From gullet to fundament, the gastrointestinal tract - stomach, small intestine, colon, rectum - is a lengthy trap for sowing tumors. Their prevalence varies geographically, but in the U.S. colorectal cancer - small intestine, colon, rectum - is a lengthy tumor trap.

In fact, colorectal cancer ranks third among all neoplasms in terms of occurrence and death rates among both male and female Americans. An estimated 147,000 patients will be diagnosed with colorectal cancer (CRC) this year in the U.S., and 57,100 will die.

"This mortality is not inevitable," emphasizes medical geneticist/oncologist William Dove, at the University of Wisconsin in Madison. "Early detection of colorectal cancer is critical for the management of this disease. The identification of useful biomarkers for detection and diagnosis of human colon cancer is a major goal of our research. Current methods for early detection, such as colonoscopy, suffer from inaccuracy and/or poor patient compliance," Dove observed.

He is senior author of an article in the Proceedings of the National Academy of Sciences (PNAS) dated Aug. 5, 2003. Its title: "Clusterin as a biomarker in murine and human intestinal neoplasia."

"Our overall finding," Dove said, "is that we've employed a new strategy in which we use a mouse that simulates human colon cancer. However, it is genetically and environmentally homogeneous in order to detect molecules that are overexpressed in colorectal cancer. So the second step of our strategy," Dove continued, "is to check these candidates by various molecular tests on human colon cancer. So not all molecules that can be tested in a purified experimental system will then show up if we look in true human CRC. This is a two-step strategy, and some years ago we developed this mouse model that very closely simulates the molecular lesions found in many human colon cancers that lack the gatekeeper. Absence of the gatekeeper gene," Dove went on, "is a metaphor. Many tumors are prevented from forming by the activity of what is otherwise called a tumor-suppressor gene - which controls transition from normal growth into neoplastic growth."

For Loss Of A Gatekeeper Tumor Suppressor

"The intestinal epithelium is continually turning over," Dove noted. "It's a highly proliferative tissue in every individual. So these gatekeeper genes prevent that normal growth from progressing into the neoplastic growth phase. When their function is lost - and there are several ways that can happen - a tumor is formed. We were able to create a mouse family that simulates the same molecular process with the loss of that gatekeeper leading to intestinal cancer.

"We created that mouse by mutagenesis in the germline," Dove recounted, "and then screening. Thus we discovered this mouse family that had incurred a mutation in its gatekeeper genes. We found it by its colon cancer phenotype.

"We compared gene expression patterns between mice with intestinal tumors and matched tumor-free mice. The analysis revealed elevated levels of clusterin in mouse tumors. We then measured clusterin levels in human colon tissue samples, and observed that levels rose early in tumor development. This suggests that clusterin may improve the early diagnosis of colon cancer.

"But this doesn't get to the endpoint: Is it going to be clinically useful in early detection? And as we are exploring different kinds of analytical chemistry to find out whether in a mouse that is expressing this gene, are there molecules related to clusterin? Can one find them in fecal material or serum? Do they provide a kind of analytic biomarker whereby one can detect the presence of colon cancer even in a nascent cancer individual who doesn't show signs of advanced CRC tumor?

"Clusterin has been studied in many labs for many years. What we're generally learning in the field is that tumors do overexpress normal molecules that are normally expressed in other bodily tissues. It has to be expressed in the tumor. Clusterin is a normal molecule found in many different tissues in the body, but it also becomes strongly expressed in intestinal tumors compared to normal intestine.

"We call it tumor-associated,' not tumor-specific.' You'll be hard pressed to find anything that when finally studied is truly tumor-specific. New mutations that cause tumors? Yes, those will be tumor-specific. You can certainly find the well-known clusterin gene in the human genome sequence and the mouse genome sequence. It's encoded protein is a glycoprotein, so the final gene product is modified above and beyond its amino-acid sequence. Whether the gene is subject to mutation in connection with colon cancer is a really important question - being studied, but not yet known."

Enzyme Puts Brakes On Tumor Growth

"Our ongoing work," Dove recounted, "is to study the genes and molecules that are associated with colon cancer, as covered in the experimental format, and then validate this in human colon cancer. We have this general way of discovering genes that modify the growth rate of colonic tumors in mice, and the famous gene called mom one, which encodes a secretory phospholipase enzyme that makes tumors grow more slowly.

"We're able to discover these genes through the methods of mouse genetics and genomics, then investigate whether this molecule has any role in the growth of human colon cancer. When we've found the genes, with the present power of mouse genomics, we can within a year or two learn what molecule was affected when that gene changed its form from sensitive to resistant. If you have a molecule of interest, as we are doing with clusterin, we can make mice that are deficient in that molecule, and ask whether the disease is affected when the molecule is affected.

"If we alter that molecule we change the colon cancer phenotype," he said. "That's our overall strategy. It's really empowered by the power of the mouse and human genome sequences becoming available. All of this strategy is made much more feasible by the establishment of these two genome sequences."

A final question: Might a simple CRC blood test replace colonoscopy in whole or part? "We'd like to get to that point," Dove replied. "It's a very important goal for the field. We're not promising that we will get there," he concluded, "but we have blood tests very much in mind."