Repression of c-Myc and inhibition of G₁ exit in cells conditionally overexpressing p300 that is not dependent on its histone acetyltransferase activity

Abstract

p300 and cAMP response element binding protein (CREB)-binding protein (CBP) are two highly homologous, conserved transcriptional coactivators, and histone acetyltransferases (HATs) that link chromatin remodeling with transcription. Cell transformation by viral oncogene products such as adenovirus E1A and SV40 large T antigen depends on their ability to inactivate p300 and CBP. To investigate the role of p300 in cell-cycle progression, we constructed stable rat cell lines, which conditionally overexpress p300 from a tetracycline-responsive promoter. When p300 was induced in these cells, serum-stimulated S-phase entry was significantly inhibited. The inhibition of S-phase induction was associated with down-regulation of c-Myc, but not of c-Fos or c-Jun. Simultaneous overexpression of c-Myc and p300 before serum stimulation reversed the inhibition of S-phase induction to a significant level, indicating that the inhibition of c-Myc to a large extent is responsible for the p300 inhibition of G₁ exit. Similar studies with stable rat cell lines that overexpress a mutant p300, which lacks the HAT activity, showed that the intrinsic HAT activity of p300 is not required for the negative regulation of c-Myc or G₁. These findings, and our previously published results (Kolli, S., Buchmann, A. M., Williams, J., Weitzman, S. & Thimmapaya, B. (2001) Proc. Natl. Acad. Sci. USA 98, 4646–4651), establish an important negative regulatory role for p300 in c-Myc expression that may be important in maintaining the cells in the G₀/G₁ phase of the cell cycle.

Transcriptional coactivators p300 and the cAMP response element binding protein (CREB)-binding protein (CBP) are two highly conserved large nuclear phosphoproteins that contain histone acetyl transferase (HAT) activity. These proteins coactivate a large number of DNA-bound sequence specific transcription factors and remodel chromatin by acetylating nucleosomal histones (reviewed in refs. 1 and 2). p300 and CBP have also been shown to acetylate certain transcription factors, thereby affecting their DNA binding and/or transcriptional activation activities (3–6). Targeted gene disruption studies have shown that p300 is essential for normal cell proliferation and development (7), and haploinsufficiency of CBP in humans gives rise to severe abnormalities characteristic of Rubinstein–Taybi syndrome (8). Similar abnormalities are observed in mice lacking one allele of CBP, but not p300, suggesting that these two proteins may have certain nonoverlapping functions (9). p300 and CBP are mutated in several cancers, suggesting a tumor suppressor function of these proteins (reviewed in refs. 1 and 10).

Growth factor-induced G₁-S transition in mammalian cells is negatively controlled by retinoblastoma tumor suppressor protein Rb that represses E2F-driven cell cycle-related genes by forming a complex with E2F and several cellular repressor complexes (reviewed in ref. 11). Earlier adenovirus E1A studies have indirectly suggested that p300 may also negatively regulate G₁ phase of the cell cycle. For example, E1A stimulates DNA synthesis in quiescent baby rat kidney cells (12–15), and also restimulates DNA replication in terminally differentiated cardiac myocytes through its p300/CBP-binding domain (16). Other studies have shown that p300 is involved in terminal differentiation in several cell types including muscles, neurons and enteroendocrine cells (reviewed in ref. 1). During terminal differentiation, p300 transactivates p21 in cooperation with Sp1, Sp3, or tissue-specific transcription factors, suggesting that p300 may play a role in keeping cells in G₀/G₁ (1, 17). To assess the role of p300 in cell-cycle control, we recently used an antisense approach, and showed that quiescent human cells depleted of p300 exit G₁ without serum. This premature G₁ exit was found to be the result of the up-regulation of c-Myc that occurs because of depletion of p300 (18). This result suggests that p300 negatively regulates c-Myc in quiescent cells, and provides an explanation, at least in part, for the effect of E1A on DNA synthesis in quiescent and terminally differentiated cells. However, some p300 functions may also be necessary for cell-cycle progression. For example, fibroblasts from p300 knockout mice are unable to replicate and appear to undergo a cell-cycle arrest (7). p300 acetylates both growth stimulatory transcription factors such as E2F (6), and growth inhibitory proteins such as p53 (3), and increases their DNA-binding activities. p300 is underphosphorylated in quiescent cells and hyperphosphorylated during S and G₂/M phases (19). Phosphorylation of CBP increases during G₁/S boundary with an associated increase in its HAT activity and entry of cells into S phase (20). Cyclin E/cyclin-dependent kinase (cdk2) binds to the C-terminal region of p300 in vivo, and this interaction results in the activation of NF-κB function, which is required for the coordinated cell-cycle progression (21). Therefore, the overall role of p300 in the cell cycle is unclear.

