Insulin receptor protein-tyrosine phosphatases. Leukocyte common

4 downloads 0 Views 1MB Size Report
Jul 15, 2015 - In this representative experiment, LAR demonstrated a striking .... J. (1990) in Handbook of Experimental Pharmacology. (Cuatrecasas, P., and ...
Communication

THEJOURNAL OF BIOLOGICAL CHEMISTRY Vol. 267, No. 20, Issue of July 15, pp. 13811-13814, 1992

0 1992 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A .

Insulin Receptor Protein-tyrosine Phosphatases

Insulin initiates its effects on cellular growth and metabolism by binding to a specific plasma membrane receptor which elicits a rapid autophosphorylation of specific tyrosine residues in several domains of the cytoplasmic portion of the LEUKOCYTE COMMON ANTIGEN-RELATED receptor @-subunit(1). Receptor autophosphorylation and the PHOSPHATASE RAPIDLYDEACTIVATES THE associated activation of the receptor kinase activity toward INSULIN RECEPTOR KINASE BY PREFERENTIAL exogenous substrates constitutes a major pathway of signal DEPHOSPHORYLATIONOF THE RECEPTOR transduction by insulin (1-5). Recent work has characterized REGULATORY DOMAIN* in detail the relationship between phosphorylation of specific receptor sites and activation of the receptor kinase activity (Received for publication, April 17, 1992) (3-7). These studies have shown that full phosphorylation of Naotake Hashimoto, Edward P. FeenerS, three tyrosyl residues (1146, 1150, and 1151) in the receptor Wei-Ren Zhang, and Barry J. GoldsteinQ @-subunit(the 3-Tyr(P)form of the Tyr-1150 domain) leads From the Research Division, Joslin Diabetes Center and to full activation of the receptor kinase toward exogenous Department of Medicine, Brighnm and Women’s Hospital substrates; the 2-Tyr(P)form of this region is associated with and Harvard Medical School, Boston, Massachusetts 02215 only minimal kinase activity (3, 5). The phosphorylation of tyrosines 1316 and 1322 in the C-terminal region does not A number of protein-tyrosine phosphatase(s) appear to influence activation of the receptor kinase (8-10). (PTPases) have been shown to dephosphorylate the Less is known about the cellular processes of receptor insulin receptor in vitro; however, it is not known dephosphorylation that determine the overall steady-state whether any individual PTPase has specificity for certain phosphotyrosine residues of the receptorthat reg- level of insulin receptor kinase activity (11). Since highly ulate its intrinsic tyrosine kinase activity.We evalu- purified, activated insulin receptors retain their phosphorylated thedeactivation of the insulin receptor kinaseby ation state andkinase activity, cellular protein-tyrosine phosthree candidate enzymes that areexpressed in insulin- phatase enzymes (PTPases,’ EC 3.1.3.48; see Ref. 12 for a review) that reverse the phosphorylation of specific regulatory sensitive rat tissues, including the receptor-like PTPases LAR and LRP, and the intracellular enzyme, sitesin the insulin receptor have been implicated in the PTPaselB. Purified insulin receptors were activated inactivation of receptor kinase activity in intactcells (13,14). by insulin and receptordephosphorylation, and kinase In addition, in vitro studies have supported this hypothesis activity was quantitated after incubation with recom- by demonstrating that conversion of the 3-Tyr(P)form of the binant PTPases from an Escherichia coli expression receptor Tyr-1150 domain to the 2-Tyr(P) form is closely system. When related to the level of overall receptor correlated with deactivation of the receptor kinase activity dephosphorylation, LAR deactivated the receptor ki- (15). Several cloned or purified PTPases have been shown to nase 3.1 and 2.1 times more rapidlythaneither dephosphorylate the insulin receptor (11,14,16-19) or insulin PTPaselB or LRP, respectively ( p < 0.03). To assess receptor-related phosphopeptides (20-22); however, virtually whether these effects were associated with preferen- noinformation is available on the potential specificity of tial dephosphorylation of the regulatory (Tyr-1150) individual PTPases for the insulin receptor regulatory domain domain of the receptor &subunit, we performed tryptic or theirability to inactivate the receptor kinase. mapping of the insulin receptor &subunit after deOur laboratory has recently identified three PTPases that phosphorylation by PTPases. Relative to the rate of are highly expressed in insulin-sensitive liver and muscle initial loss of 32Pfrom receptor C-terminalsites, LAR dephosphorylated the Tris-phosphorylated Tyr- 1150 tissue and are candidate enzymes for having a physiological role in the regulation of insulin action (23, 24). These include domain 3.5 and 3.7 times more rapidly than either two PTPases thathave a receptor-like transmembrane strucPTPaselB orLRP, respectively ( p < 0.01). The accelerated deactivation of the insulin receptor kinase by ture and tandem conserved PTPase domains, including the LAR and its relative preference for regulatoryphos- enzymes LAR and LRP (25-28) and PTPaselB, which has a photyrosine residues further support a potential role single PTPase domain and appears to be associated with the for this transmembrane PTPase in the physiological endoplasmic reticulum (29-33). Using expression of recombinant DNA constructs in an Escherichia coli system, we regulation of insulin receptors in intactcells. recently demonstrated that each of these PTPases was able to dephosphorylate activated, native insulinreceptors (23). In the present study, we examined the relative ability of these * This work was supported by National Institutes of Health Grant PTPases to functionally inactivate the receptor kinase. FurDK43396 (to B. J. G.). Molecular and biochemistry core laboratory thermore, we have shown that their ability to deactivate the services were provided by the Joslin Diabetes and Endocrinology receptor kinase correlates with a relative PTPase specificity Research Center Grant DK36836. The costs of publication of this for dephosphorylation of the 3-Tyr(P) form of the receptor article were defrayed in part by the payment of page charges. This regulatory (Tyr-1150) domain.

