Tumor suppressor p53 represses transcription of RECQ4 helicase

Article metrics

Abstract

RECQ4 is a member of the RecQ helicase family, which has been implicated in the regulation of DNA replication, recombination and repair. p53 modulates the functions of RecQ helicases including BLM and WRN. In this study, we demonstrate that p53 can regulate the transcription of RECQ4. Using nontransformed, immortalized normal human fibroblasts, we show that p53-dependent downregulation of RECQ4 expression occurred in G1-arrested cells, both in the absence or presence of exogenous DNA damage. Wild-type p53 (but not the tumor-derived mutant forms) repressed RECQ4 promoter activity. The camptothecin or etoposide-dependent p53-mediated repression was attenuated by trichostatin A (TSA), an inhibitor of histone deacetylases (HDACs). Repression of the RECQ4 promoter was accompanied with an increased accumulation of HDAC1, and the loss of SP1 and p53 binding to the promoter. The simultaneous formation of a camptothecin-dependent p53-SP1 complex indicated its occurrence outside of the RECQ4 promoter. These data suggest that p53-mediated repression of RECQ4 transcription during DNA damage results from the modulation of the promoter occupancy of transcription activators and repressors.

Introduction

Rothmund–Thompson syndrome (RTS) is an autosomal recessive disorder. Patients with RTS are characterized by short stature, congenital skeletal abnormalities, skin atrophy and talangiectsia with hyper- and hypopigmentation and manifestations of premature aging (Vennos and James, 1995). The gene responsible for RTS, RECQ4, encodes a 1208-amino-acid protein (Kitao et al., 1998). RECQ4 is a member of the highly conserved RecQ helicase family, thought to be involved in the maintenance of genomic integrity. Members of this family have been identified from Escherichia coli to human. In human, apart from RECQ4, four other genes encoding members of the RecQ helicases have been identified – WRN, BLM, RECQL and RECQ5β (Hickson, 2003; Shimamoto et al., 2004). Germline mutations in BLM, WRN and RECQ4 result in autosomal-recessive disorders, Bloom syndrome (BS), Werner Syndrome (WS) and RTS, respectively. Severe growth retardation and skin abnormalities have been reported in RECQ4-deficient mice, suggesting the defects in the RECQ4 gene may indeed be responsible for RTS (Hoki et al., 2003). RECQ4 has been shown to be part of a stable complex in the cytoplasm with the ubiquitin ligases, UBR1 and UBR2 (Yin et al., 2004). The pathological association between a subset of RTS patients and osteosarcoma has long been observed (Drouin et al., 1993; el-Khoury et al., 1997). The molecular basis of this is probably due to the mutations in the RECQ4 gene. Mutations predicted to result in the loss of RECQ4 function occur in approximately two-thirds of RTS patients who are also diagnosed with osteosarcoma (Wang et al., 2003).

The study of genomic organization indicates that the RECQ4 gene spans over 21 exons. The 5′ upstream region of RECQ4 contains several elements for constitutive transcription factors (SP1, AP1, AP2, CRE and PAE3), consistent with the possibility that RECQ4 is an essential housekeeping gene (Kitao et al., 1999). The RECQ4 protein sequence encodes a helicase domain (Kitao et al., 1998). However, the in vivo function(s) of the RECQ4 protein is not known. It has been proposed that RECQ4 also functions like BLM and WRN as a ‘genome caretaker’ (Hickson, 2003). Loss of functions in these ‘caretaker tumor suppressors’ due to germline mutations can lead to a corresponding decrease in their functional interactions with ‘gatekeeper tumor suppressors’ like p53 and Rb, potentially leading to accelerated cancer development.

The major downstream functions of the p53 tumor suppressor are the induction of growth arrest, apoptosis or senescence, with the ultimate end point depending on the cellular context, the type and duration of the extracellular stress (Vogelstein et al., 2000; Haupt et al., 2002; Oren et al., 2002; Hofseth et al., 2004; Sengupta and Harris, 2004). While in G1/G2 phases of the cell cycle, p53 primarily functions as a sequence-specific transcription activator (Ko and Prives, 1996), it can also mediate the different downstream functions by repressing many target genes (Zhao et al., 2000). Although the transactivator function of p53 is much better understood, considerable efforts have also been made in recent years to understand the trans-repression function of p53 (Ho and Benchimol, 2003).

Both BLM and WRN have been proposed to function at the interface of replication and recombination. Both BLM and WRN facilitate the repair of DNA damage and have functional inter-regulation with p53 (Hickson, 2003). BLM is a responder to DNA damage and rapidly colocalizes to presumed stalled replication forks, as visualized by its colocalization with phosphorylated H2AX (γ-H2AX), p53-binding protein 1 (53BP1) and DNA damage sensor kinases, Chk1 and ATR (Sengupta et al., 2004). BLM binds to p53 in response to ionizing radiation and replication arrest, and transports it to stalled DNA replication forks (Wang et al., 2001; Sengupta et al., 2003). Inactivation of p53 in BS cells causes a statistically significant increase of sister chromatid exchange (SCE) when compared with BS cells alone, thereby demonstrating that p53 and BLM cooperatively affect homologous recombination (HR) (Sengupta et al., 2003). During the hydroxyurea (HU)-mediated replication stress, BS cells undergo p53-dependent apoptosis (Davalos and Campisi, 2003). These studies indicate a dynamic relationship between p53 and RecQ helicases.

In this study, we have demonstrated that the functional interaction between p53 and RecQ helicases also extends to RECQ4. We demonstrate that the repression of RECQ4 during G1 arrest depended on p53. Wild-type p53, but not the tumor-derived mutants, repressed RECQ4 promoter activity. Finally, we provide evidence that p53 repressed the RECQ4 promoter by forming a complex with the transcription factor, SP1, and by regulating promoter occupancy of both itself and SP1 in concert with histone deacetylase 1 (HDAC1).

