Specific sites of lysine methylation on histones correlate with either activation or repression of transcription1,2,3. The tumour suppressor p53 (refs 4–7) is one of only a few non-histone proteins known to be regulated by lysine methylation8. Here we report a lysine methyltransferase, Smyd2, that methylates a previously unidentified site, Lys 370, in p53. This methylation site, in contrast to the known site Lys 372, is repressing to p53-mediated transcriptional regulation. Smyd2 helps to maintain low concentrations of promoter-associated p53. We show that reducing Smyd2 concentrations by short interfering RNA enhances p53-mediated apoptosis. We find that Set9-mediated methylation of Lys 372 inhibits Smyd2-mediated methylation of Lys 370, providing regulatory cross-talk between post-translational modifications. In addition, we show that the inhibitory effect of Lys 372 methylation on Lys 370 methylation is caused, in part, by blocking the interaction between p53 and Smyd2. Thus, similar to histones, p53 is subject to both activating and repressing lysine methylation. Our results also predict that Smyd2 may function as a putative oncogene by methylating p53 and repressing its tumour suppressive function.
Members of the Smyd (SET and MYND domain) family and Suv4-20 were expressed in bacteria and tested for histone and p53 methylation. Smyd2, but not Smyd3, Smyd5 or Suv4-20, methylated a peptide spanning amino acid residues 358–393 in the carboxy terminus of p53 (Supplementary Fig. S1a). Peptide mapping showed that lysine 370 (K370) is the methylation site for Smyd2 (Fig. 1a, b). Unmodified, mono-, di- and tri-methylated K370 p53 peptides (residues 361–380) were subjected to Smyd2 methylation in vitro followed by mass spectrometry (Supplementary Fig. S1b). The results showed that Smyd2 mono-methylates K370 in p53 in vitro.
Polyclonal antibodies specifically recognizing mono-methylated (p53K370me1), di-methylated (p53K370me2) and tri-methylated (p53K370me3) K370 were generated and tested for specificity in vitro (Supplementary Fig. S2a). These antibodies were used to test p53 methylation by Smyd2 in vivo. H1299 cells were transfected with Flag–p53, Flag–p53(K370A) or Flag–p53(K370R) (Fig. 2a and Supplementary Fig. S2b), either with or without Flag–Smyd2 or a methylation-defective mutant Flag–Smyd2(MD) (Supplementary Fig. S3). Flag immunoprecipitates were subjected to western blot analysis with methylation-specific antibodies. Flag–p53, but not the K370R mutant, was detected by the antibody to p53K370me1 (Fig. 2a, compare lanes 1 and 4). The extent of K370 methylation of Flag–p53, but not the K370R mutant, was greatly increased by cotransfection with Flag–Smyd2 (Fig. 2a, compare lanes 2 and 5). Detection of p53 by the antibody to p53K370me1 was greatly diminished by cotransfection with Flag–Smyd2 (MD) (Fig. 2a, compare lanes 2 and 3). The antibody to p53K370me2 recognized p53 largely without regard to its methylation status at K370, whereas the antibody to p53K370me3 did not detect any tri-methylation signal (data not shown).
The methylation status of Flag–p53 in H1299 cells was also tested by mass spectrometry (Fig. 2b). Flag–p53 was immunoprecipitated and gel-isolated after Coomassie blue staining. The results showed that p53 is mono-methylated by Smyd2 in vivo (Fig. 2b).
We analysed methylation of p53 in U2OS cells. Cotransfected Flag–p53 and Flag–Smyd2 yielded mono-methylated p53 (Fig. 2c), but not di- or tri-methylated p53 (data not shown). The ability of endogenous Smyd2 to methylate endogenous p53 was tested. The amount of p53K370me1 was reduced when cells were treated with a short interfering RNA (siRNA) targeting Smyd2 as compared with a control siRNA (Fig. 2d; see Supplementary Fig. S4b, c, for siRNA controls). These in vivo data support our in vitro findings that Smyd2 mono-methylates p53 at K370.
To address the role of Smyd2 in regulating the function of p53, we used the human fibroblast cell line BJ-DNp53, which stably expresses a dominant-negative variant of p53 to inactivate the function of endogenous p53 (ref. 9). Reducing Smyd2 concentration with Smyd2 siRNA correlated with increased p21 and mdm2 expression in control BJ cells but not in BJ-DNp53 cells (Fig. 3a and Supplementary Fig. S4a) with or without irradiation. Smyd2 reduction also resulted in an increase in mRNA (Supplementary Fig. S4b) and protein (Supplementary Fig. S4c) concentrations of p21 and mdm2 in U2OS cells. However, this effect was muted in H1299 cells, which are p53 null (Supplementary Fig. S4c). Taken together, these results show that Smyd2 negatively regulates p53-responsive genes in a p53-dependent manner.
