Their work is being published in the March 29 issue of the journal Current Biology.
"Before, we thought this gene was a classical tumor suppressor," says Louis Dubeau, professor of pathology at the Keck School and principal investigator on the paper. If that were the case, it would mean that mutation of the gene would allow the cell it's in to grow out of control and create a tumor. Instead, Dubeau notes, "What we've shown is that the gene actually acts indirectly, that it disrupts interactions between different cell types."
The well-known breast cancer gene, BRCA1, not only gives carriers of its mutated form a four in five chance of developing breast cancer, it also confers a 40 percent risk of developing ovarian cancer by the age of 70. How that risk is imparted, however, had been harder to pin down.
"We've known for a long time that ovarian cancer is associated with ovulation, in that women who have regular menstrual cycles through their life without interruption by pregnancy or oral contraceptive use are at highest risk for developing sporadic ovarian cancer," Dubeau explains. "So we had some clues that the cells that control the menstrual cycle-the ovarian granulosa cells-have an influence on ovarian cancer."
But how? Was that influence direct, or indirect? Dubeau eventually got a handle on the problem by looking at ovarian cancer rates in genetically modified mice created in collaboration with Robert Maxson, Keck School professor of biochemistry and molecular biology and director of the mouse core facility at the USC/Norris Cancer Center. "The whole project was based on creating a mouse that lacks BRCA1 in only its granulosa cells," Dubeau says. "This collaboration was essential to the project's success."
What Dubeau and his colleagues found was that while mutating the BRCA1 gene in granulosa cells did indeed give rise to ovarian tumors, those tumors did not arise in granulosa cells. Instead, when the tumor cells were analyzed, they were found to be epithelial cells very similar to those found in human ovarian cancers, with perfectly intact, functioning copies of the BRCA1 gene.
"What this says is that the cells that control the menstrual cycle, the ovarian granulosa cells, also control ovarian tumor development, but from a distance," Dubeau explains. The most likely scenario, he says, is that the granulosa cells normally give off a chemical signal that stops the epithelial cells from growing out of control. When that chemical signal disappears or is muted by a mutation in the BRCA1 gene, the epithelial cells don't get the message, and keep on growing and dividing. The result: ovarian cancer.
This finding is actually good news for scientists and physicians trying to figure out new ways to treat ovarian cancer. If the cancer had arisen in the same cells that had the BRCA1 mutation, the only way to interfere would be to correct the mutation. In this case, however, there's a mediator-a biochemical of some sort-that scientists might be able to replace in people with identified BRCA1 mutations, making their risk of ovarian cancer drop precipitously.
In addition, once the chemical messenger that's affected has been identified, it will be much easier to diagnose a predisposition to ovarian cancer or pinpoint just who is at risk, simply by measuring the chemical's levels.
"The consequence of this finding," Dubeau says, "is that ovarian cancer is the result of some biochemical problem that may be correctable or preventable. That's what makes this finding so exciting."
Dubeau points out that women with BRCA1 mutations are also predisposed to cancers of the fallopian tubes, and that the mice with mutated BRCA1 genes in their granulosa cells developed tumors there as well. "This not only underscores the relevance of our mouse model to human cancer," Dubeau notes, "But also strongly supports a theory we have formulated about the site of origin of ovarian cancers."
The work was supported by grants from the U.S. Department of Defense and the Whittier Foundation.
Rajas Chodankar, Stanford Kwang, Frank Sangiori, Hao Hong, Hai-Yun Yen, Chuxia Deng, Malcolm C. Pike, Charles F. Shuler, Robert Maxson, Louis Dubeau, "Cell-Nonautonomous Induction of Ovarian and Uterine Serous Cystadenomas in Mice Lacking a Functional Brca1 in Ovarian Granulosa Cells." Current Biology. Volume 15, Number 6, March 29, 2005.