Modulation of growth-related gene expression and cell cycle synchronization by a sialoglycopeptide inhibitor

EXPERIMENTAL

CELL RESEARCH

194,62-66

(1991)

Modulation of Growth-Related Gene Expression and Cell Cycle Synchronization by a Sialoglycopeptide Inhibitor H. K. FATTAEY,*

C. C. BASCOM,? AND T. C. JOHNSON*

*Center for Basic Cancer Research, Division of Biology, Kansas State University, Manhattan, Kansas 66506; and tDepartment of Cell Biokgy, Vanderbilt Medical School, Nashville, Tennesee 37232

tion of growth stimulators, there remains little information on the interaction of naturally occurring growth inhibitors with regard to early events essential to the transition of cells from the G,/G, phase of the cell cycle [l]. The interaction of growth factors and tumor promoters with their specific cell surface receptors initiates a cascade of intracellular events that includes a transient increase in cytosolic pH, mobilization of Ca2+, phosphoinositide metabolism, and protein phosphorylation [2,3]. A coordinate induction of the expression of genes associated with ceil proliferation accompanies these metabolic events, and it is thought that the products of these genes play a central role in the initiation and/or maintenance of the proliferative state. In some cases, the expression of these growth-related genes has been altered by the presence of growth-inhibitory levels of transforming growth factor fl and tumor necrosis factor [4-71. Whether these observations can be generalized to other cell lines and growth inhibitors is not known. We previously have described an 1%kDa cell surface sialoglycopeptide (SGP) inhibitor that was isolated from bovine cerebral cortex intact cells [8]. This SGP has been purified to homogeneity, and it has shown to be a potent inhibitor of cellular protein and DNA synthesis and to cause complete cell cycle arrest [8, 91. Its inhibitory action is reversible and nontoxic, cell cycle arrest occurs in G,,/G, , and the biological activity is directed against a wide array of eukaryotic cell species and types [9]. There is reason to suspect that the SGP initiates its action within minutes of its addition to the target cells. The calcium ionophore A23187, but not the sodium ionophore monensin, has been shown to antagonize the action of the SGP if added to the target 3T3 cells simultaneously with the SGP or within 10 min of the inhibitor [lo]. The SGP abrogated the mitogenic action of epiderma1 growth factor (EGF), 12-O-tetradecanoylphorbol13-acetate (TPA), and bombesin, but only if the inhibitor was added within minutes of the mitogens [ll-131. These observations provided us with a rationale to investigate the potential involvement of protoonco-

When an 18-kDa cell surface sialoglycopeptide (SGP) , isolated from intact bovine cerebral cortex cells, was incubated with exponentially growing Swiss 3T3 cells, cell proliferation was efllciently arrested. The inhibition was totally reversible since after removal of the SGP the arrested cells resumed their progress in the cell cycle in a synchronized manner for at least two divisions. Keaddition of the GSP 4 h after reversal of the inhibition did not, however, affect the commitment of the cells to advance through metaphase, although progress through the cell cycle was once again inhibited after the cells reentered the G1 phase. The efficient nature of the SGP-mediated cell cycle arrest in G, provided us with a basis to examine potential changes in the expression of several competence genes, and genes associated with mid and late G1, that have been implicated in cell cycle progression. Upon serum stimulation of quiescent Swiss 3T3 cells, the induction of c-myc and c-fos expression was not influenced by the SGP at concentrations highly inhibitory to cell cycling. Expression of JE was induced by serum, and the presence of the SGP had little effect on the expression of this growthrelated gene. KC expression was not appreciably stimulated by serum although, surprisingly, the addition of the SGP resulted in a significant increase in expression. In addition, we learned that the SGP did not alter expression of ornithine decarboxylase, c-ras, or thymidine kinase, which are induced later than the genes associated with the initial stages of competence. c 1s~ Academic Prese, Inc.

