EXPERIMENTAL

CELL

RESEARCH

198,250-258

(1992)

Mechanism of Induction of Cellular DNA Synthesis by the Adenovirus El A 12s cDNA Product’ EIJI HARA,* TOMOKO OHSHIMA,~ TAKAKO ISHII,$ WATARU SUGINO,~ Ko TSUTSUI,* SUSUMU NAKADA,~ NOBUO TSUCHIDA,? AND KINICHIRO ODA$*~ *Departments

of

Applied Biological Science, $Biological Science and Technology, and $Pkurmaceutical Science, Science University Noo!u, Chiba 278, Japan; and tDepartment of Molecular and Cellular Oncology, Faculty of Dentistry, Tokyo Medical and Dental University, Yushima, Bunkyo-ku, Tokyo 113, Japan

The mechanism of induction of DNA synthesis in quiescent rat 3Yl cells by the adenovirus ElA gene was investigated using the 3Y 1 derivative cell lines g12-2 1, gnl2RB1, and gn12RB2. The g12-21 cells express the ElA 12s cDNA and the latter two cells express both the ElA 125 cDNA and the human retinoblastoma susceptibility (Rb) gene at different levels in response to dexamethasone (dex). The cDNA sequences of ElA-inducible cell cycle-dependent genes, clone 3 and clone 16, were isolated by differential screening of a cDNA library constructed from dex-treated g12-21 cells. The quiescent 3Y 1 cells induced c-foe and c-myc expression within 2 h after serum stimulation and expressed clone 16 and clone 3 transiently at around 8 h before the onset of DNA synthesis (10 h). In contrast, the quiescent g12-21 cells treated with dex expressed a high level of ElA at 6 to 8 h after treatment and expressed clone 16 and clone 3 at around 8 h without stimulation of c-fos and c-myc expression, suggesting that ElA bypasses the cell cycle early in Gl. The half-maximal rate of DNA synthesis was reached in a much shorter time in dex-treated g12-21 cells (12 h) than in serum-treated 3Yl cells (18 h), suggesting that ElA also bypasses the cell cycle at the Gl/S boundary. The gnl2RBl and gn12RB2 cells were unable to induce DNA synthesis in response to dex presumably due to lower levels of ElA expression, although gn12RB2 but not gnl2RBl cells could express clone 16 and clone 3. These results suggest that the level of ElA required for bypass at the GUS boundary is higher than that required early in G 1. 0 1892 Academic Preen, Inc.

INTRODUCTION The cell cycling in quiescent cells begins with the GO/ Gl transition and proceeds through Gl to S phase in response to specific signals. The cell cycle progression ’ Supported by a grant from the Ministry of Education, Culture of Japan and by a grant from the Vehicle Racing rative Foundation to K.O. 2 To whom reprint requests should be addressed. 0014-4827/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

Science and Commemo-

of

Tokyo,

seems to take place through successive biochemical events, most of which are still unknown. Stimulation of quiescent fibroblasts with growth factors results in sequential expression of c-fos and c-myc [l-6]. The former seems to be involved in the GO/G1 transition [7] and the latter in the subsequent Gl progression [a-lo]. Induction of cell cycle progression in quiescent cells by adenovirus type 2 (Ad2) or Ad5 seems to be different from that mediated by serum factors, since Ad2 could induce cellular DNA synthesis in Gl-specific temperature-sensitive (ts) cell mutants at the restrictive temperature [ 11,121. Ad2 may have the ability to bypass the ts defect by activating the late Gl or S phase genes [13]. This bypass is presumably mediated by the ElA function, since the ElA gene has mitogenic activity [14-191. The ElA gene of Ad2 and Ad5 generates two major species of 13s and 12s mRNA, which encode similar proteins of 289 and 243 amino acids (289R and 243R), respectively, differing only by the presence of an internal 46 amino acid sequence in the 289R protein [20-231. The ability to transactivate a variety of viral and cellular promoters resides in the internal 46 amino acid sequence [24-301 and the ability for repression of the enhancer-dependent promoters [31,32] resides in the first exon [26,30]. Both of these transcriptional modulation functions, however, are unlinked to the function required for cell immortalization and transformation [33, 341. Direct linkage of the ElA function to cell transformation has been shown by its ability to form complexes with cellular proteins, including the product of the retinoblastoma susceptibility (Rb) gene [35-391. All the ElA mutants defective in the binding domains concomitantly lose the ability to transform cells [38,39]. The Rb gene known as a tumor suppressor gene encodes a nuclear phosphoprotein with DNA binding activity [40]. Involvement of RB in the control of cell growth was suggested by its cell cycle-dependent phosphorylation. RB is hypophosphorylated in the GO/G1 phase, but becomes phosphorylated at the Gl/S boundary [41-441, presumably losing its negative role in cell proliferation. 250

INDUCTION

OF DNA SYNTHESIS

In the present paper, to study the mechanism of ElAmediated cell cycle progression, the time course of the induction of DNA synthesis in quiescent cells by ElA was compared with that mediated by serum factors using rat 3Yl derivative cell lines in which expression of the ElA 12s cDNA and/or the Rb gene can be regulated by dexamethasone (dex). The cDNA sequences of ElAinducible cell cycle-dependent genes were isolated from a cDNA library constructed from a 3Yl derivative cell line expressing ElA and used as markers for cell cycle progression. The results suggest that ElA bypasses the cell cycle both early and late in the Gl phase. MATERIALS

