Oncogene (2007) 26, 781–787 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc SHORT COMMUNICATION Relationship between E1A binding to cellular proteins, c- activation and S-phase induction

S Baluchamy1,3,4, N Sankar1,3, A Navaraj1,5, E Moran2 and B Thimmapaya1

1Department of Microbiology–Immunology Feinberg School of Medicine, Northwestern University, Chicago, IL, USA and 2Fels Institute for Cancer Research and Molecular Biology, Temple University School of Medicine, Philadelphia, PA, USA

We recently showed that p300/CREB-binding protein Cell transformation by the adenoviral E1A oncoprotein (CBP) plays an important role in maintaining cells in G0/ is dependent on its binding to and inactivating several G1 phase by keeping c-myc in a repressed state. host proteins including p400, p300, pRb, and the two Consistent with this, adenovirus E1A oncoprotein induces Rb family proteins p130 and p107, PCAF, TRRAP and c-myc in a p300-dependent manner, and the c-myc GCN5 (reviewed in Moran, 1993; Barbeau et al., 1994; induction is linked to S-phase induction. The induction Frisch and Mymryk, 2002); for an extensive list of of S phase by E1A is dependent on its binding to and references on this topic please refer to the web site inactivating several host proteins including p300/CBP. To http://www.geocities.com/jmymryk.geo/. These host determine whether there is a correlation between the host proteins are found in distinct chromatin remodeling proteins binding to the N-terminal region of E1A, complexes that modulate expression either by activation of c-myc and induction of S phase, we assayed activating or repressing transcription. In most cases, the c-myc and S-phase induction in quiescent human cells specific transcriptional targets of these complexes in the by infecting them with Ad N-terminal E1A mutants with context of cell cycle and the mechanism by which E1A mutations that specifically affect binding to different interferes in their function are not known. pRb, p130 chromatin-associated proteins including pRb, p300, p400 and p107proteins in association with deacety- and p300/CBP-associated factor (PCAF). We show that lases repress E2F1 function that is critical for the the mutants that failed to bind to p300 or pRb were induction of S phase (Harbour, 2001). p400 is a severely defective for c-myc and S-phase induction. The component of the human TIP60/NuA4 complex con- induction of c-myc and S phase was only moderately taining histone acetyl transferase (HAT) activity that affected when E1A failed to bind to p400. Furthermore, modulates both transcriptional repression and activa- analysis of the E1A mutants that fail to bind to p300, and tion (Fuchs et al., 2001; Doyon et al., 2004). p300 and both p300 and PCAF suggests that PCAF may also play a CREB-binding protein (CBP) are two large highly role in c-myc repression, and that the two chromatin- homologous nuclear phosphoproteins containing associated proteins may repress c-myc independently. In HAT activity that function as transcriptional coactivators summary, these results suggest that c-myc deregulation by and link chromatin remodeling with transcription E1A through its interaction with these chromatin-asso- (Goodman and Smolik, 2000). TRRAP is found in ciated proteins is an important step in the E1A-mediated three distinct human HAT complexes including the cell cycle deregulation and possibly in cell transformation. TIP60 HAT complex and two other complexes that are Oncogene (2007) 26, 781–787. doi:10.1038/sj.onc.1209825; similar to yeast SAGA (Spt-Ada-Gcn5-acetyl transferase published online 24 July 2006 complex) containing either GCN5 or PCAF (reviewed in Sterner and Berger, 2000). Keywords: E1A; p300; c-myc repression; cell cycle Until recently, the significance of E1A binding to p300 in the E1A-mediated cell cycle deregulation and cell transformation was not known. We recently showed that in quiescent human cells, p300/CBP plays an important role in keeping c-myc in a repressed state. Depletion and induction of p300/CBP in serum-starved cells led to induction and repression, respectively, of Correspondence: Dr B Thimmapaya, Department of Microbiology— c-myc and DNA synthesis (Kolli et al., 2001; Baluchamy Immunology, Feinberg School of Medicine, Northwestern University, et al., 2003; Rajabi et al., 2005). Consistent with this, we 303 East Chicago Avenue, Olson 8452, Chicago, IL 60611, USA. E-mail: [email protected] showed that in quiescent cells, E1A induced c-myc in a 3These two authors contributed equally to this work. p300-dependent manner indicating that E1A interferes 4Current address: Department of Genetics and Biochemistry, Uni- with p300/CBP repression of c-myc (Kolli et al., 2001). versity of Illinois at Chicago, Chicago, IL, USA. To determine whether or not there is a correlation 5Current