An HDAC1-binding domain within FATS bridges p21 turnover to radiation-induced tumorigenesis

Article metrics


There is a gap between the initial formation of cells carrying radiation-induced genetic damage and their contribution to cancer development. Herein, we reveal a previously uncharacterized gene FATS through a genome-wide approach and demonstrate its essential role in regulating the abundance of p21 in surveillance of genome integrity. A large exon coding the NH2-terminal domain of FATS, deleted in spontaneous mouse lymphomas, is much more frequently deleted in radiation-induced mouse lymphomas. Its human counterpart is a fragile site gene at a previously identified loss of heterozygosity site. FATS is essential for maintaining steady-state level of p21 protein and sustaining DNA damage checkpoint. Furthermore, the NH2-terminal FATS physically interacts with histone deacetylase 1 (HDAC1) to enhance the acetylation of endogenous p21, leading to the stabilization of p21. Our results reveal a molecular linkage between p21 abundance and radiation-induced carcinogenesis.


Ionizing radiation (IR) is a well-known complete carcinogen that is able to initiate and promote neoplastic progression, resulting from its induction of a broad spectrum of DNA lesions including damage to nucleotide bases, crosslinking, and DNA single-strand breaks and DNA double-strand breaks (DSBs) that impairs chromosomal stability (van Gent et al., 2001). The latent period between radiotherapy and the appearance of second primary tumor or recurrence ranges from a few years to several decades (Holm, 1990), which is particularly the case in patients treated for childhood Hodgkin's disease (Magrath, 1997; Schellong, 1998). With the increasing use of diagnostic radiation for various conditions, concern is growing that the risk/benefit ratio for such exposure needs to be weighed, and cancer is considered the major long-term contributor to health risk at doses below causing acute tissue injury (Brenner et al., 2003). Recently, Genome-wide analysis of chromosomal aberrations has verified the similarities and differences between spontaneous and IR-induced tumorigenesis. In particular, radiation-induced tumors exhibit a high frequency of localized deletion events (Mao et al., 2005). However, mechanistic understanding of radiation-induced carcinogenesis remains a major challenge.

Common fragile sites (CFSs) are site-specific loci that are found on most human chromosomes and can be induced in all individuals of the population under replication stress. Aphidicolin, an inhibitor of DNA polymerase that can cause DNA lesions including DSBs, is an efficient inducer of CFSs. CFSs are evolutionarily conserved regions that are AT-rich and late replicating (Mangelsdorf et al., 2000; Glover et al., 2005; Schwartz et al., 2006; Zhang and Freudenreich, 2007). In general, CFSs are now being implicated as regions of high genomic instability associated with cancer (Buttel et al., 2004; Durkin and Glover, 2007). However, much less is known about their molecular nature and functional characterization of CFS genes in DNA damage signaling.

Defects in cell-cycle checkpoints result in genomic instability and the initiation of tumorigenesis (van Gent et al., 2001; Maser and DePinho, 2002; Rouse and Jackson, 2002; Kastan and Bartek, 2004). In eukaryotic DNA damage signaling pathways, tumor suppressor p53 and its transcriptional target CDKN1A (p21) have an essential role in monitoring cell-cycle checkpoints (Brugarolas et al., 1995; Deng et al., 1995; Bunz et al., 1998). The cell-cycle inhibitor p21 is an unstable protein that is subjected to proteasome-dependent degradation in both ubiquitin-dependent (Bloom et al., 2003; Bendiennat et al., 2003; Coulombe et al., 2004) and ubiquitin-independent manners (Sheaff et al., 2000; Li et al., 2007; Chen et al., 2007). Unbound p21 is directly degraded by proteasome independently of ubiquitination, as the endogenous p21 is fully acetylated at its amino terminus (Chen et al., 2004) and is therefore not a substrate for N-end rule ubiquitination (Varshavsky et al., 2000). Ubiquitin-independent turnover of p21 results from the interaction of its C terminus with the C8 α-subunit of 20S proteasome (Touitou et al., 2001) or proteasome activator REGγ (Chen et al., 2007). In addition, the inhibition of histone deacetylases (HDACs) has been known to selectively induce p21 (Johnstone, 2002; Dokmanovic et al., 2007), although the impact of acetylation on p21 turnover remains uncertain. Precise control of p21 abundance is required for its role in cell-cycle regulation, thus further elucidating the regulatory mechanism of p21 stabilization is an important priority toward a greater understanding of p21 deregulation in human cancer (Abbas and Dutta, 2009).

Here, we report the identification of a previously uncharacterized and evolutionarily conserved gene, FATS (for Fragile-site Associated Tumor Suppressor), at a frequently deleted region in IR-induced tumors. FATS functions as a regulator of p21 abundance. Furthermore, the NH2-terminal domain of FATS binds to HDAC1 and suppresses its binding to p21, leading to the enhanced acetylation of endogenous p21. We also provide new evidence for a tight functional link between acetylation and stability of p21. This study thus brings new insights into our understanding of radiation-induced carcinogenesis.


FATS is a key component of radiation-induced tumorigenesis

IR-induced mouse thymic lymphoma is one of the classic models in radiation carcinogenesis, and p53 insufficiency accelerates IR-induced tumor progression (Kemp et al., 1994). Our previous results demonstrated the markedly difference of genome-wide aberations between IR-induced and spontaneous mouse tumors (Mao et al., 2005), allowing us to systematically analyze the deletion events specific to IR-induced tumorigenesis. Using microarrays containing over 19 200 bacterial artificial chromosome (BAC) clones with insert size in the range of 170–210 kb and maximized coverage of mouse genome (Cai et al., 2002), we performed microarray-based comparative genomic hybridization analysis and identified a deletion site close to the end of mouse chromosome 7 in IR-induced tumors (Figure 1a). The DNA region, corresponding to the contig of RP23-35I7, was deleted in 15% of spontaneous tumors (n=20). Remarkably, the same region was deleted in 21 out of 29 (72.4%, P<0.001) IR-induced tumor samples (Figure 1b). Bioinformatics analysis revealed a previously uncharacterized open reading frame in this frequently deleted region as a candidate tumor suppressor. This open reading frame, encoding a protein consisting of 365 amino-acid residues, overlapped (99.4%) the largest coding exon of a previously uncharacterized gene comprised by 8 exons (Figure 1b). Its human counterpart was mapped to a CFS FRA10F at 10q26 (Figure 1c), spanning a region of allelic loss associated with cancers (Maier et al., 1997; Nagase et al., 1997). Therefore, we called it FATS for fragile-site associated tumor suppressor.

Figure 1

Identification of FATS. (a) A BAC array comparative genomic hybridization profile of chromosome 7 for a radiation-induced mouse lymphoma. An arrow indicates a deletion region of interest. (b) A BAC contig map of the FATS gene in the deleted region in panel a, and the frequencies of copy number loss of BAC contigs in spontaneous (n=20) and IR-induced (n=29) mouse thymic lymphomas. Boxes indicate exons and shaded boxes indicate coding regions. Genbank database accession number of FATS is NM_001081331. (c) Co-localization of BAC clone RP11-179O22 with a common fragile site FRA10F at 10q26. Fluorescence in situ hybridization image (red) on human chromosome 10 represents RP11-179O22 that spans 170 kb and contains all sequences of FATS ortholog. (d) The expression of FATS mRNA in normal tissues. The primer set for reverse transcription (RT)–PCR targets exon 2 and 4 respectively. (e) The mRNA levels of FATS in normal thymus and IR-induced thymic lymphomas. The same primer set as in panel d was used for RT–PCR analysis.