article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solelyto indicate this fact. $Supported by a fellowship grant from the Juvenile Diabetes Foundation. 5 To whom all correspondence should be addressed Division of Endocrinology and Metabolism, Jefferson Medical College, Rm. 349, Alumni Hall, 1020 Locust St., Philadelphia, P A 19107-6799.



The abbreviations used are: PTPase, protein-tyrosine phosphatase; LAR, leukocyte common antigen-related phosphatase; LRP, leukocyte common antigen-related phosphatase; HEPES, 4-(2-hydroxyethy1)-1-piperazineethanesulfonicacid; DTT, dithiothreitol; TPCK, L-1-tosylamido-2-phenylethylchloromethyl ketone; Tricine, N-tris(hydroxymethy1)methylglycine;SDS, sodium dodecyl sulfate.

13811

13812

Insulin Receptor Protein-tyrosine Phosphatases EXPERIMENTAL PROCEDURES

Materials-The pKK233-2 bacterial expression vector was obtained from Pharmacia LKB Biotechnology Inc. E. coli DHB4 cells were kindly provided by Dr. Haruo Saito (Dana-FarberCancer Institute, Boston, MA). Other materials were obtained as described previously (23). cDNA Expression in a Bacterial System-cDNA inserts for the full-length coding region of rat PTPaselB and the entire cytoplasmic domain of LRP and LAR were obtained by cDNA library screening and/or cDNA amplification and subcloned into the isopropylthio-#?D-galactoside-inducibleexpression vector pKK233-2 (23). For protein expression, the plasmids were transformed into E. coli strain DHB4 (AphoA phoR (F' lacIQpro)), which lacks bacterial alkaline phosphatase activity (34). Cultures were induced with 2 mM isopropylthio-#?D-galactoside, and thebacteria were sedimented, lysed with lysozyme, and subjected to three cycles of freeze-thawing in a dry ice/ethanol bath. Recombinant PTPases were recovered in supernatants prepared by treating the lysed cells with DNase I and removing insoluble material by centrifugation a t 15,000 X g for 15 min at 4 "C (35). Control extractsof induced DHB4 cells carrying the pKK233-2 vector alone were devoidof PTPase activity against eithera phosphopeptide substrate or intact, phosphorylated insulin and epidermal growth factor receptors (23). Insulin Receptor Dephosphorylation and Kinase Inactivation by Recombinant PTPases-Partially purified human insulin receptors (36 pg of protein), obtained by wheat germ agglutinin-agarose affinity chromatography of solubilized plasma membranes from transfected Chinese hamster ovary cells (36), were autophosphorylated for 2 h at 4 "C in 0.45 ml of a solution containing 50 mM HEPES buffer, pH 7.6,0.1% (v/v) Triton X-100,5mM MnC12,1p M insulin, and 180 pCi of -y-[32P]ATP(3000 pCi/mmol) with a final [ATP] of 0.1 mM. A BioGel P6 spin column (Bio-Rad) was used to remove free labeled ATP, and aliquots of the autophosphorylated insulin receptors (2 pgof protein) were incubated with 0.075 ml ofrecombinant PTPase extract at room temperature ina 0.125-mlreaction containing 50 mM HEPES buffer, pH 7.6, 1 mM DTT, and 2 mM EDTA. If necessary, recombinant PTPases were diluted in extraction buffer to provide an initial receptor dephosphorylation of 30-50% of the control level during a 10-min incubation. The reaction was terminated by adding 0.5 ml of a chilled stop solution containing 4 mM EDTA, 100 mM NaF, 5 mM sodium orthovanadate, 0.1 mg/ml aprotinin, and 2 mM phenylmethylsulfonyl fluoride in 50 mM HEPES buffer, pH 7.6. Insulin receptors were then adsorbed with 1 pgof anti-insulin