Germline mutations affecting telomere maintenance or DNA repair may, respectively, cause dyskeratosis congenita or Fanconi anaemia, two clinically related bone marrow failure syndromes. Mice expressing p53Δ31, a mutant p53 lacking the C terminus, model dyskeratosis congenita. Accordingly, the increased p53 activity in p53Δ31/Δ31 fibroblasts correlated with a decreased expression of 4 genes implicated in telomere syndromes. Here we show that these cells exhibit decreased mRNA levels for additional genes contributing to telomere metabolism, but also, surprisingly, for 12 genes mutated in Fanconi anaemia. Furthermore, p53Δ31/Δ31 fibroblasts exhibit a reduced capacity to repair DNA interstrand crosslinks, a typical feature of Fanconi anaemia cells. Importantly, the p53-dependent downregulation of Fanc genes is largely conserved in human cells. Defective DNA repair is known to activate p53, but our results indicate that, conversely, an increased p53 activity may attenuate the Fanconi anaemia DNA repair pathway, defining a positive regulatory feedback loop.
Inherited bone marrow failure syndromes are a set of clinically related yet heterogeneous disorders in which at least one haematopoietic cell lineage is significantly reduced. Among them, Fanconi anaemia (FA) and dyskeratosis congenita (DC) are caused by germline mutations in key cellular processes, that is, DNA repair and telomere maintenance, respectively1.
We recently found that p53Δ31/Δ31 mice, expressing a mutant p53 lacking its C-terminal domain, die rapidly after birth with a complete set of features of the telomere syndrome DC, including aplastic anaemia, pulmonary fibrosis, oral leukoplakia, skin hyperpigmentation, nail dystrophy and short telomeres2. Loss of the p53 C terminus increases p53 activity in mouse embryonic fibroblasts (MEFs) and in most tested tissues2,3, and p53Δ31/Δ31 MEFs exhibited decreased messenger RNA (mRNA) levels for 4 out of 10 genes implicated in telomere syndromes (Dkc1, Rtel1, Tinf2 and Terf1). Nutlin, a drug that prevents the Mdm2 ubiquitin ligase from interacting with p53, allowed to confirm that p53 activation leads to the downregulation of these four genes. These data revealed that p53 plays a major role in telomere metabolism.
We previously focused on the potential p53-mediated regulation of genes mutated in DC (Dkc1, Rtel1 and Tinf2) or implicated in aplastic anaemia, a milder form of telomere syndrome (Terf1)4. As a striking evidence for the clinical relevance of our mouse model, patients with severe DC who carry mutations affecting PARN, a negative regulator of p53, were recently shown to exhibit decreased DKC1, RTEL1 and TERF1 mRNA levels5. Importantly, however, tens of proteins are thought to be involved in the regulation of telomeres (reviewed in ref. 6). Thus, it remained possible that the impact of p53 on telomere-related genes was underestimated in our previous study. Here we tested whether p53 affects the expression of 42 additional genes implicated in telomere metabolism, and found 7 genes that are downregulated in p53Δ31/Δ31 cells. Importantly, some of these p53-regulated genes are involved in the FA DNA repair pathway. This was particularly intriguing because Rtel1, one of the four telomere-related genes we previously found regulated by p53, encodes a Fancj-like helicase7. These observations led us to evaluate whether p53 regulates more genes belonging to the FA pathway, and whether p53Δ31/Δ31 cells exhibit characteristic features of FA cells. We found that murine p53 downregulates 12 Fanc genes, that human p53 downregulates 9 FANC genes and that the capacity to repair DNA interstrand crosslinks is attenuated upon p53 activation. These data reveal an unexpected role for p53 in downregulating the FA DNA repair pathway, which may help to understand the pathological processes implicated in FA, and suggest therapeutic strategies against tumour cells that retain a functional p53 pathway.
Expression of telomere-related genes in p53Δ31/Δ31 cells
Our initial aim was to test whether, besides the four genes previously identified2, p53 could regulate other genes that might contribute to the telomere phenotype of p53Δ31/Δ31 mice. We therefore compared, in unstressed p53−/−, wild-type (WT) and p53Δ31/Δ31 fibroblasts, mRNA levels for 42 candidate genes reported to be relevant to telomere metabolism. Candidates included genes implicated in telomere syndromes (Acd/Tpp1, Apollo/Snm1b, C16orf57/Mpn1/Usb1, Naf1, Obfc1/Stn1, Parn and Sbds)5,6,8,9,10; genes mutated in diseases not primarily associated with telomere biology but for which telomere dysfunction or DC-like features were reported (Dnmt3b, Fancd2 and Recql4)6; genes encoding proteins of complexes involved in telomere biology, that is, the telomerase (Gar1/Nola1, Ruvbl1 and Ruvbl2), shelterin (Pot1a and Pot1b, Rap1/Terf2ip, and Terf2), CST (Ten1) and CIA (Ciao1, Iop1/Narfl, Mip18 and Mms19) complexes, as well as Cajal bodies (Coilin and Hot1)6,11, or proteins otherwise proposed to participate in telomere replication or maintenance (Artemis/Snm1c, Blm, Csb/Ercc6, Dek, Dna2, Ercc3/Xpb, Ercc4/Fancq/Xpf, Fancc, Fen1, Lmna/Progerin, Nbs1, Pim1, Slx4/Fancp, Timeless, Tnks1, Tnks1bp1, Upf1 and Wrn)6,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26.
For a gene to be a good candidate, we considered that the mean (from three to four independent experiments) of its mRNA levels in unstressed WT cells should fall between the means measured in p53−/− and p53Δ31/Δ31 cells; and that the means for the three genotypes should be statistically different according to an analysis of variance. Out of the 42 genes, 7 fulfilled these criteria: Blm, Dek, Fancd2, Fen1, Gar1, Recql4 and Timeless (Fig. 1a; Supplementary Fig. 1). Because RECQL4 was shown to be downregulated by p53 in human cells27, the lower Recql4 mRNAs in p53Δ31/Δ31 cells were not surprising. The decreased mRNA levels for the six other genes were not anticipated however. To specifically assay for a p53-dependent regulation, we next compared the effects of Nutlin, a drug that activates p53 by preventing its interaction with the ubiquitin ligase Mdm2. Results clearly indicated that p53 activation leads to the downregulation of these genes (Fig. 1b).
