Foxp3 Expression in p53-dependent DNA Damage Responses

Abstract

The forkhead transcription factor, Foxp3, is thought to act as a master regulator that controls (suppresses) expression of the breast cancer oncogenes, SKP2 and HER-2/ErbB2. However, the mechanisms that regulate Foxp3 expression and thereby modulate tumor development remain largely unexplored. Here, we demonstrate that Foxp3 up-regulation requires p53 function, showing that Foxp3 expression is directly regulated by p53 upon DNA damage responses in human breast and colon carcinoma cells. Treatment with the genotoxic agents, doxorubicin or etoposide, induced Foxp3 expression in p53-positive carcinoma cells, but not in cells lacking p53 function. Furthermore, knock down of endogenous wild-type p53 using RNA interference abrogated Foxp3 induction by genotoxic agents, and exogenous expression of p53 in cells lacking p53 restored the responsiveness of Foxp3 to DNA-damaging stresses. In addition, Foxp3 knock down blunted the p53-mediated growth inhibitory response to DNA-damaging agents. These results suggest that induction of Foxp3 in the context of tumor suppression is regulated in a p53-dependent manner and implicate Foxp3 as a key determinant of cell fate in p53-dependent DNA damage responses.

Introduction

The tumor suppressor protein, p53, is a transcription factor that responds to genotoxic and oncogenic stress by inducing the expression of a number of genes that regulate a range of cellular responses, including cell cycle arrest, apoptosis, and senescence (for review, see Refs. 1, 2). The expression of p53 target genes is directly attributable to the transcriptional activity of p53. This is exemplified by p21 and 14-3-3σ, whose expression is driven by p53 in response to genotoxic stress, leading to cell cycle arrest (3, 4). Additional p53-targeted genes related to cell cycle control, cell division, apoptosis, and other responses have also been identified (1). Thus, functional p53 is important in regulating a host of cellular responses. However, p53 is frequently mutated in human tumors, resulting in disruption of normal p53 function (5). Importantly in this context, p53 functional status is a key determinant of the chemosensitivity of human tumor-derived cells to genotoxic agents used in cancer therapy (6, 7).

Foxp3, a member of the forkhead family of transcription factors, plays a key role in the function of regulatory T cell (8, 9). Ectopic expression of Foxp3 has been shown to convert normal T cells into suppressor T cells (10). Foxp3-positive T cells can suppress various immune cells, such as CD4, CD8, B cells, NK, and NKT cells (11–16). Until recently, Foxp3 expression was thought to be restricted to the T cell lineage (17, 18). However, using diverse approaches, studies have since shown that the Foxp3 transcription factor is expressed in breast cancer cells, melanoma cells, virally transformed B cells, and cell lines derived from a variety of solid tumors (19–22). One research group recently reported that Foxp3 is expressed in normal breast epithelial cells, but is down-regulated in mammary cancer tissues (22, 23). In addition, Foxp3 has been shown to act as a novel transcriptional repressor of the oncogene, Skp2 (22), and to inhibit the HER-2/ErbB2 oncogene (23), which plays a key role in breast cancer progression.

Given these findings, there is great interest in the role of the Foxp3 gene in breast cancer development. Although important downstream targets of Foxp3 have been identified, the mechanisms that regulate Foxp3 expression have remained unclear. Interestingly, the tumor suppressor protein, p53, also down-regulates both ErbB2 and Skp2 genes (24, 25), suggesting a possible linkage between p53 and Foxp3 expression in tumor development. Here, using human breast and colon carcinoma cells, we establish a relationship between p53 and Foxp3, showing that Foxp3 expression is induced in a p53-dependent manner by DNA damage-inducing stimuli and may be required for certain p53-mediated cellular responses.

EXPERIMENTAL PROCEDURES

Cell Culture, Plasmids, and Transfection

Human breast cancer cell lines, MCF7, MDA-MB231, MDA-MB157, and MDA-MB453, as well as human colon cancer cell lines, HCT116-p53wt and HCT116-p53^−/−, were maintained in Dulbecco's modified Eagle's medium (Invitrogen). The Foxp3 plasmid was obtained by subcloning Foxp3 cDNA, kindly provided by Dr. H. T. Chung (University of Ulsan). Recombinant adenoviruses encoding wild-type p53 were purchased from (Vector BioLaboratories). Lipofectamine^TM RNAiMAX (Invitrogen) was used for transient transfections.

