(2001) 20, 543 ± 550 ã 2001 Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc ORIGINAL PAPERS Regulation of the Cdc25A by the human papillomavirus Type 16 E7 oncogene

Stephanie C Katich1, Karin Zerfass-Thome1 and Ingrid Ho€mann*,1

1Angewandte Tumorvirologie (F0400), Deutsches Krebsforschungszentrum, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany

Cdc25A is a tyrosine that is involved in the (Ho€mann et al., 1993). Recent work indicates that regulation of the G1/ transition by activating Cdc25B seems to play the role of a `starter phosphatase' E/Cdk2 and /Cdk2 complexes through by activating /Cdk1 in order to initiate the removal of inhibitory phosphorylations. The E6 and E7 positive feedback mechanism (Lammer et al., 1998; oncoproteins of the high-risk human papillomaviruses Karlsson et al., 1999). are main (HPV) interact with and functionally abrogate the players of the G2 arrest caused by DNA damage or in the and pRB , respectively. In the present study we presence of unreplicated DNA (Sanchez et al., 1997; have investigated the regulation of the Cdc25A promoter Peng et al., 1997). Cdc25A and B are putative human during G1 and S-phases of the and by the (Galaktionov et al., 1995, 1997). The Cdc25A HPV-16 E7 oncoprotein. Serum induction leads to a phosphatase plays a crucial role at the G1/S phase derepression of the Cdc25A promoter and can be transition as shown by microinjection experiments using mediated through two binding sites, E2F-A and Cdc25A speci®c antibodies (Ho€mann et al., 1994) and E2F-C. While E2F-A is by both E2F1 and E2F4, E2F-C by conditional overexpression of Cdc25A (Blomberg and is regulated only by E2F1. The Cdc25A promoter is Ho€mann, 1999). Cdc25A functions by activating cyclin transactivated by the E7 oncogene of HPV-16. Further- E- and cyclin A-dependent (Blomberg and more, Cdc25A levels are highly increased in E7- Ho€mann, 1999). expressing cell lines. Inducible expression of E7 leads Studies on human papillomavirus type 16 have to an immediate increase in Cdc25A levels. These demonstrated that the product of the early gene E7, data suggest that Cdc25A may be a critical target of plays a key role in the immortalization and malignant HPV-16 E7 in the disruption of the G1/S phase transformation of the host cell. The E7 protein is the transition. Oncogene (2001) 20, 543±550. major oncogenic protein produced by cervical - associated HPV-16 (reviewed in zur Hausen, 1999). The Keywords: Cdc25; cell cycle; human papillomavirus majority of the transforming potential of E7 protein has type 16; E7 oncoprotein been ascribed to its ability to bind to, and down-regulate the retinoblastoma (RB) gene product (Boyer et al., 1996; Lee et al., 1998). Loss of RB function causes the Introduction release and activation of multiple E2F transcription factor family members, leading to cell cycle progression Transitions in the mammalian cell cycle are governed by through the G1/S phase transition. (Funk et al., 1997). cyclin-dependent kinases (Cdks) whose activities are Ectopic expression of E2F1, E2F2, E2F3 and to a lesser determined by cyclin binding, small Cdk inhibitor extent E2F4 is sucient to induce S-phase in immorta- proteins and both positive and negative regulatory lized quiescent rat ®broblasts (Lukas et al., 1996; phosphorylations (Nilsson, 2000). The Cdc25 phospha- DeGregori et al., 1997). that are responsive to tases are members of the tyrosine phosphatase family E2F include well-studied cell cycle regulators including and catalyze dephosphorylation and activation of cyclin/ , cyclin A, Cdk1 and Cdk2 (Dalton, 1992; Ohtani Cdk complexes through removal of inhibitory phosphor- et al., 1995; Schulze et al., 1995; Botz et al., 1996; Arata et ylations. In human cells, three Cdc25 homologs called al., 2000). Recently it has been shown that TGFb- Cdc25A, Cdc25B and Cdc25C have been identi®ed mediated down-regulation of Cdc25A in keratinocytes is (Sadhu et al., 1990; Galaktionov and Beach, 1991). mediated by an E2F site within the Cdc25A promoter Cdc25C dephosphorylates cyclin B/Cdk1, thereby acti- (Iavarone and Massague, 1999). Cdc25A is an essential vating its activity. Cdc25C in turn becomes target for E2F1, since its activity is required for ecient phosphorylated by cyclin B/Cdk1 thus creating a induction of S phase by E2F1 and RB family members positive feedback loop leading to entry into (Vigo et al., 1999; Chen and Prywes, 1999). In this study we show that the human Cdc25A promoter can be transactivated by the E7 oncogene of human papillomavirus type 16. This transactivation is *Correspondence: I Ho€mann Received 18 May 2000; revised 14 November 2000; accepted 23 mediated through two functional E2F binding sites November 2000 within the promoter. Furthermore Cdc25A levels are Regulation of Cdc25A by HPV-16 E7 SC Katich et al 544 highly increased in E7-expressing cell lines. Inducible prototype E2F site in nucleotide sequence and protein expression of E7 leads to an immediate increase in binding. An E2F-binding site with a similar sequence Cdc25A protein levels. These results suggest that the motif was previously found in the cyclin A promoter transcriptional regulation of the Cdc25A gene con- (Schulze et al., 1995; Figure 1, middle panel). To tributes to the control of cell cycle progression in determine which of the E2F sites is functional we virally transformed ®broblasts. transfected wild-type or mutant Cdc25A promoter constructs (nt 7457/+121) driving a luciferase reporter gene into NIH3T3 cells that were serum-starved and then released from the block. Samples were taken at the Results indicated time points. In Figure 2a the percentage of cells in G1, S and G2 or M phase are shown. While the Characterization of the functional E2F sites in the human Cdc25A promoter is repressed during G0 (Iavarone and Cdc25A promoter Massague, 1999; Figure 2b), the Cdc25A wild-type The region from 7457 to +121 of the human Cdc25A promoter construct was induced threefold at 16 h after promoter contains potential binding sites for known serum-induction when the cells were in mid S-phase. transcription factors, among them an E2F binding site These results are in agreement with previously published denoted E2F-A, which is located 60 bp upstream of the results (Chen and Prywes, 1999). Mutation of the E2F- transcriptional start site (Iavarone and Massague, A binding site (MUT 1) (Figure 1, lower panel) leads to 1999). The same authors also identi®ed a second E2F a similar fold of induction as found for the wild-type site (E2F-B) spanning the transcriptional start site. We promoter activity in such an experiment. This would have isolated the human Cdc25A promoter and suggest that another site of induction is present in the identi®ed an additional putative E2F binding site in the promoter, E2F-C, which is located at 160 bp upstream of the transcriptional start site in the promoter (Figure 1, upper panel). This site di€ers from the a