Importantly, the finding that p53 downregulates Gar1, which encodes a component of the telomerase complex, strengthened our previous conclusion that p53 plays a significant role in telomere biology. However, Fancd2 appeared as the gene whose expression was most markedly affected by p53 activation (Fig. 1b). This was surprising because, even if primary cells from patients with a FANCD2 mutation may exhibit telomere dysfunction28, these patients are diagnosed with FA, a syndrome primarily characterized by defects in DNA repair. This led us to further analyse the p53-dependent regulation of Fancd2. We first verified that the relative decrease in Fancd2 mRNA levels were observed in vivo, in bone marrow cells (BMCs) from p53Δ31/Δ31 mice (Fig. 1c). We next tested whether the p53-dependent regulation of Fancd2 detected by quantitative PCR had an impact on Fancd2 protein levels. Lower Fancd2 protein levels were observed in unstressed p53Δ31/Δ31 cells compared with unstressed p53−/− or WT cells, and Nutlin treatment led to a decrease in Fancd2 proteins only in WT and p53Δ31/Δ31 MEFs, in complete agreement with quantitative PCR data (Fig. 1d; Supplementary Fig. 2).
p53 activation leads to increased E2F4 binding at Fancd2
The p53-mediated downregulation of many genes requires the cdk inhibitor p21, and occurs through the recruitment, upon p53 activation, of E2F4 repressive complexes at their promoters29,30. Notably, this mechanism would account for the p53-dependent regulation of cell cycle genes whose promoters contain CDE/CHR regulatory motifs31,32,33. Consistent with this mechanism, p53 activation had no effect on Fancd2 mRNA levels in p21−/− cells (Fig. 2a), and chromatin immunoprecipitation (ChIP) experiments with an antibody against E2F4 indicated increased E2F4 binding at the Fancd2 promoter in Nutlin-treated WT cells, compared with unstressed WT or Nutlin-treated p53−/− cells (Fig. 2b; Supplementary Fig. 3). Of note, ChIP assays for E2F4 binding at the Fancd2 promoter could not be performed in p53Δ31/Δ31 MEFs because their accelerated senescence2 prevented the recovery of sufficient amounts of chromatin, but it is likely that the p53/p21/E2F4 pathway operates similarly in p53Δ31/Δ31 cells. We next identified a candidate CDE/CHR motif in the Fancd2 promoter, and mutation of the CDE element (typically bound by E2F4) abolished the Nutlin-dependent repression of this promoter in NIH-3T3 cells (Fig. 2c), independently of cell cycle dynamics (Supplementary Fig. 4). Thus, although the expression of Fancd2 is known to vary during the cell cycle34, the differences in Fancd2 mRNA levels observed between WT and p53Δ31/Δ31 MEFs would not simply result from differences in G1/S ratios2. Rather, our results indicate that p53 activation promotes the recruitment of E2F4 at the Fancd2 gene, and that E2F4 plays a major role in the repression of Fancd2.
In the experiments above, p53 activation resulted from a treatment with Nutlin, a molecule that acts as a specific Mdm2 inhibitor. We next tested whether similar results could be obtained in response to DNA damage, by evaluating the effects of doxorubicin, a clastogenic anticancer agent. Doxorubicin treatment led to decreased Fancd2 mRNA and protein levels in WT and p53Δ31/Δ31 cells, but not p53−/− MEFs (Supplementary Fig. 5a,b). Furthermore, we observed increased E2F4 binding at the Fancd2 promoter in doxorubicin-treated WT cells, compared with unstressed WT or doxorubicin-treated p53−/− cells (Supplementary Fig. 5c). Thus, both Nutlin and doxorubicin lead to p53 activation and consecutive Fancd2 downregulation.
Interestingly, the Blm and Fen1 genes, also downregulated by p53 (Fig. 1b), respectively, encode an helicase that associates with Fanc proteins in a multienzyme complex35, and an endonuclease stimulated by a Fanc protein36. Furthermore, Rtel1, one of the four telomere-related genes we previously found regulated by p53 (ref. 2), encodes a Fancj-like helicase7. This led us to further evaluate the impact of p53 activation on the FA DNA repair pathway.
p53 downregulates many Fanc genes
Because the expression levels of four FA genes had been tested in our previous experiments—Fancc, Fancd2, Fancp/Slx4 and Fancq/Ercc4 (Fig. 1; Supplementary Fig. 1), we next compared, in unstressed p53−/−, WT and p53Δ31/Δ31 cells, mRNA levels for the 15 remaining FA genes. Strikingly, 11 were less expressed in p53Δ31/Δ31 cells (Fig. 3a). Again, Nutlin was used to confirm the p53-mediated downregulation of these genes (Fig. 3b). As for Fancd2, this p53-mediated downregulation required p21 (Supplementary Fig. 6), and p53 activation correlated with an increased binding of E2F4 near the transcription start site of each of these Fanc genes (Fig. 3c). We next used the sequence of six functional CDE/CHRs to define a positional frequency matrix, which was then used to search in silico for candidate CDE/CHRs near the E2F4-binding sites identified in ChIP assays. Using this approach, candidate CDE/CHR motifs were identified for 9 out of the 11 tested Fanc genes, with the best candidate motifs for Fanci and Fancr (Fig. 4a; Supplementary Fig. 7). These data led us to further analyse the p53-mediated regulation of Fanci and Fancr. We first verified that the relative decreases in Fanci and Fancr mRNA levels were observed in vivo, in BMCs from p53Δ31/Δ31 mice (Supplementary Fig. 8). We then found that p53 activation leads to decreased Fanci and Fancr protein levels ex vivo (Fig. 4b; Supplementary Fig. 9). Luciferase assays next showed that mutating the CDE site in each candidate CDE/CHR abolished the Nutlin-dependent repression of the Fanci and Fancr promoters (Fig. 4a,c).
