. The co-8ordinated regulation of glucose metabolism in different tissues and cells is intricatelycomplex, and involves neurohumoral factors as well as the various intermediateproducts generated in the metabolism of not only glucose, but also fat and proteins.Considering the relevance of this information to the thesis, only important aspects ofinsulin-related regulation will be discussed.Without going into the detailed pathway of glycolysis, the very first step in theoxidation of glucose is phosphorylation of glucose catalyzed by a group of enzymesknown as hexokinases. These enzymes are not significantly regulated by insulin, butinsulin does regulate the activity of an isoform of the enzyme in liver and pancreas,which is referred to as glucokinase (GK, type IV or D hexokinase, Magnums, 1990).These enzymes have a K m for glucose that is 1000 times higher than that of the otherhexokinases (>10 mM), and are therefore considered to be active only at highambient glucose concentrations, such as after a meal (Cornish-Bowden andCardenas, 1991). Insulin deficiency induces a severe decrease in the enzyme activityof glucokinase and a resultant impairment in the utilization of glucose in the liver(Weinhouse, 1976). However, the hexokinases in other tissues such as muscle andadipose tissues have a very low Km value for glucose, which would maintain a largegradient of glucose across the membrane and a constant flux of glucose-6-phosphate(G-6P) to feed the glycolytic pathway. In these cells, glucose transport through themembrane is the limiting step and the hexokinases are little affected by hormonessuch as insulin (Espinal, 1989).Glucose is degraded to supply energy for cells when needed, and ispolymerized to form glycogen when in excess. The metabolism of glycogen isaccomplished by two major enzymes, glycogen synthase (GS) and glycogenphosphorylase (GP) (reviewed in Hers, 1976; Madsen, 1986; Cohen, 1986; Roach,91986). Understandably, these enzymes are regulated allosterically by the availabilityof their substrates and concentrations of their products. But more importantly, theyare regulated coordinately by neurohumoral factors via phosphorylation, whichactivates glycogenolytic enzymes and inhibits GS (Soderling and Sheorain, 1985;Cohen et al., 1985; Cohen, 1989). GP is the first enzyme shown to have allosteric aswell as covalent regulation, and the structure and function of GP has been studied indetail (reviewed in Madsen, 1986; Johnson and Barford, 1990; Johnson, 1992;Browner and Fletterick, 1992). In the classical phosphorylation cascade establishedby Krebs and Fischer, cAMP-elevating agents activate PKA, which phosphorylatesand activates phosphorylase kinase (PhK), which in turn phosphorylates and activatesGP (Krebs, 1986, 1989). On the other hand, insulin decreases glycogenolysis byinhibiting GP, partly through inhibition of cAMP formation and subsequent decrease ofPKA activity (Stalmans, 1976; Hems and Whitton, 1980). In contrast, the regulation ofGS differs significantly from that of GP in that phosphatases appear to play a criticalrole (Miller 1985; Soderling and Sheorain, 1985). GS is phosphorylated on at leastnine sites by a minimum of 8 different enzymes collectively known as GS kinase(GSK), among which are PKA (GSK-1), cyclic GMP-dependent protein kinase (PKG),phosphorylase kinase (GSK-2), casein kinase 2 (CK-2, GSK-5), GSK-3, and GSK-4(Cohen et al., 1985; Woodgett, 1991). The most important sites appear to betargeted by GSK-3, which also phosphorylates a variety of other proteins that areintimately involved in transcriptional regulation (Woodgett, 1991). GSK-3 does notseem to be activated by hormones such as insulin, therefore the activity of GSK-3helps to keep GS in its inactive state, and the regulation of GS is realized throughphosphatases that remove phosphate from the GS. The major GS phosphatase hasbeen identified as a PP1 type, with the regulatory G subunit bound to glycogen(Cohen, 1986, Cohen and Cohen, 1989). Interestingly, while both catecholaminesand insulin are capable of phosphorylating PP1 G, their effects on the activity of PP110toward GS are opposite, apparently due to differences in the sites of phosphorylation.Catecholamines inactivate GS by activating PP1 G, which is mediated by PKA-inducedphosphorylation of the site-1 and site-2 in PP1G, while insulin specificallyphosphorylates site-1. Phosphorylation of site-2 but not site-1 is associated withdissociation of the free catalytic subunit of PP1 from PP1 G, which results in a markeddecrease in its activity, while phosphorylation of site-1 stimulates the activity of PP1.Thus, catecholamines induce the dissociation of PP1G by phosphorylating site-2, theinhibitory effect of the dissociation would outweigh the stimulatory effects of site-1phosphorylation. Furthermore, PKA also phosphorylates an endogenous inhibitor ofPP1, inhibitor-1 (1-1), which activates 1-1 and leads to its binding and inhibition of thedissociated free catalytic subunit of PP1 (Cohen, 1987). The overall effects ofcatecholamines are therefore an inhibition of GS-specific PP1, the maintenance of ahigh level of phosphorylation of GS, and a decrease in the GS activity (Cohen, 1987;Cohen, 1989; Cohen and Cohen, 1989). Insulin, on the other hand, activates aprotein kinase which phosphorylates the site-1 of PP1 G, resulting in an increasedactivity of PP1 that dephosphorylates and stimulates GS increases. This mechanismis strongly supported by the identification of an insulin-stimulated protein kinase fromrabbit skeletal muscle that specifically phosphorylates the site-1 of PP1 G under bothin vitro and in vivo conditions (Dent et al., 1990). It should be noted, however, thatother regulatory mechanisms may also exist, as implied by a recent study whichreported that the activity of GSK-3 can be regulated by tyrosine phosphorylation onsites that are equivalent to that required by mitogen-activated protein (MAP) kinasesfor full activity (Hughes et al., 1993).Gluconeogenesis is another important aspect of glucose metabolism. It is aprocess whereby glucose is synthesized from pyruvate, lactate, glycerol and certainamino acids. Liver is the major site of gluconeogenesis. However, kidney can do so11as well, especially during long term starvation (Pilkis et al., 1990). Gluconeogenesisis accomplished by _enzymes catalyzing reactions opposing those of glycolyticenzymes. Glucose-6-phosphatase (G6Pase), fructose-1,6-bisphosphatase (FBP),and phosphoenolpyruvate carboxykinase (PEPCK) promote the flux of glucoseformation, and the opposing reactions are catalyzed by glucokinase, fructose-6-phosphate kinase (PFK), and pyruvate kinase (PK), respectively. If these enzymesare active at the same time, it would form an apparent "futile cycle", which wouldcause energy waste. But the integrated regulation of all the enzymes by hormoneand substrate availability ensures that this does not happen. For instance, cAMP- orcalcium-elevating hormones such as glucagon, catecholamines and angiotensin couldincrease the activities of PEPCK, FBP and G6Pase while decreasing the activities ofPK, PFK and GK simultaneously (Pilkis and Granner, 1992). These effects may haveresulted from the same mechanism mediated by PKA or Ca2+-calmodulin-dependentprotein kinase (Pilkis et al., 1988; Pilkis and Granner, 1992). Insulin may have directinhibitory effects on the activity of the gluconeogenic enzymes, but it may also inhibitthe activity of these enzymes and stimulate the activity of glycolytic enzymes viacAMP-independent mechanisms involving actions of phosphatases (Assimacopoulos-Jeannet and Jeanrenaud, 1990; Traxinger and Marshall, 1992). Furthermore, insulinalso affects the gene expression of PEPCK (Magnums, and Granner, 1989; O'Brienand Granner, 1990; 1991). Cis-elements in the promoter of PEPCK gene have beenidentified for cAMP-mediated activation and insulin-mediated inhibition of PEPCKgene expression (O'Brien et al., 1990). Transcription factors for the insulin responseelement (IRE) in the PEPCK gene have not been identified, although the pancreatictranscription factor PTF1 may be one of the candidates (O'Brien and Granner, 1991).C) Metabolism of Fatty Acids and Proteins. Generally, insulin promotes theanabolic metabolism of fatty acid and protein. Glycolytic flux culminates in the12formation of pyruvate, which can be converted to acetyl-CoA by the enzyme pyruvatedehydrogenase (PDH). In muscle, acetyl-CoA enters the tricarboxylic acid (TCA)cycle to produce energy, while in adipose tissue and liver, acetyl-CoA is used for thesynthesis of lipids under the action of acetyl-CoA carboxylase (ACC). Thus, the keyenzymes responsible for the synthesis of lipids are PDH and ACC. Not surprisingly,the regulation of both PDH and ACC involves phosphorylation. PDH isphosphorylated by a PDH kinase, which inactivates the enzyme, anddephosphorylated by PDH phosphatases, which activates the enzyme. Insulin isbelieved to stimulate PDH by activating the PDH phosphatases (Hughes et al., 1976;Espinal, 1989). ACC was first shown to be phosphorylated in cells by Brownsey et al.(1977), who demonstrated an increase in the incorporation of 32 P into ACC afterincubating fat cells for 75 min with 32Pi. It was subsequently shown that both insulinand epinephrine increased the extent of phosphorylation in ACC, but the sites ofphosphorylation were different as indicated by peptide mapping, which might explainwhy insulin activated ACC while epinephrine inactivated the enzyme (Brownsey andDenton, 1982; Witters et al., 1983). Similar results have been confirmed in rat liverand rabbit mammary gland (reviewed in Brownsey and Denton, 1985; Witters, 1985;Brownsey et al., 1988). PKA apparently mediated the epinephrine-inducedphosphorylation of ACC, while the kinase responsible for insulin-stimulated ACCphosphorylation remains elusive. CK-1 and CK-2 (Witters et al., 1983) as well as anovel insulin-activated kinase (Brownsey and Denton, 1984) were implied by earlierstudies, and MAP kinase may also be one of the likely kinases that phosphorylateACC (Pelech et al., 1991). It should be noted, however, that conflicting results havebeen reported on the effects of insulin on the phosphorylation of ACC. Witters et al.(1988) demonstrated that in Fao hepatoma cells, activation of ACC by insulin wascorrelated with insulin-induced dephosphorylation of ACC on sites that werephosphorylated by PKA or 5'-AMP-stimulated protein kinase. The stimulation of ACC13activity by insulin in rat liver was accompanied by a 20% decrease in the level ofphosphate in ACC (Mabrouk et al., 1990). Thus, the exact mechanism of ACCregulation by insulin requires further investigation.An additional aspect of insulin-induced activation of fatty acid synthesis is theconcomitant inhibition of lipolysis, which may result from the ability of insulin to inhibitadenylate cyclase and stimulate high-affinity cAMP phosphodiesterase. Interaction ofinsulin with fat cell adrenoreceptors may also play a role (Engfeldt et al., 1988). Theeffects of insulin on lipolysis is highly significant and have broad effects on energymetabolism (e.g., "glucose-fatty acid cycle" by Randle et al.)The effects of insulin on protein metabolism will not be described in detail here.Again, insulin exerts an anabolic effect on protein synthesis in sensitive cells. Themajor site of action is inhibition of proteolysis, while stimulation of protein synthesismay play a role as well (Marshall, 1989; Marshall and Monzon, 1989), especially infetal central nervous system (Heidenreich and Toledo, 1989). Observations havebeen made that insulin also stimulates transport of amino acid into cells, which wouldcontribute to the stimulation of protein synthesis (Bonadonna et al., 1993). It isinteresting to note that intracellular application of insulin stimulated protein synthesisin Xenopus oocyte, suggesting an intracellular site of insulin action (Miller, 1988).D) Regulation of Gene Expression.  Insulin has widespread effects on geneexpression, and is capable of regulating the expression of more than 50 genes,including those mentioned in the preceding paragraphs (Magnums, and Granner,1988; Magnums, and Granner, 1989; Magnums, et al., 1989; O'Brien and Granner,1991). These genes encode enzymes (O'Brien and Granner, 1990; O'Brien et al.,1990; Manzella et al., 1991), membrane proteins (e.g., glucose transporters),14secreted proteins or hormones (Prager and Melmed, 1988; Prager and Melmed,1989; Powell et al., 1991; Stanley, 1992), transcription factors (Messina, 1990),structural proteins (Straus and Takemoto, 1987; Craig et al., 1989), etc. Effects ofinsulin on gene expression are reflected by changes in translation or transcription of aparticular gene. A typical example of transcriptional regulation is the inhibition ofPEPCK gene expression by insulin. A 10-bp sequence in PEPCK gene betweennucleotide -416 to -407 was identified as an IRE for the effects of insulin. Importantly,this sequence is highly conserved, existing in many other genes that are regulated byinsulin, suggesting that IRE is an important element in realizing the effects of insulinon gene transcription (O'Brien and Granner, 1991). Unfortunately, efforts to identifythe transcription factors that interact with IRE have not yielded any concreteinformation, except that PTF1 and the transcription factor SP1 may be involved in theregulatory effects of insulin on gene transcription.II. INSULIN RECEPTOR1 Insulin Receptor Structure and FunctionWhile the exact mechanism of insulin action is still ill-defined, it is widelyaccepted that the insulin receptor (IR) mediates the biological functions of insulin.Consequently, extensive research efforts have been focused on the study of the IR.Structural and functional characterization of the IR have been extensively reviewed(Jacobs and Cuatrecasas, 1983; White and Kahn, 1986; White and Kahn, 1988; Yipet al., 1988; Yip, 1989; Pilch et al., 1989; Zick, 1989; Carpentier, 1989; Roth, 1990;Olefsky, 1990). I will summarize some of our current knowledge in this field.15Three major approaches were used in the early characterization of IR. Thefirst approach was to cross-link radioactive-labelled insulin with IR and analyze theinsulin-receptor complex with sodium dodecyl sulfate polyacrylamide gelelectrophoresis (SDS-PAGE); the second approach was to immunoprecipitate IR frommetabolically labelled cells and analyze the immunoprecipitates; and the thirdapproach was to purify IR and characterize the purified receptors (reviewed in Roth,1990). IR has been purified from a variety of sources, including rat liver, humanplacenta, human brain, and rat skeletal muscle (reviewed in Newman and Harrison,1990; Maturo and Hollenberg, 1990). All the purification procedures involve the useof either anti-IR receptor or affinity chromatography with insulin itself as animmobilized ligand (Jacobs and Cuatrecasas 1985). The original procedure has beengreatly improved over the years, with the introduction of wheat-germ agglutinin (WGA)column chromatography and modification of elution buffer for the insulin-agarosecolumn, which yielded highly purified IR with an excellent recovery. Sequenceanalysis of these purified receptors and molecular cloning eventually led to theisolation of human IR cDNA (Ebina et al., 1985; Ullrich et al., 1985).An intact functional IR consists of 4 subunits which are held together bydisulfide bonds to yield a tetramer of a2132 (Figure 1). The disulfide bonds that link thetwo a subunits are known as class I disulfides and are characterized by theirsensitivity to reduction with low concentration of DTT, while the disulfide bonds linkinga and p subunits are designated as class II disulfides which require a high DTTconcentration for reduction. Cysteine 647 in the carboxyl-terminal region has beenmapped to be the residue responsible for the formation of class II disulfide (Cheathamand Kahn, 1992). The IR complex has a molecular weight of 350 kDa, as determinedby sizing techniques such as gel filtration, ultracentrifugation and non-reducing SDSgels. On SDS reducing gels, however, a and j3 subunits migrate at 135 kDa and 95161/of autophosphorylsitessubunit and ation.Tyr(P)-960JuxtamernbraneRegulatoryRegionC-Term nalRegionTyr(P) 1316, 1322rACCrom-FACor^ir EFACIrdrdra,080ys-1Tyr(P) 1146,1150.1151DomainATP Reg 0BindingnFigure 1, Structure of insulin receptor. The tetrameric structure of the insulinreceptor is presented showing the functional domains of  the .:Elk4=7"4".:kDa, respectively. After treatment with neuraminidase, the mobility of the subunitsdisplayed an enhanced mobility, indicating that they are glycosylated (Jacobs et al.,1980). Initial cross-linking studies with 1251-labelled insulin in intact cellsdemonstrated that only the a subunit was labelled, indicating that this subunitcontains ligand binding sites (Pilch and Czech, 1980). Similar results were obtainedusing photoaffinity labelling (Yip et al., 1980). With the deduction of the amino acidsequence of IR from its cDNA (Ebina et aL, 1985, Ullrich et al., 1985), it is clear thatthe a subunit contains 731 amino acid residues with 15 consensus sequences (Asn-X-Ser/Th r) for asparagine-linked glycosylation. Additionally, 26 cysteine residueswere found to locate in a region between residues 155 and 315, forming the so-called"cysteine-rich" domain, which was also present in the LDL, EGF, and IGF-1 receptors.Apparently, the IR is bivalent, since each IR could bind up to two molecules ofinsulin (Yip and Jack, 1992). While the precise binding sites for insulin in the asubunit has not been identified, studies using receptor chimeras suggest that a 56amino acid sequence (230-285) in the cysteine-rich domain may play an importantrole in interaction with insulin (Gustafson and Rutters, 1990), which agrees withphotoaffinity labelling studies that mapped the binding domain at amino acidsequence 205-316 (Yip, 1989; Yip et al., 1991). However, other studies with chimericinsulin/IGF-1 receptors suggest that, unlike the IGF-1 receptor, the major insulinbinding domains for the IR are in the amino- and carboxyl- terminal regions of the asubunit (Schumacher et aL, 1991; Schumacher et al., 1993). The precise bindingdomain determinants are therefore controversial and await further investigation. Theresidues in the insulin molecule responsible for interaction with the IR have also beenexamined. Residues B6-Leu and the aromatic triplet (B24-Phe, B25-Phe, and B26-Tyr) were shown to confer affinity to insulin-IR interaction (Nakagawa and Tager1991; Mirmira et al., 1991). The binding affinity of insulin to IR is apparently under the18regulation of glucose, which at high concentrations (e.g. 20 mM) enhances thebinding of insulin to IR (Traxinger and Marshall, 1990).While the a subunit is entirely extracellular, the 13 subunit displays three distinctfunctional domains indicated by crosslinking studies and cDNA-deduced amino acidsequence. The 13 subunit could be labelled with both right-side out and inside outvesicles, indicating that this subunit had intracellular as well as extracellular domains(Hedo and Simpson 1984). Importantly, insulin stimulated phosphorylation of the 13subunit on tyrosine residues (Kasuga et aL, 1982a, 1982b), while onlyphosphoserines or phosphothreonines were detected under basal conditions (Jacobsand Cuatrecasas 1983). Moreover, phosphorylation of the 13 subunit wasdemonstrated in a cell-free system using purified receptor and the phosphorylationoccurred exclusively on tyrosine residues (Kasuga et al., 1982c). Thus, IR wasrecognized as a receptor tyrosine kinase, and this finding was confirmed bysubsequent studies using IR purified from various sources (reviewed in White andKahn 1986). Subsequent studies also indicated that insulin not only stimulatedautophosphorylation of its p subunit, but also enhanced its kinase activity towardphosphorylation of exogenous substrates such as poly Glu:Tyr (Ellis et al., 1986) andhistone (Klein et al., 1986). It should be noted that autophosphorylation of the 13subunit may require the structural integrity of the receptor (Boni-Schnetzler et aL,1988; Marchand-Brustel et al., 1989), which is true for the IGF-1 receptor as well(Tollefsen et aL, 1991). These data, however, must be viewed with caution, becausecontradictory results have been reported by Tavare et al. (1991) and Levine et al.(1991), who expressed an IR cytoplasmic domain in insect cells and the 48-kDa formof IR was fully active in causing autophosphorylation.19Inspection of the complete sequence of IR revealed that the R subunit of IRcontains at least three distinct domains, i.e., an extracellular domain which contains 4glycosylation sites, a 23 amino acid hydrophobic transmembrane domain, and anintracellular domain (Ebina et al., 1985, Ullrich et al., 1985). Comparison of the IRsequence with known kinases revealed strong homology in the cytoplasmic domain ofIR, confirming that the 13 subunit of IR is indeed a kinase. The intracellular domain ofIR contains a typical consensus ATP-binding site (Gly-X-Gly-X-X-Gly), followed by aLys residue at amino acid 1018 (Figure 1) with 22 amino acids in between. Thecritical role of Lys-1018 was unequivocally proved by site-directed mutagenesisstudies which showed that IR was completely devoid of kinase activity when Lys-1018was replaced by an Ala residue (McClain et al., 1987; Chou et al., 1987). At least sixof the 13 tyrosine residues in the IR subunit are autophosphorylated, and the sitesof phosphorylation were shown to localize in three regions, i.e., the C-terminal tailregion containing Tyr-1316 and -1322; the regulatory region containing Tyr-1146 -1150, and -1151 (Tornqvist et al., 1987; 1988; White. and Kahn 1989; Zhang et al.,1991); and the juxtamembrane region containing Tyr-960 (Murakami and Rosen1991) (Figure 1). Site-directed mutagenesis studies indicated that tyrosine residuesin the C-terminal region do not play a major role in regulation of IR tyrosine kinaseactivity, while simultaneous mutation of tyrosine residues in the regulatory regionabolished the kinase activity (Murakami and Rosen, 1991). A full activation of IRtyrosine kinase (IRTK) activity is dependent upon the "tris"-phosphorylation of theregulatory domain (Wilden et al., 1992). Phosphorylation of the three tyrosineresidues in the regulatory region seems to proceed sequentially, with Tyr-1146 firstphosphorylated, followed by Tyr-1150 or 1151 (White et al., 1988, White and Kahn1989). The mechanism of IR autophosphorylation has been studied using chimericreceptors, which suggests that phosphorylation occurs mainly through inter-molecularor intra-molecular transphosphorylation (Ballotti et al., 1989; Lammers et al., 1990;20Frattali et al., 1992). While intramolecular cis-phosphorylation does occur, it does notappear to play a significant role in activating the kinase (Frattali et al., 1992).Extracellular and transmembrane domains of IR 0 subunit have been found tobe of functional importance in terms of regulating IRTK activity and insulin signaltransduction (Yamada et al., 1992; Leconte et al., 1992). Remarkably, replacing asingle amino acid at Val-938 by Asp resulted in phosphorylation of the p subunit andstimulation of IR signalling in the absence of insulin (Longo et al., 1992). However,Frattali et al. (1991) obtained exactly opposite results using a similar mutant receptorwhere Val-938 was substituted by Glu. Further studies are required to clarify thiscontroversy. The functional importance of IRTK has been firmly established by alarge number of investigations involving the use of anti-IR antibodies or receptormutants. While one study (Maddux and Goldfine 1991) suggested that binding toATP in the presence of insulin could change the conformation of IR and initiate insulinsignalling without activating IRTK, most studies strongly support a role for IRTK in thesignal transmission of the biological function of insulin. Replacement of Tyr-1150 and-1152 resulted in a dramatic reduction in insulin-stimulated autophosphorylation andkinase activity, which was associated with a decrease in insulin-activated glucoseuptake (Ellis et al., 1986). Similarly, a Lys-1018 mutant that was devoid of kinaseactivity was unable to stimulate glucose uptake, to activate S6 kinase activity, tostimulate glycogen synthesis, or to mediate thymidine incorporation into DNA (Chouet al., 1987). Mutations of the three tyrosine residues in the regulatory region of the 1subunit resulted in a biologically inactive IR, apparently due to defectiveautophosphorylation and kinase activity (Murakami and Rosen, 1991). Similarly,another mutant receptor with Gly-996 replaced by Val was also kinase-deficient andlost the ability to transmit the insulin signal (Yamamoto-Honda et al., 1990). Theresidues Tyr-1150 and -1151 were also shown to play a role in mediating the21biological function of insulin in terms of myristyl-diacylglycerol generation and glucosetransport activation (Cherqui et al., 1990). Mutations of residues other than theautophosphorylation sites, such as the Gln-1152 mutant where a glutamine replacedthe tyrosine residue, also resulted in an impaired kinase activity, which correlated wellwith a decrease in insulin sensitivity (Formisano et aL, 1993). Furthermore, whenmutant receptors devoid of kinase activity were expressed together with normal IR,they inhibited the function of normal receptors (Maegawa et al., 1988), whileantibodies that stimulated IR autophosphorylation induced the biological effects ofinsulin (Caron et aL, 1987). Taken together, these observations are consistent with arole for IRTK in mediating the effects of insulin.The physiological significance of IRTK can be further appreciated bycomparing the sequence homology of IR with other growth factor receptors andtransforming viral oncogenes or their counterpart, the cellular proto-oncogenes. It isapparent that all growth factor receptors possess a large glycosylated extracellularligand binding domain, a single hydrophobic transmembrane domain, and acytoplasmic domain that contains intrinsic tyrosine kinase activity. The growth factorreceptors are classified into four subclasses on the basis of sequence similarity andstructural characteristics. These include the EGF-receptor subclass (I), IR subclass(II), PDGF receptor subclass (III), and the FGF receptor subclass (IV). Subclass Ireceptor has a single polypeptide structure with two cysteine-rich domains in theextracellular domain, while subclass II receptor has a tetramer structure exemplifiedby IR. Subclass III and IV contain several immunoglobin-like repeats in theirextracellular domain, and a unique kinase insert region that interrupts the kinasedomain (Ullrich and Schlessinger, 1990). Comparison of oncogene products withreceptor tyrosine kinase provided further insight into the function and mechanism ofreceptor activation by its ligands, since most oncogenes are tyrosine kinases that are22constitutively active, such as v-src, v-yes, v-fgr, v-fps, v-fes, v-able, v-ros, v-erb-B,and v-fms, (Hunter and Cooper 1986), suggesting that tyrosine kinase activity iscrucial in cell transformation and proliferation. The oncogenes described to dateencode proteins that fall into four classes: growth factors (e.g., v-sis for PDGF),growth factor receptors (eg., v-erbB for EGF receptor, v-fms for colony-stimulatingfactor receptor, CSF-R), transducer of growth factor responses (e.g., v-src, v-ras, v-rat). and transcription factors (e.g., v-jun, v-fos) (Hunter, 1985; Cantley et at, 1991).Strikingly, in the case of oncogenes encoding growth factor receptors, the oncogenesgenerally represent a truncated or mutated version of the cytoplasmic domain of thegrowth factor receptors. These alterations render the tyrosine kinase activity of theoncogenes constitutively active (White and Kahn 1986). This observation suggeststhat the extracellular ligand binding domain of the growth factor receptors may exhibita tonic inhibition on the tyrosine kinase activity of its cytoplasmic domain, andinteraction of receptors with ligands release the inhibition by causing conformationalchanges in the receptor and activating the receptor kinase activity (Ruther et al).Alternatively, ligand binding may facilitate the dimerization or aggregation of cellsurface receptors, leading to autophosphorylation of receptors and activation ofreceptor tyrosine kinase activity (Ullrich and Schlessinger, 1990)2. Processing of IR and Modulation of IR Function.The number of cell surface IR is dynamic. Soon after insulin treatment, adecrease in the number of IR occurs by a process known as receptor internalizationor endocytosis (reviewed in Carpentier, 1989; Levy and Olefskey, 1990; Knutson,1991). Endocytosis of IR is initiated by insulin binding and requires the activation ofthe IRTK (McClain et al., 1987; Klein et al., 1987; Hari and Roth 1987), although aconstitutive slow internalization of IR has also been identified (Knutson, 1992;23Paccaud et al., 1992). Endocytosis serves as a mechanism of down-regulation toprevent overstimulation of cells by insulin, but the internalized receptor also play arole in mediating the effects of insulin (Khan et al., 1989). The internalized receptorcan be either degraded in the endosome or recycled back to the membrane.Excessive degradation of IR as induced by mutating Lys-460 could result in insulinresistance (Kadowaki et al., 1990). The amino acid sequence that directs IRinternalization has been mapped to a 12 amino acid juxtamembrane region (Backer etal., 1990). This was accomplished by searching for the NPXT sequence that dictatesthe internalization of LDL receptor (Rajagopalan et al., 1991).Phosphorylation of IR on serine and threonine residues was observed in earlierstudies (Gazzano et aL, 1983). Various kinases are known to phosphorylate IR,including PKC (Takayama et al., 1988; Mudd and Raizada 1989; Lewis et al., 1990a;Grunberger and Levy, 1990; Ahn et al., 1993), PKA (Roth and Beaudoin, 1987), CK-1(White and Kahn, 1986; Rapuanot and Rosen, 1991), and even the 13 subunit of IR(Baltensperger et aL, 1992). Other unidentified kinases that are tightly associatedwith IR have also been shown to phosphorylate IR in an insulin-dependent manner(Smith and Sale, 1988; Lewis et al., 1990b). The phosphorylation is generallyassociated with a decrease in the activity of IRTK, perhaps due to an increase in theKm of IRTK for ATP (Haring et al., 1986). While the physiological role ofserine/threonine phosphorylation of IR remains speculative, it may offer an additionalmechanism for the regulation of IR function .While most studies focused on the aspect of phosphorylation of IR and othercellular proteins by IRTK, the area of IR dephosphorylation has not been welldeveloped. Dephosphorylation of IR leads to inactivation of the receptor andtermination of insulin actions. Human placenta was found to contain tyrosine24phosphatase that dephosphorylates IR and other growth factor receptors (Roome etal., 1988). Several tyrosine phosphatases (PTPase) such as the receptor-likePTPases LAR and LRP as well as the intracellular PTPase1B were shown todephosphorylate the intact autophosphorylated IR (Hashimoto et al, 1992). Inanother study, IR was dephosphorylated by a PTPase that was closely associatedwith IR (Mooney and Bordwell, .1992) or anchored to the endosomal membrane(Faure et al., 1992). These PTPases did not appear to relate to PTPase1B.Furthermore, insulin was shown to regulate cytosolic and membrane phosphataseactivity that dephosphorylates IR (Meyerovitch et al, 1992).III. MECHANISM OF POST-INSULIN RECEPTOR SIGNAL TRANSDUCTIONA central theme in the study of insulin action is the mechanism by which insulintransmits signals from the exterior of cells into the interior of cells such as thenucleus. It is now well recognized that all of the biological effects of insulin aremediated by IR, and most, if not all, of these effects require the activation of IRTK.However, it is far from clear as to what happens after IRTK activation, and how theseevents are ultimately translated into the biological actions of insulin. Nevertheless,studies in the past 5 years have made exciting findings in this area. These findingsstem from the belief that as a kinase itself, IR must use phosphorylation as one of itsmost important regulatory mechanisms. Indeed, a picture has emerged from a vastliterature that a protein kinase cascade may mediate the effects of insulin, and thatthis cascade may act as a common pathway for many other growth factors as well.Although solid evidence is still lacking to correlate the phosphorylation events withany of the physiological functions of insulin, it is just a matter of time before themystery is finally solved.251. Substrates of IRTKTwo strategies are used in attempts to decipher protein phosphorylationmechanism in insulins signalling. One is to search for and identify proteins that arephosphorylated by IR directly in response to insulin, and the other is to trace theupstream components from a distal end of the phosphorylation event. Bothapproaches have yielded interesting discoveries, especially the second approach.As IRTK phosphorylates proteins on tyrosine residues, anti-phosphotyrosineantibodies were used to identify proteins that become phosphorylated on tyrosine inresponse to insulin stimulation. Two major proteins were detected in this manner, a95 kDa protein and a 185 kDa protein (White et al., 1985). The 95 kDa protein wasidentified as the 0-subunit of IR by anti-IR antibodies, while the 185 kDa protein wasapparently a new protein. p185 was shown to be a soluble protein weakly associatedwith membranes, and phosphorylation of p185 occurred within seconds of insulinstimulation, and reached maximum at 30 seconds, with identical dose-response to IRautophosphorylation. This protein, however, was not well characterized until 1991,when Rothenberg et al. (1991) succeeded in purifying p185 from rat liver andsequencing several tryptic peptides from the protein. The cDNA clone of p185 wassoon isolated and a complete sequence of p185 was obtained (Sun et al., 1991).p185 was shown to be a novel protein and was named IRS-1, for insulin receptorsubstrate-1. Interestingly, IRS-1 contains more than 10 potential tyrosinephosphorylation sites with 6 of them in the YMXM motif that is known to bind the srchomology-2 (SH2) domain (Koch et al., 1991). Indeed, IRS-1 undergoes insulin-stimulated tyrosine phosphorylation and binds phosphatidylinositol-3 kinase (PI3K),whose 85-kDa regulatory subunit contains 2 SH2 domains (Sun et al., 1991). Thisobservation suggests that IRS-1 may act as a "docking" protein or adaptor protein to26bind molecules that contain SH2 domains. When IRS-1 was subsequently expressedin Chinese hamster ovary (CHO) cells, it was found to associate with PI3K duringinsulin stimulation, and the association was more sensitive and more responsive toinsulin stimulation in cells overexpressing IRS-1 and IR (Sun et al., 1992; Backer etal., 1992; Yonezawa et al., 1992; Myers et al., 1993). In vivo association andactivation of P13K has also been observed in intact rats, where a 2-fold and a 20-foldincrease in the activity of PI3K were detected in liver and muscle, respectively (Folli etal., 1992). Thus, IRS-1 plays an important role in the activation of PI3K by insulin.The mechanism of P13K activation probably involves direct tyrosine phosphorylationof the p85 regulatory subunit of P13K by IR (Hayashi et al., 1992; 1993) as well as theformation of a binary or ternary complex between IR, IRS-1, and P13K (Kelly andRuderman 1993; Backer et aL, 1993), although this is not the only mechanism. Theexpression of IRS-1 was induced during the differentiation of 3T3-L1 fibroblasts toadipocytes, and treatment of the adipocytes with insulin, dexamethasone or 1-methyl-3-isobutylxanthine dramatically reduced the expression of IRS-1 (Rice et al., 1992).Moreover, IRS-1 was shown also to mediate the activation of P13K by IGF-1 (Giorgettiet al., 1993; Myers et al., 1993). The expression of IRS-1 exhibits a large increaseduring liver regeneration (Sasaki et al., 1993), further supporting a physiological rolefor IRS-1. Thus, IRS-1 represents the first IRTK substrate that is identified to date.Elucidation of the functions of this protein holds strong promise for our understandingof the signalling mechanism for insulin action.In addition to IRS-1, other cellular proteins are also shown to act as in vitro orin vivo substrates for IRTK. Many of these proteins are not identified, and it is difficultto assign a physiological role for these in vitro substrates. A 60-kDa protein that wastyrosine phosphorylated by IRTK was recently purified and found to be associatedwith PI3K, but the sequence of this protein is not available for its identification (Lavan27and Lienhard 1993). Another putative IRTK substrate is a 15-kDa protein, pp15,which was phosphorylated exclusively on tyrosine in response to insulin stimulationand was suggested to play a role in insulin stimulation of glucose transport (Bernier etal., 1988a; 1988b). Treatment of cells with phenylarsineoxide (PAO) blocked insulin-stimulated glucose transport activity and resulted in the accumulation of pp15,suggesting that the turnover of .pp15 may be important in the insulin-stimulatedglucose transport. This protein was purified and identified as the fatty acid bindingprotein 422(aP2) which mediates intracellular transport of fatty acids (Hresko et al.,1988). Binding of fatty acids induces conformational changes in 422(aP2) andexposes the tyrosine for IRTK which phosphorylates 422(aP2) and presumablydecreases its activity (Hresko et al., 1990). This effect may contribute to theantilipolytic action of insulin, but its relationship to glucose transport remainsunknown.2. S6 Phosphorylation and S6 KinasesWhile IRTK catalyzes phosphorylation of proteins on tyrosine residues,changes in the level of overall Ser/Thr phosphorylation greatly surpass the changes intyrosine phosphorylation in insulin-stimulated cells. It is therefore most likely thatactivation of IRTK is only an initial step that acts as a trigger for wide-spread ser/thrphosphorylation of cellular proteins that directly perform tasks for the regulation ofcellular functions (Czech et al., 1988). One of the most consistent observations ininsulin-treated cells is phosphorylation of 40S ribosomal S6 protein, which wasdetected in insulin-treated 3T3-L1 cells or serum-treated Swiss mouse 3T3 cells insome of the earlier studies (Smith et al., 1979; 1980; Rosen et al., 1981a; 1981b;Thomas et al., 1981). While the physiological function of S6 phosphorylation is notfully understood, it has been suggested to play a role in stimulating protein synthesis28by increasing the recruitment of mRNA and facilitating initiation of translation (Thomaset al., 1982; Hasson and Ingelman-Sundberg, 1985; Traugh and Pendergast, 1986;Palen and Traugh, 1987; Hershey, 1989; Roads, 1991). That S6 phosphorylationmay be physiologically important is implied by the fact that numerous neurohumoralfactors stimulate S6 phosphorylation in mammalian cells, and most of the stimuli aregrowth factors, cytokines, or transforming viruses (Pelech et al., 1990). As anapproach to delineate mitogen-signalling pathways, major research efforts have beendevoted toward the identification of protein kinases that are responsible for mitogen-induced S6 phosphorylation (reviewed in Pelech et al., 1990; Sturgill and Wu, 1991).A variety of kinases were shown to phosphorylate S6 protein in intact cells or in cell-free systems, such as PKA and PKG (Issinger and Beier, 1980; Del Grande andTraugh, 1982), PKC (Le Peuch et al., 1983; Padel and Soling, 1985; Pelech andKrebs, 1987; Evans and Farrar, 1987; Chen et al., 1990), protease-activated kinase II(PAK-II) (Lubben and Traugh, 1983; Perisic and Traugh, 1983a; 1983b; 1985), a so-called H4P kinase (Donahue et al., 1984), a CK-1 like kinase (Rosen and Cobb,1983), and a 60-70 kDa kinase (Cobb, 1986). However, none of these kinases havebeen shown to play a significant role in the mitogen-induced S6 phosphorylation.Erikson and Mailer (1985) were the first to purify an S6 kinase from Xenopuseggs. The kinase was highly specific for S6 protein, and was able to phosphorylateall the physiologically relevant sites stimulated by progesterone treatment. Accordingto the elution profile on a DEAE column, this kinase was designated S6 kinase II. Themolecular weight (M r) of the kinase was determined to be 90,000-92,000 by SDS-PAGE, varying slightly with different levels of phosphorylation (Erikson and Mailer1986). The kinase was referred to as ribosomal S6 kinase (rsk), and a syntheticpeptide sequence from Xenopus S6 kinase II facilitated the molecular cloning of twoclosely related cDNAs in Xenopus oocytes, rsk a and rsk E3, which encode for29proteins with calculated of M r of 83,000 and 74,000, respectively (Jones et al., 1988).These two cDNA clones share 91% amino acid identity between themselves, andmore than 80% identity with the three subsequently isolated homolog genes fromchicken and mouse (Alcorta et al., 1989; Sweet et al., 1990). Mammalian homologsof the Xenopus S6 kinase II cDNAs were also identified in human cells (Chen andBlenis, 1990), suggesting evolutionary conservation of the kinase. Inspection of theamino acid sequence deduced from these cDNAs reveals a unique feature of thesekinases, i.e., the existence of two catalytic domains, with the N-terminal catalyticdomain most similar to PKA and the C-terminal domain most similar to phosphorylasekinase (PhK) (Jones et al., 1988). Recently, another highly related 90 kDa kinaseknown as S6 kinase I was purified from Xenopus oocytes (Erikson and Mailer, 1991),and shares 91% homology with S6 kinase II (Erikson, 1991). Also, a relatedmammalian 91 kDa S6 kinase was recently purified from rabbit skeletal muscle(Lavoinne et al., 1991). The similar size, enzymatic properties, cross-immunoreactivity, and sequence homology indicate that these kinases belong to thesame gene family designated rsk . The gene products of rsk are collectively referredto as p9Orsk, and this terminology will be used throughout this thesis.Purification and characterization of another family of S6 kinases known asmitogen-stimulated S6 kinases from mammalian cells did not succeed until muchlater, due to the rarity of the kinase and the difficulty in preserving its activity. Novak-Hofer and Thomas (1984) first noticed that inclusion of phosphatase inhibitors andEGTA in the extraction buffer seemed to minimize the loss of S6 kinase activity ingrowth factor- or serum stimulated cells, suggesting that the S6 kinase may beregulated by a phosphorylation mechanism. This finding helped in the subsequentisolation from Swiss mouse 3T3 cells of a 70-kDa S6 kinase that was stimulated byEGF, orthovanadate, or serum (Jeno et al., 1988). This kinase seems to be the same30as the 67-kDa and 65-kDa S6 kinase previously purified from bovine liver (Tabarini etal., 1987) and chicken embryo fibroblasts (Blenis et al., 1987), respectively. It alsoappears to relate to the 70-kDa S6 kinase identified in regenerating liver (Nemenoff etal., 1988). The same mitogen-activated S6 kinase was later identified and purifiedfrom cycloheximide-treated rat liver (Kozma et al., 1989; Price et al, 1989), andinsulin-stimulated rabbit liver (Gregory et al, 1989). A variety of growth factors andcytokines including EGF, PDGF, insulin and interleukin-2 were shown to activate thiskinase, designated p70s6k, in different cell systems (for reviews, see Pelech et al,1990; Kozma and Thomas 1992). While p70s6k was originally thought to be amammalian kinase, homologous kinases were recently identified in Xenopus oocytes(Lane et al., 1992). Two cDNA clones encoding p70s6k were isolated from rathepatoma cDNA library, i.e., clone 1 and clone 2, which encodes S6 kinases withpredicted Mr. of 56,160 and 59,186, respectively (Kozma et al., 1990; Banerjee et al.,1990; Reinhard et al., 1992). The two clones are identical except that clone 2 has a23-amino acid extension at the N-terminus. The two clones appear to arise from asingle gene, perhaps through alternate splicing, and the products of both clones arefunctional in terms of their responses to growth factor stimulation (Reinhard et al.,1992; Kozma et aL, 1993). While the physiologic significance of having two isoformsof the same kinase is not clear, the presence of a nuclear localization sequence inclone 2 (p85s6k) may suggest that this kinase is targeted to the nucleus (Reinhard etal., 1992). Two human p70s6k cDNAs, designated as p70 S6 kinase al (calculatedM r 58,946) and p70 S6 kinase all (calculated M r 56,153) were also identified (Groveet al, 1991). These might be counterparts of the rat clone 2 and clone 1,respectively. Indeed, examination of the cDNA sequence reveals that the humanp70s6k shared 100% identity with the coding sequence of rat p70s 6k, while the humanp85s6k had only a two amino acid difference from the rat p85s 6k. It is noteworthy thatS6 kinase al migrates on SDS-PAGE at an apparent M r of 90,000, which may result31from the presence of six consecutive arginine residues located at the N-terminus ofthis kinase (Grove et al., 1991).Apparently, p9Orsk and p70s6k represent two distinct families of kinasesresponsible for mitogen-stimulated S6 protein phosphorylation. The structure andfunction of these kinases have been summarized in several reviews (Kozma et al.,1989; Pelech et al., 1990; Erikson, 1991; Sturgill and Wu, 1991; Thomas, 1992).p9Orsk has a relatively broad substrate specificity, capable of phosphorylating laminand glycogen synthase besides S6 protein (Erikson and Mailer, 1989; Sturgill et al.,1988). The p70s6k differs from the p9Orsk in that it has only one catalytic domainresembling that of PKC (Kozma et al., 1990). However, the catalytic domain ofp70s6k is also most closely related to the N-terminal catalytic domain of the p9Orsk,with 57% amino acid identity between each other (Erikson, 1991). Furthermore,p70s6k has a very narrow substrate specificity, with S6 protein as the only substrateidentified to date. Kinetic studies indicate that p70s6k undergoes a biphasic activationin response to EGF, PDGF or insulin, with a PKC-independent early phase and aPKC-dependent late phase activation (Susa et al., 1989; 1990; 1992).The sites of S6 phosphorylation by the S6 kinases have been shown to belocalized in a 15 amino acid segment in the C-terminus of S6 protein. Five serinesare phosphorylated in vivo and the phosphorylation appears to occur in a specificorder : S236>S235>S240>S244>S247. Both families of S6 kinases are capable ofphosphorylating all five sites in vitro (Ferrari et al., 1991; Kozma and Thomas, 1992).The C-terminal sequences of S6 protein are therefore often used to synthesizepeptide substrates for S6 kinases in assaying S6 kinase activity.32While the regulatory mechanisms for S6 kinase activation by insulin or othermitogens are incompletely understood, there is compelling evidence suggesting thattheir phosphorylation on serine/threonine residues participates in the activationmechanism (Ballou et al., 1988a ,1988b; Erikson and Mailer, 1989; Price et al., 1990).The activating kinases for p9Orsk, also referred to as rsk kinases, have beenidentified as MAP kinases (Sturgill et al, 1988; Chung et aL, 1991). The upstreamactivating kinase for p70s6k has not yet been identified, although the existence of twoTP/SP motifs in the amino acid sequence suggests that a proline-directed kinase maybe responsible for phosphorylation and activation of this kinase (Ferrari et al., 1992).Indeed, p70s6k was shown to be phosphorylated by MAP kinases under in vitroconditions (Mukhopadhyay et al., 1992), but the significance of this finding is hard toassess, since other studies have clearly demonstrated that the MAP kinase andp70s6k lie on distinct pathways (Blenis et al., 1991; Ballou et al., 1991). Based on theprimary sequence of p70s6k, Avruch et al. (1991) have proposed a regulatory modelfor the activation of the kinase. A basically charged S/T-rich segment immediately C-terminal to the catalytic domain of the p70s 6k may serve as a inhibitorypseudosubstrate domain, since this region shares 25% amino acid identity with the C-terminus of the substrate S6 protein. Thus, this autoinhibitory domain folds over tocover the catalytic domain under resting conditions, but the catalytic domain is unableto phosphorylate this segment because the Ser/Thr residues are followed directly byproline which acts as a negative substrate determinant for S6 kinase. However, uponmitogen-stimulated phosphorylation of the Ser/Thr in this segment, the autoinhibitoryregion unfolds to expose the active sites of the catalytic domain, resulting in activationof the kinase.As pg 0 rsk and p70s6k may exist in the same cell system (Sweet et al., 1990),and both phosphorylate S6 protein, it is important to understand the relationship33between the function of the two kinases and to determine which kinase plays aphysiological role in a particular cell system. The two kinases may serve to responddifferently to different stimuli. For instance, cycloheximide only activates the p70s6kbut not the p9Orsk (Erikson, 1991). However, with the use of rapamycin whichspecifically blocks the activation of p70s 6k, Chung et al. (1992) recently demonstratedthat p70rsk is the physiological kinase, at least in fibroblasts. Similar results are alsoreported in T cells (Kuo et al., 1992; Calvo et al., 1992), H4IIEC hepatoma cells (Priceet al., 1992), and COS cells (Price et al., 1992). These studies therefore cast doubton the physiological significance of p9Orsk. On the other hand, it is most likely thatother unidentified S6 kinases may also exist. For example, a 45-kDa nerve growthfactor (NGF)-stimulated S6 kinase was identified in PC12 cells (Matsuda and Guroff,1987), and a previous report also described the stimulation of S6 kinase activities inrat skeletal muscle by in vivo administration of insulin (Hecht and Straus, 1988).Furthermore, we found that one of the activated S6 kinase peaks did not appear to beeither the p90 rsk or the p70s6k (Hei et al., 1993).The existence of a protein kinase cascade leading to S6 phosphorylationbecame immediately apparent when the MAP kinase from 3T3-L1 cells was shown tophosphorylate and activate the Xenopus S6 kinase II (Sturgill et al., 1988). However,the original model of the cascade now appears too simplistic, viewing a possible directlink between IR and MAP kinase, based on the observation that the MAP kinase isphosphorylated on both threonine and tyrosine (Anderson et al., 1990; Boulton et al.,1991a). Studies in the past five years clearly indicate that the protein kinase cascadeleading to S6 phosphorylation is much more complex, and that the signalling eventsfor growth signals may involve an intricate network of protein kinases interacting witheach other (Pelech et al., 1990). Kinases and activators acting upstream of MAPkinases are being identified and characterized, and currently, MAP kinase or34extracellular-signal regulated kinase (ERK), MAP kinase kinase (MKK) or ERK kinase(MEK), oncogene product p74raf and MEK kinase, and p21 ras have been identified asactivating components upstream of rsk family S6 kinases (Lange-Carter et al., 1993;reviewed in Pelech and Sanghera, 1992a; Nishida and Gotoh 1993) (Figure 2).These components are now considered in the following sections3. MAP KinasesMAP kinase was originally identified in insulin-treated 3T3-L1 cells as a Ser/Thrkinase that phosphorylates microtubule-associated protein-2 (MAP-2) (Sturgill andRay, 1986). The kinase was purified and characterized as a 40-42-kDa cytosolickinase that was stimulated 1.5-3-fold by insulin and was phosphorylated on boththreonine and tyrosine in vivo (Ray and Sturgill, 1987; 1988). Subsequently, thisMAP-2 kinase was shown to phosphorylate and activate the phosphatase-inactivatedXenopus S6 kinase II (Sturgill et al., 1988). This finding immediately stimulatedenthusiastic research efforts on the study of MAP-2 kinase, since it represented thefirst example of a possible protein kinase cascade activated by sequential activationof multiple Ser/Thr kinases. It was soon realized that MAP-2 kinase could bestimulated by a variety of growth factors and mitogens other than insulin (Hoshi et al.,1988; Rossomando et al., 1989) (Table 1), and was therefore renamed "MAP kinase"for mitogen-activated protein kinase.In addition to the 42-kDa MAP kinase, isoforms of MAP kinase have beenidentified from various sources. These include a 44-kDa maturation-activated myelinbasic protein (MBP) kinase (p44mPk) purified from sea star oocytes (Sanghera et al.,1990), a 44-kDa MAP kinase from Swiss mouse 3T3 cells (Rossomando et al., 1991),a 43-kDa MAP kinase purified from rat1 fibroblasts (Boulton et al., 1991 a), a 54-kDa35S6 PROTEIN GSFigure 2. Mitogen-activated^ ajor shown and their relationships are inunconfirmed links, and dashed lines indicate homologous kinases inyeast pheromone signalling pathways. See text for details, (Modified fromLange-Carter et al., 1993)36Table 1. Stimulators of MAP kinases in different cells or tissues.Agents or^MAPK^Cell or^ ReferencesConditions lsoform^Tissue typeAIF4-^p42^Mouse thymoma^Anderson et al., 1991aAngiotensin II^p42/p44^Smooth Muscle^Duff et al., 1992Bombesin^p42^Swiss 3T3 cells^Pang et al., 1993Carbachol^p44^Fibroblasts^Kahan et al., 1992CD3 antibody^?