July 28, 1999
Study opens door to treatments that short-circuit signals between tumor cells and their hosts
Researchers at Dana-Farber Cancer Institute have devised a new way of eavesdropping on the "molecular conversation" that exists between melanoma cancer cells and their host body - an exchange of signals that enables cancer cells to grow their own networks of blood vessels, elude attack from the immune system, and seed the body with new colonies of tumors.
In a study in the July 29 issue of Nature, researchers led by Lynda Chin, M.D., of Dana-Farber, created a breed of mouse that has a biological "switch" for turning on and off the development of melanoma skin cancer. When the animals were given a common antibiotic in their drinking water, the drug switched on a cancer-causing gene, leading to the development of melanoma tumors in two to three months. When the antibiotic was withdrawn, the switch was turned off and the tumors shriveled up. The advance provides investigators with a new tool for understanding the development of melanoma in humans and for exploring the complex interactions between cancer cells and their host.
The study also holds some encouraging news for pharmaceutical companies working on gene therapies for melanoma. Such therapies seek to replace defective genes - dozens of which may accumulate in cancer cells - with normal ones. The study found that genetic defects that occur at the very start of the cancer process remain vital to tumor maintenance even when cells become fully cancerous. As a result, therapies for repairing just a few key defects in melanoma cells can, in theory, cause those cells to return to a more normal state.
"It's known that full-fledged tumor cells contain many - possibly hundreds - of gene mutations," Chin observes. "But scientists have wondered whether the mutations that initially cause cells to turn cancerous are also necessary for them to stay cancerous. This study indicates that those early mutations are important even in fully malignant melanoma cells."
The study builds on earlier efforts to create animal models of human melanoma. Because mice and other laboratory animals do not normally develop melanoma tumors, it has been difficult for scientists to study the disease and develop effective new treatments for it.
A few years ago, Chin and her colleagues developed a strain of mouse that has an inborn tendency to develop melanoma in the skin. The animal's pigment-producing skin cells - called melanocytes - had an overactive gene, called RAS, and a de-activated gene, called INK4a that normally acts as a brake against tumors. Both genetic defects have been found in human melanomas.
In humans, some individuals have an inherited susceptibility to melanoma. In fact, 10 percent of all melanomas are found within certain families. When such individuals acquire additional genetic damage to their melanocytes - often as a result of intense, intermittent sun exposure - the melanocytes are transformed into cancerous cells. To make the mouse model more nearly resemble the human condition - that is, acquire a mutation during adult life - Chin and her colleagues inserted a genetic switch that made it possible to turn RAS on and off, enabling researchers to control the development of melanoma tumors in the animals. The switch was triggered by giving the animals doxycycline, an antibiotic similar to tetracycline, in their drinking water (see figure 1).
"When we gave the mice doxycycline, the presence of the drug turned on the engineered switch and activated RAS - a potent cancer-causing gene - leading to rapid melanoma development," Chin says. "When the drug was removed from their water, RAS is switched off and the tumors disappeared within two or three weeks. This demonstrates that RAS is important not only at the start of the cancer process, it also plays a role in the maintenance of full-fledged melanoma cells."
The research holds clear implications for gene therapy, she continues. "We've shown that it may not be necessary to repair or reverse all the genetic defects within solid tumor cells. Reversing key defects that take place at the beginning of the cancer process may be enough."
In the second phase of the study, researchers sought to determine precisely why the tumors shrank once the doxycycline was withdrawn. "There could be a variety of reasons for tumor reduction," Chin says. "We wanted to know whether it was due to a loss of the tumors' blood supply, whether they had been attacked by the immune system, or if their natural suicide program had been re-started."
As the tumors shrank, researchers surgically removed small samples and analyzed them. "We found that the number one reason for tumor shrinkage was the loss of blood supply," Chin says. "The blood vessels that had been feeding the tumors were dying. "This work provides us with a model for understanding the intricate relationship between cancer cells and their host, and how we can intervene in that relationship in therapeutic ways." Chin continues. "Targeting the cells that respond to signals sent by tumor cells may prove more effective than targeting tumor cells themselves. It is anticipated that host cells, unlike tumor cells, will not readily develop resistance to drugs, because they have greater genetic stability. Thus, blocking the signals between tumor cells and their host represents an important new strategy against cancer."
Other Dana-Farber investigators involved in the study were: senior author Ronald DePinho, M.D.; Alice Tam; Nabeel Bardeesy, Qiong Shen, M.D.; Ronan O'Hagan; and James Horner.