We also observed that a 24-h long treatment with doxorubicin led to decreased Fanci and Fancr mRNA, and protein levels in WT and p53Δ31/Δ31 cells, but not p53−/− MEFs (Supplementary Fig. 10a). Furthermore, the nine other Fanc genes downregulated by p53 on Nutlin treatment were also downregulated in a p53-dependent manner on treatment with doxorubicin (Supplementary Fig. 10b). We then searched for confirmation of our results by analysing the data recently reported by Younger et al., who performed a genomic analysis that integrated transcriptome-wide expression levels, genome-wide p53-binding profiles and chromatin state maps to characterize the regulatory role of p53 in response to DNA damage37. Although this approach was designed to identify direct p53 targets, we reasoned that genes regulated by p53 indirectly, via p21/E2F4, might also be detected in their transcriptome-wide expression data. These experiments were performed on p53−/− and WT MEFs, treated or not with doxorubicin for 6 h (ref. 37), and our previous time-course experiments with Nutlin suggested that 6 h might be sufficient to observe a partial p53-mediated trancriptional downregulation2. Thus, we extracted the data of Younger et al. (Gene Expression Omnibus # GSE55727) to analyse the expression of the 12 Fanc genes that we had found downregulated by p53. In agreement with our results, this analysis showed that doxorubicin led to an overall decrease in the expression of Fanc genes in WT, but not p53−/− MEFs (Supplementary Fig. 11).
Transcriptome data mining was also used to find whether the downregulation of Fanc genes could correlate with p53 activation in haematopoietic cells. The Homeobox (Hox) transcription factors are important regulators of normal and malignant haematopoiesis, because they control proliferation, differentiation and self-renewal of haematopoietic cells. We analysed the data of Muntean et al. (Gene Expression Omnibus # GSE21299), who immortalized murine BMCs by transduction with Hoxa9-ER cells in the presence of tamoxifen (4-OHT), and observed that they undergo myeloid differentiation 5 days after 4-OHT withdrawal38. We found this differentiation to correlate with an induction of genes known to be transactivated by p53 (Cdkn1A/p21, Mdm2 and Fas), and with the downregulation of Fanc genes (Supplementary Fig. 12).
In sum, we found that 12 genes of the FA DNA repair pathway are downregulated by p53 via a p21/E2F4 pathway, and identified CDE/CHR motifs that are crucial for this regulation for three of these genes. Importantly, the genes are downregulated by p53 in response to Mdm2 inhibition or DNA damage, or on haematopoietic cell differentiation, and encode proteins involved in all parts of the FA DNA repair pathway, that is, proteins that belong to the FA core complex (Fanca, Fancb and Fancm) and its accessory protein (Fanct/Ube2t), the pivotal ID2 complex (Fancd2 and Fanci), or downstream effector proteins (Fancd1/Brca2, Fancj/Bach1/Brip1, Fancn/Palb2, Fanco/Rad51c, Fancr/Rad51 and Fancs/Brca1)39,40,41,42. Together, these data suggested an important role for p53 in regulating the FA pathway.
p53 activation attenuates the repair of specific DNA lesions
A typical feature of FA cells is their inability to repair DNA interstrand crosslinks, as evidenced by an increased frequency of chromosomal aberrations, and more specifically tri- and quadri-radial chromosomes, after exposure to mitomycin C (MC)39. We compared the effects, on WT and p53Δ31/Δ31 cells, of a 48-h treatment with 50 nM MC. Such a treatment procedure was previously reported to differentially affect WT MEFs and MEFs with an impaired FA pathway43. Interestingly, we found that this procedure led to a rather subtle induction of p53 (suggested by a limited increase in p21 transactivation), which correlated with a twofold decrease in Fancd2 mRNA expression in p53Δ31/Δ31 MEFs, but no significant alteration of Fancd2 mRNA levels in WT cells (Supplementary Fig. 13). We next determined the frequencies of all types of chromosomal aberrations, or of radial chromosomes, in WT and p53Δ31/Δ31 cells before or after treatment with MC. In untreated cells, no significant difference was found between the two genotypes. Strikingly, however, chromosomal aberrations, and particularly radial chromosomes, were more frequent in p53Δ31/Δ31 cells after treatment with MC, consistent with a decreased capacity to repair interstrand crosslinks in the mutant cells (Fig. 5a). Accordingly, chromosomes with sister chromatid exchanges were also more frequent in MC-treated p53Δ31/Δ31 cells than in WT cells (Fig. 5b). These results suggested that the FA DNA repair pathway is attenuated in p53Δ31/Δ31 cells, presumably because these cells exhibit an increased p53 activity. Consistent with this, p53Δ31/Δ31 cells exhibited a decreased capacity to form Rad51 foci and an increased sensitivity to MC, and the pretreatment of cells with Nutlin appeared to further impact on these cellular phenotypes (Fig. 5c,d). Further evidence that the decreased DNA repair in p53Δ31/Δ31 cells resulted from increased p53 activity (rather than a loss of the p53 CTD per se) came from analysing Mdm2+/− Mdm4+/ΔE6 MEFs. These MEFs express a WT p53 protein, but exhibit an increased p53 activity due to lower levels of p53 inhibitors44,45. Like p53Δ31/Δ31 MEFs, Mdm2+/− Mdm4+/ΔE6 cells were more sensitive than WT cells to MC (Supplementary Fig. 14). In sum, a defective FA DNA repair pathway is known to activate p53 (ref. 46), but these results indicate that an increased p53 activity might reduce the expression of several FA genes and attenuate the FA DNA repair pathway. Taken together, these data indicate the existence of a positive regulatory feedback loop (Fig. 6).