Results

Loss of RECQ4 expression during G1 arrest depends on p53

It has been previously reported that the expression of RECQ4 was upregulated in different types of transformed human cells (Kawabe et al., 2000). The level of RECQ4 was higher in proliferating fibroblasts than in cells arrested in stationary phase and decreased dramatically in senescent fibroblasts (Kawabe et al., 2000; Furuichi, 2001). To determine the cell cycle expression of RECQ4 and whether the expression depends on p53 status, we used nontransformed hTERT-immortalized normal human fibroblasts (NHF) or fibroblasts expressing human papillomavirus 16 E6 protein (NHF E6), which degrade p53. Cells were arrested in G1 phase due to contact inhibition, replated at low dilutions and grown for various time intervals. At each time point, cells collected were either analysed by flow cytometry for the cell cycle profile (Figure 1A), or lysed and probed for the expression of different proteins (Figures 1B and C). In NHF, both BLM and RECQ4 expression were not detectable in G1-arrested cells (Figure 1B). The expression of both the proteins was initiated in the early or mid S phase and further elevated in the late S and G2 phases of the cell cycle. Transcriptionally active p53 was expressed to a high level in G1-arrested cells, as demonstrated by the expression of p53 target gene, p21. As reported earlier (Gottifredi et al., 2001), p53 that was expressed in S phase was transcriptionally inactive.

Figure 1
figure1

Expression of RECQ4 in NHF and NHF E6. (A) Synchronization of cells. NHF or NHF E6 was synchronized in G1 by contact inhibition. Postsynchronization, the cells were plated at low density and released for different time periods. Cell cycle profiles were analysed by flow cytometry in FACS Calibur, the data analysed and quantitated by ModFit v3.0. (B) Expression of proteins in NHF. Same as (A), except that cell extracts were made from NHF. Western blots (50 μg of the lysate) are carried out with antibodies against (a) RECQ4; (b) BLM; (c) p53 (DO-1); (d) p21 and (e) TBP. Lane 1: G1-arrested NHF, lane 2: cells in early S (12 h postrelease), lane 3: cells in mid S (18 h postrelease), lane 4: cells in late S/G2 (24 h postrelease) and lane 5: cells in G1/G2 phases (32 h postrelease). (C) Expression of proteins in NHF E6. Same as (A), except that cell extracts were made from NHF E6. Western blots (50 μg of the lysate) are carried out with antibodies against (a) RECQ4; (b) BLM; (c) p21 and (d) TBP

To determine whether the expression of RECQ4 depended on p53, similar experiments were carried out in NHF E6 cells. Flow cytometry experiments indicated similar profiles, although the G1 arrest in NHF E6 was less stringent (Figure 1A), possibly due to the lack of p53 expression and a consequent low level of p21 protein (Figure 1C). The expression profile of RECQ4 in NHF E6 was different when compared with NHF. A low level of RECQ4 was detected in G1-arrested NHF E6, the level increased further in the S and G2 phases of the cell cycle. Interestingly, the level of BLM was similarly expressed to a higher degree in early S phase in NHF E6 cells. Together, these results indicate that the deregulation of RecQ helicases (especially the loss of RECQ4 expression during G1 arrest) depended on p53.

Repression of RECQ4 expression during DNA damage depends on p53

The p53-dependent loss of RECQ4 expression during G1 phase of the cell cycle prompted us to examine whether such a phenomenon was also observed during exposure to predominantly G1 arrest, causing DNA damage in NHF. We treated asynchronous NHF with two drugs, etoposide (Etop) or camptothecin (Camp), for 3–6 h. This duration of treatment by the topoisomerase inhibitors caused an accumulation of cells mostly in G1/S phases of the cell cycle (Figure 2A). Treatment with either drugs caused an accumulation of transcriptionally active p53 in a time and dose-dependent manner (Figure 2C, Supplementary Figure 1A and data not shown). However, as p53 was stabilized and accumulated, there was a decrease in the transcript (Figure 2B, lanes 1–3) and protein (Figure 2C) levels of RECQ4. Similar p53 stabilization and decrease in RECQ4 protein expression was also observed in two primary human fibroblast lines, GM08402 and GM03348 (Supplementary Figure 1B).

Figure 2
figure2

DNA damage-induced p53 accumulation causes decreases in the RECQ4 transcript and protein in a p53-dependent manner. (A) Cell cycle profile of NHF after DNA damage. NHF was either left untreated or treated with camptothecin or etoposide for 6 h. The cell cycle profiles were analysed by flow cytometry in FACS Calibur, the data analysed and quantitated by ModFit v3.0. (B) p53-dependent decrease in the RECQ4 transcript due to DNA damage. NHF (lanes 1–3) or NHF E6 (lanes 4–6) were either left untreated (lanes 1,4), treated with camptothecin (lanes 2,5) or etoposide (lanes 3,6) for 6 h. Total RNA (20 μg) was extracted and probed with cDNA probes for (a) RECQ4 and (b) GAPDH. (C) Etopside and camptothecin caused a decrease in the level of RECQ4 protein in NHF. NHF was treated as in (A) except that cell extracts were prepared. Western blots (50 μg of the lysate) were carried out with antibodies against (a) RECQ4, (b) p53 (DO-1), (c) p21 and (d) TBP. (D) Etopside and camptothecin did not decrease the level of RECQ4 protein in NHF E6. NHF E6 were treated as in (A). Western blots (50 μg of the lysate) were carried out with antibodies against (a) RECQ4 and (b) TBP. (E) Camptothecin caused a decrease in the RECQ4 protein level in MRC5, but not in MRC5 siRNAp53 cells. MRC5 (lanes 1,2) or MRC5 siRNAp53 (lanes 3,4) cells were treated with camptothecin for 6 h. Extracts were prepared and Western blots (50 μg of the lysate) were carried out with antibodies against (a) RECQ4, (b) p53 (DO-1) and (c) TBP

To investigate whether the decreases in the RECQ4 transcript and protein levels were p53 dependent, we carried out parallel experiments in NHF E6. Like NHF, NHF E6 cells also accumulated in the G1/S-phases of the cell cycle after 6 h of drug treatment (data not shown). However, unlike NHF, NHF E6 cells did not show any decrease in the RECQ4 transcript (Figure 2B, lanes 4–6) or protein (Figure 2D). E6 protein has other pleiotropic effects apart from degrading p53 (Munger and Howley, 2002). Therefore, we repeated the above experiment in another pair of isogenic primary human fibroblasts. In MRC5 siRNAp53, p53 expression was silenced with a previously characterized siRNA (Brummelkamp et al., 2002). Like NHF E6, the MRC5 siRNAp53 cells showed no decrease in the RECQ4 protein level after camptothecin treatment (Figure 2E). These results confirmed that the decreases in the transcript and protein levels of RECQ4 were p53-dependent.