INTRODUCTION

The mechanism(s) responsible for the regulation of cell division is of central importance to numerous areas of cell biology ranging from developmental biology to immunology and to tumorigenesis. There is little question that proliferation of normal and tumorigenic cells is a result of interactions between both positive signals (growth factors) and negative signals (growth inhibitors). Although a considerable amount of information is available concerning the structure and biological func0014-4827/91 $3.00 Copyright Q 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

62

GENE

genes that are preferentially of competence. MATERIALS

EXPRESSION

AND

expressed in the early stage

AND METHODS

Purification of the bovine SGP. The bovine SGP was purified as described previously [8]. Briefly, the SGP was released from intact bovine cerebral cortex cells by mild proteolysis, concentrated by ethanol precipitation, and extracted with chloroform/methanol (2/l, v/ v). Following overnight dialysis against water, the SGP was lyophilized, resuspended in 0.05 M sodium acetate (pH 5.5), and bound to a DEAE ion-exchange column. The unbound material was discarded and the bound proteins were eluted with 0.4 M NaCl in 0.05 M sodium acetate (pH 5.5). The eluted material was then concentrated, resuspended in PBS, and further purified by wheat germ agglutinin afEnity chromatography. The unbound fraction was vacuum-dried, resuspended in 0.1 M sodium phosphate buffer (pH 6.8), and purified by HPLC with a TSK 3000 SW size-exclusion column (Phenomenex, Ranch0 Palos Verdes, CA). The final preparation was dialyzed against dilute PBS, vacuum-dried, and stored at -1OO’C. DNA synthesis inhibition was tested with various concentrations of the purified SGP inhibitor, resuspended in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% calf serum, sterilized through a 0.22~pm filter, and serially diluted in culture medium. Medium (250 ~1) was incubated with exponentially growing Swiss 3T3 cell cultures for 20 h. DNA synthesis then was measured by incubating each culture with 2.0 &i/ml of [3H]thymidine for 1 h and then precipitating the cellular DNA with ice-cold trichloroacetic acid as described by Fattaey et al. [9]. One unit of biological activity of the inhibitor was designated as the amount that provided a 25% reduction of DNA synthesis (-0.03 PM). Cell culture. All Swiss 3T3 cells were grown as monolayer cultures in a humidified incubator with 5% CO,:95% air atmosphere. Cells were maintained in 250 ~1 of DMEM containing 10% calf serum and used during exponential growth for DNA synthesis and cell cycle arrest, and as confluent and quiescent cultures for serum stimulation. Inhibition of Swiss 3T3 cell division. Swiss 3T3 cells were seeded in 48-well plates at 4.0 X lo3 cells/cm’. After a 4- or 10-h incubation period, the medium was removed and replaced with either 250 ~1 of complete medium alone or complete medium containing the SGP inhibitor (0.12 PM). After every generation time (20 h) the cell number was determined by detaching the cells from culture wells with trypsin and diluting (1:20) in Isoton II (Coulter Electronics, Inc., Hialeah, FL). Cell numbers were measured with a Coulter counter, Model ZM (Coulter Electronics, Inc.). After two generation times, the medium was removed from all cultures, the cells were washed two times with medium, fresh medium was added, and the cell number was determined as described in specific experiments. Swiss 3T3 cells were culCell culture for &y(A)+ RNA isolation. tured in 75-cm’ Costar flasks (Costar, Cambridge, MA), as described above, and grown for 5 days to quiescence. The culture medium was then removed and the cells were washed with 10 ml of serum-free DMEM. Serum-free DMEM medium was then added and the cultures were reincubated for 48 h. Following this 48-h incubation, serum-free medium was removed from the cells and replaced with DMEM containing 10% calf serum and various concentrations ofthe SGP (0 to 0.36 a). Several types of control cultures were used in these experiments. Control cultures, used to measure gene expression in quiescent cultures, received either conditioned DMEM medium that had been incubated with confluent Swiss 3T3 cell monolayers for 4 to 5 days or serum-free DMEM. To measure serum-induced expression, another set of cultures received fresh DMEM with 10% serum, while experimental cultures received fresh DMEM with 10% serum and various concentrations (0 to 0.36 JLM) of the SGP. Northern hybridization RNA analysis by Northern hybridization. was carried out with RNA that was extracted from all cell cultures