receptor monoclonal antibody (83-7, kindly provided by Dr. Kenneth Siddle, University of Cambridge) and 2 pg ofrabbit anti-mouseIgG at 4 "C. The adsorbed receptors were then precipitated with Trisacryl protein A (Pierce Chemical Co.), washed, and boiled in sample buffer containing 100 mM DTT prior to electrophoresis in 7.5% polyacrylamide gels containing SDS (37). Protein was assayed by the method of Bradford (38). Receptor kinase inactivation was assayed by a modification of the method of King et al. (15). Insulin receptors were phosphorylated in parallel reactions exactly as described above except without -y-[3'P] ATP. The activated insulin receptors were treated with recombinant PTPases as above, the stop solution was added, and the receptors were immunoprecipitated. The pellets were washed, and thereceptor kinase activity was assayed in duplicate for 2 min a t room temperature in a 60-pl assay containing 50 mM HEPES, pH 7.6, 0.1% Triton X100, 5 mM MnC12, 10 mM MgC12, and 10 pCi of -y-[32P]ATPwith a final [ATP] of 50 PM and 0.6 mM insulin receptor Tyr-1150 domain peptide (residues 1142-1153 according to Ref. 39), which was kindly provided by Dr. Steven Shoelson (Joslin Diabetes Center). The reaction was stopped by adding 50 pl of ice-cold 10% (w/v) trichloroacetic acid, precipitated proteins were sedimented by centrifugation, and the supernatant containing the phosphorylated peptide substrate was applied to a 3 X 3-cm piece of phosphocellulose paper (Whatman). The paper was washed four times with 75 mM phosphoric acid, and the retained radioactivity was measured by Cerenkov counting. Analysis of Regional Insulin Receptor Dephosphorylation by Tryptic Peptide Mapping-Insulin receptors were autophosphorylated with -y-[32P]ATP,purified over a P-6 spin column, incubated with recombinant PTPases, andimmunoprecipitated exactly as described above except using Pansorbin (Calbiochem) to adsorb the immune complexes. After washing, the pellets were boiled in sample buffer containing 100 mM DTT (37), fractionated on a SDS-7.5% polyacrylamide gel. The gel was electroblotted onto a nitrocellulose filter, and while keeping the filter damp, the 95-kDa insulin receptor #?-subunit

was visualized by brief autoradiography on Kodak X-Omat AR film and excised. The filter piece was treated with 1 mlof 0.5% (w/v) polyvinylpyrrolidone-40 (Sigma) in 0.1 M acetic acid a t 37 "C for 60 min, washed extensively with water, cut into2-mm square fragments, and incubated in 0.1 ml of 0.1 mg/ml TPCK-treated trypsin in 100 mM NaHC03, pH 8.2, containing 5% (v/v) acetonitrile for 24 h at 37 "C. At that time, an additional 10 pgof trypsin was added, the incubation was repeated for another 24 h, 0.1 ml of a 2-fold concentrated gel sample buffer was added, and thepeptides were denatured by boiling.The phosphopeptides were fractionated by Tricine-sodium dodecyl sulfate gel electrophoresis as described by Schagger and von Jagow (401, using a 32-cm acrylamide gel consisting of 3% stacking, 10% separating, and 16.5% resolving gels. The gels were sealed with plastic wrap and exposed to Kodak X-Omat AR film with a Du Pont Cronex Lightning Plus intensifying screen at -80 "C.Autoradiograms were quantitated by scanning on a Molecular Dynamics computing densitometer. Statistical analyses using the two-tailed Mann-Whitney test were performed with Instat version 1.0 software (Graphpad, San Diego, CA). RESULTS AND DISCUSSION