Conventional, Real-time Reverse Transcription (RT)³-PCR Analyses, and Northern Blot Analysis

For conventional RT-PCR, total RNA was extracted from human breast cancer cell lines using TRIzol, and cDNA was synthesized using avian myeloblastosis virus reverse transcriptase, with oligo(dT₁₅) as a primer. Targets were amplified from cDNA using the following primers: Foxp3 (sense), 5′-CAG CAC ATT CCC AGA GTT CCT C-3′, and (antisense), 5′-GCG TGT GAA CCA GTG GTA GAT C-3′; β-actin (sense), 5′-AAA GGG TGT AAC GCA ACT AA-3′, and (antisense) 5′-GGA CCT GAC TGA CTA CCT CA-3′. The conventional PCR conditions for Foxp3 were 38 cycles of 30 s at 94 °C, 1 min 10 s at 58 °C, and 1 min at 72 °C; conditions for β-actin were 35 cycles of 1 min at 95 °C, 1 min at 50.5 °C and 1 min at 72 °C. PCR products were electrophoresed, and the density of each band was analyzed. Real-time RT-PCR was performed using a Rotor-Gene 3000 real-time cycler (Corbett Research, Sydney, Australia), with 40 cycles of 10 s at 94 °C (denaturation), 15 s at 58 °C (annealing), and 20 s at 72 °C (extension). The amplification efficiency of the target was equal to that of the endogenous reference β-actin, which was used as the endogenous reference in the comparative cycle threshold method. For Northern blot analysis, total RNA was prepared by NorthernMax-Gly kit (Ambion) and blotted onto Hybond-N nylon membrane (Amersham Biosciences Pharmacia). The following primer sequences were used: 5′-CAGCTGCCTACAGTGCCCCTAG-3′ and 5′-CATTTGCCAGCAGTGGGTAG-3′. It was labeled with [³²P]dCTP by random hexamer incorporation (MegaPrime kit; Amersham Biosciences) and used to probe the filters.

RNA Interference and Immunofluorescence Analyses

Human breast or colon cancer cells were transiently transfected with scrambled siRNA (Samchully Pharmaceutical Co.), p53-siRNA (5′-GAC TCC AGT GGT AAT CTAC-3′) or Foxp3-siRNA (5′-GCA GCG GAC ACT CAA TGAG-3′) using Lipofectamine^TM RNAiMAX (Invitrogen). Immunofluorescence staining was performed as previously described (26). Briefly, for phospho-p53 or Foxp3 staining, cells were fixed with phosphate-buffered saline and 3.7% paraformaldehyde at room temperature for 15 min and then permeabilized with phosphate-buffered saline, 0.1% Triton X-100, 0.1 mg/ml RNase A at 37 °C for 30 min. The permeabilized cells were treated with 0.5 mg/ml NaBH₄ to reduce autofluorescence. The cells were then immunostained by incubating with primary antibodies against phospho-p53 (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA) and Foxp3 (1:100 dilution; eBioscience) and then with fluorescein isothiocyanate- and Texas red-conjugated secondary antibodies (Molecular Probes, Eugene, OR) and DAPI. After washing three times with phosphate-buffered saline, coverslips were mounted on microscope slides using ProLong antifade mounting reagent (Molecular Probes), and analyzed using an Olympus DP50 digital camera (Olympus Optical, Japan).

Analysis of Cell Cycle and Cell Proliferation

For each DNA content analysis, 1 × 10⁶ cells were harvested by trypsinization and fixed by rapid submersion in 1 ml of cold 70% ethanol. After fixation at −20 °C for at least 1 h, cells were pelleted, resuspended in 1 ml of staining solution (50 μg/ml propidium iodide, 50 μg/ml RNase, 0.1% Triton X-100 in citrate buffer, pH 7.8), and then washed with phosphate-buffered saline. Fluorescence-stained cells were transferred to polystyrene tubes with cell-strainer caps (Falcon), sorted using a fluorescence-activated cell sorter (BD FACSCalibur^TM), and analyzed with Cell Quest 3.2 (Becton Dickinson) software (27). For cell proliferation analyses, cells were plated in 60-mm culture dishes and then treated with 0.3 μg/ml doxorubicin or 20 μm etoposide. Cell proliferation was analyzed by counting cells at the indicated times.