CELL

63

PROLIFERATION

using the method of Schwab et al. [14], after the incubation periods described in each experiment, following serum addition, with or without the SGP. Poly(A)+ RNA was purified by oligo(dT) chromatography, and the mRNA was separated by electrophoresis in a 1.2% agarose-formaldehyde gel. Northern blotting was performed as previously described [15]. Purified cDNA inserts of c-myc (4.7-kb BanHI-XboI insert), c-fos (l.l-kb P&I insert), KC (0.82-kb PstI insert), JE (0.75-kb PstI insert), thymidine kinase (TK) (l.l-kb EcoRI insert), c-ras (5.2- and 2.0-kb EcoRI-XbaI insert), ornithine decarboxylase (ODC) (l.l-kb P&I insert), and lB15 (0.68-kb BamHI-XbaI insert), were labeled by random primer extension [16]. Following hybridization, the blots were washed in 0.1X SSC, 0.1% SDS at 42’C, and exposed to Kodak XAR-5 film using an intensifying screen. Most of the kinetic experiments involved at least two time periods of incubation, measurements with two to three independent preparations of poly(A)+ RNA for each time period, and duplicate Northern analysis with each sample (c-fos, c-myc, KC, JE, and c-ras). TK and ODC expression were measured with independent preparations of poly(A)+ RNA isolated from cells incubated over at least four incubation periods. For all of the Northern blot analyses, hybridization with a control lB15 probe demonstrated equal loading of RNA in each lane.

RESULTS Synchronization by the SGP

of Mouse Swiss 3T3 Cell Division

The ability of the SGP inhibitor to synchronize cell populations was assessed with nonconfluent and exponentially dividing monolayer cultures of mouse Swiss 3T3 cells. The cells were incubated in 250 ~1 of medium with 0.12 pM of the SGP for a period of 40 h, and the cell number was determined at least every generation time (20 h). Compared to control cultures fed with complete medium without the SGP at the same time as the experimental cultures, within 20 h cells treated with the SGP were inhibited from dividing (Fig. 1). After a 40-h incubation period, the medium on all cultures was replaced with complete medium containing 10% calf serum. There was no significant change in the rate of proliferation in control cultures which continued exponential growth, while the number of cells in the previously SGP-treated cultures doubled abruptly within 12 h after removal of the inhibitor (Fig. 1). Following this doubling, the cell number remained essentially constant for an extensive period, after which the number of cells, again, in a synchronous fashion completely doubled. The kinetics of recovery clearly indicated that the SGP-mediated cell cycle arrest was totally reversible, and that a single exposure of Swiss 3T3 cells to the SGP inhibitor synchronized the cell populations for at least two divisions. The kinetics of inhibition and the synchronized recovery of the Swiss 3T3 cell population, as shown in Fig. 1, suggested that a single arrest point was associated with the inhibitory action of the SGP. To determine if there was a single arrest point or multiple points in the cell cycle where the SGP might act, the inhibitor was used to synchronize Swiss 3T3 cell populations, and the inhibi-

64

FATTAEY,

BASCOM,

bt t 20

40

60

00

Imabrtlon Time (h-s)

FIG. 1. Synchronization of Swiss 3T3 cells by the bovine SGP inhibitor. Swiss 3T3 cells were plated and allowed to attach to the culture wells for 10 h, and then the inhibitor (0.12 j&f) was added (first arrow). After two generation times, the medium with the inhibitor was removed and fresh medium was added to all cultures (second arrow). The data represent the means + SD of triplicate cultures. Control cultures not treated with the SGP (closed circles); SGP inhibited and reversed cultures (open circles).

tion again was reversed by feeding the cultures with complete medium without the SGP. In this series of experiments, 4 h after the SGP was removed, one set of the previously SGP-treated cells received medium with 0.12 PM of SGP while a second set received fresh medium without the inhibitor. Cells retreated with the SGP continued to traverse the cell cycle and divided, but again they were inhibited after the mitotic phase (Fig. 2). This illustrated that there was not a second arrest point between 4 h after reversal and mitosis, and that the inhibition again occurred in the G, phase of the cell cycle. The efficient cell cycle arrest that was mediated by the SGP, and the synchronous nature of cell division following reversal of the inhibition, provided a basis to assess the relationship between gene expression and cell cycling. To determine the potential influence of the SGP on the expression of c-myc and c-fos, monolayers of Swiss 3T3 cells were grown for 5 days to confluence and the quiescent cultures were then fed with serumfree DMEM and reincubated for 48 h.