AND METHODS

Cell lines. The 3Yl-B cell line, clonel-6, is a clonal line of Fischer rat embryo fibroblasts [46]. A 3Yl derivative cell line, g12-21, was established by introducing 3Yl cells with pM12SG (Fig. 1) and pSV2gpt [47]. The cell lines gnl2RBl and gn12RB2 were established by introducing g12-21 cells with pMSRB (Fig. 1) and pSV2neo [48]. All these cell lines were cultivated at 37°C in Dulbecco’s modified Eagle’s medium with 10% fetal calf serum (FCS). Construction of recombinant DNAs. The plasmid pMDSG [48] was partially digested with Hind111 and BglII, which have three and two restriction sites, respectively, and vector DNA lacking the mouse dihydrofolate reductase (dhfr) cDNA was isolated. The Ad2 ElA 12s cDNA was isolated from plasmid pAd2-12s [50] after cleavage with HindIII and BglII and ligated to the pMDSG vector DNA to generate pM12SG (Fig. 1) in which the 12s cDNA is placed downstream of the mouse mammary tumor virus (MMTV) long terminal repeat (LTR). On the other hand, the pMDSG vector DNA lacking the dhfr cDNA was circularized after ligation of the X&I linkers to its ends to generate pMSG-XbaI for cloning of a gene downstream of the MMTVLTR. The pMSG-XbaI plasmid was digested with BamHI and self-ligated after removal of the Ecogpt gene. The resulting plasmid was digested with XbaI and the BglII linker was inserted to generate pMSG-BglII. The plasmid p4.95BT, containing the human Rb cDNA [51] in the multicloning site of pBluescriptI1 KS+, was digested with KpnI and the BamHI linker was inserted after the KpnI cohesive ends were blunt-ended to generate p4.95BTB. The Rb cDNA was isolated from p4.95BTB by cleavage with BamHI and ligated to pMSG-BglII cleaved with BglII to generate pMSRB (Fig. 1). Cloning of ElA-inducible cellular genes. A cDNA library was constructed from quiescent g12-21 cells treated with 10e6 M dex for 7 to 8 h using XgtlO as a cloning vector [52]. The library, which originally contained 2 X 10s members, was amplified once and screened for ElA-inducible gene sequences by differential plaque hybridization [53]. =P-labeled cDNA probes were prepared from quiescent g12-21 and 3Yl cells treated with dex for 7 to 8 h. The XgtlO cDNA library was plated on the E. coli hflA strain at about 3000 plaques per 137mm dish. The recombinant phage plaques were transferred to duplicate sets of nylon filters (NEN Research Product). After denaturation of the phage DNA, the filters were prehybridized in 50% formamide (BRL), 1% SDS, 1 M NaCI, and 10% (w/v) dextransulfate (Pharmacia) for at least 6 h at 42°C. Hybridization was performed at 42°C for 16 to 18 h in the prehybridization buffer containing 100 pg of sonicated herring sperm DNA per milliliter and lO@cpm/ml of either one of the =P-labeled cDNA probes. After autoradiography, the intensity of the signal of each plaque was compared between these two filters and the plaques which preferentially hybridized to the cDNA probe from g12-21 cells were selected. Positive clones were further screened by two cycles of differential hybridization.

251

BY ElA

Nuclear run-on transcription assay. Cells were suspended in RSB (10 mM Tris-HCI (pH 7.4), 10 mM NaCl, 3 n&f MgCl,) containing 0.5% (v/v) Nonidet-P40 and the liberated nuclei were collected by centrifugation. After washing with the same buffer, the nuclei were resuspended in a solution of 40% (v/v) glycerol, 50 mM Tris-HCl (pH 8.3), 5 mM MgC&, 0.1 mM EDTA at a DNA concentration of 1 to 2 mg/ml, and stored at -80°C. Transcription in isolated nuclei (200 pg of DNA) was performed in the presence of [a-zaP]dCTP [54] and nascent isolated RNAs were hybridized with 1 pg each of probe DNA spotted on nitrocellulose filters. The filters were prehybridized at 42°C for 8 h in 50% formami&-5X Denhardt’s dution-o.l% SDS5~ SSPE [0.9 M NaCl, 50 mM phosphate buffer (pH 7.7), 5 mM EDTA] in the presence of 100 pglml of sonicated HS DNA. Hybridization was carried out for 16 h at 42’C in 50% formamide-10% dextran sulfate-0.1% SDS-5~ SSPE with S2P-labeled RNA at 2 X 10’ cpml ml. The filters were washed twice in 2~ SSC containing 0.1% SDS for 10 min and then twice in 0.1X SSC containing 0.1% SDS for 10 min at room temperature. Measurement of the rate of DNA synthesis. Induction of DNA synthesis in quiescent cells after treatment with lo-‘M dex or 20% FCS was measured by labeling the cells with [SH]thymidine at 5 &i/ml for 1 h. After labeling, the cell monolayers were washed with cold phosphate-buffered saline (PBS) and harvested with a rubber policeman. The cells were completely dissolved in 2 ml of PBS containing 1% SDS and the radioactivity in an aliquot of the samples was counted in a liquid scintillation counter after precipitation in cold 5% trichloroacetic acid.