Western Blot Analysis

Cell lysates were prepared with radioimmune precipitation assay lysis buffer (50 mm Tris-HCl, pH 7.5, 50 mm NaCl, 1 μm EGTA, 1% Triton X-100) containing a protease inhibitor mixture. Protein concentrations in extracts were determined using a Bradford assay, and 30 μg of total cell protein/sample was separated by SDS-PAGE and then transferred to a PolyScreen membrane (PerkinElmer Life Sciences). Membranes were blocked with 5% nonfat dry milk in TBST buffer (20 mm Tris-HCl, pH 7.4, 150 mm NaCl, 0.1% Tween 20) and probed with one of the following antibodies: anti-p53, anti-Noxa, anti-p21, anti-Bax (Santa Cruz Biotechnology), anti-phospho-p53, anti-Puma (Cell Signaling Technology, Beverly, MA), anti-Foxp3 (eBioscience), or anti-β-actin (Sigma). Horseradish peroxidase-conjugated goat anti-mouse and goat anti-rabbit secondary antibodies in conjunction with enhanced chemiluminescence detection (Amersham Biosciences) were used to develop the Western blots.

Chromatin Immunoprecipitation (ChIP) and Luciferase Activity

ChIP was performed as described in Ref. 30, using 25 μg of chromatin mixed with 3 μl of anti-p53, anti-RNA polymerase II (as a positive control) (31), or anti-IgG (as a negative control). Foxp3 primer sequences are described in Table 1. For luciferase assay, Foxp3 promoter-luc plasmid that includes Foxp3 promoter region (−1657 to −1 from transcription start (TSS)) and Foxp3 intron-luc plasmid that includes a part of Foxp3 intron region (+1 to +176 from TSS) were constructed by subcloning into the pGL3-Basic vector (luc+).

TABLE 1.

Primers for ChIP assay

Site	Sequence (5′ → 3′)	Binding region (including)
−1572 to −1355	TGGGATTTGGAGAGTCCTTG
	GGAGTGGTTGTTGGACGACT
−1353 to −1149	GATCTCCCTGCCATCACATC
	CCCTTGAGCTCCATCTCATC
−1146 to −894	TCAGCCTTTGGGGTCTTTTA
	CATGGTGAGGATTGAACATGA
−956 to −756	CCCTGGCTCCCAGAATCTAC
	TTCCTGAAACAAGAGGGTCAG
−758 to −601	TCCCCATTGCTTGAACTACC
	GGGATTCTCCGACTCTTCAA
−511 to −320	GGGCTCATGAGAAACCACAGTTTCCCATCCACACATAGAGC	NFAT (−474 to −468, −389 to −383, −335 to −329), AP-1 (−462 to −455, −330 to −324)
−288 to −73	AAGCCGCAGACCTCTCTCTT
	GTGGTGAGGGGAAGAAATCA
−17 to +175	CCTTACCTGGCTGGAATCAC
	ACGTGACAGTTTCCCACAAG

RESULTS

Foxp3 Induction by Genotoxic Agents Is Correlated with p53 Status in Various Cellular Contexts

Recent reports indicate that Foxp3 is expressed in normal breast epithelial cells but is down-regulated in mammary cancer tissues (22, 23). This observation indicates that the level of Foxp3 expression is correlated with tumor development and suggests that Foxp3 may act as a tumor suppressor to modulate cellular responses to genotoxic stress. To evaluate changes in Foxp3 expression in response to genotoxic stress, we first measured Foxp3 mRNA levels in four human breast carcinoma cell lines (MCF7, MDA-MB231, MDA-MB157, and MDA-MB453) following treatment with the DNA-damaging agent, doxorubicin, by conventional RT-PCR. Foxp3 expression was significantly increased in MCF7 cells, but not in the others (Fig. 1A). To confirm these results using a more quantitative approach, we also assessed changes in Foxp3 expression after exposure to doxorubicin using real-time RT-PCR (Fig. 1B). Consistent with the conventional RT-PCR results, Northern blot analysis and real-time RT-PCR revealed that the increase in Foxp3 expression was unique to MCF7 cells (Fig. 1B, i). Interestingly, of the four breast carcinoma cell lines tested, only MCF7 cells express wild-type p53; all the others lack functional p53. Furthermore, the pattern of Foxp3 expression analyzed by Northern blot analysis was paralleled with that of (real-time) RT-PCR (Fig. 1C). These results are consistent with the idea that doxorubicin-induced changes in Foxp3 expression are related to p53 functional status.

