|Home | About | Journals | Submit | Contact Us | Français|
In response to γ-radiation-induced DNA damage, organisms either activate cell cycle checkpoint and repair machinery or undergo apoptosis to eliminate damaged cells. Although previous studies indicated that the tumor suppressor p53 is critically involved in mediating both responses, how a cell decides which pathway to take is not well established. The zinc-finger-containing transcription factor, Krüppel-like factor 4 (KLF4), is a crucial mediator for the checkpoint functions of p53 after γ-irradiation and does so by inhibiting the transition from the G1 to S and G2 to M phases of the cell cycle. Here, we determined the role of KLF4 in modulating the apoptotic response following γ-irradiation. In three independent cell systems including colorectal cancer cells and mouse embryo fibroblasts in which expression of KLF4 could be manipulated, we observed that γ-irradiated cells underwent apoptosis if KLF4 was absent. In the presence of KLF4, the degree of apoptosis was significantly reduced and cells resorted to checkpoint arrest. The mechanism by which KLF4 accomplished this antiapoptotic effect is by activating expression of the cell cycle arrest gene, p21WAF1/CIP1, and by inhibiting the ability of p53 to transactivate expression of the proapoptotic gene, BAX. Results of our study illustrate an unexpected antiapoptotic function of KLF4, heretofore considered a tumor suppressor in colorectal cancer, and suggest that KLF4 may be an important determinant of cell fate following γ-radiation-induced DNA damage.
Cells respond to DNA damage from γ-irradiation either by activating cell cycle checkpoint and repair machinery or by triggering apoptosis. The damage to the DNA is believed to result from the production of free radicals (Lewanski and Gullick, 2001). Apoptosis is a self-destruct process with which to prevent cells from perpetuating mutations that might be harmful to the whole organism. The exact mechanism by which a cell chooses between life and death is still not well understood and believed to be dependent on the cell context. However, it is clear that an important molecule involved in both responses following γ-irradiation is the tumor suppressor p53.
p53 exerts its cellular effects by modulating expression of a host of genes (Vogelstein et al., 2000). Examples of proapoptotic target genes of p53 following DNA damage include BAX, PUMA, FAS and NOXA (Miyashita and Reed, 1995; Fei and El-Deiry, 2003; Mihara et al., 2003; Erster et al., 2004). Examples of p53-dependent cell cycle checkpoint target genes following DNA damage include those encoding the cyclin-dependent kinase inhibitor, p21WAF1/CIP1 (El-Deiry et al., 1993; Waldman et al., 1995) and 14-3-3δ (Hermeking et al., 1997), which lead to arrest at the G1/S and G2/M boundaries, respectively, in the cell cycle. We previously demonstrated that the gene encoding Krüppel-like factor 4 (KLF4; also called gut-enriched Krüppel-like factor or GKLF) (Garrett-Sinha et al., 1996; Shields et al., 1996) is transcriptionally activated by p53 following DNA damage (Zhang et al., 1998) and causes cell cycle arrest at both the G1/S and G2/M boundaries (Yoon et al, 2003; Yoon and Yang, 2004). These results indicate that KLF4 is an important factor in mediating the checkpoint functions of p53 following DNA damage. Accordingly, KLF4 has been shown to be a potential tumor suppressor in tumors of the gastrointestinal tract (Zhao et al., 2004; Katz et al., 2005; Wei et al., 2005).
In view of the evidence of the dependence of KLF4 expression on p53 following DNA damage by γ-irradiation, we decided to investigate whether KLF4 plays a role in the regulation of apoptosis during the DNA damage response. Here, we present evidence for a crucial role of KLF4 as an antiapoptotic factor, in part by suppressing BAX expression following γ-radiation-induced DNA damage.