To determine whether Smyd2 regulates the function of p53 specifically through K370 methylation, wild-type p53, p53(K370R) or an empty expression vector was transfected into H1299 cells stably expressing control or Smyd2 short hairpin RNA (shRNA; Fig. 3b). Wild-type p53 induced p21 expression to a greater extent in cells expressing Smyd2 shRNA than in cells expressing control shRNA. However, p53(K370R) showed the same transcriptional ability in both cell lines, and this ability matched the higher activity of wild-type p53 in H1299 cells expressing Smyd2 shRNA (Fig. 3b). These findings indicate that Smyd2 may downregulate the transcriptional activation ability of p53 through K370 methylation.
We investigated whether the p53K370me1 modification occurs on p53 associated with cognate genes. Double chromatin immunoprecipitation (ChIP) was done with Flag antibody for the first ChIP, followed by Flag peptide elution, and then p53K370me1 antibody for the second ChIP. The total p53 signal decreased when Smyd2 was cotransfected, whereas the p53K370me1 ChIP essentially remained the same (data not shown). Thus, the ratio of p53K370me1 signal to total p53 signal increased in response to cotransfected Smyd2 (Fig. 3c, compare lanes 1 and 2). This increase was eliminated in the Smyd2(MD) mutant (Fig. 3c, compare lanes 2 and 3). The higher ChIP signal required intact K370, because the increase was not seen in Flag–p53(K370R) immunoprecipitates (Fig. 3c, compare lanes 1–3 and lanes 4–6). The higher signal of p53K370me1, as compared with IgG, arising from p53(K370R) immunoprecipitates presumably results from the residual recognition of unmodified p53 by the methylation antibody when p53 is overexpressed (Fig. 3c, lanes 4–6).
The endogenous amounts of total p53 and K370-methylated p53 bound to the p21 promoter were assessed by ChIP assay in U2OS cells (Supplementary Fig. S5a). Although the promoter-associated amount of total p53 increased markedly on adriamycin treatment, K370-methylated p53 increased only slightly (Supplementary Fig. S5a), resulting in a decrease in the relative percentage of K370-methylated p53 associated with the p21 promoter (Supplementary Fig. S5b). These data further support the view that Smyd2-mediated K370 methylation is a repressive modification of p53.
We investigated the mechanism underlying the repressive effect of Smyd2-mediated K370 methylation. We found no change in the distribution of p53 between the nucleus and cytoplasm when Smyd2 was reduced (Supplementary Fig. S6). In addition, we detected no change in the amount of total p53 (Supplementary Figs S4c, S7 and S8d), K382-acetylated p53 and S15-phosphorylated p53 (Supplementary Fig. S7) when Smyd2 was reduced or increased.
Next, we examined the effect of increasing Smyd2 concentrations on p53 binding at the p21 gene by ChIP using U2OS cell lines inducibly expressing either empty vector or Smyd2 (Supplementary Fig. S8a). Increasing Smyd2 lowered the amount of p53 bound to the p21 promoter (Supplementary Fig. S8b). The decreasing amount of promoter-associated p53 correlated with decreasing amounts of p21 mRNA (Supplementary Fig. S8c) and p21 protein (Supplementary Fig. S8d). We also reduced Smyd2 by siRNA and observed an increase in promoter-associated p53 (Fig. 3d). These data suggest that K370-methylation of p53 reduces the DNA-binding efficiency of p53.
To address the role of Smyd2 in p53-mediated cell-cycle arrest and apoptosis, we carried out flow cytometry analysis in U2OS cells stably expressing either control or Smyd2 shRNA with or without adriamycin treatment (Fig. 3e). Reduction of Smyd2 did not cause obvious changes in the apoptotic fraction of cells in U2OS cells under the non-DNA damage condition (Fig. 3e, top). However, 24 h after DNA damage, fivefold more cells expressing Smyd2 shRNA (10.3%) entered apoptosis (sub-G1) than did cells expressing control shRNA (2.5%; Fig. 3e, middle). This difference remained after 36 h of adriamycin treatment (Fig. 3e, bottom).