FIGURE 1.

Foxp3 expression is induced in response to doxorubicin. A, breast cancer cell lines were treated with 0.3 μg/ml doxorubicin, and total RNA was extracted 24 h later. cDNA was amplified as described under “Experimental Procedures.” RT-PCR products were electrophoresed, and the density of each band was analyzed. B, total RNA was extracted from cells treated with 0.3 μg/ml doxorubicin for 24 h. Real-time RT-PCR was performed using a Rotor-Gene 3000 real-time thermocycler. Values shown are the means ± S.E. values of three independent experiments. C, each cell line was treated with or without 0.3 μg/ml doxorubicin, and then total RNA were prepared. A Northern blot analysis was performed as under “Experimental Procedures.” GAPDH, glyceraldehyde-3-phosphate dehydrogenase. D and E, lysates from MCF7 (D) or MDA-MB453 (E) cells treated with 0.3 μg/ml doxorubicin were prepared at the indicated times and analyzed by Western blotting using anti-p53, anti-phospho-p53, and anti-Foxp3 antibodies. β-Actin levels were used as loading controls. F, MCF7 cells were treated with either 5 μm cisplatin or 10 nm paclitaxel (Taxol), and cell lysates were prepared at the indicated times for Western blot analysis. G, the HCT116-p53wt human colon cancer cell line and its p53^−/− derivative were treated with 20 μm etoposide, and then cell lysates were prepared at the indicated times. Changes in p53, p53 phosphorylation status, and Foxp3 were detected by Western blotting using antibodies specific for each protein.

To confirm that Foxp3 induction in human breast carcinoma cells is correlated with functional p53, we examined Foxp3 induction after exposure to doxorubicin at the protein level, using MCF7 cells, which have functional p53 (Fig. 1D), and MDA-MB453, which lack p53 function (Fig. 1E). Doxorubicin induced a remarkable increase in Foxp3 protein in MCF7 cells that was associated with the phosphorylation of p53, but had no effect in MDA-MB453 cells (Fig. 1, D and E). We also confirmed that Skp2 suppression was dependent on Foxp3 induction (data not shown). To extend these findings, we evaluated the relationship between Foxp3 induction and functional p53 in MCF7 cells using two additional genotoxic agents: cisplatin, which functions primarily through p53-dependent mechanisms, and paclitaxel (Taxol), which may act through p53-independent pathways. We found that Foxp3 induction and p53 phosphorylation were dramatically increased after treatment with cisplatin, although neither change was observed in response to paclitaxel (Fig. 1F). These observations are consistent with the mechanism of action of these drugs and provide further support for the supposition that Foxp3 induction is dependent on p53 function. We further explored the relationship between Foxp3 and p53 using two different derivatives of the HCT116 human colon cancer cell line, one that expresses wild-type p53 (HCT116-p53wt) and one that is p53-null (HCT116-p53^−/−). Consistent with the above results, after exposure to etoposide, which can act through p53-dependent pathways, Foxp3 was induced in parallel with phosphorylation of p53 in HCT116-p53wt cells, but not in HCT116-p53^−/− cells (Fig. 1G). Thus, the induction of Foxp3 protein expression by DNA-damaging stresses is likely p53-dependent.