Human p53 also regulates FA genes
We next tested whether the FA genes that were found regulated by murine p53 were similarly regulated in human cells. We compared human primary WT cells with p53-deficient cells, and observed that out of the 12 p53-regulated FA genes identified in mouse cells, 9 are also downregulated upon p53 activation in human MRC5 cells: FANCA, FANCB, FANCD1, FANCD2, FANCI, FANCJ, FANCM, FANCR and FANCT (Fig. 7a). Interestingly, one of these genes, FANCB, was recently identified as one of 210 genes most likely to be downregulated by p53 in a E2F4-dependent manner33. Furthermore, candidate CDE/CHR motifs could be found for each of these genes (Supplementary Fig. 15a), and the CDE/CHRs in Fancd2, Fanci and Fancr were highly conserved in the human FANC homologous genes (Supplementary Fig. 15b). Consistent with this, we next found that human p53 activation leads to increased E2F4 binding at the FANCD2, FANCI and FANCR promoters (Supplementary Fig. 16a), and that mutation of the CDE/CHRs in these promoters abolished their p53-dependent regulation (Supplementary Fig. 16b). The p53-dependent downregulation of FANC genes could also be observed in response to DNA damage in MRC5 cells (Supplementary Fig. 17), and we verified that the CDE/CHR motif in FANCD2 is important for its DNA damage-induced downregulation (Supplementary Fig. 18). In addition, the data mining of a transcriptome-wide analysis were again consistent with our results (Supplementary Fig. 19). BLM, DEK, FEN1, TIMELESS and RECQL4 were also downregulated in human cells upon p53 activation, further indicating an overall conservation of the regulatory pathways identified in murine cells (Supplementary Fig. 20).
Further evidence of this conservation was obtained using the Oncomine software (www.oncomine.org). Tumour samples from the Australian Ovarian Cancer Study revealed that the p53 pathway is functional in low-grade ovarian serous tumours, but frequently lost in high-grade ovarian carcinomas. Evidence for this first came from using a transcriptomic signature of p53 target genes47. Formal demonstration was later obtained by TP53 sequencing, which identified p53 mutations in 0% of low-grade serous tumours48 and 96.7% of high-grade carcinomas49. We analysed the transcriptome data of Anglesio et al.47, who characterized 90 ovarian samples from the Australian Ovarian Cancer Study, including 60 high-grade adenocarcinomas. As expected, the expression of genes activated by p53 (CDKN1A/p21, MDM2, DDB2 and SESN1) was decreased in high-grade tumours. On the opposite FANCD2, and other genes known to be repressed by E2F4 in a p53-dependent manner (BIRC5, CDC6 and CDC25C), were more expressed in high-grade tumours (Fig. 7b). Increased FANCD2 expression also correlated with increases in the expression of other FA genes (FANCA, FANCI, FANCJ, FANCR and FANCT), as well as additional genes regulated by p53 in our experiments (BLM, FEN1 and TIMELESS; Fig. 7b). Similar results were obtained when we analysed data from liver cancers (Supplementary Fig. 21) and adrenocortical tumours (Supplementary Fig. 22), providing evidence that human p53 downregulates several genes of the FA pathway in many tissues, and that loss of p53 function leads to an increased expression of FANC genes in advanced human cancers.
We next found that Nutlin sensitized human primary WT cells, but not their p53-deficient counterparts, to MC (Fig. 7c). Likewise, the sensitivity to MC of human cancer cells expressing a WT p53 was markedly increased by Nutlin (Fig. 7d), suggesting a potential therapeutic relevance of our findings.
In this report, we further analysed the consequences of a deletion of the p53 carboxy-terminal domain. Our previous analysis indicated that most p53Δ31/Δ31 mice exhibit a full set of features characteristic of DC. At the molecular level, the increased p53 activity in p53Δ31/Δ31 MEFs correlated with the downregulation of four genes implicated in telomere syndromes: Dkc1, Rtel1, Terf1 and Tinf2 (ref. 2). Here we show that several other genes involved in telomere metabolism are downregulated in p53Δ31/Δ31 cells: Blm, Dek, Fancd2, Fen1, Gar1, Recql4 and Timeless, strengthening the notion that p53 plays a major role in the regulation of telomere metabolism.
Importantly, some of these genes are involved in DNA repair, and we next found p53Δ31/Δ31 cells to exhibit decreased mRNA levels for 11 additional genes mutated in FA, and a reduced capacity to repair DNA interstrand crosslinks. Because DC and FA are both inherited bone marrow failure syndromes in humans, these new findings raised the possibility that an attenuated FA pathway might contribute to the bone marrow failure that affects p53Δ31/Δ31 mice. Importantly, however, mice carrying knocked out alleles of Fanc genes exhibit little or no haematological abnormalities in the absence of additional stress50 (for example, aldehyde-mediated DNA damage51,52), whereas aplastic anaemia occurs spontaneously in mouse models of telomere dysfunction (for example, Pot1b−/− mTR+/− mice53) and in p53Δ31/Δ31 mice2. Furthermore, p53Δ31/Δ31 mouse cohorts of mixed genetic backgrounds previously indicated that a gene linked to the Agouti locus, on chromosome 2, had an impact on their survival2. None of the Fanc genes maps on chromosome 2, whereas mRNA levels for Rtel1, located 26 cM away from Agouti, affected the survival of mutant mice2. Rtel1 encodes a Fancj-like helicase that might participate in DNA repair54, but that mainly acts as a dominant regulator of telomere length55. Accordingly, Rtel1 is mutated in telomere syndromes, including severe DC56,57,58 and pulmonary fibrosis59. Together, these data indicate that telomere dysfunction most likely plays a predominant role in the aplastic anaemia that affects p53Δ31/Δ31 mice.