b

Figure 2 Cell cycle regulation of the Cdc25A promoter. (a) NIH3T3 cells were transfected with either the wild-type Cdc25A promoter or promoter constructs with a mutation in the E2F-A site (MUT1), E2F-C (MUT2) or mutations in both sites (MUT3). Figure 1 Sequence of the human Cdc25A promoter. Three E2F Cells were subsequently serum-starved for 3 days. After re- sites (denoted E2F-A, E2F-B and E2F-C) are boxed (upper addition of serum synchronized cells were harvested at di€erent panel). Comparison of the di€erent E2F binding sites found in the time points. Relative luciferase activity was measured. The results human Cdc25A promoter (middle panel) with the E2F consensus are the mean of three independent experiments. (b) Cell cycle sequence and a functional E2F binding site of the human cyclin A distribution. NIH3T3 cells were synchronized by serum-starvation promoter. Lower panel; Schematic representation of the E2F-site and released by addition of 10% calf serum. Samples were taken mutations within the Cdc25A promoter for ¯ow cytometry at the indicated time points

Oncogene Regulation of Cdc25A by HPV-16 E7 SC Katich et al 545 a Cdc25A promoter. To test this we mutated the E2F-C site, MUT 2 (Figure 1, lower panel), and determined the promoter activity. Mutation of both E2F binding sites (MUT3) (Figure 1, lower panel) only showed a slight increase in promoter activity (Figure 2b) which could be due to the third E2F site that is present in the promoter, E2F-B (Vigo et al., 1999). Taken together these results show that the E2F-A and E2F-C sites are both functional and required for E2F-induced S-phase entry.