p53 Functional Status Determines the Response of Foxp3 to DNA-damaging Agents

To firmly establish that Foxp3 induction by genotoxic stresses is directly attributable to p53 function, we first examined the effect of silencing endogenous wild-type p53 in MCF7 and HCT116-p53wt cell lines using siRNAs. After transiently transfecting both cell lines with p53-siRNA or scrambled siRNA constructs, we treated cells with doxorubicin or etoposide and evaluated changes in p53 and Foxp3 protein by Western blot analysis. Endogenous Foxp3 was not induced in p53-siRNA-treated cells after exposure to chemotherapeutic agents, whereas its induction was unaffected in cells treated with scrambled siRNA (Fig. 2, A and B). In addition, p53 up-regulation and phosphorylation were not induced by DNA-damaging agents in cells treated with p53-siRNA constructs, indicating that functional p53 is necessary for the induction of Foxp3 in response to genotoxic stresses. We also assessed the effect of siRNA-mediated knock down of endogenous wild-type p53 on the cellular response to genotoxic stress using an immunocytochemical approach. As shown in Fig. 2C, in cells co-stained for phosphor-p53 (red) and Foxp3 (green) in the nucleus, transfection of p53-siRNA suppressed the level of nuclear Foxp3 protein after doxorubicin treatment in MCF7 cells, whereas scrambled siRNA had no effect. Similar results were obtained in HCT116 cells, where Foxp3 induction was not observed following etoposide treatment in cells transfected with p53-siRNA (supplemental Fig. 2A). Consistent with these results, the absence of nuclear Foxp3 induction in cells transfected with p53-siRNA was paralleled by a suppression of p53 phosphorylation (Fig. 2C and supplemental Fig. 2A).

FIGURE 2.

Foxp3 response to genotoxic stress is determined by the functional status of p53. A and B, MCF7 (A) or HCT116-p53wt (B) cells were transiently transfected with a scrambled siRNA (control) or p53-siRNA 24 h before treating with 0.3 μg/ml doxorubicin or 20 μm etoposide for 48 h. Whole cell extracts of cells were prepared and analyzed by Western blotting as in Fig. 1. C, MCF7 cells were prepared as in Fig. A and immunostained for phospho-p53 and Foxp3 and stained with DAPI (blue), as described under “Experimental Procedures.” D and E, MDA-MB453 (D) or HCT116-p53^−/− (E) cells were infected with Ad-p53 and then treated with 0.3 μg/ml doxorubicin or 20 μm etoposide. Cell lysates were prepared 48 h later, and changes in proteins were analyzed by Western blotting as in Fig. 1. β-Actin levels were used as loading controls. F, HCT116-p53^−/− cells were prepared as in E and immunostained as described under “Experimental Procedures.”

We also employed the converse approach to establish a direct link between p53 function and Foxp3 induction, introducing wild-type p53 into p53-deficient cells. To this end, we infected MDA-MB453 cells, which lack p53 function, and HCT116-p53^−/− cells with adenoviral p53 (Ad-p53) to exogenously express p53 (Fig. 2, D and E). In cells infected with Ad-p53, treatment with the chemotherapeutic agents doxorubicin or etoposide induced Foxp3 expression in association with p53 phosphorylation. In contrast, Foxp3 was not induced in cells infected with control adenoviral constructs, indicating that exogenous expression of p53 in cells that have no p53 function restored Foxp3 induction after DNA damage. These results are confirmed by immunocytochemistry experiments, which show that the induction of Foxp3 by doxorubicin or etoposide in cells infected with Ad-p53 was paralleled by phosphorylation of p53 (Fig. 2F and supplemental Fig. 2B). In addition, we retried to confirm whether functional p53 is necessary for Foxp3 induction in response to a DNA-damaging agent in four cancer cells using an immunocytochemical approach (supplemental Fig. 2C). Collectively, these results indicate that functional p53 is necessary for Foxp3 induction in response to genotoxic stress.

Foxp3 Protein Is Induced by p53 Stabilization/Accumulation through the Inhibition of Mdm2

As shown on above results, the expression of Foxp3 was induced by p53 induction/phosphorylation after DNA damage. Therefore, we expected that p53 stabilization/accumulation could also induce the expression of Foxp3 in cells that have p53 function would result in an increase in Foxp3 protein. To this test, we used nutlin-3A, which can induce p53 stabilization/accumulation through inhibition of Mdm2 (32). The levels of Foxp3 protein were significantly increased by nutlin-3A in 2 cancer cell lines, MCF7 (Fig. 3A) and HCT116 cells (Fig. 3B), which have p53 function, indicating that p53 stabilization/accumulation can induce the expression of Foxp3. In addition, we examined the combined effects of nutlin-3A and doxorubicin or etoposide on Foxp3 induction (supplemental Fig. 3, A and B). MCF7 or HCT116 cells were treated with nutlin-3A alone, doxorubicin or etoposide alone, or combination of nutlin-3A and doxorubicin or etoposide. Although there are no synergistic effects on Foxp3 induction after treatment with combination in both cell lines, co-treatment with nutlin-3A and doxorubicin or etoposide was also observed the increased levels of Foxp3 protein. Therefore, these results clearly indicate that Foxp3 is regulated by DNA damage in a p53-dependent manner and by Nutlin-3, an inhibitor of Mdm2.