RESULTS

Cloning of ElA-inducible cell cycle-dependent genes. The rat 3Yl derivative cell line which express the Ad2 ElA 12s cDNA in response to dex was established by introducing rat 3Yl cells with the recombinant plasmid pM12SG (Fig. 1). The gpt+ colonies capable of expressing the 12s cDNA upon addition of dex were first selected by RNA dot hybridization [55] and then by Northern blotting. The cell line, g12-21, thus selected expressed a very low level of the 12s mRNA in the absence of dex, but expressed a high level after addition of dex. The cells were made quiescent and progression of the cell cycle was monitored by flow cytometry after treatment with dex. As shown in Fig. 2, almost all the cells were in the GO/G1 phase at time zero. At 6 h after treatment, most of the cells were still in the GO/G1 phase and no significant proportion of the cells entered S phase. Under these conditions, expression of the 12s cDNA was induced within 4 h, reaching its maximal level at 6 to 8 h as shown in the following sections. At 12 h after treatment, approximately one-half of the cell population entered the S and G2/M phases. The cell cycle progressed further and the proportion of the cells in the GO/G1 phase again increased after 24 h. To isolate the ElA-inducible cell cycle-dependent genes, a cDNA library was constructed from g12-21 cells at 7 to 8 h after treatment with dex, using XgtlO as a cloning vector. Differential plaque hybridization was performed with the 32P-labeled cDNA probes prepared from quiescent g12-21 and 3Yl cells which had been

252

HARA ET AL.

Eco-gpt

6gWBamHI hdlll

FIG. 1. Structures of recombinant plasmids. The plasmid pM12SG contains the Ad2 ElA 12s cDNA downstream of the MMTV-LTR and the Ecogpt expression unit derived from pSV2gpt [47]. The plasmid pMSRB contains the human Rb gene cDNA downstream of the MMTVLTR. The shaded, stippled regions contain the SV40 RNA splicing signals and the stippled regions contain the polyadenylation signal.

treated with dex for 7 to 8 h. The recombinant phage plaques plated on the Escherkhia coli hflA- strain were transferred in duplicate to nylon filters and hybridized with these probes, respectively. Approximately 5 X 10’ phage plaques were screened and about 200 plaques which showed stronger intensity with the cDNA probe from g12-21 cells were first isolated. These plaques were rescreened twice more by differential hybridization and several recombinant phages which repeatedly showed preferential hybridization to the g12-21 cDNA probe were finally isolated as shown in Fig. 3. Among these several cDNA clones, clone 3 and clone 16 showed a transient expression during the progression of the cell cycle from quiescence as shown below. Establishment of 3Yl derivative cell lines which express both ElA and Rb genes in response to okx. To establish 3Yl derivative cell lines which express both ElA and Rb genes in response to dex, g12-21 cells were cotransfected with pMSRB (Fig. 1) and pSV2neo. The plasmid pMSRB contains the Rb gene downstream of the MMTV hormone-inducible promotor. The cell lines designated gnl2RBl and gn12RB2 were established. Cotransfection of g12-21 cells with pSV2RB and pSV2neo developed many G418-resistant colonies comparable to those developed with pMSRB, but none of these colonies so far tested expressed a detectable level of Rb mRNA. Since the Rb gene was placed downstream of the SV40 enhancer-promoter in pSVBRB, the result suggests that constitutive expression of the exogenous Rb gene may have an inhibitory effect on the proliferation of 3Yl cells. Induction of the ElA and/or the Rb gene expression in quiescent g12-21, gnl2RB1, and gn12RB2 cells after addition of dex was analyzed by Northern blot hybridization with total cellular RNAs prepared at the times indicated (Fig. 4). Under these

conditions, no endogenous RB transcript was detected in dex-treated g12-21 cells. The level of /3-actin transcript was constant during the progression of the cell cycle (data not shown). In all the cell lines, the ElA 12s transcript began to be detected within 4 h after the addition of dex and reached its maximal level at around 8 h. The level of the transcript decreased steeply after 18 h. The level of the 12s transcript in the g12-21 cells was much higher than that in gnl2RBl and gn12RB2 cells. A moderate level of the 12s transcript was expressed in gn12RB2 cells, while the level was low in gnl2RBl cells. In contrast, a high level of the exogenous Rb transcript was induced in gnl2RBl cells, reaching its maximal level at 12 h, while the level induced in gn12RB2 cells was very low. Induction of DNA synthesis in 3Yl derivative cell lines which express ElA andlor Rb genes in response to dex. Under the same conditions, induction of DNA synthesis in these cell lines was followed by the incorporation of [3H]thymidine into the acid-insoluble fraction. For comparison, DNA synthesis was also induced by the addition of fresh medium containing 20% FCS with or without dex. As shown in Fig. 5A, DNA synthesis in g12-21 cells was induced at about 10 h after addition of serum and/or dex. The initial rate of DNA synthesis increased more quickly in the cells treated with serum plus dex than in the cells treated with serum alone. The halfmaximal rate of DNA synthesis was reached in about 12 h in g12-21 cells treated with dex or treated with serum plus dex, while that in the cells treated with serum alone was reached in about 18 h. This difference must not be due to the effect of dex treatment since the maximal rate reached in g12-21 cells by serum stimulation was inhibited by 30 to 40% by the presence of dex. The results suggest that ElA enables the quiescent cells to