To directly assess the role of p300 in G₁-S transition, we have constructed several stable Rat-1 cell lines capable of expressing p300 from the tetracycline (tet)-responsive promoter. We demonstrate that when p300 is induced in these cells, the serumstimulated induction of c-Myc and S phase are inhibited. We provide evidence that the repression of c-Myc by p300 contributes significantly to the inhibition of G₁-S transition. Furthermore, by constructing stable cell lines that conditionally overexpress a HAT-defective p300, we also show that the HAT activity of p300 that has been implicated in a number of p300 related functions is not required for this negative regulation. These findings and our published results (18) establish an important role for p300 in maintaining cells in G₁ phase of the cell cycle by preventing inappropriate expression of c-Myc. Thus, p300 may have a universal role as a checkpoint protein, preventing the untimely onset of DNA synthesis in senescent or differentiated cells.

Materials and Methods

Viral Vectors. Adenovirus (Ad) vectors expressing the tet activator/repressor (Adtet) (22), and β-gal (Adβ-gal) from the cytomegalovirus promoter were constructed as described (18). AdM4 is an Ad vector that contains a promoter-luciferase reporter cassette in which four copies of a c-Myc-binding site (E box element) are cloned upstream of the herpes simplex virus (HSV) tk minimal promoter (23). In AdM4mut the E box sequences are mutated so that c-Myc binds poorly to these sequences (23). Adc-Myc is an Ad vector that overexpresses c-Myc (24).

Construction of Rat Cell lines Overexpressing p300. The Rat-12 cell line, originally derived from Rat-1 cells (25), are indistinguishable from the Rat-1 cells with respect to their growth properties (unpublished results). Hemagglutinin epitope-tagged p300 cDNA (26) was cloned downstream of the tet-responsive promoter in a plasmid that also contains the hygromycin gene driven by the HSV tk promoter. Rat-12 cells were transfected with this plasmid, and selected for hygromycin resistance by using DMEM containing 150 μg of hygromycin per ml and 10% FBS. The drug-resistant colonies developed after 3 weeks were cloned and amplified by using standard procedures. Rat cell lines expressing a mutant p300 (p300mutAT2; ref 27), in which the HAT activity is abolished were constructed under identical conditions. The mutant p300AT2 contained a 6-aa substitution mutation (H1415A, E1423A, Y1423A, L1428S, Y1430A, and H1434A) in the HAT domain.

