One reason cancer is so difficult to understand and to treat is that tumors are a genetic muddle. A cell can become cancerous via a number of pathways, so the cancer-causing mutations found in one tumor cell may be quite different from those found in neighboring cells. One consequence of this genetic heterogeneity is that a treatment that successfully kills some cells may be ineffective against others. As these surviving cells proliferate, the tumor may become resistant to treatment.

Cancer researchers are beginning to decipher how genetic defects interact within individual cells and lead them awry. However, very little is known about the far more daunting question of whether, in the complex biological milieu that scientists call the “tumor microenvironment,” mutations in one cell can interact with those in other cells to promote cancer.

In an article in Science published in 2003, Tian Xu, Ph.D., vice chair and professor of genetics, and colleagues noted that tumors in Drosophila “display many characteristics observed in human cancers.” In order to emulate the genetic patchwork of human tumors, Xu and School of Medicine colleagues have created a “mosaic” form of the fruit fly Drosophila melanogaster in which a tiny number of cells with mutant genes can exist in a fly with mostly normal cells. In an elegant series of experiments reported in the journal Nature in January, Xu and colleagues used this model to demonstrate that interactions between separate cells carrying different mutations are indeed possible, and that they can have a profound effect on how cancers grow and metastasize.

In the new studies, Xu, along with graduate student Ming Wu, Ph.D., and postdoctoral fellow José Carlos Pastor-Pereja, Ph.D., the lead authors, first created mosaic flies in which a well-known cancer-causing gene known as Ras is expressed in some cells along with a mutant, non-functional form of scribbled (abbreviated scrib), a tumor-suppressor gene. To track the effects of these manipulations, these cells were tagged with green-fluorescent protein (GFP).

The result was explosive: GFP-tagged tumors soon engulfed the normal cells in the flies’ brains, and spread to adjacent tissues as well. In these cells, the presence of mutant scrib signals that the cells are damaged, which activates a stress-related signaling pathway called JNK (pronounced “junk”). If the cells had been carrying mutant scrib alone, the activation of the JNK pathway would have caused the cells to die, while simultaneously triggering another pathway in adjacent cells known as JAK-STAT, which promotes proliferation. This process makes up for the loss of the lost scrib mutant cells in a process is known as compensatory proliferation. However, the presence of Ras and the scrib mutation together in the same cell overrides the JNK cell-death signal, while preserving the JAK-STAT proliferation signal, putting the Ras carrying cells on a path to tumor development and progression.

The most intriguing observation, however, was that when Ras was expressed alone in some cells, and the scrib mutation alone in other, adjacent cells (with most cells being normal, as before), rampant tumor growth and metastasis was seen again, just as when the two mutants occurred together in the same cells (see photo).

This phenomenon had never been observed, and indicated that some intercellular interaction between Ras and scrib mutant cells could fuel the growth and spread of cancer. The researchers discovered that the JNK pathway is activated as usual in the scrib mutant cells, causing them to die, but JNK signaling in these cells is also somehow propagated to adjacent cells carrying Ras. In those cells, JNK activates the JAK-STAT proliferation pathway, causing tumors and metastasis.

As noted above, JNK signaling is induced by stress, such as when tissue is wounded. When the researchers damaged wing discs in Drosophila larvae with cells carrying Ras, tumors resulted, indicating that stress-induced JNK signaling alone is sufficient to turn cells carrying Ras toward cancer.

“A lot of different conditions can trigger JNK stress signaling,” says Xu. “Physical stress, emotional stress, infections, inflammation – all these things.” The Xu team’s view of stress-induced tumor growth is consistent with recent findings of other researchers, who have shown that tumors and wounds have similar traits, and that there is a close relationship between chronic inflammation and cancer.

Xu believes that his group’s findings on cell–cell interaction in tumors are a powerful demonstration of how the precise genetic control offered by Drosophila can shed new light on the biology of cancer. “The bad news is that it is much easier for a tissue to accumulate mutations in different cells than in the same cell,” says Xu, a Howard Hughes Medical Institute investigator and director of the Institute of Developmental Biology and Molecular Medicine at Fudan University in Shanghai, China. “Better understanding of the underlying mechanism causing cancer always offers new tools to battle the disease.”