We first examined the response of the human colon cancer cell line, RKO, to γ-radiation-induced DNA damage. Expression of KLF4 in RKO cells is negligible owing to hemizygous deletion of one KLF4 allele and silencing of the other allele by hypermethylation (Dang et al., 2001; Zhao et al., 2004). We previously established a derivative of RKO cells containing a stably transfected KLF4 cDNA under the control of an inducible promoter that responds to the insect hormone, ecdysone. This cell line, called EcR-RKO/pAdLoxEGI-KLF4 (Chen et al., 2001), produced a significant amount of KLF4 when treated with the inducer, ponasterone A. EcR-RKO/pAdLoxEGI-KLF4 cells were irradiated, or not, with 12Gy of γ-ray and maintained in the presence of ponasterone A, or not, for 72 h before being processed for fluorescence-activated cell sorting (FACS) analysis. Figure 1 shows that in the absence of KLF4 induction, an appreciable number of cells contained DNA in the sub-G1 fraction following irradiation (compare lane 2 with 1). In contrast, the fraction of cells in the sub-G1 phase was significantly reduced in irradiated cells that had also been induced to produce KLF4 (compare lanes 4 and 2). A similar finding was observed when the fractions of apoptotic cells were assessed by terminal deoxynucleotidyl transferase-mediated deoxyuridine tri-phosphate-biotin nick-end labeling stain (Supplementary Figure 1). These results suggest that induction of KLF4 expression suppresses apoptosis following γ-radiation-induced DNA damage.
Previous studies demonstrate that expression of p21WAF1/CIP1 is dependent on KLF4 following γ-irradiation (Yoon et al., 2003; Yoon and Yang, 2004). Studies also indicate that expression of the proapoptotic gene, BAX, is stringently regulated by p53 at the transcriptional level following DNA damage (Miyashita and Reed, 1995). To determine the levels of p21WAF1/CIP1 and BAX in relationship to the levels of KLF4 in EcR-RKO/pAdLoxEGI-KLF4 cells following γ-irradiation, we performed Western blot analysis for p53, KLF4, BAX and p21WAF1/CIP1. As shown in Figure 2, the level of p53 was increased in irradiated cells either in the absence or presence of KLF4 induction (lanes 2 and 4, respectively). KLF4 was absent from both unirradiated and irradiated cells without induction (lanes 1 and 2). In contrast, levels of KLF4 were significantly increased upon induction in both the unirradiated and irradiated states (lanes 3 and 4). The levels of p21WAF1/CIP1 correlated with those of KLF4, consistent with previous results showing that p21WAF1/CIP1 is a target of KLF4 (Zhang et al., 2000; Chen et al., 2001). BAX level was significantly increased in uninduced cells that had been irradiated (compare lane 2 to 1). However, this radiation-dependent increase was attenuated in cells that had been induced to produce KLF4 (compare lane 4 to 2). Both increased p21WAF1/CIP1 and decreased BAX levels following KLF4 induction and γ-irradiation correlated with the reduction in apoptosis as shown in Figure 1 and Supplementary Figure 1.
To validate the observation that KLF4 suppresses apoptosis following γ-irradiation, we examined the response of HCT116 human colon cancer cells to γ-radiation-induced DNA damage. HCT116 cells contain wild-type (WT) p53 and KLF4 alleles and undergo p21WAF1/CIP1-mediated cell cycle arrest following γ-irradiation (Yoon et al, 2003; Yoon and Yang, 2004). We previously developed a clonal derivative of HCT116, called HCT116-KLF4/sh2-2, in which expression of KLF4 is significantly reduced owing to the presence of small hairpin RNA (shRNA) against KLF4 (Yoon et al., 2005). Following γ-irradiation, HCT116-KLF4/sh2-2 cells exhibited a significantly higher percentage of cells in the sub-G1 fraction than the control cells, HCT116-KLF4sssh-c (Figure 3a and Supplementary Figure 2). Importantly, reduced KLF4 expression resulted in increased BAX level and decreased p21WAF1/CIP1 level in response to γ-irradiation (Figure 3b; compare lane 4 with 2). These findings provided further evidence for an antiapoptotic activity of KLF4 following DNA damage.
To investigate whether KLF4 is sufficient in reducing apoptosis following γ-irradiation, we compared the apoptotic levels of WT HCT116 cells with those null for BAX or p21WAF1/CIP1 following γ-irradiation. As shown in Figure 4a, radiation-induced apoptosis was significantly reduced in BAX-null cells relative to WT cells (compare lanes 4 and 2). In contrast, p21WAF1/CIP1-null cells exhibited an increased level of apoptosis following irradiation when compared to WT cells (compare lanes 6 and 2). In all three cell types, the induction of KLF4 was intact in response to γ-irradiation, as was that of p53 (Figure 4b). These results indicate that KLF4, although necessary, is not sufficient to inhibit apoptosis following irradiation in the absence of p21WAF1/CIP1. The results also demonstrate that BAX is required for the induction of apoptosis in response to γ-irradiation.