^Jurkat cells^Hanekom et al., 1989Nel et al., 1990a;1990bCycloheximide^p54^Rat liver^Kyriakis & Avruch, 1990Diolein^p57^Colon cells^Lee et al., 1993EGF Swiss 3T3 cells^Rossomando et al., 1989Ahn et al., 1990Pang et al., 1993CHO cells^Selva et al., 1993p42/p44^A431 cells Chao et al., 1992PC12 cells^Traverse et al., 1992Gotoh et al., 1990aB82L cells^Campos-Gonzalez &Glenney 1992TIG3 Hoshi et al., 1988ET-1^p42^Rat Mesangial^Wang et al., 1992cellsFGF^p44^Fibroblasts^Kahan et al., 1992CCL-39 cells Meloche et al., 1992L'Allemain et al., 1991ap42^Swiss 3T3 cells^Rossomando et al, 1989Ahn & Krebs 1990fMLPGHp41 erkl^Neutrophils^Grinstein & Furuya, 1992p43end44mpk^Neutrophils Torres, et al., 1993p p42^CHO cells^Moller et al., 1992p42 Preadipocytes^Anderson et al., 1992Winston & Bertics 1992Gip2^ Rat1 cell^Gupta et al., 1992GM-CSF Neutraphils Gomez-Cambronero et al., 199337Table 1. (cont'd)IGF-1 ? TIG-3 Hoshi et aL, 1988Ig cross-linking p42 Lymphocytes Gold et al., 1992IL-1 p42/p44 KB epidermal Bird et al., 1991Carcinoma cells? Foreskin cells Guy et al., 1991Insulin ? CHO cells Ando et aL, 1992Tobe et al., 1991Bartton et al., 1990Dickens et al., 1992p42 3T3-L1 cells Sturgill & Ray 1986,Ray & Sturgill 1987, 1988p42/p44 Rat liver Tobe et al., 1992p44 Rat 1 cells Boulton et al., 1991ap42/p44 HIRc/PC12 cells Boulton et al., 1991b? TIG3 cells Hoshi et al., 1988? Rat adipocytes Haystead et al., 1990LPS p42 TIG3 Hoshi et al., 19881-Methyl- p44 Sea star oocytes Jasbinder et aL, 1990AdenosineNGF p44 PC12 cells Gomez & Chen, 1991Miyasaka et al., 1990Scimeca et al., 1992Traverse et al., 1992Gotoh et al., 1990a;1990bOkadaic acid ? Rat 3Y1 cell Gotoh et al., 1990bRat adipocytes Haystead et aL, 1990PAF p42/p44 platelets Samiei et al., 1993PDGF p44 CHL cells Meloche et al., 1992PMA p42 B Cell Casillas et al., 1991Chung et al., 1991Samiei et al., 1993Progesterone p42 Xenopus oocytes Pelech et aL, 1987Serum p42/p44 Swiss 3T3 cells Chung et al., 1991TIG3 cells Hoshi et al., 1988p44 CHL cells Meloche et aL, 1992src ? Rat1 a cell Gupta et al., 199238Table 1. (cont'd)Chung et al., 1991Thapsigargin p42/p44 A431 cells Chao et al., 1992TIG3 Hoshi et al., 1988a-Thrombin p44 CHL cells Meloche et al., 1992p44 Fibroblasts Kahan et al., 1992p44 CCL-39 cells Meloche et al., 1992p42/p44 CCL-39 cells Vouret-Craviavi et al., 1993CCL-39 cells L'Allemain et al., 1991aCHO cells Tobe et al., 1991TNF Foreskin cells Guy et al., 1991Vanadate Chung et al., 1991For abbreviations, see LIST OF ABBREVIVATIONS.Note : this table is not meant to be exhaustive. It only examplifies the numerous factors thatcan stimulate the activity of MAP kinases in different model systems.39MAP kinase from cycloheximide-treated rat liver (Kyriakis and Avruch, 1990), and a57-kDa protein kinase from colon carcinoma cells (Lee et aL, 1993). The 43-kDaMAP kinase from rat1 cells was also named ERK1 for gxtracellular signal-Legulatedkinase 1. A cDNA clone encoding this kinase was isolated from rat brain (Boulton etal., 1990; Marquardt and Stabel, 1992), and its predicted amino acid sequence is 85%identical to the 44-kDa MAP kinase identified in Swiss mouse 3T3 cells and 77%identical to the sea star kinase p44mPk, suggesting that these kinases may behomologs (Rossomando et al., 1991; Pelech and Sanghera, 1992). ERK2 and ERK3cDNA are also cloned, which encode the previously characterized 42-kDa MAP-2kinase and a 64-kDa MAP kinase, respectively (Boulton et aL, 1991b). Transcriptscorresponding to the three ERK clones are distinctly regulated during celldifferentiation, suggesting that they may play different roles in different developmentalstages, in different cell types, or following exposure to different extracellular signals(Boulton et al., 1991b). Using antipeptide antibodies against MAP kinases, another45-kDa isoform of MAP kinase tentatively named ERK4 was also detected (Boulton etal., 1991b), but a cDNA clone corresponding to this kinase has not been isolated, norwas the kinase purified for detailed biochemical characterization. ERK2 cDNA wasisolated from other species, such as mouse (Her et aL, 1991), Xenopus oocytes(Gotoh et al., 1991a, Posada et al., 1991), and human (Owaki et al., 1992). TheERK2 sequence was found to share 87% similarity to rat ERK1 (Marquardt andStabel, 1992). Furthermore, p42erk2 was identified as the previously identified pp42in Xenopus oocytes, which became phosphorylated on tyrosine residues uponmitogen stimulation (Cooper, 1989; Rossomando et al., 1989; Ferrell and Martin,1990; Posada et al., 1991; Jessus et al., 1991). Thus, a family of MAP kinases existswith multiple isoforms, and the functions of these kinases are being explored (Pelechand Sanghera, 1992a).40All of the MAP kinases that have been purified share a requirement for dualphosphorylation on tyrosine and threonine (or serine) for maximal catalytic activity(Anderson et al., 1990; Sanghera et al., 1991; Scimeca et al., 1991). The sites ofphosphorylation for p42srk2 have been determined to be Thr-183 and Tyr-185 insubdomain VIII, (Payne et al., 1991). A similar sequence was also found in ERK1with Thr-204 and Tyr-206 as sites of phosphorylation (Her et al., 1991). In eithercase, the two phosphorylation sites are separated by only one amino acid Glu, hencethe phosphorylation sites are generally referred to as "TEY" (Nishida and Gotoh1993). MAP kinases also have similar substrate specificity, capable ofphosphorylating MBP or MAP-2, and the substrate recognition determinants weremapped as -Pro-Xaan-Ser/Thr-Pro- (where Xaa is a neutral or basic amino acid andn=1 or 2) (Gonzales et al., 1991; Alvarez et al., 1991; Clark-Lewis et al., 1991). Thus,MAP kinase falls into the general category of proline-directed kinases (Hall andVulliet, 1991). The MAP kinase activity can therefore be assayed with MBP or MAP-2as substrate, and can be expressed as "MBP" phosphotransferase activity. Whilebeing Ser/Thr kinases themselves (Rossomando et al., 1989; Ferrell and Martin1990), MAP kinases display a unique feature of autophosphorylating on tyrosineresidues (Seger et al., 1991; Wu et al., 1991; Crews et al., 1991; Barrett et al., 1992;Robbins et al., 1993). Interestingly, the bacterially-expressed ERK1 and ERK2 alsoautophosphorylate on serine residues, although tyrosine residues appear to be thepreferred site (Robbins et al., 1993). In addition to the p9Orsk, physiologicallysignificant substrates for the MAP kinases also include transcription factors such as c-jun (Pulverer et al., 1991; Chou et al., 1992; Pulverer et al., 1993), c-Myc (Seth et al.,1992), TALI oncoprotein (Cheng et al., 1993), and p62 TCF (Gille et al., 1992) (Table2). Moreover, smooth muscle caldesmon (Childs et al., 1992) and recombinant tauprotein (Drewes et al., 1992) were also shown to be in vitro substrates of MAPkinases. Most recently, phospholipase A2 was shown to be phosphorylated by MAP41Table 2. Putative substrates of MAP kinasesSubstrateProteinMAP KinaseIsoformSource ofSubstrateReferenceACC p42/p44 Purified fromrat adipocytesPelech et al., 1991Caldesmon p44mpk Smooth muscle Childs et al., 1992c-jun p42/p44p42Recombinant Pulvere et al., 1991,1993; Alvarez et al 1991Chou et al., 1992EGF-R KB human tumorcellsNorthwood et al., 1991Alvarez et aL, 1991Takishima et al., 1991Lamin B2 p42/p44 Chicken Peter et al., 1992c-Myc p41(human) COS-7 cells Alvarez et al., 1991Seth et al., 1992cPLA2 p42 Recombinant Nemonoff et al., 1993p62TCFp41 and purifiedPurified fromrabbitLin et al., 1993Gille et al., 1992pgorskp70S6kp42/p44p44XenopusRat liverAhn et al., 1990Mukhopadhyay et al., 1992TALI p44? T cell leukemiacellsCheng et al., 1993Tau protein p42 Recombinant Drewes et aL, 1992For abbreviations, see LIST OF ABBREVIVATIONS.kinase under both in vitro (Nemenoff et aL, 1993) and in vivo (Lin et aL, 1993)conditions. Insulin-activated MAP kinase was also able to phosphorylate raf-1 kinase,a product of proto-oncogene raf-1 (Anderson et al., 1991; Lee et al., 1991a; 1992).However, the physiological significance of these phosphorylation events have notbeen firmly established. Activation of p90rsk by MAP kinase is not sufficient for thephosphorylation of S6 protein, and may not represent a physiologically relevantpathway, while phosphorylation of raf-1 by MAP kinase had no effects on the kinaseactivity of raf-1 (Kyriakis et al., 1993), suggesting that this may not be of significance.The physiological significance of c-Jun phosphorylation by MAP kinase has also beendebated. Therefore, the challenge will be to firmly establish the physiological roles ofMAP kinase-mediated phosphorylation of these substrates in intact cell systems.The broad involvement of MAP kinase activation in growth proliferating signalssuggests that MAP kinase may play a role in the regulation of the cell cycle (Thomas,1992). Sequence analysis reveals that ERK1 has a striking similarity to KSS1 andFUSS, the two yeast protein kinases involved in cell cycle control (Boulton et aL,1990), as well as to Spk1, another yeast kinase involved in pheromone-activatedsignal transduction (Toda et al., 1991). The sea star kinase p44mPk exhibiteddramatic activation during oocyte maturation (Sanghera et al., 1990; 1991b), andXenopus MAP kinase was phosphorylated during the M phase of the oocyte cell cycle(Gotoh et al., 1991a; Jessus et aL, 1991), which is required for oocyte maturation(Posada and Cooper, 1992). Moreover, MAP kinase was shown be activated bymicrotubule disruption (Gotoh et al., 1991b), and the activated MAP kinase wascapable of causing interphase-metaphase transition of microtubule arrays in Xenopusoocytes (Gotoh et aL, 1991c), and to release oocyte from cell cycle arrest (Shibuya etaL, 1992). In mammalian cells, MAP kinase was activated during the Go-G1 transitionin fibroblasts (Posada et aL, 1991) and CHO cells (Tamemoto et al., 1992). These43studies suggest that in addition to mediating mitogen-induced responses, MAPkinases may also play an important role in regulating cell division cycle.To summarize, a large family of MAP kinases is found to exist in yeast, seastar, and Xenopus oocytes, as well as in mammalian cells. The MAP kinases arehighly conserved, as indicated by.the fact that the mammalian MAP kinase can actsimilarly to Xenopus MAP kinase in the regulation of cell cycle division (Gotoh et al.,1991c). The two major isoforms of MAP kinases, p42maPk or p42erk2, and p44mapkor p44erkl, have been cloned and characterized. Other isoforms have been identifiedbut are less well characterized. MAP kinases play a key role in the signaltransduction of growth factors and mitogens such as insulin. Regulation of MAPkinases involves phosphorylation on both tyrosine and threonine (or serine) residues,which will be further discussed in the following sections.4. MAP Kinase kinase or ERK Kinase (MEK)MAP kinase is clearly activated by a mechanism involving phosphorylation, andkinases that phosphorylate MAP kinase or activators that stimulateautophosphorylation and activation of MAP kinase have recently been identified. A45-60 kDa protein kinase(s) which facilitate tyrosyl and threonyl phosphorylation andactivation of p42maPk was first identified in fibroblasts and PC12 cells (Ahn et al.,1991; Seger et al., 1991; Gomez and Cohen, 1991; Rossomando et al., 1992). Theantigen receptor CD4-associated tyrosyl kinase p56 1ck has been shown tophosphorylate and activate sea star p44mPk (Ettehadieh et al., 1992). The MAPkinase activator was first purified from insulin-treated rabbit skeletal muscle as a 45-50-kDa phosphoprotein (Nakielny et al., 1992a; 1992b). A similar activator was alsopurified from mature Xenopus oocytes, and was shown to be an intermediate44between the maturation promoting factor (MPF) and MAP kinase (Matsuda et al.,1992). At the same time, the MAP kinase activator was successfully identified andpurified from EGF-stimulated A431 cells (Seger et al., 1992), rat 3Y1 fibroblasts(Shirakabe et al., 1992), murine T cell hybridoma (Crews and Erikson, 1992), andU937 cells (Adams and Parker 1992). These activators were demonstrated to bekinases, based on the observations that they phosphorylate MAP kinases that wereinactivated by mutation of their critical lysine residues (Nakielny et al., 1992a;Rossomando et aL, 1992); that they are able to autophosphorylate (Kosako et al.,1992); and that they were inhibited by an irreversible kinase inhibitor 5'-p-fluorosulfonylbenzoyladenosine (FSBA) (Adams and Parker, 1992). Interestingly, theMAP kinase kinase or ERK kinase (MKK or MEK) is a kinase with dual specificity,catalyzing the phosphorylation of both Thr-183 and Tyr-185 in p42erk2 (Nakielny etal., 1992a; 1992b; Alessandrini et al., 1992; Seger et al., 1992; Rossomando et al.,1992; Crews and Erikson, 1992; Kosako et al., 1992; 1993a). Thus, MEK representsthe first kinase with established dual specificity.The cDNA clones encoding MEK have been isolated from cDNA libraries ofrabbit brain (Ashworth et aL, 1992), murine pre-B cells (Crew et al., 1992), human Tcells (Seger et aL, 1992), rat kidney (Wu et al., 1993), and Xenopus ovary (Kosako etaL, 1993b). In fact, two MEK cDNAs referred to as MKK1a and MKK1b, were clonedfrom human T cells, with the MKK1b clone having a 78 nucleotide gap betweenposition 471-548 (Seger et aL, 1992). Except for the Xenopus clone which encodes a395-amino acid kinase, the human MKK1a clone, murine, and rabbit clones allencode a kinase of 393 amino acids with a calculated M r of 43.5 kDa. The sequenceidentity between the Xenopus and rat MKK clones is more than 92%, while the rat andrabbit clones differs at only three and one amino acid residues, respectively, ascompared with the human clone MKK1a, indicating that MEK is very highly45conserved. Expression of MEK clones resulted in a functional MEK whichphosphorylates and activates MAP kinase. Significantly, MEK was found to behomologous to the yeast kinases STE7 and PBS2 (in S. cerevisiae), and byrl andwisi (in S. pombe), with higher homology with STE7 and byrl, which participate in thesignal transduction of pheromone-induced yeast cell mating andconjugation/sporulation (Figure 2).(Ashworth et al., 1992; Crews and Erikson, 1992;Nakielny et al., 1992; Crews et al., 1992; Seger et al., 1992; Kosako et al., 1993; Wuet al., 1993). STE 7 has been demonstrated in genetic studies to act upstream ofKSS1 and FUS3, the yeast homologs of MAP kinases (Gartner et al., 1992; Cairns etaL, 1992; Stevenson et al., 1992; Zhou et al., 1993). FUS3 was also shown to beactivated by STE7 in vitro (Errede et al., 1993). In S. pombe, byri acts upstream ofSpk1, another yeast homolog of MAP kinase, in a similar way as STE7 actingupstream of FUS3. These studies suggest that the protein kinase cascade involvingMAP kinase and MEK is perhaps a common pathway mediating hormonal signaltransduction in all eukaryotic cells.MEK is widely expressed in rat tissues and two MEK transcripts have beenidentified, which appear to derive from a single gene, presumably through alternativesplicing (Wu et al., 1993). Sequential activation of MEK, MAP kinase, and S6 kinasewas observed in in vitro studies (Robbins et al., 1993), in NGF- or EGF-stimulatedintact PC12 cells (Traverse et al., 1992), as well as in intact animals (Tobe et al.,1992). While MEK autophosphorylates on serine, threonine as well as tyrosine(Kosako et al., 1992b), MEK activity is probably regulated only by serine or threoninephosphorylation, because PP2A treatment of MEK completely inactivated its kinaseactivity toward MAP kinase (Matsuda et al., 1992; Kosako et al., 1993). Moreover,the activation of MEK during Xenopus oocyte maturation closely correlated with itsphosphorylation, which occurs mostly on threonine and partly on serine (Kosako et46aL, 1993). Apparently, there exist other Ser/Thr kinases upstream of MEK thatcatalyze the activation of MEK. The protein kinase encoded by proto-oncogene c-raftproved to be one such kinase.5. Raf and Other MEK Kinases:Rafl is one of the few oncogenes that encode protein Ser/Thr rather than Tyrkinases. Other examples of such oncogenes are akt (rac), mos, and pim-1, and cotoncogenes. Three active raf genes are known to exist in man, mouse and chicken,namely, the A-raf, B-raf and c-raf-1 (Rapp et al., 1988; Rapp, 1991). The human andchicken raf-1s are highly homologous to the Xenopus raf-1, sharing over 87% ofsequence similarity (Guellec et al., 1991). While c-raf is ubiquitously expressed in allcell types, A-raf and B-raf are only expressed in high levels in certain tissues such asmuscle, cerebrum and testis (Rapp, 1991). A-raf, c-raf-1 and B-raf encode 68-, 74-and 73-kDa proteins, respectively (Sithanandam et al., 1990). Two additional cDNAsfor B-raf have recently been cloned which encode a 93.5-kDa and a 94-kDa proteinsin neuronal cells (Eychene et al., 1992a; 1992b). These higher molecular weight B-rafs seem to correspond to the 95-kDa B-raf detected in NGF-stimulated PC12 cells(Stephens et al., 1992). Sequence analysis of raf oncogenes indicate the presence oftwo distinct structural domains, the C-terminal catalytic domain, and the N-terminalputative regulatory domain which contains conserved regions CR1, CR2 and CR3(Rapp et al., 1988). Notably, a potential zinc finger sequence also exists in raf. Theimportance of the regulatory domain is suggested by the observation that truncationof raf N-terminal sequence resulted in an oncogenic form of raf (Rapp et al., 1988).That raf plays a significant role in mitogenic signal transduction is stronglysuggested by observations that numerous mitogenic stimuli activate raf. Treatment of47mouse 3T3 cells with PDGF or transformation of cells with v-fms, v-src, v-sis, or Ha-ras induced phosphorylation and activation of p74raf-1 , and the phosphorylationoccurred not only on serine and threonine, but also on tyrosine (Morrison et aL,1988). Phosphorylation of p74raf-1 leads to a slower migration of raf-1 on SDS-PAGEas a 76-78 kDa protein, a phenomenon known as band-shift, which is used as amarker for the activation of p74raf-I. Tyrosine phosphorylation and activation of raf-1was also induced by interleukin-2 (IL-2) treatment of murine T-cell line CTLL-2(Turner et aL, 1991), in CD3-stimulated primary human T cells (Zmuidzinas et al.,1991), as well as in erythropoietin-treated murine cell lines HCD-57 and FDC-P1/ER(Carroll et al., 1991). However, PDGF failed to stimulate tyrosine phosphorylation ofp74raf-1 in vivo in 3T3-L1 cells (lzumi et al., 1991). EGF-, CSF-1-, NGF- and insulin-induced activation of p74raf- / was associated with serine and threoninephosphorylation exclusively (Baccarini et at, 1990; Blackshear et aL, 1990; Kovacinaet al., 1990; lzumi et al., 1991; App et al., 1991; Ohmichi et aL, 1992). Thus, it isdifficult to ascertain a role for tyrosine phosphorylation in the activation of raf-1 kinase.Nevertheless, serine and threonine phosphorylation is certainly sufficient formaximum activation of raf-1 kinase. Activation of p74raf-1 may involve its associationwith growth factor receptors, which has been observed in the case of PDGF and EGFreceptors (Morrison et al., 1989; App et al., 1991). In contrast, raf-1 was found toassociate with IL-2 receptor in T-cell blasts in an active form and dissociated andtranslocated into cytosol upon IL-2 stimulation (Maslinski et al., 1992). Thus, themechanism of activation for raf-1 may be more complicated, involving more than justphosphorylation, and the role of tyrosine phosphorylation for raf-1 activation requiresfurther investigation.Not very much information is available on the function of B-raf and A-raf.Recently, activation of the 95-kDa B-raf was observed in NGF-treated PC12 cells.48Similar to activation of raf-1 in NGF-treated PC12 cells (Ohmichi et al., 1992),activation of p95B -raf was associated with exclusive serine phosphorylation (Oshimaet al., 1991; Stephens et al., 1992), suggesting that B-raf may function in a similarway to raf-1.While the above studies strongly suggest that raf-1 is an essential integralcomponent in the oncogenic and growth-factor signalling pathways (Kolch et al.,1991; Li et al., 1991b; Kizaka-Kondoh et al., 1992), functional consequences of raf-1activation remained highly speculative until it was found that raf-1 is, in fact, a kinasethat phosphorylates and activates MEK (Kyriakis et aL, 1992; Dent et al., 1992). MEKis the first physiological substrate identified for raf-1, and this finding has againexpanded the ever-growing cascade of protein kinases in the mitogenic signallingpathway. Expression of the oncogenic form of raf-1 led to constitutive activation ofMEK and MAP kinase in NIH-3T3 cells, and bacterially expressed oncogenic raf-1was able to activate the phosphatase-inactivated MEK (Kyriakis et al., 1992; Dent etaL, 1992; Howe et al., 1992). Recently, an in vivo connection between raf-1 and MEKhas also been demonstrated by genetic studies in Drosophila melanogaster (Tsuda etal., 1993). Significantly, the activation of MEK by raf-1 was associated with serineand threonine phosphorylation (Kyriakis et al., 1992; Dent et al., 1992; Howe et al.,1992).When the mammalian signalling pathway is compared with that in the yeast, itis clear that the counterparts of MEKK are STE11 in S. cerevisiae and byr2 in S.pombe (Figure 2). However, raf-1 does not share homology with these two kinases,suggesting that other kinases in addition to raf may exist and function as MEK kinase.Indeed, using sequences that are identical in STE11 and byr2, a mammalian MEKkinase (MEKK) cDNA was isolated recently (Lange-Carter et al., 1993). MEKK49shares 75% and 35% sequence similarity with STE11 and byr2, respectively.Expression of MEKK clone in COS cells resulted in phosphorylation and activation ofMEK, which in turn activated MAP kinase. MEKK encodes a protein kinase of 73 kDa,although the expressed protein migrate on SDS-PAGE at 78 kDa or 82 kDa,presumably due to phosphorylation (Lange-Carter et al., 1993). A MEKK was alsoidentified in Xenopus eggs with an apparent M r of 250,000 as determined by sucrosedensity gradient (Matsuda et al., 1993). It is not clear whether this kinase is related tothe 73-kDa MEKK, since no further characterization of the kinase has been reported.Interestingly, the proto-oncogene product mos was recently shown to be yet anotherMEKK, capable of phosphorylating and activating a purified, phosphatase-inactivatedMEK (Posada et al., 1993).As the yeast kinase cascade is initiated at the membrane by activation of Gproteins, it is proposed that MEKK may mediate the activation of MAP kinase by Gprotein-coupled receptors, while raf kinase appears to respond to signals coming fromreceptor tyrosine kinases (Lange-Carter et al., 1993). Therefore, the next critical stepwould be to identify the link between receptor tyrosine kinase to raf kinase, and Gprotein-coupled receptors to MEKK. While raf-1 is clearly activated viaphosphorylation, it is not clear what is responsible for the phosphorylation of raf.Conceivably, another kinase may act upstream of raf kinase, or an activator maystimulate the autophosphorylation of raf kinase. Indeed, MAP kinase was shown tophosphorylate raf-1 in vitro (Anderson et al., 1991; Lee et al., 1991 a; 1992), and PKCmay mediate the phosphorylation and activation of raf-1 in phorbol ester-treatedinsect cells (Sozeri et al., 1992) or the antigen-activated T cells (Siegel et al., 1990).However, the role of these kinases in raf-1 activation has not been firmly established.For instance, no evidence exists to demonstrate a role for MAP kinase in activatingraf-1 in intact cells, and the NGF-stimulated activation of raf-1 is apparently I-1 KC-50independent (Ohmichi et al., 1992). With regard to MEKK, evidence is lacking tosupport its regulation by phosphorylation, and the mechanism of activation for thiskinase is virtually unknown.While the exact link between raf and cell surface receptor tyrosine kinaseremains elusive, there is ample genetic and biochemical evidence suggesting thatp21 ras, another proto-oncogene product, may act upstream of raf (Satoh et aL, 1992;Robert, 1992). p21 ras will be considered in the following section.6. p21f in Insulin SignallingThree mammalian ras genes are known to exist and encode H-ras, K-ras, andN-ras, which are members of a superfamily of small molecular weight GTP-regulatoryproteins (smg). Oncogenic forms of these proteins are involved in cell transformationand carcinogenesis, whereas the normal cellular counterparts of these proteins maybe important in the signal transduction by growth factors (reviewed in Haubruck andMcCormick, 1991; Grand and Owen, 1991) Possible involvement of p21 ras in insulinaction was suggested initially by experiments demonstrating an interaction of IR withp21 ras (O'Brien et al., 1987), and by the observation that microinjection of antibodiesagainst p21ras blocked insulin-induced maturation of Xenopus oocytes (Korn et aL,1987). In cells overexpressing normal ras protein, there was an increase in theresponse of these cells to insulin (Burgering et al., 1989), while transfection ofNIH3T3 cells with dominant inhibitory mutants of p21ras interfered with the biologicalfunction of insulin (Medema et al., 1991). Insulin was shown to activate p21ras inNIH3T3 cells overexpressing insulin receptor (Burgering et al., 1991) and in 3T3-L1cells (Porras et al., 1992; Osterop et al., 1992). Furthermore, IR was shown to51phosphorylate p21ras in vitro as well as in vivo (Kamata et al., 1987), although thephysiological significance of the phosphorylation is not known.Activation of ras protein results in a variety of biological responses.Microinjection or expression of oncogenic p21 ras results in stimulation of MAP kinasein a cell-free system (Hattori et al., 1992) as well as in a variety of cell lines andXenopus oocytes (Gallego et al., 1992; Hattori et al., 1992; Shibuya et al., 1992;Pomerance et al., 1992; Leevers and Marshall 1992; Vries-Smits et al., 1992; Itoh etal., 1993). When a mutant inactive p21ras was expressed in rat1 fibroblasts, itinhibited the insulin- and PDGF-stimulated activation of MAP kinase (Vries-Smits etal., 1992). In addition to insulin, other growth factors or mitogens appear to requirep21ras for their signal transduction as well. For instance, the stimulation of MAPkinase by NGF in PC12 cells was shown to involve p21 ras (Robbins et al., 1992), andactivation of p21ras was observed in PC12 cells treated with NGF, FGF and IL-6(Nakafuku et al., 1992), and EGF-treated G54 cells and V59 cells (Satoh et al.,1990). Additionally, signalling by the Drosophila receptor tyrosine kinase sevenless(sev) and induction of IL-2 gene expression by T cell activation involved ras activation(Fortini et al., 1992; Rayter et aL, 1992). These observations strongly suggest a rolefor p21 ras in oncogenic and mitogenic pathways.To identify the immediate cellular targets for p21ras, Itoh et al. (1993)established a cell-free system to assay the activation of MAP kinase by p21 ras. Aras-dependent ERK-kinase stimulator (REKS) was identified this way. REKS has aMr of 150,000-200,000 as estimated by gel filtration. It is interesting, however, thatp21 ras appears to act directly upon the so-called REKS in this system to activate MEKand MAP kinase. In view of its position of being downstream of p21 ras, REKS wouldbe at the same level with raf or MEK, and may well be related to the 250-kDa putative52Xenopus MEKK. Further characterization and sequence information are required toclarify the relationship among REKS, raf, MEKK, and the 250-kDa putative MEKKfrom Xenopus oocytes.That raf-1 acts downstream of p21 ras-mediatedsignalling pathway was firstsuggested by Smith et al. (1986) and Feig and Cooper (1988), who demonstrated thatoncogenic raf caused cell transformation without the need for p21ras. Also,expression of active ras protein in NIH3T3 cells was sufficient to causehyperphosphorylation of raf-1 (Morrison et al., 1988). However, transformation ofcells with ras appeared to require raf-1 (Kolch et aL, 1991), while ras was required forgrowth factor-mediated phosphorylation of raf-1 in PC12 cells (Mamon et aL, 1991).Serum-, TPA-, and ras-induced gene expression was shown to require raf-1 by usingNIH-3T3 or HepG2 cells transfected with dominant-negative raf-1 (Bruder et al., 1992;Sakoda et al., 1992; Troppmair et al., 1992). Conversely, expression of a dominantinhibitory mutant of ras antagonized NGF- and phorbol ester-induced activation ofMAP kinase, S6 kinases, as well as phosphorylation of raf-1, suggesting that p21rasmediates NGF and PKC modulation of MAP kinase, S6 kinase, and raf-1 (Wood et al.,1992; Thomas et aL, 1992). Further evidence that raf-1 operates downstream ofp21 ras came from analysis of the Drosophila sevenless (sev) and torso receptortyrosine kinase systems, where the response to sev activity is dependent on raffunction, while constitutively activated raf protein can induce R7 cell development inthe absence of sevfunction (Dickson et al., 1992).While raf appears to be downstream of p21 ras, it is not known how rasactivates raf-1. How growth factor receptors activates p21 ras also remains an openquestion. However, new information accumulated recently has begun to shed light onthe connection between receptor tyrosine kinase and activation of p21ras (Feig 1993).53Ras protein exists in two states, one that is bound with GDP is inactive, and one thatis bound with GTP is active (Downward, 1992). The rate-limiting step in the activationof ras is the replacement of GDP by GTP, and three classes of protein have beenidentified to regulate the activity of p21 ras by altering the balance between the twostates of ras protein (Bourne et aL, 1990 Lowy et al., 1991). One class of suchproteins that stimulate the release.of guanine nucleotide from p21 ras and promote theexchange of bound-GDP for GTP are known as guanine nucleotide releasing factor(GRFs) or GDP-dissociation stimulator (GDSs). These proteins stimulate the activityof p21 ras by favouring the binding of GTP with p21 ras. The second class of proteinsthat activates the GTPase activity of p21 ras have also been identified and are knownas GAP for GTPase activating proteins, which decrease the activity of p21ras bypromoting hydrolysis of GTP. (McCormick, 1989). The third class of proteins areknown as GDI, since they inhibit the dissociation of GDP from p21 ras, and thuskeeping p21ras in inactive states. Thus, growth factor receptor tyrosine kinase mayactivate ras through its effects on the activities of GRFs, GAPs, and GDIs, which inturn determines the activation state of p21 ras, p21 ras-GAP is one of the bestcharacterized proteins that regulate ras protein. It contains a SH2 domain which isimplicated to play an important role in mitogenic signalling pathway (Koch et al., 1991;Fry, 1992). Tyrosine phosphorylation of GAP has been observed in transformed cellsor cells stimulated with growth factors such as PDGF and EGF (Molloy et aL, 1989;Ellis et aL, 1990; Margolis et al., 1990; Moran et al., 1991; Molloy et al., 1992), butinsulin and basic FGF failed to induce tyrosine phosphorylation of GAP under identicalconditions (Molloy et al., 1989; Buday and Downward, 1993a), suggesting thattyrosine phosphorylation of GAP may not be required for insulin activation of p21 ras(Porras et a!, 1992), although interaction of IR with GAP has been reported (Pronk etal., 1992). Stimulation of GAP activity was observed in PC12 cells in response toNGF (Li et al., 1992), whereas in Drosophila, GAP acts as a negative regulator of54signalling by the sev kinase (Gaul et al., 1992). Thus, GAP clearly play aphysiological role in mitogenic signalling by regulating the activity of p21 ras,The studies on GRFs and GDIs have progressed very slowly until veryrecently. Stimulation of GRF was observed in NGF-stimulated PC12 cells (Li et al.,1992) as well as insulin- and EGF-stimulated fibroblasts (Medema et al., 1993). Inhemopoietic cells, the activity of p95vav, a homolog of GRF, was recently shown to bestimulated by tyrosine phosphorylation in response to activation of T cell receptor(Gulbins et al., 1993). This study represents a significant advance in ourunderstanding of the link between tyrosine kinase receptors and ras activation,although this may be limited specifically to hemopoietic cells. Tissue specific GRFalso been reported in brain cells where p140 ras-GRF was identified as a homolog ofyeast GRF CDC25 (Shou et al., 1992; Wei et al., 1992; Margegani et al., 1992).GRFs that are ubiquitously expressed have been identified as human SOS (hSOS),mouse SOS-1 and SOS-2, which are the mammalian counterparts of the Drosophilaputative ras-GRF aon Qf aevenless (SOS) (Bowtell et al., 1992). Growth factor-induced activation of ras does not appear to involve activation of GRF activity of SOS,but rather the tyrosine phosphorylation of growth factor receptors triggersrelocalization of SOS from cytosolsol to membrane where ras resides. This isprobably mediated by adaptor proteins known as ASH (for abundant Src homology) orGrb2 (for growth factor receptor-12ound protein 2) in mammalian cells (Matuoka et al.,1992; Lowenstein et al., 1992). Homologs of these proteins have also been identifiedin Caenorhabditis elegans as Sem5 (Clark et al., 1992), and in Drosophila as Drk(Simon et al., 1993; Olivier et al., 1993). These proteins feature the existence of bothSH2 and SH3 domains. While the SH2 domains are known to bind with proteinscontaining phosphotyrosines (such as autophosphorylated growth factor receptors),SH3 domains bind proteins with proline-rich regions, which have been found in SOS-155and -2 (Rozakis-Adcock et al., 1993; Buday and Downward, 1993b). Thus, theautophosphorylation of growth factor receptors induces binding of adaptor proteinsSem5/Grb2 via their SH2 domains, which is followed by translocation of GRFs tomembrane via the interaction of their proline-rich domains with SH3 domains in theadaptor proteins (Gale et al., 1993; Li et al., 1993; Egan et al., 1993; reviewed in Feig1993; McCormick, 1993). Evidence supporting such a hypothesis has come from theobservation that EGF stimulated the formation of a ternary complex containing SOS-Sem5/Grb2-EGF receptors (Buday and Downward, 1993b; Rozakis-Adcock et al.,1993).It is clear from the above information that growth factors stimulate theactivation of p21 ras via GAPs or GRFs. Thus ras acts upstream of rail in mitogenicsignaling, and both are required to mediate the full spectrum of the biological effectsof growth factors such as insulin. It is not known, however, if the adaptor proteinsSem5/Grb2 would prove universal in the activation of ras by tyrosine kinasereceptors. A newly published report demonstrated that insulin stimulates theassociation of IRS-1 with Grb2 in CHO-HIR cells (Tobe et al., 1993). Thus, more thanone adaptor proteins may be required to link p21 ras with IR. However, furtherinvestigations are required to identify the GRFs that interact with IRS-1-Grb2 complexin the insulin-induced activation of ras protein.7. CK-2 and Other Insulin-activated KinasesIn addition to the kinases described above which constitute a network of kinasecascades, insulin also activates a number of other kinases whose regulations arecurrently unknown. Among these kinases are a class of independent multiple-substrate kinases known as casein kinases (CK), which are subdivided into CK-1 and56CK-2 according to their order of elution from DEAE column chromatography(Hathaway and Traugh, 1982). CK-2 is an oligomeric 130-kDa protein composed oftwo large catalytic subunits (a and/or a', with M r of 36,000 and 44,000) and twosmaller 13-subunits of M r 24,000-28,000. In contrast, mammalian cell CK-1 is amonomeric protein with a M r of 30,000-37,000, depending or the source ofpurification. The regulation of CK-2 is far from clear, although phosphorylation hasbeen suggested (Agostinis et al., 1987). CK-2 autophosphorylates its 13 subunit, butthis does not appear to have a significant effect on the kinase activity (Pinna, 1990).CK-2 was shown to phosphorylate numerous substrates, such as initiation factors,RNA polymerase, GS, ACC, and 1-2 (reviewed in Pinna, 1990; Pinna et al., 1991;Tuazon and Traugh, 1991). Some of these substrates may play an important role incellular function. The primary sequence of CK-2 was obtained from human, bovine,rat, Drosophila, and yeast cDNAs. A high degree of homology exists between human,rat and bovine CK-2 (Tuazon and Traugh, 1991). The expression of CK-2 a or a'subunit displays an apparent tissue specific pattern in chicken embryo (Maridor et al.,1991), and the expression of human CK-2 13 subunit is under the control of the asubunit, which results in coordinated expression of the two subunits (Robitzki et aL,1993).The primary sequence of CK-2 does not immediately suggests a role for thiskinase. However, careful inspection of a segment of the CK-2 13 subunit reveals aresemblance to the yeast protein kinase CDC28, hinting at a possible role for CK-2 incell growth and cell cycle regulation (Takio et al., 1987). Evidence has beenaccumulating over the past few years supporting such a notion. Serum-stimulatedgrowth of quiescent WI38 human lung fibroblasts induced a 6-fold activation of CK-2within 30 min, and the activity of CK-2 oscillated during a 24 h period following serumstimulation (Carroll and Marshak, 1989). CK-2 activity is increased by 3-4-fold during57mouse embryogenesis (Schneider et aL, 1986). Normal growth of mouse myelomacells or primary human fibroblasts and HeLa cells is associated with a concomitantincrease in CK-2 activity (Schneider and Issinger 1989; Lorenz et al., 1993).Moreover, the expression of CK-2 is transiently increased by dexamethasone-induceddifferentiation of 3T3-L1 cells (Sommercorn and Krebs 1987) and Theileria parva-induced transformation of bovine lymphocytes (Ole-Moiyoi et al., 1993), whiledepletion of CK-2 by antisense oligonucleotide prevented neuritogenesis inneuroblastoma (Ulloa et aL, 1993). Further analysis of CK-2 in cell growth andproliferation indicate that CK-2 phosphorylates and activates DNA ligase I (Prigent etaL, 1992) and induces c-fos expression, which was accomplished by phosphorylationand activation of p67SRF, the mammalian transcription factor that binds to the serumresponse element (SRE) in the enhancer of the c-fos proto-oncogene (Manak et aL,1990; Gauthier-Rouviere et al., 1991; Manak and Prywes, 1991). Recent studiesindicate that CK-2 phosphorylates the cell cycle-regulating kinase p34cdc2, and thisappears to occur in vivo as well, since a two-dimensional tryptic map of the G1-phasephosphorylated p34cdc2 is identical to that of a CK-2-phosphorylated syntheticpeptide spanning residues 33-50 of human p34cdc2 (Russo et al., 1992).Interestingly, CK-2 was also an in vitro substrate for p34cdc2 at mitosis during celldivision cycle (Litchfield et al., 1992). Furthermore, immunochemical studiesindicated that during interphase, CK-2 a and [3 subunits are localized in the cytoplasmbut are distributed throughout the cell during mitosis, while the a' subunits are in thenucleus during interphase and translocate into the cytoplasm during mitosis (Yu et aL,1991). It was observed that CK-2 interacts with other cell growth-related proteins,such as the tumor suppressor gene product p53 (Meek et al., 1990; Filho et al., 1992)and heat shock protein HSP90 (Miyata and Yahara, 1992). CK-2 forms a complexwith p53, which tends to dissociate upon phosphorylation of p53 by CK-2. On theother hand, association of CK-2 with HSP90 prevents the aggregation of CK-2 and58causes activation of its kinase activity. While the significance of these interactionsremains to be determined, CK-2 clearly plays a significant role in cell growth and cellcycle regulation. Additionally, CK-2 is implicated in the mitogenic signal transductionby growth factors (Krebs et al., 1988). Stimulation of CK-2 activity was observed incells treated with insulin (Sommercorn et al., 1987; Diggle et al., 1991), IGF-1(Klarlund and Czech 1988), EGF (Sommercorn et al., 1987; Ackerman and Osheroff,1989), and bombesin (Agostinis et al., 1992). Therefore, CK-2 may participate inmitogenic responses, perhaps by phosphorylating growth- or cell cycle- regulatingmolecules. It is also possible that translocation of CK-2 13 subunit may contribute tomitogenic signalling (Lorenz et aL, 1993). However, no link has so far beenestablished between CK-2 with any of the components in the mitogen-stimulatedkinase cascade.Not very much information is available on the role of CK-1 in cell growth andproliferation, less is known about its role in mitogenic signalling. The significance ofCK-1-mediated phosphorylation of GS (Singh and Huang, 1985) and IR (Rapuanoand Rosen, 1991) in insulin signal transduction is not clear. Cloning of a bovine braincDNA encoding CK-1 indicates that there are at least 4 CK-1-like kinases referred toas CK-la, CK-113, CK-1y and CK-18, suggesting that CK-1 activity in tissues or cellextracts may be composed of multiple related but distinct enzymes (Rowles et aL,1991). Major progress in understanding the role of CK-1 was made by recent findingsthat CK-1 is highly homologous to HRR25 (DeMaggio et al., 1992, Graves et al.,1993), a protein kinase from S cerevisiae that is involved in DNA repair and isrequired for normal cell growth (Hoekstra et al., 1991). CK-1a shares 68% aminoacid identity and 86% amino acid similarity with HRR25, suggesting that HRR25 is anisoform of CK-1 (DeMaggio et al., 1992; Graves et al., 1993). Using antibodies raisedagainst CK-1, it was shown that CK-1 is localized to vesicular cytosolic structures and59to the centrosome in interphase cells, but becomes associated with the mitotic spindleduring mitosis (Brockman et al., 1992). Furthermore, CK-1 was identified as thekinase that inhibits the ability of Simian virus 40 (SV40) large T antigen to initiate DNAsynthesis (Cegielska and Virshup, 1993). These studies provided the very first cluesfor deciphering the physiological functions of CK-1. Again, the molecular mechanismfor the regulation of CK-1 is virtually unknown.Stimulation of cells with insulin results in modification of various other kinasesor enzymes, some of which may involve phosphorylation or dephosphorylation.Insulin-stimulated Ser/Thr kinases other than those described above have beenpurified from rat adipocytes (Yu et aL, 1987); rat liver (Klarlund et aL, 1990; 1991),and isolated hepatocytes (Reddy et al., 1990). Activation of PKC by insulin wasobserved in BC3H-1 myocytes (Cooper et al., 1987), adipocytes (Egan et al., 1990;Ishizuka et aL, 1991), chick neurons (Heidenreich et al., 1990); and H4IIE hepatomacells (Messina et al., 1992). PKC was shown to be involved with some biologicalfunctions of insulin, such as the stimulation of glucose transport (Ishizuka et al., 1991)and induction of .c-fos expression (Messina et al., 1992). However, Blackshear et al.(1991) carefully re-examined the nature of PKC activation by insulin using fibroblaststhat over-express normal human IR and antibodies to the myristoylated, alanine-richC kinase substrate (MARCKS) protein. It was found that insulin stimulated MARCKSprotein phosphorylation by about 2-fold in HIR 3.5 cells. This represents about 14%of the response to phorbol ester and about 17% of the response to 10% fetal calfserum. In four other cell lines, HIRC-B, BC3H-1, 3T3-L1 adipocytes, and H35 rathepatoma cells, which were made to express MARCKS protein, insulin did notstimulate MARCKS protein phosphorylation significantly, yet insulin was known tostimulate other Ser/Thr kinases in these cell lines. It was thus concluded that insulin60may activate PKC to a minor extent in certain cell types that overexpress IR, butactivation of PKC is unlikely to be of physiological importance.8. G Proteins in Insulin SignallingG proteins include heterotrimeric proteins composed of three subunits: a,and y. G proteins play a key role in the signal transduction of hormones orneurotransmitters whose receptors share the common structure of seventransmembrane domains (Gilman, 1987). While it is generally believed that the asubunit that contains GTPase activity plays a major role in signal transmission fromreceptors to intracellular effectors, the 13 and 1( subunits are now being shown to alsohave a role in signal transduction across the membrane (Lefkowitz, 1992). A family ofdifferent isoforms of G proteins are known to exist, with at least 16 a subunit genes, 413 subunit genes, and 5 y subunit genes (Birnbaumer, 1992). The best characterizedG proteins are Gs and Gi which are involved in activation and inhibition of adenylatecyclase, respectively; G q is another well-studied protein that is involved in theactivation of phospholipase Cp (PLC) (Spiegel, 1992).A role for G proteins in the signalling of insulin action was suggested by studiesusing pertussis toxin (PTX) which specifically catalyzes the ADP-ribosylation of G1protein and causes the inactivation of its function. Pretreatment with PTX blocked theeffects of insulin on protein synthesis, concomitant with a blockade of a rapid transientincrease in diacylglycerol (DG) in 3T3 cells (Hesketh and Campbell, 1987). PTX pre-treatment of BC3H-1 cells markedly attenuated the effects of insulin on hexosetransport, DG production, thymidine incorporation and generation ofphosphatidylinositol glycan (GPI), the putative intracellular mediator of insulin (seebelow) (Luttrell et al., 1988). Insulin was shown to stimulate hydrolysis of GTP in61isolated myocyte membrane, which was also blocked by PTX treatment (Luttrell et al.,1990). These studies suggest that insulin may interact with a G protein in producingsome of its biological effects. A G protein also appears to play a role in the insulin-induced GLUT4 translocation to plasma membrane in rat adipose cells (Baldini et al.,1991). The putative G protein involved, however, has never been identified (Houslay,1990), although the effectiveness of PTX may suggest a Gi-like protein. On the otherhand, insulin may interact with G proteins to modify their function. IRTK was shown tophosphorylate purified Gi and G o on tyrosine in vivo (O'Brien et al., 1987; Krupinski etal., 1988), but insulin failed to stimulate phosphorylation of G1 in rat liver plasmamembrane (Rothenberg and Kahn, 1988) or intact cells (Pyne et al., 1989).Nevertheless, the ability of PTX to ADP-ribosylate Gi protein was inhibited by insulin(Rothenberg and Kahn, 1988; Pyne et al., 1989). It was observed that in bovineadrenal cortical cells, insulin enhanced the angiotensin II-induced breakdown ofphosphoinositides, probably via an increase in the amount of G p, protein (Langlois etal., 1990). Thus, a possibility exists that insulin may realize some of its functionthrough interaction with G proteins, and conversely, G protein-mediated signaltransduction may be altered following insulin stimulation.9. Second Messenger Theory  :It is well known that G protein-mediated signal transduction is generallyaccomplished by production of second messengers such as cAMP, cGMP, IP3, DG,Ca2+, and etc. Attempts were made to search for such second messengers forinsulin without much success (Seals, 1985). One particular class of chemicals mayprove to be potential mediators of insulin action, i.e., the glycosyl-phosphatidylinositol(GPI) (reviewed in Lamer, 1988; Lamer et al., 1988; 1989; Walaas and Walaas, 1988;Schwarts et al., 1988; Pollet, 1989; Houslay and Siddle, 1989). Major evidence62supporting a role for GPI as a second messenger came from the observation thatinsulin treatment of rat liver resulted in the generation of oligosaccharide inositolphosphate compounds that mimicked the stimulatory effects of insulin on cAMPphosphodiesterase (Saltiel et al., 1986). It was subsequently found that insulinactivates a GPI-specific PLC, which catalyzes the hydrolysis of GPI to produce inositolglycan (IG) and DG. While DG could activate PKC, it is probably not an importantmessenger as discussed above. On the other hand, IG was shown to mimic anumber of the biological effects of insulin, such as inhibition of lipolysis, activation ofPK, stimulation of glucose oxidation, inhibition of GP, and stimulation of DNAsynthesis (Suzuki et aL, 1991; Misek and Saltiel, 1992; also reviewed in Saltiel, 1989;1990a; 1990b; 1990c). The molecular mechanism for the generation of IG has notbeen elucidated. It could perhaps require IRTK or interaction with G proteins assuggested in the above section. Whatever it entails, further characterization of theputative mediators is required before they are generally accepted as beingphysiologically relevant.10. An Integrated ViewMajor currently known molecular events as related to the mechanism for insulinsignal transduction have been described in the above sections. The complexity of thedifferent systems involved is formidable, and we are still only at the beginning ofdeciphering the mystery. It is encouraging, however, that a clear picture of proteinkinase cascade as summarized in Figure 2 has emerged as a central component inthe realization of the pleiotropic effects of insulin. Clearly, this cascade is not uniqueto insulin signalling, nor is it only present in mammalian cells. Many other growthfactors and mitogens may interact with their respective receptors to activate the samecascade through tyrosine kinase activity or through G proteins, and comparable63systems also exist in yeast. There is no indication that we are near the completion ofdiscovering all the individual kinases that constitute this cascade, but it may not be toolong before we see at least one completed pathway leading from membrane receptorto a particular biological response.Not shown in the diagram are molecules whose role in insulin action has notbeen firmly established, such as the putative mediators of insulin. It may well be thatmultiple mechanisms are required for the expression of the full biological effects ofinsulin. However, these aspects will not be further discussed in this thesis, but adescription of the pathological conditions resulting from abnormalities in the signaltransmission for insulin is well warranted.64IV. INSULIN SIGNALLING IN DIABETES AND INSULIN RESISTANCE1. Overview of DiabetesIn view of the essential role insulin plays in the maintenance of normal cellfunction, it is not surprising that the lack of it results in profound pathological changesin cell metabolism, which is expressed in the devastating disease diabetes mellitus.Diabetes mellitus is not a single disease, but rather a heterogeneous group ofdisorders hallmarked by an elevated blood glucose. The fundamental cause ofdiabetes is a deficiency of insulin function, due to a lack of circulating insulin or adiminished response of cells to insulin. Diabetes afflicts as many as 5% of the NorthAmerican population, and is the third leading cause of death. A thoroughunderstanding of the molecular pathology of diabetes would help in the attempt tobetter manage the disease.On the basis of their clinical manifestations, two major types of diabetes can bedistinguished, i.e., the juvenile-onset (Type 1) and the maturity onset (Type 2)diabetes (Table 3). Less than 10% of diabetics are in the category of Type 1, whilemost of the rest fall into the category of Type 2. The two types of diabetes also differin their etiology and pathogenesis. On the one hand, Type 1 diabetes results from amarked decrease in insulin levels and patients require exogenous insulin in order tosurvive, hence the name insulin-dependent diabetes mellitus (IDDM). On the otherhand, Type 2 diabetes usually exhibits a normal or higher level of circulating insulin,but the peripheral tissues do not respond well to insulin. Thus, a decrease in thesensitivity of tissues to insulin, which is also known as insulin resistance, occurs inType 2 diabetes (DeFronzo, 1988, 1992). Since insulin is not required and may notbe totally effective for the treatment of Type 2 diabetes, this type of diabetes is also65Table 3. Characteristics of Type 1 and Type 2 Diabetes.Features Type 1 Diabetes^Type 2 Diabetes(IDDM)^(NIDDM)Age of onset^< 20 > 40Proportion of < 10%^ > 90%all diabeticsSeanonal trend^Fall and Winter^NoneAppearance of^Acute or subacute^slowsymptomsMetabolic^Frequent^RareketoacidosisObesity onset^Uncommon^Common13 cells^ Decreased VariableInsulin Decreased^VariableInflammatory^Present initially^Absentcells in isletsFamily history^Uncommon^CommonHLA association^Yes^ NoAntibody to Yes NoIslet cellsreferred to as non-insulin dependent diabetes (NIDDM). An alternative to insulinwould therefore be useful in the management of NIDDM.2. Etiology and Pathogenesis of DiabetesThe causes of IDDM or NIDDM have not been well defined. There is strongevidence, however, suggesting that IDDM results from destruction of pancreatic 13-cellby autoimmune reactions, while genetic factors may play a role in NIDDM (Pyke,1989; Taylor, 1989). Genetic makeup is also intimately related to the development ofIDDM.Genetic components in IDDM are suggested by the observation thatsusceptibility to IDDM is associated with major histocompatibility complex (MHC)molecules, which are known as human leukocyte-associated antigen (HLA) inhumans (Niven and Hitman, 1986; 1989; Basabe, 1989; Leslie et al., 1989). Genesencoding DR1, 3, and 4 proteins of the MHC II molecules in the D region of the MHCloci on the short arm of chromosome 6 have been identified as the three major MHCloci to associate with IDDM susceptibility, while DR2 and 5 genes confer someprotection against IDMM susceptibility (Todd, 1990; Sheehy, 1992). The associationappears to be strongest for DR3 and 4, as more than 95% of IDDM are MHC DR3/4positive, although other loci such as DQA1 and DQB1 may also be involved (Todd,1990). When the sequence from many alleles of each locus were compared, onlyresidue 57 of the DQ 13-chain correlated consistently with susceptibility (Todd, 1990).The importance of this finding can only be understood by examing the structure