RECQ4 promoter is directly repressed by p53

To determine whether p53 can directly repress the RECQ4 promoter, the 594 basepair (bp) Nru1/Nco1 fragment upstream of the ATG was cloned into the pGL2 luciferase vector (Figure 3a). The resultant construct, named RECQ4-luc, was transfected in two p53-deficient cell lines, NHF E6 or HTC116 p53(−/−). In both cell lines, wild-type p53 activated the known p53 target gene, p21 (el-Deiry et al., 1993), in a dose-dependent manner. Under the same conditions, expression of wild-type p53 repressed the expression of the RECQ4 promoter (Figures 3b and d).

Figure 3
figure3

RECQ4 is a direct target of p53-mediated repression. (a) Schematic representation of the RECQ4-luc construct. The 594 bp upstream of the start site of the RECQ4-coding region was fused from the pGL2 luciferase vector to make the RECQ4-luc construct. (b) Wild-type p53 represses the RECQ4 promoter, while activating the p21 promoter. NHF or HCT116 p53 (−/−) cells were transfected with pGL2, p21-luc or RecQL4-luc in the absence or presence of increasing amounts of wild-type p53 (p53 WT). The amounts of p53 constructs used were 0.05, 0.1, and 0.2 μg. The graph represents the fold difference obtained from three independent experiments in duplicate. (c) Effect of different p53 mutants on RECQ4 promoter. Same as (b) except that HCT116 p53 (−/−) cells were transfected with increasing amounts of transactivation-deficient p53 mutant p53 (22,23), and two tumor-derived p53 mutants, p53 (R175H) and p53 (R248W). (d) Expression of p53 proteins. HCT116 p53 (−/−) cells are either left untransfected (lane 1) or transfected with p53 (WT) (lanes 2–4); p53 (22,23) (lanes 5–7); p53 (R175H) (lanes 8–10) and p53 (R248W) (lanes 11–13). Western blots (50 μg of the lysate) were carried out with antibodies against (a) p53 (Pab 1801) and (b) TBP

p53-dependent transcriptional repression has been dissociated from the transactivation function during a physiological stress like hypoxia (Koumenis et al., 2001). Hydrophobic amino-acid residues Leu-22 and Trp-23 of p53 have been shown to be critical for p53 transactivation function (Lin et al., 1994). Hence, an artificially generated transcriptionally inactive p53 mutant, p53 (22,23) (Lin et al., 1994), was used to study the trans-repression effect of p53 on the RECQ4 promoter. As expected, p53 (22,23) was not able to transactivate the p21 promoter. However, like wild-type p53, the transactivation-deficient p53 repressed the RECQ4 promoter (Figures 3c and d).

The most common p53 mutations are in the DNA-binding domain, which lead to the generation of mutant proteins with an altered amino-acid sequences (Hollstein et al., 1991; Levine et al., 1991). Hence, most of the tumor-derived mutant p53 proteins lack the sequence-specific transactivation function of p53 (Michalovitz et al., 1991; Zambetti and Levine, 1993; Ko and Prives, 1996). Apart from being deficient in p53-dependent transactivation, mutant p53 also shows a lower or no repressing capability on various p53-repressed promoters like c-fos, interkeukin-6 (IL-6), insulin-like growth factor (IGF-1) and WRN (Santhanam et al., 1991; Webster et al., 1996; Yamabe et al., 1998). Two tumor-derived p53 missense mutants (R175H and R248W) were overexpressed to similar levels as wild-type p53. While p53 (R175H) did not at all repress the RECQ4-luc activity, p53 (R248H) repressed the RECQ4-luc activity to a much lower extent compared with wild-type p53 (Figures 3c and d), indicating that the structural integrity of the DNA-binding domain of p53 was essential for p53-mediated repression of this promoter.

p53-mediated repression of RECQ4 promoter is inhibited by TSA

To elucidate the mechanism of repression of the RECQ4 promoter by p53, NHF was incubated either with or without camptothecin (or etoposide), and in the presence or absence of TSA for 6 h. TSA is a nonspecific inhibitor of all the histone deacetylases (HDACs). The addition of TSA to camptothecin (or etoposide) did not cause any further change in G1/S arrest, as observed with the topoisomerase inhibitors alone. TSA incubation alone for 6 h did not cause any significant change in the cell cycle or in the protein or transcript levels of RECQ4 (Figures 4A–C and data not shown). Transcriptionally active p53 was induced due to the treatment with etoposide or camptothecin, either alone or along with the deacetylase inhibitor (Figures 4C (b, d) and 4D (b)).

Figure 4
figure4

Decrease in RECQ4 due to DNA damage is reversed by TSA. (A) Cell cycle profile of NHF in the presence of camptothecin, TSA and camptothecin+TSA. NHF were either left untreated or treated with camptothecin, camptothecin+TSA or TSA for 6 h. The cell cycle profiles were analysed by flow cytometry in FACS Calibur, the data analysed and quantitated by ModFit v3.0. (B) p53-dependent decrease in the RECQ4 transcript due to camptothecin was reversed by TSA treatment. NHF was either left untreated (lane 1) or treated with camptothecin (lane 2), camptothecin+TSA (lane 3), or TSA (lane 4) for 6 h. Total RNA (20 μg) was extracted and probed with a cDNA probe for (a) RECQ4 and (b) GAPDH. (C) Decrease in RECQ4 protein after camptothecin treatment was inhibited by TSA. NHF was treated as in (B), except that cells extracts were prepared. Western blots (50 μg of the lysate) were carried out with antibodies against (a) RECQ4, (b) p53 (DO-1), (c) p53 (Lys382), (d) p21 and (e) TBP. (D) Decrease in RECQ4 protein after etoposide treatment was inhibited by TSA. Same as (C), except that etoposide was used. Western blots (50 μg of the lysate) were carried out with antibodies against (a) RECQ4, (b) p53 (DO-1) and (c) TBP

In vivo, the principal site at which p53 is acetylated in response to different kinds of stress is at Lysine 382 (Appella and Anderson, 2001). Camptothecin-dependent acetylation of p53 at Lysine 382 was effectively inhibited by TSA cotreatment, indicating that at the experimental time point TSA retained its activity (Figure 4C, c). Most significantly, the camptothecin- or etoposide-induced decreases in the levels of the RECQ4 transcript and protein were reversed when NHF was incubated in the presence of both the topoisomerase inhibitors and TSA (Figures 4B–D and data not shown).