To further confirm the results obtained from RKO and HCT116 cells, both of which are cancerous, we investigated the response of mouse embryo fibroblasts (MEFs) derived from embryos that are WT (+/+) or null (−/−) for the Klf4 alleles (Katz et al., 2002). As shown in Figure 5a, Klf4−/− MEFs were highly prone to γ-radiation-induced apoptosis as compared to Klf4+/+ MEFs. We also examined MEFs heterozygous for the Klf4 alleles (Klf4+/−), which had a similar response to γ-irradiation as Klf4+/+ MEFs (results not shown). As expected, Klf4−/− MEFs lacked Klf4 expression in the absence or presence of γ-irradiation (Figure 5b; lanes 3 and 4). This was correlated with an absence of p21WAF1/CIP1 (Figure 5b). Klf4+/+ MEFs had a relatively high basal level of BAX expression that did not increase after γ-irradiation (Figure 5b; compare lanes 2 and 1). In contrast, Klf4−/− MEFs contained a low basal level of BAX that was significantly increased following γ-irradiation (Figure 5b; compare lane 4 to 3). To determine whether the propensity of the Klf4−/− MEFs to apoptosis is limited to γ-irradiation, we compared the apoptotic response of Klf4+/+ and Klf4−/− MEFs following treatment with 15μM cisplatin for 16 and 48 h (Supplementary Figure 3). As shown, Klf4−/− MEFs were more susceptible to cisplatin-induced apoptosis than Klf4+/+ MEFs. Combining the results obtained from RKO, HCT116 and MEFs, it appears that the relative levels of both BAX and p21WAF1/CIP1, which vary depending on the status of KLF4, are important in determining whether or not the cells undergo apoptosis following γ-irradiation.
To explore the possibility that KLF4 negatively regulates BAX expression at a transcriptional level, we examined the immediate 1000 base pair (bp) upstream sequence from the translation initiation site of the human BAX gene (Miyashita and Reed, 1995) for possible KLF4-binding sites based on the established consensus KLF4-binding sequence (Shields and Yang, 1997). A total of nine potential binding sites were identified (Supplementary Figure 4), with site number 5 located at nucleotide positions −464 and −458, immediately adjacent to an established p53-binding site (Miyashita and Reed, 1995).
To determine the ability of KLF4 to modulate activity of the BAX promoter, we subcloned the region of BAX gene between nucleotides positions −870 and −193 from the translation initiation site that contains the first five potential KLF4-binding sites, the p53-binding site, the TATA box and the transcription initiation site (Miyashita and Reed, 1995) into the pGL3-basic luciferase reporter, and performed co-transfection experiments with an expression construct of p53, KLF4 or both. Figure 6a shows that p53 activated and KLF4 repressed the BAX promoter-reporter (lanes 2 and 3, respectively). Moreover, increasing concentrations of KLF4 abrogated the ability of p53 to transactivate the BAX promoter (Figure 6a; lanes 4–7). These results indicate that p53 and KLF4 regulate the BAX promoter in opposite directions, that KLF4 inhibits the ability of p53 to transactivate the BAX promoter, and that the BAX gene promoter is repressed by KLF4 in the absence of p53.
We then determined whether the putative KLF4-binding site number 5, situated immediately adjacent to the established p53-binding site, is involved in modulating the effect of p53 on the BAX promoter. We subcloned a 100-bp fragment of the BAX promoter (nucleotide positions −530 to −431 relative to the translational start site) that contains the p53-binding site and the adjacent putative KLF4-binding site (site 5) into the pGL2-promoter vector that contains a minimal SV40 promoter element, and performed co-transfection studies. As seen in Figure 6b, p53 transactivated the reporter (lane 2) but KLF4 did not (lane 3). The lack of response of this short heterologous promoter to KLF4 may be due to its relatively low basal activity or because additional KLF4-binding sites in the BAX promoter are required for KLF4 to inhibit the promoter. However, and importantly, increasing concentrations of KLF4 were able to reduce the ability of p53 to transactivate the short heterologous promoter (lanes 4–7). These findings suggest that KLF4 is capable of either directly (through DNA binding) or indirectly (through interacting with p53) to interfere with the ability of p53 to regulate the BAX promoter.