of theDQ 13-chain, which indicates that amino acid 57 is positioned at a site important forinteraction with antigens. In addition to the MHC II linkage, recent studies indicate67that a faulty expression of MHC class I molecules could also contribute to IDDMsusceptibility (Faustman et al., 1991; Miyazaki et al., 1992).An important question in understanding the etiology of IDDM is what initiatesthe autoimmune attack on the pancreatic 13-cells. Environmental factors such as virusinfection may be one of the factors that trigger the autoimmune attack (Leslie et al.,1989; Ohashi et al., 1991; Oldstone et aL, 1991). Based on the accumulatedevidence, it is most likely that the autoimmune reaction occurs from the delayedexpression of islet antigens with subsequent failure to establish self-tolerancefollowed by activation of islet-specific autoimmunity (Wilson and Eisenbarth, 1990;Todd, 1990). Three major pancreatic autoantigens have been shown to exist only indiabetic humans or animal models, i.e., the cytoplasmic islet-cell autoantigen (ICA),the insulin autoantigen (IAA), and a 64-kDa pancreatic protein (Maclaren, 1988;Maclaren et al., 1989; Wilson et al., 1990; Christie et al., 1990; Palmer andMcCulloch, 1991). The 64-kDa protein was later identified as glutamic aciddecarboxylase (GAD), an enzyme that synthesizes the brain neurotransmitter 7-aminobutyric acid (GABA) (Baekkeskov et aL, 1990). The autoimmune attack isprobably triggered by a process known as molecular mimicry, in which foreignantigens resemble the pancreatic autoantigens. The attack is mediated by thesensitized cytotoxic T cells (CTL) that bind the autoantigens via MHC I molecules anddamage the p-cells (Bottazzo et al., 1989; Atkinson and Maclaren, 1990).Subsequent leakage of intracellular proteins from damaged 13-cells coupled withoverproduction of MHC I molecules cause further destruction of the 13-cells. Thesereactions lead to massive infiltration of lymphocytes into the islets, a condition knownas isulitis (islet inflammation), a cardinal sign of IDDM.68But how does the autoimmune reaction in IDDM relate to the MHC genes?MHC molecules may in fact play a "permissive role" with individuals carrying thesusceptible gene having an altered affinity for the helper T cells to bind with antigens.It was observed that when the aforementioned amino acid at position 57 of the MHC IIchain is aspartic acid, the likelihood of developing diabetes is very low (Atkinsonand Maclaren, 1990). Thus, connections between IDDM and MHC molecules may beexplained by an excessive attack on pancreatic (3—cells resulting from alterations inthe processing of antigens by the immune system.With regard to NIDDM, concerted research into the disease has providedinsight into its pathogenesis. Development of NIDDM requires the combined interplayof genetic and environmental factors. NIDDM cannot be attributed to alterations inany single gene, and the genetic basis of the disease is apparently of polygenicnature (Leiter 1989). It is generally accepted that the central issue to NIDDM isperipheral insulin resistance, but abnormalities in 3-cell function must be present totrigger the development of diabetes in the genetically predisposed subjects (Taylor,1989). In fact, development of impaired glucose homeostasis in NIDDM perhapsproceeds as a dynamic interaction between insulin action and insulin secretion, aprocess involving pancreatic n-cell, muscle and liver (DeFronzo, 1988; Garvey, 1989).Thus, NIDDM occurs as a consequence of an imbalance between insulin sensitivityand insulin secretion. The primary defect may be insulin resistance in muscles ingenetically susceptible individuals (Reaven, 1988; Beck-Nielsen et al., 1992), whichnecessitates a high rate of insulin secretion from (3-cells. When the p-cells areeventually exhausted and fail to maintain a high rate of insulin secretion, glucoseintolerance occurs and is followed by overt NIDDM (DeFronzo, 1992). The pulsatilepattern of insulin secretion is altered in NIDDM, presumably due to chronichyperglycemia, which desensitizes (3-cells and results in the so-called glucose "non-69sense" (Robertson, 1989). However, research into the molecular mechanism ofNIDDM revealed that the glucose "non-sense" in pancreatic 13-cells may have aprimary genetic basis. As discussed earlier, the glucose sensor in [3-cells is in fact thepancreatic glucokinase. In transgenic mice expressing a high activity of glucokinase,hyperinsulinemia persists in the presence of hypoglycemia (Epstein et al., 1992). Itwas also observed that the activity of glucokinase is decreased in diabetics withNIDDM, due to mutations in the glucokinase gene (Randle, 1993; Gidh-Jain et al.,1993). Lean diabetics may present with the clinical manifestations of NIDDM withprimary defects in the (3-cells. Thus, a full-blown syndrome of NIDDM occurs only inthe presence of two major defects, insulin resistance and fa-cell dysfunction(DeFronzo, 1992).3. Cellular and Molecular Defects in Diabetes and Insulin ResistanceA multitude of physiological and biochemical abnormalities have beenobserved in cells from diabetic individuals. This is not surprising in view of thepleiotropic nature of insulin action, the complex etiology, and the long duration ofdiabetes. Detailed studies on these changes entails the use of animal models ofdiabetes. Both chemically-induced diabetic animals and spontaneous diabeticanimals have been developed for diabetes research (Bell and Hye, 1983; Dulin et al.,1983). The two most commonly used compounds for induction of diabetes arestreptozotocin (STZ) and alloxan (Fischer, 1985). These compounds selectivelydestroy pancreatic 13-cells, probably through the generation of free radicals (Takasu etal., 1991). Animals treated with STZ or alloxan usually display signs of IDDM. Whengiven repeatedly at a lower dose into neonatal rats or obese male Zucker rats (fa/fa),STZ induces diabetes that resembles NIDDM. Of the many spontaneous diabeticanimal models, the most widely used ones are the BioBreeding (BB) rat, the ob and70db mouse (C57BUKsJdb), and the non-obese-diabetic (NOD) mouse. A caveat inusing these animal models of diabetes is that they do not fully represent humandiabetes in every aspects. For instance, STZ-diabetic rats are generally consideredto be models of IDDM, but these rats in reality do not require exogenous insulin forsurvival, and they do not usually display ketoacidosis which occurs often in humanIDDM if insulin is not administered, Thus, STZ-diabetic rats cannot reflect the entirepathology of human IDDM.While hallmarked by a decrease in the circulating levels of insulin, additionalfactors are definitively involved in the development of IDDM. As discussed earlier, thesignalling of insulin action is exquisitely complex, so that any alterations along thepathway of insulin signalling could contribute to the ultimate defects in insulin action.Thus, in addition to an absolute insulin deficiency, molecular defects have beenidentified in insulin signalling during diabetes. For instance, the number of IR wasincreased by 60%-200% in STZ-diabetic rats, while the activity of IRTK from musclesand liver was decreased as indicated by a decreased autophosphorylation as well asphosphorylation of exogenous substrates such as histone H2B (Burant et al., 1986;Okamoto et al., 1986; Giorgino et al., 1992). Similar alterations were also detected inthe diabetic BB rats (Okamoto et al., 1986) as well as in human diabetic subjects(Rosenzweig et al., 1990). Disparate results have also been reported showing anormal insulin binding and IRTK activity in liver membrane from STZ-diabetic rats(Amatruda and Roncone, 1985). The discrepancies may be due to differences inexperimental conditions and preparation of the samples, the substrate used to assayfor kinase activity, and the existence of inhibitory substances for IR (Block et al.,1991a, 1991b).71Defective IRTK and insulin signalling have been well recognized in NIDDM orinsulin resistance. In contrast to IDDM, the number of IR is generally decreased byhyperinsulinemia-induced down-regulation of IR. The IR isolated from skeletalmuscles of insulin resistant obese mice or obese Zucker rats displayed a decrease inIRTK activity toward autophosphorylation as well as exogenous substrate (LeMarchand-Brustel et al., 1985; Slieker et al., 1990). In human adipocytes, thecoupling of IRTK with p185 phosphorylation was shown to be normal in IR fromNIDDM individual, but the activity of IRTK was 50% lower than the controls (Thies etal., 1990). On the contrary, autophosphorylation and tyrosine kinase activity towardexogenous substrates were significantly enhanced in the liver IR from obese Zuckerrats (Slieker et al., 1990), suggesting that IRTK may be regulated differently in liverand muscle. There have also been reports indicating no changes in IRTK in skeletalmuscles of the db/db mouse (Vicario et al., 1987). While these differences may resultfrom different ways of expressing the activity of IRTK, it is also important to considerother interfering factors such as hyperlipidemia which is known to affect IRTK andinsulin binding to IR (Svedberg et al., 1992).The development of defects in IRTK during diabetes or insulin resistance hasbeen investigated using isolated rat adipocytes, where PKC was shown to play a rolein glucose-induced insulin resistance (Muller et al., 1991). Indeed, hyperglycemiawas shown to enhance the de novo synthesis of DG which activates PKC (Lee et al.,1989). However, studies involving human subjects strongly suggest a genetic basisfor the defective IRTK in NIDDM and insulin resistance. In NIDDM subjects, therewas a significant decrease in a population of IR that was immunoprecipitable byantiphosphotyrosine antibodies (Brillon et al., 1989). This is in agreement with thefinding that patients with Type A extreme insulin resistance and insulin-resistantdiabetes have mutations in IR locus that result in a decrease in the expression of IR72(Imano et al., 1991). These mutations include a substitution of Gly-996 by Val in theconsensus G-X-G-X-X-G sequence for ATP binding in IR kinase domain (Odawara etal., 1989; Yamamoto et al., 1990), replacement of Glu-981 in the kinase domain byArg (Kusari et al., 1991), introduction of a stop codon at amino acid 988 resulting indeletion of the entire kinase domain (Kusari et al., 1991), substitution of Trp-1200 forSer and Ala-1134 for Thr (Moller et al., 1991), and a mutation at nucleotide 145 whichis just before the codon for the critical ATP binding Lys-1030 in exon 17 of IR locus(Taira et al., 1989). Additional mutations are identified in IR gene which result inmutant IR with alteration in extracellular domains causing misfolding (Accili et al.,1992), or impaired processing and transportation of IR (van der Vorm et al., 1992).These observations provide strong evidence supporting a role for genetic componentsin insulin resistance syndromes and NIDDM. The various mutations that occur withinsulin resistance and NIDDM can be categorized into five classes (Taylor, 1992), i.e.,mutations that lead to a decrease in the number of IR mRNA or impair receptorbiosynthesis by introducing a premature termination codon (class 1), mutations thatimpair the transport of IR through the endoplasmic reticulum and Golgi apparatus tothe plasma membrane (class 2), mutations that decrease the affinity of IR to bindinsulin (class 3), mutations that impair IRTK (class 4), and mutations that acceleratethe rate of receptor degradation (class 5).While extensive studies have focused on the changes of structure and functionof IR per se in diabetes, very little information is available on the possible contributionof post-receptor defects in diabetes and insulin resistance. Impaired insulin signallingat steps beyond the IR was suggested by studies using STZ-diabetic rat models(Nishimura et al., 1989), and human subjects (Pima Indians) with insulin resistanceand NIDDM (Nyomba et al., 1990). Further analysis of IR from skeletal muscles ofSTZ-diabetic rats demonstrated an enhanced phosphorylation of pp185 (IRS-1)73(Giorgino et al., 1992), while livers from old female Wistar rats with glucoseintolerance exhibited a decrease in pp185 (IRS-1) phosphorylation (Nadiv et al.,1992). In addition, STZ-diabetic rats were also found to have a decreased CK-2activity (Metallo and Villa-Moruzzi, 1991) and a moderate but significant decrease inmicrosomal S6 phosphatase activity (Stephenson, 1991). Interestingly, a recentstudy using skeletal muscle samples from insulin resistant Pima Indians demonstratedan abnormal regulation of the p70 S6 kinase (Sommercorn et al., 1993). Thus, post-receptor defects as well as alterations of IRTK resulting from mutations of IR gene orpost-translational modification of IR may contribute significantly to insulin resistance inNIDDM (for reviews, see Le Marchand-Brustel et al., 1988; Kahn and Goldstein,1989; Goldfine and Caro, 1989; Taylor, 1989; Becker and Roth, 1990; Taylor et al.,1990; Olefsky, 1991; Taylor, 1992).V. VANADIUM: INSULIN-MIMETIC EFFECTS AND MECHANISM OF ACTIONVanadium is a widely distributed group Vb transition metal, with an averageconcentration of 100 ppm in the earth's crust (Boyd and Kustin, 1984; Willsky, 1990).The chemistry of vanadium is complex, owing to its multiple oxidation states,hydrolysis, and polymerization (Macara, 1980). The oxidation states of vanadium insolution vary from -1 to +5, depending on factors such as concentration-dependentequilibria and pH. The most stable forms of vanadium are the +4 (as vanadyl, V0 2+)and +5 states (as vanadate, VO4 3-), which are also the forms that occur in biologicalsystems (Willsky, 1990; Boyd and Kustin, 1984). Vanadium is recognized as anessential nutrient for higher animals, with a dietary requirement of about 10 gg/day(Ramasarma and Crane, 1981; Nielsen and Uthus, 1990). The biological propertiesof vanadium were known for more than a century, but interest in studying vanadiumwas derived from two major findings, i.e., the potent inhibition of Na+-K+-ATPase by74vanadium (Cantley et al., 1977) and the insulin-like effects of vanadium (Shechter andKarlish, 1980; Dubyak and Kleinzeller, 1980). Another significant advance is thefinding that oral administration of vanadium was able to normalize glucose levels indiabetic rats (Heyliger et al., 1985). The insulin-like effects of vanadium have sincebeen extensively studied in normal as well as diabetic animals using in vitro and invivo models (Shechter et al., 1988a; 1988b; 1990; Ramanadham et al., 1989a,1989b,1990; 1991; Shechter, 1990; Battell et aL, 1992; McNeill et aL, 1992, Cam et al.,1993).1. Biochemical Properties and Insulin-like Effects of VanadiumVanadium was found to affect an extensive list of enzymes such as ATPases,phosphatases, kinases, etc. (Ramasarma and Crane, 1981; Nechay, 1984; Nielsen,1987; Willsky, 1990), which suggests a physiological role for vanadium. Thebiochemical properties of vanadium were also exploited by using it as an investigativetool in characterizing various biological systems (Simons, 1979). Inhibition of Na -F-K -F-ATPase by vanadium was first characterized by Cantley et al. (1977) whodemonstrated that vanadate inhibited Na -F-K -F-ATPase by binding to the ATPhydrolysis site, with an IC50 as low as 40 nM. The active species in the inhibition ofNa-F-Ki--ATPase was identified as vanadate (V 5 as in VO43-), whereas vanadyl (V4as in V02-F) was ineffective. This observation was significant in that vanadate istransported across the membrane by way of an anion exchange system into thecytosol, where vanadate was reduced non-enzymatically through glutathione,NADPH, or ascorbic acid to vanadyl, which results in a loss of the inhibitory effects ofvanadium on Na -F-K -F-ATPase (Cantley and Aisen, 1979; Nechay et al., 1986). Thus,any in vivo biological effects of vanadate are unlikely due to the inhibitory effects ofvanadium on Na -F-K-F-ATPase.75Phosphoryl transfer enzymes are selective targets for vanadium, presumablydue to the fact that vanadate may serve as a phosphate analog (Gresser and Tracey,1990). Vanadate was shown to inhibit a variety of phosphatases in vitro, with specialselectivity toward phosphotyrosine phosphatase (Swarup et aL, 1982a; 1982b),although non-specific phosphatases such as alkaline and acid phosphatase were alsoinhibited by vanadate. Thus, vanadate is included in many biochemical buffersystems as one of the most potent phosphotyrosine phosphatase inhibitors (Gordon,1991).The insulin-like effects of vanadium were first observed in in vitro experiments,which demonstrated that vanadium stimulated glucose transport and oxidation in ratadipocytes, enhanced glycogen synthesis in diaphragm sections and hepatoyctes,and inhibited gluconeogenesis in hepatocytes (Tolman et al., 1979). Theseobservation were soon confirmed and extended in rat adipocytes as well as othertissues such as muscles (Shechter and Karlish, 1980; Agius and Vaartjes, 1982; Clarket al., 1985; Kellett and Barker, 1989), and were shown to be independent of theinhibitory effects of vanadium on Na -F-K÷-ATPase (Dubyak and Kleinzeller, 1980;Tamura et aL, 1984). Other insulin-mimetic effects of vanadium documented in invitro experiments include inhibition of protein degradation (Seglen and Gordon, 1981),inhibition of lipolysis (Duckworth et al., 1988), activation of glycolysis (Gomez-Foix etal., 1988), stimulation of amino acid transport (Munoz et al., 1992), stimulation of DNAsynthesis (Hori and Oka, 1980; Carpenter, 1981; Smith, 1983; Canalis, 1985), andcounteraction of the effects of glucagon (Miralpeix et al., 1989). Furthermore,vanadium displayed mitogenic effects, stimulating bone cell proliferation (Lau et al.,1988) and promoting the entrance of BC3H1 cells from Go into G1 phase (Wice et al.,1987). Vanadate was also found to prevent mouse embryo cell death (Rawson et aL,761990) and to substitute for IL-3 for the survival of a mouse IL-3-dependent cell line,IC2 (Tojo et aL, 1987). It should be noted, however, that some non-insulin-like effectsof vanadium have been observed in rat hepatocytes where vanadium was shown tostimulate GP and inhibit GS, which are effects opposite to those of insulin (Bosch etal., 1987; Villar-Palasi et al., 1989).The in vivo insulin-mimetic effects of vanadium were first observed in a study ofHeyliger et aL (1985) who demonstrated that oral administration of vanadate in thedrinking water at a concentration of 0.6-0.8 mg/ml for 4 weeks normalized theelevated glucose in STZ-diabetic rats. Moreover, treatment with vanadate alsoprevented the decline of cardiac performance usually observed in diabetic animals.These insulin-mimetic effects of vanadium could occur at a lower dose, are long-lasting, and are independent of changes in insulin concentration or insulin binding inthe rats (Meyerovitch et al., 1987; Brichard et al., 1988; Challiss et al., 1987).Vanadium treatment also exhibited anti-diabetic effects when its administration wasdelayed until diabetes was allowed to fully develop in STZ-diabetic rats (Cam et al.,1993). Most surprising, however, is the observation that the euglycemic effects ofvanadium lasted for another 13 weeks after it was withdrawn from diabetic ratsfollowing a three-week treatment (Ramanadham et al., 1989).While vanadium was shown to have non-insulin-like effects in diabetic rat liverin some studies (inactivation of GS and activation of GP) (Rodriguez-Gil et al., 1989),others have demonstrated that vanadium does produce insulin-like effects in rat liver,such as restoring glucose-induced activation of GS (Kanthasamy et al., 1988; Bollenet al., 1990), increasing the depressed basal GP activity (Pugazhenthi andKhandelwal, 1990), enhancing glucokinase activity (Gil et al., 1988), and stimulatingglycolysis (Rodriguez-Gil et aL, 1991). Vanadium was also found to inhibit hepatic77glucose output, to stimulate glycogen synthesis and glucose utilization in the skeletalmuscles diabetic rats (Blondel et aL, 1989).A prominent characteristic of vanadium is its ability to increase insulinsensitivity. Hyperglycemia associated IDDM severely impairs insulin sensitivity oftissues (Goto et al., 1988; Rossetti and Laughlin, 1989), which was shown to benormalized by vanadium treatment (Ramanadham et at., 1990; Rossetti and Laughlin,1989; Rossetti et al., 1990). Together with its effect to reduce insulin levels,vanadium would be particularly useful for NIDDM, in which insulin resistanceaccompanied by hyperinsulinemia is present. Indeed, in a very limited number ofstudies, vanadium was found to stimulate glucose oxidation, glycogen synthesis andglucose utilization in skeletal muscles from obese Zucker rats (Leighton et at., 1991),and to normalize glucose levels by partially normalizing insulin resistance whilereducing insulin levels (Pugazhenthi et al., 1991).Since vanadium alone was unable to support the survival of BB rats which donot have endogenous insulin (Battell et al., 1992), it may be that a residual amount ofinsulin is required for vanadium to produce insulin-like effects. Although most studiesindicate that vanadium does not affect insulin levels in diabetic rats but reduce insulinlevels in the control rats (Bendayan and Gingras, 1989), it is not known whethervanadium also acts directly on pancreatic p-cells. Studies on the effects of vanadiumon insulin secretion are controversial. Some indicated no effects of vanadium onbasal or stimulated insulin secretion (Serradas et aL, 1991), whereas others reporteda modest stimulation of insulin release by vanadium following pancreas perfusion(Fagin et al., 1987; Zhang et al., 1991) or by long-term vanadyl treatment (Pedersonet al., 1989). It is likely, however, that vanadium can produce beneficial effects on78pancreatic (3-cells by preventing hyperglycemia-induced exhaustive secretion ofinsulin.Most of the studies described above used vanadate or vanadyl in the treatmentof diabetic rats. In addition to vanadate and vanadyl, other compounds of vanadiumhave been developed in the hope of improving absorption and decreasing side effects(Cros et al., 1992). One of the promising compounds is bis(maltolato)oxovanadium(IV) (BMOV) which has been shown to produce insulin-like effects, is more potentthan vanadyl sulphate, and has fewer toxic effects as well as a rapid onset of action(McNeill et al., 1992).2. Mechanism of Action of VanadiumInhibition of Na+-K+-ATPase apparently does not mediate the insulin-likeeffects of vanadium. What then is the mechanism of vanadium action? To answerthis question, it is necessary to refer to the mechanism of insulin action wherebyphosphorylation plays a crucial role. Vanadium could act on any step of insulinsignalling from receptor binding to protein phosphorylation to mimic the effects ofinsulin . Many of these aspects have been explored and there is evidence that themechanism of vanadium action may be multifaceted. As we have not come to a fullunderstanding of the mechanism of insulin action, it is not yet possible to determinethe exact mode of action for vanadium.Vanadium was shown to have powerful inhibitory effects on protein tyrosinephosphatase (PTPase) as well as non-specific phosphatases under in vitro conditions(Swarup et al., 1982a, 1982b; Lopez et al., 1976; Montesano et al., 1988; Crans et al.,1989; Tessier et al., 1989). As the effects of insulin are largely mediated by IR that79contains tyrosine kinase activity, inhibition of PTPase would conceivably prevent thedephosphorylation of proteins that are phosphorylated by insulin stimulation, whichwould enhance and prolong the effects of insulin. When added to rat adipocytes, theactivation of GS by vanadium was associated with phosphorylation of the 95 kDa f3-subunit of IR (Tamura et al., 1983; 1984), suggesting that vanadium indeed affectcellular protein phosphorylation. In NRK-1 cells, addition of vanadate resulted in a40-fold increase in the level of phosphotyrosine which correlated well with vanadium-induced cell transformation (Klarlund, 1985). A general increase in the level ofphosphotyrosine phosphorylation in response to vanadium was observed in quiescentNakano mouse lens cells (Gentleman et aL, 1987), rat hepatoma (Fao) cells (Heffetzet aL, 1990), rat liver membrane (Yang et al., 1989), electropermeabilized humanneutrophils (Trudel et al., 1990; Grinstein et al., 1990; Bourgoin and Grinstein 1992),canine prostate (Tessier et al., 1989), and cultured chondrocytes (Owada et al.,1989). Vanadium was also shown to stimulate the phosphorylation of certain specificcellular proteins, such as a 50- and a 38-kDa protein in electropermeabilized platelets(Lerea et al., 1989), the pp15 (422aP2 protein) in rat adipocytes (Bernier et al.,1988b), p59fYn and p56 1ck in T cells (Secrist et al., 1993), and interestingly, pp6Osrc(Brown and Gordon, 1984; Soric and Gordon, 1988; Hutchison et al., 1992). In eachcase, the phosphorylation of these proteins was correlated with a certain specificbiological function of vanadium, such as platelet secretion, glucose transport, T cellactivation, cell transformation, etc. It is difficult, however, to assign a causalrelationship between the effects of vanadium on phosphotyrosine phosphatase andon these biological functions. More relevant to the insulin-mimetic effects ofvanadium is perhaps the stimulation of IRTK by vanadium, which was demonstratedwith purified IR preparations (Gherzi et al., 1988), in intact cells (Kadota et al., 1987a;1987b; Fantus et al., 1989), as well as in intact rats (Cordera et al., 1990). Thesestudies claim that stimulation of IRTK plays an important role in the insulin-mimetic80effects of vanadium. However, this notion has been contested by several workerswho have demonstrated that vanadium had no effect on IRTK in human lymphocytes(Torossian et al., 1988), that the anti-lipolytic effects of vanadium in rat adipocyteswas not correlated with a proportional increase in tyrosine phosphorylation of IR 13-subunit (Mooney et al., 1989), and that in vivo administration of vanadate normalizedhepatic glucose production in diabetic rat liver and stimulated glycogen synthesis inrat diaphragm without activating IRTK (Blondel et al., 1990; Strout et al., 1989).Furthermore, a 60% loss of IR number did not affect the effectiveness of vanadium tostimulate glucose transport (Green 1986), and vanadium normalized glucose levels indb/db or ob/ob mice where insulin was not effective (Meyerovitch et al., 1991). Thus,mechanisms other than activating IRTK must be sought to explain the full spectrum ofthe biological effects of vanadium.Like insulin, vanadium down-regulates cell surface IR in rat adipocytes by morethan 60% (Marshall and Monzon 1987). Similar effects of vanadium were observed incultured human lymphocytes where vanadium decreased the number of IR by 60%but inhibited the degradation of IR (Torossian et al., 1988). Vanadium was alsoshown to increase the binding of insulin and IGF to their receptors (Kadota et al.,1986; 1987a; Eriksson et aL, 1992), although the converse has been reported as well(Levin et aL, 1988). The biological significance of these observations are not clear.In addition to its effects on phosphorylation and insulin binding, it was alsoshown that vanadium stimulated cAMP PDE in adipocytes (Souness et al., 1985; Uekiet al., 1992), which would presumably counteract the effects of cAMP-elevatinghormones such as glucagon. The stimulation of glucose transport and normalizationof blood glucose by vanadium in diabetic rats could be related to the increasedexpression of glucose transporters. In diabetic rats, the liver glucose transporter81(GLUT-2) was increased, whereas GLUT-1 and insulin-sensitive GLUT-4 weredecreased (Oka et al., 1990). Administration of vanadium in vivo enhanced theexpression of GLUT-4 in skeletal muscle from STZ-diabetic rats (Strout et al., 1990)whereas in vitro application of vanadium in cultured NIH3T3 mouse fibroblasts wasshown to increase GLUT-1 expression (Mountjoy et al., 1990). In the insulin-resistant fa/fa rats, however, vanadium treatment markedly stimulated glucoseutilization without increasing the number of GLUT-4 (Brichard et al., 1992), suggestingthe involvement of mechanisms other than the regulation of GLUT-4 expression.Some interesting advances have been made recently regarding themechanism of action for vanadium. With the identification of IRS-1 as one of themajor substrates for IRTK, injection of vanadium into rat liver was found to stimulatethe phosphorylation of IRS-1 and the specific association of IRS-1 with P13K (Hadariet al., 1992). Vanadium was also found to stimulated a cytosolic tyrosine kinase in ratadipocytes (Shisheva and Shechter, 1992). This kinase was shown to be a 53-kDaprotein (determined by gel filtration) that was not stimulated by insulin. With the useof protein kinase inhibitors such as staurosporine and its analog K-252a, this kinasewas characterized as a primary post-receptor site for vanadium-mediated insulin-likeresponses (Shisheva and Shechter, 1993).While the in vitro mechanism of action for vanadium has been much studied,the in vivo mechanism of action for vanadium has not been well characterized.Available information suggests that the major defects in glucose disposal in diabetesis a lack of glycogen synthesis in response to insulin, and vanadium improved insulinsensitivity and reversed the defects in muscle glycogen synthesis (Rossetti andLaughlin, 1989; Rossetti et al., 1990). This effect appears to be specific andindependent of the normalization of glucose levels, since phlorizin, which lowers82glucose levels in diabetic rats by preventing glucose reabsorption from kidney, waswithout effect. Another possible in vivo mechanism for vanadium may involveinhibition of hepatic glucose output which is enhanced in diabetes.VI. RATIONALE, HYPOTHESES AND OBJECTIVESA protein kinase cascade has been identified which has the potential to explainmost, if not all, of the biological effects of insulin. Numerous studies in this area havebeen largely done using cultured cells or isolated tissues. Very few studies haveinvestigated the physiological significance of this cascade using an intact animalmodel, which would certainly be important if the kinase cascade was to be acceptedas universal. Furthermore, as the kinase cascade represents a fairly recentdevelopment, there has not been any study that systematically examines the kinasecascade in pathological conditions such as diabetes. This again is significant in thatany correlation between defects in the kinase cascade system and diabetes wouldlend further support to the physiological significance of the kinase cascade in insulinsignalling. Finally, the mechanism of action for vanadium remains uncertain in spiteof extensive research efforts, and an obvious possibility is that vanadium may act oncertain steps in the post-receptor kinase cascade to exert insulin-like effects.Exploration of the effects of vanadium on the kinase cascade would provide additionalinsight into the molecular mechanism of action of vanadium. This information mayalso contribute to the development of better insulin-substitutes in the management ofdiabetes. Thus, the following hypotheses are proposed and objectives of the studyare outlined accordingly.831. Hypotheses1) If the mitogen-activated protein kinase cascade is indeed the physiologicallyimportant signalling pathway for insulin action, it should be possible to demonstratethe activation of this cascade in an insulin-sensitive tissue by in vivo administration ofinsulin into an intact animal.2) Under pathological conditions where defects in insulin action exist, anychanges in the protein kinase cascade would contribute to the alterations in theactions of insulin.3) Vanadium compounds could exert their insulin-mimetic effects by affectingthe kinase activities in the protein kinase cascade.4) The protein kinase cascade that has been identified using cultured cells maynot include all the kinases that are functioning in the signal transduction pathway bygrowth factors. Kinases that have not been described may exist in rat skeletalmuscles and contribute to the realization of the biological response of insulin.2. ObjectivesBased on the above hypotheses, the objectives of this study were:1) To establish an intact animal model for the characterization of the proteinserine/threonine kinases involved in the responses to in vivo administration of insulin.842) To examine and characterize the protein kinase cascade in a model ofchronic diabetes and to correlate the alterations in the proteins kinases with theseverity and duration of diabetes.3) To investigate the effects of treating diabetic animals with vanadiumcompounds on the altered protein kinases cascade.4) To further pursue any novel findings observed during the above studies,such as purification of potentially novel protein kinases that are stimulated by insulin.85MATERIALS AND METHODSI. Materials.Bovine insulin, bovine serum albumin (BSA), f3-glycerolphosphate, EGTA,EDTA, MOPS, sodium orthOvanadate, p-methyl-aspartic acid, ATP,phenylmethylsulfonyl fluoride (PMSF), aprotinin, leupeptin, benzamidine,dithiothreitol (DTT), soybean trypsin inhibitor, pepstatin A, myelin basic protein(MBP), phosvitin, the peptide inhibitor of cAMP-dependent protein kinase (PKI),Kemptide, N,N-dimethylformamide (DMF), sodium heparin-agarose, and P81phosphocellulose filter paper were from Sigma. [y- 32P]ATP was from ICN.Acrylamide, bis-acrylamide, TEMED, ammonium persulfate, Brij 35, nitrobluetetrazolium (NBT), 5-bromo-4-chloro-indoly1 phosphate (BCIP), polylysine, HTP,PVDF membrane, Bradford protein assay kit, molecular weight standards for gelelectrophoresis and gel filtration, and p-mercaptoethanol were from Bio-Rad.S6 peptide and raf peptide (IVQQPGPQRRASDDGKKK) were kindly provided byDr. Ian Clark-Lewis (the Biomedical Research Center, Vancouver, B.C.,Canada). Ribosomal 40S subunits were prepared from rat liver by a proceduremodified from that of Krieg et al. (1988). Rabbit antibody against purified sea starMAP kinase p44mPk (anti-p44mPk) (Sanghera et al., 1991) and p44mPk peptide(GLAYIGEGAYGMV) (anti-mpk-I) (Posada et al., 1991) were prepared asdescribed. The following synthetic peptides were also prepared by Dr. Clark-Lewis for the production of rabbit antibodies that were affinity-purifed on peptide-agarose columns: CGGGGGEPRRTEGVGPGVPGEVEMVKGGC (ERK1-NT);PFENQTYCQRTLREIQILLGFRHENVIGIRDILRAPGGC (ERK1-III); CGGPKKPT-VPWKRLYGAADPKSLSLLDRILTFN P (ERK1-CT); GLAYIGEGAYGMVGC(MPK-I); LKPVKKKKIKREIKILENLRGGC (CK-2-III); MIVRNAKDTAHTKAERNIL-86EEVKHPGGC (S6K-I11); and CLVKGAMAATYS-ALNSSKPTPQLKPIESSILAQRRVRKLPSTTL (rsk-CT). Anti-raf1 antibodies were purchased from Medac(anti-c-mil, Hamburg, Germany) , Cambridge Research Biochemicals (CRB,Cambridge, U.K.) and from the National Cancer Institute (NCI, Bethesda).Monoclonal antiphosphotyrosine antibodies from mouse ascites tumor werepurchased either from I.C.N. (PY20) or from U.B.I. (4G10). Affinity-purifiedgoat-anti-rabbit IgG alkaline phosphatase conjugates and affinity-purified goat-anti-mouse IgG alkaline phosphatase conjugates were from Bio-Rad andCalBiochem, respectively. Glucose assay kits were from Boehringer-Mannheim.All other routine buffer chemicals were from BDH. Purified PP2A, MAP kinase,cdc-2 kinase and CK-2 were from Dr. Steve Pelech (Biomedical ResearchCenter and Faculty of Medicine, Vancouver, B.C.). PKC peptide substrates andPKG peptide substrates are from Peninsula Laboratory, Q Sepharose, SSepharose, MonoQ column, MonoS column, Superdex 200 column, Superose12 column, molecular weight marker for gel filtration, and FPLC are fromPharmacia. Calmodulin was from Calbiochem. Vanadyl sulfate was from FischerScientific, and BMOV was synthesized by Dr. Chris Orvig (Department ofChemistry, UBC)II. Experimental procedures.1. Preparation of animals.Wistar rats weighing about 200 g were injected intravenously with a bolusof streptozotocin (STZ, 60 mg/kg) to induce diabetes. Age-matched control ratswere injected with an equal volume of saline. The diahptic state was verified 18 h after the injection by measuring the blood glucose concentration using a87glucometer. The rats were kept for one, two or six months before carrying outthe biochemical analysis. In the 1- and 2-month studies, vanadyl sulphatetreatment was initiated at a concentration of 0.5 mg/ml in the drinking water andthe concentration was gradually increased to 1 mg/ml. The actual dose ofvanadyl sulphate fluctuated during the period of treatment due to variations inthe amount of water the rats drank. On average, the dose was about 0.45mmol/kg/day, based on calculations made according to the concentration ofvanadyl sulphate in the drinking water and the amount of fluid consumed by therats. The blood glucose levels were monitored throughout the study by samplingblood from the tail of the rat and measuring glucose with a glucometer. At theend of the treatment, the rats were fasted overnight and about half of the rats ineach group received a bolus injection of insulin (10 U/kg, iv), and the crudeextracts of skeletal muscle samples were prepared 20 min after the injection.For the 6-month study, 42 Wistar rats of similar body weight were used and 23were made diabetic by STZ injection (60 mg/kg). Eleven control and 12 diabeticrats were treated with (BMOV) by adding the compound in the drinking water(0.75 mg/ml). At the end of the 6-month period, skeletal muscles were removedto prepare crude muscle extracts. These rats did not receive insulin injection.Serum samples from all the rats were prepared at the termination of theexperiments for the determination of glucose concentration using a glucoseassay kit from Boehringer-Mannheim.2. Euglycemic clamp proceduresWistar rats of -200 g were anesthetized with sodium pentobarbital (65mg/kg), and the jugular vein and carotid arteinfusion and blood sampling, respectively. After a baseline reading of blood88glucose was taken, insulin (10 U/kg) was injected as a bolus, which was followedimmmediately by glucose infusion with a syringe pump. The blood glucoseconcentration was determined every 3 min with a glucose analyzer after insulininjection, and the rate of glucose infusion was adjusted to keep the glucoselevels at about 5.6 mM throughout the experiment (30 min). Skeletal muscleswere removed at various times after insulin injection for the preparation ofmuscle extracts.3. Determination of protein concentrationsProtein concentrations in the samples were assayed with a protein assaykit from Sigma using a procedure modified from that of Lowry et al. (1951), withBSA as a standard. Briefly, samples containing 50-200 lag protein were dilutedto 1 ml with water, to which 1 ml Lowry reagent (Sigma) was added. The samplewas then allowed to stand for 20 min and 0.5 ml of the Folin & Ciocalteu's phenolreagent working solution were added into the sample with rapid mixing. After 30min color development, the samples were read at 750 nM. Proteinconcentrations were determined with the use of a BSA standard curve.Alternatively, the samples were assayed for protein concentrations with dyereagents from Bio-Rad, which was based on the method of Bradford (1976).This method offers the advantage of ease and speed, and the micro-assayprocedure can be used to assay samples containing less than 25 pg protein. Toassay for proteins, the sample was simply mixed with the dye reagents(Coomassie brilliant blue G-250) and read at 595 nm against a BSA standardcurve.894. Preparation of 40S ribosomeThis procedure was modified from that of Krieg et al., (1988). Adult maleSD rats (200-220 g body weight) were fasted for 24 h and anesthetized withsodium pentobarbital. The livers were removed from the rats after decapitationand immediately frozen in liquid nitrogen. The frozen liver (100 g) was ground toa fine powder and suspended in 2 volumes of homogenization buffer (20 mMTris-HCI, pH 7.4, 100 mM KCI, 5 mM MgCl2, 1 mM DTT, 1% Triton X-100(v/v),1% sodium deoxycholate(w/v)). The suspension was then homogenized with apolytron at setting 5 for 20 s. The homogenate was centrifuged at 10,000 g for20 min at 4°C. The pellets were discarded and the supernatant was distributedin 29.5 ml aliquots into 38.5 ml Quick Seal tubes (Beckman) and underlayed with4 ml of buffer A (5 mM Tris-HCI, pH 7.4, 500 mM KCI, 2.5 mM MgCl2, 0.5 Msucrose, 1% sodium deoxycholate, 1%(v/v) Triton X-100, and 1 mM DTT)followed by 5 ml of Buffer B (same as buffer A except that the sucroseconcentration was 1 M) with a long glass Paster pipet. The tubes were sealedand centrifuged for 16 h at 56,000 rpm (Ti 70 rotor; L5-65 Beckmen centrifuge)at 0-2°C. The supernatant was discarded and the ribosomal pellets wereremoved and suspended in 1 ml buffer C (same as homogenization bufferexcept that no Triton X-100 and sodium deoxycholate were added). The pelletwas fully disrupted by shaking the tubes in the presence of three glass beads.The resultant 80S ribosome was then stored at -70°C.About 3000 A260 units of the 80S ribosome incubated for 30 min at 37°Cin the dissociation buffer (50 mM Tris-HCI, pH 7.4, 500 mM KCI, 3 mM MgCl2, 4mM DTT, and 2 mM puromycin). The sample solution (4.8 ml) was loaded intosucrose gradient tubes (7.5-37.5%(w/v) sucrose). The tubes were then90centrifuged at 4°C for 10 h at 28,000 rpm in a Beckman SW28 rotor. At the end,the tubes were punched at the bottom and 1 ml fractions were collected forabsorbance readings at A260. Two absorbance peaks were observed (Figure 3)and the fractions of the second peak (40S ribosome peak) were pooled andcentrifuged at 50,000 rpm for 4 h in a Beckman Ti 50.2 rotor. The 40S ribosomepellet was then resuspended in buffer C using a hand homogenizer. Proteinconcentration and A260 of the 40S ribosome sample were then determined.5. Preparation of crude rat skeletal muscle extracts.The muscle extracts were prepared from rats using a procedure modifiedfrom those previously described by Gregory et al. (1989) and Klarlund eta/.(1990). Briefly, male Wistar rats weighing about 250 g (Charles River) werefasted overnight and injected with bovine insulin (10 U/kg) dissolved inphosphate-buffered saline (10 mM sodium phosphate, pH 7.2, 145 mM NaCI,5.3 mM KCI). The rats were killed 2.5, 5, 10, 15, 20 or 30 min after the injectionby an overdose of sodium pentobarbital and the skeletal muscle from the hindlegs removed immediately, cut into small pieces and placed into 40 ml ofhomogenization buffer containing 25 mM MOPS (pH 7.2), 5 mM EGTA, 2 mMEDTA, 75 mM f3-glycerolphosphate, 1 mM Na3VO4, and 2 mM DTT. A mixtureof protease inhibitors was added into the buffer immediately beforehomogenization to yield final concentrations of 1 mM PMSF, 3 mM benzamidine,5 gM pepstatin A, 10 11M leupeptin, 10 gg/m1 aprotinin and 200 gg/m1 soybeantrypsin inhibitor. The homogenate was centrifuged at 10,000 x g for 15 min at4°C (Beckman J2-21), and the supernatant was again centrifuged at 100,000 x g(Beckman L8-60M) for another 60 min at 4°C_ The rAsultant supprnatant was stored as crude extracts at -70°C until further analysis. Blood samples were91Figure 3. Isolation of 40S ribosomes by sucrose gradient centrifugation. The 80Sribosome were dissociated with puromycin and applied to a sucrose gradient (7.5%-37.5%(w/v)) in centrifugation tubes, which were centrifuged in a Beckman swingingbucket rotor (SW 28) at 22,000 rpm for 16 h at 4°C. The centrifuge tubes were thenpunctured at the bottom to allow collection of 1 ml fractions, which were read for theirabsorbance at 260 nm. The slower migrating peak is the 40S ribosomal subunit.92Fraction Number93obtained after the skeletal muscles were excised for glucose assay using a kitfrom Boehringer and Mannheim.6. SDS-Polyacrylamide gel electrophoresisBio-Rad Protein II Electrophoresis unit or the Mini-protein II cell were usedto perform electrophoresis. The samples (crude muscle extracts, columnfractions or other proteins) containing <100 gig protein were digested by boilingfor 5 min in the sample digestion buffer (100 mM Tris-HCI, pH 6.8, 5%(w/v) SDS,10%(v/v) glycerol, 1 mM 2-mercaptoethanol, 0.02%(w/v) bromophenol blue).Molecular standards were treated similarly and the sizes of the proteinstandards were:Rabbit muscle phosphorylase b^97.4 kDaBovine serum albumin (BSA)^66.2 kDaHen egg white ovalbumin(OVA)^45 kDaBovine carbonic anhydrase^31 kDaSoybean trypsin inhibitor 21.5 kDaHen egg white lysozyme^14.4 kDaRegular slab gels or minigels were cast according to the method ofLaemmli (1971). The resolving gel contained 11%(w/v) (for immunoblotting) or12.5%(w/v) (for the S6 kinase assay with 40S ribosome as a substrate) totalacrylamide (acrylamide:N,N'-methylene bisacrylamide = 37.5:1), 375 mM Tris-HCI, pH 8.8, and 0.1% SDS(w/v). The stacking gels contained 3.5%(w/v) totalacrylamide, 125 mM Tris-HCI, pH 6.8, and 0.1%(w/v) SDS. The digested samples were loaded into the sample wells in the stacking gel and the upper and94lower tank were filled with running buffer (25 mM Tris, pH 8.3, 192 mM glycine,and 0.1%(w/v)) SDS. Electrophoresis was performed for 45-60 min at 200 V forthe minigel, or at 15-30 mA per gel for the regular gels (8-14 h). If the samplescontained radioactive ATP, the dye front was allowed to run off the bottom of thegel so that most of the free radioactive ATP would be in the lower tank runningbuffer. The gels were then stained for 20 min (minigel) or 60 min (regular gel)with Coomassie brilliant blue R-250 (0.25%(w/v) in destaining solution containing45%(v/v) methanol and 10%(v/v) acetic acid) and destained with the destainingsolution. In the case of radioactive gels, autoradiography of the gels wasperformed by exposing Kodak X-OMAT RP film with the dried gels in a plasticbag. Protein bands of interest were cut from the gel for the determination ofradioactivity.Silver staining of the gels was performed with a kit from Bio-Radaccording to the procedures suggested by the manufacturer. Briefly, the gelswere fixed with 40%(v/v) methano1/10%(v/v) acetic acid for 30 min followed byfixation twice with 10°/0(v/v) ethanol/5%(v/v) acetic acid for 15 min each. Thegels were then incubated for 5 min with oxidizer containing potassiumdichromate and nitric acid, and washed twice with distilled water for 5 min each.The gels were incubated again for 20 min with silver reagents containing silvernitrate. The gels were washed for 2 min with distilled water after the incubation,and developed by adding developer containing sodium carbonate andparaformaldehyde. When the protein bands became visible, the developingreaction was stopped by adding 5%(v/v) acetic acid.957. Protein kinase assays.A filter paper assay was used to determine the phosphotransferaseactivities of MBP kinase, CK-II, S6 kinases and p74raf 1 . Briefly, the reactionwas initiated by adding 5 pi [7- 32 P]ATP (250 IAA, specific activity was -2000cpm/pmol) into a 20 RI reaction medium, which contained 5 gl column fractionsor muscle extracts (1.25 mg protein/ml), yielding a final concentration of 20 mMMOPS (pH 7.2), 25 mM Vi-glycerolphosphate, 5 mM EGTA, 2 mM EDTA, 20 mMMgCl2, 2 mM Na3VO4, 1 mM DTT, 500 nM PKI and one of the kinasesubstrates. The assay of kinae activity in the crude tissue extracts wasperformed in the presence as well in the absence of substrate, which includedMBP (1 mg/ml), phosvitin (1 mg/ml), raf peptide (2 mg/ml) or S6 peptide (0.25mg/ml). MBP was used to assay for MAP kinase activity, since it is a goodsubstrate for MAP kinase. Phosvitin is phosphorylated well by CK-2, and isrelatively more specific with CK-2 than with CK-1. The raf peptide wassynthesized after a putative autophosphorylation site in the raf-1 sequence.The reaction was allowed to proceed for 10 min at 30°C, after which a 20 1.11aliquot was removed and spotted onto a P81 phosphocellulose filter papersquare (2x2 cm). The filter paper squares were washed extensively with 1%phosphoric acid with at least five changes, after which the paper squares werecounted by liquid scintillation counting in a Packard Tricarb 4530 counter. A pilotstudy showed that the reaction was linear for at least 20 min at a proteinconcentration of 0.25 mg/ml and also linear over a range of cytosolic proteinconcentration from 0.1-0.4 mg/ml when a 10 min reaction was performed. Theassay for phosphorylation of ribosomal S6 protein was performed in a similarway, essentially as described by Pelech and Krebs (1987), with ribosomal 40S subunits as substrate (0.35 mM) and a concentration of [y- 32 P]ATP at 15 1.M96(specific activity-10,000 cpm/pmol). After 30 min at 30°C, the reaction wasterminated by addition of 40 2x SDS sample buffer (10%(w/v) SDS, 20%(v/v)glycerol, 5%(v/v) (3-mercaptoethanol, in 0.125 mM Tris-HCI, pH 6.8). Thecontents of the reaction medium was then boiled at 100°C for 4 min and loadedonto a 12.5%(w/v) SDS-polyacrylamide gel. Gel electrophoresis was performedwith a Bio-Rad Protein II cell and the protein bands were visualized byCoomassie brilliant blue stain. The 32 kDa S6 protein band was excised fromthe gel for quantitation of radioactivity by Cerenkov counting.8. Fractionation of muscle extracts by column chromatography.For anion exchange chromatography of the muscle extracts, samplescontaining 10 mg protein were applied at a flow rate of 0.8 ml/min to aPharmacia MonoQ column (HR5/5) equilibrated with buffer A (10 mM MOPS, 25mM (3-glycerolphosphate, 5 mM EGTA, 2 mM EDTA, 2 mM Na3VO4, and 2 mMDTT). The column was developed at the same flow rate with a 20 ml linear NaCIgradient (0-800 mM) in buffer A. Fractions of 250 [LI were collected and used toassay for kinase activities or for immunoblotting.For hydrophobic interaction chromatography, samples containing 5 mgprotein were prepared in 250 mM NaCI in K-II buffer (12.5 mM MOPS, 12.5 mM(i-glycerolphosphate, 5 mM EGTA, 7.5 mM MgCl2, and 50 mM NaF), andapplied at a flow rate of 0.3 mg/ml onto a Pharmacia phenyl-Superose column(HR5/5) equilibrated with the same buffer. The column was eluted with adescending gradient of NaCI (250-0 mM) and an ascending gradient of Brij 35(0-3%,v/v) over a total volume of 20 ml. Fractions of 250 were collected and used to assay for kinase activity or for immunoblotting. A phenyl-Sepharose97column (2.5x7.5 cm) was also employed for the fractionation of muscle extracts.Muscle samples containing about 500 mg protein in 200 ml K-II buffer wereapplied at a flow rate of 1 ml/min to the column, which was equilibrated with thesame buffer. After a wash with 50 ml of K-11 buffer, the column was eluted with alinear 300 ml gradient of Brij 35 in K-I1 buffer (0-3%,v/v). Fractions (5 ml) werecollected and used for kinase assay or immunoblotting.Gel filtration chromatography was performed on a Pharmacia Superose12 column (HR10/30) or a Pharmacia Superdex 200 column. The columns werecalibrated using protein standards from Bio-Rad (Figure 4). Superose 12column was equilibrated with buffer A, and about 200 p.1 of sample (from MonoQcolumn fractions) were applied to the column at a flow rate of 0.5 ml/min and thecolumn eluted at the same flow rate with 30 ml of buffer A. Fractions (250 p.1)were collected for the kinase assay.9. Immunoblotting.Crude muscle extracts containing about 100 pg protein or aliquots offractions from column chromatography of the muscle extracts were digested with2x SDS sample buffer and applied to either 11 %(w/v) (for MAP kinases, p74raf land CK-II) or 10%(w/v) (for S6 kinases) SDS-polyacrylamide gels afterincubation for 4 min at 100°C. Gel electrophoresis was performed in a Bio-RadProtein II cell overnight at 15 mA per gel. At the conclusion of electrophoresis,the proteins in the gels were transferred onto a nitrocellulose membrane in aHoefer transfer cell at 400 mA for 3 h. The membrane was blocked for 2 h withblocking buffer containing 5%(w/v) skim milk and 0.5%(w/v) sodium azide in TTBS (20 mM Tris-HCI, pH 7.4, 0.5 M NaCI, 0.05%(v/v) Tween 20) and98Figure 4. Calibration of Superose 12 (A) and Superdex 200 (B) columns. Gelfiltration standards from Bio-rad were injected into the columns which were eluted withMOPS buffer that was used for samples. The flow rate was 0.6 ml/min for Superose12 column and 2.5 ml/min for Superdex 200 column. Elution volume for eachstandard protein was determined by absorbance at 280 nm using an in-line UVdetector. The size of the standards are shown and their logarithmic values are plottedagainst elution volumes. A linear regression was performed and the equations asshown in the graphs were used to calculate the size of unknown proteins.990)•....