Regulation of histone H2A.Z expression is mediated by sirtuin 1 in prostate cancer

Abstract

Histone variants seem to play a major role in gene expression regulation. In prostate cancer, H2A.Z and its acetylated form are implicated in oncogenes' upregulation. SIRT1, which may act either as tumor suppressor or oncogene, reduces H2A.Z levels in cardiomyocytes, via proteasome-mediated degradation, and this mechanism might be impaired in prostate cancer cells due to sirtuin 1 downregulation. Thus, we aimed to characterize the mechanisms underlying H2A.Z and SIRT1 deregulation in prostate carcinogenesis and how they interact. We found that H2AFZ and SIRT1 were up- and downregulated, respectively, at transcript level in primary prostate cancer and high-grade prostatic intraepithelial neoplasia compared to normal prostatic tissues. Induced SIRT1 overexpression in prostate cancer cell lines resulted in almost complete absence of H2A.Z. Inhibition of mTOR had a modest effect on H2A.Z levels, but proteasome inhibition prevented the marked reduction of H2A.Z due to sirtuin 1 overexpression. Prostate cancer cells exposed to epigenetic modifying drugs trichostatin A, alone or combined with 5-aza-2'-deoxycytidine, increased H2AFZ transcript, although with a concomitant decrease in protein levels. Conversely, SIRT1 transcript and protein levels increased after exposure. ChIP revealed an increase of activation marks within the TSS region for both genes. Remarkably, inhibition of sirtuin 1 with nicotinamide, increased H2A.Z levels, whereas activation of sirtuin 1 by resveratrol led to an abrupt decrease in H2A.Z. Finally, protein-ligation assay showed that exposure to epigenetic modifying drugs fostered the interaction between sirtuin 1 and H2A.Z. We concluded that sirtuin 1 and H2A.Z deregulation in prostate cancer are reciprocally related. Epigenetic mechanisms, mostly histone post-translational modifications, are likely involved and impair sirtuin 1-mediated downregulation of H2A.Z via proteasome-mediated degradation. Epigenetic modifying drugs in conjunction with enzymatic modulators are able to restore the normal functions of sirtuin 1 and might constitute relevant tools for targeted therapy of prostate cancer patients.

Figure 1. Transcriptional status of SIRT1 and H2AFZ in clinical samples (normal prostate tissues – NPT –, prostatic intraepithelial neoplasia – PIN – and prostate carcinoma - PCa) and PCa cell lines (LNCaP, DU145 and PC-3)

(A) SIRT1 downregulation and H2AFZ overexpression are early events in prostatic carcinogenesis (group analysis with Kruskal-Wallis test followed by a pairwise Mann-Whitney U test, **p<0.01, ***p<0.001 and ****p<0.0001); (B) SIRT1 and H2AFZ gene expression in PCa cell lines follow the tendency observed for clinical samples of prostatic malignancy (mean ± SD, n=3).

Figure 2. Protein profile of three PCa cell lines - LNCaP, DU145 and PC-3 – after (A1) SIRT1 overexpression, (B1) mTOR silencing and (C1) SIRT1 overexpression and exposure to bortezomib

After SIRT1 overexpression levels of (A2) sirtuin 1 increased, with concomitant decrease in (A3) pS6, (A4) c-Myc and H2A.Z. FRAP silencing induced a decrease in (B2) mTOR, (B3) pS6 and (B4) H2A.Z. When SIRT1 overexpressing cells were exposed to bortezomib, the increased levels of (C2) sirtuin 1 caused a significant decrease in H2A.Z levels (C3). However, this decrease was much less impressive than previously observed for SIRT1 upregulation without proteasome inhibition. Results were normalized to β-actin and are presented as fold variation in comparison to control (mean±SD, n=3). Dunnet's test: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 3. Epigenetic modulating drugs positively regulate H2AFZ and SIRT1 transcription

Transcript levels of (A1) H2AFZ and (A2) SIRT1 in three distinct PCa cell lines after exposure to 5-aza-dC and/or TSA. The results are presented as fold variation in comparison to the experimental control (mean±SD, n=3). Dunnet's test: **p<0.01, ***p<0.001, ****p<0.0001. ChIP assay results in LNCaP cell line regarding (B1) H2A, (B2) H2A.Z, (B3) AcH2A.Z, (B4) H3, (B5) AcH3 and (B6) H3K4me2 histones and histones marks across H2AFZ promoter and regarding (C1) H2A, (C2) H2A.Z, (C3) AcH2A.Z, (C4) AcH3, (C5) H3K4me2 and (C6) histones and histones marks along SIRT1 promoter. Results are normalized with the input of total sonicated chromatin (mean ± SD).

Figure 4. Epigenetic modulating drugs have the capacity to upregulate protein levels of sirtuin 1 and decrease levels of H2A.Z and c-Myc

(A) Western blot analysis of sirtuin 1, H2A.Z and c-Myc of three PCa cell lines exposed to 5-aza-dC and/or TSA. Optical densities of visible bands for (B1) sirtuin 1, (B2) H2A.Z and (B3) c-Myc following treatment with 5-aza-dC and/or TSA. Intensity of the bands was normalized to the loading control β-actin and presented as folding variation in comparison to the untreated control (mean ± SD, n=3). Dunnet's test: *p<0.05, **p<0.01, ***p<0.001.

Figure 5. Pharmacological inhibition or activation of sirtuin 1 promotes H2A.Z up- or downregulation

(A1) Western blot analysis of sirtuin 1 and H2A.Z of three PCa cell lines exposed to nicotinamide, alone or combined with 5-aza-dC and/or TSA. Optical densities of visible bands for (A2) sirtuin 1 and (A3) H2A.Z following drug exposure. (B1) Western blot analysis of sirtuin 1 and H2A.Z of three PCa cell lines exposed to resveratrol, alone or combined with 5-aza-dC and/or TSA. Optical densities of visible bands for (B2) sirtuin 1 and (B3) H2A.Z following drug exposure. Intensity of bands was normalized to the loading control β-actin and presented as folding variation in comparison to the untreated control (mean±SD, n=3). Dunnet's test: *p<0.05, **p<0.01, ***p<0.001.

Figure 6. PCa cells exposure to epigenetic modulating drugs and resveratrol promotes the physical and direct interaction between sirtuin 1 and H2A.Z

(A) PLA assay for PC-3 after exposure to 0.5μM TSA, alone or combined with 1μM 5-aza-dC, and 40μM resveratrol (40x magnification). (B) Analysis of PLA results represented as number of signals per cell (mean±SD, n=50). Dunnet's test: ****p<0.0001.

Figure 7. Interplay among sirtuin 1, H2A.Z, mTOR and c-Myc in prostate cancer cells

(A) Under normal conditions, PCa cells display downregulation of SIRT1, thus, sirtuin 1 is unable to promote H2A.Z negative regulation directly, through its degradation, and indirectly, through mTOR pathway repression, leading to upregulation of c-Myc and other oncogenes. (B) When exposed to TSA, alone or combined with 5-aza-dC, or resveratrol, levels of sirtuin 1 are restored, leading to decreased levels of H2A.Z, partially due to mTOR pathway inhibition, but mostly via direct interaction between sirtuin 1 and the histone variant. Similarly, c-Myc levels decrease after treatment with those drugs.