Inhibition or Down-regulation of Protein Kinase C Attenuates Late ...

Vol. 267, No.10,Issue of April 5, pp. 6905-6909,1992 Printed in US.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Inhibition or Down-regulation of Protein Kinase C Attenuates Late Phase p7OSgkActivation Induced by Epidermal Growth Factor but Not by Platelet-derived Growth Factor or Insulin* (Received for publication, October 23, 1991)

Mira Sub$, Duiica Vulevic, HeidiA. Lane, and George Thomas8 From the Friedrich Miescher Institute, P. 0. Box 2543, 4002 Basel, Switzerland

The late phaseofthetime-dependentepidermal growth factor(EGF)-inducedbiphasic activation of the ~ 7 0 ” ’ is selectively attenuatedbythe specific PKC inhibitor, CGP 41 261, a staurosporine derivative. At a 40-fold lower concentration than CGP 41 261, staurosporine inhibits both phases of S6 kinase activation to the same extent, whereas the inactive staurosporine derivative CGP 42 700 shows no effect on eitherphase.Platelet-derivedgrowth factor S6 kinase (PDGF)andinsulinalsoinducebiphasic activation, but in neither case is either phase of activation affected by the presence of CGP 41 251. This finding was unexpected in thecase of PDGF, whichis a potent activator of PKC and whose receptordirectly interacts with phospholipase C,1. However,similar results were obtained following down-regulation of PKC byprolonged 12-O-tetradecanoylphorbol-13acetate treatment. Therefore, even though EGF and PDGF induce PKC activation,PDGF, unlike EGF, does not appearto use this signaling pathway forlate phase p7OSBk activation.

to bind, phosphorylate, and stimulate phospholipase C,’ (911), leading to theproduction of diacylglycerol and the activation of PKC (12, 13). Recently the use of TPA-induced down-regulation of PKC in distinguishing this pathway has been placed in question by the finding that a number of TPA effects may instead be attributable to membrane perturbations (14) or to antagonistic effects on Ca2+channels (15). In addition, othersin the field have failed to establish a link between EGF-induced ~70”~ activation ’ and PKC by using TPA pretreatment (16, 17). These observations prompted us to test the effect of specific PKC inhibitors on EGF-induced activation of the p70*‘jkas well as their effect on two additional mitogens, one of which is coupled to the PKC pathway (PDGF (11))and one which is not coupled to this pathway (insulin (18)).The results arediscussed with regard to known signaling pathways. EXPERIMENTAL PROCEDURES