We subsequently performed chromatin immunoprecipitation (ChIP) assays in co-transfected cells to see if p53 or KLF4 binds to the BAX promoter using primers that flanked nucleotide positions between −870 and − 193 from the translation initiation site. As seen in Figure 7a, both KLF4 and p53 bound to this region of the BAX promoter (lanes 2 and 3, respectively). Importantly, transfection of cells with increasing amounts of KLF4 in the presence of a constant amount of p53 resulted in the increasing binding of KLF4 and diminishing binding of p53 to the BAX promoter (Figure 7b). This finding was confirmed by ChIP assays designed to measure the ability of KLF4 to influence the binding of p53 to the endogenous BAX and p21WAF1/CIP1 promoters using the EcR-RKO/pAdLoxEGI-KLF4-inducible system. As shown in Figure 8a, upon induction, KLF4 bound to both the BAX and p21WAF1/CIP1 promoters (lanes 3 and 4). However, the binding of p53 to the BAX and p21WAF1/CIP1 promoters varied significantly depending whether KLF4 was absent or present. In the absence of KLF4, the binding of p53 to the BAX promoter was significantly increased after irradiation (compare lanes 2 and 1). In contrast, p53 failed to bind to the p21WAF1/CIP1 promoter in either un-irradiated or irradiated cells when KLF4 was absent. The induction of KLF4 increased the binding of p53 to the p21WAF1/CIP1 promoter in un-irradiated or irradiated cells (lanes 3 and 4), indicating that the binding of p53 to the p21WAF1/CIP1 promoter requires KLF4. In contrast, the binding of 53 to the BAX promoter was diminished upon irradiation and KLF4 induction. The results suggest that KLF4 exerts a differential effect on the binding of p53 to the BAX and p21WAF1/CIP1 promoters.
In response to genotoxic stress, organisms undergo either cell cycle arrest followed by DNA repair or apoptosis. The exact mechanism by which a cell decides which path to take is still unclear. Apoptosis is an evolutionarily conserved process that enables an organism to remove unwanted or damaged cells (Yu and Zhang, 2005). Several studies have indicated that the induction of apoptosis is an essential function of the tumor suppressor, p53 (Symonds et al., 1994; Schmitt et al., 2002; Liu et al., 2004), and that this function has been shown to exist in both normal (Clarke et al., 1993; Lowe et al., 1993a, b) and cancerous cells (Bunz et al., 1999). Other studies have indicated that this function is conserved in lower eukaryotes, at least in response to DNA damage (Jin et al., 2000; Schumacher et al., 2001; Brodsky et al., 2004).
Cell cycle arrest following DNA damage is considered to be an important factor in preventing p53-dependent apoptosis. Following DNA damage, p21WAF1/CIP1 has been shown to be a major mediator of p53-dependent cell cycle arrest (Waldman et al., 1996, 1997; Chan et al., 2000; Yu et al., 2003). The p53-dependent expression of p21WAF1/CPI1 following DNA damage and the resultant cell cycle arrest have been shown to be mediated by KLF4 (Zhang et al., 2000; Yoon et al., 2003). This led us to investigate whether KLF4 plays a role in regulating apoptosis following DNA damage by γ-irradiation. Our results show that KLF4 is required for suppressing apoptosis after γ-radiation-induced DNA damage in several independent cell systems investigated including RKO, HCT116 and MEFs.