Interestingly, aplastic anaemia is not the only clinical trait shared by patients with FA and DC: abnormal skin pigmentation, short stature and testicular hypoplasia may affect patients with either syndrome. Furthermore, telomere dysfunction was reported for at least some patients with FA28,60, and cells from patients with DC appeared hypersensitive to MC in a few studies 56,61. In fact, although DC and FA are distinct clinical disorders caused by mutations in different genes, their clinical similarities initially led to some confusion62,63,64,65, and recent evidence of misdiagnosis can still be found occasionally66. As mentioned above, because a defective FA pathway may activate p53 (ref. 46), our results suggest the operation of a positive-feedback loop between p53 and an attenuated FA pathway. Likewise, short telomeres activate p53 (ref. 67), and our data may also suggest a positive-feedback loop between p53 and telomere metabolism. Together, our analyses of p53Δ31/Δ31 mutant cells raise the intriguing possibility that a sustained p53 activation might contribute to the clinical overlap between DC and FA, notably by leading to a concomitant downregulation of genes important for telomere metabolism and genes of the FA DNA repair pathway (Supplementary Fig. 23). Because the p53 pathway is affected by single-nucleotide polymorphisms in many genes including TP53, MDM2, MDM4 and CDKN1A68, we further presume that the strength of the regulatory loops that affect p53, telomere-related and FA genes should vary among humans, and that this might contribute, in patients with identical disease-causing mutations, to the variability in clinical overlap between these syndromes. Independently, our data also provide a rationale for the combination of Nutlin with therapeutic agents inducing DNA interstrand crosslinks, to efficiently kill cancer cells that retain a functional p53 pathway.
Cells and cell culture reagents
MEFs, isolated from 13.5-day embryos, were cultured for ≤6 passages in a 5% CO2 and 3% O2 incubator, in DMEM Glutamax (Gibco), with 15% fetal bovine serum (FBS; Biowest), 100 μM 2-mercaptoethanol (Millipore), 10 μM non-essential amino acids and penicillin/streptomycin (NEAA/PS, Gibco). BMCs were flushed from femurs and tibias of 3-week-old WT and p53Δ31/Δ31 mice. The isolation of MEFs and recovery of BMCs were performed according to Institutional Animal Care and Use Committee (IACUC) regulations, as supervised by the Curie Institute’s Comité d’éthique en expérimentation animale. NIH-3T3 cells were grown in the same conditions as primary MEFs. Human lung fibroblasts MRC5 and their SV40-transformed derivatives (MRC5 SV2, Sigma) were cultured in a 5% CO2 and 3% O2 incubator in minimum essential medium (Gibco), completed with 10% FBS, 2 mM L-glutamine (Gibco), 1 mM pyruvate and 10 μM NEAA/PS. Human colon carcinoma cells HCT116 and their derivatives (HCT116 p53 KO, which do not express p53α), kind gifts from Bert Vogelstein (Johns Hopkins University, Baltimore, MD, USA), were grown in a 5% CO2 incubator in McCoy’s 5A medium with 10% FBS, HEPES and penicilin–streptomycin. Cells were treated for 24 h with 10 μM Nutlin 3a or with 0.5 μg ml−1 doxorubicin before PCR with reverse transcription (RT–PCR) or ChIP assays, or 50 nM MC for 48 h before RT–PCR or metaphase spread preparations.
Total RNA, extracted using Nucleospin RNA II (Macherey-Nagel), was reverse transcribed using Superscript III (Invitrogen). Real-time quantitative PCRs (primer sequences in Supplementary Tables 1 and 2) were performed on ABI PRISM 7500 using Power SYBR Green (Applied Biosystems).
Protein detection by immunoblotting was performed using antibodies raised against Fancd2 (Abcam, ab108928, 1/500 dilution), Fanci (Abcam, ab74332, 1/500), Fancr (Calbiochem, PC130, 1/2,500), E2F4 (Santa Cruz, C-20, 1/200), p53 (Novocastra, CM5, 1/1,000), p21 (Santa Cruz, F-5, 1/250) or actin (Santa Cruz, C-4, 1/5,000) and chemiluminescence revelation was achieved with SuperSignal west dura (Perbio, France). Band quantification was performed using ImageJ, with actin as loading control. Control bands for Fancd2, Fanci and Fancr proteins were obtained using 20 μl lipofectamine 2000 according to the supplier’s procedure to transfect NIH-3T3 cells (a 10-cm dish at 60% confluency) with 3 μg of a pCAGGS expression vector (empty pCAGGS vector, pCAGGS-Fancd2, pCAGGS-Fanci or pCAGGS-Fancr—cloning details upon request), then extracting proteins after 24 h. Uncropped scans of the western blots in Figs 1d and 4b, as well as relevant controls, are presented in the Supplementary Figs 2 and 9, respectively.
ChIP analysis was performed as described69. Briefly, cells were left untreated or treated with Nutlin or doxorubicin for 24 h. Cellular proteins of 107 cells were crosslinked to chromatin with 1% formaldehyde for 10 min at 25 °C. E2F4–DNA complexes were immunoprecipitated from total extracts using an antibody against E2F4 (Santa Cruz, C-20, 30 μg) and 400–500 μg of sonicated chromatin. Rabbit IgG (Abcam) was used for control precipitation. Quantitative PCRs (primer sequences in Supplementary Table 3) were then performed on ABI PRISM 7500.
Cell cycle assays
Log phase cells, treated or not with Nutlin, were incubated for 24 h, then pulse-labeled for 1 h with bromo-deoxy uridine (BrdU) (10 μM), fixed in 70% ethanol, double stained with fluorescein isothiocyanate anti-BrdU and propidium iodide, and sorted using a LSRII cytometer. Data were analysed using FlowJo.