We next evaluated whether FATS was an expressed gene that was involved in IR-induced mouse tumors. Indeed, FATS mRNA was expressed in various tissues including thymus, brain, heart, lung, spleen and kidney, but undetectable in liver (Figure 1d). The expression of FATS was extensively silent in IR-induced lymphomas (Figure 1e), indicating its involvement in IR-induced tumorigenesis.

An exon, encoding the NH2-terminal region of FATS, is enveloped in sequences of genetic instability

Genetic instability is an important facet of carcinogenesis. We next analyzed whether the DNA sequences of FATS gene exhibit features of DNA fragility. Non-coding sequences of FATS gene are AT-rich. Interestingly, exon 2 of mouse FATS, encoding the NH2-terminal domain of FATS that is evolutionarily conserved (Figure 2a), is surrounded by AT-rich sequences inserted with AT-dinucleotide repeats (Figure 2b), which is a feature of CFSs instability (Schwartz et al., 2006; Zhang and Freudenreich, 2007). Unusually, additional dinucleotide repeats such as (CA)n and (TG)n are distributed in AT-rich sequences of both upstream and downstream of FATS exon 2 (Figure 2b). A secondary structure model was further derived by free energy data calculated using the mFOLD program ( The structure of (CA)15-(TG)21 repeat in intron 1 appears as a succession of three independent stem-loop structures (Figure 2c), which exhibit a tendency to induce replication pausing. The structure of (AT)24 in intron 2 appears as a succession of two independent stem-loop structures that are separated by a short single-stranded region (Figure 2d), which is similar to an AT-rich fragment causing DNA fragility (Zhang and Freudenreich, 2007). These dinucleotide repeats inserted in AT-rich sequences thus confer genetic instability that are susceptible to DNA lesions induced by replication-blocking and radiation (Pearson et al., 2005; Bichara et al., 2006).

Figure 2

NH2-terminal sequences of FATS protein and secondary structures of DNA surrounding its coding region. (a) Sequence alignment of NH2-terminal domains of FATS orthologs. Entrez database accession numbers are as follows: FATS_CF, GI: 73998998; FATS_BT, GI: 119917909; FATS_HS, GI: 52145312; FATS_RN, GI: 253735807; FATS_MM, GI: 124486743. CF, Canis familiaris; BT, Bos Taurus; HS, Homo sapiens; RN, Rattus norvegicus; MM, Mus musculus. (b) Genomic organization of dinucleotide repeats in AT-rich sequences surrounding exon 2 of mouse FATS gene. A shaded box indicates the coding region for NH2 terminus in panel a. (c) Secondary structure of a dinucleotide repeat in intron 1. (d) Secondary structure of a (AT)24 repeat in intron 2.

FATS deficiency results in hypersensitivity to DNA damage and severe defects in G2/M checkpoint

To provide a physiological context for our findings, we next examined whether FATS has a role in DNA damage response. We took RNA interference approach using a small interfering RNA (siRNA) that specifically knockdowned the expression of FATS mRNA (Figure 3a). The phosphorylation of Chk1 kinase, a mediator of DNA damage signaling, was more pronounced after knockdown of FATS expression in mouse embryonic fibroblast (MEF) cells after IR treatment (Figure 3b). To further verify this observation, we examined the effect of FATS knockdown on IR-induced nuclear foci of 53BP1, a mediator of DNA damage signaling (DiTullio et al., 2002). The nuclear staining of 53BP1 in unstressed cells was diffused, and a few nuclear foci of 53BP1 were observed 3 h after IR treatment in MEF cells. In contrast, FATS deficiency significantly increased the number of 53BP1 nuclear foci under the same conditions (Figure 3c). These results demonstrated that knockdown of FATS expression resulted in sensitivity to DNA damage response induced by IR.

Figure 3

Involvement of FATS in DNA damage response and functional cell-cycle checkpoint. (a) MEF cells were untreated or transfected with either FATS siRNA or control siRNA, respectively. Expression of endogenous FATS was detected by reverse transcription (RT)–PCR after 48 h. (b) MEF cells with or without FATS siRNA transfection were exposed to 10 Gy dose of IR. Cell lysates were prepared 6 h after IR and subjected to western blot analysis using an antibody to phospho-Chk1 on Serine 345. (c) Representative microscope images of IR-induced nuclear foci of 53BP1 in MEF cells with or without FATS inhibition. At least 200 cells were examined and 53BP1 foci formation was quantified. (d) Mitotic entry of FATS-inhibited MEF cells after IR. After siRNA transfection, cells were subjected to IR treatment and incubated with nocodazole (0.2 μg/ml) for 24 h. Cells were then fixed, stained and examined by fluorescence microscopy to determine the mitotic index as described earlier (Bunz et al., 1998). (e) Representative microscope images of FATS-inhibited cells entering mitosis after IR-induced DNA damage. The siRNA-treated MEF cells were subjected to IR (10 Gy) exposure. After 24 h, cells were fixed and stained with an antibody to γ-tubulin component of centrosome, and nuclear DNA was stained with DAPI (red).Arrows indicate centrosomes (green). At least 200 cells were examined and the percentage of mitotic defects after IR was shown.

Because radiosensitivity and cancer susceptibility are hallmarks of many genomic instability syndromes (Shiloh, 2003), and cell-cycle checkpoint machinery has a pivotal role in guarding genomic stability and suppressing carcinogenesis, we next investigated the mitotic entry after DNA damage in FATS-inhibited MEF cells. After irradiation, cells were treated with nocodazole, a microtubule-disrupting agent that can trap cells in mitosis for several hours. In normal cells, the mitotic index was very low after irradiation. However, FATS-inhibited cells exhibited significantly higher mitotic index after irradiation (Figure 3d). For those FATS-deficient cells entering mitosis under IR-induced genotoxic stress, severe mitotic defects in nuclear division and centrosome duplication were observed (Figure 3e), confirming that FATS is required to sustain G2/M checkpoint after DNA damage.

FATS regulates cell-cycle inhibitor p21

We further evaluated the effect of FATS expression on cell proliferation. Knockdown of FATS expression led to a significant increase in nuclear staining of Ki-67, a well-known proliferation marker (Figure 4a). In line with this observation, the growth rate of NIH3T3 cells stably transfected with a FATS-expressing vector was significantly decreased than those stably transfected with an empty vector pcDNA3 (Figure 4b). Because p21 is not only a cyclin-dependent kinase (CDK) inhibitor but also essential to sustain G2 checkpoint after DNA damage, we determined whether FATS by itself could upregulate p21. Indeed, the protein level of p21 was significantly increased in the presence of overexpressed FATS. Interestingly, FATS did not affect the protein level of p27, another CDK inhibitor (Figure 4c). To gain further insight into the properties of the NH2-terminal region of FATS, namely FATS(1–363), whose coding region was frequently deleted in IR-induced tumors and susceptible to DNA lesions caused by repeat instability (Figures 1 and 2), we generated an FATS(1–363)-expressing vector and performed similar experiments. The overexpression of FATS(1–363) alone in NIH3T3 cells was sufficient to induce p21 (Figure 4c). The activities of cyclinE-associated CDK2 kinase and cyclinB1-associated CDK1 kinase, which stimulates the progression of G1 to S phase and G2 to M phase of cell cycle, respectively (Sherr and Roberts, 1999; Ohi and Gould, 1999), were significantly suppressed after overexpression of FATS(1–363) (Figures 4d and e), supporting the role of FATS in regulating cell cycle. The overexpression of FATS induced a dominant effect on G1 arrest (Supplementary Figure S1a and S1b). However, the overexpression of FATS(1–363) in p21-null cells failed to induce G1 arrest (Figure 4f), indicating that FATS-mediated effect on cell cycle is p21-dependent.