INDUCTION

OF DNA SYNTHESIS

BY ElA

253

serum, but could not be induced by dex (Figs. 5B and 5C) presumably due to lower levels of ElA expression. Expression of ElA-inducibb cellular genes during the progression of the cell cycle. Expression of the ElA-inoh ducible cellular genes clone 3 and clone 16 during the progression of the cell cycle was analyzed by Northern blotting of total cellular RNAs prepared from quiescent I-.i .--- --Y g12-21, gnlBRB1, and gn12RB2 cells at various times after addition of dex (Fig. 6). The level of clone 16 transcript increased steeply at 8 h after treatment in g12-21 6h and gn12RB2 cells but decreased rapidly thereafter. The size of the clone 16 transcript was estimated to be about 1 kb. No clone 16 transcript was detected in gnl2RBl cells. The level of clone 3 expression also increased steeply at 8 and 12 h after dex treatment in g12-21 and gn12RB2 cells. However, the kinetics of clone 3 expression were somewhat different between 12h these cells. In g12-21 cells, the clone 3 transcript began to be detected after 8 h, reaching its maximal level at 12 h. Its expression continued until 18 h and then ceased abruptly. In gn12RB2 cells, the increase was delayed, reaching its maximal level at 18 h. The level was then decreased abruptly. The size of the clone 3 transcript 18h was estimated to be about 3 kb. No clone 3 transcript was detected in gl2RBl cells. The inability of gl2RBl cells to induce clone 16 and clone 3 expression is presumi, ably due to a low level of ElA expression and a high level of exogenous Rb expression. As a control, the expression of the cell cycle-dependent gene 2F1, isolated by 24h Liu et al. [13], was similarly analyzed to ensure the transient expression of clone 16 during the Gl phase. As shown in Fig. 6,2Fl was also expressed transiently at 8 h after treatment. This time was slightly later than that FIG. 2. Cell cycle progression of g12-21 cells after treatment reported in BHK cells. with dex. Monolayer cultures of g12-21 cells were maintained in lowTo eliminate the possibility that expression of clone serum (0.5% FCS) medium for about 40 h after they had reached 16 and clone 3 begins after entry to S phase, quiescent confluence and were treated with lo* M dex at Time 0. At the times (h) indicated on the left, the cells were stained with propidium iodide g12-21 cells were treated with dex in the presence of and analyzed by flow cytometry as previously described [16]. The aphidicolin and expression of clone 16 and clone 3 was arrows indicate the positions of cells in the GO/G1 phase and the similarly analyzed. Under these conditions, DNA synG2/M phase, respectively. thesis was completely inhibited. As shown in Fig. 7, clone 16 and clone 3 began to be expressed at 8 and 12 h after treatment, respectively, irrespective of the presence of aphidicolin and continued to be expressed for a enter S phase in a shorter time period than do serum factors. The maximal rate of DNA synthesis reached in prolonged period. These results suggest that clone 16 is g12-21 cells treated with dex was approximately 40% of expressed in dex-treated g12-21 cells at middle to late that reached in the cells treated with both serum and Gl phase and its expression ceases before entry into S dex. This difference is at least partly ascribed to the phase. The residual expression observed at 12 h in the (Fig. 6) may be due to difference in the proportion of cells entered into S absence of aphidicolin phase. The previous analysis of the labeled nuclei by asynchrony in the cell population. Expression of clone 3 seems to be induced somewhat later than that of clone autoradiography showed that nearly 40% of the gMA cells entered into S phase after induction of ElA ex- 16, presumably at the Gl/S boundary and continued pression by dex, while almost all the cells entered into S until S phase. phase after addition of serum [17]. DNA synthesis was To determine whether the transient expression of similarly induced in gnl2RBl and gnl2RBl cells by clone 3 and clone 16 is regulated at the level of trant,

t

L.-*..

A-

254

HARA ET AL.