Protein Kinase C Inhibitors and Growth Factors-Sphingosine and H7 were purchased from Sigma and stored as 25 mM stock solutions in absolute ethanol at -20 “Cuntil used at the concentrations indicated in the text. Staurosporine, CGP 41 251, and CGP 42 700 were gifts from Drs. T. Meyer, U. Regenass, and G. Caravatti, Ciba-Geigy, Previously we demonstrated that treatment of quiescent Basel, and were stored as 10 mM stock solutions in dimethyl sulfoxide a t -20 “C and applied to cells at the concentrations indicated in the Swiss 3T3 cells with EGF’ results in a time-dependent bitext. Since staurosporinederivatives appeared to be sensitive to light phasic activation of the M,70,000 40 S ribosomal protein S6 and repeated freeze-thawing, they were stored in small aliquots in kinase (1, 2) or ~ 7 0This ~ ~kinase . is able to phosphorylate dark tubes. TPA was from Sigma, EGF from Bioprocessing, PDGF four of the five S6 sites observed in vivo (3,4), is activatedby (c-sis) from Amersham (AMGEN Biologicals), and insulin from Ser/Thr phosphorylation (5))and, by sequencing of two cDNA Calbiochem. Cell Culture and Extraction-Swiss 3T3 fibroblasts weregrown clones, has been shown to be a novel member of the second messenger family of protein kinases (6,7). Thekinase respon- and cell extracts were prepared as previously described (1). S6 Kinase Assays in Crude Cell Extracts-The extracts were diluted sible for activating the ~ 7 0 remains ’ ~ ~ unidentified (8).It has 10-fold in dilution buffer and assayed for S6 kinase activity as been shown, however, that down-regulation of protein kinase described (1). C (PKC) by prolonged treatment with 12-O-tetradecanoyl- previously S6 Kinase Assays in Zmmunoprecipitates-Immunoprecipitation phorbol-13-acetate (TPA) severely suppresses the EGF-in- was performed with a rabbit monospecific polyclonal antibody against duced late phase S6 kinase response, with no effect on the a synthetic peptide derived from the sequence of the p70e6’, referred early phase (1).Loss of the latephase was accompanied by a to as M5 (33). Briefly, parallel confluent cultures on6-cm plates were reduction in S6 phosphorylation, initiation of protein synthe- washed twice with ice-cold wash buffer (120 mM NaCl, 50 mM Trissis and cell growth (1).This finding is consistent with the HC1,20 mM NaF, 1mM benzamidine, 1 mM EDTA, 5 mM EGTA, 10 ability of the ligand-activated EGF tyrosine kinase receptor mM sodium pyrophosphate, 30 mM p-nitrophenylphosphate, and 0.1 mM phenylmethylsulfonyl fluoride, pH 8.0 at 4 “C),lysed in 500 pl of * The costs of publication of this article were defrayed in part by the same buffer supplemented with 1%Nonidet P-40, and centrifuged the payment of page charges. This article must therefore be hereby a t 4 “C for 10 min a t 10,000rpm (Sorvall RB centrifuge, SS-34 rotor), marked “aduertisement” in accordance with 18 U.S.C. Section 1734 and the supernatant was then frozen in liquid nitrogen. Aliquots of 200 pl were then incubated on ice for 2 h with 5 r l of antiserum. solely to indicate this fact. 4 Present address: St. Elizabeth’s Hospital, Tufts University Next, 20 pl of 50% protein A-Sepharose was added and the mixture shaken for 30 min a t 4 “C, which was followed by centrifugation at School of Medicine, 736 Cambridge St., Boston, MA 02135. 5 TOwhom correspondence should be addressed. Tel. 41-61-697- top speed in an Eppendorf centrifuge (model 5412). Pellets were washed three times with wash buffer containing 1%Nonidet but not 3012; Fax: 41-61-697-3976. The abbreviations used are: EGF, epidermal growth factor; PKC, EDTA, EGTA, sodium pyrophosphate, then washed once with diluprotein kinase C; TPA, 12-O-tetradecanoyl-phorbol-13-acetate;tion buffer without dithiothreitol ( l ) , and finally resuspended in 5 p1 PDGF, platelet-derived growth factor; p7OMk,M,70,000 S6 kinase; of dilution buffer and assayed in the same buffer as described for EGTA, [ethylenebis(oxyethylenenitrilo)]tetraaceticacid. total cell extracts (1).