Figure 4

Overexpression of the NH2-terminal FATS is sufficient to induce p21. (a) Proliferation index of cells after FATS knockdown. Cells were fixed 24 h later and stained for a proliferation marker Ki-67 (red). The nuclei were stained with DAPI. The representative images were shown and Ki-67 staining was quantified. (b) Growth rates of NIH3T3 cells after stable transfection of a FATS-expressing vector and an empty vector pcDNA3, respectively. (c) Induction of p21 by FATS and FATS (1–363) in NIH3T3 cells. (d) CyclinE-associated kinase activity in the absence or presence of FATS(1–363). After transfection, cell lysates were subjected to immunoprecipitation (IP), and Histone H1 was used as a substrate for in vitro kinase assay. The protein level of CyclinE in lysate was examined by immunoblotting (IB). (e) CyclinB1-associated kinase activity in the absence or presence of FATS(1–363). Histone H1 was used as a substrate for in vitro kinase assay. (f) Effect of FATS (1–363) on G1 phase in the absence of p21. Wild-type (wt) and p21 null MEF cells were transfected with FATS(1–363) or an empty vector pcDNA3, respectively. After 48 h, cells were fixed and subjected to flow cytometry analysis. The percentage of G1 cells in wt MEFs and p21-null MEFs before FATS(1–363) transfection was 62.5 and 55.0%, respectively. The increase in G1 is depicted in the bar graph. Each value is the average of three independent experiments and error bars represent the s.d.

Expression of the NH2-terminal region of FATS is sufficient to stabilize p21 independently of ubiquitination

Interestingly, the expression of FATS(1–363) was capable of increasing p21 protein level in p53-null cells (Figure 5a), indicating that FATS induces p21 protein in a p53-independent manner. Because p21 is an unstable protein and its abundance is also tightly regulated post-transcriptionally, we next evaluated whether FATS could stabilize p21. In the presence of cycloheximide (CHX), an inhibitor of protein synthesis, the expression of p21 protein was diminished after CHX treatment for 4 h in cells. However, the protein level of p21 only slightly decreased after CHX treatment for 12 h in the presence of overexpressed FATS(1–363) (Figure 5b), indicating that FATS inhibits the degradation of p21. To investigate the effect of FATS on p21 abundance after DNA damage without the influence of p53-mediated transcriptional activation, we performed immunoblotting analysis to determine the expression level of p21 protein in HeLa cancer cells, which carry inactive p53 (Hoppe-Seyler and Butz, 1993). The half-life of endogenous p21 in HeLa cells was shorter than 1 h, and the expression of p21 protein in HeLa cells was abolished within 3 h after IR treatment. In contrast, the protein level of p21 in the presence of FATS(1–363) under the same conditions was not significantly changed even 24 h after IR treatment (Figure 5c), strongly suggesting that FATS is important to regulate p21 abundance under genotoxic stress. To gain more definitive insight into the mechanism of FATS-mediated stabilization of p21, we investigated the effect of FATS on p21 ubiquitination. His-tagged ubiquitin and a p21-expressing vector were co-transfected with FATS(1–363) or an empty vector into p53-null cells, respectively. Ubiquitinated proteins were subsequently purified and subjected to immunoblotting using an antibody against p21. The expression of FATS(1–363) did not inhibit the ubiquitination of p21 in p53-null cells (Figure 5d), indicating that FATS-mediated stabilization of p21 is ubiquitin-independent. These results were in agreement with our observation (Figure 4c) that FATS did not change the protein level of p27, whose turnover was strictly ubiquitin-dependent (Pagano et al., 1995).

Figure 5

Inhibition of p21 turnover by the NH2-terminal region of FATS independently of p53. (a) Induction of p21 by FATS(1-363) in the absence of p53. MEF cells (p53−/−) were transfected with FATS (1–363) or an empty vector, respectively. The protein level of p21 was evaluated by immunoblotting using an antibody to p21. Results of three independent experiments were averaged and plotted as relative p21 induction. Error bars refer to ±s.d. A representative result of western blot was shown. (b) Degradation of endogenous p21 in the presence of FATS(1–363). The stable transfectants 3T3-FATS and its control, 3T3-pcDNA3, were treated with 50 μg/ml of cycloheximide (CHX) for different periods of time as indicated, followed by immunoblotting using a p21 antibody. The results were quantified and plotted against the indicated time course. (c) Protein levels of p21 after IR treatment in p53-inactivated cells. HeLa cells were transfected with FATS(1–363) or an empty vector, respectively. After 24 h, cells were treated with 10 Gy dose of IR. Protein levels of p21 were evaluated by immonublotting at the indicated time points. (d) p53−/− cells were transfected with p21 in combination with His-tagged ubiquitin in the absence or presence of FATS (1–363). At 48 h after transfection, cells were incubated with MG132 (20 μM) for 6 h. Ubiquitinated proteins were purified with Ni-NTA (Qiagen, Hilden, Germany) beads, followed by immunoblotting against p21.

The NH2-terminal FATS contains one HDAC1-binding domain

To explore the mechanism by which FATS inhibited p21 turnover, we determined the cellular localization of p21 induced by FATS. We performed cell fractionation analysis and found that FATS-induced p21 was localized in nucleus (Supplementary Figure S2). Given that p21 is selectively induced by HDAC inhibitors (Johnstone, 2002; Dokmanovic et al., 2007) and that HDAC1 is a major deacetylase localized predominantly to the nucleus (Lagger et al., 2002; Supplementary Figure S2), we hypothesized that FATS might stabilize p21 through interacting with HDAC1. To this end, we performed coimmunoprecipitation assay to assess whether a physical association could be observed between FATS and HDAC1 in cells. A Flag-tagged HDAC1 and a myc-tagged FATS(1–363) were co-transfected into HeLa cells, and cellular proteins binding to HDAC1 were immunoprecipitated using a Flag antibody, followed by immunoblotting with a myc antibody. Indeed, the NH2-termianl region of FATS protein coimmunoprecipitated with a significant amount of HDAC1, whereas a nonspecific immunoglobulin G (IgG) did not (Figure 6a), indicating that the NH2-terminal region of FATS specifically binds HDAC1 in vivo. Such conclusion was confirmed by the results using a reciprocal immunoprecipitation and immunoblotting procedure (Figure 6a). To determine whether FATS might directly interact with HDAC1, FATS(1–363) region was tagged with glutathione-S-transferase (GST), and purified fusion protein was subjected to pull down an in vitro-translated HDAC1. GST–FATS(1–363), but not GST alone, was found to bind a significant amount of HDAC1 (Figure 6b), indicating a direct interaction between the NH2-termianl region of FATS and HDAC1.