RNA, Protein, and Cell-Cycle Analysis. To analyze RNA, cells were seeded at a density of 1 × 10⁶ cells per 100-mm dish and were serum starved for 32 h, infected at 25 pfu per cell with appropriate Ad vectors under continued serum-starvation condition. Sixteen h later, they were stimulated with DMEM containing 10% serum, and Poly(A)-containing RNA was isolated at indicated time points in the figure by using a Poly(A)-RNA isolation kit (Qiagen, Valencia, CA). The RNA was then analyzed on Northern blots by using probes specific for Rat c-Myc (a gift of L. Penn, University of Toronto, Toronto), c-Jun, and c-Fos (gifts of R. Tjian, University of California, Berkeley). The RNA bands in the autoradiograms were quantified by using a densitometer. To determine the induced expression of p300 in rat cell clones, cells were infected with Adβ-gal or Adtet and maintained in DMEM with hygromycin. Twelve hours after infection, they were labeled with 200 uCi (1 Ci = 37 GBq) of [³⁵S]methionine per ml (3,000 Ci/mmol; Amersham Pharmacia) for 10 h, lysed in radioimmunoprecipitation assay buffer (28), and equal quantities of protein were immunoprecipitated by using an anti-p300 mAb (sc-584; Santa Cruz Biotechnology), and analyzed on SDS/8% PAGE as described (13). For cell-cycle analysis, cells were serum starved, infected with appropriate Ad vectors, and serum stimulated as described above for RNA analysis. The distribution of cells in G₁, S, and G₂/M phases were quantified by fluorescence-activated cell sorter (FACS) analysis as described (18, 28).

Cdk Assays. Serum-stimulated cells were lysed at the time points shown in Fig. 3C, and 100 μg of protein from each cell lysate was immunoprecipitated with either an anti-cyclin E or cyclin A antibody (sc-751 and sc-481, respectively; Santa Cruz Biotechnology) and then assayed for kinase activity by using histones as substrates (18, 25).

Fig. 3.

(A) Northern blot showing the levels of c-Myc, c-Fos, and c-Jun RNAs in serum-stimulated clone 37 and 48 cells. Serum-starved clone 37 and 48 cells were infected with Ad vectors and stimulated with serum as shown in Fig. 2 A. Poly(A)-containing RNA was isolated at different time points shown and was analyzed by using Northern blot hybridizations. c-Jun and GAPDH in each case were analyzed after reprobing the filters used for c-Myc analysis. One microgram of poly(A)-containing RNA was used in each case. (B) Transcriptional activation activity of Myc in serum-stimulated R12G12 (control) and clone 37 and 48 cells. (Upper) Structure of the Ad vector (AdM4) containing the Myc-responsive promoter-reporter cassette is shown. The inverted terminal repeat (ITR) and 0 and 4 map units of the Ad genome are shown. Serum-starved cells were infected with Ad vectors and were superinfected 2-h later with Ad-M4 or Ad-M4mut at 10 pfu per cell. Sixteen hours later, the cells were serum-stimulated, and at the indicated times the luciferase activity in the lysates was assayed by using 5 μg of protein. The experiments were repeated twice with reproducible values, and the data of one such experiment are shown. For R12G12 cells, only 0- and 3-h time points are shown. Luc, luciferase activity. (C) Determination of the levels of cyclin E/Cdk2 and cyclin A/Cdk2 activities in clone 37 cells. Assay details are as described in Materials and Methods. Phosphorylated histones were resolved on SDS/12% PAGE, dried, and autoradiographed.

Endogenous Myc Transcriptional Activation Activity Assays. Cells were serum starved and infected with either Adtet or Adβ-gal as above, and 2 h later, were superinfected with AdM4 or AdM4mut. Serum starvation was continued for another 16 h and cells were then stimulated with serum. Cells were lysed at the indicated times, and the luciferase activity in the lysates was determined by taking equal quantities of protein and by using a luciferase assay kit (Promega).

HAT Assays. Proliferating cells from clones 37 and AT21 were infected with Adtet or Adβ-gal for 16 h and lysed, and p300 was immunoprecipitated exactly as described (29) by using a mixture of anti-p300 mAbs (catalog no. 05-267, Upstate Biotechnology). One mg of protein from each cell lysate were assayed for HAT activity by using histones H3 (3 μg) and H4 (5 μg; both from Upstate Biotechnology), 0.5 uCi of [³H]acetyl CoA (catalog no. TRK688, Amersham Pharmacia) as described (29). Mouse IgG was as used a negative control.