6905

6906

~ 7 0Activation " ~ ~ RESULTS AND DISCUSSION

Because of their key role in a number of metabolic functions and carcinogenesis, the PKCfamily of kinases has become an attractive targetfor the development of pharmaceutical inhibitors. Initially we tested several of the well known PKC inhibitors for their ability to selectively block the EGF-induced late phaseS6 kinase response, including sphingosine, a long chain lipid base (19), H7, an isoquinoline-sulfonamide derivative (20), staurosporine, a bacterial alkaloid (21), and CGP 41 251,a newly developedcompound which is a chemical derivative of staurosporine (22). Sphingosine is a competitive inhibitor of diacylglycerol and phorbol ester binding, and the other three are argued to act by preventing ATP binding. Inhibitor concentrations used were in the range reported to block PKC action in uivo (19-22). As can be seen in Table I, 60 pM sphingosine produced a moderate inhibition and 1.6 pM staurosporine an almost complete block of late phase S6 kinase activation, as measured at 45 min following EGF stimulation. Bothagents, however, led to large changes in cell morphology, with cells rounding up and beginning to detach from the plate within 1 h (Table I). After 24 h all cells had detached from the plate and were no longer viable as judged by trypan blue staining (data not shown). These results are consistent with previous reports showing that sphingosine can be cytotoxic (23) and that staurosporine is avery potent but nonspecific protein kinase inhibitor also blocking receptor tyrosine kinase activity (21, 22). It has been reported that delivery of sphingosine together with bovine serum albumin reduces its cytotoxic effects, while preserving its PKC inhibitory properties (23). Such treatment retardedcell detachment in the first hour, but by 24 h no cells were attached to the plate. Unlike sphingosine and staurosporine, 50 p~ H7 had no effect on cell morphology, attachment, or the late phase S6 kinase response (Table I), even at concentrations as high as 150 p~ (data not shown). This finding is consistent with the recent suggestion that H7 may only lead to partial inhibition of PKC (24). In contrast to the compounds described above, the newly described inhibitor CGP 41 251 led to an inhibition of the late S6 kinase response with no observed morphological or toxic effects even after 3 days of treatment (Table I, and data notshown). To determine the specificity of staurosporine derivative CGP 41 251, cells were pretreated with increasing concentrations of either the inhibitor (staurosporine) or CGP 42 700 (an inactive derivative of staurosporine) (22), followed by stimulation with EGF for either 10 min or 45 min. Staurosporine exhibited no selectivity for either phase and at the lowest concentration tested,0.16 p ~ completely , abolished S6 kinase activation at both times (Fig. lA). In contrast, the inactive derivative had no effect over the concentration range tested at either time point (Fig. 1B). Similarly, CGP 41 251 had no effect on the early S6 kinase response, but did selecTABLEI Effect of PKC inhibitors on S6 kinase activity and cell attachment Cells were pretreated with inhibitors for 30 min and stimulated with 1 nM EGF for 45 min. Cell attachment was evaluated microscopically, cells were extracted, and the extracts assayed for S6 kinase activity (1).Results presented are the means k S.D. from two separate experiments. PKC inhibitor

S6 kinase % of control

None Sphingosine (60 pM) H7 (50 p M ) staurosporine (1.6 pM) CGP 4 1 251 (1.6 pM)

100 41 f 11 94 f 11 12 f 6 17 f 5

Cell attachment

+++

0 .I6 .32

.61

1.6 3.2 6.40 .w5.05 .5

5

Log inhibitor concentration (pM)

FIG. 1. In vivo and in vitro effect of staurosporine and its derivatives on S6 kinase activity. Quiescent Swiss mouse 3T3 fibroblasts were treated for 30 min with the indicated concentrations of staurosporine (A), CGP 42 700 ( B ) ,or CGP 41 251 (C) and then treated with either 1 nM EGF for 10 min (A)and 45 min (0)or with 500 nM TPA for 60 min (A). Cell extracts were prepared and assayed for S6 kinase activity as previously described (1). A 100% S6 kinase activity refers to values obtained for EGF or TPA stimulation in the absence of inhibitors, after subtraction of the basal enzyme activity (EGF, 10 min: 41.6 +. 5.0 units; EGF, 45 min: 34.0 f 3.9 units; TPA, 60 min: 28.1 f 5.1 units; no treatment: 4.1 f 1.8 units). Results represent the mean of at least two independent experiments. Extracts from either 10 min (A)or 45 min (0)EGF-stimulated cultures were assayed for S6 kinase activity in the presence of staurosporine (D), CGP 42 700 ( E ) ,or CGP 41 251 (F).1 unit of kinase activity equals 1 pmol phosphate incorporated into S6 per min per mg of protein extract.