FIG. 3. The third screening of the cDNA library for clones carrying the ElA-inducible cellular gene sequences. After two cycles of differential screening, the recombinant phages were spotted on an agar plate containing the E. coli hflA- strain. After transferring the plaques to duplicate sets of nylon filters, hybridization was performed with 32P-labeled cDNA probes prepared from dex-treated quiescent 3Yl cells (3Yl) and dex-treated g12-21 cells (12s).

scription, a nuclear run-on assay was performed. 32P-labeled nascent RNAs were synthesized in isolated nuclei prepared from g12-21, gnlBRB1, and gn12RB2 cells at various times after the addition of dex. These RNAs were hybridized to ElA and various c-one DNAs as well as the clone 3 and clone 16 cDNAs immobilized on nylon filters (Fig. 8). In g12-21 cells treated with dex, the expression of clone 16 and clone 3 was induced in a manner identical to that observed with Northern blot analysis. Under these conditions the expression of c-myc, crus, c-fos, and c-&n was not stimulated. In quiescent 3Yl cells treated with serum, c-fos expression was induced shortly after serum stimulation and c-myc was expressed subsequently. Expression of clone 16 was similarly induced, reaching its maximal level at 8 h. The pattern of clone 3 expression was different from that observed in dex-treated g12-21 cells. The expression was induced transiently at 8 h, but ceased immediately thereafter. The patterns of ElA and Rb expression in dextreated gnl2RBl and gn12RB2 cells were essentially the same as those observed in the Northern blot analysis. The former induced a higher level of Rb than ElA, while the latter induced a higher level of ElA than Rb. The patterns of clone 3 and clone 16 expression in gn12RB2 cells were also essentially the same as those observed in Northern blot analysis. Neither of these genes was expressed in dex-treated gnl2RBl cells. These results indicate that the transient expression of

clone 3 and clone 16 is regulated at the level of transcription. The lengths of clone 3 and clone 16 cDNAs were 1.7 and 0.3 kb and were incomplete judging from the lengths of the mRNAs (3 and 1 kb, respectively). Computer analysis of these cDNA sequences showed no significant homology with any known gene sequences. The patterns of transient expression of clone 16 and 2Fl were quite similar; however, the base sequences of these genes are completely different. The 2Fl gene encodes an ADP/ATP carrier protein [65]. DISCUSSION To study the mechanism of induction of cellular DNA synthesis by the ElA gene, a rat 3Yl derivative cell line capable of inducing DNA synthesis by the expression of the ElA 12s cDNA in response to dex was established. The 12s cDNA was chosen since it lacks the ElA transactivation domain and may activate cellular growth-regulated genes through an interaction with a cellular negative regulator(s) [35-391. A cDNA library was constructed from dex-treated g12-21 cells before the onset of DNA synthesis and the cDNA sequences of ElA-inducible cellular genes were isolated by the differential screening of the library. The Rb gene linked to the hormone-inducible promoter of MMTV-LTR was further introduced to g12-21 cells and two cell lines, gnl2RBl and gn12RB2, which express the ElA and Rb

INDUCTION m-2l

OF DNA SYNTHESIS

am2m-l

255

BY ElA gnl2RBi

912-21 n A I) 12 18 24

FIG. 4. Time course of ElA and Rb transcripts synthesized in g12-21, gnl2RB1, and gnl2RB2 cells treated with dex. The cells were maintained in low-serum medium for about 40 h after they had reached confluence and treated with 10e6 M dex at Time 0. Total cellular RNAs were prepared by the AGPC (acid guanidinium thiocyanate-phenol-chloroform) extraction method [66] at the times (h) indicated aboveeach lane. Aliquots of 25 gg were used for Northern blot hybridization with the same “P-labeled ElA and Rb probe DNAs. The filters were exposed to X-ray film for 2 days at -80°C. The ElA coding region (positions 453 to 1569) was prepared by cleaving the pAd2ElA DNA with Hi&II and HpaI and the 3.8-kb Rb cDNA fragment by cleaving the p2R3.8 DNA with EcoRI.

genes at different levels, were established. The results obtained with these cell lines and ElA-inducible cDNA sequences suggested that ElA bypasses the cell cycle at two points: one early in the Gl phase and the other at the Gl/S boundary as discussed below. Bypass early in the Gl phase. The quiescent 3Yl cells stimulated by serum induced c-fos and c-myc ex-

04

04

8121824

gn12Rl3.2

- _ R- 121824 .- - -.

8121824

flA

048121824

04

8121824

FIG. 6. Time course of clone 3 and clone 16 expression in g12-21, gnl2RB1, and gn12RB2 cells after treatment with dex. The quiescent g12-21, gnlZRB1, and gn12RB2 cells were treated with lo-‘M dex at time zero. At the times (h) indicated above each lane, total cellular RNA was prepared and used for Northern blot hybridization. The l.O- and 0.3-kb DNA fragments containing the coding regions of clone 3 and clone 16 genes were isolated from the recombinant phage DNAs by cleavage with EcoRI. These DNA fragments were used as probes after 32P end-labeling with T4 polynucleotide kinase. The cell cycledependent gene 2Fl [13] used as a marker was similarly labeled.

pression sequentially within 2 h after stimulation. Expression of clone 16 and clone 3 was induced transiently before the onset of DNA synthesis. In quiescent g12-21 A

0

10

20

30

10

20

30 Time

10

20

04

8

121824

B

0

4

8

12 18 24

-I

30

(h)

FIG. 6. Induction of cellular DNA synthesis in quiescent g12-21, gnlORB1, and gn12RB2 cells treated with serum and/or dex. Confluent monolayers of g12-21, gnlPRB1, and gn12RB2 cells maintained in low-serum medium for about 40 h were treated with 20% FCS (0), 10” M dex (A), or both (0) at time zero. At the times (h) indicated, the cells were labeled with 5 &i/ml of [SH]thymidine (1.04 TBq/mmol) for 1 h and the radioactivity in an aliquot of the samples was counted after TCA precipitation.