Transactivation of the Cdc25A gene by the E7 oncoprotein of human papillomavirus type 16 Since we identi®ed functional E2F binding sites in the Cdc25A promoter we asked if expression of HPV-16 E7 would lead to an activation of Cdc25A transcrip- tion in NIH3T3 cells. An expression vector for HPV-16 E7 was co-transfected with the Cdc25A promoter construct into NIH3T3 cells and the relative luciferase activity was determined. As shown in Figure 3a a threefold increase of Cdc25A promoter activity was observed when HPV-16 E7 was co-transfected in b comparison to the co-transfection of an empty vector. In order to demonstrate that only a high-risk papillomavirus oncogene as HPV-16 E7 is able to transactivate the Cdc25A promoter we also co- transfected a protein from the low-risk papilloma virus, HPV-11 E7. As expected, HPV-11 E7 is not able to transactivate the Cdc25A promoter. The transactivation experiments shown in Figure 3a were repeated with vectors containing HPV-16 E7 or HPV- 11 E7 that were fused to a NH2-terminal epitope tag from in¯uenza hemagglutinin (HA). Similar results to c the HPV E7 proteins without a HA-tag were obtained (data not shown). The expression levels of both HPV- 11 E7 and HPV-16 E7 proteins after transient transfection into NIH3T3 cells was similar as shown by immunoblotting with the 3F10 monoclonal anti- body that detects speci®cally the HA-tagged proteins (Figure 3b). Having identi®ed two functional E2F binding sites in the Cdc25A promoter we compared the transactivation of HPV-16 E7 on the mutant Cdc25A promoter constructs. Figure 3c shows that the Cdc25A promoter activity is about ®ve times increased after transactiva- tion by E7 when either the E2F-A site (MUT 1) or the E2F-C site (MUT 2) are mutated. Only a low transactivation rate was observed when both E2F sites were mutated. These results indicate that both E2F sites can be used for E7-mediated transactivation of the Cdc25A gene.

HPV-16 E7 (lane 1) or HPV-11 E7 (lane 2) oncoproteins. Cell Figure 3 Transactivation of the human Cdc25A promoter by extracts were prepared from NIH3T3 cells after transfection. HPV-16 E7. (a) NIH3T3 cells were transfected with the Cdc25A Equal amounts of protein (40 mg) were resolved by SDS ± PAGE promoter construct and then cotransfected with either an HPV-16 and immunoblotted with an anti-HA rat monoclonal antibody E7 or an HPV-11 E7 expression vector or with an empty vector (3F10, Roche) or Cdk2 antibodies as a control. (c) Comparison of (CMV). Cells were washed after 18 h and then incubated for 24 h the transactivation of HPV-16 E7 on the Cdc25A wild-type in the presence of DMEM+0.5% calf serum. Relative luciferase promoter and the di€erent E2F-site mutations within the activities were measured. Shown are the averages of data from promoter (MUT 1 ± 3). Transient transfection experiments were three independent experiments. (b) Expression of HA-tagged performed as described in Figure 2a

Oncogene Regulation of Cdc25A by HPV-16 E7 SC Katich et al 546 Activation of Cdc25A by HPV-16 E7 a To further evaluate the role of the HPV-16 E7 oncoprotein in the modulation of Cdc25A gene expression we prepared total cell extracts from NIH3T3-derived E7-expressing (E7/2) cells (Lam et al., 1994) and control NIH3T3 cells. As expected HPV- 16 E7 is highly expressed in E7/2 cells and not present in control cells (Figure 4a). A higher expression rate was also detected for the G1 E and A in E7/2 cells. Similarly Cdc25A protein is highly expressed in the E7-overexpressing cell line which suggests that its expression may be modulated by the viral oncoprotein. To investigate the mode of action of HPV-16 E7 in cell transformation Crook et al. (1989) developed a system of inducible expression of the E7 oncoprotein in eukaryotic cells based on the ability of the mouse mammary tumor virus long terminal repeat (MMTV- LTR) to respond to glucocorticoid hormones. These cells, called 14/2 cells are derived from baby rat kidney cells and are arrested in G1 in the absence of hormone, and progress through S-phase with high synchrony upon addition of dexamethasone which induces the expression of E7. To investigate the e€ect of E7- b dependent cell transformation on Cdc25A gene activa- tion in 14/2 cells, expression of E7 was induced through addition of 1 mM of dexamethasone to the medium. Progression of cells into S-phase was monitored by FACScan analysis. Cells started to synthesize DNA at 10 h after addition of hormone (Table 1). RNA was prepared at regular time intervals after addition of dexamethasone. Expression of E7 and Cdc25A genes were monitored by Northern blotting (Figure 4b). E7 mRNA was absent in control preparations and was ®rst detected at 2 h after hormone addition at the same time as Cdc25A mRNA appeared. To examine whether there was a correspond- ing change in protein levels, cell extracts were prepared at regular time intervals after the addition of hormone. c Protein expression of E7, cyclins A and E and Cdc25A was monitored by immunoblotting (Figure 4c). The expression of E7 was detected at 3 h after the addition of dexamethasone while Cdc25A protein expression was ®rst observed after 3 ± 4 h. Cyclin E protein expression is present during the whole time course but the levels increased threefold at 3 h after hormone addition. Thus cyclin E expression is increased at the same time as Cdc25A expression ®rst appears, while cyclin A expression increases much later, at 10 h after addition of dexamethasone (Figure 4c). From the Figure 4 HPV-16 E7 dependent expression of Cdc25A and cyclins A and E. (a) Equal amounts of protein extracts (40 mg) timing of expression these results indicate that Cdc25A from NIH3T3 cells or E7/2 cells (E7-expressing NIH3T3 cells) could act as a direct target for E7-dependent induction were resolved by SDS ± PAGE and analysed by immunoblotting of S-phase entry. with antibodies against HPV-16 E7, Cdc25A, cyclins E and A. Antibodies against Cdk2 were used to monitor equal loading. (b) The HPV-16 E7 inducible cell line 14/2 was synchronized in G1 Table 1 E€ect of hormone addition on 14/2 cells arrested in G1 by withdrawal of dexamethasone. Cell proliferation was induced through addition of 1 mM dexamethasone. Samples were taken at Cell cycle Time after serum addition the time points indicated and probed for E7, Cdc25A and phase 0 h 1 h 2 h 3 h 4 h 5 h 6 h 10 h GAPDH mRNA expression. (c) Extracts were prepared at the indicated time points. Equal amounts of protein (40 mg) were Go/G1 84% 78% 79% 76% 72% 83% 77% 34% resolved by SDS ± PAGE and immunoblotted with antibodies S 17% 22% 21% 23% 22% 17% 23% 66% against E7, Cdc25A, cyclins E and A and Cdk2 G2/M 0% 0% 0% 1% 6% 0% 0% 0%