AND JOHNSON

taining 10% calf serum and various concentrations of the SGP inhibitor (0 to 0.36 PM). Poly(A)+ RNA was extracted 30 and 60 min after the various treatments and analyzed by Northern blot for both c-myc and c-fos expression. Results of Northern blot analysis showed that c-fos expression was measurably induced by serum stimulation after 30 min and that by 60 min the expression was significantly reduced. Even though highly inhibitory to cell division, the presence of 0.12 to 0.36 PM of the SGP had little, if any, influence on c-fos expression (Fig. 3). Although slight, it appeared as if the expression of c-fos was stimulated at 60 min in the presence of 0.36 PM of the SGP. Unlike c-fos, the expression of c-myc was only minimal at 30 min with an approximately 30-fold induction at 1 h after the addition of DMEM and 10% calf serum. The expression of c-myc, however, was not measurably affected, at 30 min or at 60 min, by any of the concentrations of the SGP (Fig. 4). Dose-Dependent

Induction of KC Expression by the SGP

The KC gene is a platelet-derived growth factor (PDGF)-inducible gene that commonly is measured in cell growth regulation studies. To examine the influence of the SGP on the induced expression of KC, confluent and quiescent Swiss 3T3 cell cultures were incubated in serum-free DMEM for 48 h, the medium was removed,

Joeml

-

3 0

20

40 Incsbstlos

The SGP Does Not Inhibit Serum-Induced of c-myc and c-fos

Expression

The potential influence of SGP inhibitor on seruminduced expression of c-myc and c-fos was assessed with confluent cultures of mouse Swiss 3T3 cells. After the 48-h incubation period in serum-free medium, the medium was removed and replaced with conditioned medium, serum-free medium, or complete DMEM con-

60 Tlmw

80

100

(bn)

FIG. 2. Readdition of the SGP after reversal of the inhibition of cell cycling. Swiss 3T3 cells were plated and allowed to attach to the culture plate for 10 h and the SGP (0.12 FM) was then added (first closed arrow). After two generation times, the medium from all cultures was removed and fresh medium without the SGP, was added (open arrow). Four hours following the removal of the SGP the cells previously inhibited by the SGP again received the SGP (0.12 a) (second closed arrow). The data represent the means + SD of triplicate cultures. Control cultures not treated with the SGP (closed circles); SGP inhibited, reversed, and reinhibited cultures (open circles).

GENE

EXPRESSION

AND

CELL

65

PROLIFERATION

CoadltioaaiMedinm &3mmFreeMedfu

IWbCaSt DMEM t .l2opM t .24opM

8cP

+ .36opM

ConditionedMedium SerumFreeMedium 10% CaS + DMEM

t .I20 pM

I

FIG. 3. Effect of bovine SGP on expression of c-fos in serumstimulated quiescent Swiss 3T3 cells. Swiss 3T3 cells were grown to quiescence and incubated with DMEM with 10% calf serum as described under Materials and Methods. After 30 and 60 min, poly(A)+ RNA was isolated and analyzed by Northern blots as described. Each lane contained 1.0 pg of poly(A)+ RNA that was hybridized to a 32P-labeled cDNA probe as described under Materials and Methods.

and then the cultures were fed with conditioned medium, serum-free medium, or complete DMEM containing 10% calf serum and 0 to 0.36 PM of the SGP inhibitor. Northern blot analysis of poly(A)+ RNA, isolated 30 min after serum stimulation, showed only minimal KC expression under all experimental conditions.

d

1

60’

SW

t .2M PM t .360 pM

60

c-myc

FIG. 5. Effect of bovine SGP inhibitor on the expression of KC in serum-stimulated quiescent Swiss 3T3 cells. The experiment was carried out as described in the legend to Fig. 3. Each lane contained 1.0 pg of poly(A)+ RNA that was hybridized to a 3ZP-labeled cDNA probe as described under Materials and Methods.

Surprisingly, an increase in KC gene expression was observed after 60 min incubation with the SGP (Fig. 5). In addition, the SGP-induced induction observed at 60 min was shown to be SGP dose dependent in that there was a measurable increase in KC expression with concentrations of the SGP from 0.12 to 0.36 PM. Expression of JE in the Presence of the SGP Expression of the PDGF-inducible gene JE, which also has been correlated with cell cycling, was stimulated by the replacement of serum-free medium with DMEM containing calf serum. JE expression was measurable within 30 min of the addition of medium containing calf serum although the expression was significantly increased at 60 min. After 30 min of incubation with serum-containing DMEM, the presence of 0.12 and 0.24 ,I& of the SGP did not affect the expression although there was a marked reduction with the highest concentration of the inhibitor used (0.36 a). Although at 60 min of incubation JE expression appeared to have been increased (Fig. 6), densitometric analyses of the autoradiographs did not show significant differences in gene expression with or without the SGP. The SGP Does Not Inhibit the Expression of the c-ma, ODC, and TK GenesAssociated with Mid and Late G,

FIG. 4. Effect of the bovine SGP on the expression of c-myc in serum-stimulated quiescent Swiss 3T3 cells. The experiment was carried out as described in the legend to Fig. 3. Each lane contained 1.0 pg of poly(A)+ RNA that was hybridized to a “P-labeled cDNA probe as described under Materials and Methods.