Figure 6

FATS as an HDAC1-binding protein. (a) Physical association of FATS(1–363) with HDAC1 in vivo. Cells were co-transfected with 1 μg of Flag-tagged HDAC1 and 1 μg of myc-tagged FATS (1–363). Cell lysates were subjected to immunoprecipitation (IP) followed by western blot analysis using a myc antibody. A reciprocal IP with a myc antibody followed by immunoblotting is shown in the right panel. (b) The NH2-terminal region of FATS directly interacts with HDAC1. Purified GST-FATS(1–363) protein was incubated with 35S-labeled in vitro translated HDAC1. Protein interaction was detected by autoradiography. (c) GST pull-down assay for delineation of an HDAC1-binding domain within NH2-terminal FATS. (d) Schematic diagram of FATS truncated mutants and the summarized results in panel c indicate FATS(67–175) as the binding domain to HDAC1.

In order to gain additional insight into the interaction between FATS and HDAC1, we generated truncated FATS mutants and performed GST pull-down assay. We found that FATS(67–363) and FATS (1–288) bound to HDAC1, whereas FATS(175–363) did not (Figures 6c and d). Therefore, FATS(67–175) domain within the NH2-terminal region of FATS is required for efficient interaction between FATS and HDAC1.

The NH2-terminal FATS inhibits HDAC1 binding to p21 and facilitates the acetylation of p21

We next assessed the effect of FATS–HDAC1 interaction on FATS-mediated induction of p21. Given that endogenous p21 undergoes acetylation (Chen et al., 2004), we sought to determine whether FATS(1–363) could have an impact on acetylation of endogenous p21. After transfection of FATS(1–363) into cells, the acetylated p21 was examined by immunoprecipitation with an antibody to acetylated lysine residue, followed by immunoblotting using a p21 antibody. The acetylation of p21, which was barely detectable in the absence of FATS(1–363), was much more pronounced after the overexpression of FATS(1–363) (Figure 7a), indicating that FATS-mediated stabilization of p21 is associated with enhanced acetylation modification of p21. Furthermore, a significant amount of HDAC1 was physically associated with GST–p21 protein, whereas GST protein did not bind to HDAC1 (Figure 7b), indicating that HDAC1 protein is directly associated with p21. We next examined the effect of the NH2-terminal FATS on HDAC1 binding to p21 in vivo. The interaction between HDAC1 and endogenous p21 was observed in vivo. However, such interaction was abolished in the presence of FATS(1–363) (Figure 7c). The effect of FATS(1–363) on inhibiting HDAC1 binding to p21 in vivo was further substantiated by the observation that much less amount of HDAC1 was coimmunoprecipitated with an HA-tagged p21 in the presence of the NH2-terminal FATS, in comparison with that in the absence of the NH2-terminal FATS (Figure 7d). To further characterize the role of HDAC1 in FATS-mediated regulation of p21, we used RNA interference to inhibit the expression of HDCA1. Although knockdown of HDAC1 led to an increase in basal level of p21 protein, the fold induction of p21 protein level by the NH2-terminal FATS was significantly attenuated (Figure 7e), demonstrating that FATS targets HDAC1 to stabilize p21.

Figure 7

The interaction between FATS and HDAC1 enhances acetylation modification of endogenous p21. (a) HeLa cells were transfected with FATS(1–363) or an empty vector. After transfection for 48 h, cell lysates were immunoprecipitated with an antibody specific to acetyl-lysine (Ac-K) or an immunoglobulin G control, followed by immunoblotting with a p21 antibody. (b) GST pull downs using purified GST-p21 and in vitro translated HDAC1. GST-p21 proteins in pull-down assay system were analyzed by coomasie blue staining. (c) Effect of FATS(1–363) expression on HADC1 binding to p21 in vivo. HeLa cells were untreated or transfected with Flag-tagged HDAC1 alone or in combination of FATS(1–363). After transfection for 48 h, cell extracts were subjected to immunoprecipitation with a Flag antibody, followed by immnoblotting using an antibody to p21. (d) HA-p21 was transfected into HeLa cells alone, or co-transfected with Flag-tagged HDAC1, in the absence or presence of FATS(1–363). Cell extracts were subjected to immunoprecipitation with a Flag antibody, followed by immnoblotting using an antibody to HA tag. (e) FATS(1–363)-mediated induction of p21 after knockdown of HDAC1 expression. HeLa cells were transfected with HDAC1 siRNA or a non-targeting control siRNA for 24 h. siRNA-transfected cells were subsequently transfected with FATS(1–363) and an empty vector, respectively. After 24 h, cellular p21 was examined by immunoblotting using an antibody to p21. Results from three individual experiments were plotted for comparison. The data are represented as means±s.d.

Acetylation of p21 suppresses its ubiquitin-independent proteasomal turnover linked to FATS-mediated control of p21 function

To gain further insight into the biological significance of acetylation modification of p21, we asked whether acetylation of p21 could suppress its proteasomal degradation. The treatment of trichostatin A (TSA), a HDAC inhibitor, has been known to selectively induce p21 (Johnstone, 2002; Dokmanovic et al., 2007). To rule out the possibility that p53-dependent induction of p21 under TSA treatment was involved, we investigated the effect of TSA treatment on p21 stability in HeLa cells, which carry inactive p53. Consistent with the results in Figure 5c, endogenous p21 protein in HeLa cells quickly diminished under CHX treatment and only about 10% of p21 protein was remained after CHX treatment for 1 h (Figures 8a and b). However, TSA treatment in advance for 2 h, followed by CHX treatment under the same conditions, significantly increased the protein stability of endogenous p21 and 80% of p21 protein remained intact after CHX treatment for 1 h (Figures 8a and b), indicating the relevance of p21 actetylation to its stability in vivo. The complete degradation of endogenous p21 after CHX treatment for 2 h in the presence of TSA (Figure 8a) suggested the transient effect of TSA on p21 stability in the presence of CHX and the dynamic mechanism underlying p21 acetylation.

Figure 8

Acetylation of p21 suppresses its ubiquitin-independent proteasomal turnover linked to FATS-mediated control of p21 function. (a) Effect of HDAC inhibitor on p21 protein stability. HeLa cell were incubated with or without TSA (10 μM) treatment for 2 h, and subsequently treated with CHX at 100 μg/ml for indicated times. The endogenous p21 was evaluated by immunoblotting. (b) The effect of TSA and CHX treatment on p21 degradation as described in panel a. Cells with or without TSA treatment for 2 h in panel a were subjected to immunoprecipitation with an antibody specific to Ac-K, followed by immunoblotting using an antibody to p21. (c) Schematic diagram showing the position of lysine residues in C-terminal region of p21 protein. The sequences of HA-tagged p21(139–164) peptide with or without acetylation modification on K141, K154, K161 and K163 are indicated. (d) In vitro proteolytic analysis to compare the stability of p21(139-164) and Ac-p21(139-164) in the presence of activated 20S proteasome (Chemicon) and in the absence of ATP and ubiquitin. (e) Effect of acetylation on C8-binding to p21(139–164). GST pull-down assay was performed and bound polypeptides were analyzed by immunoblotting using an HA antibody. (f) The abundance of p21 after FATS knockdown. MEF cells were transfected with FATS siRNA or control siRNA, respectively. After transfection for 24 h, the steady-state level of p21 protein was analyzed by immunoblotting. (g) The protein levels of p21 in mouse thymic lymphomas were measured by immunoblotting. The total RNA was extracted from a portion of the same sample and the mRNA levels of FATS were examined by reverse transcription–PCR. (h) Flow cytometry analysis of FATS-inhibited MEF cells after IR (10 Gy) treatment. The percentage of the population of cells was summarized. Each value is the average of three independent experiments and error bars represent the s.d.