Oncogene Regulation of Cdc25A by HPV-16 E7 SC Katich et al 547 Regulation of the Cdc25A gene by the transcription To demonstrate that E7 is able to sequester p107 or factors E2F1 and E2F4 p130 from complex A, we incubated nuclear extracts from NIH3T3 or E7/2 cells with an oligonucleotide The motifs TTTGGCGC (E2F-A) and TCCCGC containing the E2F-A motif. As shown in Figure 5d, (E2F-C) are present at positions 762/755 and right panel, complex A is absent in extracts containing 7153/7158, respectively, of the human Cdc25A functional E7 protein. promoter (Figure 1). To analyse the binding of cellular proteins to these motifs, we performed bandshift experiments using the oligonucleotide containing either Discussion the E2F-A or the E2F-C motif with nuclear extracts from NIH3T3 cells. Figure 5a, upper panel shows the To further analyse how papillomavirus oncoproteins results using the E2F-A motif. It reveals a strong band disrupt cell cycle control we investigated the 5' that we identi®ed as unbound or `free E2F' since regulatory region of the human Cdc25A gene and preincubation of the nuclear extracts from NIH3T3 analysed its activation by the E7 oncogene of human cells with GST-RB leads to a retardation of the band. papillomavirus type 16. The Cdc25A promoter con- Free E2F, the smallest E2F complexes, are hetero- tains two functional E2F sites, denoted E2F-A and dimers containing a subunit encoded by the E2F gene E2F-C. Mutation of both sites in the promoter family and a subunit encoded by the DP family of abrogated its G1/S-speci®c induction and released its genes. In addition, experiments using speci®c anti- repression during G1. Transactivation of the Cdc25A bodies against human E2F1 and E2F4 lead to a promoter by E7 can be mediated through both E2F supershift of the free E2F band. No supershifts were binding sites. The protein levels of Cdc25A are observed when antibodies against E2F2 or E2F3 were markedly increased in NIH3T3 cells that are stably used. An additional cellular complex that binds to the transfected with E7. Moreover, dexamethasone-depen- E2F-A site, was identi®ed in the Cdc25A promoter dent activation of HPV-16 E7 gene expression in 14/2 (Figure 5a upper panel, complex A). We then cells leads to an immediate activation of Cdc25A gene performed bandshift experiments with the oligonucleo- expression. Taken together, these results suggest that tide containing the E2F-C motif (Figure 5a, lower Cdc25A is a crucial target of HPV-16 E7 in virus- panel). In those experiments we also detected a `free mediated transactivation of cells. E2F' which was shifted upon incubation with GST-Rb. A large number of promoters contain potential E2F Antibodies against the E2F1 but not E2F2, E2F3 or binding sites and, in a few cases, mutation of the E2F4 supershifted the free E2F band. We found two binding sites has con®rmed that E2F is critical for cell- cellular complexes that bind to the E2F-C site (Figure cycle regulated transcription. We have shown that two 5a lower panel, complexes B and C). In addition we of the three putative E2F binding sites in the Cdc25A analysed the formation of these cellular complexes promoter, i.e. E2F-A and E2F-C, are required for E2F- binding to either the E2F-A or E2F-C site in serum- mediated S-phase entry (Figure 2). Mutation of E2F-A starved and serum-stimulated cells. Figure 5b shows and E2F-C sites prevents an increase of Cdc25A that complex A is absent in serum-starved NIH3T3 promoter activity when cells enter S-phase. These results cells but its formation is markedly increased at 24 h indicate that the third E2F-site in the Cdc25A after serum-stimulation. Similarly, the complexes B and promoter, E2F-B, has only a small e€ect on S-phase. C that bind to the E2F-C site are absent in serum- Our results are in good agreement with the results starved cells but appear at 24 h after serum-stimula- published by Iavarone and Massague (1999) who tion. showed that the E2F-B site is not a functional E2F site To test which RB family member was present in in the Cdc25A promoter. Interestingly, the E2F-C site complex A we used antisera to p107 and p130 in of the Cdc25A promoter varies from the consensus E2F supershift assays. We found that both p107 (A')and site motif in nucleotide composition. A similar E2F site p130 (A'') were present in the complexes. In agreement motif represents also the functional E2F site in the with previously published data (Chen and Prywes, Cyclin A promoter. In the cyclin A promoter cell cycle 1999), we also show by supershift experiments that RB regulation is determined by this single promoter element itself is not present in the complex. If HPV-16 E7 (Schulze et al., 1995). Our results also indicate that the directly interacts with the E2F-A site of the Cdc25A cellular complexes that bind to either functional E2F promoter, incubation of E7 with the oligonucleotide binding site of the Cdc25A promoter are di€erent containing the motif should result in an increase of (Figure 5a). Cell lines that allow an inducible expression `free' E2F band. To demonstrate this we incubated of E2F1 fused to the estrogen receptor reveal that HPV-16 E7 that was expressed as fusion protein to Cdc25A is an E2F1 target gene and that Cdc25A is not glutathione-S-transferase (GST) with nuclear extracts transactivated by either E2F2 and E2F3 (Vigo et al., from NIH3T3 cells. Figure 5d, right panel shows that 1999). In our hands addition of antibodies to both E2F1 the amount of free E2F is markedly increased due to and E2F4 lead both to a supershift of the complexes the presence of E7 while no increase was observed that are bound to this site suggesting that this particular when GST alone was used or by incubation with an E7 E2F site is regulated by both E2F proteins. The other mutant, GST-E7(Gly24), that is de®cient of binding to functional E2F binding site, E2F-C, seems to be the pocket domain of pRB. regulated only by E2F1. Taken together, the E2F-A