address: Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA, USA. between E1A binding to host proteins, and induction of Received 28 September 2005; revised 9 June 2006; accepted 11 June 2006; c-myc and S phase, we have studied a series of previously published online 24 July 2006 characterized E1A mutants for their capacity to induce c-myc activation by adenovirus E1A S Baluchamy et al 782 c-myc and to enter S phase. Our results show that the either of the mutations alone; Wang et al., 1995), it was E1A mutants that do not bind to p300 or Rb do not as defective as 928 in inducing c-myc, suggesting that the induce c-myc and S phase, whereas a mutant that does effects of RG2 mutation were further reduced by 928 not bind to p400 induces both c-myc and cell cycle mutation suggesting that WT E1A may recruit both efficiently. In addition, our studies suggest that E1A p300 and Rb simultaneously as suggested earlier (Wang binding to PCAF may also independently contribute to et al., 1995). In contrast, mutant dl2–36 (aa 2–36 c-myc induction. deleted) that binds to neither p300 nor p400 showed a Figure 1 shows a schematic of the E1A mutants dramatic decrease in c-myc RNA levels that was analysed in this study that were characterized exten- comparable to that of control samples. All of the E1A sively by several laboratories with respect to their mutants used in these studies expressed their respective capacity to bind to different host proteins in vivo, proteins, at reasonable levels as shown by a Western induce DNA synthesis and transform rodent cells in blot in Figure 2d. culture in the presence of activated ras (references cited Comparison of the c-myc induction and the protein above; also see Egan et al., 1988; Fuchs et al., 2001; binding properties of the mutant dl26–35, RG2 and dl2– Lang and Hearing, 2003). To determine the capacity of 36 suggested that host proteins other than Rb and p300 these mutants to induce c-myc and DNA synthesis in might also repress c-myc. For example, mutant dl26–35 human cells, we used MCF10A cells (an immortalized induces c-myc to near-normal levels (Figure 2a), and it non-transformed human breast epithelial cell line; Soule does not bind to p400, and also TRRAP and GCN5; its et al., 1990) that can be growth arrested by serum p300 and PCAF binding capacity is not affected (Lang starvation and show normal serum-stimulated induction and Hearing, 2003). RG2, which retains about 40% of of c-myc and cell cycle (Kolli et al., 2001). its Myc-inducing capacity, binds to p400 at normal To determine the levels of c-myc induced by the E1A levels, whereas it does not bind to p300 (TRRAP and mutants, serum-starved MCF10A cells were infected GCN5 binding is not tested). In contrast, dl2–36 that is with Ad mutants described above for 12 h, and the RNA inactive in inducing c-myc does not bind to both p400 samples were analysed using a quantitative real-time and p300. As dl2–36 also contains dl26–35 deletion, it is PCR assay as described in the legend to Figure 2. An Ad not expected to bind to TRRAP and GCN5 as well. vector expressing beta-galactosidase (Adb-gal) was used Thus, it seemed possible that the total loss of c-myc as a control, and the fold-increase in c-myc induction by induction by dl2–36 may be owing to its inability to bind the E1A mutants was compared to that of Adb-gal- to p300 and another protein such as PCAF as PCAF infected cells. As shown in Figure 2a, there was about and p300 binding regions on E1A overlap (Lang and sixfold induction of c-myc in WT-infected cells. In Hearing, 2003). To test this possibility, MCF10A cells contrast, in cells infected with RG2 (arginine-2 changed were infected for 16 h with WT, RG2 and dl2–36 along to glycine) that does not bind to p300, c-myc induction with an Ad vector expressing FLAG epitope-tagged was only about 2.5-fold. Similarly, the mutant 928 PCAF from the cytomegalovirus (CMV) promoter (cysteine-128 changed to glycine) that does not bind to (AdPCAF). The cell lysates were immunoprecipitated Rb also was severely defective for c-myc induction with with an anti-FLAG antibody followed by Western a 1.7-fold increase of c-myc RNA levels, suggesting a immunoblotting with an anti-E1A antibody. As shown role for pRb in c-myc repression. E1A mutant that does in Figure 2e, both WT and RG2 E1A proteins bound to not bind to p400 (dl1102, deletion of aa 26–35; referred PCAF at comparable levels. Reverse experiment (Im- to here as dl26–35) induced near-normal levels of c-myc munoprecipitation with anti-E1A and Western with RNA (5.5-fold). When RG2 and 928 mutations are anti-FLAG) also indicated that RG2 binds to PCAF at combined in one molecule (RG2.928; in baby rat kidney near-normal levels. A Western blot experiment shown in cells, this mutant is more defective in cell cycle than Figure 2e indicated comparable levels of expression of