FIGURE 3.

P53 stabilization/accumulation by nutlin-3A results in Foxp3 induction. MCF7 (A) or HCT116-p53wt cells (B) were treated with 5 μm or 10 μm nutlin-3A. Cell lysates were prepared at the indicated times and then subjected to Western blot analysis.

Foxp3 Expression Is Directly Regulated by p53

From the above data, we found that Foxp3 expression is regulated by DNA damage in p53-dependent manner. So we next explore whether Foxp3 is directly regulated by p53. First, we constructed the Foxp3 promoter-luc plasmid and Foxp3 intron-luc plasmid by subcloning Foxp3 promoter region (−1657 to −1 from the TSS) or Foxp3 intron region (+1 to +176 from the TSS) into the pGL3-Basic vector (luc+) to luciferase assay. Cells transfected with Foxp3 promoter-luc, Foxp3 intron-luc plasmid, or control vector were treated with doxorubicin or were not. We found that luciferase activity in cells that express Foxp3 promoter-luc was significantly increased in response to doxorubicin, but not in others (Fig. 4A).

FIGURE 4.

Foxp3 expression is directly regulated by p53 in response to DNA damage. A, MCF7 cells were tranfected with Foxp3 promoter-luc plasmid, Foxp3 intron-luc plasmid, or control vector. Cell lysates were prepared, and then luciferase activity was analyzed by a luminometer. B, schematic diagram of Foxp3 primers for ChIP assay is shown. Foxp3 primer sequences are described in Table 1. C, MCF7 cells were treated with or without 0.3 μg/ml doxorubicin and then cultured for 24 h. Chromatin was prepared to examine the binding of Foxp3 and p53 by ChIP using an anti-p53 antibody. RNA polymerase II was a positive control (28, 29), and anti-IgG was a negative control.

To evaluate further the effect of p53 on the transactivation of Foxp3, we measured DNA-binding activity of p53 on the promoter of Foxp3 in response to DNA damage by using a ChIP assay. To this study, we designed eight primers against Foxp3 promoter region including a part of its intron (Fig. 4B and Table 1). Surprisingly, p53 bound to the distal region of Foxp3 promoter upon DNA damage treatment. The binding site of p53 on Foxp3 promoter region was −1572 to −1355 from the TSS (Fig. 4C). In addition, the binding of p53 to p21 promoter was used as a positive control (supplemental Fig. 4). Previously, many studies have shown that p53 consensus sites exist in the promoter regions of p53-binding partners, such as p21, Bax, and Rb (supplemental Table 1). Thus, we analyzed whether a p53 consensus binding site could be found in the region −1572 to −1355 bp from the TSS. Interestingly, we found one motif (−1408 to −1399 bp from the TSS) with strong homology to the p53-binding consensus sequence (supplemental Fig. 5 and supplemental Table 1). This motif may be necessary for p53 binding. Thus, Foxp3 expression is directly regulated by p53. These results suggest that Foxp3 expression is directly regulated by p53 in response to DNA damage.

Foxp3 Is Necessary for p53-mediated Inhibition of Cell Growth, but not for Cell Death

