Editor's note: This is part two of a two-part series on cancer metastasis. Part one ran in Tuesday's issue.

The characteristics of metastases are strongly determined by the primary tumors that spawned them, and so the best hope for treating metastases is to develop better drugs against the primary tumor. But there appear to be metastasis mechanisms that contribute to the metastasis of multiple tumor types.

Researchers are beginning to work out how cells pull up their stakes and move, and their findings may offer clues to therapeutic approaches that could prevent metastases more effectively than currently is the case.

"More than 90 percent of cancers are epithelial in origin," Sendurai Mani told BioWorld Today. "And those tumors, when they want to migrate, have to undergo the epithelial to mesenchymal transition," said the Whitehead Institute for Biomedical Research postdoctoral researcher.

That epithelial to mesenchymal transition, or EMT, involves a change of cell characteristics in which tumor cells free themselves from the epithelial sheet that usually keeps them where they are.

The transition happens in two stages - loss of epithelial cell characteristics, followed by a gain of mesenchymal ones - with the net result that a tumor cell is able to migrate, resistant to apoptosis and capable of seeding a secondary tumor. A recent paper in the Proceedings of the National Academy of Sciences identified the transcription factor FoxC2 as a key protein responsible for the acquisition of mesenchymal-like properties by rogue tumor cells.

The authors, who are from the Whitehead Institute in Cambridge, Mass., Harvard Medical School and the Massachusetts Institute of Technology, as well as Hamamatsu University in Japan, compared and manipulated FoxC2 expression in four cell lines with varying degrees of metastatic potential. FoxC2 expression was highest in the most aggressive cell line. The authors later used microarrays containing human breast tumor samples, and found that FoxC2 specifically is overexpressed in highly aggressive breast cancers.

In cell lines, knocking down the expression of FoxC2 with short hairpin RNAs led to a decrease in metastases to the lung. Notably, those metastases that did form had FoxC2 protein, suggesting that they were the ones where the knockdown hadn't worked in the first place. On the other hand, upregulating FoxC2 expression in the more benign cell lines nearly tripled the number of metastases they formed.

Though the research is in its early stages, Mani said that FoxC2 is an attractive therapeutic target because it is turned on almost exclusively in "aggressive, highly metastatic tumors," making it a fairly specific target. The Whitehead Institute has filed for patents related to the discoveries, and Mani and his colleagues have industrial collaborations to advance the work.

Mani also noted that staining for FoxC2 has furthered the idea that the epithelial-to-mesenchymal transition mediates metastasis in the first place. Because mesenchymal-like tumor cells are hard to distinguish from regular fibroblasts, there has been some skepticism about whether EMT truly is a metastatic mechanism. But staining for FoxC2 can show the difference between a metastasizing tumor and a regular fibroblast. "Epithelial cells don't express [FoxC2], and regular mesenchymal cells don't express it," Mani said. "Only mesenchymal cells generated by EMT."

Another recent paper investigated the biomechanics of metastasis. A report in the June 11 issue of Cancer Cell suggested that tumor cells literally may bootstrap their way to the lymphatic system by expressing both chemokine receptors and secreting their ligands.

At first blush, secreting a ligand for one's own receptors might sound like a recipe for a cell that chases its own tail. But the tumor cells take advantage of a mechanism that transports those signaling molecules, as well as the cells themselves, in the direction they want to go: lymphatic system drainage.

Tumors produce excess fluid that continually percolates from the tumor toward nearby lymphatic vessels. The scientists, who are from the Swiss Federal Institute of Technology in Lausanne, Switzerland, and the Mount Sinai School of Medicine in New York, used three-dimensional cell culture systems to show that tumor cells use two cues to find the lymph system: that mechanical flow, as well as signaling molecules that are secreted both by endothelial cells lining the lymph system and by the tumor cells themselves.

Senior author Melody Swartz, associate professor at the Swiss Federal Institute of Technology, told BioWorld Today that the relative contributions of flow and signaling molecules depend on how far the lymphatic is from the tumor. If the tumor is far, then flow is important, but as the tumor cells approach the lymphatic vessel, the two signals amplify and help promote invasion. However, "we don't have quantitative data on exactly how far, just that they synergize." It's also not yet clear what the "broadcast distance" of lymphatic-secreted CCL21 is in vivo.

Swartz said that so far, the research is "trying to answer fundamental questions of how cells home to lymphatic vessels in the tissue environment." But "the new concept that cells can be the source of their own directional cues, when in a mechanically dynamic environment, may in the future help rational strategies" to prevent tumor cells from hitching a ride via the lymph system.