E1A Mutant E1A Binding Proteins p400 p300Rb PCAF CR1 CR2

1 40 80 120 140 243 WT +++ +++ +++ +++ dl26 - 35 - +++ +++ +++ 26 35 RG2 +++ - +++ +++ *2 dl2-36 - - +++ - 236 E55 +++ + +++ +++ 55 60 (Ala) 928 +++ +++ - ND 128* RG2 - 928 +++ - - ND **2 128 Figure 1 E1A mutants showing mutations used in this study and their capacity to bind to p400, p300, PCAF and pRb. Data for binding to p400, p300, PCAF and pRB are taken from published reports (references cited in text).

Oncogene c-myc activation by adenovirus E1A S Baluchamy et al 783 a d c-Myc 6

-gal T G2 G2-928 β W R dl2-36 928 R dl26-35 E55 4

E1A

Fold Increase 2 Actin

0

WT β-gal RG2 928 E55 dl2-36 dl26-35 RG2-928 b e c-jun

2-36 G 1.5 WT RG2 dl Ig IP: α-FLAG IgG WB: αE1A 1.0 E1A

IP: αE1A WB: α-FLAG PCAF

Fold Increase 0.5 E1A Input (5%) 0 PCAF

WT β-gal RG2 928 E55 dl2-36 dl26-35 RG2-928 c f

12 %) %) t (5 t (5 ) E1A FLAG E1A FLAG

5 - - - - IP: Inpu α α IgG Inpu α α IgG 10 × 8 IgG WB: α-E1A E1A 4 WB: α-FLAG PCAF

Luc activity ( WT E55 0

WT β-gal RG2 928 E55 dl2-36 dl26-35 RG2-928 Figure 2 Quantification of c-myc and c-jun mRNA levels by real-time PCR (a and b) and luciferase activity in response to c-myc activation (c). (d) Western blot showing E1A protein levels in cell lysates prepared from virus-infected cells, (e and f) PCAF interaction with the WT and the E1A mutant proteins. MCF10A cell were seeded overnight, serum-starved for 36 h and then infected with Ad variants at 10 PFU per cell as described (Rajabi et al., 2005). At 12 h after infection, total RNA was isolated and 2 mg of RNA was reverse transcribed by Retroscript First strand synthesis kit (Ambion, Austin, USA), and 1/20th volume of the cDNA was used as template for the real-time PCR assay. Reactions were carried out in a final volume of 20 ml in SYBR Green Jump Start Taq Ready mix (Sigma) using ABI Prism 7700 Sequence Detection System. Fold increase was calculated by 2ÀDDCt method (Applied Biosystems). The following primers were used for RT–PCR amplification. c-myc, forward: 50-gccacgtctccacacatcag; c-myc, reverse: 50-tggtgcattttcggttgttg; c-jun, forward: 50-aggaggagcctcagacagtg; c-jun, reverse: 50-agcttcctttttcggcactt. Glyceraldehyde 3 phosphate dehydrogenase (GAPDH), forward: 50-gtgaaggtcggagtcaacg; GAPDH, reverse: 50-tgaggtcaatgaaggggtc. The experiment was carried out in triplicate and the average values with s.d.’s are shown. (c) Myc activity was assayed by measuring the luciferase activity in virus- infected cells by using a Myc responsive Ad vector (AdM4; see text for details). Assays were carried out in triplicate and the average values with s.d.’s are shown. (e and f) Proliferating MCF10A cells were infected with WT or E1A mutants along with AdPCAF for 24 h, and 1 mg of proteins from each cell lysate was immunoprecipitated with antibodies as shown, followed by Western immunoblotting. Anti-E1A Ab M73 (obtained from E Moran) and anti-FLAG Ab (Sigma, cat. no. M4) were used in these assays. For input in (e and f), cell lysates equivalent to 50 mg protein were loaded directly onto the gel and Western immunoblotted with antibodies as shown.