Our results indicate that Foxp3 expression after exposure to stress is strictly dependent on p53, suggesting that Foxp3 might play a role in mediating p53-dependent cellular responses to genotoxic stress. To address this possibility, we analyzed the effects of altered Foxp3 expression on p53-dependent cellular responses such as cell growth inhibition or apoptotic cell death. First, we transfected MCF7 or HCT116-p53wt cells with Foxp3-siRNA or scrambled siRNA and then treated cells with doxorubicin or etoposide (Fig. 5A). Although cell cycle progression was clearly inhibited in both cells transfected with scrambled siRNA following doxorubicin or etoposide treatment, the growth inhibitory response was weak and sub-G₁ portion increased in Foxp3-siRNA-transfected cells (Fig. 5A), implying that Foxp3 may be involved in p53-mediated growth inhibition, but not apoptotic cell death. To evaluate further the role of Foxp3 on cellular responses mediated by p53, we first examined the time course of Foxp3 expression until commitment point to apoptosis induction (Fig. 5B). Cells were exposed to 0.3 μg/ml doxorubicin for 5 days. The levels of Foxp3 protein were clearly increased up to 2 days, but not after 2 days (Fig. 5B, i). Surprisingly, cell death was gradually increased from 3 days after exposure to doxorubicin (data not shown). Consistently, annexin V-positive cells were also increased from 3 days after doxorubicin treatment (Fig. 5B, ii). This seems to be correlated with the ability of Foxp3 to inhibit cell growth, but not cell death. Moreover, the levels of Foxp3 protein in cells exposed to high doses (1.5, 2 μg/ml) of doxorubicin, which can induce cell death within 2 days (Fig. 5C, i), were not almost increased (Fig. 5C, ii). Thus, the role of Foxp3 may be related to inhibition of cell proliferation, but not cell death. In addition, we examined the effect of overexpressing Foxp3 on p53-mediated growth inhibition, transfecting MCF7 cells with a Foxp3 expression plasmid or empty vector and then exposing them to doxorubicin. Surprisingly, the Foxp3-transfected cells were more sensitive to doxorubicin than cells transfected with empty vector (supplemental Fig. 6). Furthermore, growth inhibition in untreated Foxp3-transfected cells was comparable with that in empty vector-transfected cells treated with doxorubicin. Virtually identical results were obtained in experiments using the HCT116 cells and the genotoxic agent, etoposide (data not shown), implying that Foxp3 is important in p53-mediated inhibition of cell growth.

FIGURE 5.

Foxp3 is necessary for p53-induced inhibition of cell growth. A, cells transfected with scramble-siRNA (control) or Foxp3-siRNA and then treated with 0.3 μg/ml doxorubicin or 20 μm etoposide were subjected to flow cytometric analysis after staining with propidium iodide. B, cells were exposed to 0.3 μg/ml doxorubicin for 5 days. Changes in p53 phosphorylation status and Foxp3 level were detected by Western blotting using antibodies specific for either protein. B, i, annexin V-positive cells were analyzed at the indicated times after doxorubicin treatment (ii). C, cells were prepared for annexin V staining as shown in B (i), and the levels of Foxp3 protein in cells exposed to high doses (1.5, 2 μg/ml) of doxorubicin were analyzed by Western blotting using an anti-Foxp3 antibody (ii). D, cells were exposed to 1.5 μg/ml doxorubicin for the indicated times. Changes in p53 phosphorylation status, Foxp3, p21, Bax, Puma, and Noxa were detected by Western blotting using antibodies specific for each protein. E, cells transfected with Foxp3-siRNA or scrambled siRNA were treated with 0.3 μg/ml doxorubicin. F, cells infected with Ad-p53 were transfected with Foxp3-siRNA or scrambled siRNA. Twenty-four h after transfection, the indicated proteins were analyzed by Western blotting. G, Foxp3-expressing cells were transfected with p53-siRNA or scrambled siRNA. Twenty-four h after transfection, the indicated proteins were analyzed by Western blotting using appropriate antibodies.

We next examined the effect of Foxp3 on proteins that regulate cellular responses mediated by p53. First, we observed that the expression of p53 target genes was related to growth arrest and cell death of MCF7 cells in response to stress. Interestingly, Foxp3 was induced early after doxorubicin treatment, whereas transcription of proapoptotic p53 target genes such as Bax, Puma, and Noxa was elevated only after Foxp3 and p21 expression had ceased (Fig. 5D). This seemed to be correlated with the ability of Foxp3 to inhibit cell growth, but not cell death. Thus, we further studied the effect of Foxp3 silencing using siRNA. Treatment with doxorubicin preferentially induced expression of the proapoptotic p53 targets Puma and Noxa in Foxp3-siRNA-expressing MCF7 cells, but not in control cells (Fig. 5E). Also, infection of cells with adenoviral p53 followed by transfection with Foxp3-siRNA resulted in preferential induction of Noxa (Fig. 5F), implying that Foxp3 was necessary for cell growth inhibition mediated by p53. In addition, Foxp3 and p21 expression may be correlated. Foxp3 synthesis increased in parallel with p21 expression in MCF7 cells transfected with a Foxp3 expression vector (Fig. 5G). Thus, Foxp3 plays a preferential role in p53-mediated inhibition of cell growth.