tively inhibit the late phaseresponse, ICso= 0.6 ~ L (Fig. M IC). That theinhibitory effect of CGP 41 251 is mediated through PKC is further supported by the finding that TPA induction of S6 kinase activity is blocked in a dose-dependent manner similar to thatfound for EGF (Fig. 1C). It has been reported that CGP 41 251 has reduced inhibitory activity towards PKC in vitro as compared with staurosporine, but that it is more specific for PKC than for other kinases, including the EGF receptor and the~ 7 0 (22). " ~ In ~ agreement with these results, the ICm of staurosporine for S6 kinase in vitro is approximately 200-fold higher than that of CG 41 251 (Fig. 1, D and F ) , and CGP 42 700 has no effect on S6 kinase activity at the concentrations tested (Fig. 1E). Furthermore, CGP 41251 inhibits S6 kinase activity in extracts prepared from cells stimulated for 10 or 45 min, but with equal efficiency (Fig. lF), and therefore it is unlikely that theeffect of the drug in uiuo is exerted directly on late phase S6 kinase activity. To ensure that the inhibitory effect of CGP 41251 was selective for late phase S6 kinase activation, cultures' were stimulated for increasing times for up to 2 h with EGF in the absence or presence of 1.6 p~ inhibitor, a concentration which resulted in an "80% inhibition of S6 kinase activity at 45 min (Table I, Fig. IC). The datashow that thedrug acts only at late times (Fig, 2), similar to previous results obtained following down-regulation of PKC by TPA (1). This selective inhibition of late phaseS6 kinase activation is not a function of the length of time cultures are exposed to the inhibitor,

6907

p 7 P kActivation

100

80 60 40

:I 0

30

60

90

120

Time (rnin)

FIG. 2. Inhibition of EGF-induced late phase S6 kinase response by CGP 41 251. Quiescent Swiss 3T3 cells were pretreated for 30 min with 1.6 p~ CGP 41 251 (A)or without the inhibitor (0) and then stimulated with 1 nM EGF for the times indicated. Cells were extracted and the extracts assayed for S6 kinase activity (1). Results represent the mean of two independent experiments.

since the same results were obtained when cells were pretreated with CGP 41 251 for 10 min and 1h (data notshown). Similar to TPA-induced down-regulation of PKC (l), CGP 41 251 does not completely inhibit the late phase of S6 kinase activation (Fig. 2), even when higher concentrations of inhibitor were used (Fig. IC, compare EGF and TPA). Inaddition, under either condition used to block PKC action, stimulation of cells with TPA no longer induced S6 kinase activity. Finally, the kinetic appearance of the PKC-independent late phase S6 kinase response is distinct from that of the early phase, indicating thatthere is a third signaling pathway leading to S6 kinase activation. To ascertain whether S6 kinase activation by the PKCdependent pathway is a general response to mitogens, we examined the effect of PDGF andinsulin on this event. PDGF is a potentactivator of PKC (25, 26) and, like EGF, its receptor is directly coupled to phospholipase CV1( l l ) , whereas extensive efforts to demonstrate a connection between insulin and the PKCsignaling pathway have failed (18). Thus, itwas reasoned that PDGF would use the PKC pathway in activating the S6 kinase, while insulin would employ a different route. To test thispossibility parallel cultures were stimulated with PDGF, insulin, EGF, or TPA for 10 or 45 min in the absence or presence of CGP 41 251. The results show that with the exception of TPA, the inhibitor had no effect on the ability of any of the mitogens to induce S6 kinase activation at the early time (Fig. 3 A ) , and that only the EGF andTPA responses were attenuated at the later time (Fig. 3B). The results suggest that PDGF, like insulin, does not use the PKC signaling pathway for the activation of the S6 kinase. The inability of CGP 41 251 to block PDGF-induced S6 kinase activation was unexpected, since PDGF, like EGF,modulates many of its early mitogenic effects through PKC (25, 26). Such effects include the increased phosphorylation of both the M , 80,000 MARCKS protein and pp42 (27), which has been identified as the mitogen-activated protein kinase, or MAP kinase (28). To ensure that the absence of PDGFinduced S6 kinase activation throughPKC was not restricted to theuse of CGP 41 251, similar measurementsof S6 kinase activity were performed following PKC down-regulation by TPA. The results show that S6 kinase activation in control cells exhibited time-dependent biphasic kinetics of activation