• 1000Superose 12 Column1 06Thyroglobulin 670 K^Log(y)=-0257235x + 7.9185r2=0.9986Gamma globulin 158 K1 05Ovalbulin 44 K■■1Myoglobin 17 K1 04Cyanocobalamin 1.35 KII-61'06 _8^1 0^12^14^16^18^20Superdex 200 Column41)1 05104B Thyroglobulin 669 kDa^Log(Y)=-0.012672X + 7321(r2•42.9966)IgG 166.3 kDaOvalbumin 40 kDaHorse myoglobin 17 kDa0100 140 180 220 260Elution Volume (Ve, ml)100incubated with primary antibodies for 2 h or overnight. The membranes werethen washed twice with TTBS and incubated for another 2 h with the secondaryantibodies (goat anti-mouse IgG alkaline phosphatase conjugates for themonoclonal antiphosphotyrosine antibodies and goat-anti-rabbit IgG alkalinephosphatase conjugates for all the other antibodies). The membranes werewashed again with TTBS and rinsed with TBS (without Tween 20). The colorreaction was performed in 100 ml bicarbonate buffer (10 mM NaHCO3, pH 9.8, 1mM MgCl2) with the addition of the alkaline phosphatase substrates NBT (30 mgin 1 ml DMF) and BCIP (15 mg in 1 ml 70%(v/v) DMF). The color reaction wasallowed to proceed for 30 min or several h depending on the intensity of thebands.10. Purification of S6 kinaseA FPLC system from Pharmacia LKB Biotechnology Inc. was used in thepurification procedures, which were carried out at 4°C. Skeletal muscle extractscontaining 4-5 grams of protein was diluted in buffer A (10 mM Tris-HCI, pH 7.4,5 mM EGTA, 2 mM EDTA, 20 mM p-glycerolphosphate, 5 mM p-methyl asparticacid and 1 mM DTT) until the conductivity was lower than 3 mmho, and thesample was then loaded at a flow rate of 5 ml/min into a fast flow Q Sepharosecolumn (5 cm x 20 cm) pre-equilibrated with buffer A. The column was thenwashed with 1000 ml buffer A and eluted at 6 ml/min with a NaCI gradient of 0-800 mM in a volume of 3000 ml. Fractions of 14 ml were collected and 5 ml wastaken to assay for S6 kinase activity. The first peak from two Q sepharosecolumn runs was then pooled and diluted in buffer B (10 mM MOPS, pH 6.25, 25mM (3-glycerolphosphate, 5 mM EGTA, 2 mM EDTA and 1 mM DTT). The mixture of protease inhibitors were added to a final concentration as indicated for101the homogenization buffer, and the pooled fractions were then loaded at 5ml/min into a fast flow S Sepharose column (5 cm x 5 cm) pre-equilibrated inbuffer B. The column was washed with 300 ml buffer B and eluted at 6 ml/minwith a NaCI gradient of 0-800 mM in a total volume of 600 ml (80 x 7.5 mlfractions). The fractions were assayed for S6 kinase activity and peak fractionswere pooled and concentrated with an Amicon Centriprep-10 to 20 ml, and 10 mlwere loaded into the Superdex 200 column which was equilibrated in buffer C(same as buffer B except that the pH was 7.2, and contained 0.01%(v/v) Brij 35and 150 mM NaCI). Fractions of 2.5 ml were collected with a total elutionvolume of 140 ml, and assayed for S6 kinase activity, and the procedure wasrepeated one more time. Peak fractions from the two runs were pooled anddiluted in buffer A containing 0.01 %(v/v) Brij 35, and applied to a MonoQ column.The column was eluted with a 25 ml gradient of NaCI (0-800 mM). Fractions of0.5 ml were collected and assayed for S6 kinase activity. The peak fractionswere pooled and diluted in Buffer B containing 0.01 %(v/v) Brij 35, and applied toa MonoS column. Fractions of 0.5 ml were collected and assayed for S6 kinaseactivity. The peak fractions were pooled and concentrated with an AmiconCentriprep-10 into a small volume. The sample was then applied to a Superose12 column, which was equilibrated with buffer E (same as buffer C but withoutNaCI). The column was developed at 0.5 ml/ml and 0.25 ml fractions werecollected for S6 kinase assay. The peak fractions were finally pooled andapplied to a heparin-agarose column which was equilibrated in buffer Bcontaining 10% ethylene glycol and 0.01 %(v/v) Brij 35. The column was elutedwith a 60 ml gradient of NaCI (0-800 mM). Fractions of 1 ml were collected andassayed for S6 kinase activity. The peak fractions were pooled and stored in thesame buffer at -70°C until further analysis. The purification protocol is summarized below:102Crude rat skeletal muscle extractsQ Sepharose column1S Sepharose column (Pooled Peak 1)1Superdex 200 (concentrated S Sepharose peak)1MonoQ column (30 kDa peak 2)MonoS (dominant peak)Superose 12 column1Heparin-agarose columnPurified S6 kinase11. Phosphoamino acid analysis.To determine the the phosphoamino acid of specificity of the purified S6kinase, S6 protein phosphorylated by the kinase was resolved in a SDS-PAminigel and transferred to a PVDF membrane. The PVDF membrane wasstained with 0.1%(w/v) amido black (45:45:10, methanol:H20:acetic acid) andthe S6 protein band was excised. The band was cut into fine pieces and hydrolyzed with constant boiling HCI for 90 min. The sample was then dried and103washed several times in a Speed-Vac, and dissolved in about 20 gl buffer (0.5%(v/v)pyridine, 5%(v/v) acetic acid) containing 1 mg/ml each of phosphoserine,phosphothreonine, and phosphotyrosine. The sample was then electrophoresedon a TLC plastic sheet at 1000 V for 60 min. The TLC plastic sheet wasvisualized by spraying with 0.25%(w/v) ninhydrin in ethanol and drying the sheetin an oven. An autoradiogram of the TLC sheet was then obtained to determinethe identity of the phosphorylated amino acids.12. Statistical analysis.As indicated in the "Results", some data were expressed asmean±standard error of the mean (S.E.). Statistical significance was determinedby one-way ANOVA with a computer program, Number Cruncher StatisticalSystem (NCSS), and the significance level was set at p<0.05.104RESULTSPart 1. Insulin-stimulated Protein Kinases in Normal RatsI. Assay Validation :To determine the optimal conditions for kinase assays, a time course andprotein concentration curve were performed for MBP kinase, CK-II, S6 peptide kinaseand S6 protein kinase using crude skeletal muscle extracts from insulin-stimulatedsamples. The results indicated that the reaction was linear for at least 15 min at afinal protein concentration of less than 0.25 mg/ml for the paper assays using MBP,casein, phosvitin, or S6 peptide as substrates, (Figure 5-6), and linear for more than30 min for gel assay using 40S ribosome at a final protein concentration of less than 1mg/ml. (Figure 5&6)II. Characterization of Insulin-stimulated Seryl/threonyl Protein Kinases in RatSkeletal Muscle.1. Time course of glucose concentration.Insulin was intravenously injected into fasted rats at a dose of 10 U/kg. Bloodsamples were obtained and analyzed for glucose levels in the serum to demonstratethat the injections were successful and that the injected insulin was biologically active.Serum glucose dropped at 5 min after injection, and severe hypoglycemia wasobserved 15 min after insulin injection. By 30 min, the serum glucose concentrationwas lowered to 2.0±0.5 mM from 6.3±0.9 mM (Figure 7A).105Figure 5. Time course of protein kinase assays. Phosphotransferase activity in crudeskeletal muscle extracts from insulin-treated rats were assayed using MBP (A),phosvitin (B), S6 peptide (C) and 40S ribosomal subunits (D). The reaction mediumcontained 0.25 mg/ml proteins (or 1 mg/ml for the assay with 40S subunit withsubstrate) from muscle extracts and the substrates were 5 mg/ml for MBP andphosvitin, 1 mg/ml for S6 peptide, and 0.35 p.M for 40S ribosome. The reaction wasallowed to proceed for various periods of time and stopped by spotting the reactionmedium onto P81 phosphocellulose papers, or by adding SDS sample buffer (for 40Sassay). Free [y-32P]-ATP was removed by phosphoric acid washings or by SDS-PAGE for the 40S assay. The paper squares or S6 protein band excised from gelswere counted for radioactivity as described in 'MATERIALS AND METHODS". Theconcentrations of 40S ribosomal subunits were calculated according to its A260values. 1 nmol 40S subunit = 17.2 A260 units.106100075000 -4000 -90002000:BC10Ew4.wEOasO0G.0Oa)oC00000 20 3010(mFigure 6. Protein concentration curve of protein kinase assays. Kinase assays asdescribed in Figure 5 were performed with different concentrations of protein frommuscle extracts. The reaction was allowed to proceed for 10 min for the paperassays and 30 min for 40S assay.1086000 -5000-4000  -3000-2000 -1000 -E^1600 -c.1200 -QS800 -E0^400 -CU0RI?.40000.Co0^3000 -.00"a^2000 -m 1000 -1800 -4-Q1200 -800 -400 -0.00^0.90^0.60^0.90Protain Connentration (mg/ml)1.20109Figure 7. Time course of glucose levels after insulin injection.(A) Rat blood samples were obtained at various time points after intravenous injectionof insulin (10 unit/kg) when skeletal muscles were taken. The glucose concentrationin the serum was determined using an assay kit as described under "MATERIALSAND METHODS'. Data points represent mean±S.E. of 4-5 experiments. *p<0.05 ascompared to the control value by one-way ANOVA.(B) The rats were injected with insulin under a glucose clamp. Blood samples weretaken every 3 min for determination of glucose with a glucose analyzer. n=6 for timepoints up to 15 min, n=3 for 18 and 21 min, n=2 for 24-30 min.1106.04.02.00.0.^B5^10^15^20^25Time after insulin injection (min)6.04.02.00.00 30111Since a precipitous drop in serum glucose would invariably induce theactivation of glucose counterregulatory mechanisms such as release of epinephrine,glucagon, growth hormone, and cortisol (Gerich, 1988; Cryer, 1989), it was importantto ensure that the subsequent biochemical analysis was not a consequence of thiscompensatory response. A glucose clamp experiment was thus performed. Asshown in Figure 7B, blood glucose was maintained at a constant level after insulininjection when the glucose clamp was employed.2. Insulin-induced Tyrosine Phosphorylation in Rat Skeletal Musclelmmunoblotting of muscle extracts with PY-20 antibody was conducted toobserve if intravenous injection of insulin could cause tyrosine phosphorylation ofcellular proteins in the skeletal muscle. Several proteins at 106 kDa, 87 kDa, 78 kDa,55 kDa, and 32 kDa became phosphorylated on tyrosine as early as 5 min afterinsulin injection, reached maximum at 10-15 min, and decreased afterwards (Figure8). The identities of these proteins were not determined.3. Effects of Insulin Injection on Protein Kinase Activities in Rat Skeletal MuscleWhen the crude skeletal muscle extracts prepared from rats treated with insulinfor various periods of time were assayed for the activities of several serine/threonineprotein kinases, it was observed that MBP kinase and S6 peptide kinases weresignificantly activated at 10 min after insulin injection, peaked at about 15 min, andthen started to decline (Figure 9A-B). In the meanwhile, the S6 protein kinaseactivity determined with 40S ribosomal subunit as substrate continued to rise even at30 min after insulin injection (Figure 9C). However, the activity of CK-2 as assayedwith phosvitin did not show significant activation by insulin injection (Figure 9D). In112Figure 8. Insulin-stimulated protein tyrosine phosphorylation in rat skeletal muscle.Muscle extracts prepared at various time points after insulin injection were resolvedwith SDS-PAGE and transferred to a nitrocellulose membrane, which was probed withan anti-phosphotyrosine monoclonal antibody PY-20. Proteins of various sizes areshown to be tyrosine phosphorylated after insulin injection. The identities of theseproteins are not known. The sizes of molecular standards are shown on the left.113Antiphosphotyrosine Antibody PY-200 5 10 15 20 30TIME (MIN)114Figure 9. Insulin-induced activation of protein kinases in crude skeletal muscleextracts. Rat skeletal muscles were excised from rat hind legs at various time pointsafter intravenous injection of insulin (10 U/kg) and crude extracts were prepared asdescribed under "MATERIALS AND METHODS'. Phosphotransferase activitiestoward MBP (A), S6 peptide (B), ribosomal 40S S6 protein (C), phosvitin (D), and rafpeptide (E) were assayed. The data points represent mean±S.E. of 5-6 experimentsusing different animals except for raf peptide (E) where the experiment was repeatedonly once. *p<0.05 as compared to the control value by one-way ANOVA.11570605040302043302520 .03503002502001501005015913)00:3(1)Q.) Q.Cl)0ELcia5.4)c „,03 ca.ci)?:3ct.cococbZ-0-03 ci.ooco61Q3 v.:u)03 et-C.)Cti t-334(1) EU)ct3 (I)eLE-cso_u_•zteCaddition, using a peptide substrate for p74raf kinase, we observed activation of rafkinase by 10 min after insulin injection as well (Figure 9E). However, this is notdefinitive, since the experiment was done only once.The crude muscle extracts were resolved on 11cY0 SDS-PA gels and antibodiesagainst MBP kinase, CK-2, S6 kinases as well as raf kinase were used forimmunoblotting. Since a large amount of actin was present in the crude muscleextracts which bound the antibodies strongly, a strong wide dark band of actin atabout 43 kDa completely masked the MBP kinase or CK-2 bands, which have M r of-41,000-45,000. The immunoblots with antibodies against MBP kinase and CK-2 didnot yield meaningful results (data not shown). The antibody for p90rsk (rsk-CTantibody) recognized a ladder of bands of -84-92 kDa, and the antibody for p70s 6k(S6K-III) immunoreacted with a band of about 65 kDa (Figure 10A-B, respectively).More interestingly, the p65s 6k displayed an apparent band shift, whereas the p9Orskbands did not exhibit apparent band shift. Also detected were bands of -90 kDa inimmunoblots with S6K-III antibody (Figure 10B). These might be the same p90rskbands that were detected with rsk-CT, suggesting cross-reactivity of the two isoformsof skeletal muscle S6 kinases. For raf kinase, three different antibodies were used,and all of them could recognize a band at -81 kDa (Figure 10C-E). Interestingly, aband of -95 kDa (Figure 10D) and -65 kDa (Figure 10E) were detected by two of tehanti-raf antibodies. While it is possible that these bands may represent non-specificimmunoreactive proteins in the skeletal muscle extracts, they may also be thep95rafB and p68rafA, respectively. It should be noted that no apparent band shiftwas observed in these immunoblots.117Figure 10. Immunoblotting of crude muscle extracts with antibodies against S6kinases and p74ran. Proteins in skeletal muscle extracts prepared from rats atvarious time points after intravenous injection of insulin (10 unit/kg) were resolved on11%(w/v) SDS-polyacrylamide gels for visualization by Western blotting using anti-peptide antibodies against p9Orsk (rsk-CT, panel A), p70s 6k (S6K-III, panel B) andpmraf/ (Medac anti-c-raf, panel C; NCI anti-v-Raf, panel D, and CRB-raf, panel E).The experiment was done three times with essentially identical results. The kinasesare indicated by arrows on the left side and the sizes of molecular standards areshown on the right side.1181194 M of mn ro ra•BP KinIThe muscle extracts from control and insulin-treated rats were fractionated withan anion exchange column (MonoQ HR5/5), which resolved at least 6 peaks of MBPkinase activities (Figure 11A). While the first peak was inactivated by insulin, theother 5 peaks were significantly activated by insulin injection. The most dramaticactivation occurred with peaks II, Ill, IV and V. The kinetics of the inactivation of peakI and activation of the other peaks by insulin are shown in Figure 11 B-G. It isapparent that the inactivation of peak I and activation of other peaks are time-dependent, with different kinetics. While the maximal inactivation of peak I andmaximal activation of peak II and III occurred at 10 min after insulin injection, themaximal activation for peak IV and V was at 15 min. Although it appeared as if thepeak VI was maximally activated at 5 min, this was not very definitive since thefractionation of the 5 min muscle extracts was performed only once.5. Identification of MBP kinase isoforms by immunoblotting and columnchromatography. The MBP kinase peak fractions from both control and insulin-treated sampleswere probed with antibodies raised against peptides synthesized after varioussegments of the rat brain MAP-2 kinase (ERK1 gene). Most strikingly, in the fractionsof peak II MBP kinase from insulin-treated samples (Figure 11A), all the antibodies forMBP kinases immunoreacted with a protein band of either 44 kDa or 42 kDa, or both(Figure 12B-D). Immunoblotting with antiphosphotyrosine antibody 4G10 indicatedthat the 42 kDa band in the peak II fractions were also tyrosine phosphorylated byinsulin stimulation (Figure 12E), as no immunoreactivity was detected with the samefractions in the control samples (data not shown). The strongest immunoreactivity120Figure 11. MonoQ column chromatography of insulin-regulated MBP kinases andtheir time course of activation. (A) Crude muscle extracts (10 mg) from control (0min) and rats treated with an intravenous injection of insulin (10 U/kg) for 15 min werefractionated on a MonoQ column using a Pharmacia FPLC system with a lineargradient of NaCI (0-800 mM). The gradient was applied over 80 fractions of 0.25 mleach, and the fractions were assayed for MBP phosphotransferase activity. Theexperiment was performed several times with extracts from control rats and from ratstreated with insulin for various period of time (5-30 min). The fold changes in the sizeof the 6 individual peaks of MBP kinase activities (I, II, Ill, IV, V and VI) from samplesthat had received insulin treatment for different time periods were compared withthose from the control samples and averaged as shown in panel B, C , D, E, F andG, respectively. The number in brackets above the data points denotes the numberof times the experiment was repeated for that time point.121(3)(3 )(1)(6)I. I.I( 1 )(3)(1) J. (5)11(6) (3)F^Peak V( )0.0^• 10 51 1•I a^I nf.!10 15 20 25 30Time (min)' I^' I '^I ' I^• 1^0.010 15 20 25 30Time (min).^10 51221.201.000.800.600.400.200.001.51.00.5I' I' I.^0.0- 0.5Peak VI^- 2.5- 2.0- 1.5(6)(5 )^(3)^(3.)-I 1.0- 0.5. I^I^I.^I.I^. 1^0- Peak II6Peak IV^- 2.5- 2.0(3)- 2D^Peak HI(6)(5)( 1 )^(3)• tat^•Ia^I•I^AI(3 )300E• --- 2504.-- c.)cu L-. 200(3)$3) aco^150ca Q,...--.-32 ..^100Q- °CO E 50Z 00 10^20^30^40^50MonoQ fraction number60Figure 12. Immunoblotting of MonoQ-fractionated MBP kinases. Muscle extractsfrom rats treated with insulin were fractionated with a MonoQ column as described inthe legend to Fig.4. aliquots (1000 of certain fractions were resolved on 11% SDS-polyacrylamide gels for Western blotting with antibodies raised against the peptidespatterned after the N-terminus (erkl-NT, panel A), subdomain III (erk1-111, panel C),and the C-terminus (erkl-CT, panel D), of rat brain p44erki and against the ATP-binding domain of MAP kinase (mpk-I, panel B). Antiphosphotyrosine antibody (4G10,panel E) was also used for the immunoblotting of the same fractions. The position ofthe MBP kinase peaks are indicated above the panels. The experiments wererepeated at least four times with essentially identical results.123CMBP Kinase PeaksAaerki- NTBa mpk- Iaerkl-IIIaerki- CT-44K-44K-44K-44K-44K4 8 13 15 17 18 19 20 2123 25 27 29 34 40 46 50MonoQ fraction numbera P-Tyr4G10EIIMOW Oar OOP 4watki.ENO ON. 0111N,,IV V VIaft 16110 0101iiii ow.was observed with either anti-erk1 III antibody or anti-erkl CT antibody (Figure 12C-D). An apparent band shift was evident for the 42 kDa band (Figure 12C-D) as wellas for the 44 kDa band (Figure 12B). In the fractions of MBP kinase peak III and IV,the 42 kDa immunoreactive band became very faint, but the 44 kDa band stillappeared to be present, albeit only in the blots where anti-erki-111 and anti-erkl-CTantibodies were used. In fractions of peak V MBP kinase, only the anti-erk/-IIIantibodies recognized a distinct band of about 44 kDa, while no immunoreactive bandwas detected in the peak VI. Additionally, except for a higher M r band (about46,000) visualized with anti-erk/-11I antibodies (Figure 12C), no new immunoreactivebands were detected for the peak I MBP kinase that became inactivated in responseto insulin.To further identify the MBP kinases in rat muscle extracts, the samples fromboth control and insulin-treated rats were fractionated first by hydrophobic interactionchromatography using a phenyl-Sepharose column, and then the resolved peakswere fractionated by anion exchange chromatography with a MonoQ column. Withthe phenyl-Sepharose column, two peaks of MBP kinase activities were observed. Asharp peak was eluted first and a broad peak subsequently (Figure 13A). Theactivation of MBP kinase activities in the first peak (PS-I) was quite small, while asignificantly greater stimulation was seen in the second broad peak (PS-II). The PS-Iand PS-II were then each subjected to MonoQ column chromatography, whichresolved a single dominant peak for both PS-I and PS-II (Figure 13B-C). The positionof PS-I on MonoQ corresponded to the position of the peak V MBP kinase in theprofile of MonoQ column chromatography of the crude muscle extracts (see Figure11A), while the position of PS-II on MonoQ column corresponded to the peak III ofMBP kinase as designated in Figure 11A.125Figure 13. Sequential phenyl-Sepharose (PS) and MonoQ fractionation of MBPkinases. (A) Crude muscle extracts from control (0 min, o) and insulin-treated rats (15min, •) were loaded onto a phenyl-Sepharose column (2.5x10 cm), which was thenwashed with 50 ml starting buffer and eluted with a gradient of Brij 35 (0-3%) over 80fractions of 5 ml each. The fractions were assayed for the MBP phosphotransferaseactivity as described under "MATERIALS AND METHODS'. The two MBP kinasepeaks resolved on the column, PS-I and PS-II, were then further fractionated with aMono Q column (HR5/5) with a NaCI gradient (0-800 mM) as shown in panel B and C.The Roman numbers above the peaks of MBP kinase resolved on MonoQ columnindicate their position relative to the MonoQ profile of the crude extracts as shown inFigure 11.126A^PS-I^—0- o min-11- 15 minPS-II•350E'(.1ccs300250Q)Cb 200RS z150a-100E 500C III PS-II350300250200150100500>, 250cts 200ti cx 150RSSC.aca E1005000 10 20 30 40 50 60 70 80Phenyl-Sepharose fraction number10 20 30 40 50 60^10 20 30 40 50 60MonoQ fraction number MonoQ fraction number127When the peak fractions from MonoQ column were blotted with anti-MAPkinase antibodies, only a band of about 44 kDa was detected with the threeantibodies (Figure 14). However, the 44 kDa bands in peak III and peak V appear tobe distinct from each other, as indicated by the difference in their immunoreactivitywith different antibodies. For example, the antibodies against sea star p44mPk (anti-mpkl)) and the anti-erkl-CT antipeptide antibodies reacted poorly with the 44 kDaband in the peak III fractions, while a strong immunoreactivity was observed for thepeak V fractions. On the other hand, the anti-erk1-11I antipeptide antibodyimmunoreacted with the 44 kDa band in both peak III and peak V fractions, with astronger immunoreactivity for the peak III fractions. Furthermore, immunoblotting withantiphosphotyrosine antibodies indicated that the 44 kDa band in peak III fraction wasnot tyrosine phosphorylated in the control samples, and insulin produced a dramaticincrease in the tyrosine phosphorylation of this band. In contrast, the 44 kDa band inpeak V appeared as a dublet in both control and insulin-treated samples, suggestingthat this protein was already partially tyrosine phosphorylated in the control samples.It is noted, however, that insulin further stimulated tyrosine phosphorylation of thisprotein, as suggested by the increase in the upper band of the doublet in insulin-treated samples (Figure 14).MBP kinase isoforms have also been previously resolved by phenyl-Superosecolumn (Rossomando et al., 1991), another column of hydrophobic interactionchromatography. Therefore, we also used a phenyl-Superose column to fractionatethe muscle extracts from both control and insulin-treated rats. Two peaks of MBPkinase were resolved by this column (Figure 15A). The first peak appeared as abroad peak with very little activation by insulin, and it was followed by a second sharppeak which was greatly activated by insulin (Figure 15A). The peak fractions fromboth control and insulin-treated samples were again blotted with antibodies against128Figure 14. Immunoblotting of the MBP kinases resolved sequentially on phenyl-Sepharose and MonoQ. The antibodies used were the same as those described inthe legend to Figure 5 except that the antiphosphotyrosine antibody was PY-20instead of 4G10. Ins=insulin; con=control. Arrow indicates the 44-kDa protein in thepeak MAP kinase MonoQ fraction.129130Figure 15. Phenyl-Superose chromatography and immunoblotting of MBP kinases.Crude muscle extracts from control (0 min, o) and insulin-treated rats (15 min, •) werefractionated on a phenyl-Superose column with a linear gradient of Brij 35 (0-3%) over80 fractions of 0.25 ml each. The.. fractions were assayed for MBP kinase activity aswell as for Western blotting with the antibodies against MAP kinases as describedunder "MATERIALS AND METHODS'. The antibodies used are the same as those inFigure 12.131300"S• — 250u. 200a) Ct.c 150100ECa S 500 I •Er• • i• • • • , • • • •^• • • •^•^; • • • • • • 1 10 5 10 15 20 25 30 35Phenyl-Superose fraction40 45 50numberControlaerk1-111Insulinaerk1-111Controlaerki-CTInsulinaerk1-CTControla P-Tyr4010Insulina P-Tyr4010-44K-44K-44K-44K'-44KI-44K2 6 10 12 15 18 21 24 27 29 31 32 33 34 36 39 42 4546Phenyl-Superose fraction number132MBP kinases as well as antiphosphotyrosine antibodies (Figure 15B-G), whichindicated that a 44 kDa band was present in the fractions in first peak, while thefractions in the second peak contained a 42 kDa immunoreactive band. Both the 44kDa band and the 42 kDa band reacted strongly with the anti-erkl-CT antibody(Figure 15d-E), whereas the anti-erk/-11I antibody only reacted with the 44 kDa bandin the fractions in the first peak. Moreover, only the 42 kDa band in the second peakshowed an increase in the tyrosine phosphorylation in response to insulin stimulation(Figure 15F-G). An apparent band shift was shown by the anti-erkl-CT antipeptideantibodies for both the 44-kDa band and the 42-kDa band in response to insulinstimulation (Figure 15D-E).6. Insulin-stimulated S6 kinases and its column profile.With ribosomal 40S subunit as a substrate, two distinct peaks of S6 kinaseactivity were detected in the fractions obtained from MonoQ column chromatographyusing skeletal muscle from both control (0 min) and insulin-treated rats (15 min)(Figure 16A). The first peak (peak I) was eluted at NaCI concentration of about 100mM, and is activated by insulin by at least 4-6 fold, while the second peak (peak II)was eluted at NaCI concentration of about 350 mM with an activation by insulin ofabout 5-fold. The fold change in the size of each peak in response to insulin waspooled from several experiments with samples at different time points to yield a timecourse of activation for both peak I and peak II (Figure 16B-C), which was similar forthe two peaks, with peak II exhibiting slightly longer duration of activation.When the same fractions were assayed for S6 peptide kinase activity, at leastfive peaks were resolved by the MonoQ column, of which at least 4 peaks appear tobe activated by insulin treatment (Figure 17). The position of peak I corresponds to133Figure 16. MonoQ chromatography of S6 kinases. Crude muscle extracts fromcontrol (0 min, o) and insulin-treated rats (15 min, •) were resolved using a MonoQcolumn (HR5/5) with a linear NaCI gradient (0-800 mM) over 80 fractions of 0.25 mleach. The fractions were assayed for S6 protein kinase activity with ribosomal 40Ssubunit as a substrate, as described under "MATERIALS AND METHODS'. A typicalprofile of S6 kinases is shown in panel A. Muscle extracts from different time pointswere resolved in the same manner to determined the time course of activation of S6protein kinase peaks I (panel B) and II (panel C).134(3)(5)135Peak I Peak II(3 )(3)^(1)0^10^20^30^40^50MonoQ fraction number60-2(6)^ (3)(1) .i.,•, .,•1.1.1.I.1 ^05 10 15 20 25 30 0 5 10 15 20 25 30Time (min)^Time (min)10.08.06.04.02.00.0A^1^ —0-- 0 min—0-- 15 minFigure 17. MonoQ chromatography of S6 peptide kinases. Crude muscle extractsfrom control (0 min, o) and insulin-treated rats (15 min, .) were fractionated on aMonoQ column (HR5/5) with a linear gradient of NaCI (0-800 mM) over 80 fractions of0.25 ml each. Panel A, S6 peptide kinase profile as assayed using RRLSSLRA as asubstrate for phosphorylation as described under "MATERIALS AND METHODS'.Panel B-F, the time course of S6 peptide peaks. Crude muscle extracts preparedfrom different rats treated with intravenous injection of insulin for various durationswere resolved by a MonoQ column and the size of individual peak of S6 peptidekinases was compared with that of the control extracts. The fold changes in the sizeof the various peaks at different time points is plotted. Numbers in the bracketsrepresent the number of times the experiment was repeated.136)(6)- 3.0(1 )( 1 )(6) (5)^(3)(3 )Peak I Peak II(3).^1^•^1^4^1^It^1^.^1^A0.0i.1 .1 -^1.1^.1^  0.05 10 15 20 25 30Time (min)Peak V(1)(6)(5)^(3)(3)1.51.00.54.03.02.01.0^E^100.._,q) c):13 E•■••...^ca,q) 50Ct..COCr)^00^10^20^30^40^50^60MonoQ fraction number0 5 10 15 20 25 30Time (min)137 -A—0— 0 min—0— 15 minthat of peak I for the S6 protein kinase activity as determined with the 40S subunit asa substrate, although the fold activation was not as high as that shown by S6 proteinkinase activity. The second and third peak (peak II and peak III) fell into the positionwhere the peak II and peak III MBP kinase activity would appeared, suggesting thatthe same kinase may be responsible for these two peaks. The position of the fourthpeak (peak IV) was similar to that of the peak II S6 protein kinase activity as shown inFigure 16. There was a very minor activation of the fifth peak by insulin.The time course of activation for each individual peak was plotted the sameway as for MBP kinase peaks (Figure 17B-F). While peak I and II showed a gradualactivation by insulin treatment and reached peak activation by 15 min (Figure17B-C),the activation of peak III and IV appeared to be earlier (Figure 17D-E), being maximalat about 5 min. Peak V did not show significant activation during the 30 min period(Figure 17F).7. Immunoblotting of MonoQ Column Fractions with Antibodies for S6 Kinases.Two antipeptide antibodies for S6 protein kinases, anti-S6K-III and anti-rsk-CT,were used to probe for the S6 kinases in the MonoQ fractions obtained using samplesfrom both control and insulin-treated muscle extracts. The anti-S6K-III antibodyrecognized a band at about 100 kDa (indicated by an arrow) which was present in thefractions of S6 protein kinase peak II or S6 peptide kinase peak IV, and another bandat about 65 kDa in fractions where there was no apparent peaks of either S6 proteinor peptide kinase (Figure 18B-D). While the 100-kDa protein may be tentativelyidentified as an isoform of the p9Orsk based on its size and elution profile on MonoQ,the 65-kDa protein may be an p70rsk isoform in rat skeletal muscle. However, no S6kinase activity was detected in the fractions that contained the 65-kDa protein,138Figure 18. Immunoblotting of S6 kinases resolved on MonoQ. Fractions from aMonoQ column (HR5/5) were resolved on 10% SDS-polyacrylamide gels andtransferred to a nitrocellulose membrane for immunoblotting. The antibodies raisedagainst the peptide synthesized according to the amino acid sequence of thesubdomain III of the 70 kDa S6 kinase (S6K-III), and C-terminal sequence of thepgorsk (rsk-CT) were used to probe the fractions obtained using muscle extractsfrom both control (panels A, B, E, and F) and 15 min insulin-treated rats (panels C, D,G, and H). Arrows indicate the 100-kDa protein band that was putatively identified asa pgorsk.1393 5^STD 11 12 13 14 15 16 18 20 21 22 23 24 31 33 34 35Controla S6K-IIIInsulina S6K-IIIControla rsk-CTInsulina rsk-CTSTD 11 12 13 14 15 16 18 20 21 22 23 24 31 33 34 35Mono() fraction number1 3 5suggesting that this is either the p65s6k that was not activated by insulin, or a proteinthat non-specifically reacted with the antibody. Additionally, a band of lower M r(about 52,000) also appeared in these blots where no peaks of S6 kinase activitywere observed. The identities of these proteins cannot be determined in ourexperiment. In the washthrough fractions (1-5), a faint immunoreactive band of about54 kDa was detected by anti-S6K-III antibody (Figure 18A-C). The anti-rsk-CTantibody detected a similar band of about 100 kDa in the same fractions where the100 kDa immunoreactive band with anti-S6K-III antibody appeared (Figure 11 F-H),and the darkness of the band corresponded with the S6 kinase activity assayed witheither ribosomal 40S subunit (peak II in Figure 16) or with S6 peptide (peak IV inFigure 17), suggesting that this band may be responsible for the S6 kinase activity.Moreover, another immunoreactive band of about 105 kDa with anti-rsk-CT antibodyappeared to be present in fractions 20-22, but only in the insulin-stimulated samples(Figure 18H). The anti-rsk-CT antibody also recognized a band of lower M r at about52,000 in fractions where there were no S6 kinase peaks. In the washthroughfractions, anti-rsk-CT antibody immunoreacted strongly with two protein bands ofabout 90 kDa and 54 kDa (Figure 18E-G). The 54-kDa band appeared to be thesame band that was detected by the anti-S6K-III antibody.8. Gel Filtration Profile of S6 KinasesSince no immunoreactive bands were detected with either of the two antibodiesfor the S6 protein kinases in the fractions in the first peak S6 protein kinase as shownin the MonoQ column profile (Figure 16A), we attempted to estimate the size of thekinase in the first peak as well as the second peak using a gel filtration column(Pharmacia Superose 12 HR10/30). As shown in Figure 19A, the peak I S6 proteinkinase contained at least two species with a M r of about 40,000 and 90,000,141Figure 19. Gel filtration of the MonoQ S6 kinase peaks. The two peaks of S6 proteinkinases resolved on MonoQ column as determined with 40S as a substrate (see Fig.17) were further fractionated with a Superose 12 column (HR10/30) as describedunder "MATERIALS AND METHODS". The fractions obtained were assayed usingboth ribosomal 40S subunit (panels A and B for MonoQ peaks I and II) and S6peptide (panels C and D for MonoQ peaks I and II) as substrates. The elutionpositions are indicated for the marker proteins: thyroglobulin (669 K),immmunoglobulin G (156 K), ovalbumin (40 K) and cabonic anhydrase (17 K).1424800000006000000080060000000000080000Superose traction number-.'Q.:....)2its'O'Q.'8..Q.coco-.-.Q..._.()2'ci-o.'..O'cxQ..cbb•i—a)Q.toU)respectively, while the peak II S6 protein kinase also contains two species with M r of120,000 and 40,000, respectively (Figure 19B). These different species of S6 proteinkinases showed distinct substrate specificity, as indicated by the fact that theribosomal 40S subunit was a better substrate for the 40 kDa form S6 kinase than forthe 90 kDa form S6 kinase in the peak I S6 protein kinase (Figure 19A), whereas theS6 peptide served as a better substrate for the 90 kDa form of S6 protein kinase thanfor the 40 kDa form of S6 protein kinase in the peak I S6 protein kinase (Figure 19C).With the peak II S6 protein kinase, only the 40 kDa form S6 protein kinasephosphorylated both ribosomal S6 protein as well as S6 peptide, while the 120-kDaform S6 protein kinase did not seem to phosphorylate S6 peptide very well (Figure19B-D). Furthermore, additional species of kinases at about 90 kDa and 170 kDawere detected by S6 peptide but not by ribosomal 40S subunit (Figure 19B-D).9. MonoQ Chromatography and lmmunoblotting of Raf kinase in Rat Skeletal MuscleThe fractions from MonoQ column chromatography of muscle extracts fromboth control (0 min) and insulin-treated rats (15 min) were assayed with a syntheticpeptide based on a putative autophosphorylation site in p74raf. The MonoQ columnresolved three peaks that could phosphorylate the Raf peptide (Figure 20A). Themajor peak at fraction 34 appeared to be an S6 kinase, based on the ability of thesame fractions to phosphorylate S6 peptide and ribosomal S6 protein (data notshown). The position of this peak also coincided with the position of peak II S6protein kinase as determined with ribosomal 40S subunit (Figure 16A). When probedwith antibodies against raf kinase, the fractions in the second peak, which showed aslight activation by insulin stimulation, contain several immunoreactive bands, and themost prominent ones were -68-72 kDa (indicated by arrows in Figure 20B-C). Aclose examination of the immunoblots with the column profile revealed that the144Figure 20. MonoQ chromatography and immunoblotting of p74rafl. Crude muscleextracts from control (0 min, o), and insulin-treated (15 min, •), rats were fractionatedon a MonoQ column with a linear gradient of NaCI (0-800 mM) over 80 fractions of0.25 ml each. The fractions were used for assay of p74raf 1 using a peptide patternedafter the putative autophosphorylation site in Raf-1 as a substrate (panel A). Thefractions from both control (B) and insulin-treated extracts (A) were also resolved on11°/0(w/v) SDS-PA gels for immunoblotting with an antibody against a peptidesequence from subdomain Ill in human p74raf 1 . Arrows indicate the same proteinband that was detected in the control as well as the insulin-treated samples. A bandshift is apparent when the marked bands were compared between the control (B) andthe insulin-treated samples (C).1456010^20^30^40^50MonoQ fraction numberB*116.^Valliir411,16116° law" loy ill, IP It IIC .1111111...1••2 4 6o ^08 10 1112 1314 15 161718 1920212223 24MonoQ fraction number-97K-66K-50K-97K-66K-50K125100755025Controlarafl - IllInsulina raft- III11=10111 111111111 111111111,11111111111 it IIIintensity of these bands corresponded with the raf peptide kinase activity. Moreinterestingly, a band shift was apparent when the immunoblots from control sampleswere compared with those of the insulin-stimulated samples (compare the bandsindicated by arrows in Figure 20B with Figure 20C). While only a single band of -68kDa was present in the fractions in the second raf peptide peak from the controlsamples, two bands of -68 kDa as well as 72 kDa were detected in the samefractions from the insulin-treated samples. The immunoreactive bands at -95 kDacould be the raf isoform p95rafB. The same antibodies did not detect any bands ofthe same molecular weight in the fractions in the third and major peak of raf peptidekinases (data not shown). While these data suggest that raf kinase may be activatedin rat skeletal muscle, this has to be viewed with caution, since the peptide substrateis apparently not strictly specific, and the antibodies to raf detected multiple bandsbeside the activity peaks.10. MonoQ column profile and immunoblotting of CK-ll in rat skeletal muscle.After MonoQ fractionation of muscle extracts, the casein kinase activity in thefractions from control (0 min) and insulin-treated (15 min) rats were determined withthe use of phosvitin as a substrate. A single sharp peak of CK-2 activity wasobserved which did not differ between control and insulin-treated samples (Figure21A). The peak fractions were then probed with an antibody for the a subunit of CK-2, which detected a protein band at about 43 kDa (Figure 21B-C), indicating that thephosvitin kinase peak was indeed CK-2. Since this is the only peak of phosvitinkinase activity, the majority of the phosvitin activity in crude muscle extracts wouldrepresent the activity of CK-2. The amount of CK-2 protein in the sample from controland insulin-treated sample did not seem to differ significantly, as estimated from theintensity of the bands, although the CK-2 bands in the control fractions were seen147Figure 21. MonoQ column chromatography and immunoblotting of casein kinase-2.Crude muscle extracts from control (0 min, o) and insulin-treated (15 min, •) rats werefractionated with a MonoQ column using a linear NaCI gradient (0-800 mM) over 80fractions of 0.25 ml each, and the fractions were used to assay for CK-2 with phosvitinas a substrate (panel A). Aliquots of the fractions of the CK-2 peak were resolved on11 %(w/v) SDS-PA gels for immunoblotting with antibodies raised against the peptidesequence of the subdomain III in CK-2 (panel B and C for control and insulin-treatedsamples, respectively). In addition, the experiment was repeated with extracts fromdifferent time points and the fold changes in the peak size were plotted to show theactivity of the CK-2 peak as a function of treatment of the rats with insulin (panel D).148250""t* 200cu0°) o.(1) 150cuE 100G 0.0) E 50Q.00A —o- 0 min-fp- 15 min10 20 30 40 50 60MonoQ fraction number-50KControla CK2-IIIInsulina CK2-III-33K45 46 47 48 49 50 51 52MonoQ fraction number-44K-33K-44K1.4 -D1.21.00.80.60.40.20.0 ^0(2)( 1 )^(5)^(4)(2)• T^I^I5^10 15 20 25 30Time (min)149over -5 fractions, while the CK-2 bands in the insulin-treated samples wereconcentrated in 3 fractions with wider bands. One possible explanation may be aband-shift in the insulin-treated sample, which resulted in a wider band (possibledoublet). The time course of the CK-2 peak as determined from MonoQchromatography of samples from different time points indicated that the activity of CK-2 did not change significantly during the 30 min period after insulin injection (Figure21D).11.Insulin-stimulated Kinases in Euglycemic Clamp StudyMonoQ chromatography was performed using samples prepared from rats thatwere injected with insulin under a glucose clamp to maintain a constant level ofglucose. The MonoQ profile of MBP kinase peaks appeared similar to the profile inrats that did not undergo the glucose clamp (Figure 22A). It was noted, however, thatpeak III was smaller whereas peak IV was larger in the clamped rats (compare Figure22A with Figure 11A). Similar to that in non-clamped rats, two S6 kinase peaks andfive S6 peptide peaks were resolved in the clamped samples by MonoQ columns(Figure 22B-C), but the peak I of S6 kinase appeared to be smaller than that of thenon-clamped rats, although the size of peak II S6 kinase did not change significantly(compare Figure 22B with Figure 16A, note the difference in the scale of Y axis).12.Tissue Distribution of KinasePhosphotransferase activities toward MBP, phosvitin, S6 peptide, and 40Sribosome were examined in brain, spleen, skeletal muscle, and testes from Wistarrats. A marked difference was noted in the specific activities of the kinases indifferent tissues (Figure 23). Strikingly, spleen had the highest levels of activity for all150Figure 22. MonoQ chromatography of insulin-regulated MBP kinases in skeletalmuscles of glucose-clamped rats. The skeletal muscle extracts were prepared before(o), 10 min (A), and 15 (•) after insulin injection into rats that were under glucoseclamp. For each time point, 10 mg of protein were loaded onto a MonoQ column andfractionated as described in the legend to Fig 11. Column fractions were assayed forphosphotransferase activity toward MBP (panel A), 40S ribosomal protein S6 (panelB), and S6 peptide RRLSSLRA (panel C). Similar results were obtained in twoseparate experiments.151Z,-&--:r ,..Cu cua5 Q-0Qs c))C:Sc-0- 0ca EZ ::?:-.—s-.--_-. .--U mcu cxcbco --as EQ .....,--^R--(0CJ) ■131-:Z`-"S1.5 :"--RI Ea) L-(i) zca 1:)---• c-sc ---• ,..:-23 0•-4-61 Eco Q-ci.caco1501005005.04.03.02.01.00.0500400300200010Figure 23. Tissue distribution of protein kinases. The activities of MBP kinase(MBPK), CK-2, S6 peptide kinase (S6 PK), and S6 protein kinase (S6K) weredetermined in skeletal muscle, spleen, brain, and testes from Wistar rats that were nottreated with insulin. Protein kinase assays were performed as described in"MATERIALS AND METHODS'. All of the kinase activities were normalized againstprotein concentrations in the samples. Values represent the average from 6 rats andthe kinase assays were done in duplicate.15312001000800600400200606040302010150012009008009000.700.600.500.400.300.200.100.00404^faS6 PK154of the kinases, while skeletal muscle, an insulin-sensitive tissue, had the lowestactivity for all of the kinases (Figure 23). The specific activity of MBP kinase, S6peptide kinase, and S6 kinase were nearly 20- to 30-fold higher in the spleen than inskeletal muscles. It was also noted that the activity of casein kinase was comparableamong spleen, brain and testes, but much lower in skeletal muscles.The MBP kinase and casein kinase activity in spleen were further analyzedwith MonoQ chromatography. A single distinct peak of MBP kinase activity (Figure24A) and a phosvitin kinase peak (Figure 24B) were resolved by MonoQ column, andeluted at 400 mM and 450 mM NaCI, respectively. Importantly, both peaks weresignificantly activated in response to insulin, with the MBP kinase peak and phosvitinpeak stimulated by 2- and 5.5-fold, respectively (Figure 24). The position of elutionfor the MBP kinase peak appeared to correspond to the peak V of the skeletal muscleMonoQ profile, while the phosvitin peak corresponded to the CK-2 peak of theskeletal muscle MonoQ profile, indicating that the activity of phosvitin kinase in spleencan also reliably represent the activity of CK-2 (refer to Figure 11 and 21).Part II. Insulin-stimulated Protein Kinases in Diabetic Rats and the Effects ofVanadium Treatment.I. General characteristics of the animals.1. One-month StudyThe diabetic rats in both the one-month groups had significantly elevatedplasma glucose levels and decreased body weight. One week after the treatmentwith vanadyl sulphate in the one month diabetic rats, the plasma glucose levels weresignificantly lowered and by four weeks, the glucose levels were complelely155Figure 24. MonoQ chromatography of spleen MBP kinase (A) and casein kinase (B).Spleen extracts were prepared from control (o) and insulin-treated rats (•) (20 min),and were resolved on a MonoQ column with a linear NaCI gradient (0-800 mM) over80 fractions of 250 1.1.1 each. The fractions were used to assay for MBP (A) andphosvitin (B) phosphotransferase activity.1562500 -"a'74.  2000ia 1503o-t,••■1>Z 1 000 --:C)WA4in^500-^/A0 --6-0-9-6-0-*-0-60-04-0-9-0—^0^10^20^30^40^50^60^70^80•\ii'Ct1..1P.'—•— Insulin—0— CONCosa 120^ •a 100 -Ei880-IIIITa^ I •60•-al,Fraction Number20Id^•\U 0 9-47"—abwir-:-"72-?4)-44144V. 11116 • t— • 9 o0^10^20^30^40^50^60^70^80157normalized by the treatment (Figure 25). The glucose levels in control rats were notaltered by the treatment.2. Two-month and Six-month StudyHalf of the rats in this study.received vanadyl sulphate or BMOV treatment. Asthe vanadium compounds were administered orally in drinking water, the actual dosefluctuated on a daily basis throughout the study (Figure 26A). On average, however,the diabetic rats treated with vanadyl sulphate received about 0.45 mmol/kg/day.The initial body weight was not different between control and diabetic rats(212±4 g), but the body weight gain was much lower in the diabetic rats than thecontrols, and by the end of the 2-month period, the untreated control rats had a muchhigher body weight (541±55 g) than did the untreated diabetic rats (368±28 g) (Figure26B). The control treated rats also had a significantly lower body weight (424±57 g),while the treatment did not improve the body weight gain in the diabetic rats (385±61Before vanadium treatment, the serum glucose levels were significantly higherin the two-month diabetic rats (16.9±1.1 mM) than those in the control rats (6.0±0.2mM). The untreated diabetic rats had a high level of blood glucose throughout thecourse of study (-18 mM), while the vanadyl sulphate-treated rats showed a lowerglucose levels after one-week treatment, and the glucose levels in these rats werenearly normal by the end of the second week (-7.8 mM) (Figure 27A). The glucoselevels were completely normalized at the end of the 2-month treatment (5.6±0.6 mM).Similar results were observed in the 6-month study, although 4 out of 12 BMOV-treated diabetic rats did not respond well to the treatment with BMOV and had15825140o --,^20_‘4^ a...cdig^ao 15 - InO *^i ^oC.)O 10m0o ff------" l^i--^0C.7^5 -mEl—0— CON- CON+V--0-- DIA- DIA+V0^1^2^3^4^5Weeks After Vanadyl Sulphate TreatmentFigure 25. Time course of glucose levels in 1-month control (CON) and diabetic (DIA)rats. V indicates vanadyl sulphate treatment. Blood samples were taken from the rattails and plasma glucose concentration was analyzed by a glucometer. *p<0.05 vsCON, #p<0.05 vs DIA, n=7-13Figure 26. The dose of vanadyl sulphate (A) and body weight (B) of 2-month control(CON) and diabetic (DIA) rats. V indicates vanadyl sulphate. Vanadyl sulphate wasadministered in the drinking water and the dose was calculated according to theconcentration of vanadyl sulphate, the water consumption and the body weight of therats. Data points represent mean±SEM from 5-8 rats.160E 500-E3O 450-.E 400-Oco350->+o 300-1T■r"r•x.-474 131-'go-tr250-CON -"e• CON+V-3•• DIA —11 . DIA+V600550-200 ^AGs 1.50-co-NC1.20-E0.90-.cO"aGI•^0.30-O>0.00 ^0a—0— CON+V—E— DIA+VTIT112-11\TTT,..TrTTT•0 10^20^30^40^60^60^70TTOI^•^I^•^I^•^I^•^I^I^•^IDays of Treatment0 10^20^30^40^50^60^70Days of Treatment161Figure 27. Time course of glucose levels (A) and terminal glucose concentrations (B)in 2-month control (CON) and diabetic (DIA) rats. V indicates vanadyl sulphatetreatment. For the time course (A), blood samples were taken from rat tail veins andthe glucose concentration was analyzed with a glucometer. At the end of theexperiments, half of the rats in each group received an insulin injection. The bloodsamples were taken 20 min after insulin injection and serum was prepared for glucoseassay using an assay kit. *p<0.05 vs the corresponding -insulin group, #p<p0.05 vsCON-insulin group, n=4-8.162.E. 90Ec 25 -0Tf..C. 20-8c0 15C.)a)GT 1 0 -05 50iT30 o .^0.^.^.^.^.^.10^20^30^40^60^60^70--0— CON—s— CON+V—0— DIA—.— DIA+VDays of Treatment2 25EcO 20.77.Pi,C.00 15C00o 10000=6 6E= M -InsulinM. +InsulinCON CON+V^DIA^DIA+Vglucose concentrations >14 mM. These rats were referred to as "non-responders",even though they responded partially to the treatment.The untreated diabetic rats exhibited severe insulin resistance, as indicated bythe fact that 20 min after a large dose of intravenous insulin (10 U/kg), the serumglucose concentrations remained unchanged (Figure 27B). The treatment withvanadyl sulphate for two months not only normalized the basal serum glucose levels,but also increased the sensitivity of glucose-lowering effects of insulin in the diabeticrats.II. Protein Kinase Activities in the Crude Muscle Extracts1. One-month control and diabetic rats.It was observed in this study that the basal levels of MBP kinase activities werenot signficantly altered in the diabetic rats, and neither insulin injection nor vanadylsulphate treatment had significant effects on MBP kinase activities (Figure 28A). Wedid not observe any activation of CK-II by insulin injection into rats (Figure 28B).Insulin injection caused slight activation of S6 peptide kinase in control and diabeticrats but not in vanadyl sulphate-treated control rats, the fold activation was notdifferent between control and diabetic rats (Figure 28C). When 40S ribosome wasused as a substrate, significant activation of S6 kinases was observed, and theactivation was significantly higher in the control compared to the diabetic rats. Theactivation was normalized following one-month of treatment with vanadyl sulphate(Figure 28D).164Figure 28. Protein kinase activities in the crude muscle extracts from the 1-monthcontrol (CON) and diabetic (DIA) rats. V indicates vanadyl sulphate treatment.Skeletal muscle extracts were prepared from rats that were not injected with insulin aswell as rats that received insulin injection 20 min before removing skeletal muscles.The skeletal muscle extracts were assayed for MBP (A), phosvitin (B), S6 peptide(RRLSSLRA) (C), and S6 protein (D) phosphotransferase activity. Kinase activity wasexpressed as Pi transferred to substrates per min per mg protein except for S6 kinasewhere the incorporation of 32P into S6 protein (cpm) was used. *p<0.05 vscorresponding -insulin groups, n=3-6.165* **aEcE0E0.