Transcription activators and repressors bind RECQ4 promoter in a coordinated manner

The above results indicate that the p53-mediated repression on the RECQ4 promoter may be due to chromatin remodeling, mediated by the transcriptional repressors, the HDACs. To determine whether the HDACs are indeed bound with the RECQ4 promoter (Figure 5A) in a DNA damage-dependent manner, we carried out chromatin immunoprecipitation (ChIP) analysis on the RECQ4 promoter with the HDAC1 antibody. Both of the topoisomerase inhibitors show similar p53-dependent repression on the RECQ4 promoter, therefore, the ChIP assays were performed only with camptothecin. An increasing amount of HDAC1 was reproducibly recruited to the promoter in the presence of camptothecin (Figures 5B, lanes 17,18). Interestingly, HDAC1 was no longer bound to the RECQ4 or p21 promoter in the presence of TSA alone or TSA in combination with camptothecin (Figures 5B, C, lanes 19,20 and D, lanes 15,16). This indicated that the inhibition of HDAC1 activity by TSA probably led to a change in its conformation, subsequently causing a loss in the deacetylase binding to the promoter. A similar loss of HDAC1 binding to promoters in the presence of TSA has been reported (Mishra et al., 2001; Ghoshal et al., 2002).

Figure 5
figure5

DNA damage-dependent displacement of p53 and SP1 from the RECQ4 promoter. (A) Schematic representation of the RECQ4 promoter. The proximal RECQ4 promoter, 594 bp upstream of the start codon, consists of three SP1 sites. The primers used for ChIP assays are depicted. (B) Displacement of SP1 and p53 during repression of the RECQ4 promoter in NHF. ChIP assays were carried out in NHF in the absence (−) or presence of 6 h treatment of camptothecin (C), camptothecin+TSA (C+T) or TSA (T) for the RECQ4 promoter. The antibodies used were the following: anti-SP1 (lanes 5–8), anti-p53 (DO-1, Pharmingen) (lanes 9–12), anti-pol II (lanes 13–16), anti-HDAC1 (lanes 17–20) and anti-IgG (lanes 21–24). Input (lanes 1–4): genomic DNA prior to immunoprecipitation. The primers used for the amplification of the RECQ4 promoter were F2 and R2. (C) The accumulation of p53 and p21 during activation of the p21 promoter. Same as (B) except that ChIP assays were carried on genomic DNA isolated from NHF grown in the absence (−) or presence of camptothecin (C), campothecin+TSA (C+T) or TSA (T) for the p21 promoter. The antibodies used were the following: anti-SP1 (lanes 5–8), anti-p53 (DO-1, Pharmingen) (lanes 9–12), anti-pol II (lanes 13–16), anti-HDAC1 (lanes 17–20) and anti-IgG (lanes 21–24). The primers spanning the distal (for p53 binding sites) and proximal (for SP1-binding sites) regions on the p21 promoter have been previously described (Koshiji et al., 2004). (D) SP1 is not displaced from the RECQ4 promoter in NHF E6 cells. Same as (B) except that ChIP assays were carried out on genomic DNA isolated from NHF E6 cells. The antibodies used were the following: anti-SP1 (lanes 5–8), anti-pol II (lanes 9–12), anti-HDAC1 (13–16) and anti-IgG (lanes 17–20). The primers used for the amplification of the RECQ4 promoter were F2 and R2. (E) p53 and SP1 physically interacted in the presence of camptothecin. NHF was left untreated (lane 1), or treated with camptothecin (lane 2) or camptothecin+TSA (lane 3). Lysates (400 μg) were immunoprecipitated (IPed) with p53 (DO-1) (lanes 4–6), SP1 (lanes 7–9) antibodies or in the presence of the corresponding IgG (lane 10). Input (lanes 1–3) indicates that 10% of the lysate was used for IP. Co-IPed proteins were analysed with (a) SP1 and (b) p53 (DO-1) antibodies. Lysate (40 μg) was used as input in the Western blots

p53 and the transcription factor, SP1, have both been implicated in WRN repression. Two SP1 elements proximal to the transcription initiation site are indispensable for WRN promoter activity and bound specifically to SP1 protein. Deletion or mutations of the SP1-binding sites prevented repression by wild-type p53 (Yamabe et al., 1998). Three SP1 sites were also present in the proximal RECQ4 promoter. Using different combinations of primers, we found that the SP1 site nearest to the ATG codon did not bind the SP1 protein in ChIP assays (data not shown). Hence, the amplification of the RECQ4 promoter during the ChIP assays was carried out using the primers F2 and R2, which amplified a 230 bp fragment on the RECQ4 promoter (Figure 5A). We found that p53 and SP1 showed very similar profiles of RECQ4 promoter occupancy. Under conditions where the promoter was not repressed (as assessed by increased binding of RNA polymerase II to the RECQ4 promoter in Figure 5B and demonstrated at the RNA and protein level in Figure 2B, C), both p53 and SP1 strongly bound to the RECQ4 promoter (Figure 5B, lanes 5–12). However, in the presence of camptothecin, when the RECQ4 promoter was repressed (as assessed by decreased binding of RNA polymerase II to the RECQ4 promoter in Figure 5B and reduced levels of RECQ4 protein and transcript as seen in Figures 2B and C), both p53 and SP1 were concurrently unbound from the promoter (Figure 5B, compare lane 6 with lanes 5, 7, 8 and lane 10 with lanes 9,11,12). This effect of p53 and SP1 was specific for the RECQ4 promoter and not for a p53-activated distal p21 promoter, where the level of bound p53 increased while there was no change in the level of bound SP1 after camptothecin treatment (Figure 5C, lanes 5–12). Under the same experimental conditions, there was no change in the level of SP1 bound to the RECQ4 promoter in NHF E6 cells, indicating that its binding to the promoter directly depended on p53 (Figure 5D, lanes 5–8). The p53-dependent repression of the RECQ4 promoter was further confirmed by utilizing the MRC5 siRNAp53 cells (Supplementary Figure 2), where similar occupancy profiles for RECQ4 promoter were observed. Confirming the protein expression profiles (Figures 2D and E), the RECQ4 promoter was not repressed in either NHF E6 or MRC5 siRNAp53 cells, as observed by the equivalent RNA polymerase II level in the ChIP assay in all the tested conditions (Figure 5D and Supplementary Figure 2, lanes 9–12).

