Chemically Silenced Genes Give Clues to Cancer, Lead to New Tests
Bill Nelson, M.D., Ph.D., the Nancy and Jim O’Neal Scholar, has spent much of the last two decades making sense of the subtle, tiny clues that lead to prostate cancer. His pioneering work led to the discovery that a critical gene called GSTPi(pronounced “GST pie”) fails early on in prostate cancer.
In recent years, he has learned why this happens: Because of chemical warfare on a very small scale — a genetic process called methylation. Now, with Srinivasan Yegnasubramanian, M.D., Ph.D., the Dr. and Mrs. Peter S. Bing Scholar, Nelson is discovering new ways to target this process.
GSTPi is a casualty of methylation. A methylated gene is chemically made useless— like a zipper given an extra tooth, or a ball that’s covered with quills like a porcupine.
GSTPi is silenced early on in prostate cancer, for good reason: This gene is one of the good guys. It’s a bodyguard, one of the few defenses standing between the prostate and cancer. With GSTPi out of the way — it is knocked out in more than 90 percent of men with prostate cancer — cancer can proceedmuch more easily.
Nelson and other Brady scientists, including Don Coffey and Angelo De Marzo, have made many groundbreaking discoveries involving the epigenetics — small but significant changes, such as methylation, in gene expression — of prostate cancer. Looking at methylation changes in genes, they have found other important landmarks for cancer — particularly, abnormal clumps of DNA called “hypermethylated CpG islands.” These appear on GSTSTP1 before the gene is silenced. On the other hand, hypomethylated (undermethylated)
DNA is very active, interacting with many different proteins. (Note: Their work has been reported in previous issues of Discovery, and is available at our website: http://urology.jhu.edu).
In prostate cancer, it seems that methylation
is all over the map: “By the time most cancers become life-threatening, there is increased methylation in some regions (the hypermethylated CpG islands in GSTSTP1 and other genes), and decreased methylation in others,” says Nelson. He and colleagues recently carried out a definitive analysis of hypomethylation, published in November in Cancer Research.
In sophisticated gene-profiling research, Nelson, Yegnasubramanian, and colleagues isolated genes from prostate tissue that were found only in cancer, not in normal cells. Of these, they identified several genes that were undermethylated, which were expressed at high levels in prostate cancer cells. “Some of these genes are already being targeted by anti-cancer vaccines now in clinical trials,” Nelson says.
This research also seems to have given some chronological order to the methylation changes that can be found in the cells of men with prostate cancer. Hypermethylation happens earlier, and is seen in localized, easily curable cancer. But the presence of hypomethylation is a bad sign, Nelson says. “This appears when cancer is more advanced, and is likely to have spread to distant
Based on Brady methylation research, a new test looks at cancer-negative prostate biopsies, and predicts the likelihood that a future biopsy might show the presence of cancer.
The “Melting Pot” of Advanced Cancer
As cancer matures, like a good stock portfolio, it diversifies. Instead of one or a few kinds of cells — which are much easier to target and cure — it is a malignant mix of cells, a bad melting pot. “This is called tumor heterogeneity,” explains Yegnasubramanian, “and it makes it very difficult to develop targeted therapies aimed at killing so many different kinds of tumor cells. But now that we have discovered that hypomethylation can be a cause of this heterogeneity, we can develop new weapons to help us control these genes.”
Nelson, Yegnasubramanian, and colleagues including Alan Partin and Bruce Trock are translating what they’ve learned about DNA methylation changes into clinically useful tests. Based on the Brady research, a company called LabCorp has produced a GSTP1 methylation test that looks at cancer-negative prostate biopsies, and predicts the likelihood that a future biopsy might show the presence of cancer. “For the future,” says Nelson, “we hope to identify DNA methylation changes that can be easily detected in blood or urine. We also hope to be able to stratify a man’s risk of prostate cancer into groups, and even to predict which treatments will work best.”