Relationship Between Insulin Receptor Dephosphorylation by PTPases and Inactivation of the Receptor Kinase-We recently demonstrated that LAR, LRP, and PTPaselB were active in thedephosphorylation of intact, autophosphorylated insulin receptors (23). However, these studies did not assess the potential specificity of these candidate PTPases for certain regulatory sites on the autophosphorylated insulin receptor or, importantly, differences in their possible functional role in reversing the tyrosine kinase activity of the insulinactivated receptor. In order to compare the relative ability of each of the PTPases to deactivate the receptor kinase, we examined the time course of dephosphorylation of the intact receptors as well as the inactivation of the receptor kinase toward an exogenous peptide substrate after dephosphorylation of insulin-stimulated receptors by recombinant PTPases (Fig. 1). The loss of receptor kinase activity induced by PTPaselB or LRP generally paralleled the overall dephosphorylation of the receptor P-subunit throughout the entire reaction. LAR, on the other hand, demonstrated a rapid initial inactivation of the receptor kinase at the earliest stages of the dephosphorylation reaction. The initial rate of kinase deactivation was then calculated from the time course studies and normalized to the rate of receptor dephosphorylation in five replicate experiments (Table I). For a similar degree of receptor dephosphorylation, LAR was found to inactivate the receptor kinase 3.1 and 2.1 times more rapidly than PTPaselB and LRP, respectively ( p < 0.03), suggesting that LAR might initially catalyze the dephosphorylation of certain regulatory phosphotyrosine residues in the receptor kinase domain. The enzymatic activity of PTPaselB and LRP appeared to be less discriminating and more slowly inactivated the receptor kinase as multiple sites on the receptors were dephosphorylated. Dephosphorylation of Specific Domains of the Insulin Receptor by PTPases-We then tested the hypothesis that LAR preferentially dephosphorylated regulatory sites on the insulin receptor by performing phosphopeptide mapping of tryptic fragments of autophosphorylated insulin receptors after treatment with each of the PTPases (Fig. 2). The identity of tryptic phosphopeptides representing the Tyr-1150 and C terminal domains fractionated by this technique has been established' by analysis of insulin receptors with site-directed mutations in specific autophosphorylation sites as well as by comparison with results obtained from phosphopeptide maps

'E. P. Feener, T. Shiba, K-Q. Hu, P. A. Wilden, M. F. White, and G. L. King, submitted for publication.

Insulin Receptor Protein-tyrosine Phosphatases

*** 0

75

25

5

75

-.

" "

LAR

-

13813

PTPaselB

FIG.2. Dephosphorylation of specific insulin receptor do-

Ob

5

10

15 25 20

30

'

mains analyzed by tryptic phosphopeptide mappingof labeled insulin receptors in Tricine-SDS-polyacrylamide gels. Insulin receptors were autophosphorylated with -y-["'P]ATP and incubated for the indicated period of time with recombinant PTPases as described under "Experimental Procedures." After immunoprecipitation, the receptors were purified by SDS-polyacrylamide gel electrophoresis and blotted onto a nitrocellulose filter. The labeled receptor ,%subunit was excised, digested twice with TPCK-treated trypsin, and phosphopeptides were fractionated in Tricine-SDS-polyacrylamide gels (40).Representative autoradiograms of phosphopeptide maps produced by receptor treatment with LAR or PTPaselB are shown. The identity of tryptic phosphopeptides derivedfrom the receptor C terminus as well as the tris(P) and bis(P) forms of the Tyr-1150 domain have been verified by Feener et aLZ

Time (min)

FIG.1. Time course of insulin receptor dephosphorylation and loss of receptor kinase activity after treatment of activated receptors with recombinant PTPases. Human insulin receptors were autophosphorylated by insulin stimulation in the presence of T-[~'P]ATP,purified by spin column chromatography, incubated with individual recombinant PTPases for the indicated period of time, and immunoprecipitated with a monoclonal antibody. After electrophoresis in SDS-polyacrylamidegels, the degree of receptor dephosphorylation was quantitated by densitometric scanning of gel autoradiograms. Receptor kinase activity toward a synthetic peptide substrate was also assayed in parallel reactions as described under "Experimental Procedures." 0, receptor phosphorylation, 0, receptor kinase activity. Panel A, PTPaselB; panel B, LRP; panelC, LAR.