It is noteworthy that the C terminus of p21, namely p21 (139–164), which was bound to both C8 α-subunit of 20S proteasome (Touitou et al., 2001) and REGγ proteasome activator (Chen et al., 2007), has an essential role in direct degradation of p21 by proteasome (Sheaff et al., 2000; Chen et al., 2007; Li et al., 2007). In addition, there are four lysine residues (K141, K154, K161 and K163) in the C terminus of p21 protein, which raising the possibility that FATS-mediated acetylation might be directly correlated with p21 stabilization. Owing to the difficulties in obtaining a constitutively acetylated protein in vivo, we generated an HA-tagged Ac-p21(139–164) peptide, in which K141, K154, K161 and K163 were in vitro acetylated. The HA-tagged p21(139–164) peptide was also generated as its control (Figure 8c). The same amount of Ac-p21(139–164) or p21(139–164) peptide was incubated with purified 20S proteasome, respectively, in the absence of adenosine triphosphate and ubiquitin. After incubation at 37 °C for 10 min, p21(139–164) peptide was completely degraded. In contrast, the amount of Ac-p21(139–164) peptide only slightly decreased after incubation with 20S proteasome for 10 min, and Ac-p21(139–164) was detectable even after incubation with 20S proteasome for 60 min (Figure 8d), strongly demonstrating that acetylation of p21 is critical to extend the half-life of free p21. To further validate this mechanism, we performed GST pull-down assay to evaluate the binding of C8 subunit to p21(139–164) peptide with or without acetylation. Purified GST-C8 protein strongly associated with p21(139–164) peptide. However, only tiny amount of Ac-p21(139–164) peptide was bound to GST-C8, which did not bind a nonspecific BSA protein (Figure 8e). Therefore, acetylation of lysine residues (K141, K154, K161 and K163) in the C terminus of p21 remarkably suppresses the direct binding of C8 α-subunit of 20S proteasome to p21, leading to significantly increased stability of p21 protein.

Specifically, knockdown of FATS mRNA expression significantly decreased the steady-state level of p21 protein (Figure 8f), and reduced protein levels of p21 in thymic lymphomas were observed in those samples with downregulated expression of FATS (Figure 8g), which strongly demonstrate the essential role of FATS in controlling p21 abundance. Furthermore, in contrast to control cells that exhibit G1 arrest and significantly decreased S phase entry after IR treatment, the inhibition of S phase entry after DNA damage was significantly compromised in cells after FATS knockdown (Figure 8h), indicating that G1 cell-cycle arrest after DNA damage is defective in FATS-deficient cells. These observations were consistent with our results as shown in Figures 3 and 4, emphasizing that the tumor suppressor role of FATS is linked to, at least in part, its control of p21 function in cell-cycle checkpoints.


Identifying cancer genes and understanding their involvement in tumorigenesis are critical steps in controlling this disease. Despite tremendous works in genome-wide screening, these attempts were often hampered by the relatively low resolution of banded chromosomes and a plethora of genomic alterations in late-stage human tumors with genetic heterogenicity, making it difficult in identifying events at early stages of tumorigenesis. The identification and functional characterization of FATS suggest the values of dissecting frequent changes in IR-induced mouse tumors (Mao et al., 2005), which will allow us to uncover new cancer genes involved in early tumorigenesis. We verify that human ortholog of FATS is a CFS gene at FRA10F, one of 76 aphidicolin-induced CFSs that have previously not been cloned and characterized at the molecular level (Durkin and Glover, 2007). CFS stability is regulated by DNA damage checkpoints ATR (Casper et al., 2002), indicating that some level of replication stalling occurs at CFS regions and they could be prone to cause DSBs after radiation treatment in normal cells. Interestingly, CFS stability is not regulated by another checkpoint kinase ATM (for ataxia-telangiectasia mutated) that responds primarily to DSBs (Casper et al., 2002). As DSBs are the principal lesions of importance in the induction of chromosomal abnormalities and gene mutations, our study brings new insight into the function of some CFS genes, if not all, in linking DSBs sensor to DNA damage checkpoint machinery, and provides the first evidence that a CFS gene actively monitors DNA damage response and genomic stability. The discovery of a radiation-susceptible gene FATS that has an important role in maintaining genomic stability through regulating p21 abundance adds new insights into our understanding of radiation-associated cancer risks.

As a negative regulator of the cell cycle, p21 is uniquely positioned to function as a central inhibitor of CDK1 and CDK2, both in unstressed cells and after genotoxic stresses, leading to growth arrest in the G1 and G2 phase of cell cycle. The abundance of cellular p21 is tightly controlled at both transcriptional and post-transcriptional levels. Given that p21 is unstable and its transcription can be activated by many oncogenic factors other than p53 tumor suppressor (Macleod et al., 1995), the stabilization of p21 protein has a critical role in regulating p21 abundance and its function in cell-cycle control. Our finding reveals FATS as a potent regulator of p21 abundance. Following induction of DSBs by IR, FATS-inhibited cells exhibit radiosensitivity as shown by more pronounced Chk1 phosphorylation (Figure 3b) and distinct 53BP1 foci formation at the sites of DNA damage (Figure 3c), a character of a group of inherited chromosomal instability syndromes including ataxia telangiectasia (AT) and Nijmegen breakage syndrome with a predisposition to cancer. In addition, defects in mitosis and centrosome duplication after IR in FATS-deficient cells (Figure 3e) are more severe than that in p21-deficient cells (Bunz et al., 1998), supporting that FATS acts upstream of p21 and is required for G2/M checkpoint function. The tumor suppressor activity for p21 is demonstrated by the genetic evidence (Martin-Caballero et al., 2001), and p21 is a major determinant of tumor suppression by p53, especially in case p53 loses its capacity in inducing apoptosis (Barboza et al., 2006; Efeyan et al., 2007). It is noteworthy that p21 is often deregulated in cancer, which is frequently associated with its cytoplasmic expression due to phosphorylation at Thr145 (Zhou et al., 2001; Winter et al., 2003; Xia et al., 2004). FATS-mediated acetylation of p21 may not only inhibit p21 degradation by proteasome, but also interfere with its phosphorylation by some oncogenic kinase to retain its nuclear localization and proper function to regulate cell cycle. The importance of acetylation modification of p21 is further emphasized by the fact that the C terminal p21 interacts with PCNA, a proliferating cell nuclear antigen (Touitou et al., 2001). The stabilization of p21 by FATS may also facilitate p21-mediated regulation of cell-cycle progression by inhibiting PCNA, independently of CDK inhibition.