PCAF in WT- and the mutant-infected cells (input in (Lang and Hearing, 2003). As shown in Figure 2a, c-myc Figure 2e). In summary, the significantly reduced c-myc induction in E55-infected cells is reduced to about 50% induction in RG2-infected cells and the total lack of of the WT, consistent with its weak interaction with c-myc induction in dl2–36-infected cells suggested that p300. To ensure that E55 binding to PCAF in virus- PCAF might also contribute to c-myc repression. infected MCF10A cells was unaffected, cells were We also studied another E1A mutant in which aa coinfected with E55 and AdPCAF, lysed and immuno- 55–60 were changed to alanine (E55); it binds to PCAF precipitated with an anti-E1A antibody. The immuno- at normal levels whereas it binds to p300 very weakly precipitated proteins were Western immunoblotted with

Oncogene c-myc activation by adenovirus E1A S Baluchamy et al 784 anti-FLAG antibody. As shown in Figure 2f, and in To determine whether E1A binding to different host agreement with a previous report (Lang and Hearing, proteins affects another early response gene, the RNA 2003), the amount of PCAF coimmunoprecipitated with samples used in the c-myc experiments were analysed for anti-E1A antibody was comparable in WT- and E55- c-jun expression using real-time PCR assay. The results infected cell lysates. A separate Western blot experiment presented in Figure 2b show that the induction levels of confirmed that E55 binds to p300 very weakly (data c-jun in Adb-gal-, WT- and the mutant-infected cells not shown). Thus, the p300 and PCAF binding pro- were all comparable (less than 1.5-fold difference). perties and the Myc induction capacity of RG2 and These results are consistent with our previous report E55 mutants are very similar. Because RG2 and E55 (Baluchamy et al., 2003) and suggest that relief of retain almost normal capacity to bind to PCAF, and repression of by E1A is not global. are defective for p300 binding, it suggests that E1A We also assayed Myc activity levels in virus-infected can bind to p300 and PCAF independently, which is cells by infecting cells with WT and various E1A in agreement with previous reports (Reid et al., 1998; mutants along with a c-myc reporter Ad vector Lang and Hearing, 2003). Further, these results (AdM4; Baluchamy et al., 2003) and assaying luciferase also suggest that these two chromatin proteins repress activity at 14 h after infection. As shown in Figure 2c c-myc independently. The repression effects may be Myc activity in RG2- and E55-infected cells dropped to additive as a mutant that does not bind either p300 38 and 42%, respectively. In cells infected with dl2–36 or PCAF (dl2–36) lacks c-myc induction capacity and 928, Myc activity levels dropped to control levels. completely. The mutant dl26–35 that is defective for binding to

a 40

35

30

25

20

15

10 % of cells in S phase

5

0 WT WT WT -gal -gal -gal 928 928 928 E55 E55 E55 RG2 RG2 RG2 β β β dl2-36 dl2-36 dl2-36 dl26-35 dl26-35 dl26-35 RG2-928 RG2-928 RG2-928

20 24 hr 28 hr

b

G0/G1:93 G0/G1:62 G0/G1:87 G0/G1:91 S:02 S:35 S:08 S:04 G2/M:05 G2/M:03 G2/M:05 G2/M:05

β - gal WT RG2 dl2-36

G0/G1:92 G0/G1:89 G0/G1:84 G0/G1:85 S:04 S:05 S:33 S:10 G2/M:04 G2/M:07 G2/M:03 G2/M:05

928 RG2-928 dl26-35 E55 Figure 3 Induction of S phase in the WT- and the mutant-infected cells. (a) Bar diagram showing the number of WT- or the E1A mutant-infected cells in S phase at 20, 24 and 28 h after virus infection. (b) Cell cycle profiles for cells infected with WT and the mutant viruses at 28 h post-infection. Serum-starved MCF10A cells were infected with Ad variants at 10 PFU per cell as described above, and at 20, 24 and 28 h after infection, cells were harvested, and the distribution of cells in G1, S and G2/M phases was quantified by flow cytometry as described. Experiments were carried out twice in triplicate and the average values with s.d.’s are shown in (a).