The expression of c-ras and ODC, genes associated with mid G, , was measured in the absence and presence 0.36 piU SGP at 1,6, 12, and 18 h after serum stimulation of quiescent Swiss 3T3 cells. The expression of c-

66

FATTAEY,

BASCOM,

JE Conditioned Serum

Medium

Free Medium

t

lO%CaS

Conditioned Serum

10% CaS t DMEM .I20

t

.240 t~bf

30

Medium

Free Medium

t

1

DMEM

I 40

pM

+ .56oPM

FIG. 6. Effect of bovine SGP inhibitor on the expression of JE in serum-stimulated quiescent Swiss 3T3 cells. The experiment was carried out as described in the legend to Fig. 3. Each lane contained 1.0 c(g of poly(A)+ RNA that was hybridized to a “P-labeled cDNA probe as described under Materials and Methods.

ras was measured with 0.36 pit4 SGP through the 18-h period following serum stimulation. Although C-ras was significantly stimulated by the addition of serum, similar to the early competence genes, the presence of the SGP inhibitor did not influence expression (Fig. 7). Similar results were obtained with studies carried out over the same time period with the expression of the ODC (data not shown). With TK, a gene associated with late G, , expression was minimal at 6 h followed by a gradual increase to 24 h, and Northern blot analyses of the isolated poly(A)+ RNA showed that the SGP had little, if any, influence on TK expression (Fig. 7).

AND

JOHNSON

to as competence, entry, progression, and assembly and are separated by points of entry C, V, and R, culminating at the S phase [17]. The c-fos, c-myc, JE, and KC genes are associated with the initial competence subphase. In light of the early indication that the SGP antagonizes the mitogenic action of growth factors and promoters, as well as the abrogation of the inhibitory activity when the ionophore A23187 was added within minutes of the SGP, this set of early genes was the initial focus of the present study. One of the first genes to be stimulated after the addition of growth factors is c-fos [18, 191. The measured induction of c-fos expression within 30 min and the relatively short half-life of the transcripts (Fig. 3) were consistent with earlier reports by others [l&20,21]. There was little influence by the SGP on the expression and the half-life of c-fos transcripts with the exception of a slight, but perceptible, sparing at 0.36 PM. Upon growth factor stimulation of confluent cell cultures the induction of c-myc follows that of c-fos [20,22,23]. Similar to the induced expression of c-fos, the presence of 0.12 to 0.36 pM SGP had no measured effect on the induction of c-myc expression. Unlike c-fos and c-myc, KC gene expression was only slightly stimulated by the addition of serum, but the expression was significantly enhanced by the SGP in a concentration-dependent manner. The expression of JE, another PDGF-dependent gene associated with entry of the Swiss 3T3 cells into the mitotic cycle [24],

c- ras

TK

5 7

2 3

Incubation Time (b)

DISCUSSION

The relationship between protooncogene expression and the commitment of quiescent cell populations to enter the cell cycle has been an intensive topic of study. In the absence of purified naturally occurring cell growth inhibitors, however, the important links between inhibitory signals and mitogenic cues (growth factors and tumor promoters) have been difficult to establish. Previous studies have shown that the bovine SGP inhibitor arrested Swiss 3T3 cells in the G,/G, phase of the mitotic cycle [9], and the present study indicates that there is not a second arrest point between the original G,/G, block and cell division. There are very few “landmarks” that serve to map the G, phase of the cell cycle [17] and that would serve to establish the position of the SGP-mediated arrest site. Sequential subphases, however, have been identified and referred

FIG. 7. Effect of the SGP inhibitor on the expression of c-ras and TK in serum-stimulated quiescent Swiss 3T3 cells. The experiments were conducted as described in the legend to Fig. 3 except that poly(A)+ RNA was isolated 1 through 18 h (c-ras) or 6 through 24 h (TK) following serum stimulation, and that only 0.36 &f SGP was used in the samples indicated. Each lane contained 1.0 pg of poly(A)+ RNA that was hybridized to a **P-labeled cDNA probe as described under Materials and Methods.