>1>15<0cocdc2c00a60504030201002015105060604030201001000. D800 -600 -400 -200 -CON CON+V^DIA^DIA+V1662. Two-month control and diabetic rats. The basal MBP kinase activities were not different between the control anddiabetic rats, and were not significantly stimulated 20 min after insulin injection ineither the control or the diabetic rats (Figure 29A). Casein kinase activity was notchanged in diabetic rats (Figure .28B). The basal S6 peptide kinase activity wassignificantly increased in the diabetic rats (1.6 fold), and was reduced to near controlvalues by the vanadyl sulphate treatment (Figure 29C). Insulin slightly stimulated theS6 peptide kinase activity in both control and diabetic rats, and the magnitude ofstimulation was 1.1-fold in the diabetic rats and 1.3-fold in the control rats (Figure29C). When 40S ribosome was used as a substrate, the basal S6 kinase activity wasincreased by 2.8-fold in the diabetic rats, and was normalized by treatment withvanadyl sulphate (Figure 29D). As well, the activation of S6 kinase activity by insulinwas much higher in the control rats (8.2-fold) than in the diabetic rats (1.6-fold)(Figure 29D). Treatment with vanadyl sulphate largely restored the activation of S6kinase by insulin to 6.3-fold (Figure 29D).3. Six-month control and diabetic rats.In contrast to the two-month studies where no changes in the basal MBPkinase activity were noted in the diabetic rats, there was a significant decrease (30%)in the basal MBP kinase activity in the 6-month diabetic rats, and the treatment withBMOV completely normalized the basal MBP kinase activity in the "responders"(Figure 30A). A similar trend was noted in the basal S6 peptide kinase activity, albeitnot statistically significant (Figure 30A). The activity of CK-2 did not appear to changesignificantly in the diabetic rats (Figure 30B). However, a remarkable decrease in thebasal S6 protein kinase activity was observed in the 6-month diabetic rats (40%), and167Figure 29. Protein kinase activities in the crude muscle extracts from the 2-monthcontrol (CON) and diabetic (DIA) rats. V indicates vanadyl sulphate treatment. Theskeletal muscles were removed from rats that did not receive insulin injection or ratsthat were injected with insulin 20 min prior to removing the skeletal muscles. Muscleextracts were prepared and assayed for MBP(A), phosvitin(B), S6 peptide(C) as wellas S6 protein(D) phosphotransferase .activity as described in "MATERIALS ANDMETHODS'. Values represent mean±SEM from 4-8 rats. ( *P<0.05 versus CON,#P<0.05 versus CON+Insulin)168^ -Insulin^+Insulin13 70E60E 50o.403020to10a.ca 035-Ec 30 -Eo 25-E 20 -D.Z"5 15 -< 10-Y5-** *S6PK^ CCON^CON+V^DIA^DIA+V169Figure 30. Protein kinase activities in the crude muscle extracts from the 6-monthcontrol (CON) and diabetic (DIA) rats. V indicates the BMOV-treatment; DIA+V(R),the BMOV-treated diabetic rats which responded well to treatment; DIA+V(NR), theMBOV-treated diabetic rats which did.. not respond well to the treatment. Muscleextracts were prepared from the 6-month rats and assayed for MBP and S6 peptide(A), phosvitin (B), and S6 protein (C) phosphotransferase activity. The valuesrepresent mean±SEM of 4-11 rats. ( *P<0.05 versus CON; analyzed by ANOVA).170Phosvitin B10 -I40S RibosomeI* *100-80-60-40-20 -= CON^122) CON+V :•It• DIA^EMI DIA+V(R) El 01A+V(NR)6050403020100MBP^S6 Peptide171again, the treatment with BMOV completely restored the activity in the "responders"(Figure 30C). In the "non-responders" in which plasma glucose concentrationsremained high (>14 mM), the BMOV treatment did not have any effects on either theMBP kinase or S6 kinase activities (Figure 30A-C).Ill. Chromatographic Fractionation and Immunoblotting1. Two-month control and diabetic rats.After fractionation of the pooled extracts from the 2-month rats with a MonoQanion exchange column, several peaks of MBP kinase activity were resolved andnumbered in Figure 31 according to the previous study shown in Figure 11A. PeakII/III and IV contained the majority of the MBP phosphotransferase activity (Figure31A-C). Peaks II and III were activated 1.2- and 1.3-fold, respectively, in the insulin-treated control rats (Figure 31A), while there was little or no activation in the diabeticrats, either treated or not treated with vanadyl sulphate (Figure 31 B-C). Peaks I, V,and VI did not appear to be stimulated by insulin injection. However, as shown inFigure 11 B-G, the effects of insulin on these MBP kinases were reversed within 20min of insulin treatment.The MonoQ fractions were further analyzed by immunoblotting with an anti-MAP kinase antibody erk1-Ill. In the peak I fractions, the antibody recognized a bandof 39 kDa, a doublet of 42-43-kDa, as well as several bands of higher molecular mass(55 kDa, 60 kDa and 88 kDa) (Figure 32A-C). The identities of these bands are notknown, although a clear increase in the 39- and 43-kDa bands in fraction 17 (peak I)was noted in the diabetic rats (Figure 32A-B). In the fractions of peak II and III, twoimmunoreactive bands of 42- and 44-kDa were detected (Figure 32A-C, indicated172Figure 31. MonoQ profile of MBP kinase in the 2-month control (CON)(A), diabetic(DIA)(B), and vanadyl sulphate-treated diabetic rats (DIA+V)(C). The crude muscleextracts from 5-6 rats were pooled and 15 mg of the sample were applied to a MonoQcolumn in buffer A at 0.8 ml/min. The column was eluted with a 15 ml linear gradientof NaCI (0-800 mM) over 60 fractions of of 0.25 ml each. The fractions were assayedfor MBP phosphotransferase activity as described in "MATERIALS AND METHODS'.173II:A CON Iv """O^ Untreated--^— +Insulin25020015010050 i0.B DIA:c DIA+Vm 250a.cE 2000En 150c01000r0 50tQn,2502001501005010^20^30^40 MonoQ fraction number0:^50 60 174Figure 32. Immunoblotting of the MonoQ fraction of the 2-month control (A), diabetic(B) and vanadyl sulphate-treated diabetic rats (C). MonoQ fractions shown in Figure31 were resolved further with SDS-PAGE, transferred to nitrocellulose membranesand probed with anti-MAP kinase antibody erk1-Ill. Arrows indicate p42erk2 andp44erkl .175106K-80K-66K-50K-45K -33K-106K-80K-66K-50K -45K -33K-106K-80K-66K-50K -45K -33K -STD 12 15 17 19 21 23 24 2526 27 28 29 30 32 34 36 38 40 42Mono() fraction number176with arrows). These bands were previously identified as p42erk 2 and p44erk 1 ,respectively (Cobb et al., 1991a; Hei et al., 1993). The intensity of these bands didnot change significantly in the diabetic rats with or without vanadyl sulphate treatment.Another 44-kDa band was also detected in fraction number 30 just after peak IIIfractions, but the intensity of this band did not correlate well with the MBP kinaseactivity. The highest activity was in fraction 32, while the band was most intense infraction 30 where little MBP kinase activity was detected. Faint bands of 39 kDa and42 kDa were also detected in other fractions that did not correlate with the MBPkinase activity. These may represent a non-specific interaction of the antibody withother cellular proteins.MonoQ fractionation also resolved two peaks of S6 kinase activity eluted atabout 100 mM NaCI (peak I) and 400 mM NaCI (peak II), respectively. Both peakswere significantly stimulated by insulin in the control rats (6.1-fold and 7.1-fold forpeak I and peak II, respectively) (Figure 33A). In the diabetic rats, however, therewas a great increase in the basal activity of the peak I (4.1-fold over the control rats),but the magnitude of stimulation by insulin was much less (1.9-fold; Figure 33B).Remarkably, there was no stimulation of the peak II S6 kinase activity by insulin in thediabetic rats (Figure 33B). The treatment with vanadyl sulphate reduced the basalactivity of peak I to near normal value and the degree of stimulation by insulin to 3.6-fold. Furthermore, the stimulation of peak II was also partially recovered with amagnitude of 2.4-fold (Figure 33C).Two antibodies, i.e., S6K-III and rsk-CT, raised against peptide sequences inthe S6 kinases were employed to probe the fractions containing S6 kinase activities.The rsk-CT antibody recognized a 90-kDa band which correlated with the secondmajor peak of S6 phosphotransferase activity (Figure 34A, C, E). This 90-kDa band177Figure 33. MonoQ profile of S6 kinase in the 2-month control (CON)(A), diabetic(DIA)(B), and vanadyl sulphate-treated diabetic rats (DIA+V)(C). The crude muscleextracts from 5-6 rats not treated with insulin (o) or with insulin (•) were pooled and 15mg of the sample were applied to a MonoQ column in buffer A at 0.8 ml/min. Thecolumn was then eluted with a 15 ml linear gradient of NaCI (0-800 mM) over 80fractions of 0.25 ml each. The fractions were assayed for S6 kinase activities with40S ribosome as substrate as described in "MATERIALS AND METHODS'. I and IIindicate the two S6 kinase peaks.178A CON o^ Untreated—41-- +InsulinI.B DIA-II- C DIA+V5.04.03.02.01.00.0EL..el- 5.0.cEo 4.0E;_21._co3.0o-Q2.0Q.r.f)o.cQ.1.0ci) 0.05.04.03.02.01.00.00^10^20^30^40^50^60MonoQ fraction number179Figure 34. Immunoblotting of S6 kinases in the MonoQ fractions of the 2-monthcontrol (A&B), diabetic (C&D), and vanadyl sulphate-treated diabetic rats (E&F).MonoQ fractions from Figure 33 were resolved further with SDS-PAGE, transferred tonitrocellulose membranes and probed with anti-S6 kinase antibodies RSK-CT (A,C,E)and S6K-III (B,D,F). Arrows indicates the 100-kDa putative p90rsk S6 kinase isoformin rat skeletal muscle.180106K-80K-66K-50K-45K -33K-106K-80K-66K-50K45K -33K -106K -80K -66K -50K -45K -33K -rsk-CT Antibody^S6K-ItI Antibody12 15 17 19 23 29 32 34 36 12 15 17 19 23 29 32 34 36MonoO fraction numbermay represent a p9Orsk in the rat skeletal muscle, although we cannot exclude thepossiblity that this kinase may be an isoform of the p70s6k, since the S6K-III antibodyalso recognized this band (Figure 34B, D, F). No changes were noted in the intensityof this band in the diabetic rats, indicating that the amount of this 90-kDa S6 kinasedid not change significantly in diabetes. Also detected with rsk-CT antibody in thepeak II fractions (#32-34) was a. strongly immunoreactive band at 41 kDa, and aweaker immunoreactive band of about 70 kDa. The identities of these bands as wellas other immunoreactive proteins outside of the S6 kinase peak fractions areunknown.2. Six-month control and diabetic rats.The muscle extracts from the 6-month study were pooled and fractionated withan MonoQ anion exchange column, which resolved at least six peaks of the MBPkinase activity (Figure 35A). With the exception of peak V, all the other peaks weresignificantly smaller in the diabetic rats (Figure 35A). The treatment with BMOVincreased the size of all of the peaks to control levels (Figure 35A). When the 40Sribosome was used as a substrate, only the first S6 protein kinase peak was observedafter MonoQ fractionation of the extracts from the older rats (Figure 35B), and itsbasal activity level was over 10-fold higher than that observed with the 2-monthcontrol rats (Figure 33A). The S6 phosphotransferase activity of this peak wasdecreased by over 4-fold in the diabetic rats, and the BMOV treatment maintainedthe peak at control levels (Figure 35B).lmmunoblotting of the the MonoQ fractions from the 6-month rats with erk1-Illantibody permitted the detection of protein bands that were comparable to thosevisualized in the Western blots of the 2-month rats (Figure 36). The intensity of the182Figure 35. MonoQ profile of MBP kinase (A) and S6 protein kinase (B) in the 6-monthcontrol (0), diabetic (•) and BMOV-treated diabetic rats (A). The crude muscleextracts from 8-11 rats were pooled and 15 mg of the sample were applied to aMonoQ column in buffer A at 0.8 ml/min. The column was then eluted with a 15 mllinear gradient of NaCI (0-800 mM) over 60 fractions of 0.25 ml each. The fractionswere assayed for kinase activities with MBP as a substrate as described in"MATERIALS AND METHODS'. MBP and S6 phosphotransferase activity peaks areindicated by Rome numbers (I, II, Ill, IV, V, VI).1831502rts^120• c..2Q. 90o .zCL^60et.eQ300.o6.oCt.Qoc/3• 4.0Co EQ.2.00.008.000Figure 36. Immunoblotting of MAP kinase in the MonoQ fractions of the skeletalmuscle extracts from the 6-month control (A), diabetic (B) and BMOV-treated diabeticrats (C) that responded well to the treatment. MonoQ fractions were further resolvedwith SDS-PAGE and transferred to nitrocellulose membranes, which were probed withthe anti-MAP kinase erkl -Ill antibody. Arrows indicate p42 erk2 and p44erk1.185106K-80K-66K-50K -45K -33K -106K -80( -66K -50K -45K -33K106K -80K -66K -50K -45K -33K -STD 12 15 17 19 21 23 24 25 26 27 28 29 30 32 34 36 38 40 42Mono() fraction numbervarious immunoreactive proteins in the MBP kinase peak I (fraction #17) did notchange significantly in the diabetic rats, but the p42maPk and the p44erk 1 bands(indicated with arrows) in peaks II and III (fraction #24-27) were significantly lessintense from the diabetic rats, indicating a reduction in the amount of MAPK protein inthese animals (Figure 36B). This decrease in p42maPk and p44erk 1 protein mayrepresent a specific loss of these MAP kinase associated with long-term diabetes, andit may account, at least in part, for the decreased MBP phosphotransferase activity inthese rats. However, the intensity of these MAPK were comparable to that of controlrats after the 6-month treatment with BMOV (Figure 36C).With the use of the rsk-CT antibody, the 90-kD S6 kinase band (indicated witharrows) was clearly detected in the fractions corresponding to the peak II S6 kinase(Figure 37A, C, & E). Therefore, the lower basal S6 phosphotransferase activity inthis peak from the 6-month as compared to the 2-month rats did not appear to reflectless enzyme protein. The immunoreactive intensity of the 90-kDa protein in peak IIS6 kinase fractions was slightly lower in the diabetic rats, and was preserved by theBMOV treatment (indicated by arrows in Figure 37). Similar to the results in the 2-month rats, the S6K-III antibody recognized the same 90-kDa band as the rsk-CTantibody, and there was an apparent decrease in the intensity of the band in thediabetic rats (Figure 37B & D). It is also evident from the immunoblots that a 62-kDaband (fraction #29-32) that was detected by the S6K-III antibody completelydisappeared in the diabetic rats, and the treatment with BMOV only caused a partialrecovery of the band (Figure 37D, B & F). The identity of this band is unknown.187Figure 37. Immunoblotting of S6 kinases in the MonoQ fractions of the skeletalmuscle extracts from the 6-month control (A&B), diabetic (C&D) and BMOV-treateddiabetic rats (E&F) that responded well to the treatment. MonoQ fractions werefurther resolved with SDS-PAGE and transferred to nitrocellulose membranes, whichwere probed with anti-S6 kinase antibody RSK-CT (A,C,E) and anti-S6 kinaseantibody S6K-III (B,D,F). Arrows indicate the 100-kDa protein band that wastentatively identified as a p9Orsk.188106K-80K-66K-50K -45K -33K -106K -80K -66K -50K -45K -33K -106K -80K -66K -50K -45K -33K -rsk-CT Antibody^SO•HI Antibody12 15 17 19 23 29 32 34 36 12 15 17 19 23 29 32 34 36Mono() fraction ntunber189r riff i•n •f A P en fall N K e Fr• - • II1. Purification of S6 Kinase.As we sought to purify the first MonoQ peak of S6 kinase activity, we used acolumn with similar properties but larger capacity than MonoQ column, i.e., the fastflow Q Sepharose. Rat skeletal muscle extracts containing about 4-5 grams ofprotein were loaded into the column, and the column profile was identical to that seenwith the MonoQ column, with two peaks of S6 kinase activity (Figure 38A). While thisstep only yielded a modest 3-fold purification of the kinase, it removed more than 93%of the contaminating proteins (Table 4). The loss of S6 kinase activity is simply due tothe removal of the second S6 kinase peak. The peak fractions from the first peakwere pooled and loaded into an S Sepharose column, which resolved a single S6kinase peak that eluted at about 450 mM NaCI (Figure 38B). This step eliminatedanother 94% of the contaminating proteins and afforded a further 8-fold purification ofthe S6 kinase (Table 4). The peak fractions were concentrated and fractionated on agel filtration column (Superdex 200). Two peaks of S6 kinase activity were resolvedwith this column, with the first smaller peak eluted at an apparent Mr of 73,000 andthe second larger peak at 30 kDa (Figure 38C). The second 30-kDa peak was pooledfor further purification. This column produced a 9-fold purification of the kinase byremoving another 90% of the contaminating proteins (Table 4). The pooled fraction ofthe second peak were concentrated and subjected to MonoQ chromatography, whichresolved a S6 kinase peak at the beginning of NaCI gradient (Figure 38E), confirmingthat this is indeed the S6 kinase in the first MonoQ S6 kinase peak. This columneliminated another 92% protein and achieved a less than 2-fold purification (Table 4).The pooled peak fractions were further purified by a MonoS column, which resolved asingle sharp peak of S6 kinase activity (Figure 38E) and recovered more kinase190Figure 38. Chromatographic procedures in the purification of a potentially novel S6kinase. Skeletal muscle extracts prepared from insulin-treated rats were used forsequential chromatography using Q Sepharose column (A), S Sepharose column (B),Superdex 200 column (C), MonoQ column (D), MonoS column (E), and Superose 12column (F), and heparin-agarose column (G). S6 kinase activity (•) in columnfractions were assayed with 40S ribosome subunits and the protein concentration (o)was monitored with a UV-detector in-line with the chromatographic columns. Thewavelength of the UV detector was set at 280 nm. The estimated size of the S6kinases in the gel filtration columns (C&F) are indicated. For details of theexperimental conditions, see "MATERIALS AND METHODS'.1911.00.80.60.40.20.0•60 oficociO063.0O 2.0rta_. 1.00.00005?.12.10O008• E0.060.04 CiO0.020.00E0 08OcoOci6•1050OCOCOO8 1=H•Sap^z96.11co^3s—co000.25C.c0OOaE 100.150.100.050.0012E9tS6304032241680(3.co0.1OO0 3CMono(] fraction num5EHco2.081.51.0 °m0.5 ciOT bl 4. 6 Kin.^ P rific. ion TableSteps Volume(ml)Total Protein(mg)Specific Activity(pmol/min/mQ)Total Activity(pmol/min)Recovery(%)Purification(Fold)Crude 4850 10665 0.25 2664.00 -- 1Q Sepharose 670 737 0.82 601.68 100.00 3.27S Sepharose 125 40 6.11 244.39 40.62 24.46Superdex 200 60 3.825 59.01 225.70 37.51 236.22McnoQ 5 0.27 97.38 26.29 4.37 389.86Mc noS 4 0.108 409.72 44.25 7.35 1640.27Superose 12 10 0.035 1301.14 45.54 7.57 5208.95Heparin Agarose 7 0.0231 783.84 18.11 3.01 3138.03suiting in a more than 4-fold purification (Table 4). Only about 60% of the proteinsapplied were removed by this column. The peak fractions were then applied to aSuperose 12 column which removed about 70% of the loaded protein but recoveredall the activity loaded, yielding a further 3-fold purification. Surprisingly, the S6 kinasepeak eluted at an apparent Mr of 17,000 (Figure 38F), presumably as a result of thekinase binding to the column. At, this stage, the kinase was purified by more than5,000-fold (Table 4). However, when the fractions were assayed with Kemptide, itwas apparent that PKA was co-purified with the S6 kinase, since the PKA inhibitorpeptide, PKI, effectively inhibited the phosphotransferase activity toward the Kemptidebut not S6 protein in the 40S ribosome (data not shown). Thus, an additional heparin-agarose column was used to remove PKA from the S6 kinase. As shown in Figure39A, the S6 phosphotransferase activity was successfully resolved from the Kemptidephosphotransferase activity, which was completely inhibited by PKI. Thus, thecontaminating PKA was removed from the purified S6 kinase preparation. The totalrecovery of the purification was about 4.5% of the starting activity.The S6 kinase peak fractions from the heparin-agarose column were examinedfor purity by SDS-PAGE and silver stain. The S6 kinase peak corresponded well witha predominant 32-kDa protein in the silver-stained gel. Although a 25-kDa band alsoappear to coincide with the S6 kinase activity, silver stain of the Superose 12 columnfractions indicated that only the 32-kDa protein conincided with the kinase activity.Thus, the S6 kinase in the MonoQ peak I appeared to be this 32-kDa protein.2. Properties of the Purified Kinases.The kinetic parameters were determined with respect to ATP and 40Sribosomal subunit using a Lineweaver-Burk double-reciprocal plot. The 32 kDa S6194Figure 39. Resolution of S6 kinase from cAMP-dependent protein kinase by Heparin-agarose column chromatography (A) and silver stain of the S6 kinase peak fractionsfrom heparin-agarose chromatography (B). The heparin-agarose column fractionswere assayed with both 40S ribosome and Kemptide as substrates. The Kemptidekinase peak was apparently resolved from the S6 kinase peak, and the Kemptidekinase activity is completely inhibited by the inclusion of 1 liM PKI. Aliquots of 20 pipeak S6 kinase fractions from the column were resolved on 12% SDS-PAGE andsilver-stained as described in "MATERIALS AND METHODS'. The 32-kDa kinase isindicated by an arrow.1952.52.01.51.00.5:A•• •^• AO A 'IL •0.00^^10 20 30 40 50 60 70Heparin-agarose fraction number350 ;Ili300 co250 .111 (LE,o eL200 -Qct..Q150 Q.100 :0 E50 E0—II— 32 kDa30 32 33 34 35 36 37 38 40Fraction Number196kinase had a Km value of 0.8 ill‘A for 40S subunit and 15 1.t.M for ATP (Figure 40A-B).The specific activity of the kinase, at 0.35 gM 40S subunit and 15 ATP, wasestimated to be 7 nmol/min/mg protein. Phosphorylation of S6 protein by the 32-kDaS6 kinase was linear for at least 30 min at 30°C. (Figure 41).The purified S6 kinase required Mg 2+ for optimal activity, and Mn2+ was notable to substitute for Mg2+ (Figure 42). The optimal concentration for Mg 2+ wasbetween 5-10 mM, and higher Mg2+ concentration caused inhibition of the kinaseactivity.Both anti-rsk-CT and anti-S6K-III antibodies were used to probe the 32-kDa S6kinase. Neither antibody was able to detect this kinase (data not shown), indicatingthat this kinase may not be a member of the two S6 kinase families. The identity ofthis kinase is therefore unknown.3. Substrate Specificity of the Purified Kinases.The 32-kDa S6 kinase was very specific for the 40S ribosome. In the presenceof Mg2+, the kinase phosphorylated S6 protein and histone type-VIIIS, and casein toa minor extent (Figure 43A). In the presence of Mg 2+ as well as Mn2+, however, thekinase failed to phosphorylate S6 protein, but phosphorylated casein very well (Figure43B). The sites that are phosphorylated in S6 protein by the 32-kDa kinase weredetermined to be serine by phosphoamino acid analysis (Figure 44).A number of peptides that were synthesized as substrates for S6 kinase orother protein kinases were also used to test the substrate specificity of the 32-kDapurified S6 kinase. Surprisingly, the raf peptide appears to be by far the best197Figure 40. Determination of kinetic parameters of the purified S6 kinase. The K mand Vmax values for substrates 40S ribosomal subunits (A) and ATP (B) weredetermined by assaying the S6 kinase activity at various concentrations of 40Sribosomal subunits and ATP. The concentration of 40S ribosome was 0.35 p.M whenthe ATP concentration was the variable, and the ATP concentration was 15 pM when40S ribosome was the variable. S6 phosphorylation was plotted against substrateconcentration as well as a Lineweaver-Burk plot (inset). Kinetic parameters werecalculated by a computer program (ENZFITTER).198.11VITI414,1.^4iTilr0.00 0.40 0.80 1.20 LO 2.00 2.40 2.80 3.20 3.60 4.0040S Subunit <Ain)199Time Course of S6 Phosphorylation2500 ^F;"2000 -c0.(71 1500 -g‘-cs. 000(i)0500coCo00^10^20^30Time (min)Figure 41. Time course of S6 phosphorylation by the purified S6 kinase. The purifiedS6 kinase was incubated with 40S ribosome for up to 30 min in the presence of [7-32P]ATP. The reaction was linear within the duration of assay reaction (30 min).200o ^0 20 40 5010 302000Ea.01600}--ul^o-twco"2^400CoCO1200 -Effects of Mg and Mn on S6 KinaseConcentration (mM)Figure 42. Requirement for divalent metal ions by the purified S6 kinase.. S6phosphotransferase activity of the purified S6 kinase was assayed in the presence ofvarious concentrations of Mg 2+ (o) or Mn2+ (•). The assay was allowed to proceedfor 30 min at 30°C in the presence of 50 pAil [y-32 1:1ATP.201Figure 43. Phosphorylation of protein substrates by the purified S6 kinase. S6phosphotransferase activity of the purified 32-kDa kinase was assayed in thepresence of either 10 mM Mg2+ alone (A) or a mixture of 10 mM Mg 2+ and 10 mMMn2+ (B). The substrates used are : 40S ribosome (S), MBP (M), casein (C),phosvitin (P), protamine (PT), histones H1 (H1), HIIA (H2A), HIIIS (H3S), and HVIIIS(H8S). The concentration of the substrates was 1 mg/ml for all the substrates with theexception of 40S ribosome, the concentration of which was 0.35 g.M. The reactionwas performed with 2.5121 of the purified kinase, 5011M [y- 32 P]ATP, and was allowedto proceed for 30 min.202(kDa)97 —66 —45 —31  —22 --H8S H3S H2A H1 PT P C M SS H8S H3S H2A H1 PT P C M203Figure 44. Phosphoamino acid analysis of S6 protein phosphorylated by the purifiedS6 kinase. The S6 protein in the 40S ribosome was phosphorylated by the kinase (5;11) in the presence of 15 IINA ['Y--32P]-ATP. After SDS-PAGE, the phosphorylated S6protein was transferred to a PVDF membrane, which was digested with HCI andsubject to thin layer paper electrophoresis as described in the "MATERIALS ANDMETHODS'. Phosphoserine, phosphothreonine and phosphotyrosine standardswere also run together with the digested samples, and the final positions of thesestandards are indicated.204205substrate, followed by a peptide of S6 protein C-terminal sequence (S6P-2) (S6protein residues 241-259) (Figure 45). The major phosphorylation sites appeared tobe the second of the two adjacent seryl residues, since the peptides that lacked thefirst seryl residue were phosphorylated to an equivalent degree (compare S6P-4 toS6P-9). Threonyl residue could replace seryl residue (S6P-2, S6P-6). On the C-terminal side of the phosphoacceptor site (position 0), acidic and basic residues weresimilarly tolerated in place of leucine at the +1 position (compare S6P-7 and S6P-8 toS6P-4). The comparable degree of Kemptide phosphorylation indicated that residuesbeyond the +2 position were dispensable. On the N-terminal side of thephosphoacceptor site, an arginine at the -3 position was consistent in those peptidethat were phosphorylated by the 32-kDa S6 kinase, and residues further upstream didnot appear to be necessary (compare S6P-5 to S6P-3). These results imply aminimum consensus phosphorylation site recognition sequence for this S6 kinase asArg-X-X-SeriThr. Such a motif is repeated twice in those peptides that werephosphorylated to a higher degree by the S6 kinase (i.e., S6P-1, S6P-2 and Raf1-C2). Additional studies with other peptides will be necessary to establish an optimumconsensus phosphorylation sequence for the 32-kDa kinase, and to confirm whetherthere is a strict requirement for an arginine at -3 position or if it can be replaced byother amino acid residues.4. Modifiers of the Purified KinaseVarious reagents were tested for their effects on the 32-kDa S6 kinase. DTTappear to enhance the activity of the S6 kinase activity (Table 5), suggesting that the -SH groups may be important in maintaining the kinase activity. PKI did not inhibit thekinase activity at a concentration of 2 p.M, indicating that the kinase was not PKA.The kinase was not inhibited by 5 µg/ml heparin, thus excluding the possibility of CK-206Figure 45. Phosphorylation of peptide substrates by the purified S6 kinase. Peptidessynthesized according to the C-terminal sequence of S6 protein (S6P1-10) or used toassay other protein kinases (PKA, PKC and raf kinase) were phosphorylated by 2.5 IIIof the purified 32-kDa S6 kinase in the presence of 50 [y-32P]ATP. Theconcentrations were 1 mg/ml for the S6 peptides and raf peptide, and 200 for theother peptides. The phosphorylation reaction was performed for 10 min at 30°C. Theamino acid phosphorylated by the S6 kinase was tentatively assigned with an"underline" and the * indicates phosphorylation consensus sequence.207Peptide^Sequence S6P-1 AKRRALSaLAASISKSESSQKS6P-2^RRRLSaLRASISKSESSQKS6P-3 RRRLABLRAGGRRS6P-4^RRLAILRAGGRRS6P-5 RLAaLRAGGRRS6P-6^RRLAILRAGGRRS6P-7 RRLAaRAGGRRS6P-8^RRLAaERAGGRRS6P-9 RRLSaLRAGGRRS6P-10^RRLSaLRAo^ Kernptide^LRRAaLGcoRaf1-C2^RORORaTaTPNVHMVSTTGCPKC-1^RFARKGBLRQKNVPKC-2^PLSRTLIVAAKKPKG-1 RKRaRAEPKG-2^RK I aASGPDRPLR32P Incorporation into Peptides (cpm/10 min)0^1000 2000 3000 4000 5000Table 5. Modifiers of S6 KinaseP-32 incorporated into S6Agents Concentration cpm okNone(Control) 3546 100PKI 2 pM 3557 100cAMP 10 pM 3332 94cGMP 10 pM 3751 106DTI' 20 mM 6371 18050 mM 5189 146EGTA 5 mM 1041 2910 mM 544 1525 mM 724 20EDTA 1 mM 4827 1365 mM 1477 4210 mM 561 1620 mM 680 19Heparin 5 pg/mI 3627 10250 pg/mI 1655 47Polylysine 25 pg/mI 2607 7450 pg/mI 2656 75B-Methyl Aspartic Acid 10 pM 2687 76100 pM 3661 1031 mM 3402 96NaCl 5 mM 4113 11610 mM 3554 10050 mM 3774 106100 mM 1508 43NaF 1 mM 3061 8650 mM 3546 100100 mM 3145 89CaCl2 0.1 mM 3540 1000.5 mM 2564 721 mM 2730 775 mM 1261 36CaM 10 pg/mI 4042 11425 pg/mI 3623 102CaM + Ca2+(0.1 mM) 10 pg/mI 2392 6725 pg/mI 4069 115B-Glycerophosphate 20 mM 3966 11250 mM 1573 44100 mM 898 25ZnSO4 0.5 mM 741 211 mM 360 105 mM 353 102092. EGTA and zinc sulphate were highly inhibitory for the purified S6 kinase.Additionally, at higher concentrations, CaCl2 and n-glycerol phosphate were alsoinhibitory to the kinase. The kinase does not appear to be related to PKC orcalmodulin-dependent protein kinase, since Ca 2+ and/or calmodulin did not havesignificant effects on the kinase activity.5. Regulation of the Kinases by PhosphorylationIncubation of the 32-kDa S6 kinase in the presence of [y-32P]-ATP failed tolabel the kinase despite repeated attempts (data not shown), indicating that thekinase does not autophosphorylate. However, the regulation of this S6 kinase activityappear to involve phosphorylation. Treatment of the kinase with PP2A for 60 minresulted in a loss of about 70% of the kinase activity (Figure 46). In the presence of aspecific inhibitor of PP2A, okadaic acid, the kinase activity was preserved during a 60min incubation.To further investigate the regulation of the 32-kDa S6 kinase, the effects of 44-kDa sea star MAP kinase (p44mPk), and CK-2 on the S6 kinase activity wereexamined. Neither of the kinases appeared to have significant effects on thephosphorylation of S6 protein by the 32-kDa kinase (Figure 47A-B). Moreover, theyfailed to phosphorylate the 32-kDa kinase (Figure 48A), while the autophosphorylationof the sea star MAP kinase and CK-2 were affected by the 32-kDa kinase (48A-B).The autophosphorylation of MAP kinase was slightly stimulated whereas that CK-2was moderately inhibited by the 32-kDa kinase.210100▪ 80• 60cts^4020(/)00^20^40^60Time (min)Figure 46. Phosphatase 2A (PP2A) inactivation of the purified S6 kinase. Thepurified kinase (5 p.1) was preincubated in 25 mM Hepes buffer (pH 7.2) and 1 mMDTT in the presence of 15 I.LM [y-32 13]ATP for 30 or 60 min with PP2A (100 U/m1) inthe absence ( • ) or presence (A) of 5 gM okadaic acid (OA). Substrate (40Sribosome) was then added, and the reaction was allowed to proceed for another 30min. For kinase incubated in the absence of OA, OA was added together with 40Sribosome at the end of preincubation to the same final concentration as the reactionsperformed in the presence of OA. OA was dissolved in DMSO which did not have anynoticeable effects on the kinase activity.211Figure 47. Effects of MAP kinase (MAPK) and CK-2 on the activity of the purified S6kinase. The S6 phosphotransferase activity of the purified kinase (5 RI) was assayedin the presence of 5 III MAPK or CK-2 purified from sea star in the presence of 15 11M[y-32P]ATP. After 30 min reaction, the reaction medium was resolved by SDS-PAGEand autoradiogram of the gel was obtained (A). The S6 protein band (indicated by anarrow) was excised from the gel and counted for radioactivity (B).2128000Eo 50004000O 3000.ca.coO 2000.cILco 1000(kDa)9766453122S6 KinaseLI LIMAPK CK-2None^MAPK^CK-2miiJM -S6 Kinase +S6 Kinase213Figure 48. Interaction of MAP kinase (MAPK) and CK-2 with the purified S6 kinase.MAPK and CK-2 (30 111) purified from sea star were incubated together with 10 gl ofthe purified kinase in the presence of 15 tiM ly- 3211ATP for 30 min. The reactionmedium was resolved on SDS-PAGE, and an autoradiogram was then obtained (A).The autophosphorylated MAPK (p44mPk) and the a-subunit and 0-subunit of CK-2bands are indicated by arrows. These bands were excised and counted forradioactivity (B).214mpkCK-2aCK-213S6 Kinase1^I^1^IMAPK CK-2E20000• 1500O• 1000O.co 500MAPK-S6 KinaseCK-2a^CK-281111111 +S6 Kinase215DISCUSSION1. Insulin-stimulated protein kinases in normal rat skeletal muscle.The major insulin-sensitive tissues are skeletal muscle, adipose tissue, andliver. These tissues serve distinct functions and are differentially regulated by insulin.Thus, insulin stimulates glucose utilization to provide energy in muscle, to promotestorage of energy in the form of triglycerides in adipose tissue, and to generateglucose from other substrates or to polymerize glucose in the form glycogen forenergy storage in the liver. The specificity of these responses in different tissues isapparently conferred by the intercellular effectors that are distal to the early signallingpathway of insulin action and is specific to different cell types, since all of the diversebiological response of insulin in different cells are mediated by the same IR.Numerous studies have investigated the effects of insulin in isolated tissues orcultured cells. These studies have the advantage of being able to easily controlexperimental conditions to keep the cells in a simple environment devoid ofinterference from unknown sources. It is also possible to manipulate experimentalconditions with ease and precision when cultured cells are used. Furthermore, it ispossible to get a single uniform cell type in cell culture experiment, which eliminatedthe possible diverse effects that might occur with different types of cells. However,these advantages may become disadvantages, since the cells in vivo are bathed withfluid containing a huge spectrum of different chemicals under physiological conditions.The response of tissues to a specific signal may well be altered by some of the co-existing factors. It is therefore imperative to perform parallel investigations using an invivo model in order to better understand the physiological function of insulin.216In this study, we chose to use intact rat as a convenient animal model systemto study the mechanism of insulin action. Insulin was injected intravenously to elicitbiological responses and rat skeletal muscles were chosen because they are insulin-sensitive and a relatively large quantity of the tissue can be obtained for biochemicalanalysis. In pilot studies, rat liver was also used but the liver was not as responsive toinsulin as skeletal muscle in terms of protein kinase activation. The majority of thestudy was therefore performed with skeletal muscle, except when the tissuedistribution of kinases was examined. It should be noted, however, that otherinvestigators have recently observed stimulation of MAP kinase and S6 kinases byinsulin in liver by intra-portal vein injection with the use of immunoprecipitationtechniques (Tobe et al., 1992).Protein kinase activities in the rat skeletal muscle were assayed by measuringthe ability of muscle extracts to transfer Pi from ATP into protein substrates. Twomethods were used to separate the labelled protein substrate from free ATP. Onemethod involves the use of p81 phosphocellulose papers to bind the labelled proteinand the other utilizes SDS-PAGE to separate the labelled protein from free ATP (seeMethods). The gel method is more tedious and is limited by its capacity, but offers amore accurate determination, because the loss of substrate protein is minimal. Theadvantage of the filter paper assay is that a large number of samples can be assayedsimultaneously with relative ease, but the binding of substrates to the filter paper maynot be complete, especially when peptides are used as substrates, because theadsorption of peptides to phosphocellulose paper is usually incomplete and is stronglydependent on the amino acid composition (Toomik et al., 1992). This results in someuncertainty in the peptide kinase assays. However, when the same peptide is used toassay different samples, this uncertainty should not be a problem.217A relatively large dose of insulin was injected into the rats, producinghypoglycemia in these rats as expected. It is known that a series of compensatoryresponse known as glucose counter-regulation is evoked when glucose levels in thearterial blood are dropped to 3-3.8 mM (Gerich, 1988; Cryer, 1989). These includerelease of glucagon, epinephrine, growth hormone, cortisol, and hepaticautoregulation, all of which act collectively to increase hepatic glucose output byglycogen breakdown initially and by gluconeogenesis later on. These counter-regulatory hormones may significantly affect the kinase activities that we chose tomeasure, and hypoglycemia per se may also have significant impact on cellularmetabolism. Additionally, insulin-mediated signalling could also be affected by bloodglucose concentrations. Thus, it is imperative to have proper controls for thehypoglycemic effects of insulin. To examine whether these factors affected the proteinkinase activities that were determined in insulin-treated rats, a euglycemic clampstudy was performed in which the glucose level was maintained constant after insulininjection. The protein kinases that are activated by insulin were characterized in bothun-clamped and clamped rats.The drop in blood glucose represents one of the biological responses of anintact rat to insulin. As a tyrosine kinase, IR should also elicit phosphorylation ofproteins on tyrosine residues. The skeletal muscle extracts were thus probed withantibodies raised against phosphotyrosine. It was observed that several proteins ofunknown identities were phosphorylated on tyrosine residues under basal conditions,and the level of phosphorylation increased 5 min after insulin injection, peakedbetween 10-15 min, and declined thereafter. By 30 min, the level of tyrosinephosphorylation in skeletal muscle extracts had returned to the basal level (Figure 8).These proteins do not include IR subunits, since they are entirely cytosolic proteins.They may represent potential cytosolic candidates for IRTK, and play important roles218in mediating the effects of insulin, but further experiments such as affinitychromatography using immobilized anti-phosphotyrosine antibody will be required inorder to purify and identify these proteins. These observation do suggest, however,that tyrosine phosphorylation of proteins occurs in intact animal after insulinstimulation, and that this may represent an early response to insulin and serve totrigger subsequent cellular events which produce the biological effects of insulin.Several serine/threonine protein kinases in the crude skeletal muscle extractswere significantly activated after intravenous injection of insulin into an intact rat,including MBP kinase and S6 kinase. While MBP serves as a substrate for otherkinases, the assay medium included PKI and lacked Ca 2+ and lipid, which eliminatedthe kinase activity of PKA and PKC. Thus, the MBP kinase activity as assayed in theskeletal muscle may represent mainly the MAP kinase activity. It would be useful touse the MBP peptide substrate that contains the Thr-97 phosphorylation site which isnow commercially available. Phosphotransferase activity toward peptide substratesfor p74rafl and S6 kinase were also shown to be activated. However, casein kinasewas not activated under our experimental conditions, although it was reported to bemoderately stimulated by insulin in cultured cells (Sommercorn et al., 1987).While MAP kinase activity was originally detected in insulin-treated 3T3-L1cells as a kinase that phosphorylated microtubule-associated protein-2 (Ray andSturgill, 1987), the activation of MAP kinase in these cells was rather modest (1.5- to3-fold). Similarly, the insulin-induced activation of MAP kinase in the crude cytosolicextracts of Swiss 3T3 cells was much lower than that of EGF (Ahn and Krebs, 1990).Insulin alone failed to activate MAP kinase in quiescent Balb/c 3T3 cells (Kawakami etal., 1991). These differences may have resulted from the number of insulin receptorspresent at the cell surface. It was observed that insulin was not capable of stimulating219MAP kinase in rat 1 fibroblasts, but a 5-fold activation of MAP kinase was achieved byinsulin treatment of rat 1 cells that overexpressed normal human insulin receptors(Boulton et al., 1991 a). In the present study, in vivo administration of insulin activatedMAP kinase by about 2-fold in the insulin-sensitive tissue, skeletal muscle. Bycontrast, little activation of MAP kinase was detected when the livers from the insulin-treated rats were analyzed in a similarfashion (data not shown). The stimulation ofskeletal muscle MAP kinase activity by intravenous injection of insulin followed asimilar time course to that of tyrosine phosphorylation. A similar time course ofactivation for MAP kinase was also observed in cultured cells treated with insulin, withmaximal activation at —10-15 min (Ray and Sturgill, 1987). However, this wasconsiderably slower than EGF-induced stimulation, in which case the maximalactivation of MAP kinase occurred within 5 min (Ahn and Krebs, 1990). Some of thephysiologically-relevant substrates for MAP kinases have been shown to be S6kinases, while p74rafl has been shown to be an in vitro substrate for MAP kinase(Sturgill et al., 1990; Anderson et al., 1991; Chung et al., 1991; Lee et al., 1992).Both p42 mapk and p44erkl are capable of activating the p9Orsk family of S6 kinases(Sturgill et al., 1990; Ahn et al., 1990; Chung et al., 1991a), which was demonstratedto be activated in rat skeletal muscle as well (Figure 9B&C). The activation timecourse for protein kinases assayed with S6 peptide and ribosomal 40S subunit assubstrates in rat skeletal muscle were similar to those observed in Swiss 3T3 cells,with S6 peptide kinase having a faster time course of activation than the S6 proteinkinase (Ahn et al., 1990). The antibodies to p9Orsk reacted strongly with an array ofprotein bands with molecular masses of 88-92 kDa (Fig. 10), consistent with thepresence of p9Orsk S6 kinases in rat muscle extracts. Since p9Orsk is known to bephosphorylated at multiple sites in response to insulin, the tightly clustered array ofbands might represent different phosphorylation states of the p90rsk. Unfortunately,we were unable to differentiate the intensity of each individual band during the time220course of insulin injection, which would otherwise suggest a time-dependent patternof phosphorylation following insulin treatment. The antibody to the p70s6k recognizeda protein of -65 kDa, which may be an isoform of the p70s 6k in the muscle. Theintensity of this band seemed to decrease with time of insulin treatment,concomitantly with the appearance of a protein band at -66 kDa. Typically, this wouldsuggest a band shift due to protein phosphorylation and, therefore, activation of thep65s6k in skeletal muscle. However, we were unable to demonstrate activation ofp70s6k with S6 protein as a substrate following MonoQ chromatography.Nevertheless, these observations suggest the existence of both the rat homolog ofXenopus p90rsk (Erikson, 1991) and mitogen-activated p70s6k (Avruch et al., 1991) inrat skeletal muscle. It should be noted that the pattern of the immunoblot is verycomplex, and the existence of other immunoreactive bands make it less certain thatthe 65-kDa band is indeed the p65s 6k.To further characterize the MAP kinases, chromatographic fractionation ofmuscle extracts was performed with a MonoQ column. Many isoforms of MAP kinasehave been identified by biochemical and molecular cloning and immunoblottingtechniques (Cobb et al., 1991 a, Pelech and Sanghera, 1992). However, most studiesonly detected activation of two MAP kinase peaks using MonoQ chromatography,which correspond to p42erk2 and p44erk 1 , respectively (Ahn et al., 1991a, 1991b,Cobb et al., 1991a) (see Table 1). Using MBP as a substrate, we observed at least 6peaks of MBP phosphotransferase activity in MonoQ fractionated rat skeletal muscle,which was significantly more than for cultured Swiss 3T3 cells, where only two distinctpeaks of MBP-phosphotransferase activity were evident (Ahn and Krebs, 1990; Ahnet al., 1991). Apparently, not all of these peaks represent authentic MAP kinases.For instance, the first peak was inactivated rather than activated in response to insulin221injection. There are no reports of such an insulin-inactivated peak in any other modelsystems. The identity of the kinase in this peak is thus unknown.A close examination of the kinetics of inactivation or activation of the individualpeaks indicated that each had a slightly different time course of regulation by insulin.MonoQ peaks II and III were stimulated earlier, while MonoQ peaks IV and Vexhibited a more protracted activation, suggesting that they may represent distinctkinases. The greatest activation of MBP kinase occurred in MonoQ peak II, whichdefined a position of elution (-160 mM NaCI) between the elution of ERK2 (-120-150mM NaCI) and ERK1 (180-200 mM NaCI) as observed in other model system (Ahn etaL, 1990; Boulton et aL, 1991b; Cobb et al, 1991a; ), suggesting that this peak maycontain either of the two MAP kinase isoforms.The identities of the MBP kinase peaks in the MonoQ column profile wereprobed with antibodies raised against peptides patterned after rat brain ERK1.Fractions of peak II MAP kinase activity were shown to contain two strongly-immunoreactive proteins of 42 and 44 kDa with affinity-purified anti-erkl-CT and anti-erk1-Ill antipeptide antibodies. Another antipeptide antibody, anti-erk1-NT alsorecognized a band of 44 kDa in the same fractions, as did an antibody to a MAPkinase ATP-binding domain (anti-mpk-I). These observations indicate that bothp42maPk and p44ark 1 are present in rat skeletal muscle and they co-elute in theMonoQ column under our experimental conditions. The 42-kDa (p42maPk) protein,but not the 44-kDa (p44erk 1) protein, was shown to be tyrosyl phosphorylated uponinsulin stimulation with the use of a monoclonal anti-phosphotyrosine antibody 4G10.A band shift with this protein was also evident, suggesting that p42maPk wasphosphorylated in response to insulin injection. However, since both p42ark 2 andp44erkl co-elute in the same peak, it is not known which isoform is responsible for the222MBP phosphotransferase activity in peak II or whether both isoforms are activated toequivalent degree. Further analysis was performed with the use of different types ofcolumns.When the skeletal muscle extracts were applied to a phenyl-Sepharosecolumn, two peaks of MBP kinase activity (PS-I and PS-II) were resolved.Subsequent fractionation of the two peak fractions on MonoQ column indicated thatPS-I actually eluted at MonoQ MBP kinase peak V position, and PS-II at MonoQ MBPkinase peak III position. However, the MonoQ MBP kinase peak II was missing fromthis column, indicating that some MBP kinase activity did not bind to the phenyl-Sepharose column under our experimental conditions. This is surprising in view ofthe results of Rossomando et al., (1991) who demonstrated that both p42erk2 andp44erkl were bound to phenyl-Sepharose. The difference may have resulted fromslight different experimental conditions between that study and our study.The fractions of MBP kinase peaks from MonoQ column following pheny-Sepharose column were also probed with anti-MAP kinase antibodies. Interestingly,both the MonoQ resolved PS-1 and PS-II fractions contained a 44-kDa band thatreacted with three anti-MAP kinase antibodies. Furthermore, the 44-kDa band inMonoQ peak III underwent dramatic tyrosine phosphorylation in response to insulininjection, while the 44-kDa band in MonoQ peak V had relatively high level of basaltyrosine phosphorylation that was only slightly stimulated after insulin stimulation(Figure 14). Thus, these 44-kDa kinases have immunoreactivity toward anti-MAPkinase antibodies but are distinct from the well-characterized p42erk 2 or p44erk1,suggesting that they may be potentially novel MAP kinases in rat skeletal muscle.This has to be confirmed by additional experiments.223As we failed to detect ERK1 and ERK2 with a phenyl-Sepharose column,another hydrophobic interaction column was employed to resolve the skeletal muscleextracts. With the use of phenyl-Superose chromatography, two peaks of MBPkinase activity were resolved which appeared different from those in phenyl-Sepharose profile (compare Figure 13 with Figure 15). When the fractions wereimmunoblotted with anti-MAP kinase antibodies, a 44-kDa and a 42-kDa doublet wereobserved in peak I and peak II of the phenyl-Superose MBP kinase activity,respectively. These results are comparable to that of Rossomando et al., (1989;1991), and the two MBP kinase peaks represent p44erki and p42maPk, respectively.An apparent band shift for both p42maPk and p44erki was evident on theseimmunoblots, (Fig. 15C&E), implying that both proteins might have beenphosphorylated in response to insulin. However, the tyrosyl phosphorylation occurredmainly on p42maPk, and the tyrosyl phosphorylation of p44erki was barely detectablewith the 4G10 antiphosphotyrosine antibody (Fig. 15G). In accordance with thisobservation, the extent of activation in response to insulin was much greater in thesecond peak (p42 mapk) than that in the first peak (p44erki). Therefore, wesuccessfully resolved the two major isoforms of MAP kinases in rat skeletal muscleand demonstrated that a major role for p42maPk in the overall stimulation of MAPkinase in skeletal muscle by insulin injection.While these experiments indicated the existence and activation of ERK1 andERK2 in rat skeletal muscle, which resided mainly with MonoQ peak II, at least 4additional peaks of MBP kinase activity were shown to be activated by insulin, albeitto different degree. While this might reflect the phosphorylation of MBP by otherprotein kinases that were activated by insulin, or altered binding affinity of the sameisoform of MAP kinase as a consequence of different phosphorylation states, it is alsopossible that novel isoforms of MAP kinase that have not hitherto been described may224be present in rat skeletal muscle. This notion is supported by the detection of 44-kDaproteins in MonoQ peaks III and V with anti-erkl-CT and anti-erk1-11I antibodies.These 44-kDa proteins appeared to be distinct from the 44-kDa protein in peak II,based on their chromatographic and immunoreactive behaviors. Firstly, they eluteddifferently from MonoQ. Secondly, the immunoreactivity of the three 44-kDa proteinsdiffered significantly from each other with the panel of MAP kinase antibodies.Thirdly, from a phenyl-Sepharose column, the 44-kDa protein in MonoQ peak Veluted early as a sharp peak, while the 44-kDa protein in MonoQ peak III eluted laterin a broad peak. Fourthly, the tyrosyl phosphorylation of the 44-kDa protein in MonoQpeak III occurred only after insulin stimulation, like p44erki in MonoQ peak II, while asignificant portion of the 44 kDa protein in MonoQ peak V was alreadyphosphorylated on tyrosyl residues in the absence of insulin stimulation (Fig. 14).These potentially novel 44-kDa MBP kinases also contributed to the overallstimulation of MBP phosphotransferase activity in rat skeletal muscle in response toinsulin injection. In view of the size of these proteins, it is possible that one of the 44-kDa proteins may be the ERK4 MAP kinase that was immunologically detected byBoulton et al. (1991b).Using 40S ribosome subunit as a substrate, we detected two distinct peaks ofS6 phosphotransferase activity in the MonoQ fractions of skeletal muscle extractswhich were significantly activated by insulin. These two peaks may contain the samekinase that also phosphorylated S6 peptide in peak II and peak IV of S6 peptidephosphotransferase activity. The peak I S6 kinase activity appeared to be activatedearlier than the peak II, but the activity of both peaks declined at 30 min after insulininjection. This differs significantly from the experiment with crude extracts as well asfrom EGF-stimulated Swiss 3T3 cells where the activation of S6 kinase reached225maximum at 30 min (Ahn and Krebs 1990). The difference between this study andthe study of Ahn and Krebs in cultured cells could presumably be due to the fact thatin cultured cells, the added growth factors would be present in the sameconcentration throughout the experiment, while the injected insulin would beeliminated from the blood stream quickly, with a half-life of less than 3-5 min. Thus,the decline in the activity of the activated S6 kinase may, at least in part, reflect adecrease in the concentration of insulin during the course of study. The differencemay also be due to the difference in the response of tissue to insulin and theresponse of cultured cells to EGF. Furthermore, the in vivo experimental conditionsdifferred dramatically from the in vitro conditions. However, it is not known why theMonoQ fractionated S6 kinase activation time course differs from that of the crudesamples. It is possible that S6 protein may also be phosphorylated by unrelatedkinases in the crude extracts which is not stable under our chromatographicconditions. It is also possible that a certain amount of the S6 kinase that wasstimulated later after insulin injection did not bind with the column, as discussed in thefollowing.Activation of S6 kinases in rat skeletal muscle in response to insulin has beenpreviously reported by Hecht and Straus (1988), who described the detection of twoactivated peaks of S6 kinase activity with DEAE chromatography of rat skeletalmuscle extracts following intramuscular insulin injection. However, there was nocharacterization nor identification of the S6 kinase peaks in their report. We thereforefurther characterized the two MonoQ peaks of S6 protein kinase by immunoblottingand gel filtration. The detection with anti-rsk-CT antibody of a 100-kDa protein in theMonoQ peak II S6 kinase supported its identification as a member of the p9Orskfamily. This assignment is also supported by the fact that the second MonoQ S6kinase peak also phosphorylated the S6 peptide RRLSSLRA, which is a good226substrate for p90rsk but is poorly phosphorylated by p70s 6k which may require theexistence of 4 basic amino acids in front of the phosphorylatable serine. However,the same 100-kDa protein appeared to cross-react with an antibody developedagainst a peptide sequence in the subdomain Ill of the p70s6k, thus making it difficultto assign the 100-kDa protein as a rsk homolog. The situation is further complicatedby the identification of a human homolog of p70s 6k that had an apparent molecularmass of 90 kDa on SDS-polyacrylamide gels (Grove et al., 1991). Since 8 of the 13residues in a portion of the rat S6K-III peptide are identical to those in acorresponding segment found in the N-terminal catalytic domain of murine p9Orsk, it isfeasible that the anti-S6K-III antibody might recognize a p9Orsk isoform. In the flow-through fractions from MonoQ chromatography, the anti-rsk-CT antibodyimmunoreacted with a protein of 90 kDa which was not detected with the anti-S6K-IIIantibody, indicating that a rsk homolog did not bind to MonoQ. These fractionsexhibited a minor activation of S6 kinase activity by insulin injection (data not shown),suggesting that this 90-kDa rsk homolog may have contributed partially to thestimulated S6 kinase activity in response to insulin injection. While two recent studiespresented weak evidence suggesting the S6 kinase peak eluting later in the NaCIgradient (-300 mM NaCI) may be a 70-kDa p70s6k (Dickens et al., 1992;Sommercorn et al., 1993), we did not detect a protein of similar size in the S6 kinasepeak II fractions that reacted with anti-p70 S6 kinase antibodies. Thus, at least twop9Orsk isoforms with apparent molecular masses of 90,000 and 100,000 may