ANN ARBOR, MI - Like a killer charged with more than one murder, a tiny protein that has already been linked to deadly prostate cancer is now being implicated in lethal breast cancer. And it may soon help doctors tell cancer patients just how dangerous their tumors are.
The double-duty offender, called EZH2, appears to help cancer cells invade nearby tissue and form colonies, according to a new study in the Proceedings of the National Academy of Sciences. This makes it crucial for aggressive, metastatic forms of breast and prostate cancers, both of which are regulated by steroid hormones.
But like many killers who get caught, EZH2 also leaves "fingerprints" -- copies of the protein that can easily be detected in cancerous tissue. In the new study, led by scientists from the University of Michigan Comprehensive Cancer Center, levels of EZH2 in a patient's tumor corresponded to the tumor's level of danger. The more EZH2 there was, the deadlier the cancer and the worse the patient's outcome.
The new results linking EZH2 to breast cancer, and describing the mechanism of its role in cancer's spread, are based on exhaustive analysis of tissue samples and medical records from 280 U-M breast cancer patients, and studies in cell cultures.
The results will be published this week in the online version of PNAS by a team from U-M, the University of Amsterdam and Harvard Medical School. Members of the team previously reported EZH2's role in aggressive and metastatic prostate cancer.
"EZH2 may serve as an excellent biomarker for determining a breast cancer patient's prognosis more precisely than current methods," says lead author Celina G. Kleer, M.D., an assistant professor of pathology at the U-M Medical School. "Just as with prostate cancer, its association with the severity of a patient's disease, and clinical outcome, is striking. And it is easy to detect with an antibody stain."
The actual use of EZH2 as a clinical tool is still several years away, though the U-M team is planning a prospective clinical trial to test its use in breast cancer patients.
If validated, the EZH2 test may join other tools that physicians use to determine how aggressively to treat breast cancer patients. These include the staging system that relies on tumor size and presence of cancer cells in nearby lymph nodes, and tests for estrogen and progesterone receptors that determine whether a cancer might be susceptible to certain drugs. But all these tests, like other protein biomarkers that have been considered for breast cancer, don't consistently give a precise prognosis.
Because EZH2 can be stained using an antibody specific to the protein, it can clearly be seen both in conventional slides of tissue and in tiny samples arranged in grids called microarrays.
Using sophisticated high-density tissue microarrays, Kleer and her colleagues were able to determine the levels of EZH2 protein in normal and cancerous tissue samples from the U-M Health System's breast cancer library. The samples came from breast cancer patients with a broad range of cancer types and stages, and the researchers were also able to cross-reference the levels of EZH2 for 194 of the patients with information on their post-treatment survival time.
Tissue from invasive breast cancers, which had spread to blood vessels or lymph nodes, had the highest levels of EZH2, although some invasive cancers did not have extra EZH2. Cancer metastases, or satellite tumors that formed elsewhere in the bodies of breast cancer patients, had the highest levels of EZH2 -- just as prostate cancer metastases had the highest levels in previous studies.
Normal tissue and mild forms of breast abnormalities (atypical hyperplasia and ductal carcinoma in situ) had much lower levels of the protein, although DCIS, which is often a precursor to invasive cancer, did have some EZH2.
The team checked their results against an existing database of gene expression rates in a certain type of invasive breast cancer. They found that levels of EZH2 messenger RNA, a measure of EZH2 gene activity, were much higher in samples of tissue from cancers that spread within five years of diagnosis, when compared with cancers that did not metastasize.
When the researchers looked at how well EZH2 levels correlated with a patient's clinical outcome, they found a clear connection. Of the 194 patients for whom survival data were available, 21.6 percent had died of breast cancer. Based on this, the researchers calculated that the expected five-year survival rate for the entire group was 60 percent, and the ten-year survival was 39 percent.
They also recorded how well EZH2 protein levels correlated with existing measures of cancer severity, including tumor diameter, stage of disease, lymph node status, and estrogen and progesterone receptor status.
In all, Kleer says, the link was clear. "We found a strong association between EZH2 protein levels and patient outcome. Patients with higher levels had a shorter disease-free interval after surgery, lower overall survival, and a higher probability of death due to breast cancer."
At the end of 10 years, about 25 percent of the patients with high EZH2 levels were alive, while 59 percent of those with low levels were still living.
In addition to the studies of human breast cancer tissue, the researchers performed laboratory experiments using cell cultures to look at the actual cellular mechanism that EZH2 helps control. They inserted the EZH2 gene into an adenovirus, and a modified form of the gene into a different virus, and infected normal immortalized breast cells with them. Cells that received the extra EZH2 gene produced more of the EZH2 protein, but did not divide any faster than those that got the altered gene.
However, the EZH2-enhanced cells formed colonies, while the other cells did not -- and colony formation is important to help cancer get a foothold outside of its original location. And, EZH2-enhanced cells successfully "invaded" cell membranes, while others did not.
The researchers believe that EZH2 achieves these effects by an action called "gene silencing", which controls a cell's genetic memory and interferes with gene transcription.
In the new study, EZH2 over-expression increased the activity of an enzyme called histone deacetylase, or HDAC. Previous studies also found that inhibition of HDAC interfered with EZH2's action. Since some anti-cancer drugs can inhibit HDAC, Kleer and her colleagues note that those drugs might be effective against tumors with high levels of EZH2.
In all, Kleer says, EZH2 appears to be a promising biomarker that could one day be used to help determine a patient's risk of cancer progression. "If our work is confirmed through carefully controlled clinical trials, testing for EZH2 would be a relatively straightforward and feasible way to judge a patient's prognosis and help determine her best course of treatment," she says. "For tens of thousands of women a year, that could mean a lot."
In addition to Celina Kleer, the study's authors are: senior author Arul Chinnaiyan, M.D., Ph.D., assistant professor of pathology and urology at the U-M; U-M co-lead authors Qi Cao and Sooryanarayana Varambally; U-M co-authors Ronglai Shen, Ichiro Ota, Scott A. Tomlins, Debashis Ghosh, Daniel Hayes, Michael S. Sabel, Donna Livant, and Stephen J. Weiss; Richard G. Sewalt and Arie Otte of the Swammerdam Institute for Life Sciences at the University of Amsterdam; and Mark Rubin of Brigham and Women's Hospital, Harvard University.
The research was funded by grants from the Department of Defense and the National Institutes of Health, a Munn grant from the University of Michigan, and by the Breast Cancer Research Foundation, the Mary Kay Ash Foundation and the V Foundation.
Patients interested in breast cancer and prostate cancer research at the University of Michigan Comprehensive Cancer Center may call the U-M Cancer AnswerLine, at 1-800-865-1125.
Note to editors: Images showing EZH2 concentrations in stained human cancer tissue, and EZH2-enhanced cancer cells invading chicken embryo tissue, are available upon request.
Reference: "EZH2 is a Marker of Aggressive Breast Cancer and Promotes Neoplastic Transformation of Breast Epithelial Cells", Proceedings of the National Academy of Sciences, online publication week of Sept. 8, 2003 at www.pnas.org.