GENE

EXPRESSION

AND

however, was not influenced by the presence of the SGP inhibitor. Since expression of these competence genes, allowing the cells to traverse this initial substage of G, , does not require protein synthesis [17,25]), the protein synthesis inhibitory activity of the SGP would not be expected to play a critical role in arresting cells at the C point. However, we considered the possibility that the SGP did not increase the transcription of the KC gene but rather mediated a diminished rate of protein synthesis that decreased the stability of the transcripts. It already is known that the protein synthesis inhibitory activity of the SGP can be measured within minutes of its addition to target cells [8, 261. This explanation does not seem plausible, however, since a sparing influence on c-fos, c-myc, and JE transcripts would have been expected. It is clear that the bovine SGP inhibitor does not influence competence gene expression like a global protein synthesis inhibitor. Superinduction of c-fos expression has been reported with Swiss 3T3 cells treated with a combination of PDGF and cycloheximide [27], and a superinduction of KC, JE, and c-myc expression has been measured in the presence of cycloheximide [24, 281. Unlike cycloheximide, growth inhibitory levels of the SGP only superinduced KC expression; whether this unusual observation is related in any fashion to cell growth control remains to be resolved. Although previous observations have shown that the inhibitory action of the SGP is rapidly mediated when abolishing the action of growth factors and TPA [ll131, the inhibition is not associated with a transcriptional block of at least these four competence genes. This is consistent with a recent report that transforming growth factor fl could inhibit serum-induced cell cycling of murine primary fibroblasts or lectin-stimulated T-lymphocytes, although inhibition was not accompanied by an altered expression of competence genes [29, 301. Even the expression of mid G, (ODC and c-ras) or late G, (TK) genes was not inhibited by the SGP. It is possible that the translation of these mRNA transcripts, in the later entry and progression subphases of G,, actually leads to the cell proliferation arrest [31-331. It is interesting, however, that cycloheximide has been shown to inhibit the induction of ODC mRNA synthesis [34], while the SGP was found to have little influence on ODC expression. Clearly, a simple translational block, in a manner like cycloheximide, would not explain the mitotic arrest mediated by the SGP inhibitor. The most likely site of inhibition in this case would be in the progression subphase of G, because there is a requirement for rapid protein synthesis during this latter part of G, [ 17,351. The kinetics of reversal, with a 12 h delay between the removal of the SGP and subsequent mitosis, is consistent with a mid or late G, arrest point. The ability of the SGP to synchronize 3T3 cell populations permits a systematic examination of the effects

CELL

67

PROLIFERATION

that the inhibitor imposes on the cell cycle and/or cell growth control mechanisms. Recent experiments that measured the ability of the SGP to inhibit the mobilization of cellular Ca2+ andcytosolic alkalinization suggest that early signal transduction is the initial site of the inhibitor’s action (in preparation). It is therefore possible that rapidly occurring events initially trigger the subsequent mitotic arrest, although the cells are not stalled in their progression through the cell cycle until a later stage, i.e., mid or late G, near the R point. The authors gratefully acknowledge the excellent technical assistance of Ruth Williams. This study was supported by Grant CD-323 from the American Cancer Society, Contract NAGW-1197 from NASA for A Center for the Commercial Development of Space, the Wesley Foundation, and the Kansas Agricultural Experiment Station. This is Contribution 90-129-J from the Kansas Agricultural Experiment Station, Kansas State University.

REFERENCES 1.

Lau, L. F., and Nathans,

D. (1985) EMBO J. 4,3145-3151.

2. Eppel, D., and Dub& F. (1987) in Control of Animal Cell Prolifer-

3.

ation (Boynton, A. L., and 393, Academic Press, New Hesketh, R., Smith, G. A., of Animal Cell Proliferation Eds.), Vol. 2, pp. 395-431,

Leffert, H. L., Eds.), Vol. 2, pp. 364York. and Metcalfe, J. C. (1987) in Control (Boy&on, A. L., and Leffert, H. L., Academic Press, New York.

4. Kronke, M., Schluter, C., and Pfizenmaier,

(1987) Proc. Natl.