Oncogene c-myc activation by adenovirus E1A S Baluchamy et al 785 24h

80 G0/G1-S-G2/M

β-gal + β-gal 97/1/2

60 β-gal + c-Myc 62/32/3

β-gal + WT 62/35/3

WT + c-Myc 25/ 73/2 40 β

% S-phase cells S-phase % -gal + RG2 84/9/6

RG2 + c-Myc 15/81/4 20 β-gal + dl2-36 93/4/3

dl2-36 + c-Myc 27/67/6 0

β-gal β-gal β-gal β-gal + Myc + Myc -gal + MycWT + Myc β-gal + WT + RG2 + β RG2 dl2-36 + dl2-36 Figure 4 Reversal of the S-phase induction defect in RG2- and dl2–36-infected cells by c-myc overexpression at 24 h post infection. Serum-starved cells were coinfected with Ad WT or the E1A mutants with Adc-myc that overexpress c-myc (Mitchell and El-Deiry, 1999). Experimental conditions were essentially as described earlier in the legend to Figure 2. Adb-gal was used where appropriate to keep the multiplicity of infection constant. The experiment was carried out in triplicate and the average values with s.d.’s are shown. p400, TRRAP and GCN5 showed only a moderate human MCF10 cells, DNA synthesis was severely decrease. Thus the Myc RNA levels detected in real-time affected when either of the sites is inactive (Figure 3a PCR assay correlated with the Myc activity levels. and b). This could be owing to more stringently To determine the capacity of the E1A mutants to controlled growth regulation of human cells compared induce S phase, serum-starved MCF10A cells were to that of BRK cells. In summary, our results show that infected with different E1A mutants, and beginning the inability of the mutants RG2, 928, E55 and dl2–36 with 20 h post-infection, cells in G1, S and G2/M to induce S phase in quiescent cells correlates with their fractions were quantified by flow cytometry. The impaired capacity to induce c-myc. number of cells in S phase at 20, 24 and 28 h time points To determine whether the cell cycle defect associated is shown in Figure 2a and representative FACS profiles with RG2 and dl2–36 mutants is the result of reduced of these samples at 28 h are shown in Figure 3b. As c-myc synthesis, we performed a Myc reversal experi- evident, the WT- and dl26–35-infected cells showed a ment in which we infected cells with WT or the RG2 robust S-phase induction (about 38 and 34% at 28 h mutant along with an Ad vector that overexpresses respectively, Figure 3a). In contrast, the S-phase c-myc (detailed in Kolli et al., 2001; also see the legend induction by all other mutants was greatly reduced to Figure 4). Cells were harvested at 24 h after infection, (5–10%). These results are in broad agreement with and the distribution of cells in different cell cycle RNA and Myc activity levels (Figure 2a and c). Under fractions was determined by flow cytometry. Data these conditions, less than 10% of the virus-infected cells presented in Figure 4 indicate that overexpression of c- were in the sub G1 fraction (cell debris and apoptosing myc in RG2- and dl2–36-infected cells abrogated the cell cells, Figure 3b), indicating that there was no significant cycle defect associated with these mutations. For apoptosis. As RG2 mutant was severely defective in example, at 24 h after infection without c-myc over- inducing S phase in human cells, the additive effects, expression, about 35% of the WT-infected cells were in if any, by the second mutation that prevents binding to S phase, whereas only about 5–10% of the RG2- and pRb (RG2.928) or PCAF (dl2–36) was not evident in dl2–36-infected cells moved to S phase. In contrast, these studies. We note here that the cell cycle defect when c-myc was overexpressed in these cells, the number displayed by the E1A mutants presented here is more of cells in the S-phase fraction was comparable between severe than that reported earlier using primary BRK WT- and the mutant-infected cells. Note that all the cells (Wang et al., 1993; Nees et al., 2000). For example, c-myc-overexpressing samples moved into S phase faster the RG2 or 928 mutations individually were moderately than the corresponding control samples (table in defective for DNA synthesis in BRK cells, whereas, the Figure 4) presumably because of mitogenic effects of two mutations when combined together in one E1A the Myc protein. molecule (RG2.928 mutant) led to total inactivation of In summary, our studies show that the capacity of DNA synthesis (Wang et al., 1993). In contrast, in E1A to induce c-myc in quiescent cells is significantly

