|
||||||||
|
PATHOLOGY No Innocent BystandersEven Normal Cells Act as Cancer Cell Accomplices On their infamous route to immortality, cancer cells devise ways to break the rules of cellular fair engagement. They need oxygen from their host, so they pilfer it by growing new blood vessels. If they do not get this vital nourishment, molecular sensors within them activate their suicide machinery. To prevent this, neoplasms subvert their own cell-death program.
Extracellular matrix proteins, such as thrombospondin-1, have a profound effect on tumor growth and survival, according to Jack Lawler and collaborators. Photo by Pam Murray According to Jack Lawler, HMS associate professor of pathology at Beth Israel Deaconess, there are no innocent bystanders in these processes. "Previous research focused on the genetics of the cancer cell. Now we have to think about the whole tissue and the various other cell types present." Malignancies evolve not only because a single cell takes the renegade path to cancer but also because that cell solicits help from normal cells. Lawler and his coworkers, M. Luisa Iruela-Arispe from UCLA, Roderick Bronson of Tufts University and HMS, and Richard Hynes from MIT, have addressed the interaction between cancer cells and their accomplices by studying the effects of thrombospondin-1 (TSP-1) on tumor growth and angiogenesis. Thrombospondin-1 interacts with a host of other molecules: CD36, CD47, integrins, gelatinases, TGF-beta--names often dropped when the subject is cancer. Under normal conditions TSP-1 lies low. But if the tissue is stressed, it is upregulated. In other words, when things go awry, thrombospondin-1 steps in to keep order. In a study published in the November American Journal of Pathology, Lawler and colleagues provide the proof of principle that when thrombospondin-1 is absent from the extracellular matrix, the advantage goes to the tumor. The Lawler group transplanted cancer cells into mutant mice that did not express thrombospondin-1. Tumors grew roughly twice as fast as in wild-type mice, demonstrating that when host cells express TSP-1, the tumor microenvironment is less permissive. "But most tumors don't start with 250,000 fully transformed cancer cells," Lawler said. A Generic EffectIn order to mimic the natural and spontaneous process of tumorigenesis and at the same time manipulate thrombospondin-1 levels, Lawler teamed up with Iruela-Arispe, formerly at Beth Israel Deaconess, and Hynes. They used knockout mice lacking TSP-1 gene expression and crossed them with two cancer-prone mouse strains. The first study used mice with a mutated p53 gene, which succumbs to a range of cancers. The offspring expressing thrombospondin-1 survived longer. "Looking at the TSP-1 null effect broken down into lymphomas, carcinomas, and sarcomas, it is clear that the difference in survival is across the board," Lawler said. "This is a protein that has a very generic effect."In a second study, appearing in the Oct. 23 Proceeding of the National Academy of Sciences, senior author Iruela-Arispe made a mouse that overexpressed thrombospondin-1, then crossed it with mice prone to mammary cancer. As in the mutant-p53 mice, tumors grew larger and faster in the absence of TSP-1. "In addition, if you overexpress TSP-1 the mice survive even longer," Lawler said. Differences in the tumor vascular architecture also correlated with TSP-1 expression. In TSP-1 null mice, for example, capillary perimeter increased twofold compared to the "normal" mammary mouse. The Multitasker's MethodsFurther experiments shed light on what thrombospondin-1 is doing. Tumors acquire the ability to form blood vessels through a conversion called the "angiogenic switch." The switch may be more like a scale that tips in favor of stimulatory effectors. Normally, thrombospondin-1 acts as the prototypical angiogenesis inhibitor. Its effect is counterbalanced by growth factors such as the vascular endothelial growth factor (VEGF), which can stimulate blood vessel formation. During tumorigenesis, this careful equilibrium between positive and negative effects is likely upset through mechanisms that are actively being investigated."Until this study," said Lawler, "it was generally accepted that thrombospondin-1 acted through CD36, a membrane receptor." The researchers found that thrombospondin-1 can also act in another way: by preventing the conversion of gelatinase B to an active form. But how can increased levels of this protease explain the more invasive phenotype, rapid growth, and greater vascularity displayed by the thrombospondin-1 null tumors? It turns out that without the policing effect of thrombospondin-1, gelatinase B modifies the extracellular matrix. "What the active gelatinase does is free up VEGF, the stimulatory effector," said Lawler. In support of this, visual evidence abounds. "The tortuous, dilated, and abundant vascular networks characteristic of the TSP-1 minus tumors are reminiscent of phenotypes attributed to elevated VEGF." Clearly, a therapeutic approach to augment TSP-1 expression--as in the case of the overexpressor transgenic strain--might help. In a third study, in the November issue of Cancer Research, the Lawler lab bioengineered fragments of TSP-1 and tested their effects on carcinomas. Two molecules differing by a single change in the amino acid sequence RFK (arginine-phenylalanine-lysine) were of particular interest. "Both inhibited angiogenesis, as had been reported before," said Lawler. But histological specimens showed that apoptosis doubled in the presence of the RFK sequence. Lawler suspects that the tumor cell growth regulator TGF-beta, which binds to the RFK sequence, has something to do with this. Experiments with Lewis lung carcinomas support this idea. When experimental tumors from this cell line, which are insensitive to TGF-beta, are treated with the RFK protein, there is no effect on cell death. --Anne Mahon |