These results revealed that two concurrent processes possibly mediate p53-dependent RECQ4 repression: enhanced binding of the transcriptional repressor, HDAC1, and decreased binding of p53 and SP1 to the promoter. Hence, the possibility remained that during DNA damage, the promoter-excluded p53 and SP1 can form a complex that, in turn, titrated away SP1 from binding to the RECQ4 promoter. To examine this hypothesis, coimmunoprecipitations were carried out in NHF in the absence or presence of camptothecin (or camptothecin and TSA). TSA treatment alone did not induce any complex formation between p53 and SP1 (data not shown). A strong and specific endogenous p53–SP1 complex was observed by reciprocal coimmunoprecipitations, only in the presence of camptothecin (Figure 5E, lanes 5 and 8). The lack of a complex formation during camptothecin plus TSA treatment (Figure 5E, lanes 6 and 9) indicated that other HDAC-mediated events, like protein acetylation, might play a role in p53–SP1 complex formation.

Discussion

Using genetically defined nontransformed human cell lines, we have demonstrated that like BLM and WRN, RECQ4 also has a functional interaction with p53. The reciprocal relationship between p53 and RECQ4 exists during G1/S arrest, induced either without (Figure 1) or with (Figure 2) DNA damage. Wild-type p53 (but not the tumor-derived mutant forms) can repress the RECQ4 promoter (Figure 3). The p53-mediated repression can be relieved by the deacetylase inhibitor, TSA (Figure 4) and involved the concurrent removal of both p53 and SP1 from the RECQ4 promoter (Figure 5). Hence, we propose that RECQ4 is also a direct target of p53-mediated transcriptional repression.

Various mechanisms have been proposed for p53-mediated transacriptional repression. One of the mechanisms, called recruitment inhibition model, involves the interaction of p53 with activators at the promoter of the target genes, thus interfering with the function of the activator by formation of protein complexes (Ho and Benchimol, 2003). For example, the formation of the SP1–p53 complex was proposed to be the mechanism for the repression of the WRN promoter (Yamabe et al., 1998). Apart from WRN, the p53–SP1 interaction has been demonstrated to be involved in the repression of other p53-target genes. p53 bound and repressed the SP1-stimulated human immunodeficiency virus-long-terminal repeat (HIV-LTR) or human polymerase δ subunit gene (POLD1) promoter activity. This repression is largely due to the loss of sequence-specific interaction between SP1 proteins and the SP1-binding sites on the promoter, which overlaps with the p53-binding sites (Bargonetti et al., 1997; Li and Lee, 2001). The p53–SP1 complex prevents the binding of SP1 to the vascular permeability factor/vascular endothelial growth factor (VPF/VEGF), SV40 large T and human telomerase reverse transcriptase (hTERT) promoters (Perrem et al., 1995; Xu et al., 2000; Pal et al., 2001).

The second mechanism of p53-dependent repression involves the interference of the basal transcription machinery by p53 without an apparent need to act though gene-specific activators or their binding sites (Ho and Benchimol, 2003). For instance, cyclin B was downregulated in a p53-dependent manner even when binding sites for known regulators like SP1 and NF-Y were mutated (Krause et al., 2000). This is possibly because p53 can interact with TATA-binding protein (TBP) and certain TAFs (Seto et al., 1992; Horikoshi et al., 1995; Thut et al., 1995). p53 can also compete with TBP for binding to a promoter fragment from the cyclooxygenase-2 (Cox-2) gene (Subbaramaiah et al., 1999), thereby providing a possible mechanism to explain the p53-dependent repression of Cox-2.

The third mechanism for p53-dependent repression involves the alteration of the chromatin structure of the promoters by recruiting proteins like HDACs (Ho and Benchimol, 2003). The transcriptional repression by wild-type p53 at the Map4 and stathmin promoters utilizes HDAC1 and is mediated via the interaction with corepressor mSin3a (Murphy et al., 1999). TSA was shown to inhibit p53-mediated repression on the Map4, stathmin (Murphy et al., 1999) and the survivin (Mirza et al., 2002) promoters. Therefore, p53-mediated repression on some of the target promoters may involve the recruitment of chromatin modifying factors (like HDAC1) and the chromatin-remodeling complex, SWI/SNF, with which p53 is known to interact in vivo (Lee et al., 2002).

Based on our results, it is possible that p53 can repress the RECQ4 promoter by any of the above three or a combination of the above mechanisms. For example, it is possible that the enhanced targeting of HDACs during drug-mediated repression possibly modifies the chromatin around the RECQ4 promoter, which precludes p53 and SP1 from binding to the promoter itself. Instead, p53 and SP1 form a complex outside the promoter, which can in turn, bind to the promoter regions of p53-activated genes like p21. However, repression of the RECQ4 promoter solely by interfering with the basal transcription machinery or by inhibiting the recruitment of the transcriptional activators are also possible.

Various post-translational modifications, especially phosphorylation and acetylation, are critical for the activation of p53 (Brooks and Gu, 2003). p53 has been shown to be acetylated on multiple lysine residues (Lysine 370, Lysine 371, Lysine 372, Lysine 381 and Lysine 382) of the carboxy-terminal regulatory domain by p300/CREB-binding protein (p300/CBP) and, to a lesser extent, by p300/CBP-associated factor (PCAF) (Appella and Anderson, 2001; Brooks and Gu, 2003). Once the transcriptional activation of p53-target genes is no longer needed, p53 undergoes deacetylation. The role of HDAC activity on p53 deacetylation is not yet fully understood. An adaptor protein known as PID (for p53 target protein in the deacetylase complexes) or MTA2 (for metastasis-associated protein 2) can enhance HDAC1-mediated deacetylation of p53 (Luo et al., 2000). The TSA-resistant, NAD-dependent histone deacetylase, Sir2α (also called SIRT1), can deacetylate p53 and attenuate its transacriptional activity (Luo et al., 2001; Vaziri et al., 2001; Langley et al., 2002). It can be hypothesized that the acetylation of p53 can also play a role in the repression of its various target genes. p53 acetylation on Lysine 382 in the presence of campthothecin is concordant with p53-dependent repression of RECQ4. Hence, a possibility exists for a role of p53 acetylation during transcriptional repression of RECQ4.