Among the key regulators of the p53-dependent apoptosis is BAX, which is a member of the proapoptotic BCL-2 family proteins. The expression of BAX has been shown to be directly induced by p53 at a transcriptional level (Miyashita and Reed, 1995). Results of our studies indicate that BAX and p21WAF1/CIP1 levels in RKO and HCT116 cells are inversely correlated with the levels of KLF4 following γ-irradiation. Thus, increased KLF4 levels are correlated with increased p21WAF1/CIP1 and decreased BAX levels. The opposite is also true, that is, reduced KLF4 levels either owing to absence of endogenous expression in RKO cells or shRNA inhibition in HCT116 cells are correlated with reduced p21WAF1/CIP1 and increased BAX levels. Moreover, corresponding findings are noted in MEFs with WT or null alleles for Klf4 following γ-irradiation. In each of the three cell systems examined, reduced KLF4 levels are always associated with increased apoptosis after γ-irradiation. These results suggest that the ratio between BAX and p21WAF1/CIP1 levels may be crucial in determining whether the cells will undergo apoptosis or cell cycle arrest. In the absence of KLF4, the relative ratio of BAX to p21WAF1/CIP1 is high following γ-irradiation and cells undergo apoptosis. In contrast, if KLF4 is present, the relative ratio of BAX to p21WAF1/CIP1 is low after γ-irradiation and apoptosis becomes significantly reduced. The importance of the ratio of BAX to p21WAF1/CIP1 ratio is further substantiated by the finding that HCT116 null for p21WAF1/CIP1 cells are more sensitive and that HCT116 cells null for BAX are resistant to γ-irradiation-induced apoptosis compared to WT HCT116 cells. Our results also indicate that the protective effect of KLF4 against apoptosis is not limited to one type of DNA damage, at least in MEFs.
The finding of an inverse relationship between KLF4 and BAX led us to postulate that KLF4 might have a direct role in regulating BAX expression. Indeed, in co-transfection experiments, we show that whereas p53 transactivates the BAX promoter as previously demonstrated (Miyashita and Reed, 1995), KLF4 inhibits the same promoter. The ability of KLF4 alone to suppress the longer BAX promoter that contains multiple potential KLF4-binding sites (Figure 6a) and the inability of KLF4 to suppress the short, heterologous BAX promoter with a putative KLF4-binding site adjacent to the p53-binding site (Figure 6b) suggest that KLF4 requires full-length BAX promoter to exert its inhibitory effect. Be that as it may, KLF4 is able to reduce p53-mediated activation of the BAX promoter of either length in a dose-dependent fashion. A potential mechanism by which KLF4 accomplishes this task is to compete for the binding of p53 to the BAX promoter as shown by the ChIP experiments in Figures 7 and and8.8. It is also of interest to note from a previous study that KLF4 physically interacts with p53 (Zhang et al., 2000). Thus, it is possible that KLF4 exerts the inhibitory effect on p53’s ability to activate the BAX promoter by sequestering p53 from the BAX promoter to the p21WAF1/CIP1 promoter following γ-irradiation.
Accordingly, we propose a model for the role of KLF4 in response to γ-irradiation-induced DNA damage (Figure 9). In this model, p53 is upregulated following γ-irradiation. p53 then activates KLF4 (Zhang et al., 2000), p21WAF1/CIP1 (El-Deiry et al., 1993) and BAX (Miyashita and Reed, 1995). KLF4 has a dual function – to synergistically induce the expression of p21WAF1/CIP1 with p53, leading to cell cycle arrest (Zhang et al., 2000) and to suppress BAX expression, both directly and indirectly by inhibiting activation of BAX by p53, thus reducing apoptosis (this study). This model may explain the context-dependent nature by which KLF4 functions as either a tumor suppressor or an oncogene as reported by various groups (Foster et al., 1999, 2005; Zhao et al., 2004; Rowland et al., 2005; Wei et al., 2005; Rowland and Peeper, 2006). It may also explain the recent finding that inactivation of p21WAF1/CIP1 by oncogenic RASV12 neutralizes the cytostatic action of KLF4, converting the latter to a transforming protein owing to its antiapoptotic activity (Rowland et al., 2005)
The stably transfected human cancer cell line, RKO, that contained the inducible KLF4 construct, EcR-RKO/pAdLox-EGI-KLF4, was previously established (Chen et al., 2001). Cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. To induce KLF4 expression, cells were treated with 5 μM ponasterone A for the periods of time specified in the figure legends. Control cells were treated with vehicle (ethanol) alone. The HCT116 colon cancer cell lines transfected with shRNA against KLF4 (HCT116-KLF4/sh2-2) or vector-alone control (HCT116-KLF4/sh-c), were previously established (Yoon et al., 2005). HCT116 cells that are either WT, homozygous null for BAX or homozygous null for p21WAF1/CIP1 were also used. Cells were cultured in McCoy’s medium supplemented with 10% FBS and 1% penicillin/streptomycin. MEFs were isolated from E14 embryos from parents who were heterozygous for Klf4 (Katz et al., 2002). Isolated MEFs were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin. The cells were genotyped for being WT (+/+), heterozygous (+/−) or homozygous null (−/−) for Klf4 by PCR using primers and amplification conditions as described before (Katz et al., 2002).