Luciferase expression assays
To construct the Luciferase reporter plasmids, we cloned a 2-kb fragment (for Fancd2) or 1-kb fragment (for Fanci, Fancr, FANCD2, FANCI or FANCR) centred around the transcription start site upstream of the firefly luciferase gene in a pGL3-basic vector (Promega), or a variant fragment generated by PCR mutagenesis of the putative CDE/CHR motif (details on request). Next, 106 NIH-3T3 cells were transfected using lipofectamine 2000 by 3 μg of a Fanc-luciferase reporter plasmid and 30 ng of renilla luciferase expression plasmid (pGL4.73, Promega) for normalization, and treated or not with 10 μM Nutlin 3a or 0.5 μg ml−1 doxorubicin. Transfected cells were incubated for 24 h, then trypsinized, resuspended in 75 μl culture medium with 7.5% FBS and transferred into a well of an optical 96-well plate (Nunc). The dual-glo luciferase assay system (Promega) was used according to the manufacturer’s protocol to lyse the cells and read firefly and renilla luciferase signals. Results were normalized, then the average luciferase activity in cells transfected with a WT Promoter and not treated with Nutlin were assigned a value of 1.
Metaphase spread preparation and analyses
Cells were plated in duplicate, then untreated or treated with 50 nM MC for 48 h, and treated with 0.1 mM nocodazole for 3 h to arrest cells in metaphase. Cells were submitted to hypotonic shock (75 mM KCl), fixed in a (3:1) ethanol/acetic acid solution, dropped onto glass slides and air-dried slides were stained with Giemsa to score for chromosome aberrations. To analyse sister chromatid exchanges, cells plated in duplicate and treated or not with MC were, 1 h after plating, treated with 10 μM (BrdU 1/3 BrdC) for 48 h, then metaphase spreads were prepared as above. Air-dried slides were stained with 10 μg ml−1 Hoescht 33258 for 20 min, submitted to ultraviolet at 365 nm while heated at 55 °C during 30 min, then stained with Giemsa. Images were acquired using a Zeiss Axiophot (X63) microscope.
Cells were spread onto coverslips, treated or not with Nutlin 10 μM, then MC 0.1 μg ml−1 for 1 h, and left to recover for 12 h. Twenty-four hours after Nutlin treatment, cells were fixed and permeabilized. Coverslips were incubated with a Rad51 antibody (Ab-1 Calbiochem) for 1 h at 37 °C in a humid chamber, then with secondary Alexa Fluo anti-rabbit antibody (Invitrogen). Slides were mounted in Vectashield with 0.2 μg ml−1 4,6-diamidino-2-phenylindole. Images were captured on a Zeiss Axioplan2 microscope using equal exposure times for all images.
Cellular sensitivity to mitomycin C
Cells were seeded into wells of a 96-well plate (500 cells per well, in triplicates). After adhesion, cells were treated or not with Nutlin 2.5 μM for 24 h, then with MC for 48 h at 0, 0.01, 0.1 and 1 μg ml−1. Cells were then counted using the CyQUANT kit (Life technologies) and a microplate reader according to the supplier’s recommendations.
Differences between two groups were analysed by Student’s t-test, difference between three groups were analysed by one-way analysis of variance, and values of P≤0.05 were considered significant.
How to cite this article: Jaber, S. et al. p53 downregulates the Fanconi anaemia DNA repair pathway. Nat. Commun. 7:11091 doi: 10.1038/ncomms11091 (2016).
Khincha, P. P. & Savage, S. A. Genomic characterization of the inherited bone marrow failure syndromes. Semin. Hematol. 50, 333–347 (2013).
Simeonova, I. et al. Mutant mice lacking the p53 C-terminal domain model telomere syndromes. Cell Rep. 3, 2046–2058 (2013).
Hamard, P. J. et al. The C terminus of p53 regulates gene expression by multiple mechanisms in a target- and tissue-specific manner in vivo. Genes Dev. 27, 1868–1885 (2013).
Armanios, M. & Blackburn, E. H. The telomere syndromes. Nat. Rev. Genet. 13, 693–704 (2012).
Tummala, H. et al. Poly(A)-specific ribonuclease deficiency impacts telomere biology and causes dyskeratosis congenita. J. Clin. Invest. 125, 2151–2160 (2015).
Holohan, B., Wright, W. E. & Shay, J. W. Telomeropathies: an emerging spectrum disorder. J. Cell Biol. 205, 289–299 (2014).
Wu, Y., Suhasini, A. N. & Brosh, R. M. Jr Welcome the family of FANCJ-like helicases to the block of genome stability maintenance proteins. Cell. Mol. Life. Sci. 66, 1209–1222 (2009).
Kocak, H. et al. Hoyeraal-Hreidarsson syndrome caused by a germline mutation in the TEL patch of the telomere protein TPP1. Genes Dev. 28, 2090–2102 (2014).
Codd, V. et al. Identification of seven loci affecting mean telomere length and their association with disease. Nat. Genet. 45, 422–427 (2013).
Jain, D., Malik, A. A., Kumar, A., Malik, B. K. & Raina, V. Variations in exon-2 of SBDS gene and its association with aplastic anemia. Int. J. Lab. Hematol. 36, e88–e90 (2014).
Venteicher, A. S., Meng, Z., Mason, P. J., Veenstra, T. D. & Artandi, S. E. Identification of ATPases pontin and reptin as telomerase components essential for holoenzyme assembly. Cell 132, 945–957 (2008).
Rooney, S. et al. Defective DNA repair and increased genomic instability in Artemis-deficient murine cells. J. Exp. Med. 197, 553–565 (2003).
Du, X. et al. Telomere shortening exposes functions for the mouse Werner and Bloom syndrome genes. Mol. Cell. Biol. 24, 8437–8446 (2004).
Ting, A. P., Low, G. K., Gopalakrishnan, K. & Hande, M. P. Telomere attrition and genomic instability in xeroderma pigmentosum type-b deficient fibroblasts under oxidative stress. J. Cell. Mol. Med. 14, 403–416 (2010).
Ivanauskiene, K. et al. The PML-associated protein DEK regulates the balance of H3.3 loading on chromatin and is important for telomere integrity. Genome Res. 24, 1584–1594 (2014).