How the different players are targeted to the RECQ4 promoter is a subject of ongoing research. Canonical p53-response element(s) are not present on the RECQ4 promoter. However, p53 can still be targeted to the promoter either via its carboxy terminal region (that interacts nonspecifically with DNA) or due to interaction with a novel, yet unknown, DNA-binding site. It is also possible that at least another not yet identified protein is required to target p53 to the RECQ4 promoter. It is interesting to note that p53 can form a complex with HDAC1 and mSin3A on the Map4 promoter, another p53-repressed gene (Murphy et al., 1999). From the ChIP data (Figure 5B), it is possible that a similar basal level of p53–HDAC1 interaction also exists on the RECQ4 promoter. The enhanced binding of the HDAC1 to the RECQ4 promoter during repression can be due to several mutually nonexclusive mechanisms. HDAC1 recruitment to promoters is inversely correlated with histone H4 acetylation status on specific lysines (Ferreira et al., 2001). Interestingly, decrease in histone acetylation has been linked to DNA methylation (Eden et al., 1998) and methylation of CpG islands is associated with transcriptional silencing of genes (Jones et al., 1998). It is quite possible that the CpG islands present within the SP1-binding sites on the RECQ4 promoter undergo drug or cell-cycle-dependent changes in the acetylation and methylation status, which in turn, resulted in enhanced recruitment of HDAC1.

So why should p53 repress the RECQ4 helicase? Biological functions of RECQ4 are still largely unknown; therefore, a definite postulation of the reason and consequences of p53-mediated repression on RECQ4 may be premature. However, it can be speculated that the repression of RECQ4 can act as a ‘facilitator’ for p53 tetramerization and stabilization. It is also possible that RECQ4 (and WRN) repression is a temporal process and is a way for the cells to create ‘back-up systems’ to maintain genomic integrity. During the G1 phase of the cell cycle or during DNA damage, p53 maintains genomic fidelity by cell cycle arrest or apoptosis. At this step, helicase expressions are not obligatory, thus the levels of WRN or RECQ4 are low or repressed. Instead, p53 switches onto its role as a transactivator, redistributing the transcription factor pool (like SP1) inside the cells to transactivate genes involved in its downstream functions. This can be constituted as the ‘first line of defense’ by the cell against genomic instability. However, if the cells do progress beyond the G1/S checkpoint into the S phase without the DNA damage being completely removed, back-up processes, mediated by WRN and RECQ4, may be switched on in another attempt to correct the lesions, thereby constituting the ‘second line of defense’. During this phase, p53 becomes transcriptionally inactive and no longer represses the promoter, allowing the transcription of RECQ4. Both RTS and Li–Fraumeni syndrome (with mutated p53) patients have been found to have an increased risk of having osteosarcoma (Fuchs and Pritchard, 2002). Hence, lack of repression by the tumor-derived p53 mutants (Figure 3) of the RECQ4 promoter may indeed have functional significance in tumor progression.

Materials and methods

Antibodies

Anti-p53: monoclonal, DO-1 and 1801 (Santa Cruz and Pharmingen). Anti-p53 (Lys382): polyclonal, Ab-1 (Oncogene Research Product). Anti-RECQ4: monoclonal, K6312 (Kawabe et al., 2000). Anti-TBP: monoclonal, 58C9 (Santa Cruz). Anti-p21WAF1/CIP1: monoclonal, Ab-1 (Oncogene Research Product). Anti-SP1: monoclonal, 1C6 (Santa Cruz). Anti-pol II: monoclonal, 7C2 (Puvion-Dutilleul et al., 1997). Anti-HDAC1: polyclonal, H51 (Santa Cruz). Pol II antibody was generously provided by John Bradsher (NCI, NIH).

Recombinants

The GST-p53 constructs (Sengupta and Wasylyk, 2001; Wang et al., 2001), p21–luc (el-Deiry et al., 1993), wild-type p53 (Hinds et al., 1990) and p53 (22,23) (Lin et al., 1994) have been described earlier and were generously provided by Arnold J Levine (Cancer Institute of New Jersey Robert Wood Johnson School of Medicine, New Brunswick, NJ, USA) and Bert Vogelstein (Johns Hopkins University, Baltimore, MD, USA). RECQ4-luc was generated by cloning the proximal 594 bases of the RECQL4 promoter into NruI/NcoI sites of the pGL2 construct.

Cells, synchronization, culture conditions and treatments

The hTERT-immortalized normal human fibroblast strains. GM07532-hTERT (NHF) and NHF E6 were maintained as described (Sengupta et al., 2003). Primary human diploid fibroblasts GM08402, GM03348 (both purchased from Coriell Cell Repositories), and MRC5 (purchased from ATCC) were maintained according to the supplier's specifications. p53 depletion in MRC5 was carried out with a published siRNA (Brummelkamp et al., 2002).

Cells were synchronized by growth to confluence and replated at low density. Parallel aliquots of the cells were used either for lysis and subsequent Western blotting, or for flow cytomenty. Cells were treated with camptothecin (Sigma, 500 ng/ml) or etoposide (Sigma, 20 μg/ml) as indicated. Coincubation with TSA (Sigma, 20 nM) was carried out during the entire duration of the drug treatment.

IP and Western blots

IPs were carried out as previously described (Sengupta et al., 2003). The cells were lysed in a modified RIPA buffer (1 mM Tris HCl, pH 7.4, 150 mM NaCl, 1% sodium deoxycholate, 0.1% SDS, 1 mM PMSF and protease inhibitors by two rounds of freeze–thawing, followed by centrifugation. IPs were carried out with 400 μg of lysates with the primary antibody coupled Protein G Sepharose 4 (Amersham) in the presence of PBS and 0.1% NP-40 for 2 h at 4°C. Western blots were carried out by standard protocols using 50 μg of the total cell lysate.

Flow cytometry

Flow cytometry was carried out according to the previously published protocol (Koshiji et al., 2004). In short, the cells were washed with PBS containing no serum and resuspended. The cells were incubated in an icebath for 15 min and fixed in ice-cold 70% ethanol. The fixed cells were stored for a minimum of 2 h and subsequently treated with RNase (100 U). Cells were subsequently incubated with propidium iodide (diluted in PBS, 50 μg/ml) in the dark for 30 min. Cells were analysed in a FACS Calibur and the data was analysed and quantitated by ModFit v3.0.