Oncogene Regulation of Cdc25A by HPV-16 E7 SC Katich et al 548

ab

c

d

Figure 5 Cellular proteins interacting with the E2F-A and E2F-C binding sites of the Cdc25A promoter. (a) Identi®cation of free E2F binding to the E2F-A and E2F-C sites. A 32P-labelled oligonucleotide containing either the E2F-A site (upper panel) or the E2F-C site (lower panel) was incubated with nuclear extracts from NIH3T3 cells. Competition with excess unlabelled oligonucleotide carrying the wild-type or mutated version of the E2F-A site was performed to assess speci®city. The presence of free E2F was revealed by addition of 100 ng of a GST-RB pocket fusion protein. Antibodies to E2F1, E2F2, E2F3 and E2F4 were added as indicated. Speci®c complexes are marked. (b) NIH3T3 cells were serum starved and then induced with serum for the indicated time points. Extracts were incubated with a 32P-labelled oligonucleotide containing either the E2F-A site or the E2F-C site. (c) Nuclear extracts were incubated with a 32P-labelled oligonucleotide containing the E2F-A site and with antibodies against p107 or p130 as indicated. Speci®c complexes are marked on the left. (d) Left panel, E€ect of HPV-16 E7 on the levels of `free' E2F. Nuclear extracts from NIH3T3 cells were pre-incubated with GST alone, GST/HPV-16 E7 and GST/HPV-16 E7 (Gly24). A 32P- labelled oligonucleotide containing the E2F-A site was then incubated with these nuclear extracts. Right panel, analysis of the complexes binding to the E2F-A site in cell lines with or without expression of E7. A 32P-labelled oligonucleotide containing the E2F-A site was incubated with nuclear extracts of either NIH3T3 or E7/2 cells

