BERKELEY, CA (UroToday.com) - Prostate cancer is a disease with a 5-year relative survival rate of nearly 100% and a 10-year relative survival rate of over 90% in the U.S.
However, these numbers are just the silver lining of a dark and ominous cloud that is prostate cancer management. Due to the high prevalence of the disease, with a low proportion progressing to an aggressive form, coupled with inaccurate screening methods and the inability to accurately predict tumor progression, prostate cancer is a massive financial burden on the U.S. healthcare system. Two large-scale, randomized trials of current screening methods suggest that prostate cancer is overdiagnosed and overtreated; one of these studies indicated that in order to save one life, 1,420 men would need to be screened and 49 surgeries be performed.
To improve the ability to screen patients and avoid unnecessary biopsies, and to improve prognostic power to avoid unnecessary treatment, researchers have turned to characterizing the disease at the molecular level. Thus far the most common molecular event identified is a DNA hypermethylation event in the CpG island near the transcriptional start site of a tumor suppressor gene, GSTP1. Following this discovery, a handful of other DNA methylation alterations have been described, using candidate gene-based approaches. Many of these alterations were tumor specific, while others were implicated in the more aggressive form of the disease. In our recently published study, we took a genomic-scale approach to identify DNA methylation alterations, using the Illumina HumanMethylation27 microarray platform.
To explore the prostate DNA methylome, we profiled 95 primary prostate tumors and 86 benign adjacent prostate tissues, including 70 matched pairs. An analysis of these data led to several key observations. First, hierarchical clustering of the entire data revealed two main clusters, which largely corresponded to the tumor samples and the benign adjacent tissues. Both clusters showed relatively short branch lengths indicating homogeneity within the tumors and within the benign adjacent tissues. Second, 31% of all assayed CpGs showed statistically different methylation between the two sample types (FDR<0.8%). Third, of these 8,063 differentially methylated sites, three-quarters showed increased methylation in tumors while one-quarter showed decreased methylation. Finally, we were able to generate a list of 87 CpG sites whose methylation status could accurately distinguish tumor from benign adjacent tissues. While GSTP1 hypermethylation was part of this list, it was ranked 57th. Therefore we identified 56 mostly previously uncharacterized molecular events that are even better identifiers of prostate cancer than GSTP1.
To search for possible prognostic markers of aggressive tumor, we combined the DNA methylation profiles of the tumors and available clinical data of these patients. We did not observe any methylation change that correlated with Gleason grades of the tumor at the time of surgery. However, we identified 69 DNA methylation alterations that correlated with the patient's time to biochemical recurrence (FDR<26.8%). While we identified these handful of methylation events that may be identifiers of aggressive tumors, this was a strikingly low number compared to the methylation events that separate tumor from adjacent benign, and was consistent with the relative homogeneity among tumors observed in the initial hierarchical clustering.
These data together illustrate an interesting picture of the relationship between DNA methylation alterations and prostate cancer progression. On the one hand, it appears that the global DNA methylome architecture is drastically altered in the tumors with nearly one-third of the CpGs we assayed showing differential methylation. However, the comparison among tumors suggests homogeneity, despite the diverse backgrounds of these patients and histology of their tumors. Based on these observations, we speculate that global DNA methylation alteration is a common and early event. Because of the widespread DNA methylation alterations, we suspected that a disruption in the regulation of DNA methyltransferases (DNMT) and/or their binding partners might provide the mechanism. Real-time quantitative PCR of mRNA samples from the same samples indeed revealed increased transcript levels of three genes involved in DNA methylation, DNMT3A2, DNMT3B and EZH2. Strikingly, the levels of all three of these transcripts had a positive correlation with the methylation levels of the hypermethylated CpGs.
To interrogate the role of DNMT in prostate cancer, we overexpressed these genes in primary prostate cells and obtained their methylation profiles again using the HumanMethylation27 microarrays. The overexpression of DNMT or EZH2 in these cells was able to largely recapitulate the methylation patterns we observed in tumors.
Our data strongly suggest that DNMT and their interacting partners play a major role in the early events of prostate cancer. With increasing number of reports of aberrant DNA methylation in various cancer types, the studying these enzymes, and the consequences of their misregulation, is likely to shed considerable new light on cancer biology in general.
Yuya Kobayashi, PhD as part of Beyond the Abstract on UroToday.com. This initiative offers a method of publishing for the professional urology community. Authors are given an opportunity to expand on the circumstances, limitations etc... of their research by referencing the published abstract.