FIG. 7. Effect of aphidicolin on clone 3 and clone 16 expression in quiescent g12-21 cells treated with dex. Northern blot analysis of the clone 3 and clone 16 transcripts was performed as described in the legend to Fig. 6, except the cells were treated with dex in the presence of 4 pg/ml of aphidicolin (Sigma, A0781). (A) Clone 16. (B) Clone 3.

256

HARA gl2-21

0

4

8

I2

18

24

FIG. 8. Nuclear run-on transcription of clone 3, clone 16, and various c-one genes. The quiescent g12-21, gnl2RB1, and gn12RB2 cells were treated with dex and 3Yl cells with serum at time zero. At the times (b) indicated above each lane, the nuclei were isolated and 32P-labeled nascent RNAs synthesized were hybridized with 1 gg each of the probe DNA spotted on nitrocellulose filters. The probes for c-one genes were prepared as follows and “P-labeled by the multilabeling system. c-myc, the 0.8-kb BgliI-HpaI fragment of pSmBH [62] containing the rat c-myc exon 3; c-fos, the 5-kb EcoRI fragment cleaved from the plasmid, pc-fos[mouse]-3 1631;c-jun, the 5.4-kb fragment cleaved from the plasmid, ph-c-jun with EcoRI; c-r(ls, the 0.8-kh BarnHI fragment cleaved from the plasmid, pNCOlO2 1641 carrying the human melanoma c-Ha-ras-1. The plasmid ph-c-jun was a gift from Dr. Y. Taya.

cells treated with dex, ElA expression was induced within 4 h, reaching its maximal level after 6 to 8 h without activation of c-fos and c-myc expression. Clone 16 was expressed transiently at around 8 h. Expression of clone 3 was induced somewhat later and continued until S phase. Since expression of clone 16 was induced shortly after the level of ElA reached a maximum, it may be a direct target for ElA. These results indicate that ElA bypasses early in the Gl phase to enter the cell cycle from quiescence. The results are consistent with the previous reports that infection of cellular ts mutants with adenovirus or SV40 bypassed a temperature-sensitive block in mid-G1 phase at the restrictive temperature [ll, 12, 561 presumably by activating late Gl or S phase genes [13,57]. It has been recently reported that SV40 large T antigen acts downstream of c-fos in the signal-transducing cascade to induce DNA synthesis in human endometrial stromal cells [58]. Induction of clone 16 and clone 3 expression by the ElA 12s cDNA product lacking the transactivation domain suggests that the induction occurred through an interaction with a cellular negative regulator(s) for cell proliferation. This transactivation occurred at the level of transcription since the nuclear run-on assay gave essentially the same results as those obtained with North-

ET

AL.

ern blotting. The induction of clone 16 and clone 3 expression may occur at a relatively low level of ElA, since expression of these cellular genes was induced in dextreated gn12RB2 cells expressing a moderate level of ElA and a low level of exogenous Rb. However, gn12RB2 cells could not induce DNA synthesis in response to dex, suggesting that further progression of the cell cycle beyond the restriction point requires a higher level of ElA. It is also possible that the low level of RB may override the DNA synthesis-inducible activity of ElA regardless of the level of ElA. This possibility could be sorted out by fusing different promoters to these genes. The level of ElA expression in gnl2RBl cells was too low to induce expression of clone 16 and clone 3. In addition, the cells expressed a high level of the exogenous Rb in response to dex. The transient expression of clone 16 and clone 3 may be a prerequisite to override the restriction point. Bypass at the GlIS boundary. The inability of gn12RB2 cells to induce DNA synthesis in spite of having the ability to express clone 16 and clone 3 suggests that there is another step at which ElA bypasses the cell cycle. A higher level of ElA must be required for this bypass, since a much higher level of ElA was expressed in g12-21 cells than in gn12RB2 cells in response to dex. The restriction point may reside near the Gl/S boundary, since clone 16 and clone 3 seem to be expressed at middle to late Gl phase. The maximal level of clone 16 was expressed in both dex-treated g12-21 cells and serum-treated 3Yl cells at around 8 h after treatment, suggesting that these cells were cycling at the same stage in middle to late Gl phase. In contrast, the halfmaximal rate of DNA synthesis in dex-treated g12-21 cells was reached 12 to 13 h after treatment, while that in serum-treated 3Yl and g12-21 cells was reached 16 to 18 h after treatment. This difference in the time period for entering S phase from late Gl suggests the presence of the restriction point near the GllS boundary at which ElA enables the cells to override it efficiently. The products of the tumor suppressor genes RB and p53 seem to be involved in the transition of the cell cycle from late Gl to S phase. The hypophosphorylated form of Rb in GO/G1 phase has been assumed to lose its negative role in cell proliferation upon phosphorylation at the GUS boundary and enables late Gl cells to enter S phase [41-441. A similar cycle of hypophosphorylated and phosphorylated forms of p53 has been observed. Human p53 and RB are phosphorylated by ~34”~’ kinase at the GUS boundary [45, 59, 601. The entry of serum-stimulated 3Yl cells to S phase may be delayed until the cellular machinery including the phosphorylation system become ready to work. In the case of ElAinduced cell cycle progression, the hypophosphorylated form of RB may be directly inactivated by forming the complex with ElA.