Results

Conditional Overexpression of p300 in Stable Rat Cell Lines. p300 expressed from the tet-responsive promoter in stable rat cell lines (created as described in Materials and Methods) was induced by infecting them with Adtet. As a control, we used one of the Rat-12 clones (R12G12) that were selected for only the empty vector (25). In the initial screening, nearly half of the drug-resistant clones (5 of 11) were found to contain the inducible p300 gene. Immunoprecipitation analysis shows that the levels of p300 in cells derived from clones 5, 23, 35, 37, and 48, but not in R12G12 control cells infected with Adtet, increased by ≈8- to 10-fold when compared with those infected with control vector Adβ-gal (Fig. 1A). Immunoprecipitation of the cell extracts using the HA epitope-specific antibody confirmed that the large increase in p300 levels in clone 37 cells is solely because of the newly introduced p300 gene (Fig. 1 A; data for other clones are not shown). The uninduced levels of p300 in these clones were similar, which were comparable to that of R12G12 cells (<2-fold difference), indicating that basal level expression of the tet-responsive promoter-driven p300 is minimal (see legend to Fig. 1 for comparison of p300 levels between different lanes). These rat cells expressed normal levels of the endogenous p300 as judged by immunoprecipitation analysis (data not shown). p300 that is overexpressed in these cells coactivated cAMP response element binding protein in a reporter assay, indicating that we did not select cells that overexpress p300 with altered properties (data not shown). Growth rates of clones 5, 35, 37, and 48 were in general comparable to that of R12G12 (Fig. 1B). Even though p300 levels in clone 23 were comparable to that of other clones, the growth rate of these cells was significantly slower. Reasons for their poor growth rate are not clear at present; it could be due to integration of the exogenous p300 sequences in a locus whose function may be critical for the normal growth of the cell.

Fig. 1.

Induction of p300 in rat cell clones and their growth properties. (A) Immunoprecipitation analysis of p300 (see Materials and Methods for details). Note that p300 levels for clones R12G12, 23, 48, 5, and 35 in lanes 1–10 and clones 37 and 48 cells in lanes 11–14 were assayed in separate experiments. For comparison of p300 between lanes, the levels of p300 in lanes 11–14 were normalized to clone 48 values in lanes 5–6. (B) Growth properties of rat cell clones.

Overexpression of p300 in Growth-Arrested Cells Inhibits the G₁/S Transition. The effects of overexpression of p300 on serum-stimulated S-phase induction were studied as described in Materials and Methods (shown schematically in Fig. 2A). High-level accumulation of p300 in clones 23 and 48 before serum stimulation was confirmed in a parallel time course experiment by inducing p300, followed by immunoprecipitation analysis as shown in Fig. 2B (a 7- to 8-fold increase by ≈16 h after Adtet infection in both cases; see Fig. 2 legend for details). Distribution of cells in G₁, S, and G₂/M fractions at various time points after serum stimulation were determined by FACS analysis. Each clone was analyzed in three independent experiments that yielded very similar results (except clone 5, which was analyzed only twice), and the representative data are shown in Fig. 2C. Control R12G12 cells emerged from quiescence at an identical rate when infected with either Adtet or Adβ-gal (Fig. 2C). Clones 35, 37, and 48 infected with control vector also exited G₁ at rates comparable to that of R12G12. For some unknown reasons, clone 5 cells emerged from G₁ slightly faster than from clone 37 or 48. Consistent with its poor proliferation rate, clone 23 infected with Adβ-gal exited G₁ significantly more slowly when compared with other clones (Fig. 2C). Induction of p300 in all p300-overexpressing clones, including clone 23, significantly inhibited the S-phase entry with the levels of inhibition ranging from 50 to 60% at the 18-h time point (for clone 5, 50% inhibition at 16 h after stimulation).

Fig. 2.