20

100

80 60 40 20

1

2

3

4

FIG. 3. CGP 41 261 does not inhibit S6 kinase induced by PDGF and insulin. Quiescent Swiss 3T3 cells were pretreated for 30 min with 1.6 p~ CGP 41 251 and then stimulated with either 1 nM PDGF (I), 10 nM insulin (Z),1 nM EGF (3), or 500 nM TPA (4) for 10 min (panel A ) or for 45 min (panel B ) . Extracts were then prepared and assayed for S6 kinase activity (1). A 100% S6 kinase activity refers to values obtained from cells not treated with CGP 41 251, after correction for the basal S6 kinase activity (PDGF, 10 min: 42.2 -e 4.0 units, or 45 min: 40.3 & 9.0units; insulin, 10 min: 40.6 k 11.2 units, or 45 min: 33.2 & 6.0 units; EGF, 10 min: 32.8 f 6.8 units, or 45 min: 26.6 ? 4.0 units; TPA, 10 min: 37.5 f 10.1 units, or 45 min: 16.6 f 3.5 units; no stimulation 7.0 & 4.6 units. Results represent the mean of at least two independent experiments.

similar to that observed with EGF (Fig. 4). However, the response elicited by PDGF (Fig. 4) as compared with EGF (Fig. 2 and Ref. 1)was stronger and of longer duration. This finding is in accordance with PDGF being a more potent mitogen in these cells (25). Consistent with the results obtained with CGP 41 251, PKC down-regulation did not significantly alter the early or late phase S6 kinase response (Fig. 4). Similar results were obtained with insulin (data not shown), arguing again that insulin does not exploit the PKC pathway to induce S6 kinase activation. Since PDGF stimulates S6 kinase activity to a higher level than EGF, the lack of inhibition of PDGF-induced S6 kinase activation in the presence of CGP 41 251 could be a consequence of multiple redundant signaling pathways leading to S6 kinase activation. To test thispossibility, we also measured the effect of CGP 41 251 or long term TPA treatment on S6 kinase activation at concentrations of PDGF which led to a similar degree of kinase activation as observed with EGF. However, no inhibition of either the early or late phase S6 kinase response was observed (data not shown). Recently it was demonstrated that thepp9OTSk family of S6 protein kinases is also activated during the G,/G, transition of the cell cycle (reviewed in Ref. 29). Although it has been argued by Blenis and co-workers (30) that thepp90mkactivity

p7@'jkActivation

6908

" " " " "

"

. A

2. ul

Log PDGF (nM)

30-j

0.1 -,.'

c

P

(D

UJ

I

0.51 I

I

5 I

20-3 10-

g

20

u)

UJ

0

I

0

1

30

60

1

90

120

Time (min)

FIG. 4. Down-regulation of PKC does not inhibit PDGFinduced S6 kinase activity. Quiescent cells were pretreated with either 5 p~ TPA (A)or with 0.1% dimethyl sulfoxide (0)and then stimulated with 1 nM PDGF for the times indicated. Extracts were prepared and then assayed for S6 kinase activity (1). Inset, doseresponse curve for S6 kinase activity after a 30-min treatment.

.""""" -70 k

4 7 0k

4S6

1

2

3

4

5

6

7

8

9

1

u

FIG. 5. EGF- but notPDGF-induced ~ 7 0 ' 'activity ~ is inhib251. Quiescent Swiss 3T3 cells were pretreated ited with CGP 41 for 30 min with 1.6 p~ CGP 41 251 (lanes 5, 6, 9, and 10) or not treated with theinhihitor (lunes 1-4, 7, and B ) , followed by no treatment (lunes I and 2 ) or stimulation with either1 nM EGF (lunes 3-6) or 1 nM PDGF (lunes 7-10) for 45 min. Extracts were prepared and incubated with preimmuneserum (oddnumbers) or immune serum (even numbers), immunoprecipitated with protein A-Sepharose, and assayed for S6 kinase activity as described under "Experimental Procedures." The lower arrowhead indicates the position of S6 protein; the upper arrowhad shows 70-kDa reRion of the gel. Inset, 70-kDa region of the same autoradiograph after longer exposure.