Oncogene Regulation of Cdc25A by HPV-16 E7 SC Katich et al 549 site in the Cdc25A promoter seems to be regulated MgSO4,4mM EGTA, 1 mM DTT, 1% Triton X-100) and di€erently from the second functional E2F site in the luciferase activity was measured in triplicate samples. promoter (E2F-C). In the future, it will be interesting to Variations in transfection eciency were accounted for by perform a detailed analysis of the molecular compo- normalizing to a cotransfected cytomegalovirus promoter- nents that are bound to the E2F-C site. driven b-galactosidase control plasmid (pCMV-bgal) (Ren- trop, DKFZ, Heidelberg). Overexpression of Cdc25A in rat-1 cells leads to an immediate activation of cyclin E and A- dependent kinase activities (Blomberg and Ho€mann, Reporter plasmids and expression vectors 1999) suggesting that these Cdk/cyclin complexes act as in vivo substrates of Cdc25A. Inducible over- To obtain the Cdc25A wildtype promoter construct, a human pCYPAC library (Ioannou et al., 1994) was screened with a expression of HPV-16 E7 in the BRK cell line 14/2 32P-labelled 386 bp probe (1620 ± 2006 bp of the Cdc25A leads to a rapid increase in Cdc25A protein levels at open reading frame). The DNA of a positive clone was used the same time when cyclin E protein levels augment as a template in a PCR using the primers PROM1 (5'- but way before cyclin A protein levels increase. CTCGAATCCGGCCAGACCTCC-3') and PROM2 (5'-AA- Overexpression of the Cdc25A gene induced by E7 GCTTCGAGGCAACGGCCCAG-3'). After sequencing, the would then lead to an increase in Cdc25A PCR product was cloned into the pBL luciferase reporter phosphatase activity which in turn activates cyclin vector (Hoppe-Seyler and Butz, 1992). The promoter E and cyclin A dependent kinases leading to entry constructs MUT1, MUT2 and MUT3 were also generated into S-phase. This study provides new insights into by PCR using the additional primers E2FM1 (5'- the mechanisms of how viral oncogenes may over- TTCCTAGTTGGAAGGTAACGGAATCCA-3'), E2FM2 (reverse to E2F1), E2FM3 (5'-AAAGCGGCTCGATATCT- come checkpoint control mechanisms during G1 to CTGCCAGGT-3') and E2FM4 (reverse to E2F3) to delete induce entry into S-phase. E2FA, E2FC or both sites. The expression vector pX/HPV-16 E7 was a gift of P Jansen-DuÈ rr (Innsbruck, Austria). In order to detect its proper protein expression, a hemagglutinin (HA)-tag was Materials and methods cloned in front of the HPV-16 E7 open reading frame. The expression vectors pX/HPV-11 E7 and pX/HPV-11 Cell culture and cell cycle analysis HA-E7 were constructed by PCR using HPV-11 E7 contain- ing DNA as a template. The resulting PCR products were NIH3T3 cells were obtained from the American Type Culture cloned into the pX vector (G Superti-Furga, EMBL Collection. The E7/2 clone, expressing wild type HPV-16 E7 Heidelberg, Germany). was obtained from K Vousden (MD, USA). Both cell types were grown in Dulbecco's modi®ed Eagle's medium (DMEM) supplemented with 10% calf serum, 2 mM glutamine, Electrophoretic mobility shift assay penicillin (100 U/ml) and streptomycin (100 mg/ml). The BRK cell line 14/2 was a kind gift from L Banks (Trieste, The bandshift shift assays were performed using NIH3T3 or Italy). The cell line was grown in DMEM, supplemented as E7/2 nuclear extracts. Pellets of exponentially grown cells above, containing 1 mM dexamethasone. 14/2 cells were were resuspended in ice cold bu€er A (10 mM HEPES, synchronized by hormone withdrawal for 48 h; cell prolifera- pH 7.9; 10 mM KCl; 0.1 mM EDTA; 0.1 mM EGTA; 1 mM tion was induced by readdition of 1 mM dexamethasone. PMSF) and incubated 15 min on ice. To lyse the cells, 10% Cell cycle progression was monitored by ¯uorescence- NP40 was added carefully, immediately mixed and left on activated cell sorter (FACS) analysis, using a Becton ice for another 5 min. The lysate was then centrifuged at Dickinson FACScan system, the Cell Quest and ModFit 2.0 48C at 13 000 rpm for 2 min. The remaining pellet contain- software. The blocked or induced cells were trypsinized, ing the nuclei was resuspended in bu€er C (20 mM HEPES washed and resuspended in 0.2 ml phosphate-bu€ered saline pH 7.9; 0.4 M NaCl; 1 mM EDTA, 1 mM EGTA, 1 mM (PBS). Fixation occurred by addition of 0.8 ml of ice-cold DTT, 1 mM PMSF), incubated on ice for 20 min and methanol and incubation at 48C for at least 20 min. After centrifuged at 48C at 13 000 rpm. The supernatant represent- centrifugation the pellet was resuspended in 1 ml of PI mix ing the nuclear extract was then tested by Western blot for (50 mg of propidium iodide per ml, 10 mM Tris-HCl, pH 7.5; purity. 6 5mM MgCl2;10mg of RNase A per ml) per 10 cells. Double stranded oligonucleotides E2F1B/2B (5'-GAT- CACCTGGCAGAGTCCCGCGAGCCGCT-3') or E2F3B/ 4B (5'-GATCTGGATTCCGTTTGGCCCAACTAGGAA-3') Transient transfections and luciferase assay were 32P-end-labelled with Klenow and diluted in NIH3T3 cells were plated on six well plates at a density of 50% glycerol to a concentration of 20 000 c.p.m./ml. Two ml 16105 cells and transfected using the E€ectene Kit from of the labelled probe was mixed with 10 mg of nuclear Qiagen (Hilden, Germany). Two mg of DNA were diluted in extract, 100 or 200 ng of sonicated salmon sperm DNA and EC bu€er to a ®nal volume of 100 ml. To condense the DNA 1 mlof20mM MgCl2 to a ®nal volume of 10 ml and Enhancer (8 ml/mg DNA) was added and mixed. Then the incubated at room temperature for 10 min. The reaction mix transfection reagent E€ectene (2 ml/mg DNA) was added and was separated by electrophoresis in a 4.5% native poly- incubated for 10 min. Six hundred ml of media was added to the acrylamide gel at 48C. After ®xation in a 10% methanol/ transfection mix and distributed to each well containing 1.6 ml 20% ethanol solution for 15 min and drying, the gel was media. After 15 ± 20 h the cells were washed twice with serum exposed to ®lm. free media and fresh media containing 10% calf serum, or for For supershifts, the extracts were preincubated with transactivation assays 0.5% serum was added for another 24 h. speci®c antibodies prior to the addition of labelled Cells were lysed in Triton bu€er (25 mM Tris pH 7.5, 15 mM oligonucleotides for 45 min at 48C. Monoclonal antibodies