Inhibition of the S-phase entry in p300-overexpressing cells. (A) Schematic representation of the time course of serum starvation, vector infection, and harvesting of cells for FACS analysis. (B) Induction of p300 in serum-starved cells. Serum-starved cells were infected 16 h before serum stimulation with Ad vectors, labeled, and harvested at the indicated times, and p300 in the lysates was analyzed by immunoprecipitation. (C) Kinetics of S-phase entry of serum-stimulated cells. Cells were seeded, starved, and infected with Ad vectors as in A. The distribution of cells in G₁, S, and G₂/M phases after serum stimulation was determined by FACS analysis as described (18). The assays were done in triplicate, and the average number of cells in S phase with SD are shown.

c-Myc, but Not c-Fos or c-Jun, Is Down-Regulated in Cells Overexpressing p300. We showed (18) that when p300 is depleted in quiescent human cells, c-Myc is up-regulated, and such cells exited G₁ without serum. To examine whether p300-overexpressing cells contain reduced Myc expression, p300 was overexpressed in serum-starved cells, was serum stimulated as described for cell-cycle analysis, and then Poly(A)-containing RNA was isolated at 30-min intervals (for c-Fos, 15-min intervals) up to 3 h after serum stimulation, and analyzed on Northern blots by using rat cDNA probes. As shown in Fig. 3A, c-Myc RNA levels were barely detectable at the 0-h time point in clones 37 and 48 infected with Adβ-gal, appeared at significant levels at 1 h, peaked at ≈2 h, and declined thereafter. In cells overexpressing p300, c-Myc RNA levels were reduced by ≈4- to 5-fold, which was evident both at the 1- and 2-h time points (compare lanes Adβ-gal with Adtet in Fig. 3A for c-Myc). In contrast, c-Fos and c-Jun RNA levels were comparable in control and p300-overexpressing cells. Thus, we conclude that increased p300 levels in serum-starved cells led to reduced steady-state levels of c-Myc RNA, but not of c-Fos or c-Jun RNAs.

To determine whether the reduced c-Myc RNA levels correlate with the endogenous c-Myc transcriptional activation activity, p300 was induced in the cells as shown in Fig. 2 A, infected with an Ad vector containing a promoter-reporter cassette in which four copies of the Myc-binding sites (23) were cloned upstream of a minimal promoter (AdM4; AdM4mut contained mutations in the Myc-binding sites; see Materials and Methods). Sixteen h later, they were stimulated with serum, harvested at 30-min to 1-h intervals, and then assayed for luciferase activity. The promoter-reporter activities in R12G12 cells infected with Adβ-gal or Adtet rose steadily with comparable increase, and peaked at ≈3 h (Fig. 3B and data not shown). As expected, the levels of the mutant promoter-reporter activities in cells infected with the β-gal vector was ≈4- to 5-fold lower than that of the WT promoter-reporter. In clone 37 and 48 cells infected with control vector, the WT promoter-reporter activity peaked at ≈3 h after serum stimulation (Fig. 3B and data not shown). In contrast, when p300 was induced in these cells, the reporter activity was significantly reduced. These experiments were repeated twice with reproducible values and data shown are of one such experiment. These results correlate well with the RNA data (Fig. 3A), and indicate that p300-overexpressing cells synthesize reduced amounts of Myc protein.

Cells Overexpressing p300 Show Decreased Levels of Cdk Activities. Several S-phase-specific genes, including cyclin E and Cdk2, are direct targets of Myc (30). To determine whether the repression of Myc and S phase by p300 is associated with a decrease in the levels of Cdk activities, the levels of cyclins E- and A/Cdk2 activities in clone 37 cells were determined (see Materials and Methods for details). As shown in Fig. 3C, in control cells that transit from G₁ to S normally, induction of cyclin E/Cdk2 and cyclin A/Cdk2 activities was detected ≈9 h and 12 h after stimulation, respectively (data for time points earlier than 9 h not shown), consistent with the normal coordinated S-phase progression. When p300 was induced in these cells, reduction of cyclin E/Cdk2 activity was observed as early as 9 h after serum stimulation when compared with that of control cells. Decrease in cyclin A/Cdk2 activity became evident at the 12-h time point, which is ≈3 h later than that observed for cyclin E/Cdk2 activity. These results are consistent with the reduced Cdk2 activities, contributing to the progression of S phase in p300 overexpressing cells. It is likely that the reduction of Myc protein in p300-overexpressing cells contributed to the reduction of Cdk2 activity, which could be rate limiting for the coordinated S-phase progression.