bepresent in rat skeletal muscle, and these are resolvable by anion exchangechromatography. These findings are reminiscent of reports from the Blenis laboratory(Chen and Blenis, 1990; Chung et al., 1991b), in which p9Orsk was detected in theMonoQ flow-through fractions as well as in the NaCI gradient fractions.227The antibody specific for the p70s 6k detected a 65-kDa protein in the crudemuscle extracts that underwent band shift upon insulin stimulation. The sameantibody also detected a protein of about 65 kDa in the MonoQ fractions (data notshown). However, no S6 protein kinase activity was detected in these fractions.While these fractions did contain S6 peptide kinase activity (peak Ill), it remainsunclear why there was no S6 protein kinase activity apparent in these fractions, sinceif this 65-kDa protein is indeed p70s 6k, it would phosphorylate S6 protein rather thanS6 peptide. The possibility exists that the 70-kDa S6 kinase was present in ratskeletal muscle but was not activated by insulin injection.While previous studies suggest that the first S6 kinase peak in the MonoQprofile may be a member of p9Orsk, we failed to detect a protein of about 90 kDa inthe fractions of the first S6 kinase peak. The size of the S6 kinase in MonoQ peak IS6 protein kinase was therefore estimated by gel filtration analysis. At least two S6protein kinases were found to be present, a major species of 40 kDa and a minorspecies of 90 kDa (Fig. 19A). While we cannot completely exclude the possibility thatthe 40-kDa species was a proteolytic product of the 90-kDa species, this is unlikely tobe the case. First of all, we routinely included a spectrum of protease inhibitors in oursample preparations to minimize proteolysis. Furthermore, the 90-kDa and the 40-kDa protein kinases exhibited differential substrate specificities, as evident bycomparison of Fig. 12A with Fig. 12D. The 40-kDa protein kinase preferred S6protein as a substrate to S6 peptide, while the converse was true for the 90-kDakinase. The identity of these two protein kinases remains to be established.However, they do not appear to be closely related to either of the two families of S6kinases that have been so far identified. Since they may be novel S6 kinases, we dida further experiment and purified a 32-kDa S6 kinase from S6 kinase peak I. As228discussed in the last section, this kinase appears to be novel based on severalcriteria.Two protein kinases that phosphorylated S6 protein were also resolved by gelfiltration chromatography of the MonoQ peak II S6 kinase. One was a 120-kDa S6kinase, which could be the rat muscle p9Orsk isoform that was detected by the anti-rsk-CT antibody as a 100-kDa polypeptide, while the identity of the 40-kDa S6 kinaseis unknown. Using S6 peptide as a substrate, at least three proteins were shown tobe present in this peak which could phosphorylate the S6 peptide, indicating that thepeptide was not a specific substrate. It is therefore evident that multiple S6 kinasesmay exist in rat skeletal muscle, including species that may not belong to the p9Orskor p70s6k families.Recent studies have suggested that isoforms of S6 kinases may bedifferentially activated or regulated by diverse stimuli (Mogi and Guroff, 1991; Blenisat al., 1991). The activation of p9Orsk is generally more rapid than for p70s 6k (Sturgilland Wu, 1991), and the stimulation of p70s6k activity by EGF was shown to bebiphasic (Susa et al., 1989), with the first phase of activation occurring between 10-15min and the second phase at 30-60 min. We restricted our experiments to 30 mintreatment with insulin, so we could have missed a second phase activation for thep7orsk. While the finding that rapamycin selectively blocks activation of the p70s 6khas allowed the establishment that 70-kDa S6 kinase is the physiological kinase in avariety of mammalian cells (Chung et al., 1992; Price et al., 1992; Calvo et al., 1992;Kuo et al., 1992), our results suggest that this may not be the case in rat skeletalmuscle where the predominant S6 kinases are probably a p9Orsk and another novelS6 kinase that has not been previously reported.229The activity of p74raf1 , as assayed with a synthetic peptide, appeared toincrease significantly at 10 min after insulin injection and remained activated after 30min (Fig. 2E). This is substantially different from the observation in cultured cells,where a very transient time course of activation for p74raf 1 kinase (5-8 min) wasobserved (Lee et al., 1991). While this difference may arise from the use of differentmodel systems, it could also reflect the non-specificity of the raf-1 peptide that weused as a substrate for p74rafl, and the possibility that other kinases which becomeactivated later by insulin may also utilize this peptide as a substrate. For instance,this peptide may have served as a substrate for S6 kinase as well. In fact, a declinein the raf-1 peptide phosphotransferase activity was noted at —15-20 min, which wasfollowed by another apparent increase in the raf-1 peptide kinase activity. Thisbiphasic time course may represent distinct kinases that are activated with differentkinetics. After MonoQ chromatography, an apparent bandshift appeared in fractionswhich exhibited increased phosphotransferase activity toward raf-1 peptide, implyingthat p74raf1 may have been phosphorylated as a consequence of insulin treatment,which was in accord with previous reports that p74raf 1 was phosphorylated onserine/threonine residues in H35 cells after insulin stimulation (Blackshear et al.,1990; Lee et aL, 1991). While the substrates and function of p74rafl remain to bedefined, it is likely that p74rafl may play a significant role in realization of the fullmitogenic response to insulin. The significance of this experiment has to beconsidered with caution, in view of the fact that the peptide we used was relativelynon-specific, and that the immunoblots with anti-raf antibodies in both crude muscleextracts as well as the MonoQ fractions were rather complex. Other possibleexplanations cannot be excluded.Our observation that the activities of MAP kinases and S6 kinases wereactivated in rat skeletal muscle in response to in vivo administration of insulin230represents a significant finding, especially in view of the fact that several groupsincluding that of Cohen's have failed to detect activation of these kinases in anotherintact animal model, i.e., the injection of insulin into rabbit, with skeletal muscle as atarget tissue. While this may be due to difference in the species of animals used, itmay well be that the experimental conditions such as the age of the animals and thecomponents of buffers used differed in our study. In this sense, our study pioneeredthe establishment of a good physiologically relevant model system to study signaltransdution by insulin.One of the major concerns in experiments with intact animals is the possibleinteraction of many unknown factors. For instance, hypoglycemia-induced glucosecounter-regulation may have profound impact on the kinase cascade we intend tostudy. We therefore did control experiments administrating insulin under aeuglycemic clamp technique. When samples from these rats were analyzed byMonoQ chromatography, it was observed that the general profile of kinase activityremained largely the same. However, it should be pointed out that certain alterationshave been noted as well. Typically, the MBP kinase peak III appeared smaller afterthe glucose clamp, while peak IV and VI MBP kinase was further stimulated after theglucose clamp (Compare Figure 11 with Figure 22). The S6 peptide kinase profile didnot change significantly, while the S6 protein kinase profile differed from theunclamped rats in that the peak I S6 kinase was much smaller than in the unclampedrats, although significant activation of this kinase (>3-fold) still existed. Apparently,the compensatory mechanisms during hypoglycemia did affect some of the kinasesthat were activated by insulin injection. Since growth hormone also stimulatestyrosine phosphorylation in cells (Anderson et al., 1992; Moller et al., 1992; Winstonand Bertics, 1992), it is possible that the kinase activity in unclamped rats might haveresulted from a combined effects of insulin and growth hormone released by231hypoglycemia, while the release of catecholamine during hypoglycemic responsesmay cause additional effects on the kinase activities. In addition to the counter-regulatory hormones that were released during hypoglycemic reactions, an extremelylow concentration of glucose in the blood would certainly affect the energymetabolism of the cells, further complicating the experiment. Finally, it is most likelythat glucose concentration have direct impact on the signaling of insulin action.Taken together, we acknowledge that the results that derived from using the un-clamped rats may have been obscured to some extent by these factors that were notcontrolled during the course of the study. However, the experiments with rats thatunderwent euglycemic clamp did support our conclusion that the majority of theeffects that we examined after insulin injection may result from the actions of insulin.We explored the possible activation of CK-2 by insulin stimulation in rat skeletalmuscle. Perhaps due to The very low level of CK-2 activity in skeletal muscle (Krebset al., 1988), we could not demonstrate a consistent activation of this kinase afterinsulin injection. While a role for CK-2 in the insulin signaling mechanisms in ratskeletal muscle could not be excluded by our results, it seems likely that it is minor.However, it is likely that in tissues where a high level of CK-2 exists, insulin might elicitactivation of CK-2.The tissue distribution of CK-2 was described in a previous reports (Nakajo etal., 1983). However, the tissue distribution of MAP kinase and S6 kinase activitieshas not been reported. Thus, the distribution of these kinases was examined in thisstudy. In agreement with the study of Nakajo et al., the highest level of CK-2 activitywas observed in the spleen which also exhibited the highest level of activities of MBPkinase and S6 kinase. This is surprising in view of the fact that spleen is notconsidered to be an insulin-sensitive organ. The high level of activity of these232mitogen-activated kinases in spleen may reflect a strong cell growth-related activity inspleen where lymphocytes are produced and old cells are eliminated. The lowerkinase activity in skeletal muscles may have been due to the fact that most of thecellular proteins in the muscle are contractile proteins such as myosin and actin,which would invariably lead to a lower specific activity of the kinases which wasnormalized against protein concentrations in the samples.When the spleen extracts from control and insulin-treated rats were analyzedby MonoQ chromatography, a CK-2 peak was resolved and importantly, the peak wasstimulated 5.5-fold by insulin injection. Thus, insulin is indeed able to stimulate theactivity of CK-2 in a tissue that contains high level of CK-2. The MonoQ fraction ofspleen extracts from insulin-treated rats also showed enhanced MBPphosphotransferase activity. It was unexpected, however, that only one MBP kinasepeak was noted in the MonoQ column profile of spleen extracts. This peak eluted at aposition different from the p42 erk2 or the p44erk 1 , but resembled the potentially novel44-kDa MAP kinase which was also detected in peak V of the MonoQ profile ofmuscle extracts (see Figure 11). This observation strongly suggests that although theMBP kinase activity was more than 30-fold higher in the spleen than in the skeletalmuscle, the major species of MAP kinase in spleen may differ from those in themuscle. A kinase different from ERK1 and ERK2 is responsible for the extremely highlevel of MBP kinase activity in the spleen. The significance of this finding is not clearat this time, but it may reflect functional differences between different tissues. Itshould be noted, however, that the dose of insulin that was injected was very high,which could lead to activation of IGF-1 receptors. Since spleen is not considered an"insulin target" tissue in terms of its role in the regulation of substrate metabolism,activation of IGF-1 receptors in the hemopoetic cells or lymphocytes in the spleenmay contribute significantly to the activation of MAP kinase or CK-2 activity.233Unfortunately, the activity of S6 kinase in the MonoQ fractions of spleen extracts wasnot determined.Activation of p74raf 1 , MAP kinase, and S6 kinase have been observed incultured cells with in vitro experiments as well as in intact cells (reviewed in Pelech etaL, 1990; Sturgill and Wu, 1991; ). We have demonstrated for the first time thatisoforms of all these kinases appear to be activated simultaneously in an intact animalfollowing insulin injection. Thus, this is an attractive model system for theinvestigation of the early events in insulin signal transduction. It may facilitate thepurification and characterization of novel protein kinases that participate in the insulinresponse, since large amounts of tissues can be obtained without the need forlaborious and expensive cell culture. Furthermore, the consequences ofmanipulations of the intact animal, such as streptozotocin-induced diabetes, can beassessed in this regard. This will be discussed in the following sections.2. Insulin-stimulated protein kinases in diabetic rats and the effects of vanadiumTo investigate whether the insulin signalling pathway is altered in diabetes, aSTZ-induced diabetic rat model was used to examine MAP kinase and S6 kinaseactivities in these rats. Vanadium treatment was also examined in this experiment todetermine if treatment with an insulin-mimetic agent would affect any changes thatmight occur in diabetes.While vanadate has been demonstrated to have insulin-like effects in diabeticrats, it is poorly absorbed and has some toxicity. Vanadyl sulphate and BMOV wereused in the treatment, since vanadyl sulphate, although also poorly absorbed, wassuggested to have less toxic effects and was better tolerated by both control and234STZ-diabetic rats (Ramanadham et aL, 1989). BMOV was synthesized as an organicvanadium compound that has significantly less toxic side effects and a 50% higherpotency than vanadyl sulphate (McNeill et al., 1992). The rapid onset of thiscompound also suggests that BMOV was better absorbed from the gastrointestinaltract.The STZ-induced diabetic rats, as used in this study, represent a model ofpoorly-controlled Type I diabetes, where the circulating insulin level is markedlyreduced by the free-radical mediated destruction of pancreatic p-cells. Exogenousinsulin is required to maintain normal metabolism but not required for the survival ofthese rats. A number of diabetes-related complication often occur in the STZ-diabeticrat, such as hyperlipidemia, cardiac disorders, cataracts, retinopathy and nephropathyand neuropathy (Dulin et al., 1983). However, insulin resistance also becomesapparent with prolonged duration of the disease, presumably secondary to persistenthyperglycemia (Burant et al., 1986; Rossetti et al., 1987; Nishimura et al., 1989;Blondel et al., 1989; Giorgino et al., 1992). A severe insulin resistance wasreproduced in the 2-month STZ-induced chronic diabetic rats in the present study.Considering the large dose of insulin that was injected into the diabetic rats, it isremarkable that the glucose levels in these rats remained unchanged 20 min after theinjection. The molecular mechanism for insulin resistance is not completelyunderstood, but any changes along the signalling pathway of insulin action couldaffect the biological effects of insulin, giving rise to insulin resistance (Olefsky, 1991).Various genetic abnormalities have been shown to be responsible for thedevelopment of insulin resistance in human NIDDM (see introduction), while theinsulin resistance in IDDM tends to be overlooked since it is secondary. It should benoted that with the prevalent use of insulin in the treatment of IDDM, insulinresistance may develop via down-regulation of IR with long-term use of insulin. In the235STZ-diabetic rats, however, changes in the number of IRs can not explain insulinresistance, since the number of insulin receptor is increased in these rats as aconsequence of receptor up-regulation (Burant et al., 1986; Okamoto et al., 1986;Giorgino et aL, 1992). Therefore, the insulin resistance associated with the STZ-diabetic rats could only be explained by defects in the post-receptor mechanisms.Decreases in the IR tyrosine kinase activity, the autophosphorylation of the 13-subunit of the insulin receptor as well as the phosphorylation of IR substrate 1 (IRS-1)have been previously reported in the STZ-diabetic rats in response to insulinstimulation (Kadowaki et aL, 1984; Burant et al., 1986, Block et al., 1991; Giorgino etaL, 1992). A natural consequence of such alterations would be a defect in the latersteps of signalling mechanisms for insulin. In this study, we used the long term STZ-diabetic rats to determine the activities of the MAP kinases and S6 kinases, which arecritical components in the insulin-stimulated protein kinase cascade. The resultsindicate that both MAP kinases and S6 kinases are significantly affected in thediabetic rats. The changes in these kinases were apparently time-dependent. In the1-month diabetic and 2-month diabetic rats, the basal level of S6 kinase activity waselevated while the MAP kinase activity was not changed. In the 6-month diabetic rats,however, the basal MAP kinase activity was significantly reduced (30%), as was theactivity of the S6 kinases (40%). This suggest that the S6 kinases may be moresensitive to the conditions of diabetes, and the increase in the activity of S6 kinase inthe 2-month diabetic rats may have resulted from a compensatory mechanism thatattempted to overcome the lack of insulin stimulation. With prolonged diabetes,however, the compensatory mechanism may have failed, resulting in a markeddecrease in the activity of the S6 kinases in the 6-month diabetic rats.236Stimulation of the MAP kinase activities was not detected in the crude muscleextracts from the 1-month and 2-month control as well as the diabetic rats. This maybe due to the fact that the maximal stimulation of MAP kinases by insulin occurs at 15min in the intact animal model as shown in the previous section, while the muscleextracts used in this experiment were prepared 20 min after insulin injection, at whichtime the kinase activity may have declined. It is also possible that some non-specificphosphorylation of MBP could have masked a moderate activation of the MAPkinases in the crude samples. This is supported by the fact that fractionation of thecrude muscle extracts of the 2-month rats with an MonoQ column revealed amoderate activation of the MAP kinases in the control rats. With the 2-month diabeticsample, however, no activation of MBP kinases was detected even after thefractionation, indicating an apparent change in the stimulation of MAP kinases in thediabetic rats. Considering the importance of MAP kinases in the signal transductionpathway for growth factors including insulin (Cobb et al., 1991; Thomas, 1992, Pelechand Sanghera, 1992), the impairment in the activation of MAP kinase would certainlyresult in a defective response of cells to insulin stimulation. The severe insulinresistance in the 2-month diabetic rats could at least be partially accounted for by theimpaired MAP kinase activation. The changes in the activation of MAP kinases couldoccur as a result of diabetes-induced alterations in the gene expression of thekinases, the abnormalities in the enzyme properties, or the changes in the inputsignals that activate these kinases. Since our immunoblotting data indicate noapparent change in the amount of MAP kinases in the 2-month diabetic rats,components upstream of the MAP kinases are likely responsible for the defectiveMAP kinase activation.It should be noted that in these rats where MAP kinase activation was notapparent, S6 kinase activity was significantly stimulated under the same experimental237conditions. This may imply that pathways other than MAP kinase could be involved inactivating the S6 kinases.As shown in the previous sections, two S6 kinase peaks were identified in theMonoQ profile of insulin-stimulated rat skeletal muscles. It was found that the basallevel of the peak I S6 kinase activity was greatly enhanced in the 2-month diabeticrats, which is consistent with the results of the crude extracts, although the magnitudeof activation by insulin was markedly diminished, suggesting that this kinase washighly regulated in diabetes, and that the upstream activating pathway for this kinasewas defective in diabetes. More significant, the activation of the 100-kDa S6 kinase inthe second peak was completely inhibited or obliterated in the 2-month diabetic rats.As the immunoblotting data indicated that the amount of 100-kDa S6 kinase proteindid not change in the 2-month diabetic rats, the lack of insulin stimulation of thiskinase might be explained by defects further upstream in the activating pathway thatinvolves MAP kinases. It is also possible that abnormalities may have occurred in thestructure of the S6 kinase that prevent its activation in response to insulin. This canonly be assessed by studying the kinase purified from both control and diabetic rats.It should be noted, however, that the decrease in the activity of MAP kinase and S6kinase in diabetic rats appeared to be specific changes related to diabetes rather thana reflection of a general decrease in the cellular proteins. This is supported by twopieces of evidence. First, in immunoblots of MonoQ fractions from control anddiabetic rats, the intensity of some non-specific reacting protein bands was notchanged with diabetes, while the intensity of MAP kinase bands was reduced.Secondly, when cellular proteins from both control and diabetic rats were resolved bySDS-PAGE and silver stained, there was no apparent change in the pattern ofproteins in the gel (data not shown).238The time-dependency in the changes of the MAP kinases and the S6 kinasesduring diabetes becomes evident when the MonoQ profile from the 2-month and 6-month rats are compared. Under basal conditions, all the MBP kinase activity peakswere lower in the 6-month rats than the 2-month rats, while the S6 kinase peak I wasmuch higher in the 6-month rats than that of the 2-month rats. Significantly, all theMBP kinase peaks, except for peak V, were decreased by more than 3-fold in the 6-month diabetic rats. Immunoblotting of the MonoQ fractions indicate that the amountof the MAP kinase proteins (both p42erk2 and p44erkl) did not change in the 2-monthdiabetic rats, but was significantly decreased in the 6-month diabetic rats. This isconsistent with the observation that the basal MBP kinase activities did not change inthe 2-month diabetic rats, but markedly decreased in the 6-month diabetic rats. Thebasal peak I S6 kinase activity was markedly higher in the 2-month diabetic rats,whereas a significant decrease in the peak I S6 kinase activity was noted in the 6-month diabetic rats. Likewise, the amount of the 100-kDa S6 kinase in the peak II S6kinase activity was not changed in the 2-month diabetic rats, while a decrease in thisprotein was noted in the 6-month diabetic rats. These observations clearly indicatethat with a longer duration of diabetes, there were more pronounced alterations in theprotein kinase cascade which mediates the actions of insulin.The mechanism for the decrease in MAP kinase and S6 kinase proteins is notapparent from the current study. The expression of these kinases in tissues and itsregulation has not been investigated, and the genomic structure of these kinases areunknown. However, as MAP kinase and S6 kinase are growth-related, it isconceivable that growth factors may have significant effects on the regulation of theexpression of these proteins. It is possible that certain response elements exist in thepromoter of these kinases and mediate the transcriptional regulation of these kinases.Alternatively, translational regulation and post-translational modification of these239kinases could also be affected in diabetes. For instance, S6 kinase is implicated inthe enhanced translation of proteins by increase mRNA recruitment (see"INTRODUCTION'). A decrease in this kinase could also affect the translation ofitself. The genomic DNA structure of these kinases is required in order to elucidate ofthe regulatory mechanism of their expression.The insulin-like effects of vanadium compounds have been well documented inin vitro experiments and in vivo models of diabetes. The mechanism of action forvanadium, however, is not completely understood. It is unlikely that the insulin-likeeffects of vanadium would result from a specific action of vanadium on the IR. Sincevanadium compounds can be transported into cells, it is possible that vanadium couldact directly on the intracellular components that mediate the effects of insulin.Previous studies have demonstrated inhibition of tyrosine phosphatase (Swarup et al.,1982; Klarlund, 1985), stimulation of protein tyrosine phosphorylation in intact cells(Tamura et aL, 1984; Gentleman et al., 1987; Yang et al., 1989; Lerea et al., 1989;Heffetz et al., 1990; Grinstein et al., 1990), and a direct stimulation of IR tyrosinekinase activity (Kadota et al, 1987a; Kadota et al., 1987b; Gherzi et al., 1988; Fantuset aL, 1989) by vanadium compounds. These observations suggest a role for IR inthe insulin-like effects of vanadium. However, this view has been challenged by otherreports which showed that the stimulation of glucose transport in rat adipocytes wasnot affected by a 95% loss of cell surface insulin receptors (Green, 1986); that theanti-lipolytic effects of vanadate in rat adipocytes were not correlated with theactivation of tyrosine phosphorylation (Mooney et al., 1989); and that the vanadatestimulated glucogenesis in rat diaphragm without activating insulin receptor kinase(Strout et aL, 1989). Most recently, vanadium was shown to activate a cytosolictyrosine protein kinase that was not responsive to insulin (Shisheva and Shechter1993). Also observed are the effects of vanadate on the release of insulin (Zhang et240al., 1991), insulin binding to IR (Eriksson et al., 1992), and the processing of IR in thecell (Marshall et al., 1987; Torossian et al., 1988). Thus, the molecular mechanismsfor the insulin-like effects of vanadium remain uncertain, and multiple targets distal toinsulin receptors are likely involved (Green 1986; Blondel et al., 1990, also see areview by Shechter, 1990).In the present study, the possibility that vanadium may act directly on theprotein kinases that lie in the insulin signalling pathway was explored. The resultsindicate that vanadyl sulphate or BMOV were able to partially or completely restorethe abnormal activities of MAP kinases and S6 kinases in the diabetic rats skeletalmuscle, and these changes correlated well with the euglycemic effects of vanadiumcompounds. In the 2-month diabetic rats, vanadyl sulphate treatment not onlynormalized the glucose levels, but also the insulin sensitivity, which agrees well withprevious reports (Rossetti et al., 1987; Blonde) et al., 1989; Ramanadham et al.,1990). Moreover, the impaired activation of S6 kinase activity in the diabetic ratskeletal muscle was largely corrected by vanadyl sulphate treatment. In the 6-monthdiabetic rats, the decrease in MAP kinase activity as well as S6 kinase activity wasincreased to control values after BMOV treatment. However, in the "non-responders"where the blood glucose levels remained high, the MAP kinase and S6 kinaseactivities were not changed by the treatment, suggesting a correlation between thekinase activities and the blood glucose levels. While a detailed relationship betweenprotein kinase/protein phosphorylation and glucose homeostasis is still lacking, recentstudies suggest that MAP kinases but not S6 kinases may play an important role inthe stimulation of glucose transport activity by insulin (Fingar et al., 1993; Inoue et al.,1993). As the correction of glucose levels alone could normalize the tissue sensitivityto insulin (Rossetti et al., 1987), the present study cannot rule out the possibility thatthe restoration of kinase activities actually is a result of the normalized glucose levels.241We consider this unlikely, since vanadate has been shown to have much moreextensive biological effects on cells and is able to produce many of the insulin-likeeffects that are not directly attributed to correction of hyperglycemia (Rossetti andLaughlin, 1989; Blonde) et al., 1990; Cordera et al., 1990). In any case, this can onlybe clarified by further experiments using such agents as phlorizin which lowers theblood glucose levels through the inhibition of glucose re-uptake from kidney.The existence of "non-responders" has been observed in vanadyl sulphatetreated rats as well (Cam et al., 1993). A very high degree of variability was noted inthe response of diabetic rats to vanadyl sulphate or BMOV treatment, which mayreflect the individual difference of rats to the treatment. Indeed, all of the diabetic ratsresponded to vanadium treatment when the dose of vanadium compounds was raised(Cam et al., personal communication). Unfortunately, toxic effects of the compoundsbecome increasingly frequent with higher doses, suggesting that vanadiumcompounds have a relative narrow therapeutic index. It is important to note that theblood glucose level was used as the sole parameter here to distinguish "responders"from "non-responders". This does not necessarily indicate a failure of vanadium tohave any effects on the biological functions in the treated rats. It has been shown thatvanadyl sulphate was able to normalize the depressed activity of GP in diabetic ratsthat were still hyperglycemic (Liu and McNeill, manuscript in preparation), suggestingthat the euglycemic effects of vanadium may involve a number of mechanisms, all ofwhich have to be affected for vanadium to successfully lower blood glucose levels.It is interesting to note that the effects of vanadium on the S6 kinases appearto depend on the basal kinase activities. Treatment with vanadyl sulphate returnedthe greatly enhanced basal MonoQ peak I S6 kinase activity in the 2-month diabeticrats to near normal values, whereas BMOV normalized the basal MonoQ peak I S6242kinase activity which was markedly decreased in the 6-month diabetic rats. Whileintraperitoneal injection of vanadate was shown to stimulate the p70s 6k activity(Kozma et al., 1989), the MAP kinase activity (Tobe et al., 1992), and a S6 peptideactivity (Tobe et al., 1992) in rat liver, it does not appear that these direct stimulatoryeffects of vanadium compounds are responsible for the insulin-like effects of orally-administered vanadium compounds. Firstly, the concentration of vanadium in thetissue may not be high enough to stimulate the kinase activities after oraladministration of vanadium. Secondly, the acute administration of vanadyl sulphateor BMOV by the intraperitoneal route failed to stimulate the activities of the MAPkinases or S6 kinases in rat skeletal muscle (data not shown), and thirdly, we did notdetect the activity of p70s6k in rat skeletal muscle , the isoform of the S6 kinase thatwas stimulated by vanadate in rat liver. This could result from the actions ofvanadium compounds on the expression of the MAP kinases and the S6 kinases, orthrough its actions on the activities of these kinases. The immunoblotting datademonstrated that the amount of protein was directly altered by the treatment withvanadium and the changes were relatively specific for these kinases, suggesting thatvanadium affected the gene expression or translation of the MAP kinases or S6kinases in the diabetic rats. This is not surprising in view of the studies which showedthat vanadium increased the gene expression of GLUT-1 in NIH3T3 cells (Mountjoyand Flier, 1990) and enhanced the expression of GLUT-4 at a pre-translational levelby increasing the amount of GLUT-4 mRNA in STZ-diabetic rats (Strout et al., 1990).To summarize, alterations in the basal activities and activation by insulin of theMAP kinases and S6 kinases have been demonstrated in long-term diabetic rats.The insulin-resistance observed in these rats appeared to correlate with the changesin these kinases, and furthermore, treatment with vanadium compounds returned allthe changes in these kinases to near normal or completely normal values. The fact243the insulin partially activated MAP kinase and S6 kinase in the 2-month diabetic ratswithout reducing their glucose levels may suggest that the degree of kinase activationis not sufficient to induce stimulation of glucose disposal, or that glucose metabolismmay lie in a separate pathway from MAP kinase-S6 kinase cascade. The molecularbasis for these changes awaits further studies. The results of this study providefurther support for the use of vanadium compounds in the treatment of diabetes. Inview of the various adverse complications associated with insulin therapy (Home etal., 1989), it would be of significance to have substitutes for insulin that are orallyactive. Vanadium compounds or compounds derived from research into themechanism of vanadium may prove to be valuable agents in the management ofdiabetes.3. Purification of a novel S6 kinase from rat skeletal muscle. While two major families of S6 kinases have been identified and characterizedfrom Xenopus oocytes as well as from mammalian cells and tissues, it is not known ifthese are the only S6 kinases that exist. Various S6 kinases have been identified andpurified from different sources, including a 65-kDa S6 kinase from developing chickenembryo (Blenis et al., 1987), a 67-kDa S6 kinase from bovine liver (Tabarini et al.,1987), and a 45-kDa S6 kinase from NGF-stimulated PC12 cells (Matsu and Guroff,1987). Insulin-stimulated S6 kinase activity in rat skeletal muscle has been previouslyreported but not characterized (Hecht and Straus, 1988). As mentioned previously,two S6 kinase peaks were identified in the MonoQ profile of insulin-stimulated rats.The second peak was tentatively identified as a member of the p90rsk, while the S6kinase in the first peak has been purified to near homogeneity. This kinase does notappear to be a member of either p9Orsk or p70s6k, based on physical andimmunological properties. It is unlikely to be a catalytically active fragment of p9Orsk244or p70s6k, since these have not been reported to undergo proteolytic activation.Furthermore, S6K-III antibody, which is based on the sequences within the catalyticdomain of p70s6k, did not cross-react with the 32-kDa kinase, although this antibodywas shown to recognize p9Orsk. This kinase is also distinct from PKC, since it did notphosphorylate histone H1 and MBP which are preferred substrates for PKC (Pelech etal., 1991). Interestingly, like p9Orsk and p7o.s6k, this kinase is regulated byphosphorylation on serine/threonine, which indicates that it may also be a componentin the protein kinase cascade that is activated in response to insulin.The 32-kDa kinase bears some resemblance to the NGF-stimulated S6 kinasethat was partially purified from rat PC12 cells (Matsuda and Guroff, 1987). Like the• 32-kDa kinase, the NGF-stimulated kinase was insensitive to cyclic nucleotides andcalmodulin; it was inhibited by EGTA and 13-glycerolphosphate; it phosphorylated S6exclusively on serine; it poorly phosphorylated casein, histones, and MBP; and it wasinactivated by phosphatase pretreatment. However, the NGF-stimulated kinase had amolecular weight of 45,000 on gel filtration and was much more sensitive to NaF(Ki=30 mM) than the 32 kDa kinase. The NGF-stimulated kinase was largelyinsensitive to activation by EGF in PC12 cells, although EGF did stimulate anotherunidentified S6 kinase in these cells (Mutoh et al., 1988).An insulin-activated S6 kinase of about 35,000 was reported by Cobb andRosen (1983). This kinase eluted at low NaCI concentration from a DEAE cellulosecolumn, and like the NGF-stimulated kinase, was also sensitive to NaF (Ki=25 mM).It was referred to as a CK-1-like enzyme due to its copurification with caseinphosphotransferase activity. The 32-kDa kinase that was purified in this studydisplayed properties that are similar to CK-1. For instance, CK-1 elutes from DEAEcellulose at relatively low ionic strength (70-130 mM NaCI); it is a monomer of M r24530,000 to 37,000,; it requires Mg 2+ for activity and has a Km value for ATP in therange of 7-25 gM; and it phosphorylated proteins on serine residues (Hathaway et al.,1983; Tuason and Traugh, 1991). However, the 32-kDa kinase appears to be distinctfrom CK-1 based on other criteria. Firstly, it poorly phosphorylated casein in thepresence of Mg 2+, although casein was substantially phosphorylated by this kinase inthe presence of Mn2+. Secondly, the 32-kDa kinase was not affected by 50 mM NaF,while 90% of CK-1 activity was inhibited by 20 mM NaF (Simkowski and Tao, 1980).Thirdly, CK-1 can autophosphorylate (Tao et al., 1980; Dahmus 1981; Nakajo et al.,1993), whereas the 32-kDa kinase did not. Fourthly, there are no reports of CK-1inactivation following treatment with a phosphatase, as was observed for the S6kinase. Fourthly, the consensus phosphorylation site sequence for CK-1 (i.e., thepresence of acidic or phosphorylated residues at the -3 and -4 positions relative to thephosphoacceptor site; Flotow and Roach 1991) was opposite to what has beendeduced for the S6 kinase (i.e., the presence of a basic residue at the -3 positionrelative to the phosphoacceptor site). The question remains, however, as to why the32-kDa kinase requires Mg 2+ to phosphorylate ribosomal S6 protein whereas Mn 2+ isrequired to phosphorylate casein.The assignment of Arg-X-X-Ser/Thr as the minimum consensusphosphorylation site sequence for recognition by the 32-kDa S6 kinase requiresfurther confirmation by experiments with additional peptides. This consensussequence, however, is similar to that for p90rsk (Erikson and Mailer 1988). The 32-kDa S6 kinase consensus motif is repeated three times in the C-terminus of S6 (i.e.,BRLS235 , RLSS236 , BAST241 ). The precise sites of S6 phosphorylation by the 31-kDa S6 kinase remain to be established, although Ser-236 is likely to be thepredominant site based on the phosphorylation rates with S6 peptide analogs. Itshould be noted that the level of phosphorylation appeared to be fairly low for all the246peptides, although many of the peptides were synthesized after the C-terminalsequence of the S6 protein. However, the kinase did phosphorylate S6 proteinreadily when 40S ribosome was used as a substrate, suggesting that specificconformation of the protein may be required for the phosphorylation reaction, whilethe peptide cannot satisfy the conformational requirement.The specific activity of the purified 32-kDa kinase (1.2 nmol/min/mg) may be anunderestimate. The kinase was assayed at 0.35 40S ribosome and 15 p.M ATP,which are not saturating concentrations. As well, the assay was performed in thepresence of 75 mM p-glycerolphosphate which is inhibitory to the kinase. One of thedifficulties in using saturating concentration of 40S is the limitation of its solubility.Purified pgorsk and p70s6k exhibited much higher specific activity (20-600nmol/min/mg), but the Km values of these kinases for 40S (5-10 iiM) are higher thanthe 32-kDa kinase, suggesting that the 32-kDa kinase is a bona fide S6 kinase in vivo.The regulation of the 32-kDa S6 kinase is complelely unknown. Since PP2Awas able to inactivate the kinase activity, phosphorylation is apparently involved inactivating the enzyme. The MAP kinase purified from sea star failed to phosphorylateor activate the kinase activity, neither did CK-2 have any effects on the kinase activityof the 32-kDa S6 kinase, suggesting that this kinase may be in a different activationcascade from the MAP kinase pathway. On the other hand, when MAP kinase andCK-2 were incubated with the 32-kDa S6 kinase in the presence of [y-32N-ATP, therewere an increase in the autophosphorylation of MAP kinase and a decrease in theautophosphorylation of CK-2. The implications of these observation are not clear, butthey may have resulted from certain components in the buffers that was used tocontain the 32-kDa S6 kinase.2474. Future directionsThis study has pioneered the investigation into insulin-stimulatedserine/threonine kinase cascade in an intact animal model under normal and diabeticconditions. It also explored the effects of vanadium compounds on the kinasecascade in intact rats. Furthermore, a potentially novel kinase that is activated byinsulin has been purified from insulin-treated rats. However, only preliminaryanalyses of this model system have been performed. The study opened up a newavenue of research in the understanding of insulin signalling mechanisms undernormal and pathological conditions, but it also raised many unanswered questions.Future studies are required to gain a further understanding of insulin signalling anddiabetes-associated alterations in this model system. The following directions couldbe considered.1) While previously identified MAP kinases are shown to exist in rat skeletalmuscle, this study provided evidence that other forms of MAP kinases are alsopresent in this tissue. It would be significant to further characterized and identifysome of these novel forms of MAP kinases and establish their position in thesignalling pathway of insulin.2) An intriguing finding in this study is the unexpected high level of MBP kinaseactivity in spleen. The activity does not appear to be attributed to the known MAPkinases (p42erk2 or p44 erk1). It would be interesting to analyze this MBP kinaseactivity further, which may lead to identification of other novel MAP kinases.2483) A novel 32-kDa S6 kinase has been purified from rat skeletal muscle. Howthis kinase is regulated in vivo is not known, and the role of this kinase in thesignalling of insulin remains to be established. Cloning and sequencing a cDNA forthis kinase would allow a better understanding of the molecular structure of thiskinase.4) MAP kinase and S6 kinase activity was defective under basal and insulin-stimulated conditions in chronic diabetic rats. The mechanism for these changes arenot known. It would provide further information on the pathogenesis of diabetes toinvestigate the molecular mechanism in the regulation of insulin-stimulated proteinkinases in diabetic rats. Specifically, the MAP kinase and S6 kinases can be purifiedfrom control and diabetic rats and compared with each other to determine whether thekinases per se are altered in diabetes. When the genomic structures of these kinasesare elucidated, it would be important to examine the transcriptional regulation of MAPkinase and S6 kinases.5) Vanadium was shown to have effects on MAP kinase and S6 kinases in thisstudy. It is possible that these effects are mediated through the regulation of geneexpression of these kinases rather than a direct stimulatory effects of vanadium onthese enzymes. Consequently, studies should be extended to determine the effectson vanadium on these kinases at the molecular levels. It would also be interesting tostudy whether the recently identified vanadium-stimulated cytosolic tyrosine kinase isinvolved in the effects of vanadium on MAP kinase and S6 kinase. Since seleniumhas also been shown to have insulin-like effects, the biological function of this elementwould also be examined in this model system as well.2496) Since the commencement of this study, tremendous progress has beenmade regarding the protein kinase cascade in insulin signalling. A number of kinaseshave been recently identified but not examined in the current study, such as MAPkinase kinase. Therefore, it would be significant to investigate these components inthis in vio model system. Novel findings are likely to be revealed by such studies. Itwould also be possible to extend the study in other tissues such as liver andadipocytes.SUMMARY CONCLUSIONS1. Administration of insulin into intact rats activated simultaneously several proteinser/thr kinases including MAP kinases, S6 kinases, and p74raf in skeletal muscles.CK-2 was not significantly activated in skeletal muscle, presumably due to the relativelow levels of its activity in this tissue. This is the first physiologically relevant modelsystem that proved activation of these kinases in response to in vivo administration ofinsulin.2. Using immunoblotting techniques with antibodies against MAP kinases, We haveidentified p42erk2 and p44erki as the major isoforms of MAP kinase in rat skeletalmuscle. P42erk2 appears to constitute a considerable larger portion of the insulin-stimulated MAP kinase activity in rat skeletal muscle. Other related MAP kinases mayalso exist in rat skeletal muscle and spleen, some of which may be novel MAPkinases.3. Similarly, a 100-kDa protein in the second MonoQ peak of S6 kinase activity wastentatively identified as an isoform of p9Orsk in rat skeletal muscle. The S6 kinase in250the first MonoQ peak of S6 kinase activity has been purified to near homogeneity andidentified as a 32-kDa protein. This kinase is a potentially novel S6 kinase in ratskeletal muscle since it did not react with antibodies against the p90rsk or the p70s6k.4. In STZ-induced diabetes, the basal activities of MAP kinase and S6 kinase wereselectively down-regulated, and the severity of the alteration correlated well with theduration of the disease. There was a lack of activation of MAP kinase in the crudemuscle extracts from control rats as well as control rats that were treated with vanadylsulfate in the two-month study, although the S6 kinase was still significantly activatedunder the same experimental conditions. This may suggest that the activation of S6kinase may be mediated, at least in part, by pathways that do not depend on MAPkinases.5. MonoQ fractionation of crude muscle extracts suggest that the activation of MAPkinase and S6 kinase was impaired in the two-month STZ-induced diabetes.6. Treatment of diabetic rats with vanadium compounds such as vanadyl sulphate andBMOV was able to restore the altered MAP kinase and S6 kinase activities and thestimulation of these kinases by insulin injection in long-term diabetic rats. It ispossible that these effects contribute to the glucose-lowering effects of vanadium,although this needs to be substantiated by further experiments.251BibliographyAccili, D., Kadowaki, T., Kadowaki, H., Mosthaf, L., Ullrich, A., and Taylor, S.I.(1992) lmmunoglobulin heavy chain-binding protein binds to misfoldedmutant insulin receptors with mutations in the extracellular domain. J. Biol.Chem. 267, 586-590Ackerman, P., and Osheroff, N. (1989) Regulation of casein kinase II activity byepidermal growth factor in human A-431 carcinomal cells. J. Biol. Chem.264, 11958-11965Adams, P.D., and Parker, P.J. (1992) Activation of mitogen-activated proteinkinase by a MAP kinase-kinase. J. Biol. Chem. 267, 13135-13137Agius, L., and Vaartjes, W.J. (1982) The effects of orthovanadate on fatty acidsynthesis in isolated rat hepatocytes. Biochem. J. 202, 791-794Agostinis, P., Goris, J., Pinna, L.A., and Merlevede, W. (1987) Regulation ofcasein kinase 2 by phosphorylation/dephosphorylation. Biochem. J. 248,785-789Agostinis, P., Van Lint, J., Sarno, S., DeWitte, P., Vandenheede, J.R., andMerlevede, W. (1992) Rapid stimulation of Ser/Thr protein kinasesfollowing treatment of Swiss 3T3 cells with bombesin. J. Biol. Chem. 267,9732-9737Ahn, N.G., and Krebs, E.G. (1990) Evidence for an epidermal growth factor-stimulated protein kinase cascade in Swiss 3T3 cells. Activation of serinepeptide kinase activity by myelin basic protein kinases in vitro. J. Biol.Chem. 265, 11495-11501Ahn, N.G, Seger, R., Bratlien, R.L., Diltz, C.D., Tonks, N.K., and Krebs, E.G.(1991a) Multiple components in an epidermal growth factor-stimulatedprotein kinase cascade. J. Biol. Chem. 266, 4220-4227Ahn, N.A., Seger, R., Bratlien, R.L., and Krebs, E.G. (1991b) Growth factor-stimulated phosphorylation cascade: activation of growth factor-stimulatedMAP kinase. In: Interaction among cell signalling systems. Chichester:Wiley and Son Ltd., pp113-131Ahn, J., Donner, D.B., and Rosen, O.M. (1993) Interaction of the human insulinreceptor tyrosine kinase from the baculovirus expression system withprotein kinase C in a cell-free system. J. Biol. Chem. 268, 7571-7576Alcorta, D., Crews, C.M., Sweet, L.J., Bankston, L., Jones, S.W., and Erikson,R.L. (1989) Sequence and expression of chicken and mouse rsk:homologs of xenopus laevis ribosomal S6 kinase. Mol. Cell. Biol. 9, 3850-3859Alessandrini, A., Crews, C.M., and Erikson, R.L. (1992) Phorbol ester stimulatesa protein tyrosine/threonine kinaes that phosphorylates and activatestheErk1 gene product. Proc. Natl. Acad. Sci. USA. 89, 8200-8204Alvarez, E., Northwood, I.C., Gonzalez, F.A., Latour, D., Seth, A., Abate, C.,Curran,T., and Davis, R.J.(1991) Pro-Leu-Ser/thr-Pro is a consensus252primary sequence for substrate protein phosphorylation. J. Biol. Chem.266, 15277-15285Amatruda, J.M., and Roncone, A.M. (1985) Normal hepatic insulin receptorautophosphorylation in non-ketotic diabetes mellitus. Biochem. Biophys.Res. Comm. 129, 163-170Anderson, N.G., Kigour, E., and Sturgill, T.W. (1991a) Activation of mitogen-activated protein kinase in BC3H1 myocytes by fluoroaluminate. J. Biol.Chem. 266, 10131-10135Anderson, N.G., Li, P., Marsden, L., Williams, N., Roberts, T., and Sturgill T.W.(1991 b) Raf-1 is a potential substrate for mitogen-activated protein kinasein vivo. Biochem. J. 277, 573-576Anderson, N.G., MaIler, J.L., Tonks, N.K., Sturgill, T.W. (1990) Requirement forintegration of signals from two distinct phosphorylation pathways foractivation of MAP kinase. Nature. 343, 651-653Ando, A., Momomura, K., Tobe, K., Yamamoto-Honda, R., Sakura, H., Tamori,Y., Kaburagi, Y., Koshio, 0., Akanuma, Y., Yazaki, Y., Kasuga, M., andKadowaki, T. (1992) Enhanced insulin-induced mitogenesis and mitogen-activated protein kinase activities in mutant insulin receptors withsubstitution of two COOH-terminal tyrosine autophosphorylation sites byphenylalanine. J. Biol. Chem. 267, 12788-12796App, H., Hazan, R., Zilberstein, A., Ullrich, A., Schlessinger, J., and Rapp, U.(1991) Epidermal growth factor stimulates association and kinase activityof raf-1 with the EGF recptor. Mol. Cell. Biol. 11, 913-919Ashworth, A., Nakielny, S., Cohen, P., and Marshall, C. (1992) The amino acidsequence of a mammalian MAP kinase kinase. Oncogene 7, 2555-2556Assimacopoulos-Jeannet, F., and Jeanrenaud, B. (1990) Insulin activates 6-phosphofructo-2-kinase and pyruvate kinase in the liver. J. Biol. Chem.265, 7202-7206Atkinson, M.A., and Maclaren, N.K. (1990) What causes diabetes. ScientificAmerican. (7), 62-71Avruch, J. (1989) The insulin receptor kinase: structure-function relationships. In:Draznin, B., Melmed, S., and LeRoith, D., eds., Insulin Action. New York:Alan R. Liss, Inc., pp65-77Avruch, J., Tornqvist, H.E., Gunsalus, J.R., Yurkow, E.J., Kyriakis, J.M., andPrice, D.J. (1990) Insulin regulation of protein phosphorylation. In :Cuatrecasas, P. and Jacobs, S. (eds), Insulin. Berlin: Springer-Verlag,pp314-366Avruch, J., Price, D.J., Banerjee, P., Ahmad, M.F., Grove, J.R., Mukhopadhyay,N., Balasubramanyam, A., and Kyriakis, J.M. (1991). The structure andregulation of the insulin-activated S6 (ser/thr) protein kinases : signalingintermediates in insulin action. In Shigeta, Y., Kobayashi, M., and Olefsky,253J.M., eds. Recent advances in insulin action and its disorders.Amsterdam, Elsevier Science Publishers B.V. pp63-79Baccarini, M., Sabatini, D.M., App, H., Rapp, U.R., and Stanley, E.R. (1990)Colony stimulating factor-1 (CSF-1) stimulates temperature dependentphosphorylation and activation of the raf-1 proto-oncogene product.EMBO J. 11, 3649-3657Baccarini, M., Gill, G.N., and Stanley, E.R. (1991) Epidermal growth factorstimulates phosphorylation of raf-1 independently of receptorautophosphorylation and internalization. J. Biol. Chem. 266, 10941-10945Backer, J.M., Kahn, C.R., Cahill, D.A., Ullrich, A., and White, M.F. (1990)Receptor-mediated internalization of insulin requires a 12-amino acidsequence in the juxtamembrane region of the insulin receptor b-subunit.J. Biol. Chem. 265, 16450-16454Backer, J.M., Myers, M.G., Jr., Shoelson, S.E., Chin, D.J., Sun, X.J., Miralpeix,M., Hu, P., Margolis, B., Skolnik, Y., Schlessinger, J., and White, M.F.(1992) Phosphatidylinositol 3'-kinase is activated by association with IRS-1 during insulin stimulation. EMBO J. 11, 3469-3479Backer, J.M., Myers, M.G., Sun, X.J., Chin, D.J., Shoelson, S.E., Miralpeix, M.,and White, M.F. (1993) Association of IRS-1 with the insulin receptor andthe phosphatidylinositol 3-kinase. J. Biol. Chem. 268, 8204-8212Baekkeskov, S., Aanstoot, H.J., Christgau, S., Reetz, A., Solimena, M.,Cascalho, M., Folli, F., Richter-Olesen, H., and Camillli, F.D. (1990)Identification of the 64K autoantigen in insulin-dependent diabetes as theGABA-synthesizing enzyme glutamic acid decarboxylase. Nature 347,151-156Baldini, G., Hohman, R., Charron, M.J., and Lodish, H.F. (1991) Insulin andnonhydrolyzable GTP analogs induce translocation of GLUT4 to theplasma memebrane in a-toxin-permeabilized rat adipose cells. J. Biol.Chem. 266, 4037-4040Ballotti, R., Lammers, R., Scimeca, J.C., Dull, T., Schlessinger, J., Ullrich, A.,and Van Obberghen, E. (1989) Intermolecular transphosphorylationbetween insulin receptors and EGF-insulin receptor chimerae. EMBO J. 8,3303-3309Ballou, L.M., Jeno, P., and Thomas, G. (1988a) Protein phosphatase 2Ainactivated the mitogen-stimulated S6 kinase from Swiss mouse 3T3 cells.J. Biol. Chem. 263, 1188-1194Ballou, L.M., Siegmann, M., and Thomas, G. (1988b) S6 kinase in quiescentSwiss mouse 3T3 cells is activated by phosphorylation in response toserum treatment. Proc. Natl. Acad. Sci. USA. 85, 7154-7158Baltensperger, K., Lewis, R.E., Woon, C.W., Vissavajjhala, P., Ross, A.H., andCzech, M.P. (1992) Catalysis of serine and tyrosine autophosphrylationby the human insulin receptor. Proc. Natl. Acad. Sci. USA. 89, 7885-7889254Banerjee, P., Ahmad, M.F., Grove, J.R., Kozlosky, C. Price, D.J., and Avruch, J.