Public Release: 

A red flag for lethal prostate cancer

EZH2 genetic biomarker detects most aggressive cells, say U-M scientists

University of Michigan Health System

ANN ARBOR, MI - High levels of activity in a gene called EZH2 could be a warning sign of metastatic prostate cancer, according to scientists at the University of Michigan's Comprehensive Cancer Center.

In an article published in the Oct. 10 issue of Nature, U-M scientist Arul M. Chinnaiyan, M.D., Ph.D., and his research team used advanced DNA microarray technology to show that EZH2 expression was at "the top of the list" of 55 genes found to be more active in metastatic prostate cancer than in localized prostate cancer. This is the first study linking EZH2 to solid tumors, according to Chinnaiyan.

When the EZH2 gene is active, the cell uses its coded instructions to produce EZH2 protein. U-M scientists believe a future diagnostic test for high levels of this protein could serve as a red flag for physicians and help save the lives of men with the most dangerous form of the disease.

"We found the greatest EZH2 overexpression in metastatic prostate cancer tissue. At this point, it's unclear whether the gene plays a role in cancer's development or is simply an indicator of lethal progression," says Chinnaiyan, an assistant professor of pathology and urology in the U-M Medical School.

EZH2 is one of several related proteins that control a cell's genetic memory and interfere with transcription - the process cells use to transcribe or copy their genetic code. According to Chinnaiyan, it is similar to a gene recently shown to shut down transcription in fruit flies.

If additional research and human clinical trials confirm the U-M results, a test for EZH2 protein could, for the first time, allow physicians to identify accurately those men who need immediate, aggressive treatment to prevent the cancer from spreading outside the prostate. Once prostate cancer metastasizes, or spreads, to other organs, it is usually incurable.

"Over the past 50 years, there has been no significant improvement in clinical outcome for men diagnosed with advanced prostate cancer and no way to tell ahead of time which cancers will spread and which cancers will remain localized," says Mark A. Rubin, M.D., a former U-M faculty member, who is now an associate professor of pathology at Brigham and Women's Hospital and director of the Dana Farber Harvard Cancer Center Tissue Microarray Core. "It is exciting to think that we may have finally found something to help the 30,000 men who die every year from metastatic prostate cancer."

To validate their DNA microarray results, U-M researchers analyzed levels of EZH2 protein in more than 1,000 prostate tissue samples. They included normal prostate tissue, tissue with non-malignant cell changes, and tissue with localized and advanced cancer. Four hundred of these tissue samples were donated by U-M Health System patients who died from hormone-refractory metastatic prostate cancer.

"We never would have discovered this molecule without gene expression tumor profiles from these patients who donated their tissue," Rubin says. "We owe them a real debt of gratitude. Because of them, scientists now can study EZH2 and other important genes that some day will help other patients with prostate cancer."

When U-M scientists compared the prostate tissue samples, the intensity of EZH2 protein staining increased steadily as the samples progressed from benign to clinically localized prostate cancer to metastatic disease. The highest EZH2 concentrations were found in metastatic prostate cancer.

"One of the differences in our study is that we correlated EZH2 expression in prostate cells with clinical outcome," Rubin explains. "We analyzed 278 tissue samples from 64 men for EZH2 protein expression, as well as other common prognostic indicators used by pathologists, such as Gleason score, tumor stage or PSA levels. We found EZH2 protein expression to be significantly better at predicting clinical outcome than any other factor."

Another intriguing finding reported in the Nature paper was the use of a new technique called RNA interference to inhibit production of EZH2 protein in cancer cells. RNA is the messenger molecule that copies DNA's genetic instructions and delivers them to the cell's protein-production factory. In 2001, scientists discovered that short sequences of RNA could be used to shut down the activity of specific genes in mammalian cells.

"We ordered RNA nucleotides specifically targeted for the EZH2 gene and tested them on two different prostate cancer cell lines," Chinnaiyan says. "The first thing we noticed is the cells stopped growing. After 120 hours, 80 percent to 90 percent of the cultured cells containing the RNA nucleotides targeted for EZH2 had stopped dividing. When cells can't divide and grow, they die. This suggests that EZH2 could play an important role in the progression of prostate cancer."

In future research, Chinnaiyan and Rubin hope to discover exactly how the EZH2 gene regulates cell division and whether EZH2-targeted RNA interference will have the same effect on cancer cells in living animals.

Rubin and Chinnaiyan emphasized it is unlikely there will be just one biomarker for prostate cancer. In previous research published in the April 3, 2002 issue of the Journal of the American Medical Association, they described another protein called AMACR, which is overexpressed in several types of cancer. Used in combination with screening tests for AMACR and other biomarkers, they believe a test for EZH2 could help physicians diagnose cases of prostate cancer earlier and determine the most effective and least invasive treatment for each patient.


Research on the genetic and molecular profile of prostate cancer is part of a major initiative underway in the U-M's Comprehensive Cancer Center. Its goal is to link molecular genetics and clinical outcome for all types of cancer.

This research was supported by the National Cancer Institute, CaPCURE and the U-M Prostate Specialized Program of Research Excellence (SPORE), directed by Kenneth Pienta, M.D., a study co-author and professor of internal medicine and surgery in the U-M Medical School. The U-M has applied for a patent on prostate cancer gene expression profiles for future diagnostic and therapeutic use.

Other U-M scientists participating in the study include: Sooryanarayana Varambally, Ph.D., and Saravana M. Dhanasekaran, Ph.D., research fellows; Ming Zhou, M.D., Ph.D., former U-M pathology fellow; Terrence R. Barrette, research associate; Martin G. Sanda, M.D., associate professor of surgery and of internal medicine; and Debashis Ghosh, Ph.D., assistant professor of biostatistics in the U-M School of Public Health.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.