Overexpression of c-Myc Reverses the p300 Inhibition of S-Phase Induction. To determine whether overexpression of Myc would reverse p300-repression effects, p300 and Myc were overexpressed in serum-starved clone 48 and 37 cells simultaneously by coinfecting cells with Adtet and an Ad vector overexpressing c-Myc (ref. 24 and Fig. 4A). We note here that these experiments work well only when cells from early passages are used. The serum-stimulated S-phase entry of cells was then determined by FACS analysis. The average number of cells obtained from three independent experiments with SD at 18 h after serum stimulation (a time point at which cells continue to accumulate in S phase, but do not yet move to G₂/M phase) are shown in Fig. 4 B and C. These data suggest that overexpression of Myc significantly abrogates the inhibition of S-phase induction caused by excess p300 in both cell clones. For example, in clone 48, when only Myc was overexpressed, 62 ± 11% (mean and SD, respectively) of the cells accumulated in S phase (Fig. 4B; lane 6). In contrast, only 19 ± 6% of the p300-overexpressing cells accumulated in S phase at this time point (lane 7). However, when Myc and p300 were overexpressed simultaneously, the number of S-phase-specific cells increased to 64 ± 9%, suggesting that overexpression of Myc in this clone completely abrogated the p300-inhibitory effects (lane 8). Overexpression of Myc in clone 37 also showed similar results, although the reversal effect was slightly less efficient. For example, 57 ± 9% and 19 ± 6% of the Myc-, and p300-overexpressing cells, respectively, accumulated at 18 h after serum stimulation (Fig. 4C, lanes 6 and 7). When Myc and p300 together were overexpressed, the number of cells accumulated in S phase increased to 52 ± 11% (lane 8). The differences between the mean cell numbers in lanes 7 and 8 for both clones are significant, as validated by statistical analysis (see legend to Fig. 4B). The Myc reversal effects were also evident at the 20-h time point (data not shown). Data presented in Fig. 4 B and C also show that the Myc-overexpressing cells move faster than the control cells, presumably because of the mitogenic effects of Myc.

Fig. 4.

Abrogation of p300-mediated S-phase inhibition by Myc. (A) Schematic representation of the time course of vector infection and harvesting of cells for FACS analysis. (B) Serum-starved cells were coinfected with Ad vectors as shown (all at 25 pfu per cell). Sixteen hours later, they were stimulated with serum and were harvested at indicated time points. Distribution of cells in G₁, S, and G₂/M phase were determined by FACS analysis as described (18). Adβ-gal was used where appropriate to keep the multiplicity of infection constant. Average values obtained in three independent experiments with SD are shown. For statistical validation, the mean cell numbers in lane 7 (cells infected with Adtet plus Adβ-gal) were compared with those of lane 8 (cells infected with Adtet plus Adc-Myc) by independent sample Student's two-tailed t test. P values for lane 7 vs. lane 8 for clone 48, 0.007; for clone 37, 0.011. (C) Determination of cyclin E/Cdk2 and cyclin A/Cdk2 activities in clone 37 cells at 15 h after infection with Ad vectors as shown in B.