Northern analysis

Northern blots were carried out by standard protocols using 20 μg of total RNA. The full-length RECQ4 cDNA was used as the probe for Northern blot (Kitao et al., 1999). The experiments were repeated at least twice and representative blots are shown.

Transfection and reporter assays

Transfections were carried out in six-well cluster plates with Fugene 6 transfection reagent (Roche) according to the manufacturer's protocol, with 1.5 μg of the luciferase reporter and 0.1 μg of pEYFP-Nuc (Clontech) for normalization. The amounts of p53 constructs used were 0.05, 0.1 and 0.2 μg. The total amount of DNA transfected per well was 4 μg. Lysates were prepared 24 h post-transfection, and luciferase and green fluorescence were assayed. ‘Post-transfection’ is defined as the time lapsed after adding the Fugene 6–DNA complex to the cells.

ChIP assays

ChIP assays were performed using the ChIP assay kit (Upstate), following the manufacturer's protocol. The primers used for RECQ4 promoter are the following:F1 (5′-IndexTermCTGGACCCTCCGCTCTTT-3′; −528 to −545);F2 (5′-IndexTermCTCCCAAATGCAGCCACT-3′; −373 to −389);F3 (5′-IndexTermCTCCCATTGGCTGCTTGT-3′, −149 to −166);R1 (5′-IndexTermGATCGTCCAGCGAATCTCC-3′, −35 to −54);R2 (5′-IndexTermCTCGGACAAGCAGCCAAT-3′, −144 to −161).

The primers spanning the distal (for p53-binding sites) and proximal (for SP1-binding sites) regions on the p21 promoter have been previously described (Koshiji et al., 2004).

References

  1. Appella E and Anderson CW . (2001). Eur. J. Biochem., 268, 2764–2772.

  2. Bargonetti J, Chicas A, White D and Prives C . (1997). Cell. Mol. Biol. (Noisy-le-grand), 43, 935–949.

  3. Brooks CL and Gu W . (2003). Curr. Opin. Cell. Biol., 15, 164–171.

  4. Brummelkamp TR, Bernards R and Agami R . (2002). Science, 296, 550–553.

  5. Davalos AR and Campisi J . (2003). J. Cell Biol., 162, 1197–1209.

  6. Drouin CA, Mongrain E, Sasseville D, Bouchard HL and Drouin M . (1993). J. Am. Acad. Dermatol., 28, 301–305.

  7. Eden S, Hashimshony T, Keshet I, Cedar H and Thorne AW . (1998). Nature, 394, 842.

  8. el-Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Trent JM, Lin D, Mercer WE, Kinzler KW and Vogelstein B . (1993). Cell, 75, 817–825.

  9. el-Khoury JM, Haddad SN and Atallah NG . (1997). Br. J. Radiol., 70, 215–218.

  10. Ferreira R, Naguibneva I, Mathieu M, Ait-Si-Ali S, Robin P, Pritchard LL and Harel-Bellan A . (2001). EMBO Rep., 2, 794–799.

  11. Fuchs B and Pritchard DJ . (2002). Clin. Orthop., 40–52.

  12. Furuichi Y . (2001). Ann. N. Y. Acad. Sci., 928, 121–131.

  13. Ghoshal K, Datta J, Majumder S, Bai S, Dong X, Parthun M and Jacob ST . (2002). Mol. Cell. Biol., 22, 8302–8319.

  14. Gottifredi V, Shieh S, Taya Y and Prives C . (2001). Proc. Natl. Acad. Sci. USA, 98, 1036–1041.

  15. Haupt Y, Robles AI, Prives C and Rotter V . (2002). Oncogene, 21, 8223–8231.

  16. Hickson ID . (2003). Nat. Rev. Cancer, 3, 169–178.

  17. Hinds PW, Finlay CA, Quartin RS, Baker SJ, Fearson ER, Vogelstein B and Levine AJ . (1990). Cell Growth Differ., 1, 571–580.

  18. Ho J and Benchimol S . (2003). Cell Death Differ., 10, 404–408.

  19. Hofseth LJ, Hussain SP and Harris CC . (2004). Trends Pharmacol. Sci., 25, 177–181.

  20. Hoki Y, Araki R, Fujimori A, Ohhata T, Koseki H, Fukumura R, Nakamura M, Takahashi H, Noda Y, Kito S and Abe M . (2003). Hum. Mol. Genet., 12, 2293–2299.

  21. Hollstein M, Sidransky D, Vogelstein B and Harris CC . (1991). Science, 253, 49–53.

  22. Horikoshi N, Usheva A, Chen J, Levine AJ, Weinmann R and Shenk T . (1995). Mol. Cell. Biol., 15, 227–234.

  23. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J and Wolffe AP . (1998). Nat. Genet., 19, 187–191.

  24. Kawabe T, Tsuyama N, Kitao S, Nishikawa K, Shimamoto A, Shiratori M, Matsumoto T, Anno K, Sato T, Mitsui Y, Seki M, Enomoto T, Goto M, Ellis NA, Ide T, Furuichi Y and Sugimoto M . (2000). Oncogene, 19, 4764–4772.

  25. Kitao S, Lindor NM, Shiratori M, Furuichi Y and Shimamoto A . (1999). Genomics, 61, 268–276.

  26. Kitao S, Ohsugi I, Ichikawa K, Goto M, Furuichi Y and Shimamoto A . (1998). Genomics, 54, 443–452.

  27. Ko LJ and Prives C . (1996). Genes Dev., 10, 1054–1072.

  28. Koshiji M, Kageyama Y, Pete EA, Horikawa I, Barrett JC and Huang LE . (2004). EMBO J., 23, 1949–1956.

  29. Koumenis C, Alarcon R, Hammond E, Sutphin P, Hoffman W, Murphy M, Derr J, Taya Y, Lowe SW, Kastan M and Giaccia A . (2001). Mol. Cell. Biol., 21, 1297–1310.

  30. Krause K, Wasner M, Reinhard W, Haugwitz U, Dohna CL, Mossner J and Engeland K . (2000). Nucleic Acids Res., 28, 4410–4418.

  31. Langley E, Pearson M, Faretta M, Bauer UM, Frye RA, Minucci S, Pelicci PG and Kouzarides T . (2002). EMBO J., 21, 2383–2396.