HDAC1 is a major deacetylase whose genetic deletion causes a significant reduction in total HDAC activity (Zupkovitz et al., 2006), and its involvement in cancer has been established (Lin et al., 1998). In addition, HDAC1 has been implicated in ATM-mediated sensing of IR-induced DNA damage (Kim et al., 1999), implying the potential crosstalk between FATS and ATM. Although we cannot rule out the possibility that the interaction between FATS and HDAC1 may also abolish HDAC1-mediated repression of p21 transcription, which acts through a general transcription factor Sp1 (Lagger et al., 2002; Lagger et al., 2003), such effect on FATS-mediated induction of p21 is negligible, as FATS did not affect the expression of p27, whose expression is also mediated by Sp1 (Wei et al., 2003; Cen et al., 2008). Finally, FATS may have potential application in cancer therapeutics and the development of new generation of HDAC inhibitors.

Materials and methods

Microarray-based comparative genomic hybridization

Mouse whole-genome BAC arrays consisting of more than 19 200 BAC clone DNAs were used to perform comparative genomic hybridization as previously described (Cai et al., 2002). Heterozygous p53+/− mice were exposed to a single dose of 4 Gy by whole-body γ-radiation to induce tumors as previously described (Mao et al., 2005). Mouse Cot-1 DNA was used for blocking repetitive sequences in BAC clones and genomic probes. The Cy5/Cy3 ratios were plotted along individual chromosomes. For each mouse tumor sample, two experiments were carried out with reversal of Cy3/Cy5 labeling to remove any ratio artifact.

Reagents, plasmids and cell culture

NIH3T3 and HeLa cells were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA), MEF and p53-null cells were obtained from Dr A Balmain (San Francisco, CA, USA). p21-null cells were kindly provided by Dr Lozano G (Houston, TX, USA). Cells were grown in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal calf serum.

To generate pcDNA3-FATS, a FATS fragment from MEF cDNA was generated by PCR using the following primers: 5′-IndexTermCACACTCCTGGGAGCCTTAC-3′ and 5′-IndexTermCAGATCCAGGGCTAGCAGAG-3′ and subsequently inserted into pcDNA3.1/Topo vector (Invitrogen). A FATS(1–363) fragment was obtained by PCR amplification from MEF cDNA using the following primers: 5′-IndexTermCTGGCATCACAGAACACAAGAATGA-3′ and 5′-IndexTermCTACTCACCAGCCCTGTAACTCCAG-3′, and subsequently inserted into pcDNA3.1/Topo vector to generate pcDNA3-FATS(1–363). All the constructs were verified by sequence analysis. GST-FATS(1–363) plasmid was produced by in-frame inserting a FATS(1–363) fragment from pcDNA3-FATS (1–363) into EcoRI/XhoI sites of pGex-6p1 vector (Amersham Biosciences, Buckinghamshire, UK). A vector pCMV-Tag5 (Stratagene, La Jolla, CA, USA) was used to generate myc-FATS(1–363). 3T3-FATS and 3T3-pcDNA3 cells were generated by transfecting pcDNA3-FATS or pcDNA3 (Invitrogen) into NIH3T3 cells, respectively, followed by G418 selection. Plasmids were transfected into cells by lipofectamine (Invitrogen) or an electroporator (Amaxa, Cologne, Germany). Flag-HDAC1 was kindly provided by Dr RA DePinho (Boston, MA, USA). GST-p21 and HA-p21 were gifts from Dr H Lu (Indianapolis, IN, USA). GST-C8 was kindly provided by Dr Y Long (Shanghai, China). Antibodies including anti-p21, anti-p27, and anti-Ki-67 were purchased from BD Pharmigen (San Diego, CA, USA); antibodies to Flag, β-actin, and γ-tubulin were from Sigma (Buchs, Switzerland); an antibody to cyclinB1 was from Santa Cruz Biotechnology (Santa Cruz, CA, USA); an antibody to acetyl-lysine was from Upstate (Charlottesville, VA, USA). The antibodies to Chk1 and Chk1 (phospho S345) were purchased from Abcam (Cambridge, UK). HA-tagged p21(139–164) and Ac-p21(139–164) peptides were synthesized by Shanghai Science Peptide Biological Technology (Shanghai, China). The sequences were as follows: IndexTermYPYDVPDYAGRKRRQTSMTDFYHSKRRLIFSKRKP and IndexTermYPYDVPDYAGRK(Ac)IndexTermRRQTSMTDFYHSK(Ac)IndexTermRRLIFSK(Ac)IndexTermRK(Ac)P. TSA, MG132 and CHX were purchased from Sigma.

RNA interference

FATS siRNA was generated using a Block-iT RNAi topo transcription kit (Invitrogen) and Block-iT dicer RNAi kits (Invitrogen) following the manufacturer's instructions. The primers for FATS were 5′-IndexTermCTCAGCCTCCGCTGTAGTTC-3′ and 5′-IndexTermCCTTCCAGTGACCACCTTGT-3′. The siRNA pool for FATS was obtained by cleaving ds-RNA with Dicer enzyme. Purified FATS-siRNA was diluted to 0.1 μg/μl and stored at −80 °C. A non-targeting siRNA pool from Dharmacon (D-001206-13, Lafayette, CO, USA) was used as control. Lipofectamine (Invitrogen) was used for siRNA transfection.

Reverse transcription–PCR

First-strand cDNA was synthesized using M-MuLV reverse transcriptase (New England BioLabs, Ipswich, MA, USA) from 0.5 μg of total RNA. The primers specific to mouse FATS were as follows: 5′-IndexTermGTGACAGTGGGGTCTTCGTT-3′ and 5′-IndexTermGGTGTTGAAGTCCCAGCAAT-3′. The primers specific to a housekeeping gene glyceraldehyde-3-phosphate dehydrogenase were as follows: 5′-IndexTermTGCCTCCTGCACCACCAACT-3′ and 5′-IndexTermCGCCTGCTTCACCACCTT-3′.

GST pull-down assay

GST fusion proteins were expressed and purified as described previously (Li et al., 2004).

Immunoprecipitation and western blot

Immunoprecipitation and western blot analysis was performed as described previously (Li et al., 2004).

Fluorescence in situ hybridization and immunofluorescence microscopy

Peripheral blood lymphocytes from two healthy individuals were cultured and treated with aphidicolin (0.2 μM) for 24 h. Then metaphase chromosomes were prepared according to standard procedures. A BAC clone RP11-179O22 was selected from the UCSC database ( for fluorescence in situ hybridization analysis as described (Liehr et al., 2002).

Cells were fixed with 4% paraformaldehyde for 20 min, permeabilized with 0.1% Triton X-100 solution, and then immunostained with a primary antibody (1:100 dilution) overnight at 4 °C, then incubated with a fluorescein isothiocyanate- or cy3-conjugated secondary antibody (1:200 dilution) for 1 h. Images were taken with a deltavision deconvolution microscope (Olympus, Tokyo, Japan).

MTT assay

Cells (1 × 104) were plated onto 24-well plates. At the indicated times, 5 mg/ml of MTT [3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide] (Sigma) was added into media at 1:10 dilution and incubated for 5 h. Then supernatants were removed, and 1 ml of acidic isopropanol containing 0.04 M HCl was added to dissolve intracellular purple formazan. Absorbance was measured at 570 nm. The reference wavelength was 650 nm.