(1990) Proc. Natl. Acad. Sci. 87, 8550-8554Bark, S., and Banerjee, A.K. (1992) Phosphorylation by cellular casein kinase IIis essential for transcriptional activity of vesicular stomatitis virusphosphoprotein P. Proc. Natl. Acad. Sci. USA. 89, 6570-6574Barrett, C.B., Erikson, E., and Mailer, J. (1992) A purified S6 kinase kinase fromXenopus eggs activates S6 kinase II and autophosphorylates on serine,threoine, and tyrosine residues. J. Biol. Chem. 267, 4408-4415Basabe, J.C. (1989) The immune system and islet function. In:Larkins, R.,Zimmet, P., and Chisholm, D., eds., Diabetes 1988, Amsterdam: ElsevierScience Publishers, B.V., pp37-43Battell, M.L., Yuen, V.G., and McNeill, J.H. (1992) Treatment of BB rats withvanadyl sulphate. Pharmacol. Comm. 1, 291-301Beck-Nielsen, H., Vaag, A., Damsbo, P., Handberg, A., Nielsen, 0.H.,Henriksen, J.E., and Thye-Ronn, P. (1992) Insulin resistance in skeletalmuscles in patients with NIDDM. 15, 418-429Becker, A.B., and Roth, R.A. (1990) Insulin receptor structure and function innormal and pathological conditions. Annu. Rev. Med. 41, 99-115Begum, N., Leitner, W., Reusch, J.E.B., Sussman, K.E., and Draznin, B. (1993)GLUT4 phosphorylation and its intrinsic activity. J. Biol. Chem. 268, 3352-3356Bell, R.H., and Hye, R. (1983) animal models of diabetes melluts: physiology andpathology. J. Surg. Res. 36, 433-460Benabe, J., Echegoyen, L.A., Pastrana, B., Martinez-Maldonado, M. (1987)Mechanism of inhibition of glycolysis by vanadate. J. Biol. Chem. 262,9555-9560Bendayan, M., and Gingras, D. (1989) Effect of vanadate administration onblood glucose and insulin livels as well as ton the exocrine pancreaticfunction in streptozotocin-diabetic rats. Diabetolog. 32, 561-567Berggren, P.O., Hallberg, A., Welsh, N., Arkhammar, P., Nilsson, T., and Welsh,M. (1989) Transfection of insulin-producing cells with a transforming c-Ha-ras oncogene stimulates phospholipase C activity. Biochem. J. 259, 701-707Bernier, M., Laird, D.M., and Lane, M.D. (1988b) Effect of vanadate on thecellular acccumulation of pp15, an apparent product of insulin receptortyrosine kinase action. J. Biol. Chem. 263, 13626-13634Bernier, M., Laird, D.M., and Lane, M.D. (1988a) Identification and role of a 15-kDa cellular target of the insulin receptor tyrosine kinase. In: Goren, J.,Hollenberg, M.D., and Roncari, K.K., eds., Insulin Action and Diabetes.New York: Raven Press, pp117-128255Bidwai, A.P., Hanna, D.E., and Glover, C.V.C. (1992) Purification andcharacterization of casein kinase II from CKa1 and cka2 S. cerevisiaerescued by Drosophila CKII subunits. J. Biol. Chem. 267, 18790-18796Bird, T.A., Sleath, P.R., deRoos, P.C., Dower, S. K., and Virca, G.D. (1991)Interleukin-1 represents a new modality for the activation of extracellularsginal-regulated kinases/microtubule-associated protein-2 kinases. J. Biol.Chem. 266, 22661-22670Birnbaumer, L. (1992) Receptor-to-effector signalling through G proteins : Rolesfor (3'y dimers as well as a subunits. Cell 71, 1069-1072Blackshear, P.J., Haupt, D.M., and Stumpo, D.J. (1991) Insulin activation ofprotein kinase C: a reassessment. J. Biol. Chem. 266, 10946-10952Blackshear, P.J., Haupt, D.M., Ap, H., and Rapp, U.R. (1990) Insulin activatesthe Raf-1 protein kinase. J. Biol. Chem. 265, 12131-12134Blenis, J. (1991) Growth-regulated signal transduction by the MAP kinasesRSKs. Cancer Cells. 3, 1-5Blenis, J., Kuo, C.J., and Erikson, R.L. (1987) Identification of a ribosomalprotein S6 kinase regulated by transformation and growth-promotingstimuli. J. Biol. Chem. 262, 14373-14376Block, N.E., Komori, K., Robinson, K.A., Dutton, S.L., Lam, C.F., and Buse, M.G. (1991a) Diabetes-associated impairment of hepatic insulin receptortyrosine kinase activity : a study of mechanisms. Endocrinol. 128, 312-322Block, N.E., Komori, K., Dutton, S., Robinson, K.A., and Buse, M.G. (1991b)Skeletal muscle insulin-receptor kinase. Diabetes 40, 1691-1700Blondel, 0., Bailbe, D., and Portha, B. (1989) In vivo insulin resistance instreptozotocin-diabetic rats-evidence for reversal following oral vanadatetreatment. Diabetol. 32, 185-190Blondel, 0., Simon, J., Chevalier, B., and Portha, B. (1990) Impaired insulinaction but normal insulin receptor activity in diabetic rat liver : effect ofvanadate. Am. J. Physiol. 258, E459-467Bogden, J.D., Higashino, H., Lavenhar, M.A., Bauman, J.W., Kemp, F.W.,, andAviv, A. (1982) Balance and tissue distribution of vanadium after shortterm ingestion of vanadate. J. Nutr. 112, 2279-2285Bollen, M., Miralpeix, M., Ventura, F., Toth, B., Bartrons, R., and Stalmans, W.(1990) Oral administration of vanadate to streptozotocin-diabetic ratrestores the glucose-induced activation of liver glycogen synthase.Biochem. J. 267, 269-271Bonadonna, R.C., Saccomani, M.P., Cobelli, C., and DeFronzo, R.A. (1993)Effect of insulin on system A amino acid transport. J. Clin. Invest. 91, 514-521256Boni-Schnetzler, M., Kaligian, A., Vecchio, R.D., and Pilch, P.F. (1988) J. Biol.Chem. 263, 6822-6828Bosch, F., Arino, J., Gomez-Foix, A.M., and Guinovart, J.J. (1987)Glycogenolytic noninsulin-like effects of vanadate on rat hepatocyteglycogen synthase and phosphorylaase. J. Biol. Chem. 262, 218-222Bottazzo, G.F., Bosi, E., Bonifacio, E., Mirakian, R., Todd, L, and Pujol-Borrell,R. (1989) Pathogenesis of Type I (insulin-dependent) diabetes: possiblemechanisms of autoimmune damage. Brit. Med. Bull. 45, 37-57Boulton, T.G., Yancopoulos, G.D., Gregory, J.S., Slaughter, C. Moomaw, C.,Hsu, J., and Cobb, M.H. (1990) An insulin-stimulated protein kinasesimilar to yeast kinases involved in cell cycle control. Science. 249, 64-67Boulton, T.G., Gregory, J.S., and Cobb, M.H. (1991a) Purification and propertiesof extracellular signal-regulated kinase 1, an insulin stimulatedmicrotubule-associated protein 2 kinase. Biochem. 80, 278-286Boulton, T.G., Nye, S.H., Robbins, D.J., Ip, N.Y., Radziejewska, E.,Morgenbesser, S.D. Depinho, R.A., Panayotatos, N., Cobb, M.H.,Yancopoulos, G.D. (1991b) ERKs: a family of protein serine/threoninekinases that are activated and tyrosine phosphorylated in response toinsulin and NGF. Cell. 65, 663-675Boulton, T.G., and Cobb, M.H. (1991c) Identification of multiple extracellularsignal-regulated kinases (ERKs) with antipeptide antibodies. CellRegulation. 2, 357-371Bourgoin, S., and Grinstein, S. (1992) Peroxides of vanadate induce activation ofphospholipase Din HL-60 cells. J. Biol. Chem. 267, 11908-11916Bourne, H.R., Wrischnik, L., and Kenyon, C. (1990) Some signal development.Nature 348, 678-679Bowtell, D., Fu, P., Simon, M., and Senior, P. (1992) Identification of murinehomologues of the Drosophila son of sevenless gene: potential activatorsof ras. Proc. Acad. Sci. USA. 89, 6511-6515Boyd, D.W., and Kustin, K. (1984) Vanadium: a versatile biochemical effectorwith an elusive biological function. Adv. Inorg. Biochem. 6, 311-365Bradford, M.M. (1976) A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dyebinding. Anal. Biochem. 72, 248-254Brichard, S.M., Okitolonda, W., and Henquin, J.C. (1988) Long termimprovement of glucose homeostasis by vanadate treatment in diabeticrats. Endocrinol. 123, 2048-2053Brichard, S.M., Assimacopoulos-Jeannet, F., and Jeanrenaud, B. (1992)Vanadate treatmnet markedly increases glucose utilization in muscle ofinsulin-resistant fa/fa rats without modifying glucose transporterexpression. Endocrinol. 131, 311-317257Brillion, D.J., Freidenberg, G.R., Henry, R.R., and Olefsky, J.M. (1989)Mechanism of defective insulin-receptor kinase activity in NIDDM.Diabetes 38, 397-403Brockman, J.L., Gross, S.D., Sussman, M.R., and Anderson, R.A. (1992) Cellcycle-dependent localization of casein kinas I to mitotic spindles. Proc.Natl. Acad. Sci. USA. 89, 9454-9458Brown, D.J., and Gordon, J.A. (1984) The stimulation of pp60v-src kinase activityby vanadate in intact cells accompanies a new phosphorylation state ofthe enzyme. J. Biol. Chem. 259, 9580-9586Browner, M.F., and Fletterick, R.J. (1992) Phosphorylase: a biologicaltransducer. Trends. Biochem. Sci. 17, 66-71Brownsey, R. W., Quayle, K.A., and Mui, A.L. (1988) Studies on the regulation ofacetyl-CoA carboxylase by insulin. In: Goren, H.J., Hollenberg, M.D., andRoncari, D.A., eds., Insulin Action and Diabetes. New York: Raven Press,173-184Brownsey, R.W., and Denton, R.M. (1985) Role of phosphorylation in theregulation of acetyl CoA carboxylase activity. In: Czech, M.P., ed.Molecular Mechanism of Insulin Action. New York: Plenum Press, pp297-314Brownsey, R.W., Edgell, N.J., Hopkins, T.J., and Denton, R.M. (1984) Studies oninsulin-stimulated phosphorylation of acetyl CoA carboxylase, ATP citratelyase, and' other proteins in rat epididymal adipose tissue. Evidence foractivation of a cyclic AMP-independent protein kinase. Biochem. J. 218,733-743Brownsey, R.W., and Denton, R.M. (1982) Evidence that insulin activates fat cellacetyl CoA carboxylase by increased phosphorylation at a specific site.Biochem. J. 202, 77-86Brownsey, R.W., Hughes, W.A., and Denton, R.M. (1977) Demonstration ofphosphorylation of aceyl-CoA carboxylase within intact rat epididymal fatcells. Biochem. J. 168, 441-445Bruder, J.T., Heidecker, G., and Rapp, U.R. (1992) Serum-,TPA, and ras-induced expression from Ap-1/ETs-driven promoters requires raf-1kinase. Gene Devel. 6, 545-556Brunetti, A., Maddux, B.A., Wong K.Y., Hoffmann, C., Whittaker, J., Sung, C.,and Goldfine, I.D. (1989) Monoclonal antibodies to the human insulinreceptor mimic a spectrum of biological effects in transfected 3T3/HIRfibroblasts without activating receptor kinase. Biochem. Biophys. Res.Comm. 165, 212-218Budday, L., and Downward, J. (1993a) Epidermal growth factor regulates theexchange rate of guanine nucleotides on p21ras in fibroblasts. Mol. Cell.Biol. 13, 1903-1910258Buday, L., and Downward, J. (1993b) Epidermal growth factor regulates p21rasthrough the formation of a complex of receptor, Grb2 adapter protein, andSos nucleotide exchange factor. Cell 73, 611-620Burant, C.F., Treutelaar, M.K., and Buse, M. (1986) Diabetes-induced functionaland structural changes in insulin receptors from rat skeletal muscle. J.Clin. Invest. 77 : 260-270Burgering, B.M.T., Snijders, A.J., Maassen, J.A. Van der Eb, A., and Bos, J.L.(1989) Possible involvement of normal p21 H-ras in the insulin/insulinlikegrowth factor signal transduction pathway. Mol. Cell. Biol. 9, 4312-4322Burgering, B.M.T., Medema, R.H., Maassen, J.A., Van de Wetering, M.L., Vander Eb, A.J., McCormick, F., and Bos, J.L. (1991) Insulin stimulation ofgene expression mediated by p21ras activation. EMBO J. 10, 1103-1109Buse, J.B., Yasuda, K., Lay, T.P., Seo, T.S., Olson, A.L., Pessin J.E., Karam,J.H., Seino, S., and Bell, G. I. (1992) Human GLUT4/muscle-fat glucosetransporter gene. Diabetes 41, 436-445Cairn, B.R., Ramer, S.W., and Kornberg, R.D. (1992) Order of action ofcomponents in the yeast pheromone response pathway revealed with adominant allele of the STE11 kinase and the multiple phosphorylation ofthe STET kinase. Genes and Devel. 6, 1305-1318Calvo, V., Crews, C.M., Vik, T.A., and Bierer, B.E. (1992) Interleukin wstimulation of p70 S6 kinase activity is inhibited by the immunosuppresantrapamycin. Proc. Natl. Acad. Sci. USA 89, 7571-7575Cam, M.C., Pederson, R.A., Brownsey, R.W., and McNeill, J.H. (1993)Effectiveness of vanadyl sulphate as an antididabetic agent when chronictreatment is delayed following streptozotocin-induced diabetes. Diabetol.36, 218-224Campa, M.J., Glickman, J.F., Yamamoto, K, and Chang, K.J. (1992) Theantibiotic azatyrosine suppresses progesterone or [Va112]p21 Ha-ras/insulin-like growth factor I-induced germinal vesicle breakdown andtyrosine phosphorylation of Xenopus mitogen-activated protein kinase inoocytes. Proc. Natl. Acad. Sci. USA 89, 7654-7658Campos-Gonzalez, R., and Glenney, J.R. (1992) Tyrosine phosphorylation ofmitogen-activated protein kinase in cells with tyrosine kinase negativeepidermal growth factor receptors. J. Biol. Chem. 267, 14535-14538Canalis, E. (1985) Effect of sodium vanadate on deoxyribonucleic acid andprotein synthesis in cultured rat calvariae. Endocrinol. 116, 855-862Cantley, L.C., Josephson, L., Warner, R., Yanaisawa, M., Lechene, C., andGuidotti, G. (1977) Vanadate is a potent Na, K-ATPase inhibitro found inATP derived from muscle. J. Biol. Chem. 252, 7421-7423Cantley, L.C., and Aisen, P. (1979) The fate of cytoplasmic vanadium J. Biol.Chem. 254, 1781-11784259Cantley, L.C., Auger, K.R., Carpenter, C., Duckworth, B., Graziani,A., Kapeller,R., and Soltoff, S. (1991) Oncogenes and signal transduction. Cell 64,281-302Caron, M., Melin, B., Wicek, D., Capeau, J., and Picard, J. (1989) Insulin actionis mimicked by polyclonal antireceptor antibodies that activate the insulinreceptor tyrosine kinase. Horm. Metabol. Res. 21, 295-300Carpenter, G. (1981) Vanadate, epidermal growth factor and the stimulation ofDNA synthesis. Biochem. Biophys. Res. Comm. 102, 1115-1121Carpentier, J.L. (1989) The cell biology of insulin receptor. Diabetol. 32, 627-635Carreras, M., Bassols, A.M., and Climent, F. (1988) Effects of vanadate andinsulin on glucose 1,6 Pa and fructose 2,6 Pa levels in rat skeletal muscle.Biochem. Int. 17, 359-366Carroll, D., and Marshak, D.R. (1989) Serum-stimulated cell growth causedoscillations in casein kinase II activity. J. Biol. Chem. 264, 7345-7348Carroll, M.P., Spivak, J.L., McMahon, M., Weich, N., Rapp, U.R., and May, W.S.(1991) Erythropoietin induces raf-1 activation and raf-1 is required forerythropoietin-mediated proliferation. J. Biol. Chem. 266, 14964-14969Casillas, A., Hanekom, C., Williams, K., Kata, R., and Nel, A.E. (1991)Stimulation of B-cells via the membrane immunoglobulin receptor or withphorbol myristate 13-acetate induceds tyrosine phosphorylation andactivation of a 42-kDa microtubule-associated protein 2 kinase. J. Biol.Chem. 266, 19088-19094Cegielska, A., and Virshup, D.M. (1993) Control of simian virus 40 DNAreplication by the HeLa cell nuclear kinase casein kinase I. Mol. Cell. Biol.13, 1202-1211Challiss, R.A.J., Leighton, B., B., Lozeman, F.J., Budohoski, L., and Newsholme,E.A. (1987) Effects of chronic adiminstration of vanadate into the rat onteh sensitivity of glycolysis an glycogen synthesis in skeletal muscle toinsulin. Biochem. Pharmacol. 36, 357-361Chan, S.J., Keim, P., Steiner, D.F. (1976) Cell free synthesis of ratpreproinsulin:characterization and partial amino acid sequencedetermination. Proc. Natl. Acad. Sci. USA. 73, 1964-1968Chao, T.S.O., Byron, K.L., Lee, K.M„ Villereal, M., and Rosner, M.R. (1992)Activation of MAP kinases by calcium-dependent and calcium-independent pathways. J. Biol. Chem. 267, 19876-19883Chatani, Y., Tanaka, E., TObe, K., Hattori, A., Sato, M., Tamenoto, H.,Nishizawa, N., Nomoto, H., Takeya, T., Kadowaki, T., Kasuga, M., andKohno, M. (1992) Mitogen-induced tyrosine phosphorylated 41- and 43-kDa proteins are family members of extracellular sginal regulated proteinkinases/microtubule-associated protein kinase. J. Biol. Chem. 267, 9911-9916260Cheatham, B., and Kahn, C.R. (1992) Cysteine 647 in the insulin receptor isrequired for normal covalent interaction between a and b subunits andsignal transduction. J. Biol. Chem. 267, 7108-7115Chen, R.H., Chung, J.Y., and Blenis, J. (1990) Regulation of pp90rskphosphorylation and S6 phosphotransferase activity in Swiss 3T3 cells bygrowth factor-, phorbol ester-, and cyclic AMP-mediated signaltransduction. Mol. Cell. Biol. 11, 1861-1867Cheng, J.T., Cobb, M.H., and Baer, R. (1993) Phosphorylation of the TALIoncoprotein by the extracellular signal-regulated protein kinase ERK1.Mol. Cell. Biol. 13, 801-808Cherqui, G., Reynet, C., Caron, M., Melin, B., Wicek, D., Calauser, E., Capeau,J., and Picard, J. (1990) Insulin receptor tyrosine residues 1162 and 1163control insulin stimulation of nynstoyl-diacylglycerol generation andsubssequnet activation of glucose transport. J. Biol . Chem. 265, 21254-21261Childs, T.J., Watson, M.H., Sanghera, J.S., Campbell, D.L., Pelech, S.L., andMak, A.S. (1992) Phosphorylation of smooth muscle caldesmon bymitogen-activated protein kinase and expression of MAP kinase indifferentiated smooth muscle. J. Biol. Chem. 267, 22854-22859Chistie, M.R., Pipeleers, D.G., Lernmark, A., and Baekkeskov, S. (1990) Cellularand subcellular localization of an Mr 64,000 protein autoantigen in insulin-dependent diabetes. J. Biol. Chem. 265, 376-381Chou, C.K., Dull, T.J., Russel, D.S., Gherzi, R., Lebwohl, D., Ullrich, A, andRosen, O.M. (1987) Human insulin receptors mutated at the ATP-bindingsite lack protein tyrosine kinase activity and fail to mediate postreceptoreffects of insulin. J. Biol. Chem. 262, 1842-1847Chou, S.Y., Baichwal, V., Ferrell, J.E. Jr. (1992) Inhibition of c-jun DNA bindingby mitogen-activated protein kinase. Mol. Biol. Cell. 3, 1117-1130Chung, J., Pelech, S.L., and Blenis, J. (1991a) Mitogen-activated Swiss mouse3T3 RSK kinases I and II are related to pp44mpk from sea star oocytesand participate in the regulation of pp90rsk activity. Proc. Natl. Acad. Sci.USA. 88, 4981-4985Chung, J., Chen, R.H., and Blenis, J. (1991b) Coordinate regulation of pp90rskand a distinct protein serine/threonine kinase activity that phosphorylatesrecombinant pp90rsk in vitro. Mol. Cell. Biol. 11, 1868-1874Chung, J., Kuo, C.J., Crabtree, G.R., and Blenis, J. (1992) Rapamycin-FKBPspecificically blocks growth-dependent activation of and signalling by the70 kd S6 protein kinases. Cell. 69, 1227-1236Cicirelli, M.F., Tonks, N.K., Diltz, C.D., Weiel, J.E., Fischer, E.H., and Krebs,E.G. (1990) Microinjection of a protein-tyrosine phosphatase inhibitorinhibits insulin action in Xenopus oocytes. Proc. Natl. Acad. Sci. USA. 87,5514-5518261Clancy, B.M., and Czech, M.P. (1990) Hexose transport stimulation andmembrane redistribution of glucsoe transporter isoforms in response tocholera toxin, dibutyryl cyclic AMP, and insulin in 3T3-L1 adipocytes. J.Biol. Chem. 265, 1 2434-1 2443Clancy, B.M., Harrison, S.A., Buxton, J.M., and Czech, M.P. (1991) Proteinsynthesis inhibitors activate glucose transport without increasing plasmamembrane glucose transporters in 3T3-L1 adipocytes. J. Biol. Chem. 266,10122-10130Clark, A.S., Fgan, J.M., and Mitch, W.E. (1985) Selectivity of the insulin-likeaction of vanadate on glucse and protein metabolism in skeletal muscle.Biochem. J. 232, 273-276Clark, S.G., Stern, M.J., Horvitz, H.R. (19932) C. elegans cell-signalling genesem5 encodes a protein with SH2 and SH3 domains. Nature, 356, 340-344Clark-Lewis, I., Sanghera, J.S., and Pelech, S.L. (1991) Definition of aconsensus sequence for peptide substrate recognition by p44mpk, themeiosis-activated myelin basic protein kinase. J. Biol. Chem. 266,15180-15184Cobb, M.H., and Rosen, O.M. (1983) Description of a protein kinase derivedform insulin-treated 3T3-L1 cells that catalyzes the phosphorylation ofribosomal protein S6 and casein. J. Biol. Chem. 258, 12472-12481Cobb, M.H. (1986) An insulin-stimulated ribosomal protein S6 kinase in 3T3-L1cells. J. Biol. Chem. 261, 12994-12999Cobb, M.H., Boulton, T.G., and Robbins, D.J. (1991a) Extracellular signal-regulated kinases : ERKs in progress. Cell Reg. 2, 965-978Cobb, M.H., Robbins, D.J. and Boulton, T.G. (1991b) ERKs, extracellular signal-regulated MAP-2 kinases. Current Opinion in Cell Biology. 3, 1025-1032Cohen, P., Parker, P.J., and Woodgett, J.R. (1985) The molecular mechanismby which insulin activates glycogen synthase in mammalian skeletalmuscle. In: Czech, M.P., ed., Molecular Basis of Insulin Action. New York:Plenum Press, pp213-233Cohen, P. (1986) Muscle glycogen synthase. The Enzymes 17, 462-499Cohen, P. (1987) Molecular mechanisms involved in the control ofglycogenolysis in skeletal muscle by calcium ion and cyclic AMP.Biochem. Soc. Trans. 15, 999-1001Cohen, P. (1989) The structure and regulation of protein phosphatases. Annu.Rev. Biochem. 58, 453-508Cohen, P., and Cohen, P.T.W. (1989) Protein phosphatase come of age. J. Biol.Chem. 264, 21435-21438262Cooper, D.R., Konda, T., Standaert, M., Davis, J.S., Pollet, R.J., and Farese,R.V. (1987) Insulin increases membrane and cytosolic protein kinase Cactivity in BC3H-1 myocytes. J. Biol. Chem. 262, 3633-3639Cooper, J. (1989) Related proteins are phosphorylated at tyrosine in response tomitogenic stimulin and at meiosis. Mol. Cell. Biol. 9, 3143-3147Cordera, R., Andraghetti, G., DeFronzo, R.A., and Rossetti, L. (1990) Effect of invivo vanadate treatment on insulin receptor tyrosine kinase activity inpartially pancreactomized diabetic rats. Endocrinol. 126, 2177-2183Cordle, S.R., Whelan, J., Henderson, E., Masuoka, H., Weil, P.A., and Stein, R.(1991) Insulin gene expression in nonexpressing cells appears to beregulated by multiple negative-acting control elements. Mol. Cell. Biol. 11,2881-2886Cornish-Bowden, A., and Cardenas, M.L. (1991) Hexokinase and glucokinase inliver metabolism. Trends Biochem. Sci. 16, 281-282Corvera, S., Jaspers, S., and Pasceri, M. (1991) Acute inhibition of insulin-stimulated glucose transport by the phosphatase inhibitor, okadaic acid. J.Biol. Chem. 266, 9271-9275Craig, R.G., Rowe, D.W., Petersen, D.N., and Kream, B.E. (1989) Insulinincreases the steady state level of a-1(I) procollagen mRNA in theosteoblast-rich segment of fetal rat calvaria. Endocrinol. 125, 1430-1437Crans, D., Simone, C.M., Saha, A., and Glew, R.H. (1989) Vanadate monomerand dimer both inhibit the human prostatic acid phosphatase. Biochem.Biophys. Res. Comm. 165, 246-250Crews, C.M., Alessamdrini, A.A., and Erikson, R.L. (1991) Mouse ERK-1 geneproduct is a serine/threonine protein kinase that has the potential tophosphorylate tyrosine. Proc. Natl. Acad. Sci. USA. 88, 8845-8849Crews, C.M., Alessandrini, A., and Erikson, R.L. (1992a) The primary structureof MEK, a protein kinase that phosphorylates the ERK gne product.Science 258, 478-480Crews, C.M., and Erikson, R.L. (1992b) Purification of a murine proteintyrosine/threonine kinase that phosphorylates and activates the ERK1gene product: relationship to the fission yeast byrl gene product. Proc.Natl. Acad. Sci. USA. 89, 8205-8209Cros, G., Mongold, J.J., Serrano, J.J., Ramanadham, S., and McNeill, J. H.(1992) Effects of vanadyl derivatives on animal models of diabetes. Mol.Cell. Biochem. 109, 163-166Cryer, P.E. (1989) Counterregulatory hormones and the maintenance ofnormoglycemia. In: Larkins, R., Zimmet, P., and Chisholm, D., eds.Diabetes 1988. Amsterdam: Elsevier Science Publishers B.V., pp561-563Czech, M.P., Klarlund, J.D., Yagaloff, K.A., Bradford, A.P., and Lewis, R.E.(1988) Insulin receptor signaling. J. Biol. Chem. 263, 11017-11020263Czech, M.P., Clancy, B.M., Pessino, A., Woon, C.W., and Harrison, S.A. (1992)Complex regulation of simple sugar transport in insulin-responsive cells.Trends. Biochem. Sci. 17, 197-201Damus, M.E. (1981) Purification and properties of calf thymus casein kinase Iand II. J. Biol. Chem. 256, 3319-3325De Vries-Smits, A.M.M., Burgering, B.M.R., Leever, S.J., Marshall, C.J., andBos, J.L. (1992) Involvement of p21ras in activation of extracellular signalregulated kinase 2. Nature 357, 602-604DeFronzo, R.A. (1988) Triumvirate : 13-cell, muscle, liver, a collusion responsiblefor NIDDM. Diabetes 37, 667-687DeFronzo, R.A. (1992) Pathogenesis of NIDDM. Diabetes Care 15, 318-368Del Grande, R.W., and Traugh, J.A. (1982) Phosphorylation of 40S ribosomalsubunits by cAMP-dependent, cGMP-dependent and protease-activatedprotein kinases. Eur. J. Biochem. 123, 421-428DeMaggio, A.J., Lindberg, R.A. Hunter, T., and Hoekstra, M.F. (1992) Thebudding yeast HRR25 gene product is a casein kinase I isoform. Proc.Natl. Acad. Sci. USA. 89, 7008-7012Dent, P., Lavoinne, A., Nakielny, S., Caudwell, F.B., Watt, P., and Cohen, P.(1990) The molecular mechanism by which insulin stimulates glycogensynthesis in mammalian skeletal muscle. Nature 348, 302-308Dent, P., Haser, W., Haystead, A.J., Vincent, L.A., Roberts, T.M., and Sturgill,T.W. (1992) Activation of mitogen-activated protein kinase kinase by v-rafin NIH 3T3 cells and in vitro. Science 257, 1404-1407Derewenda, U., Deresenda, A.S., Dodson, G.G., and Hubbard, R.E. (1990)Insulin structure. In : Cuatrecasas, P., and Jacobs, S., ed. Insulin. Berlin :Springer-Verlag, pp23-39Dickens, M., Chin, J.E., Roth, R.A., Ellis, L., Denton, R.M., and Tarare, J.M.(1992) Biochem. J. Characterization of insulin-stimulated proteinserine/threonine kinases in CHO cells expressing human insulin receptorswith point and deletion mutations. Biochem. J. 287, 201-209Dickson, B., Sprenger, F., Morrison, D., and Hafen, E. (1992) Raf functionsdownstream of ras1 in the sevenless signal transduction pathway. 360,600-603Diggle, T.A., Schmitz-peiffer, C., Borthwick, A.C., Welsh, G.I., and Denton, R.M.(1991) Evidence that insulin activates casein kinase 2 in rat epididymal fatcells and that this may result in the incresed phosphorylation of an acid-soluble 22 kDa protein. Biochem. J. 279, 545-551Domingo, J.L., Gomez, M., Llobet, J.M., Corbella, J., and Keen, C.L., (1991) Oralvanadium administration of streptozotocin-diabetic rats has marked264negative side effects which are independent of the form of vanadiumused. Toxicol. 66, 279-287Donahue, M.J., and Masaracchia, R.A. (1984) Phosphorylation of ribosomalprotein S6 at multiple sites by cyclic AMP-independent protein kinase fromlymphoid cells. J. Biol. Chem. 259, 435-440Downward, J., Graves, J.D., Warne, P.H., Rayter, S., and Cantrell, D.A., (1990)Stimulation of p21 ras upon T cell activation. Nature 346, 719-723Downward, J. (1992) Ras regulation : putting back the GTP. Curr. Biol. 2, 329-331Drewes, G., Lichtenberg-Kraag, B., Doring, F., Mandelkow, E.M., Biernat, J.,Goris, J., Doree, M., and Mandelkow, E. (1992) Mitogen-activated proteinkinase transforms tau protein into an Alzheimer-like state. EMBO J. 11,2131-2138Dubyak, G.R., and Kleinzeller, A. (1980) The insulin-mimetic effects of vanadatein isolated rat adipocytes. J. Biol. Chem. 255, 5306-5312Duckworth, W.C., Solomon, S.S., Liepnieks, J., Hamel, F.G., Hand, S., andPeavy, D.E. (1988) Insulin-like effects of vanadate in isolated ratadipocytes. Endocrinol. 122, 2285-2289Duff, J.L., Berk, B.C., and Corson, M.A. (1992) Angiotensin II stimulates thepp44 and pp42 MAP kinases in cultured rat aortic smooth muscle cells.Biochem. Biophys. Res. Comm. 188, 257-264Dulin, W.E., Gerritsen, G.C., and Chang, A.Y. (1983) Experimental andspontaneous diabetes in animals. In: Ellenberg, M., and Riffain, H., ed.Diabetes Mellutus: theory and practice (3rd ed.), New York: MedicalExamination Publishing Co., Inc., pp361-408Ebina, Y., Ellis, L., Jarnagin, K., Edery, M., Graf, L., Cluaser, E., Ou, J.H.,Masiarz, F., Kan, Y.W., Goldfine, I.D., Roth, R.A., and Rutter, W.J. (1985)The human insulin receptor cDNA: the structural basis for homone-activated transmembrane signalling. Cell 40, 747-758Efrat, S., Surana, M., and Fleischer, N. (1991) Glucose induces insulin genetranscription in a murine pancreatic b-cell line. J. Biol. Chem. 266, 11141-11143Egan, J.J., Saltis, J., Wek, S.A., Simpson, I.A., and Londos, C. (1990) Insulin,oxytocin, and vasopressin stimulate protein kinase C activity in adipocyteplasma membrane. Proc. Natl. Acad. Sci. USA. 87, 1052-1056Egan, S.E., Giddings, B.W., Brooks, M.W., Buday, L., Sizeland, A.M., andWeinberg, R.A. (1993) Association of sos Ras exchange protein with Grb2is implicated in tyrosine kinaes signal transduction and transformation.Nature 363, 45-50Ellis, L., Clauser, E., Morgan, D.O., Edery, M., Roth, R.A.,and Rutter, W.J.(1986) Replacement of insulin receptor tyrosine residues 1162 and 1163265compromises insulin-stimulated kinase activity and uptake of 2-deoxyglucose. Cell 45, 721-732Ellis, C., Moran, M., McCormick, F., and Pawson, T. (1990) Phosphorylation ofGAP and GAP-associated proteins by transforming and mitogenic tyrosinekinases. Nature 343, 377-380Ellion, E.A., Brill, J.A., and Fink, G.R. (1991) FUS3 supresses CLN1 and CLN2and in conert with KSS1 promotes signal transduction. Proc. Natl. Acad.Sci. USA. 88, 9392-9396Engfeldt, P., Hellmer, J., Wahrenberg, H., and Arner, P. (1988) Effects of insulinon adrenoceptor binding and the rate of catecholamine-induced lipolysisin isolated human fat cells. J. Biol. Chem. 263, 15553-15560Epstein, P., Boschero, A.C., Atwater, I., Xingang, C., and Overbeek, P.A. (1992)Expression of yeast hexokinase in pancreatic b cells of transgenic micereduces blood glucose, enhances insulin sectetion, and decreasediabetes. Proc. Natl. Acad. Sci. USA 89, 12038-12042Erikson, E., and Mailer, J. L. (1986) Purification and characterization of a proteinkinase from Xenopus eggs highly specific for ribosomal protein S6. J. Biol.Chem. 261, 350-355Erikson, E., and Mailer, J.L. (1988) Second Messengers Phosphoproteins. 12,135-143Erikson, E., and Mailer, J.L. (1989) In vivo phosphorylation and activation ofribosomal protein S6 kinases during xenopus oocyte maturation. J. Biol.Chem. 264, 13711-13717Erikson, E., and Mailer, J.L. (1991) Purification and characterization of ribosomalprotein S6 kinase I from xenopus eggs. J. Biol. Chem. 266, 5249-5255Erikson, R.L. (1991) Structure, expresion, and regulation of protein kinasesinvolved in the phosphorylation of ribosomal protein S6. J. Biol. Chem.266, 6006-6010Erikson, R.L., and Mailer, J.L. (1991) Purification and characterization ofribosomal protein S6 kinase I from Xenopus eggs. J. Biol. Chem. 266,5249-5255Eriksson, J.W., Lonnroth, P., and Smith, U. (1992) Vanadate increases cellsurface insulin binding and improves insulin sensitivity in both normal andinsulin resistant rat adipocytes. Diabetol. 35, 510-515Eriksson, J.W., Lonnroth, P., and Smith, U. (1992) vanadate increases cellsurface insulin binding and improves insulin sensitivity in both normal andinsulin-resistant rat adipocytes. Diabetol. 35, 510-515Errede, B., Gartner, A., Zhou, Z., Nasmyth, K., and Ammerer, G. (1993) MAPkinase-related FUS3 from S. cerevisiae is activated by STE7 in vitro.Nature 362, 261-264266Espinal, J. (1989) Understanding Insulin Action : Principles and MolecularMechanisms. Chichester : Ellis Norwood Ltd.Ettehadieh, E., Sanghera, J.S., Pelech, S.L., Hess-Bienz, D., Watts, J., Shastri,N., and Aebersold„P_ (1992) Tyrosyl phosphorylation an activation ofMAP kinases by p56 10c. Sciences. 255, 853-855Evans, S.W., and Farrar, W.L. (1987) Interleukin 2 and diacyiglycerol stimulatephosphorylation of 40S ribosomal S6 protein. J. Biol. Chem. 262, 4624-4630Eychene, A., Barnier, J.V., Dezelee, P., Marx, M., Laugier, D., Calogeraki, I., andCalothy, G. (1992a) Quail neuroretina c-Rmil(Braf) proto-oncognencDNAs encode two proteins of 93.5 and 95 kDa resulting from alternativesplicing. Oncogene 7, 1315-1323Eychene, A., Barnier, J.V., Apiou, F., Dutrillaux, B., and Calothy, G. (1992b)Chromosomal assignment of two human B-raf(Rmil) proto-oncogene loci:B-raf-1 encoding the p94braf/Rmil and B-raf-2, a processed pseudogene.Oncogene. 7, 1657-1660Fagin, J.A., lkejiri, K., and Levin, S.R. (1987) Insulinotropic effects of vanadate.Diabetes 36, 1448-52Fantus, G.I., Kadota, S., Deragon, G., Foster, B., and Posner, B.I. (1989)Pervanadate (peroxides of vanadate) mimics insulin action in ratadipocytes via activation of the insulin receptor tyrosine kinase. Biochem.28, 8864-8871Faure, R., Baquiran, G., Bergeron, J.J.M., and Posner, B.I. (1992) Thedephosphorylation of insulin epidermal growth factor receptors. J. Biol.Chem. 267, 11215-11221Faustman, D., Li, X., Lin, H.Y., Fu, Y., Eisenhart, G., Avruch, J., and Guo, J.(1991) Linkage of faulty major histocompatibility complex class I toautoimmune diabetes. Science 254, 1756-1761Feig, L. A., and Cooper, G.M. (1988) Inhibition of NIH3T3 cell proliferation by amutant ras protein with perferential affinity for GDP. MoI. Cell. Biol. 8,3235-3241Feig, L.A. (1993) The many roads that lead to ras. Science 260, 767-768Ferrari, S., Bandi, H.R., Hofsteenge, J., Bussian, B.M., and Thomas, G. (1991)Mitogen-activated 70 K S6 kinase. J. Biol. Chem. 266, 22770-22775Ferrari, S., Bannwarth, W., Morley, S.J., Totty, N.F., and Thomas, G. (1992)Activation of p70s6k is associated with phosphorylation of four clusteredsites displaying Ser/Thr-pro motifs. Proc. Natl. Acad. Sci. USA 89, 7282-72861 90 Identification of a 42-kDa • hosprotein as a serine(threonine) protein kinase by renaturation.Biol. 10, 3020-3026.10 14.11 hot roso e267Filhol, 0., Baudier, J., Delphin, C., Lou-Mackenbach, P., Chambaz, E.M., andCochet, C. (1992) Casein kinase II and the tumor suppressor protein p53associate in a molecular complex that is negatively regulated upon p53phosphorylation. J. Biol. Chem. 267, 90577-20583Fingar, D.C., Hausdorff, S.F., Blenis, J., and Birnbaum, M.J. (1993) Dissociationof pp70 ribosomal protein S6 kinase from insulin-stimulated glucosetransport in 3T3-L1 adipocytes. J. Biol. Chem. 268, 3005-3008Fischer, L.J. (1985) Drugs and chemicals that produce diabetes. Trends.Pharmacol. Sci. 6, 72-75Flier, J.S., and Matsouka, P. (1989) Insulin action and the glucose transportergene. In: Draznin, B., Melmed, S., and LeRoith, D., eds., Insulin Action.New York : Alan R. Liss, Inc., pp175-182Flores-Riveros, J.R., McLenithan, J.C, Ezaki, 0., and Lane, M.D. (1993) Insulindown-regulates expression of the insulin-responsive glucose transporter(GLUT4) gene : efects on transcriptional and mRNA turnover. Proc. Natl.Acad. Sci. USA. 90, 512-516Flotow, H., and Roach, P.J. (1991) Substrate recognition determinants of themitogen-activated 70K S6 kinase from rat liver. J. Biol. Chem. 266, 3724-3727Folli, F., Saad, M.J.A., Backer, J.M., and Kahn, C.R. (1992) Insulin stimulation ofphosphatidylinosital 3-kinase activity and association with insulin receptorsubstrate 1 in liver and muscle of the intact rat. J. Biol. Chem. 267, 22171-22177Formisano, P., Sohn, K.J., Miele, C., Finizio, B.D., Petruzziello, A., Riccardi, G.,Behuinot, L., and Beguinot, F. (1993) Mutation in a conserved motif nextto the insulin receptor key autophosphorylation sites de-regulates kinaseactivity and impairs insulin action. J. Biol. Chem. 268, 5241-5248Fortini, M.E., Simon, M.A., and Rubin, G.M. (1992) Signalling by the sevenlessprotein tyrosine kinase is mimicked by ras 1 activation. Nature 355, 559-561Frattali, A.L., Treadway, J.L., and Pessin, J.E. (1991) Evidence supporting apassive role for the insulin receptor transmembrane domain in insulin-dependent signal transduction. J. Biol. Chem. 266, 9829-9834Frattali, A.L., Treadway, J.L., and Pessin, J.E. (1992) Transmembrane signalingby the human insulin receptor kinase. J. Biol. Chem. 267, 19521-19528Freed, E., and Hunter, T. (1992) A 41-kDa protein is a potential substrate for thep210-bcr-abl protein tyrosine kinase in chronic myelogenous leukemiacells. Mol. Cell Biol. 12, 1312-1323- •^I D.S.Downes D.L. Frisell W.R.,Caro, J.F., and Dohm, G.L. (1991) Immunolocalization of glucosetransporter GLUT4 within human skeletal muscle. Diabetes 40, 150-154II • - • I I -^•268Fry, M.J. (1992) Defining a new GAP family. Curr. Biol. 2, 78-80Fukumoto, H., Kayano, T., Buse, J.B., Edwards, Y., Pilch, P.F., Bell, G.I., Seino,S. (1989) J. Biol. Chem. 264, 7776-7779Funasaka, Y., Boulton, T. Cobb, M., Yarden, Y., Fan, B., Lyman, S.D., Williams,D.E., Anderson, Zakut, R., Mishima, Y., and Halaban, R. (1992) c-kit-kinase induces a cascade of protein tyrosine phosphorylation in normalhuman melanocytes in response to mast cell growth factor and stimulatesmitogen-activated protein kinase but is down-regulated in melanomas.Mol. Biol. Cell. 3, 197-209Gale, N.W., Kaplan, S., Lowenstein, E.J., Schlessinger, J., and Bar-Sagi, D.(1993) Grb2 mediates the EGF-dependent activation of guaninenucleotide exchange on ras. Nature 363, 88-92Gallego, C., Gupta, S., Heasley, L.E., Qian, N.X., and Johnson, G. L. (1992)Mitogen-activated protein kinase activation resulting from selectiveoncogene expression in NIH 3T3 and rat 1 a cells. Proc. Natl. Acad. Sci.USA 89, 7355-7359Gartner, A., Nasmyth, K., and Ammerer, G. (1992) Signal transduction insaccharomyces cerevisiae requires tyrosine and threoninephosphorylation of FUSS and KSS1. Genes and Devel. 6, 1280-1292Garvey, W.T., Olefsky, J.M., Matthaei, S., Marshall, S. (1987) Glucose andinsulin coregulate the glucose transport system in primary culturdadipocytes. J. Biol. Chem. 162, 189-197Garvey, W.T., Maianu, L., Hancock, J.A., Golichowski, A.M., and Baron, A.(1992) Gene expression of GLUT4 in skeletal muscle from insulin-resistant patients with obesity, IGT, GDM, and NIDDM. Diabetes 41, 465-475Gaul, U., Mardon, G., and Rubin, G.M. (1992) A putative ras GTPase activatingprotein acts as a negative regulator of signaling by the sevenless receptortyrosine kinase. Cell 68, 1007-1019Gauthier-Rouviere, C., Basset, M., Blanchard, J.M., Cavadore, J.C., Fernandez,A., and Lamb, N.J.C. (1991) Casein kinase II induces c-fos expression viathe serum response element pathway and p67SRF phosphorylation inliving fibroblasts. EMBO J. 10, 2921-2930Gazzano, H., Kowalski, A., Fehlmann, M., and Van Obberghen, E. (1983) Twodifferent protein kinase activities are associated with the insulin receptor.Biochem. J. 216, 575-582Gentleman, S., Reid, T.W., and Martensen, T.M. (1987) Vanadate stimulation ofphosphotyrosine protein levels in quiescent Nakano mouse lens cells.Exp. Eye Res. 44, 587-594 269Gerich, J.E. (1988) Glucose counterregulation and its impact on diabetesmellitus. Diabetes 37, 1608-17Gherzi, R., Caratti, C., Andraghetti, G., Bertolini, S., Montemurro, A., Sesti, G.,and Cordera, R. (1988) Direct modulation of insulin receptor proteintyrosine kinase by vanadate and anti-insulin receptor monoclonalantibodies. Biochem. Biophys. Res. Comm. 152, 1474-1480Ghosh, A., Ronner, P., Cheong, E., Khalid, P., and Matschinsky, F.M. (1991).Rhte role of ATP and free ADP in metabolic coupling during fuel-stimulated insulin release from islet b-cells in the isolated perfused ratpancreas. J. Biol. Chem. 266, 22887-22892Gidh-Jain, M., Takeda, J., Xu, L.Z., Lange, A.J., Nionnet, N., Stoffel, M., Froguel,P., Velho, G., Sun, F., Cohen, D., Patel, P., Lo, Y.M., Hattersley, A.T., etal., (1993) Glucokinase mutations associated with non-insulin-dependent(type 2) diabetes mellitus have decreased enzymatic activity: implicationsfro structure/function relationship. Proc. Nat. Acad. Sci. USA. 90, 1932-1036Gil, J., Miralpeix, M., Carreras, J., and Bartrons, R. (1988) Insulin like effects ofvanadate on glucokinase activity and fructose 2,6-bisphosphate leveiss inthe liver of diabetic rats. J. Biol. Chem. 263, 1868-1871Gille, H., Sharrocks, D., and Shaw, P.E. (1992) Phosphorylation of transcriptionfactor p62TCF by MAP kinase stimulates ternary complex formation at c-fox promoter. Nature 358, 414-417Gilman. A. G. (1987) G proteins: transducers of receptor generated signal. Ann.Rev. Biochem. 56, 615-649Giorgetti, S., Ballotti, R., Kowalski-Chauvel, A., Tartare, S., and Van Obberghen,E. (1993) The insulin and insulin-like growth factor-1 receptor substrateIRS-1 associates with and activates phosphatidylinositol 3-kinase in vitro.J. Biol. Chem. 268, 7358-7364Giorgino, F., Chen, J., and Smith, R.J. (1992) Changes in tyrosinephosphorylation of insulin receptor and a 170,000 molecular weightnonreceptor protein in vivo in skeletal muscle of streptozotocin-induceddiabetic rats : effects of insulin and glucose. Endocrinol. 130 : 1433-1444Gold, M.R., Sanghera, J. S., Stewart, J., and Pelech, S.L. (1992) Selectiveactivation of p42 MAP kinase in murine B lymphocytes by membraneimmunoglobulin cross-linking. Evidence for protein kinase C-independentand dependent mechanism of activation. Biochem. J. 287, 269-276Goldfine, I.D., and Caro, J.F. (1989) Molecular and cellular biology of the insulinreceptor in diabetes mellitus. In: Draznin, B., Melmed, S., and LeRoith, D.eds. Insulin action. New York: Alan R. Liss, Inc., pp11-22Gomez, N., and Cohen, P. (1991) Dissection of the protein kinase cascade by-173II^•^• • Al I^•^•^- •270Gomez-Cambronero, J., Colasanto, J.M., and Sha'afi, R.I. (1993) Directstimulation by tyrosine phosphorylation of microtubule-associated protein(MAP) kinase activity by granulocyte-macrophage colony-stimulatingfactor in human neutraphils. Biochem. J. 291, 211-217Gomez-Foix, A.M., Rodriguez-Gil, J.E., Fillat, C., Guinovart, J.J., and Bosch, F.(1988) Vanadate raises fructose 2,6-bisphosphate concentrations andactivates glycolysis in rat hepatocytes. Biochem. J. 255, 507-512Gonzalez, F.A., Raden, D.L., and Davis, R.J. (1991) Identification of substraterecognition determinants for human ERK1 and ERK2 protein kinases. J.Biol. Chem. 266, 22159-22163Gordon, J. (1991) Use of vanadate as protein phosphotyrosine phosphataseinhibitor. Methods Enzymol. 201, 477-482Goto, Y., Kida, K., Kaino, Y., Inoue, T., Ikeuchi, M., Miyagawa, T., and Matsuda,H. (1988) Insulin action on glucose uptake by soleus muscles ofnonobese diabetic mice and streptozotocin diabetic rats. Metabolism 37,74-78Gotoh, Y., Nishida, E., Yamashita, T., Hoshi, M., Kawakami, M., and Sakai, H.(1990a) Microtubule-associated protein kinase activated by nerve growthfactor and epidermal growth factor in PC12 cells. Eur. J. Biochem. 193,661-669Gotoh, Y., Nishida, E., and Sakai, H. (1990b) Okadaic acid activatesmicrotubule-associated protein kinase inquiescent fibroblastic cells. Eur.J. Biochem. 193, 671-674Gotoh, Y., Moriyama, K., Matsuda, S., Okumura, E., Kishimoto, T., Kawasaki, H.,Suzuki, K., Yahara, I., Sakai, H., and Nishida, E. (1991a) Xenopus Mphase MAP kinase, isolation of its cDNA and activation by MPF. EMBO J.10, 2661-2668Gotoh, Y., Nishida, E., Hoshi, M., and Sakai, H. (1991b) Activation ofmicrotubule-associated protein kinase by microtubule disruption inquiescent rat 3Y1 cells. Exptl. Cell Res. 193, 161-166Gotoh, Y., Nishida, E., Matsuda, S., Shiina, N., Kosako, H., Shiokawa, K.,Akiyama, T., Ohta, K., and Sakai, H. (1991c) In vitro effects onmicrotubule dynamics of purified Xenopus M phase activated MAP kinase.Nature 349, 251-254Gottschalk, W. K. (1991) The pathway mediating insulin's effects on pyruvatedehydrogenase pypasses the insulin receptor tyrosine kinase. J. Biol.Chem. 266, 8814-8819Grand, R.J.A., and Owen, D. (1991) The biochemistry of ras p21. Biochem. J.279, 609-631Graves. P.R.. Haas. D.W.. Hagedorn, C.H., DePaoli-Roach, A.A., and P.Roach. (1993) Molecular cloning, expression, and characterization of a 49-kDacasein kinase I isoform from rat testis. J. Biol. Chem. 268, 6394-6401271Green, A. (1986) The insulin-like effect of sodium vanadate on adipocyte glucosetransport is mediated at a post-insulin-receptor level. Biochem. J. 238,663-669Gregory, J.S., Boulton, G.G. Sang, B.C., and Cobb, M.H. (1989) An insulin-stimulated ribosomal protein S6 kinase from rabbit liver. J. Biol. Chem.264, 18397-18401Gresser, M.J., and Tacey, A.S. (1990) Vanadates as phosphate analogs inbiochemistry. In: Chasteen, N.D., ed. Vanadium in biological systems.Amsterdam: Kluwer Academic Publishers, pp63-79Grinstein, S., Furuya, W., Lu, D.J., and Mills, G. (1990) Vanadate stimulatesoxygen consumption and tyrosine phosphorylation in electropermeabilizedhuman neutrophils. J. Biol. Chem. 265, 318-327Grinstein, S. and Furuya, W. (1992) Chemoattractant-induced tyrosinephosphorylation and activation of microtubule-associated protein kinaseinhuman neutrophils. J. Biol. Chem. 267, 18122-18125Grove, J.R., Banergee, P., Balasubramanyam, A., Coffer, P.J., Price, d.J.,Avruch, J., and Woodgett, J.R. (1991) Clonning and expression of twohuman p70 S6 kinase polypeptides differing only at their amino termini.Mol. Cell. Biol. 11, 5541-5550Grunberger, G., and Levy, J. (1990) Diacylglycerols modulate phosphorylation ofthe insulin receptor form human mononuclear cells. Eur. J. Bochem. 187,191-198Guellec, R.L., Couturier, A., Guellec, K.L, Paris, J., Fur, N.L., and Philippe, M.(1991) Xenopus c-raf proto-oncogene: cloning and expression duringoogenesis and early development. Biol. Cell. 72, 39-45Gulbins, E., Coggeshall, K.M., Baier, G., Katzav, S., Burn, P, and Altman, A.(1993) Tyrosine kinase-stimulated guanine nucleotide exchange activity ofvav in T cell activation. Science 260, 822-825Gupta, S.K., Gallego, C., Johnson, G.L., and Heasley, L.E. (1992) MAP kinase isconstitutively activated in gipw and src transformed rat 1 a fibroblasts. J.Biol. Chem. 267, 7987-7990Gustafson, T.A., and Rutters, W.J. (1990) The cysteine-rich domains of theinsulin and insulin-like growth factor I receptors are primary determinantsof hormone binding specificity. J. Biol. Chem. 265, 18663-18667Guy, G.R., Chua, S.P., Wong, N.S., Ng, S.B., and Tan, Y.H. (1991) Interleukin 1and tumor necrosis factor activate common multiple protein kinases inhuman fibroblasts. J. Biol. Chem. 266, 14343-14352Hadari, Y.R., Tzahar, E., Nadiv, 0., Rothenberg, P., Roberts, C.T., LeRoith, D.,Yarden, Y., and 7ink,_K4t.p92) Insulin and insulinomimetic agents induce.acitvation of phosphatidylinositol 3'-kinase upon its association with pp185(IRS-1) in intact rat livers. J. Biol. Chem. 267, 17483-17486272Hayashi, H., Nishioka, Y., Kamohara, S., Kana, F., Ishii, K., Fukui, Y., Shibasaki,F., Takenawa, T., Kido, H., Katsunuma, N., and Ebina, Y. (1993) The a-.^_ I^•^SIG 1111. • I . 11111111^• ,Hainque, B., Guerre-Millo, M., Hainault, I., Moustaid, N., Wardzala, L.J., andLavau, M. (1990) Long term regulation of glucose transporters by insulinin mature 3T3-F442A adipose cells. J. Biol. Chem. 265, 7982-7986Hall, F.L., and Vulliet, P.R. (1991) Proline-directed protein phosphorylation andcell cycle regulation. Curr. Opinion Cell Biol. 3, 176-184Hanekom, C., Nel, A., Gittinger, C., Rheeder, A., and Landreth, G. (1989)Complexing of the CD-22 subunit by a monoclonal antibody activates amicrotubule-associated protein 2 serine kinase in Jurkat cells. Biochem. J.262, 449-456Hansson, A. and Ingelman-Sundberg, M. (1985) Insulin-stimulated proteinsynthesis in adipocytes: enhanced rate of initiation associated withincreased phosphorylation of ribosomal protein S6. Eur. J. Biochem. 151,97-100Hari, J., and Roth, R.A. (1987) Defective internalization of insulin and its receptorin cells expressing mutated insulin receptors lacking kinase activity. J.Biol. Chem. 262, 15341-15344Haring, H., Kirsch, D., Obermaier, B., Ermel., B., and Machicao, F. (1986)Tumor-promoting phorbol esters increase the Km of the ATP-binding siteof the insulin receptor kinase from rat adipocytes. J. Biol. Chem. 261,3869-3875Hashimoto, N., Feener, E.P., Zhang, W.R., and Goldstein, B.J. (1992) Insulinreceptor protein tyrosine phosphatases. J. Biol. Chem. 267, 1 381 1-1 381 4Hathaway, G.M., and Traugh, J. A. (1982) Casein kinases--multipotential proteinkinases. Curr. Top. Cell. Reg. 21, 101-127Hathaway, G.M., Tuazon, P.T., and Traugh, J.A. (1983) Casein kinase I.Methods Enzymol. 99, 308-317Hattorin, S. Fukuda, M., Yamashita, T., Nakamura, S., Gotoh, Y., and Nishida,E. (1992) Activation of mitogen-activated protein kinase and its activatorby ras in intact cells and in a cell free system. J. Biol. Chem. 267, 20346-20351Haubruck, H., and McCormick, F. (1991) Ras p21: effects and regulation.Biochim Biophys. Act. 1072, 215-229Hayashi, H., Kamohara, S., Nishioka, Y., Kanai, F., Miyake, N., Fukui, Y.,Shibasaki, F., Takenawa, T., and Ebina, Y. (1992) Insulin treatmentstimulates the tyrosine phosphorylation of the a-type 85 kDa subunit ofphosphatydylinositol 3-kinase in vivo. J. Biol. Chem. 267, 22575-22580tyrosine 368, 580, and 607 by the insulin receptor. J. Biol. Chem. 268,7107-7117273Haystead, T.A., Weiel, J.E., Litchfield, D.W., Tsukitani, Y., Fischer, E.H., andKrebs, E.G. (1990) Okadaic acid mimics the action of insulin in stimulatingprotein kinase activity in isolated adipocytes. J. Biol. Chem. 265, 16571-16580Hecht, L.B., and Straus, D.S. (1988) insulin-stimulated protein kinase activity inrat skeletal muscle tht phosphorylates ribosomal protein S6. Biochem.Biophys. Res. Comm. 152, 1200-1206Hedo, J.A., and Simpson, I.A. (1984) Internalization of insulin receptors intheisolated rat adipose cell. J. Biol. Chem. 259, 11083-11089Heffetz, D., Bushkin, I., Dror, R., and Zick, Y. (1990) The insulinomimetic agentsH202 and vanadate stimulate protein tyrosine phosphorylation in intactcells. J. Biol. Chem. 265, 2896-2902Hei, Y.J., McNeill, J.H., Sanghera, J.S., Diamond, J., Bryer-Ash, M., and Pelech,S.L.(1993) Characterization of insulin stimulated seryl/threonyl proteinkinases in rat skeletal muscle. J. Biol. Chem. 268, 13203-13213.Heidenreich, K.a., and Toledo, S.P. (1989) Insulin receptors mediate growtheffects in cultured fetal neurons. I. Rapid stimulation of protein synthesis.Endocrinol. 125, 1451-1457Heidenreich, K.A., Toledo, S.P., Brunton, L.L., Watson, M.J., Daniel-Issakani, S.,and Strulovici, B. (1990) Insulin stimulates the activity of a novel proteinkinase C, PKC-e, in cultured fetal chick neurons. J. Biol. Chem. 265,15076-15082Heller-Harrison, R.A., and Czech, M.P. (1991) Enhanced casein kinase II activity .in COS-1 cells upon overexpression of either its catalytic or noncatalyticsubunit. J. Biol. Chem. 266, 14435-14439Hems, D.A., and Whitton, P.D. (1980) Control of hepatic glycogenolysis. Physiol.Rev. 60, 1-50Her, J.H., Wu, J., Sturgill, T.W., and Weber, M.J. (1991) Sequence of p42/MAPkinase, a serine/threoine kinase regulated by tyrosine phosphorylation.Nucleic Acids. Res. 19, 3743Hers, H. G. (1976) The control of glycogen metabolism in the liver. Annu. Rev.Biochem. 45, 167-189Hershey, J.W. (1989) Protein phosphorylation controls translation rates. J. Biol.Chem. 264, 20823-20826Hesketh, J.E., and Campbell, G. P. (1987) Effects of insulin, pertussis toxin, andcholera toxin on protein synthesis and diacylglycerol production in 3T3fibroblasts: evidence for a G protein mediated activation of phospholipaseC in the insulin signal mechanism. Bioscience Report. 7, 533-541 .274Heyliger, C.E., Tahiliani, A.G., and McNeill, J.H. (1985) Effect of vanadate onelevated blood glucose and depressed cardiac performance of diabeticrats. Science 227, 1474-1477Hitman, G.A., and Niven, M.J. (1989) Genes and diabetes mellutus. Brit. Med.Bull. 45, 191-205Hoekstra, M.J., Liskay, R.M., Ou, A.C., DeMaggio, A.J., Burbee, D.G., andHeffron, F. (1991) HRR25, a putative protein kinase from budding yeast :Association with repair of damaged DNA. Science 253, 1031-1034Holz, G.G.,and Habener, J.F. (1992) Signal transduction crosstalk in theendocrine system : pancreatic b-cells and the glucose competenceconcept. Trends. Biochem. Sci. 17, 388-393Home, P.D., Thow, J.C., Tunbridge, F.K.E. (1989) Insulin treatment : a decade ofchange. Brit. Med. Bull. 45, 92-110Hori, H., and Oka, T. (1980) Vanadate enhances the stimulatory action of insulinon DNA synthesis in cultured mouse mammary gland. Biochem. Biophys.Act. 610, 235-240Hoshi, M., Nishida, E., and Sakai, H. (1988) Activation of a Ca-inhibitable proteinkinase that phosphorylates microtubule-associated protein 2 in vitro bygrowth factors, phorbol esters, and serum in quiescent cultured humanfibroblasts. J. Biol. Chem. 263, 5396-5401Hoshi, M., Nishida, E., Nagaki, M., Gotoh, Y., and Sakai, H. (1990) Activation ofa serine/threonine kinase that phosphorylates microtubule-asociatedprotein 1B in vitro by growth factors and phorbol esters in quiescent ratfibroblastic cells. Eur. J. Biochem. 193, 513-519Houslay, M.D., and Siddle, K. (1989) Molecular basis of insulin receptor function.Brit. Med. Bull. 45, 264-284Houslay, M.D. (1990) Insulin and its interaction with G proteins. In: lyengar, R.,and Birnbaumer, L., eds. G Proteins. California: Academic Press, pp521-553Howe, L.R., Leevers, S.J., Gomez, N., Nakielny, S., Cohen, P., and Marshall,C.J. (1992) Activation of the MAP kinase pathway by the protein kinaseraf. Cell 71, 335-342Howell, S.L., and Bird, G.S. J. (1989) Biosynthesis and secretion of insulin. Brit.Med. Bull. 45, 19-36Hresko, R.C., Bernier, M., Hoffman, R.D., Flores-Riveros, J.R., Liao, K., Laird,D.M., and Lane, M.D. (1988) Proc. Natl. Acad. Sci. USA 85, 8835-8839Hresko, R.C., Hoffman, R.D., Flores-Reveros, J., and Lane, M.D. (1990) Insulinreceptor tyrosine kinase-catalyzed phosphorylation of 422(aP2) protein. J.Biol. Chem. 265, 21075-21085 275Hu, E., and Rubbin, C.S. (1990) Expression of wild-type and mutated forms ofcatalytic subunit of C. elegans casein kinase II in E. Coli. J. Biol. Chem.265, 20609-20615Hu, E, and Rubin, C.S. (1991) Casein kinase II from C. elegans. J. Biol. Chem.266, 19795-19802Hughes, K., Nikolakaki, E., Plyte, S.E., Totty, N.F., and Woodgett, J. R. (1993)Modulation of the glycogen synthase-3 family by tyrosine phosphorylation.EMBO J. 12, 803-808Hunter, T. (1985) Oncogenes and growth control. Trends Biochem. Sci. 5, 215-225Hunter, T., and Cooper, J.A. (1986) Viral oncogenes and tyrosinephosphorylation. Enzymes 17, 191-246Hutchison, K.A., Stancato, L.F., Jove, R., and Prattt, W.B. (1992) The protein-protein complex between pp60v-src and hsp90 is stabilized by molybdate,vanadate, tungstate, and an endogenous cytosolic metal. J. Biol. Chem.267, 13952-13957Imam, E., Kadowaki, H., Kadowaki, T. (wawa, N., Watarai, T., Kawamori, R.,Kamada, T., and Taylor, S.I. (1991) Two patients with insulin resistancedue to decreaesd levels of insulin-receptor mRNA. Diabetes 40, 548-57Inoue, G., Kuzuya, H., Hayashi, T., Okamoto, M., Yoshimasa, Y., Kosaki, A.,Kono, S., Maeda, I., Kubota, M., and Imura, H. (1993) Effects of ML-9 oninsulin stimulation of glucose transport in 3T3-L1 adipocytes. J. Biol.Chem. 268, 5242-5278Ishizuka, T., Cooper, D.R., Arnold, T., Hernandez, H., and Farese, R.V. (1991)Downregulation of protein kinase C and insulin-stimulated 2-deoxyglucoseuptake in rat adipocytes by phorbol esters, glucose, and insulin. Diabetes.40, 1274-1281Issinger, 0.G., and Beier, H. (1980) Comparison of phosphorylation of ribosomalproteins from HeLa and Krebs ll ascites-tumor cells by cAPM-dependentand cGMP-dependent protein kinases. Biochem. J. 185, 89-99ltkes, A.V., Imamova, L.R., Alexandrova, N.M., Favorova, 0.0., and Kisselev,L.L. (1990) Expression of c-myc gene in human ovary carcinoma cellstrated with vanadate. Exp. Cell. Res. 188, 169-171Itoh, T., Kaibuch, K., Masuda, T., Yamamoto, T., Matsuura, Y., Maeda, A.,Shimizu, K., and Takai, Y. (1993a) The post-transtlational processing ofras p21 is critical for its stimulation of mitogen-activated protein kinase. J.Biol. Chem. 268, 3025-3028Itoh, T., Kaibuchi, K., Masuda, T., Yamamoto, T., Matsuura, Y., Maeda, A.,Shimizu, K., and Takai, Y. (1993b) A protein factor for ras p21-dependentactivation of mitogen-activated protein kinase through MAP kinase kinase. Proc. Natl. Acad. Sci, USA. 90, 975-979276Izumi, T., Tamemoto, H., Nagao, M., Kadowaki, Y., Takaku, F., and Kasuga, M.