of t h e ~ 7 0 " ' peptide ~ used to generate antibodies (data not shown), arguing that this protein most prohahly represents the autophosphorylated ~ 7 0 " ' (Fig. ~ 5). Thus, in aveement with others (30), the S6 kinase activity followed in whole cell extracts apparently represents ~70""'. In addition, the fact that suppression of the PKC pathway significantly reduces PDGF-inducedMAP2kinasephosphorylation (27, 28) but not p7Wfikactivation, supports the recent argument that these two enzymes lie on distinct signaling pathways(8). The results presented here suggest the existenceof a t least one additional signaling pathway leadingto ~ 7 0 "activation ~ by EGF. It is not clear whether this pathway is the same employed by either PDGFor insulin. Indeed, the only signaling pathway which can be clearly assigned is that of PKC in the EGF-induced late phase ~ 7 0 response. " ~ ~ The failure of others to detect this EGF-induced pathway may be due to the time points examined, since no assays were conducted a t 3060 min post-EGF stimulation in PKC down-regulated cells (16, 17). As stated earlier, the results showinga link between EGFbut not PDGF-induced ~ 7 0 " activation "~ and PKC were surprising, especially in light of previous reports indicating that PDGF is a much stronger activator of PKC than EGF (2S, 26). This result, however, is consistent with the fact that TPA-induced down-regulation of PKC inhihh EGF (1) hut not PDGF-induced cell growth (32). Together these findings indicate that a small activation of a specific isoform of PKC by EGF may be more significant for mitogenesis than extensive activation of otherPKC isoforms by PDGF.Inthis context it would be of interest to know whetherthere is differential activationof PKC isoforms by EGF uersur PDGF. In the future such studieswould be greatly facilitated by the identification of the ~70""' kinase(s) responsible for modulating this response. Indeed, the possibility still exists that the ~ 7 0 " is ~ ' directly activatedby a unique PKC isoform, since to date we have only tested the a, 8, and y forms.' Obviously, the questionswhich now need to be addressed are the location of the phosphorylation sites leading to ~ 7 0 " 'activation ~ and whether all mitogens lead toactivationthroughthesame sites. Acknowledgments-We are indebted to Peter Hides for his technical assistance in the latter part of these studies. In addition, we would like to thank Dm. T. Meyer, U. Regenass. and G. Caravatti (Ciha-Geigy, Rasel) for their expertise in the synthesis and use of staurosporine derivatives. Finally. we would like to thank Dm. S. Kozma, A. Matus, and D. Reddy for their critique of this manuscript, and C. Wiedmer for secretarial assistance.

is not easily measured in whole cell extracts without specific REFERENCES antibodies, its presence has raised questions concerning the 1. &;a, M., Olivier, A. R., Fabbro. D., and Thomas, G. (1989) CelI identity of the S6 kinase which is being measured in cell 57,817-824 extracts (31). During these studies monospecific polyclonal 2. $&a, M., and Thomas, G. (1990) Proc. Natl. Acad. Sci. C! S.A. antibodies whichcould immunoprecipitate p70"" activity 87, 7040-7044 from cell extracts became available (33). These antibodies 3. Jeno, P., Ballou, L. M., Novak-Hofer, I., and Thomas. G. (1988) Proc. Natl. Acad. Sci. U.S.A. 8 5 , 406-410 immunoprecipitate activated~ 7 0 " 'from ~ extracts of 3T3 cell cultures treated for 10 min or 45 min with either EGF, insulin, 4. Ferrari, S.,Bandi, H. R., Hofsteenge, d., Russian, R.. and Thomas. G. (1991) J. Hiol. Chem. 266,22770-22775 or PDGF (data not shown). To test whether total S6 kinase 5. Ballou, L. M., Siegmann. M.. and Thomas. G . (1988) Proc. Notl. activity in cell extracts reflects ~70"" activation, cells were Acad. Sci. U.S. A. 8 5 , 7154-7158 stimulated with EGF or PDGF for 45 min in the absence or 6. Kozma, S. C., Ferrari, S., Rassand, P., Siegmann, M., Tottv. N.. presence of CGP 41 251. The results demonstrate that inS6 andThomas, G . (1990) Proc. Natl. Acnd. Sci. 11. S. A. 87, kinase assays of immunoprecipitates, activation of the ~ 7 0 " ~ ~ 7365-7369 7. Banejee, P., Ahrnad, M. F., Grove, J. R.. Kozloskv. C., Price, D. was inhibited in EGF- but not in PDGF-stimulated cells (Fig. .J., and Avruch, .J. (1990) Proc. Natl. Acad. Sct. I / . S'. A. 87, 5 ) , as was observed in whole cell extracts (Fig. 3 B ) . It should 8550-8554 also be noted that the labeling intensity of a 70-kDa band 8. Ballou, L. M., Luther, H., and Thomas, G . (1991) Nature 349. closely paralleled S6 kinase activity (Fig. 5 , inset). This band 348-350 wasnotpresentinsamplesprecipitatedwithpreimmune serum (Fig. 5, inset) and was competed away in the presence * M. his, S. Morley, and G. Thomas, data not shown.