I

I

O

40

1

20

1

TABLE I Relative effect of PTPases on insulin receptor (ZR) dephosphorylation and deactivation of receDtor k i m e activitv PTPase

O' 0

2

4

6

78

2

4

6

8

Relative initial rate (kinase loss/IR dephosphorylation)"

LAR 4.06 f 0.81 PTPaselB 1.29 f 0.17 LRP 1.93 f 0.35 a Mean f S.E.;n = 5. * p< 0.03 versus PTPaselB or LRP.

1

20 1 I

0

generated by reverse phase high pressure liquid chromatography (3) and thinlayer electrophoresis (6). The fully autophosphorylatedinsulin receptors exhibit only the 3-Tyr(P) form of the Tyr-1150 region, which appears as two peptides because of alternative tryptic cleavage sites (Fig. 2). After treatment with recombinant PTPases, the loss of labeling from multiple phosphopeptides is evident, including the dephosphorylation of both peptides representing the 3Tyr(P) form of the Tyr-1150 domain and the appearance of the 2-Tyr(P) form of the Tyr-1150 domain with continued PTPase exposure. In this representative experiment, LAR demonstrated a striking, rapid conversion of the 3-Tyr(P) to the 2-Tyr(P) form of the receptor Tyr-1150 domain, in con-

\

40

0

-

;

"4

6

I

8

Time (min)

FIG.3. Time course of regional insulin receptor dephosphorylation after treatment of autophosphorylatedreceptors with recombinant PTPases. Tricine-SDS gel electrophoresis of tryptic insulin receptor phosphopeptides was performed as described in Fig. 2 after treatment of activated insulin receptors with recombinant PTPases for the indicated period of time. The amount of radioactivity remaining in the receptor C-terminal fragment ( 0 )and the 3-Tyr(P) form of the Tyr-1150 domain (0)was quantitated by densitometric scanning. Panel A, PTPaselB; panel B, L R P panel C, LAR. trast to the much slower conversion of this domain induced by treatment with PTPaselB. Since these results could result from variation in the

Insulin Receptor Protein-tyrosine Phosphatases

13814 TABLE I1

Relative initiul rateof dephosphorylationof insulin receptor 3Tvr(P)-1150 and C-terminal domainsbv recombinant PTPases PTPase

Relative initial dephosphorylation rate (3-Tyr(P)-1150/C-terminal domains)'

LAR 5.79 2 1.49' PTPaselB 1.64 f 0.38 LRP 1.55 f 0.15 Mean 2 S.E.; n = 5. 'p < 0.01 versus PTPaselB or LRP.