(1991) Insulin and platelet-derived grwoth factor stimulate phosphorylationof the c-raf product at serine and threonine residues in intact cells. J. Biol.Chem. 266, 7933-7939Jackson, T., Salhanick, A.K., Sparks, J.D., Sparks, C.E., Bolognino, M., andAmatruda, J.M. (1988) Insulin-mimetic effects of vanadate in primarycultures of rat hepatocytes. Diabetes 37, 1234-40Jacobs, S., Hazum, E., and Cuatrecasas, P. (1980) The subunit structure of ratliver insulin receptor. J. Biol. Chem. 255, 6937-6942Jacobs, S., and Cuatrecasas, P. (1983) Insulin Receptors. Ann. Rev. Pharmacol.Toxicol. 23, 461-479Jacobs, S., and Cuatrecasas, P. (1985) Purification of insulin receptor fromhuman placenta. Methods Enzymol. 109, 399-405Jeno, P., Ballou, L.M., Novak-Hofer I and Thomas G (1988) Identification andcharacterization of a mitogen-activated S6 kinase. Proc. Natl. Acad. Sci.USA. 85, 406-410Jeno, P., Jaggi, N., Luther, H., Siegmann, M., and Thomas, G. (1989)Purification and characterization of a 40S ribosomal protein S6 kinaseform vanadate-stimulated Swiss 3T3 cells. J. Biol. Chem. 264, 1293-1297Jessus, C., Rime, H., Haccard, 0., Van Lint, J., Goris, J., Merlevede, W., andOzon, R. (1991) Tyrosine phosphorylation of p34cdc2 and p42 duringmeiotic maturation of xenopus oocyte. Development. 111, 813-820Johnson, L.N. (1992) Glycogen phosphorylase: control by phosphorylation andallosteric effectors. FESEB J. 6, 2274-2282Johnson, L.N., and Barford, D. (1992) Glycogen phosphorylase:the structuralbasis of the allosteric response and comparison with other allostericproteins. J. Biol. Chem. 265, 2409-2412Jones, S. W., Erikson, E., Blenis, J., Mailer, J., and Erikson, R.L. (1988) AXenopus ribosomal protein S6 kinase has two apparent kinase domainsthat are each similar to distinct protein kinases. Proc. Natl. Acad. Sci.USA. 85, 3377-3381Joost, H.G., Weber, T.M., Cushman, S.W., and Simpson, I.A. (1987) Activity andphosphorylation state of glucose transporters in plasma membrane frominsulin-, isoproterenol, and phorbol ester-treated rat adipose cell. J. Biol.Chem. 262, 11261-11267Joost, H.G., Weber, T.M., Simpson, I.A., and Cushman, S.W. (1988)Phosphorylation state and intrinsic activity of glucose transporters inplasma membranes from insulin-, isoproternol-, and phorbol ester-treatedrat adipocytes. In: Goren, H.J., Hollenberg, M.D., and Roncari, D. A., eds.,Insulin Action and Diabetes. New York: Raven Press, pp157-162 277Kadota, S., Fantus, I.G., Hersh, B., and Posner, B. I. (1986) Vanadatestimulation of IGF binding to rat adipocytes. Biochem. Biophys, Res.Comm. 138, 174-178Kadota, S., Fantus, I.G., Deragon, G., Guyda, H.J., and Posner, B.I. (1987a)Stimulation of insulin-like growth factor II receptor binding and insulinreceptor kinase activity in rat adipocytes. J. Biol. Chem. 262, 8252-8256Kadota, S., Fantus, I.G., Deragon, G., Hersh, B., Guyda, H.J., and Posner, B.I.(1987b) A novel and potent insulin-mimetic agent which activates theinsulin receptor kinase. Biochem. Biophys. Res. Comm. 147, 259-266Kadowaki, T., Kasuga, M., Akanuma, Y., Ezaki, 0., Takaku, F. (1984)Decreased autophosphorylation of the insulin receptor kinase instreptozotocin-induced diabetic rats. J. Biol. Chem. 259, 14208-14216Kadowaki, H., Kadowaki, T., Cama, A., Marcus-Samuels, B., Rovira, A., Bevines,C.L., and Taylor, S.I. (1990) Mutagenesis of lysine 460 in the humaninsulin receptor. J. Biol. Chem. 265, 21285-21296Kahn, C.R., Sethu, S., Reddy, K., Shoelson, S., Goren, H.J., and While, M. F.(1988) Regulation of the insulin receptor by muti-site phosphorylation. In:Goren, H.J., Hollenberg, M.D., and Roncari, D.A.K., eds., Insulin Actionand Diabetes. New York: Raven Press, pp43-52Kahn, C.R. and Goldstein, B.J. (1989) Molecular defects in insulin action.Science 245 : 7Kahn, B.B., and Flier, J.S. (1990) Regulation of glucose transporter geneexpression in vitro and in vivo. Diabetes Care 13, 548-564Kahn, C., Seuwen, K., Meloche, S., and Pouyssegur, J. (1992) Coordinatebiphasic activation of p44 mitogen-activated protein kinase and S6 kinaseby growth factors in hamster fibroblasts. J. Biol. Chem. 267, 13369-13375Kamata, T., Kathuria, S. and Fujita-Yamaguchi, Y. (1987) Insulin stimulates theposphorylation levelof v-Ha-ras protein in membrane fraction. Biochem.Biophys. Res. Comm. 144, 19-25Kanthasamy, A., Sekar, N., and Govindasamy, S. (1988) Vanadate substitutesinsulin role in chronic experimental diabetes. Ind. J. Exp. Biol. 26, 778-780Kasuga, M., Karlsson, F.A., and Kahn, C.R. (1982a) Insulin stimulates thephosphorylation of the 95000-dalton subunit of its own receptor. Science215, 185-186Kasuga, M., Zick, Y., Blith, D.L., Karlsson, F.A., Haring, H. U., and Kahn, C. R.(1982b) Insulin stimulation of phosphorylation of the b-subunit of theinsulin receptor. J. Biol. Chem. 257, 9891-9894Kasuga, M., Zick, Y., Blith, D.L., Crettaz, M., and Kahn, C.R. (1982c) Insulin^stimulates tyrosine phosphorylation of the insulin mnsptnr in s es11 -frss system. Nature 298, 667-669Kawakami, M., Nishida, E., Tobe, K., Hoshi, M., Kadowaki, T., Kasuga, M., andSakai, H. (1991) Activation of MAP kinase and enhanced phosphorylationof the 350-kDa protein by mitogenic stimuli in quiescent Balb/c 3T3 cells.Exptl. Cell. Res. 193, 120-126Kelada, A.S.M., Macaulay, S.L., and Proietto, J. (1992) Cyclic AMP acutelystimulates translocation of the major insulin-regulatable glucosetransporter GLUT4. J. Biol. Chem. 267, 7021-7025Kellett, G.L., and Barker, E.D. (1989) The effect of vanadate on glucosetransport and metabolism in rat small intestine. Biochem. Biophys. Act.979, 311-35Kelly, K.L., and Ruderman, N.B. (1993) Insulin-stimulated phosphatidylinositol 3-kinase. J. Biol. Chem. 268, 3291-3298Khan, M., Baquiran, G., Brule, C., Burgess, J., Foster, B., Bergero, J.M., andPosner, B.I. (1989) Internalization and activation of the rat liver insulinreceptor kinase in vivo. 264, 12931-12940Kikkawa, U., Mann, S.K.O., Firtel, R.A., and Hunter, T. (1992) Molecular cloningof casein kinase Ila subunit from Dictyostelium discoideum and itsexpression in the life cycle. Mol. Cell. Biol. 12, 5711-5723Kizaka-Kondoh, S., Sato, K., Yamura, K., Nojima, H., and Okayama, H. (1992)Raf-1 protein kinase is an integral component of the oncogenic signalcascade shared by epidermal growth factor and platelet derived growthfactor. Mol. Cell. Biol. 12, 5078-5086Klarlund, J.K. (1985) Transformation of cells by an inhibitor of phosphataseacting on phosphotyrosine in proteins. Cell 41, 707-717Klarlund, J.K., and Czech, M.P. (1988) Insulin-like growth factor I and insulinrapidly increase casein kinase II activity in BALB/c 3T3 fibroblasts. J. Biol.Chem. 263, 15872-15875Klarlund J.K, Bradford, A.P., Milla, M.G., and Czech, M.P. (1990). Purification ofa novel insulin-stimulated protein kinase from rat liver. J.Biol. Chem. 265,227-234Klarlund, J.K., Jaspers, S.R., Khalaf, N., Bradford, A.P., Miller, T.B., and Czech,M.P. (1991) An insulin-stimulated kemptide kinase purified from rat liver isdeactivated by phosphatase 2A. J. Biol. Chem. 266, 4052-4055Klein, H.H., Freidenberg, G.R., Kladde, M., and Olefsky, J.M. (1986) Insulinactivation of insulin receptor tyrosine kinase in intact rat adipocytes. J.Biol. Chem. 261, 4691-4697Klein, H.H., Freidenberg, G.R., Matthaei, S., and Olefsky, J.M. (1987) Insulinreceptor kinase following internalization in isolated rat adipocytes. J. Biol.ChArn. 269, 10557-10564 279Klip, A., and Douen, A. (1989) Role of kinases in insulin stimulation of glucosetransport. Memb. Biol. 111, 1-23Knutson, V.P. (1991) Cellular trafficking and processing of the insulin receptor.FASEB J. 5, 2130-2138Knutson, V.P. (1992) Ligand-independent internalization and recycling of theinsulin receptor. J. Biol. Chem. 267, 931-937Koch, C.A., Anderson, D., Moran, M.F., Ellis, C., and Pawson, T. (1991) SH2and SH3 domains : elements that control interactions of cytoplasmicsignalling proteins. Science 252, 668-674Koivisto, U., Martinez-Valdez, H., Bilan, P.J., Burdett, E., Ramlal, T. and Klip, A.(1991) Differential regulation of the GLUT1 and GLUT4 glucose transportsystems by glucose and insulin in L6 muscle cells in cultre. J. Biol. Chem.266, 2615-2621Kolch, W., Heidecker, G., Lloyd, P., and Rapp, U.R. (1991) Raf-1 protein kinaseis required for growth of induced NIH/3T3 cells. Nature 349, 426-428Korn, L.J., Siebel, C.W., McCormick, F., and Roth, R. A. (1987) Ras p21 as apotential mediator of insulin action in Xenopus oocytes. Science 236, 840-843Kosako, H., Gotoh, Y., Natsuda, S., Ishikawa, M., and Nishida, E. (1992)Xenopus MAP kinase activator is a serine/threonine/tyrosine kinaseactivated by threonine phosphorylation. EMBO J. 11, 2903-2908Kosako, H., Nishida, E, and Gotoh, Y. (1993) cDNA cloning of MAP kinasekinase reveals kinase cascade pathways in yeasts to vertebrates. EMBOJ. 12, 787-794Kovacina, K.S., Yongezawa, K., Brautigan, D.L., Tonks, N.K., Rapp, U.R. (1990)Insulin activates the kinase activity of the Raf-1 proto-oncogene byincreasing its serine phosphorylation. J. Biol. Chem. 265, 12115-12118Kozka, I.J., Clark, A.E., and Holman, G.D. (1991) Chronic treatment with insulinselectively down-regulates cell-surface GLUT4 glucose transporters in3T3-L1 adipocytes. J. Biol. Chem. 266, 11726-11731Kozma, S.C., Lane, H.A., Ferrari, S., Luther, H., Siegmann, M., and Thomas, G.(1989) A stimulated S6 kinase from rat liver : identity with the mitogenactivated S6 kinase of 3T3 cells. EMBO J. 8, 4125-4132Kozma, S.C., Ferrari, S., Bassand, P., Siegmann, M., Totty, N., and Thomas, G.(1990) Cloning of the mitogen-activated S6 kinase from rat liver revealsan enzyme of the second messenger subfamily. Proc. Natl. Acad. Sci. 87,7365-7369Kozma, S.C., and Thomas, G. (1992) Serine/threonine kinases in thepropagation of the early mitogenic response. Rev. Physol. Biochem.Pharmacol. 119, 123-155280Kozma, S.C., McGlynn, E., Siegmann, M., Reinhard, C., Ferrari, S., andThomas, G. (1993) Active baculovirus recombinant p70s6k and p85s6kproduced as a function of the infectious response. J. Biol. Chem. 268,7134-7138Krebs, E.G. (1986) The enzymology of control by phosphorylation. The Enzymes17, 3-20Krebs, E.G., Eisenman, R.N., Kuenzel, E.A., Litchfielf, D.W., Lozeman, F.J.,Luscher, B., and Sommercorn, J. (1988) Casein kinase II as a potentiallyimportant enzyme concerned with signal transduction. Cold SPringHarbor Symposia on Quantitative Biology. 53,77-84Krebs, E.G. (1989) Role of the cyclic AMP-dependent protein kinase in signaltransduction. J. Am. Med. Assoc. 262, 1815-1818Krieg J., Olivier A.R., and Thomas G. (1988). Analysis of 40S ribosomal proteinS6 phosphorylation during mitogenic response. Methods in Enzymol.164, 575-581.Krolewski, A.S., and Warram, J.H. (1989) current issues in the etiology of type Idiabetes and its complications: an epidemiologic approach. In: Larkins,R., Zimmet, P., and Chisholm, D., eds., Diabetes 1988, Amsterdam:Elsevier Science Publishers, B.V., pp321-324Krupinski, J., Rajaram, R., Lakonishok, M., Benovic, J., and Cerione, R.A>(1988) Insulin-dependent phosphorylation of GTP-binding proteins inphospholipid vesicles. J. Biol. Chem. 263, 12333-12341Kuo, C.J., Chung, J., Fiorentino, D.F., Flanagan, W.M., Blenis, J., and Crabtree,G.R. (1992) Rapamycin selectively inhibits interleukin-2 activation of p70S6 kinas. Nature 358, 70-73Kusari, J., Takata, Y., Hatada, E., Freidenberg, G., Kolterman, 0., and Olefsky,J.M. (1991) Insulin resistance and diabetes due to different mutations inthe tyrosine kinase domain of both insulin receptor gene alleles. J. Biol.Chem. 266, 5260-5267Kyriakis, J.M., and Avruch, J. (1990) pp54 microtubule-associated protein 2kinase. J. Biol. Chem. 265, 17355-17363Kyriakis, J.M., Brautigan, D.L., lngebretsen, T.S., and Avruch, J. (1991) pp54microtubule-associated protein 2-kinase requires both tyrosine andserine/threonine phosphorylation for activity. J. Biol. Chem. 266, 10043-10046Kyriakis, J.M., App, H., Zhang, X.F., Banerjee, P., Brautigan, D L., Rapp, U.R.,and Avruch, J. (1992) Raf-1 activates MAP kinase kinase. Nature 358,417-421L'Allemain, G., Pouyssegur, J., and Weber, M. (1991a) p42/mitogen-activatedpmtPin kinasP as a converging target for different growth factor signallingpathways: use of pertussis toxin as a discrimination factor. Cell. Reg. 2,675-684281L'Allemain, G., Sturgill, T.W., and Weber, M.J. (1991b) Defective regulation ofmitogen-activated protein kinase activity in a 3T3 cell variant mitogenicallynonresponsive to tetradecanoyl phorbol ester. Mol. Cell. Biol. 11, 1002-1008Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of thehead of bacteriophage T4 Nature 227, 680-685Lammers, R., Van Obberghen, E., Ballotti, R., Schlessinger, J., and Ul[rich, A.(1990) Transphosphorylation as a possbile mechanism for insulin andepidermal growth factor receptor activation. J. Biol. Chem. 265, 16886-16890Lane, H.A., Morley, S.J., Doree, M., Kozma, S.C., and Thomas, G. (1992)Identification and early activation of a Xenopus laevis p70s6k followingprogesterone-induced meiotic maturation. EMBO J. 11, 1743-1749Lange-Carter, C.A., Pleiman, C.M., Gardner, A.M., Blumer, K.J., and Johnson,G.L. (1993) A divergence in the MAP kinase regulatory network definedby MEK kinase and raf. Science 260, 315-319Langlois, D., Hinsch, K.D., Saez, J.M., and Begeot, M. (1990) Stimulatory effectof insulin and insulin-like growth factor I on Gi proteins and angiotensin-Il-induced phosphoinositide breakdown in cultured bovine adrenal cells.Endocrinol. 126, 1867-1872Laporte, R.E. (1989) Aetiology and epidemiology of IDDM/NIDDM. In: Larkins,R., Zimmet, P., and Chisholm, D., eds., Diabetes 1988, Amsterdam:Elsevier Science Publishers, B.V., pp1171-1178Lamer, J. (1988a) Insulin-signalling mechanism. Diabetes 37, 262-275Lamer, J. (1988b) Insulin-signalling mechanisms: lessons from the old testamentof glycogen metabolism and the new testament of molecular biology.Diabetes. 37, 262-275Lamer, J., Huang, L.C., Tang, G., Suzuki, S., Schwartz, C.F.W., Romero, G.,Roulidis, Z., Zeller, K., Shen, T.Y., Oswald, A.S., and Luttrell, L. (1988)Insulin meidators: structure and formation. Cold Spring Harbor Symp.Quant. Biol. 53, 965-971Lamer, J., Huang, L.C., Tang, G., Suzuki, S., Schwartz, C.F.W., Romero, L.,Luttrell, L., Zeller, K., Shen, T.Y., and Oswald, A.S. (1989) In: Larkins, R.,Zimmet, P., and Chisholm, D., eds., Diabetes 1988, Amsterdam: ElsevierScience Publishers, B.V., pp133-145Lau, K.H.W., Tanimoto, H., adn Baylink, D.J. (1988) Vanadate stimulates bonecell proliferation and bone collagen synthesis in vitro. Endocrinol. 123,2858-2867Lavan, B.E., Kuhne, M.R., Garner, C.W., Anderson, D., Reedijk, M., Pawson, T.,and Lienhard, G.E. (1992) Association of insulin-elicited phosphotyrosineproteins with src homology 2 domains. J. Biol. Chem. 267, 11631-11636282Lavan, B.E., and Lienhard, G.E. (1993) The insulin-elicited 60-kDaphosphotyrosine protein in rat adipocytes is associated withphosphotidylinositol 3-kinase. J. Biol. Chem. 268, 5921-5928Lavoinne, A., Erikson, E., Mailer, J.L., Price, D.J., Avruch, J. and Cohen, P.(1991) Purification and characterization of the insuin-stimulated proteinkinase from rabbit skeletal muscle; close similarity to S6 kinase II. Eur. J.Biochem. 199, 723-728Lawrenc, J.C., Jr., Hiken, J.F., and James, D.E. (1990) Stimulation of glucosetransporter phosphorylation by okadaic acid in rat adipocytes. J. Biol.Chem. 265, 19768-19776Le Marchand-Brustel, Y., Gremeaux, Ballotti, R., and Obberghen, E. (1985)Insulin receptor tyrosine kinase is defective in skeletal muscle of insulin-resistant obese mice. Science 315, 676-679Le Marchand-Brustel, Y., Ballotti, R., and Van Obberghen, E. (1988) Insulinreceptor kinase activity in states with altered insulin action. In:; Kahn,C.R., and Harrison, L.C., eds., Insulin receptors, part A: Methods for thestudy of structure and function. New York: Alan R. Liss, Inc., pp163-179Le Marchand-Brustel, Y., Ballotti, R. Gremeaux, T., Tanti, J.F., Branderburg, D,and Van Obberghen, E. (1989) Functional labeling of insulin receptorsubunits in live cells. J. Biol. Chem. 264, 21316-21321Le Peuch, C.J., Ballester, R., and Rosen, O.M. (1983) Purified rat brain calcium-and phospholipid-dependent protein kinase phosphorylates ribosomalprotein S6. Proc. Natl. Acad. Sci. USA. 80, 6858-6862Leconte, I., Auzan, C., Debant, A., Rossi, B., Clauser, E. (1992) N-linkedoligosaccharide chains of the insulin receptor b-subunit are essential fortransmembrane signalling. J. Biol. Chem. 267, 1 741 5-1 7423Lee, T.S., Saltsman, K.A., Ohashi, H., and King, G.L. (1989) Activation of protienkinase C by elevation of glucose concentration : proposal for amechanism in the development of diabetic vascular complications. Proc.Natl. Acad. Sci. USA. 86, 5141-5145Lee, R.M., Rapp, U.R., and Blackshear, P.J. (1991) Evidence for one or moreraf-1 kinase activated by insulin and polypeptide growth factors. J. Biol.Chem. 266, 10351-10357Lee, R.M., Cabb, M.H., and Blackshear, P.J. (1992) Evidence that extracellularsignal-regulated kinases are the insulin-activated raf-1 kinase kinases. J.Biol. Chem. 267, 1088-1092Lee, H., Ghose-Dastidar, J., Winawer, S., and Friedman, E. (1993) Signaltransduction through extracellular signal-regulated kinase pp57 blocked indifferentiated cells having low protein kinase cp activity. J. Biol. Chem.268, 5255-5263283Leevers, S.J., and Marshall, C.J. (1992) Activation of extracellular signal-regulated kinase ERK2, by p21ras oncoprotein. EMBO J. 11, 569-574Lefkowitz, R.J. (1992) The subunit story thickens. Nature 358, 372Leighton, B., Cooper, G.J.S., DaCosta, C., and Foot, E. (1991) Peroxiovanadatehave full insulin-like effects on glycogen synthesis in normal and insulinresistant skeletal muscle. Biochem. J. 276, 289-292Leiter, E.H. (1989) The genetics of diabetes susceptibility in mice. FASEB J. 3,2231-2241Lerea, K.M., Tonks, N.K., Krebs, E.G., Fischer, E.H., and Glomset, J.A. (1989)Vanadate and molybdate increase tyrosine phosphorylation in a 50-kDaprotein and stimulate secretion in electropermeabilized platelets.Biochem. 28, 9286-9292LeRoith, D., Lowe, W.L. Jr., and Roberts, C.T. Jr. (1987) Evolution of insulin andinsulin receptors. In : Raizada, M.K., Phillips, M.I., and LeRoith, D., eds.Insulin, Insulin-like Growth Gactors and Their Receptors in the CentralNervous System. NY:Plenum Press, pp107-120LeRoith, D., Lowe, W.L.,Jr., and Roberts, C.T., Jr. (1989) Evolution of insulin andthe insulin receptor. In: Draznin, B., Memed, S., and LeRoith, D., eds.,Insulin Action. New York: Alan R. Liss, Inc., pp1-9Leslie, R.D.G., Lazarus, N.R., and Vergani, D. (1989) Aetiology of insulin-dependent diabetes. Brit. Med. Bull. 45, 58-72Levin, S.R., Aftergood, D.E., Fanous, E., and Whitson, R.H. (1988) Inhibition ofinsulin receptor by vanadate and ouabain. Life Sci. 42, 1889-1896Levine, B.A., Tavare, J.M., Alejos, E., Clack, B., Sayed, N., and Ellis, L. (1991)Autophosphorylation of soluble insulin receptor protein-tyrosine kinases.J. Biol. Chem. 266, 13405-13410Levy, J.R., and Olefsky, J.M. (1990) Receptor-mediated internalization andturnover. In: Jacobs, S., and Cuatrecasas, P., eds. Insulin, Berlin:Springer-Verlag, pp237-266Lewis, R.E., Cao, L., Perregaux, D., and Czech, M.P. (1990a) Thronine 1336 ofthe human insulin receptor is a major target for phosphorylation by proteinkinase C. 29, 1807-1813Lewis, R.E., Wu, G.P., MacDonald, R.G., and Czech, M.P. (1990b) Insulin-sensitive phosphorylation of serine 1293/1294 on the human insulinreceptor by a tightly associated serine kinase. J. Biol. Chem. 265, 947-954Li, P., Wood, K., Mamon, H., Haser, W., and Roberts, T. (1991) Raf-1: A kinasecurrently without a cause but not lacking in effects. Cell. 64, 479-482284Li, B.Q., Kaplan, D., Kung, H.F., adn Kamata, T. (1992) Nerve growth factorstimulation of the ras-guanine nucleotide exchange factor and GAPactivity. Science 256, 1456-1459Li, H., and Roux, S.J. (1992) Casein kinase II protein kinase is bound to lamina-matrix and phosphorylates lamin-like protein in isolated pea nuclei. Proc.Natl. Acad. Sci. USA. 89, 8434-8438Li, N., Batzer, A., Daly, R., Yajnik, V., Skolnik, E., Chardin, P., and Sagi, D.B.(1993) Guanine-nucleotide-releasing factor hSos1 binds to Grb2 and linksreceptor tyrosine kinases to ras signalling. Nature 363, 85-86Lin, W.J., Tuazon, P.T., and Traugh, J.A. (1991) Characterization of the catalyticsubunit of casein kinase II expresed in E. Coli and regulation of activity. J.Biol. Chem. 266, 5664-5669Lin, Y.J., Harady, S., Loten, E.G., Smith, R.M., and Jarett, L. (1992) Directstimulation of immediate-early genes by intranuclear insulin in trypsin-treated H35 hepatoma cells. Proc. Natl.Acad. Sci. USA 89, 9691-9694Lin, L.L., Wartmann, M., Lin, A.Y., Knopf, J.L., Seth, A., and Davis, R.J. (1993)cPLA2 is phosphorylated and activated by MAP kinase. Cell 72, 269-278Litchfield, D.W., Lozeman, F.J., Piening, C., Sommercorn, J., Takio, K., Walsh,K.A., and Krebs, E.G. (1990) Subunit structure of casein kinase II frombovine testis. J. Biol. Chem. 265, 7638-7644Litchfield, D.W., Lscher, B., Lozeman, F.J., Eisenman, R.N., and Krebs, E.G.(1992) Phosphorylation of casein kinase II by p34cdc2 in vitro and atmitosis. J. Biol. Chem. 267, 13943-13951Liu, M.L., Olson, A.L., Moye-Rowley, W.S., Buse, J.B., Bell, G.I., and Pessin,J.E. (1992) Expression and regulation of the human GLUT4/muscle fatfacilitative glucose transporter gene in transgenic mice. J. Biol. Chem.267, 11673-11676Longo, N., Shuster, R.C., Griffin, L.D., Langley, S.D., and Elsas, L.J. (1992)Activation of insulin receptor signalling by a single amino acid substitutionin the transmembrane domain. J. Biol. Chem. 267, 12416-12419Lopez, V., Stevens, T., and Lindquist, R.N. (1976) Vanadium inhibition ofalkaline phosphatase-catalyzed phosphate ester hydrolysis. Arch.Biochem. Biophys. 175, 31-38Lorenz, P., Pepperkok, R., Ansorge, W., and Pyerin, W. (1993) Cell biologicalstudies with monoclonal and polyclonal antibodies against human caseinkinase II subunit b demonstrate participation of the kinases in mitogenicsignalling. J. Biol. Chem. 268, 2733-2739Lowenstein, E.J., Daly, R.J., Batzer, A.G., Li, W., Margolis, B., Lammers, R.,Ullrich, A., Skolnik, E.Y., Bar-Sagi, D., and Schlessinger, J. (1992) TheSH2 and SH3 domain-containing protein GRB2 links receptor tyrosine kinase to ras signalling. Cell 70, 431-442285Lowy, D.R., Zhang, K., Declue, J.E., and Willumsen, B. M. (1991) Regulation ofp21 ras activity. Trends Gen. 7, 346-351Lu, K., and Campisi, J. (1992) Ras proteins are essential and selective for theaction of insulin-like growth factor 1 late in the G1 phase of the cell cycleinBALB/c murine fibroblasts. Proc. Natl. Acad. Sci. USA 89, 3889-3893Lubben, T. H., and Traugh, J. A. (1983) Cyclic nucleotide-independent proteinkinase from rabbit reticulocytes. J. Biol. Chem. 258, 13992-13997Luttrell, L.M., Hewlett, E.L., Romero, G., and Rogol, A. (1988) Pertussis toxintreatment attenuates some effects of insulin in BC3H-1 murine myocytes.J. Biol. Chem. 263, 6134-6141Luttrell, L.M., Kilgour, E., Larner, J., and Romero, G. (1990) A pertussis toxin-sensitive G protein mediates some aspects of insulin action in BC3H-1murine myocytes. J. Biol. Chem. 265, 16873-16879Mabrouk, G.M., Helmy, I.M., Thampy, K., Wakil, S.J. (1990)Acute hormonal control of acetyl CoA carboxylase. J.Biol. Chem. 265, 6330-6338Macara, I.G. (1980) Vanadium: an element in search of a role. Trends Biochem.Sci. 5, 92-94Macara, I.G. (1986) Activation of 45Ca2+ influx and 22Na+/H+ exchange bypeidermal growth factor and vanadate in A431 cells is independent ofphosphoatidylinositol turnover and its inhibited by phorbol ester anddiacylglycerol. J. Biol. Chem. 261, 9321-9327MacDonald, M.J. (1990) Elusive proximal signals of b-cells for insulin secretion.Diabetes 39, 1461-1466Maclaren, N.K.(1988) How, when, and why to predict IDDM. Diabetes 37, 1591-94Maclaren, N., Schatz, D., Drash, A., and Grave, G. (1989) Initial pathogeniceffents in IDDM. Diabetes 38, 534-538Maddux, B.A., and Goldfine, I.D. (1991) Evidence that insulin plus ATP mayinduce a confirmational change in the n-subunit of the insulin receptorwithout inducing receptor autophosphorylation. J. Biol. Chem. 266, 6731-6736Madsen, N.B. (1986) Glycogen phosphorylase. The Enzymes 17, 365-394Maegawa, H., Olefsky, J.M., Thies, S., Boyd, D., Ulirich, A., and McClain, D.A.(1988) Insulin receptors with defective tyrosine kinase inhibit normalreceptor function at the level of sustrate phosphorylation. J. Biol. Chem.263, 12629-12637Magnuson, M.A., and Granner, D.K. (1988) Analyzing the effects of insulin ongene expression. In: Kahn, U.H., and Harrison, L.U., eds., Insulin286Receptors (Part B) : Clinical Assessment, Biological Responses, andComparison to the IGF-1 Receptor, pp189-209Magnuson, M.A., and Granner, D.K. (1989) Unse of fusion gene to analyze theeffect of insulin on PEPCK gene transcription. In: Draznin, B., Melmed,S., and D. LeRoith, eds, Insulin Action, Alan R. Liss Inc., pp153-162Magnuson, M.A. (1990) Glucokinase gene structure. Diabetes 39, 523-527Malasisse, W.J. (1990) Regulation of insulin release by the intracellularmediators cyclic AMP, Ca2+, inositol 1,4,5-trisphosphate, and diacylglycerol. In : Cuatrecasas, P., and Jacobs, S., ed. Insulin. Berlin :Springer-Verlag, pp113-124Mailer, J.L. (1990) Xenopus oocytes and the biochemistry of cell division.Biochem. 29, 3157-3165Mamon, H., Williams, N., Wood, K., Frazier, A.L., Li, P., Zmuidzinas, A., Kremer,N., D'Acangelo, G., Qi, H., Smith, K., Feig, L., Piwnica-Worms, H.,Halegoua, S., and Roberts, T. (1991) New perspective on raf-1: theinvolvement of pl2ras in the activation of raf-1 and a potentail role for raf-1 in events occuring later in the cell cycle. Cold Spring Harbor Symp.Quant. Biol. 56, 251-263Manak, J.R., Bisschop, N., Kris, R.M., and Prywes, R. (1990) Casein kinase IIenhances the DNA binding activity of serum response factor. Genes andDevel. 4, 955-967Manak, J.R., and Prywes, R. (1991) Mutation of serum response factorphosphorylation sites and the mechanism by which its DNA-bindingactivity is increased by casein kinase II. Mol. Cell. Biol. 11, 3652-3659Manzella, J.M., Rychlik, W., Rhoads, R.E., Hershey, W.B., and Blackshear, P.J.(1991) Insulin induction of ornithine decarboxylase. J. Biol. Chem. 266,2383-2389Margolis, B., Li, N., Koch, A., Mohammadi, M., Hurwitz, D.R., Zilberstein, A.,Ullrich, A., Pawson, T., and Schlessinger, J. (1990) The tyrosinephosphorylated carboxyterminus of the EGF receptor is a binding site forGAP and PLC-y. EMBO J. 9, 4375-4380Maridor, G., Park, W., Krek, W., and Nigg, E.A. (1991) Casein kinase II. J. Biol.Chem. 266, 2362-2368Marqudardt, B., adn Stabel, S. (1992) Sequence of a rat cDNA encoding theERK-1 MAP kinase. Gene 120, 297-299Marshall, S., Garvey, W.T., and Monzon, R. (1987) Shunting of insulin from aretroendocytotic pathway to a degradative pathway by sodium vanadate.J. Biol. Chem. 262, 12005-12012Marshall, S., and Monzon, R. (1987) Down-regulation of cell surface insulinreceptors in primary cultured rat adipocytes by sodium vanadate.Endocrinol. 121, 1116-1122287Marshall, S., and Monzon, R. (1989) Amino acid regulation of insulin action inisolated adipocytes. J. Biol. Chem. 264, 2037-2042Martegani, E., Vanoni, M., Zippel, R., Coccetti, P., Brambilla, R., Ferrari, C.,Sturani, E., and Alberghina, L (1992) Cloning by functionalcomplementation of a mouse cDNA encoding a homologue of CDC25, aSaccharomyces cerevisiae Ras activator. EMBO J. 11, 2151Maslinski, W., Remillard, B., Tsudo, M., and Strom, T. B. (1992) Interleukin-2induces tyrosine kinase-dependent translocation of active raf-1 from theIL-2 receptor into cytosol. J. Biol. Chem. 267, 152881-14284Matsuda, E., and Guroff, G. (1987) Purification and mechanism of activation of anerve growth factor sensitive S6 kinase from PC12 cells. J. Biol. Chem.262, 2832-2844Matsuda, S., Kosako, H., Takenaka, K., Morriyama, K., Sakai, H., Akiyama, T.,Gotoh, Y., and Nishida, E. (1992) Xenopus MAP kinase activator:identification and function as a key intermediate in the phosphorylationcascade. EMBO J. 11, 973-982Matsuda, S., Gotoh, Y., and Nishida, E. (1993) Phosphorylation of Xenopusoocytes mitogen-activated protein kinase kinase by MAP kinase kinasekinase and MAP kinase. J. Biol. Chem. 268, 3277-3281Matthaei, S., Olefsky, J., and Karnieli, E. (1988) Cycloheximde decreasesglucose transporters in rat adipocyte plasma membrane without affectinginsulin-stimulated glucose transport. Biochem. J. 251, 491-497Matuoka, K., Shibata, M., Yamakawa, A., and Takenawa, T. (1992) Proc. Natl.Acad. Sci. USA. 89, 9015-9019Maturo, J.M., and Hollenberg, M.D. (1990) Insulin receptor purification.In:Litwack, G., ed. Receptor Purification. Clifton: Humana Press, pp419-433McClain, D.A., Maegawa, H., Lee., J., Dull, T.J., Ulrich, A., and Olefsky, J.M.(1987) A mutant insulin receptor with defective tyrosine kinase displays nobiologic activity and does not undergo endocytosis. J. Biol. Chem. 262,14663-14671McCormick, F. (1989) ras GTPase activating protein: signal transmitter andsignal terminator. Cell 56, 5-8McDonald, M.J. (1990) Elusive proximal signals of b-cells from insulin secretion.Diabetes 39, 1461-1466McDonald, 0.B., Hannun, Y.A., Reynolds, C.H., and Sahyoun, N. (1991)Activation of casein kinase II by sphingosine. J. Biol. Chem. 266, 12733-21776288McNeill, J.H., Yuen, V.G., Hoveyda, H.R., and Orvig, C. (1992)Bis(maltolato)oxovanadium (IV) is a potent insulin mimic. J. Med. Chem.35, 1489-91Medema, R.H., Wubbolts, R., and Bos, J.L. (1991) Two dominant inhibitorymutants of p21 ras interfere with insulin-induced gene expression. Mol.Cell. Biol. 11, 5963-5967Medema, R. H., De Vries-smits, A.M.M., Van Der Zon, G.C.M., Maassen, J.A.,and Bos, J. L. (1993) Ras activation by insulin and epidermal growthfactor through enhanced exchange of guanine nucleotides on p21ras.Mol. Cell. Biol. 13, 155-162Meek, D.W., Simon, S., Kikkawa, U., and Eckhart, W. (1990) The p53 tumoursuppressor protein is phosphorylated at serine 389 by casein kinase II.EMBO J. 9, 3253-3260Meisner, H., and Czech, M.P. (1991) Phosphorylation of transcriptional factorsand cell-cycle-dependent proteins by casein kinase II. Curr. Opinion CellBiol. 3, 474-483Meloche, S., Seuwen, K., Pages, G., Cobb, M.H., and Pouyssegur, J. (1992a)Biphasic and synergistic activation of p44mapk(EKR1) by growth factors:correlation between late phase activation and mitogenicity. Endocrinol.Biphasic and synergistic activation of p44mapk(ERK1) by growth factors:correlation between late phase activation and mitogenicity. Mol.Endocrinol. 6, 845-854Meloche, S. Pages, G., and Pouyssegur, J. (1992b) Functional expression andgrowth factor activation of an epitope-tagged p44 mitogen-activatedprotein kinase p44mapk. Mol. Biol. Cell. 3, 63-71Messina, J.L. (1990) Insulin's regulation of c-fos gene transcription in hepatpmacells. J. Biol. Chem. 265, 11700-11705Messina, J.L., Standaert, M.L., Ishizuka, T., Weinstock, R.S., and Farese, R.V.(1992) Role of protein kinase C in insulin's regulation of c-fostranscription. J. Biol. Chem. 267, 9223-9228Metallo, A., and Villa-Moruzzi, E. (1991) Protein phosphatase-1 and -2A, kinaseFA, and casein kinase II in skeletal muscle of streptozotocin diabetic rats.Arch. Biochem. Biophys. 289, 382-386Meyerovitch, J., Farfel, Z., Sack, J., and Shechter, Y. (1987) Oral administrationof vanadate normalizes blood glucose levels in streptozotocin-treatedrats. J. Biol. Chem. 262, 6658-6662Meyerovitch, J., Backer, J.M., Csermely, P., Shoelson, S.E., and Kahn, C.R.(1992) Insulin differentially regulates protein phosphotyrosinephosphatase activity in rat hepatoma cells. Biochem. 31, 10338-10344Miller, T.B. Jr.(1985) Insulin regulation of glycogen synthase in rat heart andfiver. In: Czech, M.P., ed. Molecular Basis of Insulin Action. New York:Plenum Press, pp247-261289Miller, D.S. (1988) Stimulation of RNA and protein synthesis by intracellularinsulin. Science 240, 506-509Miralpeix, M., Gil, J., Rosa, J.L., Carreras, J., and Bartrons, R. (1989) Vanadatecounteracts glucagon effects in isolated rat hepatocytes. Life Sci. 44,1491-1497Mirmira, R.G., Nakagawa, S.H., and Tager, H.S. (1991) Imortance of thecharacter and configuration of residues B24, 25, and B26 in insulin-receptor interaction. J. Biol. Chem. 266, 1428-1436Misek, D.E., and Saltiel, A.R. (1992) An inositol phosphate glycan derived from atrypanosoma brucei glycosyl-phosphatidylinosital mimics some of themetabolic actions of insulin. J. Biol. Chem. 267, 16266-16273Miyasaka, T., Chao, M.V., Sherline, P., and Saltiel, A. (1990) Nerve growthfactor stimulates a protein kinase in PC-12 cells that phosphorylatesmicrotubule-associated protein 2. J. Biol. Chem. 265, 4730-4735Miyata, Y., and Yahara, I. (1992) The 90-kDa heat shock protein, HSP90, bindsand protects ccasein kinase II from self-aggregation and enhances itskinase activity. J. Biol. Chem. 267, 7042-7047Miyazaki, T., Matsuda, Y., Toyonaga, T., Miyazaki, J.I., Yazaki, Y., andYama9ura, K.I. (1992) Prevention of autoimmune insulitis in nonobesediabetic mice by expression of major histocompatibility complex class I Ldmolecules. Proc. Natl. Acad. Sci. 89, 9519-9523Molina, E, Plana, M., and Itarte, E. (1991) Heterogeneity of rat liver cytosolcasein kinase 2. Biochem. J. 277, 811-818Moller, D.E., Benecke, H., and Flier, J.S. (1991) Biological activities of naturallyoccurring human insuin receptor mutations. J. Biol. Chem. 266, 10995-11001Moller, C., Hansson, A., Enberg, B., Lobie, P.E., Norstedt, G. (1992) Growthhormone induction of tyrosine phosphorylation and activation of mitogen-activated protein kinases in cells transfected with rat GH receptor cDNA.J. Biol. Chem. 267, 23403-23408Molloy, C.J., Bottaro, D.P., Fleming, T.P., Marshall, M.S., Gibbs, J.B., andAaronson (1989) PDGF induction of tyrosine phosphorylation of GTPaseactivating protein. Nature 342, 711-714Molloy, C.J., Fleming, T.P., Bottaro, D.P., Cuadrado, A., and Aaronson, S.A.(1992) Platelet-derived growth factor stimulation of GTPase activatingprotein tyrosine phosphorylation in control and c-H-ras-expressing NIH313 cells correlates with p21ras activation. Mol. Cell. Biol. 12, 3903-3909Montesano, R., Pepper, M.S., Bellin, D., Vassalli, J.D., and Orci, L. (1988)Induction of angiogenesis in vitro by vanadate, an inhibitor ofphosphotyrosine phosphatase. J. Cell. Physiol. 134, 460-466290Mooney, R.A., Bordwell, K.L., Luhowskyj, S., and Casnellie, J.E. (1989) Theinsulin-like effect of vanadate on lipolysis in rat adipocytes in notaccomplished by an insulin-like effect on tyrosine phosphorylation.Endocrinol. 124, 422-429Mooney, R.A., and Bordwell, K.L. (1992) Differential dephosphorylation of theinsulin receptor and its 160-kDa substrate (pp160) in rat adipocytes. J.Biol. Chem. 267, 14054-14060Moran, M., Polakis, P., McCormick, F, Pawson, T., and Ellis, C. (1991) Proteintyrosine kinases regulate the phosphorylation, protein interaction,subcellular distribution, and activity of p21 ras GTPase-activating protein.Mol. Cell. Biol. 11, 1804-1812Morrison, D.K., Kaplan, D.R., Rapp, U., and Roberts, T.M. (1988) Signaltransduction from membrane to cytoplasm:growth factors and membrane-bound oncogene products increase Raf-1 phosphorylation and associatedprotein kinase activity. Proc. Natl. Acad. Sci. 85, 8855-8859Morrison, D.K., Kaplan, D.R., Escobedo, J.A., Rapp, U.R., Roberts, T.M., andWilliams, L.T. (1989) Direct activation of the serine/threonine kinaseactivity of raf-1 through tyrosine phospohrylation by the PDGF n-receptor.Cell 58, 649-657Mountjoy, K.G., and Flier, J.S. (1990) Vanadate regulates glucose transporter(Glut-1) expression in NIH3T3 mouse fibroblasts. Endocrinol. 126, 2778-2787Mudd, L.M., and Raizada, M.K. (1989) Regulation of growth factor receptors byprotein kinase C. In: ;LeRoith, D., and Raizada, M.K., eds., Molecular andcellular biology of insulin like growth factors and their receptor. New York:Plenum Press, pp315-325Mueckler, M. (1990) Family of glucose-transporter genes. Implications forglucose homeostasis and diabetes. Diabetes 39, 6-11Mueckler, M. (1993) Glucokinase, glucose sensing, and diabetes. Proc. Proc.Natl. Acad. Sci. USA. 90, 784-785Mukhopadhyay, N.K., Price, D.J., Kyriakis, J.M., Pelech, S., Sanghera, J.S., andAvruch, J. (1992) J. Biol. Chem. 267, 3325-3335Muller, H.K., Kellerer, M., Ermel, B., Muhlhofer, A., Obermaier-Kusser, B., Vogt,B., and Haring, H.U. (1991) Prevention by protein kinase C inhibitors ofglucose induced insulin-receptor tyrosine kinase resistance in rat fat cells.Diabetes 40, 1440-48Munoz, P., Guma, A., Camps, M. Furriols, M., Testar, X., Palacin, M., andZorzano, A. (1992) Vanadate stimulates system A amino acid transportactivity in skeletal muscle. J. Biol. Chem. 267, 10381-10388Murakami, M. Rosenautop osphorylation in signal transduction. J. Biol. Chem. 266, 22653-22660• i • • I • • I291Mutoh, T., Rudkin, B.B., Koizumi, S., and Guroff, G. (1988) Nerve growth factor,a differentiating agent, and epidermal growth factor, a mitogen, increasethe activity of different S6 kinase in PC12 cells. J. Biol. Chem. 263,15853-15856Myers, M.G., Jr., Sun, X.J., Cheatham, B., Jachna, B.R., Glasheen, E.M.,Backer, J.M., and White, M.F. (1993) IRS-is a common element in insulinand insulin-like growth factor-I signalling to the phosphatidylinositol 3'-kinase. Endocrinol. 132, 1421-1430Nadiv, 0., Cohen, 0., and Zick, Y. (1992) Defects in insulin's signal transductionin old rat livers. Endocrinol. 180, 1515-1524Nakafuku, M., Satoh, T., and Kaziro, Y. (1992) Differentiation factors, includingnerve growth factor, fibroblast growth factor and interleukin-6, induce anaccumulation of an active ras-GTP complex in rat pheochromocytomaPC12 cells. J. Biol. Chem. 267, 19448-19454Nakagawa, S., and Tager, H.S. (1991) Implications of invariant residue LeuB6 ininsulin-receptor interaction. J. Biol. Chem. 266, 11502-11509Nakajo, S., Hagiwara, T., Nakaya, K, and Nakamura, Y. (1986) Tissuedistribution of casein kinases. Biochem. Int. 13, 701-707Nakielny, S., Campbell, D.G., and Cohen, P. (1992a) MAP kinase kinase fromrabbit skeletal muscle. FEBS Lett. 308, 183-189Nakielny, S., Cohen, P., Wu, J., and Sturgill, T. (1992b) MAP kinase activatorfrom insulin-stimulated skeletal muscle is a protein threonine/tyrosinekinase. EMBO J. 11, 2123-2129Nechay, B.R. (1984) MEchanisms of action of vanadium. Ann. Rev. Pharmacol.Toxicol. 24, 501-24Nechay, B.R., Nanninga, L., Nechay, P.S.E., Post, R.L., Grantham, J.J., Macara,I. G., Kubena, L.F., Phillips, T.D., and Nielsen, F.H. (1986a) Role ofvanadium in biology. Fed. Proc. 45, 123-132Nechay, B.R., Nanninga, L.B., and Nechay, P.S.E. (1986b) Vanadyl andvanadate binding to selected endogenous phosphate, carboxyl, andamino ligands; calculations of cellular vanadium species distribution. Arch.Biochem. Biophys. 251, 128-138Nel, A.E., Poliack, S., Landreth, G., Ledbetter, J.A., Hultin, L., Williams, K,Katz,.R., and Akerley, B. (1990a) CD-3-mediated activation of MAP-2kinase can be modified by ligation of the CD4 receptor. J. Immunol. 145,971-979Nel, A.E., Hanekom, C., Rheeder, A., Williams, K., Poliack, S., Katz, R.,Landreth, G.E. (1990b) Stimulation of MAP2 kinase activity in TI m• hoc es • .1 or anti-Ti monoclonal antibody is partially• ependent on protein kinase C. J. Immunol. 144, 2683-2689292Nemenoff, R.A., Price, D.J., Mendelsohn, M.J., Carter, E.A., and Avruch, J.(1988) An S6 kinase activated during liver regeneration is related to theinsulin-stimulated S6 kinase in H4 hepatoma cells. J. Biol. Chem. 263,1945-19460Nemonoff, R.A., Winitz, S., Qian, N.X., Putten, V.V., Johnson, G.L., and Heasley,L.E. (1993) Phosphorylation and activation of a high molecular weightform of phospholipase A2 by p42 microtubule-associated protein 2 kinaseand protein kinase C. J. Biol. Chem. 268, 1960-1964Newman, J.D., and Harrison, L.C. (1990) Insulin receptor purification. ln:Litwack,G., ed. Receptor Purification. Clifton: Humana Press, pp403-417Nielsen, F.H. (1987) Vanadium. In: Mertz, ed. Trace elements in human andanimal nutrition. San Diego: Academic Press, pp275-300Nielsen, F.H., and Uthus, E.O. (1990) The essentiality and metabolism ofvanadium. In: Chasteen, N.D., ed. Vanadium in biological systems.Amsterdam: Kluwer Academic Publishers, pp51-62Nishida, E., and Gotoh, Y. (1993) The MAP kinase cascade is essential fordiverse signal transduction pathways. Trends. Biochem. Sci. 18, 128-131Nishimura, H., Kuzuya, H., Okamato, M., Yamada, K., Kosaki, A., Kakehi, T.,Inoue, G., Kono, S., and Imura, H. (1989) Postreceptor defect in insulinaction in streptozotocin-induced diabetic rats. Am. J. Physiol. 256 : E624-630Niven, M.J., and Hitman, G.A. (1986) The molecular genetics of diabetesmellitus. Bios. Rep. 6, 501-512Northwood, I.C., Gonzalez, F.A., Wartmann, M., Raden, D.L., and Davis, R. J.(1991) Isolation and characterization of two growth factor-stimulatedprotein kinases that phosphorylate the epidermal growth factor atthreonine 669. J. Biol. Chem. 266, 15266-15276Novak-Hofer, I., and Thomas, G. (1984) An activated S6 kinase in extracts fromserum- and epidermal growth factor-stimulated Swiss 3T3 cells. J. Biol.Chem. 258, 5995-6000Nyomba, B.L., Ossowski, V.M., Bogardus, C., and Mott, D.M. (1990) Insulin-sensitive tyosine kinase: relationship with in vivo insulin action in humans.Am. J. Physiol. 258, E964-974O'Brien, R.M., Houslay, M.D., Milligan, G., and Siddle, K. (1987) The insulinreceptor tyrosyl kinase phosphorylates holomeric forms of the guaninenucleotide regulatory proteins Gi and Go. FEBS Lett. 212, 281-288O'Brien, R.M., and Granner, D.K. (1990) PEPCK gene as model of inhibitoryeffects of insulin on gene transcription. Diabetes Care 13, 327-339O'Brien, R.M., and Granner, D.K. (1991) Regulation of gene expression byinsulin. Biochem. J. 278, 609-619293O'Brien, R.M., Lucas, P.C., Forest, C.D., Magnuson, M.A., and Granner, D.K.(1991) Identification of a sequence in the PEPCK gene that mediates anegative effect of insulin on transcription. Science 249, 533-537Odawara, M., Kadowaki, T., Yamamoto, R., Shibasaki, Y., Tobe, K., Accili, D.,Bevins, C., Mikami, Y., Matsuura, N., Akanuma, Y., Takaku, F., Taylor,S.I., and Kasuga, M. (1989) Human diabetes associated with a mutationin the tyrosine kinase domain of the insulin receptor. Science. 245, 66-68Ohashi, P.S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C.T., Odermatt, B.,Malissen, B., Zinkernagel, R.M., and Hengartner, H. (1991) Ablation of"tolerance" and induction of diabetes by virsu infection in viral antigentransgenic mice. Cell 65, 305-317Ohmichi, M., Pang, L., Decker, S.J., and Saltiel, A.R. (1992) Nerve growth factorstimulates the activities of the raf-1 and the mitogen-activated proteinkinases via the trk protooncogen. J. Biol. Chem. 267, 14604-14610Oka, Y., Asano, T., Shibasaki, Y., Lin, J.L., Stukuda, K., Akanuma, Y., andTakaku, F. (1990) Increaesd liver glucose-transporter protein and mRNAin streptozocin-induced diabetic rast. Diabetes 36, 441-46Okamoto, M., White, M.F., Maron, R., and Kahn, C.R. (1986)Autophosphorylation and kinase activity of insulin receptor in diabetic rats.Am. J. Physiol. 251, E542-550Oldstone, M.B.A., Nerenber, M., Southern, P., Price, J., and Lewicki, H. (1991)Virus infection triggers insulin-dependent diabetes mellitus in a transgenicmodel: role of anti-self (virus) immune response. Cell 65, 319-331Ole-Moiyoi, O.K., Brown, W.C., lams, K.P., Nayar, A., Tsukamoto, T., Macklin,M.D. (1993) Evidence fro the induction of casein kinase II in bovinelymphocytes transformation by the intracellular protozoan parasiteTheileria parva. EMBO J. 12, 1621-1631Olefsky, J.M. (1990) Insulin receptor: a mutifunctional protein, Diabetes 39,1009-1016Olefsky, J.M. (1991) Insulin action and insulin resistance in non-insulindependent diabetes mellitus. In Shigeta, Y., Kobayashi, M., and Olefsky,J.M., eds, Recent advances in insulin action and its disorders. Amsterdam: Elsvier Science Publisher, B.V., pp81-118Olivier, J.P., Raabe, T., Henkemeyer, M., Dickson, B., Mbamalu, G., Margolis,B., Schlessinger, J., Hafen, E., and Pawson, T. (1993) A Drosophila SH2-SH3 adaptor protein implicated in coupling to sevenless tyrosine kinase toan activator of ras guanine nucleotide exchange, sos. Cell 73, 179-191Orci, L., Vassalli, J.D., and Perrelet, A. (1988) The insulin factory. ScientificAmerican 259 (3), 85-94294Oshima, M., Sithanandam, G., Rapp, U.R., and Guroff, G. (1991) Thephosphorylation and activation of B-raf in PC12 cells stimulated by nervegrowth factor. J. Biol. Chem. 266, 23753-23760Osterop, A.P.R.M., Medema, R.H., Bos, J.L., Zon, G.C.M., Moller, D.E., Flier,J.S., Moller, W., and Maassen, J.A. (1992) Relation between the insulinreceptor number in cells, autophosphorylation and insulin-stimulated ras-GTP formation. J. Biol. Chem. 267, 1 4647-1 4653Owada, M.K., Iwamoto, M., Koike, T., and Kato, Y. (1989) Effects of vanadate ontyrosine phosphorylation and the pattern of glycosaminoglycan synthesisin rabbit condrocytes in culture. J. Cell. Physiol. 138, 484-492Owaki, H., Makar, R., Boulton, T.G., Cobb, M.H., and Geppert, T.D. (1992)Extracellular signal-regulated kinases in T cells: characterization ofhuman ERK1 and ERK2 cDNAs. Biochem. Biophys. Res. Comm. 182,1416-1422Paccaud, J.P., Siddle, K., and Carpentier, J.L. (1992) Internalization of thehuman insulin receptor. J. Biol. Chem. 267, 13101-13106Padel, U., and Soling, H.D. (1985) Phosphorylation of the ribosomal protein S6during agonist-induced exocytosis in exocrine glands is catalyzed bycalcium-phospholipid-dependent protein kinase (protein kinase C). Eur. J.Biochem. 151, 1-10Palen, E., and Traugh, J.A. (1987) Phosphorylation of ribosomal protein S6 bycAMP-dependent protein kinase and mitogen-activated S6 kinasedifferentially alters the translation of globin mRNA. J. Biol. Chem. 262,3518-3523Palmer, J.P., and McCulloch, D.K. (1991). Prediction and prevention of IDDM:1991. Diabetes 40, 943-47Pang, L., Decker, S.J., and Saltiel, A.R. (1993) Bombesin and epidermal growthfactor stimulate the mitogen-activated protein kinase through differentpathways in Swiss 3T3 cells. Biochem. J. 289, 283-287Payne, D.M., Rossomando, A.J., Martino, P, Erickson, A.K., Her, J.H.,Shabanowitz, J., Hunt, D.F, Weber, M.J., and Sturgill, T.W. (1991)Identification of the regulatory phosphorylation sites in pp42/mitogen-activated protein kinase (MAP kinase). EMBO J. 10, 885-892Pederson, R.A., Ramanadham, S., Buchan, M.J.,, and McNeill, J.H. (1989)Long-term effects of vanadyl treatment on streptozocin-induced diabetesin rats. Diabetes 38, 1390-95Pelech, S.L., Tinker, D.A., Chan, C.P., and Krebs, E.G. (1987) Role of proteinphosphorylation in growth factor signal transduction. In : Raizada, M.K.,Philips, M.I., and LeRoith, D. (eds) Insulin, insulin-like growth factors, andtheir receptors in the central nervous system. Plenum PublishingCorporation, New York. pp27-46 295Pelech S.L., and Krebs, E.G. (1987) Mitogen-activated S6 kinas is stimulatedvia protein kinase C-dependent and independent pathways in Swiss 3T3cells. J. Biol. Chem. 262, 11598-11606Pelech, S.L, Sanghera, J.S., and Daya, M. (1990) protein kinase cascades inmeiotic and mitotic cell cycle control. Biochem. Cell Biol. 68, 1297-1330Pelech, S.L., Sanghera, J.S., Paddon, H.B., Quayle, K.A., and Browsey, R.W.(1991) Identification of a major maturation-activated acetyl CoAcarboxylase kinase in sea star oocytes as p44mpk. Biochem. J. 274, 759-67Pelech, S.L., and Sanghera, J.S. (1992a) MAP kinases: charting the regulatorypathways. Science 257 : 1355-56Pelech, S.L., and Sanghera, J.S. (1992b) Mitogen-activated protein kinases,versatile transducers for cellular signalling. Trends. Biochem. Sci. 17,233-238Perisic, 0., and Traugh, J. A. (1983a) Protease-activated protein kinase IImediates multiple phosphorylation of ribosomal protein S6 inreticulocytes. J. Biol. Chem. 258, 13998-14002Perisic, 0., and Traugh, J. A. (1983b) Protease-activated protein kinase II as thepotential mediator of insulin-stimulated phosphorylation of ribosomalprotein S6, J. Biol. Chem. 258, 9589-9592Perisic, 0., and Traugh, J. A. (1983c) Protease-activated protein kinase II as themediator of epidermal growth factor-stimulated phosphorylation ofribosomal protein S6. FEBS lett. 183, 215-218Pessin, J.E., and Bell, G.I. (1992) Mammalian facilitative glucose transporterfamily : structure and molecular regulation. Ann. Rev. Physiol. 54, 911-930Peter, M., Sanghera, J.S., Pelech, S.L., and Nigg, E.A. (1992) Mitogen-activatedprotein kinases phosphorylate nuclear lamins and display sequencespecificity overlapping that of mitotic protein kinase p34cdc2. Eur. J.Biochem. 205, 287-294Pilch, P.F., and Czech, M.P. (1980) The subunit sttrucure of the high affinityinsulin receptor. Evidence for a disulfide-linked receptor complex in fat celland liver plasma memebrnaes. J. Biol. Chem. 255, 1722-1728Pilch, P.F., Waugh, S.M., and O'Hare, T. (1989) The structural basis for insulinreceptor function. Adv. Reg. Cell. Growth Activation. 1, 203-232Pilkis, S.J., El-Maghrabi, M.R., and Claus, T.H. (1988) Hormonal regulation ofhepatic gluconeogenesis and glycolysis. Ann. Rev. Biochem. 57, 755-83Pilkis, S.J., El-Maghrabi, M.R., and Claus, T.H. (1990)^Fructose-2,6-bisphosphate in control of hepatic gluconeogenesis. Diabetes Care 13,582-599296Pilkis, S.J., and Granner, D.K. (1992) Molecular physiology of hepaticgluconeogenesis and glycolysis. Annu. Rev. Physiol. 64, 885-909Pillion, D.J., Kim, S.J., Kim, H., and Meezan, E. (1992) Insulin signaltransduction : the role of protein phosphorylation. Am. J. Med. Sci. 303,40-52Pinna, L. A. (1990) Casein kinase 2 : an "eminence grise" in cellular regulation?Biochem. Biophys. Act. 1054, 267-284Pinna, L.A., Meggio, F., and Marchiori, F. (1991) Type-2 casein kinases :general properties and substrate specificity. In: Kempt, B.E., ed. Peptidesand protein phosphorylation. Boca Raton: CRC press, pp145-170Pollet, R.J. (1989) Glycerolipid mediators of insulin action. In: Draznin, B.,Melmed, S., and LeRoth, D., eds. Insulin action. New York: Alan R. Liss,Inc. pp107-116Pomerance, M., Schweighoffer, F., Tocque, B., and Pierre, M. (1992) Stimulationof mitogen-activated protein kinase by oncogenic ras p21 in Xenopus000cytes. J. Biol. Chem. 267, 16155-16160Porras, A., Nebreda, A.R., Benito, M., and Santos, E. (1992) Activation of ras byinsulin in 3T3 L1 cells does not involve GTPase-activating proteinphosphorylation. J. Biol. Chem. 267, 21124-21131Posada, J., Sanghera, J., Pelech, S., Aebersold, R., and Cooper, J. (1991)Tyrosine phosphorylation and activation of homologous protein kinasesduring oocyte maturation and mitogenic activation of fibroblasts. Mol.Cell. Biol. 11, 2517-2528Posada, J, and Cooper, J.A. (1992a) Molecular signal integration, interplaybetween serine, threonine, and tyrosine phosphorylation. Mol. Biol. Cell3, 583-592Posada, J., and Cooper, J.A. (1992b) Requirement for phosphorylation of MAPkinase during meiosis in Xenopus oocytes. Science 255, 212-215Posada, J., Yew, N., Ahn, N.G., Vande Woode, G.F., and Cooper, J. (1993) Mosstimulates MAP kinase in Xenopus oocytes and activates a MAP kinasekinase in vitro. Mol. Cell. Biol. 13, 2546-2553Powell, D.R., Suwanichkul, A., Cubbage, M.L., DePaolis, L.A., Snuggs, M.B.,and Lee, P.D.K. (1991) Insulin inhibits transcription of human gene forinsulin-like growth factor-binding protein-1. J. Biol. Chem. 266, 18868-18876Prager, D., and Melmed, S. (1988) Insulin regulates expression of the humangrowth hormone gene in transfected cells. J. Biol. Chem. 263, 16580-16585Prager, D., and Melmed, S. (1989) Insulin regulation of growth hormone geneexpression. In: Draznin, B., Melmed, S., and 1.). LeHoith, eds, InsulinAction, Alan R. Liss Inc.,pp163-173297Rajagopalan, M., Neidigh, J.L., and McClain, D.A. (1991) Amino acid sequencesGly-Pro-Leu-Tyr and Asn-Pro-Glu-Tyr in the submembrane domain of thermal endoc tosis. J. Biol. Chem. 266I - ^I • " II^I .Prasad, K.V.S., and Rudd, C. A raf-1 related p110 polypeptide associates withthe CD4-p56Ick complex in T cells. Mol. Cell. Biol. 12, 5260-5267Price, D.J., Grove, J.R., Calvo, V., Avruch, J., and Bierer, B.E. (1992)Rapamycin-induced inhibition of the 70-kDa S6 kinase. Science 257, 973-977Price, D.J., Gunsalus, R., and Avruch, J. (1990) Insulin activates a 70-kDa S6kinase through serine/threonine-specific phosphorylation of the exzymepolypeptide. Proc. Natl. Acad. Sci. 87, 7944-7948Price, D.