~ 7 0Activation " ~ ~ 9. Wahl, M. I., Daniel, T. O., and Carpenter, G. (1988)Science 241, 968-970 10. Margolis, B., Rhee, S. G., Felder, S., Mervic, M., Lyall, R., Levitzki, A., Ullrich, A., Zilberstein, A., and Schlessinger, J. (1989)Cell 67,1101-1107 11. Meisenhelder, J., Suh, P.-G., Rhee, S. G., and Hunter, T . (1989) Cell 67,1109-1122 12. Nishizuka, Y. (1988)Nature 334,661-665 13. Berridge, M. J., and Imine, R. F.(1989)Nature 341,197-204 14. Sha'afi, R. I. (1989)Bwchem. J. 261,688 15. Hockberger, P., Toselli, M., Swandulla, D., and Lux, H. D. (1989) Nature 338,340-342 16. Pelech, S.L., and Krebs, E. G . (1987)J. Biol. Chem. 262,1159811606 17. Chen, R.-H., Chung, J., and Blenis, J. (1991)Mol. Cell. Biol. 11, 1861-1867 18. Blackshear, P. J., Haupt, D. M., and Stumpo, D. J. (1991)J. Biol. Chem. 266,10946-10952 19. Hannun, Y.A., and Bell, R. M.(1987)Science 236,670-674 20. Hidaka, H., Inagaki, M., Kawamoto, S., and Sasaki, Y. (1984) Biochemistry 23,5036-5041 21. Tamaoki, T., and Nakano, H. (1990)Biotechnology 8,732-735

6909

22. Meyer, T., Regenass, U., Fabbro, D., Alteri, E., Rosel, J., Muller, M., Caravatti, G., and Matter, A. (1989)Znt. J. Cancer 43, 851-856 23. Lambeth, J. D., Burnham, D. N., and Tyagi, S. R. (1988)J. Bwl. Chem. 263,3818-3822 24. Krishnamurthi, S., and Joseph, S. (1989)Biochem. J. 261, 687688 25. Rozengurt, E. (1986)Science 234, 161-166 26. Blackshear, P. J., Witters, L. A., Girard, P. R., Kuo, J. F., and Quamo, S. N. (1985)J. Bwl. Chern. 260,13304-13315 27. Kazlauskas, A., and Cooper, J. A. (1988)J. Cell Biol. 106,13951402 28. Rossomando, A. J., Payne, D. M., Weber, M. J., and Sturgill, T. W. (1989)Proc. Natl. Acad. Sci. U. S. A. 86,6940-6943 29. Erikson, R. L. (1991)J. Biol. Chem. 266,6007-6010 30. Chen, R.-H., and Blenis, J. (1990)J. Mol. Cell. Biol. 10, 32043215 31. Sturgill, T. W., and Wu, J. (1991)Biochim. Biophys. Acta 1092, 350-357 32. Coughlin, S. R., Lee, W. M. F., Williams, P. W., Giels, G . M., and Williams, L. T . (1985)Cell 43,243-251 33. Lane, H. A., Morley, S. J., DorCe, M., Kozme, S. C., and Thomas, G. (1992)EMBO J.,in press