amount of each PTPase enzyme used in the incubation, we compared the dephosphorylation of the 3-Tyr(P)-1150 domain to thedephosphorylation of the receptor C terminus as an internal control (Fig. 3). Time courses of dephosphorylation of these two regions of the receptor generated by treatment with each of the recombinant PTPases were quantitated by densitometric scanning of the phosphopeptide mapping gels. The initial rate of dephosphorylation of the 3-Tyr(P)1150 domain was then normalized to the rate of dephosphorylation of the receptor C terminus in five separate experiments (Table11).PTPaselB and LRPdemonstrated asimilar initial rate of dephosphorylation of the receptor 3-Tyr(P)1150and C-terminaldomains. In contrast,LAR preferentially dephosphorylated the 3-Tyr(P)-1150 domain over the simultaneous dephosphorylation of the receptor C terminus at a relative rate calculated to be 3.5 and 3.7 times more rapid than that demonstrated by PTPaselB or LRP, respectively ( p e 0.01). Implications for the Regulation of Insulin Receptor Function-This study is the first examination of insulin receptor kinase regulation and site-specific receptor dephosphorylation by candidate PTPases that are expressed in insulin-sensitive tissues. The observation that LAR deactivated the insulin receptor kinase more rapidly than either PTPaselB or LRP was substantiated by finding that LAR also preferentially dephosphorylated phosphotyrosine residues in the 3-Tyr(P) form of the receptor Tyr-1150 domain that have been implicated in the regulation of the receptor kinase. Among normal rat tissues, we have shown that mRNA for LAR is predominantly expressed in liver and is one of the major PTPase homologs expressed in skeletal muscle (14, 24, 41). In contrast, PTPaselB and LRP,while also having high levels of mRNA expression in liver and muscle, appear to be morewidely distributed (23, 24,26,27,30, 42). The tissue expression of LAR along with its accelerated deactivation of the insulin receptor kinase and relative preference for regulatory phosphotyrosine residues support a potential role for this PTPase in thephysiological regulation of insulin receptors in intact cells. Furthermore, as a transmembrane protein that is likely to be glycosylated (25), LAR also has several characteristics of the major insulin receptor dephosphorylating activity present in extracts of liver and hepatoma cells, which is localized to the particulate fraction and is enriched in specific activity by lectin affinity chromatography (13, 14). It may be important to note that themembrane-bound form of LAR that presumably exists in situ may react differently with the insulin receptor than thesoluble cytoplasmic domain studied in this report. Further work involving expression of cDNA constructs for LAR and other candidate PTPases in insulin-sensitive cells will help to provide more direct data

regarding their potential role in the regulation of signaling through the insulin receptor. Acknowledgments-We thank Drs. Steven Shoelson and Peter Wilden for helpful discussions and Dr. Morris White for providing the insulin receptor transfected Chinese hamster ovary cells. REFERENCES 1. Rosen 0.M. (1987) Science 237,1452-1458 2. Kahn,'C. R., and White, M. F. (1988)J. Clin. Inuest. 82,1151-1156 3. White M. F. Shoelson S. E. Keutmann, H., and Kahn, C.R. (1988)J. Bioi Chem: 263,2969-2986 4. Tornqvist, H. E., Pierce, M. W., Frackelton, A.R., Nemenoff, R. A., and Avruch, J. (1987)J. Biol. Chem. 262,10212-10219 5. Flores-Riveros, J. R., Sibley, E., Kastelic, T., and Lane, M. D. (1989)J. Biol. Chem. 264,21557-21572 6. Tavare, J. M., and Denton, R. M. (1988)Biochem. J. 262,607-615 7. Avruch, J., Torn vist, H. E., Gunsalus, J. R., Yurkow, E. J., Kyriakis, J. M., and Price,%. J. (1990)in Handbook of Experimental Pharmacology (Cuatrecasas, P., and Jacobs, S., eds) Vol. 92, pp. 313-366, SpringerVerla Berlin 8. Myers, G., Backer, J. M., Siddle, K., and White, M. F. (1991)J . Biol. Chern. 266,10616-10623 9. Maegawa, H., McClain, D. A,, Freidenberg, G., Olefsky, J . M., Napier, M., Lipari, T., Dull, T. J., Lee, J., and Ullrich, A. (1988)J. Biol. Chem. 263, 8912-8917 10. Takata, Y., Webster, N. J. G., and Olefsky, J. M. (1991)J. BWL Chem. 266,9135-9139 11. Goldstein, B. J. (1992)J . Cell. Biochem. 48,33-42 12. Fischer, E. H., Charbonneau, H., and Tonks, N. K. (1991) Science 263,

&.

"_ "_

41-11 -4013

13. Sale, G. J. (1991)Adu. Protein Phos hatases 6, 159-186 14. Goldstein, B. J., Meyerovitch, J., Zgang, W. R., Backer, J. M., Csermely, P., Hashimoto, N., and Kahn,C. R. (1991)Adu. Protein Phosphatases 6, 1-17 15. King, M. J., Sharma, R. P., and Sale, G. J. (1991)Biochem. J. 276, 413418 16. Tappia, P. S., Sharma, R. P., and Sale, G. J. (1991)Biochem. J. 278, 6974 17. Tonks, N. K., Diltz, C. D., and Fischer, E. H. (1988)J. Biol. Chem. 263,

-."