To address whether the effect of Smyd2 on apoptosis is p53 dependent, we reduced the amounts of p53 protein and p53 mRNA in U2OS cells stably expressing control or Smyd2 shRNA (Supplementary Figs S9 and S4c). Ablation of p53 almost completely eliminated the difference in apoptosis between U2OS cells expressing control shRNA and those expressing Smyd2 shRNA (Fig. 3f). Because reintroduction of p53 into H1299 is known to induce apoptosis10, we transfected an empty or p53 expression vector into H1299 cells expressing control or Smyd2 shRNA cells and subjected them to γ-irradiation (Supplementary Fig. S10). We found ectopically expressed p53 caused modest cell apoptosis without DNA damage (Supplementary Fig. S10, top, compare lanes 3 and 4 to lanes 1 and 2). On DNA damage, the addition of p53 resulted in more apoptosis in H1299 cells expressing Smyd2 shRNA than in those expressing control shRNA (Supplementary Fig. S10, bottom, compare lanes 4 and 3). We were unable to use this p53 overexpression approach in H1299 cells to test further the prediction that the K370R mutant would cause even more cell death, because the apoptosis caused by overexpressed p53 was already severe. Taken together, we conclude that Smyd2 is involved in p53-dependent cell-cycle arrest and apoptosis.
We examined the possible interplay between the activating modification p53K372me1 (ref. 8) and the repressing modification p53K370me1. We first tested whether there is any mutual effect on enzyme activity by using peptides and recombinant enzymes. Methylation by Set9 was not altered on peptides that were mono- or di-methylated at K370 (Fig. 4a, top). We also tested the ability of Set9 to mono-, di- or tri-methylate p53 peptides at K372, an issue that has not been reported8. We found that Set9 can mono- and di-methylate p53 at K372 in vitro (Supplementary Fig. S10a), and that methylation by Smyd2 is lowered on both K372me1 and K372me2 peptides as compared with unmodified peptides (Fig. 4a, bottom). Taken together, our in vitro data suggest a ‘one-way’ cross-talk, in which K372 methylation inhibits K370 methylation.
We therefore tested whether there is a mutual effect in vivo. Whereas higher concentrations of p53K372me1 resulted in greatly reduced amounts of p53K370me1, higher concentrations of p53K370me1 did not lower the basal amounts of p53K372me1 (Fig. 4b, top). Furthermore, even the great increase in p53K372me1 obtained by cotransfection with Set9 was not lowered by Smyd2 cotransfection (Fig. 4b, bottom).
A trivial explanation for the observed Set9/K372me1 inhibition of p53K370me1 could be physical blocking of the p53K370me1 antibody by methylation at the neighbouring K372 residue. To test this possibility, we synthesized peptides bearing the double modification K370me1 and K372me1. We found that detection of K370me1 was comparable to that of the singly modified K370me1 peptide (Supplementary Fig. S10b).
We also found that higher concentrations of p53K370me1 occur when Set9 is decreased by siRNA in H1299 cells that ectopically express p53 (Fig. 4c). Finally, we found that lowering endogenous Set9 increases endogenous amounts of p53K370me1 in U2OS cells (Fig. 4d). On the basis of the in vitro and in vivo results, we conclude that Smyd2-mediated methylation of K370 is inhibited by Set9-mediated methylation of K372.
We considered that the inhibitory effect of p53K372me1 on p53K370me1 might be caused by blocking the interaction between p53 and Smyd2. We found that Flag–p53 coprecipitated Smyd2 in H1299 cells (Fig. 4e). Cotransfection of Set9 decreased the interaction between p53 and Smyd2 (Fig. 4e, compare lanes 3 and 4). Therefore, Set9 prevents Smyd2 from binding to p53.
Our results suggest that there is an equilibrium between promoter-bound and free p53 (Fig. 4f): Smyd2-mediated methylation of K370 shifts the equilibrium towards dissociation of p53 from DNA, whereas Set9-mediated methylation of K372 enhances the association of p53 with promoter by blocking Smyd2-mediated methylation of K370 and promotes activation of the p21 and mdm2 genes.
Histone modification cross-talk occurs, such that methylation of histone H3 on K9 (H3-K9) precludes H3-K4 methylation or phosphorylation of histone H3 on S10 (H3-S10)11,12. We have shown here that the activating modification p53K372me1 causes interference to the repressing modification p53K370me1, which is reminiscent of the finding that Set9-mediated methylation of H3-K4 precludes Suv39h1-mediated methylation of H3-K9 (ref. 13). This non-proteolytic negative regulation through methylation of K370 may provide a pool of p53 protein to permit a quick response to DNA damage or cell stress.