γ-Irradiation of cultured cells was performed using a 137Cs γ-irradiator at 0.8Gy/min for 15 min, for a total of 12Gy (Yoon et al., 2003, 2005; Yoon and Yang, 2004). Cells were harvested at 72h after γ-irradiation for subsequent assays.
Cell cycle analysis was performed as previously described (Yoon et al., 2003). Cells were rinsed twice in phosphate-buffered saline (PBS), treated with trypsin and resuspended in their corresponding medium containing 10% FBS. Cells were then collected by centrifugation, washed with PBS, collected again by centrifugation, resuspended in 70 % ethanol and fixed at −20°C overnight. Cells were pelleted once again by centrifugation and resuspended in a solution containing 50μg/ml propidium iodide, 50μg/ml RNase A, 0.1 % Triton X-100 and 0.1 mm ethylene diaminetetra acetic acid (EDTA) at room temperature for 30 min. Flow cytometry was performed on a FACSCalibur cytometer (Becton Dickinson, Franklin Lakes, NJ, USA).
Cell protein extraction and Western blot analyses were performed using standard procedures. Protein samples were mixed with loading buffer (100 mM Tris-HCl, pH 6.8, 2% SDS, 100 mM dithiothreitol, 0.01% bromphenol blue and 10%) glycerol), heated at 100°C for 5 min, and loaded onto a SDS-polyacrylamide gel in electrophoresis buffer containing 25 mM Tris-HCl, pH 8.3, 250 mM glycine and 0.1% SDS. Protein was then transferred to polyvinylidene difluoride membranes using the Trans-Blot semidry system (Bio-Rad, Hercules, CA, USA). The membranes were immunoblotted with primary antibodies against KLF4, p53, p21WAF1/CIP1, BAX or actin (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). Following incubation with the secondary antibody (horseradish peroxidase-conjugated goat anti-rabbit IgG; or rabbit anti-goat IgG at 1:10 000 dilution; Santa Cruz Biotechnology), proteins were visualized with the SuperSignal West Pico chemiluminescent substrate kit (Pierce, Woburn, MA, USA).
Co-transfection experiments of COS-7 cells were performed with the following constructs: BAX677-pGL3-luciferase: nucleotide positions −870 to −193 relative to the translation initiation site of the BAX promoter linked to the pGL3-basic luciferase reporter (Promega, Madison, WI, USA); BAX100-pGL2-luciferase: nucleotide positions −530 to −431 of the BAX promoter linked to the pGL2-promoter vector (Promega); the expression construct containing WT p53, pC53-SN3 (Zhang et al., 2000); the expression construct containing Myc-tagged KLF4, pCMV-Myc-KLF4; and an internal control Renilla luciferase. The pCMV-Myc-KLF4 was generated by introducing a Myc tag into the amino-terminus of mouse KLF4 and inserting the fusion construct into the EcoRV–SalI site of pCMV-Script (Stratagene, La Jolla, CA, USA). The total amount of DNA in each co-transfection assay was adjusted to a final 3 μg per six-well plate using pCMV empty vector. The BAX promoter, BAX677, was amplified from human genomic DNA using primers 5′-TGGCTCAAGCCTGTAATCTCAGCA-3′ and 5′-ACTGTCCAATGAGCATCTCCCGAT-3′. The BAX100 fragment was synthesized. Luciferase activity was determined 1 day following transfection using the manufacturer’s recommendation (Invitrogen, Carlsbad, CA, USA). As a control, cells were transfected with empty pGL3-basic or pGL2-promoter luciferase vector alone or co-transfected with either KLF4 or p53 expression vectors. All firefly luciferase activity was standardized to the Renilla luciferase internal control.