DISCUSSION

Using a wide variety of approaches, recent studies have provided evidence that Foxp3 plays an important role in tumor development in addition to its well established function in the immune system (19–22). Both HER-2/ErbB2 and Skp2 oncogenes, which contribute to breast cancer progression, are repressed by Foxp3 (22, 23) and can be up-regulated by its inactivation. A critical unresolved issue regarding this process is the nature of the upstream regulatory events that lead to Foxp3 expression. In this study, we present evidence that p53 function is necessary and sufficient for Foxp3 expression in response to genotoxic stress.

It has recently been shown that the tumor suppressor protein, p53, suppresses HER-2/ErbB2 and Skp2 oncogenes (24, 25). On the basis of these observations, we hypothesized that Foxp3 expression is related to the p53 response to genotoxic stress. First, we showed that Foxp3 expression is induced in response to doxorubicin, which can induce p53-dependent DNA damage responses (Fig. 1). Our analysis of the response of several human breast carcinoma cell lines to doxorubicin supports an important role for p53 in Foxp3 induction. In addition, Foxp3 expression was not induced by paclitaxel, which can induce p53-independent cellular responses. Second, we showed that Foxp3 was induced in a colon carcinoma cell line containing wild-type p53 after exposure to etoposide. Importantly, siRNA-mediated knock down of p53 virtually abrogated Foxp3 induction, whereas exogenous expression of p53 in cells lacking p53 function dramatically enhanced Foxp3 induction (Fig. 2). More importantly, treatment of nutlin-3A that induces p53 stabilization/accumulation through inhibition of Mdm2 (33) induced Foxp3 in two cancer cell lines, HCT116 and MCF7 (Fig. 3). In addition, immunostaining analyses revealed that p53 was co-localized with Foxp3 in the nucleus of cells exposed to stress. These findings are significant because they strongly suggest that p53 is a key regulator of Foxp3 expression in response to DNA-damaging stresses. The details of this mechanism, however, were not elucidated; thus, further study will be required.

p53 function has been recognized as an important component of the cellular response to DNA-damaging stresses. Here, we have shown that siRNA-mediated Foxp3 knock down rescued cells from inhibition of cell growth induced by phosphorylated p53 after exposure to genotoxic stress, but resulted in sensitivity to apoptosis (Fig. 5). Foxp3 induction was correlated with p53-mediated cell cycle arrest and was also dependent on p53 function, consistent with the recent finding that Foxp3 suppresses the growth of the tumor cell line, MCF7 (23). However, it seems to be correlated with the ability of Foxp3 to inhibit cell growth, but not cell death (Fig. 5). Foxp3 expression in a cellular condition that induces cell growth arrest was induced, but not in a cell death condition. In addition, Foxp3 was induced at early times after doxorubicin treatment, whereas proapoptotic p53 target genes such as Bax, Puma, and Noxa were induced after expiration of Foxp3 expression. Thus, Foxp3 preferentially play a key role in cell cycle arrest mediated by p53.

Furthermore, we found that Foxp3 protein was highly expressed in 16 of 21 samples of normal human breast tissue, but was only weakly expressed in breast cancer tissues (data not shown). We have also observed a significant relationship between Foxp3 and p53 expression in normal breast tissue samples (supplemental Fig. 1). More importantly, Foxp3 expression is directly regulated by p53 upon DNA damage treatment (Fig. 4). Promoter analyses revealed that p53 could bind directly to the distal region in promoter of Foxp3 gene (supplemental Fig. 5). This motif that bind to p53 has a strong homology to the p53-binding consensus (supplemental Table 1). So we need to study further for p53-binding motif in promoter of Foxp3 gene. These findings, which collectively indicate an important role for Foxp3 in p53-dependent DNA-damage responses, suggest that therapeutic strategies designed to regulate Foxp3 expression may represent attractive cancer treatment options.