Note added in proof: Smyd2 was recently shown to methylate histone H3 K36 (ref. 15).
Histone methyltransferase assay
Histone methyltransferase (HMTase) assays were done as described8. In brief, 2 μg of peptide was incubated with 2 μg of recombinant enzyme and 1.1 μCi of S-adenosyl-methionine at 30 °C for 30 min.
Cells were transfected with Lipofectamine 2000 (Invitrogen) for plasmid and with DharmaFECT 1 (Dharmacon) for siRNA (see Supplementary Information for details).
RT–PCR and real-time PCR
ChIP and sequential ChIP assays
We carried out ChIP assays as described14. Real-time PCR was used to measured signals in input material and immunoprecipitates. The percentage of input was calculated as the immunoprecipitate signal over the input signal. Sequential ChIP assays are described in the Supplementary Information.
Lentiviral RNAi and expression systems were purchased from Invitrogen. shRNA oligonucleotide sequences were designed with BLOCK-iT RNAi Designer and are given in the Supplementary Information. Lentivirus bearing p53 shRNA was purchased from Sigma.
Zhang, Y. & Reinberg, D. Transcription regulation by histone methylation: interplay between different covalent modifications of the core histone tails. Genes Dev. 15, 2343–2360 (2001)
Lachner, M., O’Sullivan, R. J. & Jenuwein, T. An epigenetic road map for histone lysine methylation. J. Cell Sci. 116, 2117–2124 (2003)
Schotta, G. et al. A silencing pathway to induce H3–K9 and H4–K20 trimethylation at constitutive heterochromatin. Genes Dev. 18, 1251–1262 (2004)
Olivier, M., Hussain, S. P., Caron de Fromentel, C., Hainaut, P. & Harris, C. C. TP53 mutation spectra and load: a tool for generating hypotheses on the etiology of cancer. IARC Sci. Publ. 157, 247–270 (2004)
Vogelstein, B., Lane, D. & Levine, A. J. Surfing the p53 network. Nature 408, 307–310 (2000)
Levine, A. J. p53, the cellular gatekeeper for growth and division. Cell 88, 323–331 (1997)
Prives, C. & Hall, P. A. The p53 pathway. J. Pathol. 187, 112–126 (1999)
Chuikov, S. et al. Regulation of p53 activity through lysine methylation. Nature 432, 353–360 (2004)
Vaziri, H. & Benchimol, S. Reconstitution of telomerase activity in normal human cells leads to elongation of telomeres and extended replicative life span. Curr. Biol. 8, 279–282 (1998)
Luo, J. et al. Negative control of p53 by Sir2α promotes cell survival under stress. Cell 107, 137–148 (2001)
Wysocka, J., Myers, M. P., Laherty, C. D., Eisenman, R. N. & Herr, W. Human Sin3 deacetylase and trithorax-related Set1/Ash2 histone H3–K4 methyltransferase are tethered together selectively by the cell-proliferation factor HCF-1. Genes Dev. 17, 896–911 (2003)
Rea, S. et al. Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 406, 593–599 (2000)
Nishioka, K. et al. Set9, a novel histone H3 methyltransferase that facilitates transcription by precluding histone tail modifications required for heterochromatin formation. Genes Dev. 16, 479–489 (2002)
Kent, J. R. et al. During lytic infection herpes simplex virus type 1 is associated with histones bearing modifications that correlate with active transcription. J. Virol. 78, 10178–10186 (2004)
Brown, M. A., Sims, R. J., Gottlieb, P. D. & Tucker, P. W. Identification and characterization of Smyd2: a split SET/MYND domain-containing histone H3 lysine 36-specific methyltransferase that interacts with the Sin3 histone deacetylase complex. Mol. Cancer 5, 26 (2006)
We thank N. Barlev for the Set9 expression vector and D. Reinberg for the p53K372me1 antibody; T. Waibel for assistance in cloning Smyd2; S. Benchimol for the BJ and BJ-DNp53 cell lines; and members of the T.J. and S.L.B. laboratories for discussions. Research in the laboratory of T.J. is supported by the IMP through Boehringer Ingelheim and by grants from the European Union and the Austrian GEN-AU initiative, which is financed by the Austrian Ministry of Education, Science and Culture. Research support to S.L.B. was provided by a grant from the NIH. B.J.P. was supported by a Wistar Cancer Training Grant.
About this article
Cell Death & Disease (2019)
Nature Communications (2019)
Nature Reviews Molecular Cell Biology (2019)
Cell Death & Disease (2019)