Oncogene Regulation of Cdc25A by HPV-16 E7 SC Katich et al 550 to p130(C20x), and E2F1(C-20x) E2F2(C-20x), E2F3(N-20x) Western blotting were obtained from Santa Cruz Biotechnology, hybridoma supernatants E2F4 (WUF11) and p107 (SD15) were gifts Whole cell extracts were prepared by lysing cell pellets in from N Dyson (Charlestown, MA, USA). RIPA bu€er (10 mM Tris-HCl pH 8.0; 150 mM NaCl; 1 mM For competition assays, 100 ng of unlabelled wild-type or EDTA; 1% NP-40; 0.1% SDS; 50 mM NaF, 0.1 mM mutated oligonucleotides (E2FM1/E2FM2) were preincu- Na3VO4,1mM DTT, 0.1 mM PMSF). After clearing o€ the bated in the same way as the antibodies. lysate by centrifugation, the protein content in the lysate was One hundred ng of recombinant proteins GST, GST-RB determined by the Bradford method. A 40 mg quantity of (379 ± 928, containing the pocket domain of pRB), GST- protein extract were separated by SDS ± PAGE, blotted to HPV16 E7 or GST-Gly24 (an HPV-16 E7 mutant, which is nitrocellulose and probed with monoclonal antibodies to unable to bind the pRB pocket domain) were preincubated in Cdc25A (F6, Santa Cruz Biotechnology), HPV-16 E7 (8C9, the same way as the antibodies. Zymed), cyclin E (M-20, Santa Cruz Biotechnology), HA-tag (3F10, Roche) and polyclonal antisera to cyclin A (Pagano et al., 1992). Equal loading of the samples was determined by Northern blotting staining with Ponceau S and antibodies against Cdk2. Total cellular RNA was prepared by the guanidinum Samples were detected with a peroxidase-linked chemilumi- thiocyanate-acid phenol method. Total RNA samples nescence detection system (Santa Cruz Biotechnology). (10 mg) were separated on formaldehyde agarose gels and transferred to Hybond N+ membranes (Amersham Pharma- cia) by capillary transfer. Acknowledgments A 340 bp EcoRI ± HindIII E7 cDNA fragment was used for We are grateful to H zur Hausen for support and helpful detection of HPV-16 E7 expression. Expression of Cdc25A comments on the manuscript. L Banks, N Dyson, P was monitored with a human probe, and glyceraldehyde-3- Jansen-DuÈ rr, G Superti-Furga and K Vousden are thanked phosphate dehydrogenase (GAPDH) expression was analysed for providing reagents. We thank F RoÈ sl and C O'Lough- by using a rat cDNA probe (Spitkovsky et al., 1996). lin for critically reading the manuscript.