Oncogene c-myc activation by adenovirus E1A S Baluchamy et al 786 impaired when it fails to bind to p300, Rb and PCAF, tion in vitro (Santoso and Kadonaga, 2006). There are and this in turn affects its capacity to induce S phase. also several published reports that show that these two These interactions are not redundant because inactiva- proteins can independently regulate gene expression (see tion of any one of the above mentioned host protein for example, Puri et al., 1997). Currently, we do not binding sites in E1A results in significant loss of both c- know the mechanism by which p300 or PCAF represses myc and S-phase induction, indicating that c-myc c-myc. It is conceivable that pRb, p300 and PCAF may repression by each of the chromatin-associated proteins be present in the cell as a single or in separate complexes is independent. The pocket family proteins in associa- that are also associated with histone deacetylases. In tion with histone deacetylases block the E2F1 activity. quiescent cells, one or more Myc promoter-specific The promoter proximal region of the c-myc promoter transcription factors may recruit these complexes and contains an E2F site, and E2F4/5 in association with repress the promoter. Finally, the observation that dl26– p107has been shown to repress this promoter in 35 mutation has no significant effect on c-myc induction response to TGF-beta treatment (Chen et al., 2002). or DNA synthesis is interesting. This mutation abro- The mutation 928 selectively inactivates Rb binding, gates E1A binding to p400, TRRAP and GCN5, and whereas its p107binding is unaffected and binds to p130 studies have shown that these proteins exist as a at significant levels (Wang et al., 1993). Rb appears to complex that modifies gene expression in the cell actively participate in c-myc repression, which is (Taubert et al., 2004). These data combined with 928, consistent with our published data in which we showed RG2 and dl2–36 results suggest that specific chromatin that overexpression of Rb in quiescent MCF10A cells complexes may be involved in c-myc repression. represses c-myc (Buchmann et al., 1998). A role for PCAF (first isolated as p300/CBP-associated factor) in Myc repression is intriguing and appears to be independent of that of p300, as E1A mutants with Acknowledgements mutations that abrogate p300 but not PCAF binding This work was supported by the PHS grant CA74403. We are (RG2 and E55) are significantly impaired in their grateful to S Bayely (U McMaster, Canada) for providing capacity to induce c-myc. Further, overexpression of a certain adenovirus mutants, and W El-Deiry (U Pennsylvania) mutant p300 that cannot bind PCAF represses c-myc as for providing Adc-myc. We sincerely apologize that because of efficiently as the WT p300 (our unpublished results). It space limitations, we could not include all the relevant has been shown that purified p300 can repress transcrip- references in this report.

