Researchers Discover Why Melanoma Is So Malignant

"Other cancers need to learn how to spread, but not melanoma," saysWhitehead Member Robert Weinberg, senior author of the paper that willbe published September 4 in the early online edition of the journalNature Genetics. "Now, for the first time, we understand the geneticmechanism responsible for this."

Metastasis (the spread of disease to an unconnected body part)is a highly inefficient, multi-step process that requires cancer cellsto jump through many hoops. The cells first must invade a nearbytissue, then make their way into the blood or lymphatic vessels. Nextthey must migrate through the bloodstream to a distant site, exit thebloodstream, and establish new colonies. Researchers have wondered whymelanoma in particular is able to do this not only more efficientlythan other cancers, but at a far earlier stage. This new study showsthat as melanocytes--cells that protect the skin from sun damage byproducing pigmentation--morph into cancer cells, they immediatelyreawaken a dormant cellular process that lets them travel swiftlythroughout the body.

Central to this reawakened process is a gene called Slug(named after the bizarre embryo shape that its mutated form can causein fruit flies). Slug is active in the neural crest, an early embryoniccluster of cells that eventually gives rise to a variety of cell typesin the adult, including dermal melanocytes. In this early embryonicstage, Slug enables the neural crest cells to travel, and then settle,throughout the developing embryo.

"Slug is a key component of the neural crest's ability tomigrate," says Piyush Gupta, a MIT graduate student in Weinberg's laband first author on the paper. "Following its activation duringembryonic development, Slug is shut off in adult tissues." But whenskin cells in, say, an individual's mole, become malignant, theyreadily reactivate Slug and gain the ability to spread--something thatother cancers can spend decades trying to do.

Weinberg's team demonstrated this through a number ofexperiments. In the first, they created models of various cancer typesby introducing cancer-causing genes into normal human cells and theninjecting the tumor cells underneath the skin of mice. Mice injectedwith breast cancer cells or with fibroblast (connective tissue) cancercells developed tumors, but the tumors didn't spread. Those injectedwith melanoma cells immediately developed invasive tumors throughouttheir body, spreading everywhere from the lungs to the spleen. Thisstrongly supported the suspicion that melanoma is so metastatic in partdue to properties intrinsic to melanocytes themselves, and not simplybecause it is external and thus uniquely exposed to environmentalstresses.

For the second experiment, the team used microarray technology(chips covered with fragments of DNA that can measure gene levels) andfound that Slug was expressed in human melanoma. "Really, this isn'tthat surprising," says Gupta, "when you consider that melanocytes inthe skin are direct descendants of the neural crest." In fact, Guptapoints out that occasionally physicians discover that perfectly benignmelanocytes will sometimes manage to migrate through a patient's bodyinto, say, the lymph nodes. This phenomenon isn't related to cancer,but rather demonstrates the latent ability of melanocytes to travel

Finally, the research team found that when Slug was knocked outin melanoma cells, the cancer was unable to metastasize when placedinto a mouse.

"This work is yet another demonstration of the notion thatcertain embryonic genes normally involved in transferring cells fromone part of the body to another are also involved in enabling cancercells to spread," says Weinberg, who is also a professor of biology atMIT.

An earlier demonstration of this notion was reported in a June 2004 paper in the journal Cell (http://www.wi.mit.edu/news/archives/2004/rw_0624.html).Here, Weinberg and colleagues described how a similar embryonic genecalled Twist becomes active in certain forms of breast cancer, therebyaiding metastasis. However, there are two important distinctions:First, Twist isn't reactivated until later in the tumor's life. Andsecond, while breast cancer cells appear to hijack Twist, melanomatumors kick-start a process that was once natural to their cellularancestors.

This research was supported by a grant from the U.S. National Institutes of Health and the National Cancer Institute.