Abstract
Prostate cancer is the second most common cause of cancer and the sixth leading cause of cancer fatalities in men worldwide (Ferlay et al., 2010). Genetic abnormalities and mutations are primary causative factors, but epigenetic mechanisms are now recognized as playing a key role in prostate cancer development. Epigenetics is defined as the study of mitotically and/or meiotically heritable changes in gene function that do not involve a change in DNA sequence (Dupont et al., 2009). Epigenetics encompasses several processes, such as DNA promoter methylation, histone modifications and, recently, changes in microRNA expression. Here we focus on histone modifications, specifically the H3K27me3 biomarkers. H3K27me3 is carried out by enzyme EZH2 that interacts with target gene promoters, leading to transcriptional repression in prostate cancer. The purpose of this study was to map H3K27me3 biomarkers in peri-tumoral tissue (PTT) from prostate cancer patients compared to normal tissue. We assessed the proportion of H3K27me3 biomarkers in six genes involved in prostate cancer: RARβ2 (retinoic acid receptor β2), RGMA (repulsive guidance molecule A), ERα (estrogen receptor α), PGR (progesterone receptor), EZH2 (enhancer of zeste homolog 2), and SRC3 (steroid receptor co-activator 3).
Chromatin immunoprecipitation (ChIP) was performed as previously described (Dagdemir et al., 2013) on 45 human prostate biopsies to assess variation in the repressive biomarkers H3K27me3 between 22 PTT biopsies from prostate cancer patients and 23 normal biopsies. Table 1 reports the clinicopathological characteristics of patients studied.
NM, nonmalignant tissues; PSA: prostate-specific antigen; PTT, peri-tumoral tissues.
All 45 tissues were ground down for chromatin extraction and shearing. ChIP was performed using an Auto ChIP kit (Diagenode) following the manufacturer's instructions on a Diagenode SX-8G IP-Star® Automated System. Specific target reactions used rabbit anti-human H3K27me3 antibody (Cat. No. pAb-069-050, Diagenode). The negative control contained nonimmune rabbit IgG (Cat. No: kch-504-250, Diagenode).
TaqMan quantitative real-time PCR was then performed using ChIP products or total DNA (input) in a 25 μL reaction containing 1X TaqMan Mix (Applied Biosystems) and 400 nM of each forward and reverse primer, 250 nM of probe, and 4.25 μL water. QPCR was carried out in triplicate for the target genes EZH2, RARβ2, PGR, ERα, RGMA, and SRC3. Table 2 reports the oligonucleotide primers and probes used.
ChIP efficiency on each genomic locus was calculated from qPCR data and reported as percentage of starting materiel. Data were analyzed by Student's t-test to assess variation in the expression of H3K27me3 biomarkers between PTT and normal prostate tissues.
The results (Fig. 1) showed different levels of H3K27me3 biomarkers for each gene. For EZH2, there was no significant difference (Student's t-test; p=0.83) between PTT and nonmalignant (NM) tissue. This suggests that H3K27me3 modification does not regulate EZH2 expression in normal or peri-tumoral tissue (PTT). Conversely, there was significantly more H3K27me3 biomarkers on RARβ2 gene in PTT than normal prostate tissue (p=4.62337E-06). However, patients with a high percentage of H3K27me3 had a Gleason score of 8 synonymous with advanced disease.
ChIP analysis of H3K27me3 enrichment in nonmalignant (NM; n=23) and peri-tumoral tissues (PTT; n=22). Mean H3K27me3 levels on RARβ2, ERα, PGR, and RGMA were found to be lower in PTT than NM tissues. H3K27me3 levels on SRC3 and EZH2 in PTT showed no statistically significant difference compared to normal tissue. *p<0.05. Black bars indicate means between biopsies. The y-axis represents percentage of total input.
H3K27me3 biomarkers were also significantly higher on ERα (p=1.5726E-06) and PGR (p=3.91452E06) in PTT than in normal tissue, and the difference correlated positively with Gleason score. Likewise, RGMA gene showed significantly higher H3K27me3 levels (p=1.143E-05) in PTT than in normal tissue, but three distinct groups emerged with different Gleason scores: patients with a high proportion of H3K27me3 biomarkers had a higher Gleason score of ≥8, whereas patients with a low proportion of H3K27me3 biomarkers had a Gleason score of ≤6. Patients with intermediate levels had a Gleason score=7. Finally, H3K27me3 biomarkers in SRC3 were at similar levels in PTT and NM, with the difference remaining nonsignificant. These results suggest that H3K27me3 histone modifications can explain gene silencing in prostate cancer.
Although DNA methylation plays a key role in gene repression, histone modification—explicitly methylation and acetylation—also plays an essential role in many molecular and cellular alterations associated with the development and progression of prostate cancer (Bannister and Kouzarides, 2004). The change in histone methylation (hyper- or hypo-) is associated with a change in gene expression regulation (up or down). In prostate cancer, EZH2 is upregulated at both transcriptional and protein levels. Previous studies were able to map H3K27me3 modifications in both prostate cancer and normal cell lines. However, in prostate cancer cells, H3K27me3 was shown to silence the expression of certain genes by EZH2-mediated H3K27me3, independently of gene promoter methylation (Ke et al., 2009), compared to normal cells. Here, we also demonstrated that H3K27me3 biomarkers increased in prostate cancer, as H3K27me3 was significantly higher in PTT than NM tissue. Further investigation of epigenetic changes in PTT could help identify potential markers for diagnosis and prognosis and potential targets for prostate cancer treatment. The development of histone methyltransferase inhibitors therefore looks a promising way forward. Note that 3-dezaneplanocin A (DZNep), an S-adenosylhomocysteine hydrolase inhibitor that indirectly inhibits EZH2, decreased H3K27me3 biomarkers and induced apoptosis in prostate cancer cells.
Although highly promising, these results do come with certain limitations. The small number of samples, with only 22 PTT biopsies, means that statistical power remains limited, which could explain why we did not find a great difference between grades 7 and 8, due to the limited availability of advanced-stage tissue. It would be interesting to increase the sample size in each patient group in order to validate the prognostic power of histone modification in prostate cancer.
Footnotes
Acknowledgments
This study was supported by “La Ligue contre le Cancer—Comités de la région Auvergne”. Aslihan Dagdemir received a grant from Protema Saglik Hizm A.S (Istanbul, Turkey).
Author Disclosure Statement
The authors have no competing financial interests to declare.