Flow cytometry analysis

Cells were harvested and fixed in 70% cold ethanol for 30 min at room temperature. Cells were stored at 4 °C until ready to stain. Cells were recovered by centrifugation, then stained with propidium iodide containing 100 μg/ml of RNAse and subjected to analysis with a flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA).

In vitro proteolytic analysis

The proteolytic analysis was performed using a 20S proteasome activity assay kit (Chemicon, Cat. no. APT280, Temecula, CA, USA). In brief, 2 μg of purified HA-tagged p21(139–164) or Ac-p21(139–164) peptides was incubated with 1 μl of 20S proteasome (Chemicon, Part no. 90205) in assay buffer (25 mM HEPES, (pH 7.5), 0.5 mM EDTA, 0.05% NP-40 and 0.001% SDS) at 37 °C for different time course. The reaction was stopped by adding 10 μl of SDS sample buffer (65 mM Tris, pH 8.0; 10% glycerol, 2% SDS, 1% DTT, 0.01% bromophenol blue) and heating at 95 °C for 5 min. An aliquot of the reaction was analyzed by western blotting using an antibody against HA.

In vitro kinase assay

Cell extracts were harvested and subjected to immunoprecipitation using an antibody against cyclin B1 at 4 °C overnight. Immnuocomplex was then suspended in buffer containing 25 mM Tris (pH 7.5), 10 mM MgCl2, 10 μCi [γ-32P]-adenosine triphosphate and 0.1 μg histone H1. After incubation at 30 °C for 30 min, samples were boiled and loaded on SDS–polyacrylamide gel electrophoresis gel. The phosphorylation of histone H1 was detected by autoradiography.

Accession codes




  1. Abbas T, Dutta A . (2009). p21 in cancer: intricate networks and multiple activities. Nat Rev Cancer 9: 400–414.

  2. Barboza JA, Liu G, Ju Z, El-Naggar AK, Lozano G . (2006). p21 delays tumor onset by preservation of chromosomal stability. Proc Natl Acad Sci USA 103: 19842–19847.

  3. Bendiennat M, Boulaire J, Jascur T, Brickner H, Barbier V, Sarasin A et al. (2003). UV irradiation triggers ubiquitin-dependent degradation of p21(WAF1) to promoter DNA repair. Cell 114: 599–610.

  4. Bichara M, Wagner J, Lambert IB . (2006). Mechanisms of tandem repeat instability in bacteria. Mut Res 598: 144–163.

  5. Bloom J, Amador V, Bartolini F, DeMartino G, Pagano M . (2003). Proteasome-mediated degradation of p21via N-terminal ubiquitinylation. Cell 115: 71–82.

  6. Brenner DJ, Doll R, Goodhead DT, Hall EJ, Land CE, Little JB et al. (2003). Cancer risks attributable to low doses of ionizing radiation: assessing what we really know. Proc Natl Acad Sci USA 100: 13761–13766.

  7. Brugarolas J, Chandrasekaran C, Gordon JI, Beach D, Jacks T, Hannon GJ . (1995). Radiation-induced cell cycle arrest compromised by p21 deficiency. Nature 377: 552–557.

  8. Bunz F, Dutriaux A, Lengauer C, Waldman T, Zhou S, Brown JP et al. (1998). Requirement for p53 and p21 to sustain G2 arrest after DNA damage. Science 282: 1497–1501.

  9. Buttel I, Fechter A, Schwab M . (2004). Common fragile sites and cancer: targeted cloning by insertional mutagenesis. Ann NY Acad Sci 1028: 14–27.

  10. Cai WW, Mao JH, Chow CW, Damani S, Balmain A, Bradley A . (2002). Genome-wide detection of chromosomal imbalances in tumors using BAC microarrays. Nat Biotechnol 20: 393–396.

  11. Casper AM, Nghiem P, Arlt MF, Glover TW . (2002). ATR regulates fragile site stability. Cell 111: 779–789.

  12. Cen B, Deguchi A, Weinstein IB . (2008). Activation of protein kinase G increases the expression of p21CIP1, p27KIP1, and histidine triad protein 1 through Sp1. Cancer Res 68: 5355–5362.

  13. Chen X, Chi Y, Bloecher A, Aebersold R, Clurman BE, Roberts JM . (2004). N-acetylation and ubiquitin-independent proteasomal degradation of p21(Cip1). Mol Cell 16: 839–847.

  14. Chen X, Barton LF, Chi Y, Clurman BE, Roberts JM . (2007). Ubiquitin-independent degradation of cell-cycle inhibitos by the REGγ proteasome. Mol Cell 26: 843–852.

  15. Coulombe P, Rodier G, Bonneil E, Thibault P, Meloche S . (2004). N-terminal ubiquitination of extracellular signal-regulated kinase 3 and p21 directs their degradation by the proteasome. Mol Cell Biol 24: 6140–6150.

  16. Deng C, Zhang P, Harper JW, Elledge SJ, Leder P . (1995). Mice lacking p21CIP1/WAF1 undergo normal development, but are defective in G1 checkpoint control. Cell 82: 675–684.

  17. DiTullio Jr RA, Mochan TA, Venere M, Bartkova J, Sehested M, Bartek J et al. (2002). 53BP1 functions in an ATM-dependent checkpoint pathway that is constitutively activated in human cancer. Nat Cell Biol 4: 998–1002.

  18. Dokmanovic M, Clarke C, Marks PA . (2007). Histone deacetylase inhibitors: overview and perspectives. Mol Cancer Res 5: 981–989.

  19. Durkin SG, Glover TW . (2007). Chromosome fragile sites. Annu Rev Genet 41: 169–192.

  20. Efeyan A, Collado M, Velasco-Miguel S, Serrano M . (2007). Genetic dissection of the role of p21Cip1/Waf1 in p53-mediated tumor suppression. Oncogene 26: 1645–1649.

  21. Glover TW, Arlt MF, Casper AM, Durkin SG . (2005). Mechanisms of common fragile site instability. Hum Mol Genet 14: R197–R205.

  22. Holm LE . (1990). Cancer occurring after radiotherapy and chemotherapy. Int J Radiat Oncol Biol Phy 19: 1303–1308.

  23. Hoppe-Seyler F, Butz K . (1993). Repression of endogenous p53 transactivation function in HeLa cervical carcinoma cells by human papillomavirus type 16 E6, human mdm-2, and mutant p53. J Virol 67: 3111–3117.

  24. Johnstone RW . (2002). Histone-deacetylase inhibitors: novel drugs for the treatment of cancer. Nat Rev Drug Discov 1: 287–299.

  25. Kastan MB, Bartek J . (2004). Cell-cycle checkpoints and cancer. Nature 432: 316–323.

  26. Kemp CJ, Wheldon T, Balmain A . (1994). p53-deficient mice are extremely susceptible to radiation-induced tumorigenesis. Nat Genet 8: 66–69.

  27. Kim GD, Choi YH, Dimtchev A, Jeong SJ, Dritschilo A, Jung M . (1999). Sensing of ionizing radiation-induced DNA damage by ATM through interaction with histone deacetylase. J Biol Chem 274: 31127–31130.