Lin, W. et al. Mammalian DNA2 helicase/nuclease cleaves G-quadruplex DNA and is required for telomere integrity. EMBO J. 32, 1425–1439 (2013).
Munoz, P., Blanco, R., Flores, J. M. & Blasco, M. A. XPF nuclease-dependent telomere loss and increased DNA damage in mice overexpressing TRF2 result in premature aging and cancer. Nat. Genet. 37, 1063–1071 (2005).
Batenburg, N. L., Mitchell, T. R., Leach, D. M., Rainbow, A. J. & Zhu, X. D. Cockayne Syndrome group B protein interacts with TRF2 and regulates telomere length and stability. Nucleic Acids Res. 40, 9661–9674 (2012).
Cottage, C. T. et al. Increased mitotic rate coincident with transient telomere lengthening resulting from pim-1 overexpression in cardiac progenitor cells. Stem Cells 30, 2512–2522 (2012).
Rhee, D. B. et al. FANCC suppresses short telomere-initiated telomere sister chromatid exchange. Hum. Mol. Genet. 19, 879–887 (2010).
Saharia, A. et al. Flap endonuclease 1 contributes to telomere stability. Curr. Biol. 18, 496–500 (2008).
Benson, E. K., Lee, S. W. & Aaronson, S. A. Role of progerin-induced telomere dysfunction in HGPS premature cellular senescence. J. Cell Sci. 123, 2605–2612 (2010).
Hou, Y. Y., Toh, M. T. & Wang, X. NBS1 deficiency promotes genome instability by affecting DNA damage signaling pathway and impairing telomere integrity. Cell Biochem. Funct. 30, 233–242 (2012).
Leman, A. R. et al. Timeless preserves telomere length by promoting efficient DNA replication through human telomeres. Cell Cycle 11, 2337–2347 (2012).
Chawla, R. et al. Human UPF1 interacts with TPP1 and telomerase and sustains telomere leading-strand replication. EMBO J. 30, 4047–4058 (2011).
Crabbe, L., Jauch, A., Naeger, C. M., Holtgreve-Grez, H. & Karlseder, J. Telomere dysfunction as a cause of genomic instability in Werner syndrome. Proc. Natl Acad. Sci. USA 104, 2205–2210 (2007).
Sengupta, S. et al. Tumour suppressor p53 represses transcription of RECQ4 helicase. Oncogene 24, 1738–1748 (2005).
Joksic, I. et al. Dysfunctional telomeres in primary cells from Fanconi anemia FANCD2 patients. Genome Integr. 3, 6 (2012).
Lohr, K., Moritz, C., Contente, A. & Dobbelstein, M. p21/CDKN1A mediates negative regulation of transcription by p53. J. Biol. Chem. 278, 32507–32516 (2003).
Benson, E. K. et al. p53-dependent gene repression through p21 is mediated by recruitment of E2F4 repression complexes. Oncogene 33, 3959–3969 (2014).
Quaas, M., Muller, G. A. & Engeland, K. p53 can repress transcription of cell cycle genes through a p21(WAF1/CIP1)-dependent switch from MMB to DREAM protein complex binding at CHR promoter elements. Cell Cycle 11, 4661–4672 (2012).
Fischer, M., Quaas, M., Wintsche, A., Muller, G. A. & Engeland, K. Polo-like kinase 4 transcription is activated via CRE and NRF1 elements, repressed by DREAM through CDE/CHR sites and deregulated by HPV E7 protein. Nucleic Acids Res. 42, 163–180 (2014).
Fischer, M., Quaas, M., Steiner, L. & Engeland, K. The p53-p21-DREAM-CDE/CHR pathway regulates G2/M cell cycle genes. Nucleic Acids Res. 44, 164–174 (2016).
Mjelle, R. et al. Cell cycle regulation of human DNA repair and chromatin remodeling genes. DNA Repair (Amst) 30, 53–67 (2015).
Meetei, A. R. et al. A multiprotein nuclear complex connects Fanconi anemia and Bloom syndrome. Mol. Cell. Biol. 23, 3417–3426 (2003).
Qian, L., Yuan, F., Rodriguez-Tello, P., Padgaonkar, S. & Zhang, Y. Human Fanconi anemia complementation group a protein stimulates the 5' flap endonuclease activity of FEN1. PLoS ONE 8, e82666 (2013).
Younger, S. T., Kenzelmann-Broz, D., Jung, H., Attardi, L. D. & Rinn, J. L. Integrative genomic analysis reveals widespread enhancer regulation by p53 in response to DNA damage. Nucleic Acids Res. 43, 4447–4462 (2015).
Muntean, A. G. et al. The PAF complex synergizes with MLL fusion proteins at HOX loci to promote leukemogenesis. Cancer Cell 17, 609–621 (2010).
Longerich, S., Li, J., Xiong, Y., Sung, P. & Kupfer, G. M. Stress and DNA repair biology of the Fanconi anemia pathway. Blood 124, 2812–2819 (2014).
Sawyer, S. L. et al. Biallelic mutations in BRCA1 cause a new Fanconi anemia subtype. Cancer Discov. 5, 135–142 (2015).
Hira, A. et al. Mutations in the gene encoding the E2 conjugating enzyme UBE2T cause Fanconi anemia. Am. J. Hum. Genet. 96, 1001–1007 (2015).
Wang, A. T. et al. A dominant mutation in human RAD51 reveals its function in DNA interstrand crosslink repair independent of homologous recombination. Mol. Cell 59, 478–490 (2015).
Crossan, G. P. et al. Disruption of mouse Slx4, a regulator of structure-specific nucleases, phenocopies Fanconi anemia. Nat. Genet. 43, 147–152 (2011).
Montes de Oca Luna, R., Wagner, D. S. & Lozano, G. Rescue of early embryonic lethality in mdm2-deficient mice by deletion of p53. Nature 378, 203–206 (1995).
Bardot, B. et al. Mice engineered for an obligatory Mdm4 exon skipping express higher levels of the Mdm4-S isoform but exhibit increased p53 activity. Oncogene 34, 2943–2948 (2015).