INDUCTION

OF DNA SYNTHESIS

To isolate and characterize the function of cellular genes which preferentially express late in Gl and to the GUS boundary, we are currently isolating the complete cDNA sequences for these cellular genes including clone 16 and clone 3 in a subtracted cDNA library [61] constructed from g12-21 cells in late Gl phase. We thank Dr. T. Grodzicker for Ad2 12s cDNA, Dr. T. I?. Dryja for the human Rb gene cDNA, and Dr. R. Baserga for 2Fl cDNA.

REFERENCES 1. Bravo, R., Macdonald-Bravo, H., Muller, R., Hubsch, D., and Almendral, J. M. (1987) Exp. Cell Res. 170,103-115. 2. Cochran, B. H., Zullo, J., Verma, I. M., and Stiles, C. D. (1984) Science 226,1080-1082. 3. Greenberg, M. E., and Ziff, E. B. (1984) Nature (Londot~) 311, 433-438. 4. Kelly, K., Cochran, B. H., Stiles, C. D., and Leder, P. (1983) Cell 35,603-610. 5. Kruijer, W., Cooper, J. A., Hunter, T., and Verma, I. M. (1984) Nature (London) 32,711-716. 6. Muller, R., Bravo, R., Burckhardt, J., and Curran, T. (1984) Nuture(London)

312,716-720.

7. Nishikura, K., and Murray, J. M. (1987) Mol. Cell. Biol. 7,639649. 8. Heikkla, R., Schwab, G., Wickstrom, E., Lobe, S. L., Pluznik, D. H., Watt, R., and Neckers, L. M. (1987) Nature (London) 328,445-449. 9. Iguchi-Ariga, S. M. M., Itani, T., Kyi, Y., and Ariga, H. (1987) EMEO

J. 6,2365-2371.

10. Studzinski, G. P., Brelvi, Z. S., Feldman, S. C., and Watt, R. A. (1985) Science 234,467-470. 11. Rossini, M., Weinmann, R., and Baserga, R. (1979) PFOC. Natl. Acad.

Sci. USA

76,4441-4445.

12. Rossini, M., Jonak, G. J., and Baserga, R. (1981) J. Virol. 38, 982-986.

13. Liu, H. T., Baserga, R., and Marcer, W. E. (1985) Mol. Cell. BioL 5.2936-2942. 14. Bellet, A. J. D., Li, P., David, E. T., Mackey, E. J., Braithwaite, A. W., and Cutt, J. R. (1985) Mol. Cell. Biol. 5, 1933-1939. 15. Kaczmarek, L., Ferguson, B., Rosenberg, M., and Baserga, R. (1986) Virology 152, l-10. 16. Nakajima, T., Masuda-Murata, M., Hara, E., and Oda, K. (1987) Mol.

Cell. Biol.

7, 3646-3852.

17. Oda, K., Masuda-Murata, M., Shiroki, K., and Handa, H. (1986) J. VirvL 58, 125-133. 18. Stabel, S., Argos, P., and Phillipson, L. (1985) EMBOJ. 4,23292336. 19. Quinlan, M. P., and Grodzicker, T. (1987) J. Viral. 61,673-682. 20. Berk, A. J., and Sharp, P. A. (1978) Cell 14,695-711. 21. Chow, L. T., Broker, T. R., and Lewis, J. B. (1979) J. Mol. Biol.

134,265-303. 22. Kitchingman, G. R., and Westphal, H. (1980) J. Mol. Biol. 137, 23-48. 23. Pericaudet, M., Akusjarvi, G., Virtanen, A., and Petterson, U. (1979) Nature (London) 281,694-696.

BY ElA

257

24. Berk, A. J. (1986) Annu. Rev. Genet. 20,45-79. 25. Carlock, L. R., and Jones, N. C. (1981) J. ViroL 40,657-664. 26. LiIIie, J. W., Green, M., and Green, M. R. (1986) CeU 46,10431051. 27. LiiIie, J. W., Green, M., Lowenstein, M., and Green, M. R. (1987) Cell 50,1091-1100. 28. Montell, C., Fisher, E. F., Caruthers, M. H., and Berk, A. J. (1982) Nature (London) 295,380-384. 29. Moran, E., Zerler, B., Harrison, T. M., and Matthews, M. B. (1986) Mol. Cell. Biol. 6, 3470-3480. 30. Moran, E., and Mathews, M. B. (1987) Cell 48, 177-178. 31. Borelli, E., Hen, R., and Chambon, P. (1984) Nature (London)