  28. Lagger G, O'Carroll D, Rembold M, Khier H, Tischler J, Weitzer G et al. (2002). Essential function of histone deacetylase 1 in proliferation control and CDK inhibitor repression. EMBO J 21: 2672–2681.

  29. Li X, Amazit L, Long W, Lonard DM, Monaco JJ, O'Malley BW . (2007). Ubiquitin- and ATP-independent proteolytic turnover of p21 by the REGγ-proteasome pathway. Mol Cell 26: 831–842.

  30. Li Z, Day CP, Yang JY, Tsai WB, Lozano G, Shih HM et al. (2004). Adenoviral E1A targets Mdm4 to stabilize tumor suppressor p53. Cancer Res 64: 9080–9085.

  31. Liehr T, Heller A, Starke H, Claussen U . (2002). FISH banding methods: applications in research and diagnosis. Expert Rev Mol Diagn 2: 217–225.

  32. Lin RJ, Nagy L, Inoue S, Shao W, Miller WH, Evans RM . (1998). Role of the histone deacetylase complex in acute promyelocytic leukaemia. Nature 391: 811–814.

  33. Macleod KF, Sherry N, Hannon G, Beach D, Tokino T, Kinzler K et al. (1995). P 53-dependent and independent expression of p21 during cell growth, differentiation, and DNA damage. Genes Dev 9: 935–944.

  34. Magrath IT . (1997). The treatment of pediatric lymphomas: paradigms to plagiarize? Ann Oncol 8: 7–14.

  35. Maier D, Comparone D, Taylor E, Zhang Z, Gratzl O, van Meir EG et al. (1997). New deletion in low-grade oligodendroglioma at the glioblastoma suppressor locus on chromosome 10q25-26. Oncogene 15: 997–1000.

  36. Mangelsdorf M, Ried K, Woollatt E, Dayan S, Eyre H, Finnis M et al. (2000). Chromosomal fragile site FRA16D and DNA instability in cancer. Cancer Res 60: 1683–1689.

  37. Mao JH, Li J, Jiang T, Li Q, Wu D, Perez-Losada J et al. (2005). Genomic instability in radiation-induced mouse lymphoma from p53 heterozygous mice. Oncogene 24: 7924–7934.

  38. Martin-Caballero J, Flores JM, Garcia-Palencia P, Serrano M . (2001). Tumor susceptibility of p21(Waf1/Cip1)-deficient mice. Cancer Res 61: 6234–6238.

  39. Maser RS, DePinho RA . (2002). Connecting chromosomes, crisis, and cancer. Science 297: 565–569.

  40. Nagase S, Yamakawa H, Sato S, Yajima A, Horii A . (1997). Identification of a 790-kilobase region of common allelic loss in chromosome 10q25-q26 in human endometrial cancer. Cancer Res 57: 1630–1633.

  41. Ohi R, Gould KL . (1999). Regulating the onset of mitosis. Curr Opin Cell Biol 11: 267–273.

  42. Pagano M, Tam SW, Theodoras AM, Beer-Romero P, Del Sal G, Chau V et al. (1995). Role of the ubiquitin-proteasome pathway in regulating abundance of the cyclin-dependent kinase inhibitor p27. Science 269: 682–685.

  43. Pearson CE, edamura KN, Cleary JD . (2005). Repeat instability: mechanisms of dynamic mutations. Nat Rev Genet 6: 729–742.

  44. Rouse J, Jackson SP . (2002). Interfaces between the detection, signaling, and repair of DNA damage. Science 297: 547–551.

  45. Schellong G . (1998). Pediatric Hodgkin's disease: treatment in the late 1990s. Ann Oncol 9: S115–S119.

  46. Sherr CJ, Roberts TM . (1999). CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev 13: 1501–1512.

  47. Shiloh Y . (2003). ATM and related protein kinase: safeguarding genome integrity. Nat Rev Cancer 3: 155–168.

  48. Schwartz M, Zlotorynski E, Kerem B . (2006). The molecular basis of common and rare fragile sites. Cancer Lett 232: 13–26.

  49. Sheaff RJ, Singer JD, Swanger J, Smitherman M, Roberts JM, Clurman BE . (2000). Proteasomal turnover of p21Cip1 does not require p21Cip1 ubiquitination. Mol Cell 5: 403–410.

  50. Touitou R, Richardson J, Bose S, Nakanishi M, Rivett J, Allday MJ . (2001). A degradation signal located in the C-terminus of p21WAF1/CIP1 is a binding site for the C8 alpha-subunit of the 20S proteasome. EMBO J 20: 2367–2375.

  51. Van Gent DC, Hoeijmakers JH, Kanaar R . (2001). Chromosomal stability and the DNA double-stranded break connection. Nat Rev Genet 2: 196–206.

  52. Varshavsky A, Turner G, Du F, Xie Y . (2000). Felix HoppeSeyler Lecture 2000. The ubiquitin system and the N-end rule pathway. Biol Chem 381: 779–789.

  53. Wei Q, Miskimins WK, Miskimins R . (2003). The Sp1 family of transcription factor is involved in p27(Kip1)-mediated activation of myelin basic protein gene expression. Mol Cell Biol 23: 4035–4045.

  54. Winter ZE, Leek RD, Bradburn MJ, Norbury CJ, Harris AL . (2003). Cytoplasmic p21WAF1/CIP1 expression is correlated with HER-2/neu in breast cancer and is an independent predictor of prognosis. Breast Cancer Res 5: R242–R249.

  55. Xia W, Chen JS, Zhou X, Sun PR, Lee DF, Liao Y et al. (2004). Phosphorylation/cytoplasmic localization of p21C/p1/WAF1 is associated with HER2/neu overexpression and provided a novel combination predictor for poor prognosis in breast cancer patients. Clin Cancer Res 10: 3815–3824.

  56. Zhang H, Freudenreich CH . (2007). An AT-rich sequence in human common fragile site FRA16D causes fork stalling and chromosome breakage in S. cerevisiae. Mol Cell 27: 367–379.

  57. Zhou BP, Liao Y, Xia W, Spohn B, Lee MH, Hung MC . (2001). Cytoplasmic localization of p21Cip1/WAF1 by Akt-induced phosphorylation in HER-2/neu-overexpressing cells. Nat Cell Biol 3: 245–252.

  58. Zupkovitz G, Tischler J, Posch M, Sadzak I, Ramsauer K, Egger G et al. (2006). Negative and positive regulation of gene expression by mouse histone deacetylase 1. Mol Cell Biol 26: 7913–7928.

Download references


We are grateful to Mien-Chi Hung for critical discussions. We thank Xiangwei He and Qi Gao for assistance in microscopic imaging; Tao Jiang and Qian Li for technical assistance. This study was supported in part by the following grants: Tianjin Medical University Cancer Institute and Hospital Start-up 08Y01 (to Z Li); Ministry of Science and Technology of China 973-program Concept Award 2009CB526407 (to Z Li); Tianjin Municipal Science and Technology Foundation (to Z Li); Department of defense of United States FG02-03ER63630 (to A Balmain); IZKF Jena Start-up S16 (to T Liehr).

Author information

Correspondence to Z Li or W-W Cai.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article


  • radiation
  • p21
  • acetylation
  • carcinogenesis

Further reading