5.

J., and Hamilton, 144,1197-1205. G. R. (1987) Cell

6.

K. Acad. Sci. USA 64,469-473. Fernandez-Pol, J. A., Talkad, V. D., Klos, D. P. D. (1987) Biochem. Biophys. Res. Commun. Takehara, K., LeRoy, E. C., and Grotendorst,

49.415-422. 7. Coffey, R. J., Bascom, C. C., Sipes, N. J., Graves-Deal,

R., Weissman, B. E., and Moses, H. L. (1988) Mol. Cell. Biol. 8, 3088-3093. 8. Sharifi, B. G., Johnson, T. C., Khurana, V. K., Bascom, C. C., Fleener, T. J., and Chou, H. H. (1986) J. Neurochem. 46,461469. 9. Fattaey, H., Johnson, T. C., and Chou, H.-H. (1989) J. Cell Physiol. 139, 269-274. 10. Sharifi, B. G., Bascom, C. C., and Johnson, T. C. (1986) Rio&em. Biophys. Res. Commun. 136,976-982. 11. Bascom, C. C., Sharifi, B. G., and Johnson, T. C. (1987) J. Cell. Biochem. 34,283-291. 12. Chou, H-H. J., Sharifi, B. G., Bascom, C. C., Johnson, T. C., and Perchellet, J. P. (1987) Cancer L&t. 35, 119-128. 13. Johnson, T. C., and Sharifi, B. G. (1989) B&hem. Biophys. Res. Commun. 161,468-474. 14. Schwab, M., Alitalo, K., Varmus, H. E., and Bishop, J. M. (1983) Nature (London) 303,497-501. 15. Thomas, P. S. (1980) Proc. Natl. Acad. Sci. USA 77,5201-5205. 16. Taylor, J. M., Illmensee, R., and Summers, J. (1976) B&him. Biophys. Acta 442,324-330. 17. Pardee, A. B. (1989) Science 246,603-608. 18. Kruijer, W., Cooper, J. W., Hunter, T., and Verma, I. M. (1984) Nature (London) 312,711-716. 19. Prywes, R., and Roeder, R. G. (1986) CeU 47,777-784. 20. Denhardt, T. D., Edwards, D. R., and Parfett, C. L. J. (1986) B&him. Biophys. Acta 865,83-125.

FATTAEY, 21.

BASCOM,

AND

Miirike, M., Quaieer, A., Muller, D., and Montenarh, M. (1988) 0ncogen.e 3,X1-158. Greenberg, M. E., and Ziff, E. B. (1984) Nature (London) 311, 433-438.

28.

Muller, R., Bravo, R., Burckhurdt, J., and Curran, T. (1984) Nature (London) 312,716-720. Cochran, B. H., Reffel, A. C., and Stiles, C. D. (1983) Cell 33, 939-947.

30.

25.

Rittling, S. R., Gibson, C. W., Ferrari, S., and Baserga, R. (1985) Biochem. Biaphys. Res. Commun. 132,327-335.

32. 33.

26.

Bascom, C. C., Shari& Physiol. 128,202-208.

27.

Cochran, B. H., Zullo, J., Verma, I. M., and Stiles, C. D. (1984) Science 226. 1080-1082.

22. 23. 24.

B. G., and Johnson,

Received September 20,199O Revised version received December

21,199O

29.

31.

T. C. (1986) J. Cell 34. 35.

JOHNSON

Kelly, K., Cochran, B. H., Stiles, C. D., and Leder, P. (1983) Cell 35,603-610. Sorrentino, R., and Bandyopadhyay, S. (1989) Oncogenc 4,569574. Morris, D. R., Kuepfer, C. A., Ellingsworth, L. R., Ogawa, Y., and Rabinovitch, P. S. (1989) Exp. Cell Res. 185, 529-534. Moelling, K. (1986) in Oncogenes and Growth Control (Kahn, P., and Graf, T. Eds.), pp. 271-277, Springer-Verlag, Heidelberg. Rozengurt, E. (1986) Science 234,161-166. Palmiter, R. D., and Brinster, R. L. (1986) Annu. Reu. Genet. 20, 465-499. Katz, A., and Kahana, C. (1987) Mol. Cell. Biol. 7, 2641-2643. Coppock, D. L., and Pardee, A. B. (1987) Mol. Cell. Bid. 7,29252932.