  32. Lee D, Kim JW, Seo T, Hwang SG, Choi EJ and Choe J . (2002). J. Biol. Chem., 277, 22330–22337.

  33. Levine AJ, Momand J and Finlay CA . (1991). Nature, 351, 453–456.

  34. Li B and Lee MY . (2001). J. Biol. Chem., 276, 29729–29739.

  35. Lin J, Chen J, Elenbaas B and Levine AJ . (1994). Genes. Dev., 8, 1235–1246.

  36. Luo J, Nikolaev AY, Imai S, Chen D, Su F, Shiloh A, Guarente L and Gu W . (2001). Cell, 107, 137–148.

  37. Luo J, Su F, Chen D, Shiloh A and Gu W . (2000). Nature, 408, 377–381.

  38. Michalovitz D, Halevy O and Oren M . (1991). J. Cell Biochem., 45, 22–29.

  39. Mirza A, McGuirk M, Hockenberry TN, Wu Q, Ashar H, Black S, Wen SF, Wang L, Kirschmeier P, Bishop WR, Nielsen LL, Pickett CB and Liu S . (2002). Oncogene, 21, 2613–2622.

  40. Mishra SK, Mandal M, Mazumdar A and Kumar R . (2001). FEBS Lett., 507, 88–94.

  41. Munger K and Howley PM . (2002). Virus Res., 89, 213–228.

  42. Murphy M, Ahn J, Walker KK, Hoffman WH, Evans RM, Levine AJ and George DL . (1999). Genes Dev., 13, 2490–2501.

  43. Oren M, Damalas A, Gottlieb T, Michael D, Taplick J, Leal JF, Maya R, Moas M, Seger R, Taya Y and Ben-Ze’Ev A . (2002). Ann. N. Y. Acad. Sci., 973, 374–383.

  44. Pal S, Datta K and Mukhopadhyay D . (2001). Cancer Res., 61, 6952–6957.

  45. Perrem K, Rayner J, Voss T, Sturzbecher H, Jackson P and Braithwaite A . (1995). Oncogene, 11, 1299–1307.

  46. Puvion-Dutilleul F, Besse S, Diaz JJ, Kindbeiter K, Vigneron M, Warren SL, Kedinger C, Madjar JJ and Puvion E . (1997). Gene Expr., 6, 315–332.

  47. Santhanam U, Ray A and Sehgal PB . (1991). Proc. Natl. Acad. Sci. USA, 88, 7605–7609.

  48. Sengupta S and Harris CC . (2005). Nat. Rev. Mol. Cell. Biol., 6, 44–55.

  49. Sengupta S, Linke SP, Pedeux R, Yang Q, Farnsworth J, Garfield SH, Valerie K, Shay JW, Ellis NA, Wasylyk B and Harris CC . (2003). EMBO J., 22, 1210–1222.

  50. Sengupta S, Robles AI, Linke SP, Sinogeeva NI, Zhang R, Pedeux R, Ward IM, Celeste A, Nussenzweig A, Chen J, Halazonetis TD and Harris CC . (2004). J. Cell Biol., 166, 801–813.

  51. Sengupta S and Wasylyk B . (2001). Genes Dev., 15, 2367–2380.

  52. Seto E, Usheva A, Zambetti GP, Momand J, Horikoshi N, Weinmann R, Levine AJ and Shenk T . (1992). Proc. Natl. Acad. Sci. USA, 89, 12028–12032.

  53. Shimamoto A, Sugimoto M and Furuichi Y . (2004). Int. J. Clin. Oncol., 9, 288–298.

  54. Subbaramaiah K, Altorki N, Chung WJ, Mestre JR, Sampat A and Dannenberg AJ . (1999). J. Biol. Chem., 274, 10911–10915.

  55. Thut CJ, Chen JL, Klemm R and Tjian R . (1995). Science, 267, 100–104.

  56. Vaziri H, Dessain SK, Ng Eaton E, Imai SI, Frye RA, Pandita TK, Guarente L and Weinberg RA . (2001). Cell, 107, 149–159.

  57. Vennos EM and James WD . (1995). Dermatol. Clin., 13, 143–150.

  58. Vogelstein B, Lane D and Levine AJ . (2000). Nature, 408, 307–310.

  59. Wang LL, Gannavarapu A, Kozinetz CA, Levy ML, Lewis RA, Chintagumpala MM, Ruiz-Maldanado R, Contreras-Ruiz J, Cunniff C, Erickson RP, Lev D, Rogers M, Zackai EH and Plon SE . (2003). J. Natl. Cancer Inst., 95, 669–674.

  60. Wang XW, Tseng A, Ellis NA, Spillare EA, Linke SP, Robles AI, Seker H, Yang Q, Hu P, Beresten S, Bemmels NA, Garfield S and Harris CC . (2001). J. Biol. Chem., 276, 32948–32955.

  61. Webster NJ, Resnik JL, Reichart DB, Strauss B, Haas M and Seely BL . (1996). Cancer Res., 56, 2781–2788.

  62. Xu D, Wang Q, Gruber A, Bjorkholm M, Chen Z, Zaid A, Selivanova G, Peterson C, Wiman KG and Pisa P . (2000). Oncogene, 19, 5123–5133.

  63. Yamabe Y, Shimamoto A, Goto M, Yokota J, Sugawara M and Furuichi Y . (1998). Mol. Cell. Biol., 18, 6191–6200.

  64. Yin J, Kwon YT, Varshavsky A and Wang W . (2004). Hum. Mol. Genet., 13, 2421–2430.

  65. Zambetti GP and Levine AJ . (1993). FASEB J., 7, 855–865.

  66. Zhao R, Gish K, Murphy M, Yin Y, Notterman D, Hoffman WH, Tom E, Mack DH and Levine AJ . (2000). Genes Dev., 14, 981–993.

Download references

Acknowledgements

We thank Bert Vogelstein, Arnold J Levine and John Bradsher for the gifts of the recombinants and antibody. We also thank Dorothea Dudek for editorial assistance and Karen MacPherson for bibliographic assistance.

Author information

Correspondence to Curtis C Harris.

Additional information

Supplementary Information accompanies the paper on Oncogene website (http://www.nature.com/onc)

Supplementary information

Supplementary Figure 1 (TIF 8252 kb)

Supplementary Figure 2 (TIF 9761 kb)

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • repression
  • histone deacetylases
  • trichostatin A
  • Rothmund–Thomson syndrome
  • SP1
  • promoter

Further reading