ChIP assays were performed based on a published protocol (Yoon and Yang, 2004). COS-7 cells were transfected with BAX664-pGL3-luciferase and pC53-SN3, pCMV-Myc-KLF4 or both. Twenty-four hours after transfection, cells were collected and washed twice with PBS, followed by the addition of 1 % formaldehyde in PBS, and incubated for 15 min at room temperature. At the end of the incubation, the cross-linking was terminated by the addition of glycine to a final concentration of 125 mM. Cells were then washed twice with PBS and lysed with 1 ml of radioimmune precipitation assay buffer that contained 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 5 mM EDTA, 1% Nonidet P-40, 0.5% deoxycholate and protease inhibitors. Cell lysates were then sonicated, insoluble materials were removed by centrifugation, at 13 000 g for 10 min and the supernatant was transferred to a new tube. The supernatant was diluted 10-fold with dilution buffer (20 mM Tris-HCl, pH 8.1, 150 mM NaCl, 2 mM EDTA and 1% Triton X-100), and the diluted extracts were pre-cleared by incubation with 60μl of protein G-Sepharose beads (Sigma, St Louis, MO, USA) and sheared salmon sperm DNA. After centrifugation for 5 min at 7000 g, the supernatant was transferred to a fresh tube. Immunoprecipitation was performed by incubating the pre-cleared cell lysates with the appropriate antibodies overnight at 4°C with gentle shaking. Immune complexes were then precipitated with Ezview Red Protein G-Sepharose beads (Sigma) and sheared salmon sperm DNA. The beads were collected by centrifugation and washed twice sequentially with radioimmune precipitation assay wash buffer I (20 mM Tris-HCl, pH 8.1, 150 mM NaCl, 0.1% SDS, 1% Triton X-100 and 2 mM EDTA), wash buffer II (20 mM Tris-HCl, pH 8.1, 500 mM NaCl, 0.1% SDS, 1% Triton X-100 and 2 mM EDTA), wash buffer III (10 mM Tris-HCl, pH 8.1, 0.25 M LiCl, 1% Nonidet P-40, 1% deoxycholate and 1 mM EDTA) and wash buffer IV (10 mM Tris-HCl, pH 8.1 and 1 mM EDTA). The immunoprecipitates were eluted by adding a total of 500 μl of elution buffer (1% SDS, 0.1 M NaHCO3) to the washed beads followed by adding 20 μl of 5 M NaCl and the mix was incubated at 65°C overnight. After centrifugation for 5 min at 7000 g, the supernatant was collected and 10 μl of 0.5 M EDTA, 20 μl of 1 M Tris-HCl, pH 6.5 and 2 μl of 10 mg/ml proteinase K were added and incubated for 1 h at 56°C. DNA was recovered from the supernatant by phenol/chloroform extraction and ethanol precipitation. The recovered DNA was re-suspended in ddH2O and 3-μl DNA sample was then subjected to amplification using pGL3-basic vector specific primers (RVprimer3 as forward primer and GLprimer2 as reverse primer) supplied by the manufacturer (Promega). Interaction of KLF4 or p53 with the endogenous BAX and p21WAF1/CIP1 promoters in the inducible RKO cell system was also assessed using ChIP assay as described above. For the endogenous BAX promoter, 3-μl of the recovered DNA sample was subjected to seminested PCR amplification using primers encompassing nt −870 to −193 of the BAX promoter (forward – 5′-TGGCTCAAGCCTGTAATCTCAGCA-3′ and reverse – 5′-ACTGTCCAATGAGCATCTCCCGAT-3′) for the first PCR round, and nt −683 to −193 (forward – 5′-ATTCCAGACTGCAGTGAGCCATGA-3′ and reverse –5′-ACTGTCCAATGAGCATCTCCCGAT-3′) for the second round. For the p21WAF1/CIP1 promoter, the recovered DNA sample was subjected to seminested PCR using primers encompassing nt −283 to +553 of the proximal region of the p21WAF1/CIP1 promoter (forward – 5′-AAGCCAGATTTGTGGCTCACTTCG and reverse – 5′-AGACGAACTTACTCCACTCCGCTT-3′) for the first PCR round and nt − 199 to + 553 (forward – 5′-TGTGCTGCGTTCACAGGTGTTTCT-3′ and reverse – 5′-AGACGAACTTACTCCACTCCGCTT-3′) for the second round. As a control, PCR using primers specific for the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter (ActiveMotif) were conducted.
We thank B Vogelstein for providing the pC53-SN3 expression construct and the HCT116BAX−/− and p21WAF1/CIP1−/− cell lines. This work was in part supported by grants from the National Institutes of Health (DK52230, DK64399 and CA84197). VWY is the recipient of a Georgia Cancer Coalition Distinguished Cancer Clinician Scientist Award.