312,608-612. 32. Velcich, A., and Ziff, E. (1985) CeU 40,705-716. 33. Subramanian, R., Kappuswamy, M., Nasr, R. J., and Chinnadurai, G. (1988) Oncogene 2,105-112. 34. Velcich, A., and Ziff, E. (1988) Mol. Cell. Biol. 8, 2177-2183. 35. Yee, S., and Branton, P. E. (1985) Virology 147,142-153. 36. Harlow, E., Whyte, P., Franza, B. R., Jr., and Schley, C. (1986) Mol. Cell. Biol. 6, 1579-1589. 37. Whyte, P., Buchkovich, K. J., Horowitz, J. M., Friend, S. H., Raybuck, M., Weinberg, R. A., and Harlow, E. (1988) Nature (London) 334, 124-129. 38. Eagn, C., Jelsma, T. N., Howe, J. A., Bayley, S. T., Ferguson, B., and Branton, P. E. (1988) Mol. Cell. BioL 8,3955-3959. 39. Whyte, P., Williamson, N. M., and Harlow, E. (1989) Cell 56, 57-75. 40. Lee, W.-H., Shew, J.-Y., Hong, F., Sery, T. W., Donoso, L. A., Young, L.-J., Bookstein, R., and Lee, E.Y.-H.P. (1987) Nature (London) 329,642-645. 41. Buchkovich, L., Duffy, L. A., and Harlow, E. (1989) Cell 58, 1097-1105. 42. Chen, P.-L., Scully, P., Shew, J.-Y., Wang, J. Y. J., and Lee, W.-H. (1989) Cell 58, 1193-1198. 43. DeCaprio, J. A., Ludlow, J. W., Lynch, D., Furukawa, Y., Griffin, J., Piwnica-Worms, H., Huang, C., and Livingston, D. M. (1989) Cell 58,1085-1095.

44. Ludlow, J. W., DeCaprio, J. A., Huang, C.-M., Lee, W.-H., Paucha, E., and Livingston, D. M. (1989) CeU 56,57-65. 45. Lin, B.T.-Y., Gruenwald, S., Morla, A. O., Lin, W.-H., and Wang, J. Y. J. (1991) EMBO J. 10,857-864. 46. Kimura, G., Itagaki, A., and Summers, J. (1975) Znt. J. Cancer

15,649-706. 47. Mulligan, R. C., and Berg, P. (1981) PFOC. Natl. Acad. Sci. USA 78,2072-2076. 48. Southern, P. J., and Berg, P. (1982) J. Mol. Appl. Genet. 1,327341. 49. Lee, F., Mulligan, R., Berg, P., and Ringold, G. (1981) Nature (Zmadon)

294,228-232.

50. Zerler, B., Maruyama, K., Moomaw, J., Grodzicker, T., and Ruley, H. E. (1986) Mol. Cell. Biol. 6, 887-899. 51. Friend, S. H., Bernards, R., Rogeli, S., Weinberg, R. A., Rappaport, J. M., Albert, D. M., and Dryja, T. P. (1986) Nature (London)

323,643-646.

52. Huynh, T. V., Young, R. A., and Davis, R. W. (1985) in DNA Cloning. A Practical Approach (Glover, D., Ed.), Vol. 1, pp. 4778, IRL, Oxford. 53. Benton, W. D., and Davis, R. W. (1977) Science 196,180-182. 54. Ayusawa, D., Shimizu, K., Koyama, K., Kaneda, S., Takeishi, K., and Seno, T. (1986) J. MoL BioL 190,559-567.

258

HARA ET AL.

55. Manser, T., and Gefter, M. L. (1984) Proc. Natl. Acad. Sci. USA 81.2470-2474. 56. Ohno, K., and Kimura, G. (1984) Somatic Cell Genet. 10,29-36. 57. Koniecki, J., Nugent, P., Kordowska, J., and Baserga, R. (1991) Cancer Res. 51,1465-1471. 58. Rinehart, C. A., Haskill, J. S., Morvis, J. S., Butler, T. D., and Kaufman, D. G. (1991) J. Virol. 65, 1456-1485. 59. Biechoff, J. R., Friedman, P. N., Marshak, D. R., Prives, C., and Beach, D. (1990) Proc. Natl. Acad. Sci. USA 87,4766-4770. 60. Sturzbecher, H.-W., Maimets, T., Chumakov, P., Brain, R., Addison, C., Simanis, V., Rudge, K., Philip, R., Crimaldi, M., Court, W., and Jenkins, J. R. (1990) Oncogene 6,795-801. Received July 29, 1991 Revised version received September 27, 1991

61. Hara, E., Kato, T., Nakada, S., Nakajima, T., Tsurui, H., and Oda, K. (1990) Nucleic Acids Symp. Ser. 22.29-30. 62. Hayashi, K., Makino, R., Kawamura, H., Arisawa, A., and Yoneda, K. (1987) Nucleic Acids Res. 15,6419-6436. 63. Miller, A. D., Curran, T., and Verma, I. M. (1984) Cell 36,51-60. 64. Sekiya, T., Fushimi, M., Hori, H., Hirohashi, S., Nishimura, S., and Sugimura, T. (1984) Proc. Natl. Acad. Sci. USA 81,47714775. 65. Battini, R., Ferrari, S., Kaczmarek, L., Calabretta, B., Chen, S.-T., and Baserga, R. (1987) J. Biol. Cbn. 262,4355-4359. 66. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156-159.