Ceccaldi, R. et al. Bone marrow failure in Fanconi anemia is triggered by an exacerbated p53/p21 DNA damage response that impairs hematopoietic stem and progenitor cells. Cell Stem Cell 11, 36–49 (2012).
Anglesio, M. S. et al. Mutation of ERBB2 provides a novel alternative mechanism for the ubiquitous activation of RAS-MAPK in ovarian serous low malignant potential tumours. Mol. Cancer Res. 6, 1678–1690 (2008).
Hunter, S. M. et al. Molecular profiling of low grade serous ovarian tumours identifies novel candidate driver genes. Oncotarget 6, 37663–37677 (2015).
Ahmed, A. A. et al. Driver mutations in TP53 are ubiquitous in high grade serous carcinoma of the ovary. J. Pathol. 221, 49–56 (2010).
Bakker, S. T., de Winter, J. P. & te Riele, H. Learning from a paradox: recent insights into Fanconi anaemia through studying mouse models. Dis. Model Mech. 6, 40–47 (2013).
Langevin, F., Crossan, G. P., Rosado, I. V., Arends, M. J. & Patel, K. J. Fancd2 counteracts the toxic effects of naturally produced aldehydes in mice. Nature 475, 53–58 (2011).
Garaycoechea, J. I. et al. Genotoxic consequences of endogenous aldehydes on mouse haematopoietic stem cell function. Nature 489, 571–575 (2012).
Hockemeyer, D., Palm, W., Wang, R. C., Couto, S. S. & de Lange, T. Engineered telomere degradation models dyskeratosis congenita. Genes Dev. 22, 1773–1785 (2008).
Uringa, E. J. et al. RTEL1 contributes to DNA replication and repair and telomere maintenance. Mol. Biol. Cell 23, 2782–2792 (2012).
Ding, H. et al. Regulation of murine telomere length by Rtel: an essential gene encoding a helicase-like protein. Cell 117, 873–886 (2004).
Ballew, B. J. et al. A recessive founder mutation in regulator of telomere elongation helicase 1, RTEL1, underlies severe immunodeficiency and features of Hoyeraal Hreidarsson syndrome. PLoS Genet. 9, e1003695 (2013).
Walne, A. J., Vulliamy, T., Kirwan, M., Plagnol, V. & Dokal, I. Constitutional mutations in RTEL1 cause severe Dyskeratosis congenita. Am. J. Hum. Genet. 92, 448–453 (2013).
Le Guen, T. et al. Human RTEL1 deficiency causes Hoyeraal-Hreidarsson syndrome with short telomeres and genome instability. Hum. Mol. Genet. 22, 3239–3249 (2013).
Stuart, B. D. et al. Exome sequencing links mutations in PARN and RTEL1 with familial pulmonary fibrosis and telomere shortening. Nat. Genet. 47, 512–517 (2015).
Leteurtre, F. et al. Accelerated telomere shortening and telomerase activation in Fanconi's anaemia. Br. J. Haematol. 105, 883–893 (1999).
Nagasawa, H. & Little, J. B. Suppression of cytotoxic effect of mitomycin-C by superoxide dismutase in Fanconi's anemia and dyskeratosis congenita fibroblasts. Carcinogenesis 4, 795–799 (1983).
McDonald, R. & Goldschmidt, B. Pancytopenia with congenital defects (Fanconi's anaemia). Arch. Dis. Child. 35, 367–372 (1960).
Bodalski, J., Defecinska, E., Judkiewicz, L. & Pacanowska, M. Fanconi's anaemia and dyskeratosis congenita as a syndrome. Dermatologica 127, 330–342 (1963).
Steier, W., Van Voolen, G. A. & Selmanowitz, V. J. Dyskeratosis congenita: relationship to Fanconi's anemia. Blood 39, 510–521 (1972).
Sirinavin, C. & Trowbridge, A. A. Dyskeratosis congenita: clinical features and genetic aspects. Report of a family and review of the literature. J. Med. Genet. 12, 339–354 (1975).
Ghemlas, I. et al. Improving diagnostic precision, care and syndrome definitions using comprehensive next-generation sequencing for the inherited bone marrow failure syndromes. J. Med. Genet. 52, 575–584 (2015).
Chin, L. et al. p53 deficiency rescues the adverse effects of telomere loss and cooperates with telomere dysfunction to accelerate carcinogenesis. Cell 97, 527–538 (1999).
Grochola, L. F., Zeron-Medina, J., Meriaux, S. & Bond, G. L. Single-nucleotide polymorphisms in the p53 signaling pathway. Cold Spring Harb. Perspect. Biol. 2, a001032 (2010).
Simeonova, I. et al. Fuzzy tandem repeats containing p53 response elements may define species-specific p53 target genes. PLoS Genet. 8, e1002731 (2012).
Lee, B. K., Bhinge, A. A. & Iyer, V. R. Wide-ranging functions of E2F4 in transcriptional activation and repression revealed by genome-wide analysis. Nucleic Acids Res. 39, 3558–3573 (2011).
We thank J. Leemput for technical help in ChIP assays. The ‘Genetics of Tumour Suppression’ laboratory is an ‘Equipe Labellisée Ligue Nationale Contre le Cancer’, with support from the Ligue Headquarters and from the Comité du Cantal. The project was initiated by grants from the Fondation de France (Comité Tumeurs), the Ligue Nationale contre le Cancer (Comité Ile de France) and the Fondation ARC. PhD candidates were supported by fellowships from the Ministère de l’Enseignement Supérieur et de la Recherche (S.J and E.T.), and the Ligue Nationale contre le Cancer (S.J).
The authors declare no competing financial interests.
About this article
Cite this article
Jaber, S., Toufektchan, E., Lejour, V. et al. p53 downregulates the Fanconi anaemia DNA repair pathway. Nat Commun 7, 11091 (2016). https://doi.org/10.1038/ncomms11091