References

Arata Y, Fujita M, Ohtani K, Kijima S and Kato JY. (2000). Karlsson C, Katich S, Hagting A, Ho€mann I and Pines J. J. Biol. Chem., 275, 6337 ± 6345. (1999). J. Cell. Biol., 146, 573 ± 584. Blomberg I and Ho€mann I. (1999). Mol. Cell. Biol., 19, Lam EW, Morris JD, Davies R, Crook T, Watson RJ and 6183 ± 6194. Vousden KH. (1994). EMBO J., 13, 871 ± 878. Botz J, Zerfass-Thome K, Spitkovsky D, Delius H, Vogt B, Lammer C, Wagerer S, Sa€rich R, Mertens D, Ansorge W Eilers M, Hatzigeorgiou A and Jansen-Durr P. (1996). and Ho€mann I. (1998). J. Cell. Sci., 111, 2445 ± 2453. Mol. Cell. Biol., 16, 3401 ± 3409. Lee JO, Russo AA and Pavletich NP. (1998). Nature, 391, Boyer SN, Wazer DE and Band V. (1996). Cancer Res., 56, 859 ± 865. 4620 ± 4624. Lukas J, Petersen BO, Holm K, Bartek J and Helin K. (1996). Chen X and Prywes R. (1999). Mol. Cell. Biol., 19, 4695 ± Mol. Cell. Biol., 16, 1047 ± 1057. 4702. Nilsson I and Ho€mann I. (2000). Prog. Cell Cycle Res., Crook T, Morgenstern JP, Crawford L and Banks L. (1989). Meijer L (ed.), 4, 107 ± 115. EMBO J., 8, 513 ± 519. Ohtani K, DeGregori J and Nevins JR. (1995). Proc. Natl. Dalton S. (1992). EMBO J., 11, 1797 ± 1804. Acad.Sci.USA,92, 12146 ± 12150. DeGregori J, Leone G, Miron A, Jakoi L and Nevins JR. Pagano M, Pepperkok R, Verde F, Ansorge W and Draetta (1997). Proc. Natl. Acad. Sci. USA, 94, 7245 ± 7250. G. (1992). EMBO J., 11, 961 ± 971. Funk JO, Waga S, Harry JB, Espling E, Stillman B and Peng CY, Graves PR, Thoma RS, Wu Z, Shaw AS and Galloway DA. (1997). Genes Dev., 11, 2090 ± 2100. Piwnica-Worms H. (1997). Science, 277, 1501 ± 1505. Galaktionov K and Beach D. (1991). Cell, 67, 1181 ± 1194. Sadhu K, Reed SI, Richardson H and Russell P. (1990). Galaktionov K, Lee AK, Eckstein J, Draetta G, Meckler J, Proc. Natl. Acad. Sci. USA, 87, 5139 ± 5143. Loda M and Beach D. (1995). Science, 269, 1575 ± 1577. Sanchez Y, Wong C, Thoma RS, Richman R, Wu Z, Gasparotto D, Maestro R, Piccinin S, Vukosavljevic T, Piwnica-Worms H and Elledge SJ. (1997). Science, 277, Barzan L, Sulfaro S and Boiocchi M. (1997). Cancer Res., 1497 ± 1501. 57, 2366 ± 2368. Schulze A, Zerfass K, Spitkovsky D, Middendorp S, Berges Ho€mann I, Clarke PR, Marcote MJ, Karsenti E and J, Helin K, Jansen-Durr P and Henglein B. (1995). Proc. Draetta G. (1993). EMBO J., 12, 53 ± 63. Natl. Acad. Sci. USA, 92, 11264 ± 11268. Ho€mann I, Draetta G and Karsenti E. (1994). EMBO J., Spitkovsky D, Jansen-DuÈ rr P, Karsenti E and Ho€mann I. 13, 4302 ± 4310. (1996). Oncogene, 12, 2549 ± 2554. Hoppe-Seyler F and Butz K. (1992). Nucl. Acids Res., 20, VigoE,MullerH,ProsperiniE,HateboerG,CartwrightP, 6701 ± 6706. Moroni MC and Helin K. (1999). Mol. Cell. Biol., 19, Iavarone A and Massague J. (1999). Mol. Cell. Biol., 19, 6379 ± 6395. 916 ± 922. zur Hausen H. (1999). Semin. Cancer Biol., 9, 405 ± 411. Ioannou PA, Amemiya CT, Garnes J, Kroisel PM, Shizuya H, Chen C, Batzer MA and de Jong PJ. (1994). Natl. Genet., 6, 84 ± 89.

Oncogene