CyclinA/Cdk2 and CyclinE/Cdk2 activities in clone 37 cells shown in Fig. 4C (lanes 5–8) were examined at 15 h after serum stimulation to determine whether overexpression of Myc reverses the inhibition of the Cdk2 activities. Data presented in Fig. 4D clearly show that the cyclin E/Cdk2 and cyclin A/Cdk2 activities in cells overexpressing p300 plus Myc increased considerably when compared with cells overexpressing only p300, and they were comparable to that observed in cells expressing only Myc. These data support the FACS data discussed above, and suggest that overexpression of Myc was able to overcome the p300 repression of the S-phase entry.

Inhibition of Induction of c-Myc and S phase by p300 Is Independent of Its HAT Activity. The HAT activity of both p300 and CBP regulates gene expression by acetylating the transcription factors and the chromatin (1). To determine whether the HAT activity of p300 contributes to the inhibition of G₁ exit and Myc induction, we constructed rat cell lines as described above that conditionally overexpress a mutant p300 that lacks the HAT activity (ref. 27; see Materials and Methods). Two such clones were analyzed (AT21, AT41) in which p300 could be induced to levels comparable to that of clone 37 cells that expresses the WT p300 (Fig. 5A; ≈6- to 7-fold in this experiment), and these cells grew at rates similar to that of clone 37 (Fig. 5B). Data presented in Fig. 5C suggest that the inhibition of S-phase induction in cells overexpressing mutant p300 are comparable to that observed in cells overexpressing the WT p300 (clone 37; range of inhibition 50–70%). Thus, overexpression of the HAT-defective mutant p300 continues to inhibit the S-phase induction in two independently isolated clones to the same extent as the WT p300. Another clone overexpressing the HAT-defective p300 (AT10) also showed similar S-phase inhibition, although this clone grew more slowly compared with other clones (data not shown). The serum-stimulated Myc activity was also reduced in clone AT21 and AT41 cells to the same extent as that of clone 37 (≈60–65% reduction; Fig. 5D). To confirm that AT21 and AT41 cells overexpress the HAT-defective p300, p300 in proliferating cells was induced, immunoprecipitated, and assayed for HAT activity in vitro as described in Materials and Methods. Fig. 5E shows a 3.5-fold increase in radiolabeling of histones in p300-induced cells, as compared with uninduced cells for clone 37, but not for clone AT21 cells (induction of p300 in these clones was comparable as determined by immunoprecipitations, data not shown). Thus, we conclude that the HAT activity of p300 does not contribute to the inhibition of the serum-stimulated induction of Myc or S phase.

Fig. 5.

HAT activity of p300 is not required for the inhibition of Myc- and S-phase induction. (A) Immunoprecipitation of p300 in cell clones AT21, AT41, and 37. (B) Comparison of growth properties of cell clones AT21 and AT41 with R12G12 and 37. (C) Induction of S phase in clones AT21 and AT41 with and without induction of p300. Details of this experiment were as in Fig. 2C. Assays for AT21 and AT41 were done in triplicate, for which the average values and SD are shown. (D) Determination of Myc activity in serum-stimulated clones 37, AT21, and AT41. Details are as in Fig. 3B. Values shown are luciferase activity expressed as percent of Adβ-gal controls. Averages values with SD obtained from three independent experiments are shown. (E) HAT activity of p300 in clone 37 and AT21 with induced and uninduced p300 levels. See Materials and Methods for details. Average values obtained from two experiments are shown. The radioactivity obtained from IgG incubations was subtracted from each matching sample. The values obtained for Adβ-gal infected cells were taken as 1 in calculating the fold induction.

Discussion

In this article, we showed that the down-regulation of c-Myc by p300 significantly contributes to the inhibition of G₁-S transition, because overexpression of Myc along with p300 abrogates the p300 inhibitory effects in both clones 37 and 48 cells. These results are consistent with our recently published data (18), in which we showed that induction of Myc and S phase in cells because of depletion of p300 can be reversed by overexpressing antisense c-Myc in the cells. These results indicate that critical levels of p300 in G₁ phase of the cell cycle negatively regulate c-Myc and maintain cells in G₁. Our results also provide a molecular basis for the need for viral oncoproteins E1A and SV40 large T to inactivate p300 during cell transformation, and the p300-dependent effect of E1A on DNA synthesis in quiescent (15), and in terminally differentiated cells (reviewed in ref. 1). It is conceivable that p300 may contribute to cell differentiation by keeping c-Myc in a repressed state.

