A Better Mouse
One would expect that the Institute’s world-renowned engineers could build a better mousetrap, but a better mouse? In fact, scientists at MIT’s cancer center are genetically engineering laboratory mice ideally suited for cancer research. In the fight against cancer, animal models are vital because they allow researchers to pinpoint the genetic sources of the disease, to study cancer cells as they grow within living tissue, and to analyze the effectiveness of treatments–something they cannot do in human beings. In the past, researchers might have cultured cancer cells in a petri dish, implanted them under the skin of a mouse, and then observed their growth. But genetic engineering has made that kind of lab work largely obsolete. Scientists now know that mutations in certain genes can lead to human cancer. But the nature of the mutations–are the genes damaged? turned off? hyperactivated?–is not always well understood. Researchers can begin to answer some of those questions by tweaking one or several genes in mice and then observing whether the animals become more susceptible to cancer. The models are informative because humans and mice are biologically similar. “In the end, we’re big mice, or mice are small people,” says Robert Weinberg ‘64, PhD ‘69, director of the Whitehead Institute for Biomedical Research and a member of the cancer center since its inception.
Weinberg has many “firsts” on his long list of professional accomplishments. He discovered the first human oncogene, a mutated form of a normal gene that can contribute to cancer. He was the first to turn a normal human cell into a cancer cell. And he was the first to clone a tumor suppressor gene, which normally restricts cell growth but when mutated allows cells to proliferate. His lab was also the first to successfully graft human breast tissue into the mammary glands of mice. The tissue produced human milk and in some cases developed tumors similar to those found in humans. For the first time, scientists could study the early stages of breast cancer in human tissue without having to examine a human being.
“We’re becoming increasingly refined in our ability to model human disease [in mice],” says Jacks. “We can produce in a specific way the kind of mutations that occur in human cancer.” Last December, for example, one of his teams announced that it had developed two mouse models that, with only slight genetic differences between them, resulted in two distinct forms of female reproductive disease. In the first model, the researchers mutated a known cancer-causing gene called K-ras, which caused the animals to develop endometriosis–a gynecologic disease that afflicts about five million women in the United States. In the second model, the researchers combined the mutated K-ras gene with an altered form of a gene known as Pten, which normally works to suppress cancer. The result was a mouse that developed a specific type of ovarian cancer. These mice represented the first animal models of the two diseases. “The end result is two models that are very powerful in how they might be used for two important diseases of the female reproductive organs,” says Jacks.
How they might be used depends on who’s interested. Jacks’s team, for example, uses its mouse models to study the effects of genetic mutation on tumor development, extracting cells from the animals’ tissue and examining how genes within the cells function. But people outside of MIT can request the animal models for use in their own research, including tests of potential anticancer drugs. That’s because almost all of the mouse strains developed in Jacks’s lab–and in anyone else’s lab at the Center for Cancer Research, for that matter–are placed in repositories such as the Mouse Models of Human Cancers Consortium Repository in Frederick, MD. Funded by the NCI, the repository makes models available for free to researchers at other institutions. It also manages a database that contains information relevant to each mouse strain. “The NCI has been aggressive in building [repositories] to make it easier for groups to share data,” says Jacks. This is important, he says, because it can take a lab a year or two to develop a specific genetic cancer model, including six months to breed mice that embody it. Singer says that MIT stands out among members of the consortium, not only for developing useful models but for “setting a new paradigm in how we need to be doing collaborative science.”