Tomlins developed an analysis method that made it possible to find the needle in the haystack, the abnormalities that had previously gotten lost in a deluge of data.
A diagnosis of prostate cancer may be a man’s worst nightmare. It can start with burning during urination, difficulty gaining an erection, pain during ejaculation or pain in the lower back, hips or thighs.
Unfortunately the cure is often worse than the illness. Surgery may leave a man incontinent, impotent or infertile; radiation therapy isn’t much better; and hormone therapy can open a Pandora’s box of problems, including decreased libido.
Prostate cancer is the most commonly diagnosed form of cancer in American men, striking one in six, and more men die from its ravages than from any other cancer except lung cancer. Researchers are at a loss as to how to prevent it and until last fall knew little about what caused it. A discovery by a recent Willamette graduate is providing new insights.
Scott Tomlins ’01, a student in the University of Michigan Medical Scientist Training Program, is the lead author of an article published in the October issue of Science, one of the world’s most prestigious science publications. He worked with a team of 16 scientists from the University of Michigan Medical School, Harvardaffiliated Brigham and Women’s Hospital, and the University of Ulm in Germany.
“Most of the big scientific projects are now collaborative efforts,” he says. “The number of authors on papers has gone up recently because research is getting so complex.” Individuals or small teams rarely have the expertise and specialized techniques required for sophisticated research, so collaborations now cross institutional and geographic boundaries. Tomlin’s colleagues from Germany, for example, had access to tissue samples not available in the United States.
The Michigan team, under the direction of Arul Chinnaiyan, MD, PhD, discovered that fused genes may trigger the development of prostate cancer. The prostate is a walnut-sized gland between a man’s penis and bladder that produces seminal fluid. Prostate cancer occurs when the cells of the prostate begin to grow and divide uncontrollably. Most prostate tumors grow slowly and stay confined to the prostate, where in many cases they do little harm. But sometimes cancerous cells begin to replicate rapidly and invade other areas of the body.
Tomlins’ work was based on the observation that in some cancers, such as leukemia, genes normally located on separate chromosomes fuse when the chromosomal order becomes scrambled. These fused genes have been shown to have a causal role in cancer development. Similar fusions in solid tumors, such as breast, colon and prostate cancer, had not been identified, in part, because previous researchers lacked an efficient method to sift through the massive amounts of data generated by current studies. Where scientists used to analyze individual genes one at a time, they now use microarrays that can analyze 20,000 genes simultaneously.
Although Tomlins is a graduate student among seasoned professionals, he came up with a novel idea. “People had focused on discovering which genes are always switched on when cancer occurs,” he says. “We were more interested in finding the genes that were switched on at very high levels in a fraction of cancer cases.” Tomlins and fellow graduate student Daniel Rhodes developed an analysis method—the Cancer Outlier Profile Analysis (COPA)—that made it possible to find the needle in the haystack, the abnormalities that had previously gotten lost in a deluge of data. COPA is a statistical algorithm, which they applied to a database of numerous DNA microarray studies.
Neighbors: In this prostate cancer cell, the ETV1 gene (red) and the TMPRSS2 gene (green) are fused (yellow) on one chromosome.
Tomlins pored over the data in the lab and began picking up consistent patterns of abnormal gene activity. Two genes that control cell division, ERG and ETV1, are tightly regulated in normal cells. However, the COPA algorithm identified inappropriate activation of ERG and ETV1 in a fraction of prostate cancers. Using additional techniques, Tomlins determined that in cases with inappropriate activation, ERG or ETV1 were fused to a gene named TMPRSS2, which is normally turned on at high levels in the prostate. After the fusion, ERG and ETV1 adopt the behavior of TMPRSS2 and switch on in prostate cells, leading to uncontrolled cell division. “We found that the fusion is what’s driving the inappropriate expression of those genes,” says Tomlins.
He took only two short breaks during an intense summer. “It all came together pretty quickly, which is unusual for a large study,” he says. Tomlins and Rhodes’ method, based on bioinformatics, allowed the Michigan team to combine large data sets, make a strong hypothesis and arrive at a powerful conclusion—that gene fusion is likely the cause of prostate cancer.
A leading prostate cancer expert at Johns Hopkins University School of Medicine, William Issacs, says the discovery “will invigorate the field in terms of looking for these kinds of fusions in other common cancers.” His assessment is seconded by Jacob Kagan of the National Cancer Institute, who believes the finding may lead to a deeper understanding of the mechanisms involved in breast, colon and lung cancer.
Tomlins says the discovery provides a target in fighting cancer. “Most cancer treatments non-selectively kill cells that are dividing, which makes it unlikely they’ll kill all the cancer cells,” he says. “That’s why cancer often reoccurs after treatment.” When researchers discovered that chronic myelogenous leukemia is caused by a gene fusion, they were able to develop the first rationally designed chemotherapeutic, which is now the standard of care. “There’s an intense research effort now to find rational, targeted treatments.”
In the near future, a test to screen for gene fusions could act as a second line of defense for men who turn up positive on the prostate-specific antigen (PSA) test, which is notoriously unreliable and therefore controversial. When caught and treated early, prostate cancer has a cure rate of more than 90 percent. Tomlins’ research may be key in early detection and more effective treatment.
“He always kept it really interesting. Whether he agreed with you or not, he challenged you to think and defend your reasoning for your arguments,” Tomlins says. “In philosophy, you analyze arguments, and in science, you analyze data. You need to be able to answer questions in both.”
Talbott retired this spring after teaching at Willamette for 34 years. He always approached his classes as a place for good dialogue, “an arena in which we struggle with ideas,” he says. “I like it to be focused. What we want in the end is a clear argument, with a lot of vigorous discussion as we proceed.”
Talbott doesn’t see retirement as being much different than teaching, though he’ll be able to devote more time to his passion for writing. He has already published numerous articles and a book, The Inescapable Love of God. He is working on a second book, this one about the controversy regarding free will and determinism.
“All of my life, I’ve been connected to universities,” Talbott says. “All of my life, I’ve lived for summers and sabbaticals. So I don’t have any problem knowing what I’ll do in retirement. I’ll do the same thing I’ve been doing now, except I won’t be grading papers.”