fi79l-fi727 ".-,

18. Cicirelli, M. F., Tonks, N. K., Diltz, C. D., Weiel, J. E., Fischer, E. H., and Krebs, E. G. (1990)Proc. Natl. Acud. Sci. U. S. A. 87,5514-5518 19. Tonks, N. K., Diltz, C. D., and Fischer, E. H. (1990)J. Btol. Chem. 266, 10674-10680 20. Cho H J h e r S. E Itoh M Winkler D. G Kitas E Bannwarth d.,Bu& P., Sdto, H,: and'wa'ish, C. T. (l99l)'Biochehi.s&y 30,6210; G 9 l fi

21. ChaiGgee, S. Goldstein, B. J., Csermely, P., and Shoelson, S. E. (1992)in Peptdes: Chemistry and Biology (Rivler, J. E., and Smith,J. A., eds) pp. 553-555,Escom Science Publishers, Leiden, Netherlands 22. Cho, H. J., Ramer, S. E., Itoh, M., Kitas, E.? Bannwarth, W., Burn, P., Saito, H., and Walsh, C. T. (1992)Btochemlstry 31,133;138 23. Hashimoto, N., Zhang, W. R., and Goldstein, B. J. (1992)Bwchem. J. 284, in press 24. Zhang, W. R., and Goldstein, B. J. (1991)Biochem. Biophys. Res. Commun. 178,1291-1297 25. Streuli, M., Krue er, N X , Hall, L. R., Schlossman, S. F., and Saito, H. (1988)J . EX g e d . ies; 1523-1530 26. Matthews, R. Cahir, E.D., and Thomas, M. L.(1990)Proc. Natl. Acud. Sei. U. S. A. 87,4444-4448 27. Sap, J., DEustachio, P., Givol, D., and Schlessinger, J. (1990) Proc. NatL Acud. Sci. U. S. A. 87,6112-6116 28. Krueger N.X. Streuli M and Saito H. (1990)EMBOJ. 9 3241-3252 29. Guan, d. L., daun, R.'S., 'Watson, S.'J., Geahlen, R. L., and Dixon, J. E. (1990)Proc. NatL Acud. Sei. U. S. A. 87,1501-1505 30. Chernoff J Schievella A. R. Jost C.A. Erikson R. L., and Neel, B. G. (199O)'P&. NatL Ach. Sci.' U. S: A. S f , 2735-2$39 31. Pallen. C. J.. Lai. D. S.. Chia. H. P.. Boulet., I.., and Tong. -_P. H. (1991) . . Bioihem. J . 276,315-323 ' 32. Woodford. T. A.. Guan. K. L.. and Dixon. J. E. (1991) . . Adu. Protein ' Phosph;rtases 6',503-524 33. Frangioni J. V Beahm P. H., Shifrin, V., Jost, C. A., and Neel, B. G. (1992)bell 68,545-560 34. Boyd, D., Manoil, C., and Beckwith, J. (1987)Proc. Natl. Acud. Sci. U. S. A. 84,8525-8529 Natl. 35. Streuli, M., Krueger, N. X., Tsai, A. Y., and Saito, H. (1989) Proc. Acud. Sci. U. S. A. 86,86984702 36. Kasuga, M., White, M. F., and Kahn, C. R. (1985)Methods Enzymol. 109, 609-621 37. Laemmli U. K. (1970)Nature 227,680-685 Bradford' M. M. (1976)Anal. Biochem. 72,248-254 38. Bell, J. R., Chen, E. Y., Herrera, R., Petruzzelli, L. M., Dull, T. 39. Ullrich, J.. Grav. A.. Coussens. L.. Liao. Y.-C.. Tsubokawa. M.. Mason. A.. Seeburi,'P. H . , Grunfeld, C.;Rosen, 0. M:, and Ramachandran,J. (i985j Nature-313, 756-761 40. Schiigger H and von J ow G. (1987)Anal. Biochem. 168,368-379 41. Goldsteii. B.'J.. Zhane. R..Hashimoto.. N... and Kahn. C. R. (1992)Mol. Cell. B&heni. 109,i07-113 42. Jirik. F. R.. Janzen. N. M.. Melhado. I. G.. and Harder. K. W. (1990)FEBS Lett. 273,239-242

f;

A.,

3.