References

Baluchamy S, Rajabi HN, Thimmapaya R, Navaraj A, Harbour JW. (2001). Molecular basis of low-penetrance Thimmapaya B. (2003). Repression of c-myc and inhibition retinoblastoma. Arch Ophthalmol 119: 1699–1704. of G1 exit in cells conditionally overexpressing p300 that is Kolli S, Buchmann AM, Williams J, Thimmapaya B. (2001). not dependent on its histone acetyltransferase activity. Proc Antisense-mediated depletion of p300 in human cells leads Natl Acad Sci USA 100: 9524–9529. to premature G1 exit and up-regulation of c-myc. Proc Natl Barbeau D, Charbonneau R, Whalen SG, Bayley ST, Branton Acad Sci USA 98: 4646–4651. PE. (1994). Functional interactions within adenovirus E1A Lang SE, Hearing P. (2003). The adenovirus E1A oncoprotein protein complexes. Oncogene 9: 359–373. recruits the cellular TRRAP/GCN5 histone acetyltrans- Buchmann AM, Swaminathan S, Thimmapaya B. (1998). ferase complex. Oncogene 22: 2836–2841. Regulation of cellular in a chromosomal context by Mitchell KO, El-Deiry WS. (1999). Overexpression of c-myc the retinoblastoma tumor suppressor protein. Mol Cell Biol inhibits p21WAF1/CIP1 expression and induces S-phase 18: 4565–4576. entry in 12-O-tetradecanoylphorbol-13-acetate (TPA)-sensi- Chen C, Kang Y, Siegel PM, Massague J. (2002). E2F4/5 and tive human cancer cells. Cell Growth Differ 10: 223–230. p107as Smad cofactors linking the TGFbeta receptor to Moran E. (1993). Interaction of adenoviral proteins with pRB c-myc repression. Cell 110: 19–32. and . FASEB J 7: 880–885. Doyon Y, Selleck W, Lane WS, Tan S, Cote J. (2004). Nees M, Geoghegan JM, Munson P, Prabhu V, Liu Y, Structural and functional conservation of the NuA4 histone Androphy E et al. (2000). Human papillomavirus type 16 E6 acetyl transferase complex from yeast to humans. Mol Cell and E7proteins inhibit differentiation-dependent expression Biol 24: 1884–1896. of transforming growth factor-beta2 in cervical keratino- Egan C, Jelsma TN, Howe JA, Bayley ST, Ferguson B, cytes. Cancer Res 60: 4289–4298. Branton PE. (1988). Mapping of cellular protein-binding Puri PL, Sartorelli V, Yang X, Hammamor Y, Ogryzko V, sites on the products of early-region 1A of human Howard B et al. (1997). Differential roles of p300 and PCAF adenovirus type 5. J Virol 8: 3955–3959. acetyltransferases in muscle differentiation. Mol Cell 1: 35–45. Frisch SM, Mymryk JS. (2002). Adenovirus-5 E1A: paradox Rajabi HN, Baluchamy S, Kolli S, Nag A, Srinivas R, and paradigm. Nat Rev Mol Cell Biol 6: 441–4452. Raychaudhari P et al. (2005). Effects of depletion of CREB- Fuchs M, Gerber J, Drapkin R, Sif S, Ikura T, Ogryzko V binding protein on c-myc regulation and cell cycle G1–S et al. (2001). The p400 target is an essential E1A transition. J Biol Chem 280: 361–374. transformation target. Cell 106: 297–307. Reid JL, Bannister AJ, Zegerman P, Martinez-Balbas MA, Goodman RH, Smolik S. (2000). CBP/p300 in cell growth, Kouzarides T. (1998). E1A directly binds and regulates the transformation, and development. Genes Dev 14: 1553–1577. P/CAF acetyltransferase. EMBO J 17: 4469–4477.

Oncogene c-myc activation by adenovirus E1A S Baluchamy et al 787 Santoso B, Kadonaga JT. (2006). Reconstitution of chromatin Taubert S, Gorrini C, Frank SR, Parisi T, Fuchs M, Chan HM transcription with purified components reveals a chromatin- et al. (2004). E2F-dependent histone acetylation and specific repressive activity of p300. Nat Struct Mol Biol 2: recruitment of the Tip60 acetyltransferase complex to 131–139. chromatin in late G(1). Mol Cell Biol 24: 4546–4556. Soule HD, Maloney TN, Wolman SR, Peterson WDJ, Brenz Wang H, Moran E, Yaciuk P. (1995). E1A promotes R, McGrath C et al. (1990). Isolation and characterization association of p300 and pRB in multimeric complexes of a spontaneously immortalized human breast epithelial required for normal biological activity. JVirol69: 7917–7924. cell line MCF-10. Cancer Res 50: 6075–6086. Wang HG, Rikitake Y, Carter MC, Yaciuk P, Abraham SE, Sterner DE, Berger SL. (2000). Acetylation of and Zerler B et al. (1993). Identification of specific adenovirus transcription-related factors. Microbiol Mol Biol Rev 64: E1A N-terminal residues critical to the binding of cellular 435–459. proteins and to the control of cell growth. JVirol67: 476–488.

Oncogene