The HAT activity of p300 does not contribute to the negative regulation of G₁ or Myc expression, because overexpression of a HAT-defective mutant continues to inhibit Myc and G₁ to the same extent as the WT p300. p300 has been shown to down-regulate gene expression by acetylating transcription factors. For example, acetylation of HMG(I)Y by p300 shuts off IFN-β gene expression during the postinduction period (4). Drosophila dCBP acetylates dTCF [homolog of human T cell factor (TCF)], and down-regulates its activity as acetylation interferes in its interaction with armadillo (human homolog of β-catenin) that provides the transcription activation function (5); human c-Myc promoter is also activated by TCF4 (31). Thus, we considered it possible that p300 could acetylate and down-regulate the DNA binding and/or the transcriptional activation activity of the Myc promoter-specific transcription factors that could contribute to Myc repression. It is unlikely that repression of Myc and S-phase entry by the HAT-defective mutant is due to other p300-associated cellular HAT proteins such as PCAF, which are present at normal levels in these cells, because the induced p300 in AT21 did not show an increase in HAT activity (Fig. 5E). However, we cannot completely rule out the possibility that other cellular acetylases associating with high levels of p300 may contribute to the effects reported here. Nonetheless, there are several examples in which specific requirement of the HAT activities of p300/CBP, but not that of PCAF (and vice versa), has been demonstrated (32–35). The HAT proteins such as PCAF and p300/CBP also exhibit differences in substrate specificities with respect to histones (36).

Our data showing only c-Myc, but not c-Fos or c-Jun transcription, is negatively regulated by p300 in G₀/G₁ (this article and ref. 18), suggests that the certain Myc promoter-specific transcription factors recruit p300 in G₀/G₁ and repress gene transcription. Studies have shown that in transient assays, c-Myc promoter is repressed by transcription factors including CTCF (37), E2F4 (38), Mxi (39), MBP-1 (40), MAZ (41), CDP/cut (42), and Blimp-1 (43, 44). E1A also can activate Myc promoter through the PRF element in a p300-dependent manner, suggesting that p300 may cooperate with factors binding to the PRF site and repress Myc promoter (45). Recent studies (46) suggest that E1A relieves YY1 repression by simultaneously interacting with p300 and YY1, raising the possibility that in serum-starved cells, p300 may cooperate with YY1 in keeping the Myc promoter in a repressed state. The majority of the transcriptional repressors mentioned above have been shown to interact with histone deacetylases (HDACs). p300 interacts with at least some of these transcription factors, although a functional link between p300 and HDAC by means of a transcription factor in repressing gene transcription not been established. One possibility is that in quiescent cells, p300 is recruited to the promoter by these transcriptional repressors that may then form an adapter to bridge the transcription factors and the chromatin remodeling complexes that function as repressors. Several such chromatin remodeling complexes that function as repressors have been described, including hBRM/BRG-1 (human homologs of yeast SWI/SNF), HDAC-containing Sin 3 complexes mSin3a and mSin3b, and the histone methylase complex SUV39H1/HP1 (reviewed in refs. 47 and 48). Repression of the cyclin E and A, and the c-Fos promoters are mediated by these repressors (49–51).

Finally, we are intrigued by our observations that we were unable to develop stable human cell lines conditionally overexpressing p300. For reasons that are not clear at present, Rat-1 cells could tolerate a newly introduced p300 gene, and could readily grow as stable cell lines. At present we do not know whether this difference is related to species difference or because of the embryonic origin of these cells.