NSAID drug protects against intestinal tumors in mice, despite poor diet and gene losses
In mouse models of intestinal cancer, use of an anti-inflammatory drug eliminated all of the cancer-causing risks produced by a high-fat Western-style diet - even when several genetic brakes to cancer formation were missing in the animals, say researchers from the Albert Einstein Cancer Center.
The investigators, who presented their findings at the 96th Annual Meeting of the American Association for Cancer Research, say that while the results do not yet have relevance for preventing human colon cancer, they do illustrate the interplay between genes and common nutritional and medicinal agents in development of cancer in the intestines.
The drug they tested, sulindac, was a highly effective chemoprevention agent, the researchers say, because it worked to induce expression of the p21 gene, which they found put a firm stop on tumor formation even though the mice were missing two key tumor suppressor genes (p27 and APC) and were fed a diet high in fat and low in calcium and vitamin D.
"It appears that p21 activation through sulindac offers protection against both a lack of tumor suppressor genes as well as poor diet," says the lead author, WanCai Yang, M.D., an assistant professor of medicine. While the drug is a NSAID (non-steroidal anti-inflammatory drug) and a COX-2 inhibitor, Yang believes its chemoprotective effects come via novel pathways that affect p21 expression.
The study builds upon a body of research conducted by Yang and Leonard Augenlicht, Ph.D., at the Albert Einstein Cancer Center that has used knockout mouse models to explore the role of genes and diet, including the finding that inactivation of the p21 gene accelerated tumor formation and that loss of this gene eliminated the ability of sulindac to inhibit tumor formation. Their earlier studies also showed that mice that lacked p27, another tumor suppressor gene, had a higher risk of developing tumors in both the small and large intestine.
Now, however, they have linked that risk to a high-risk diet. In the first part of the study, they found that mice lacking a p27 gene that were fed with a corn-oil rich diet (labeled AIN-76A, also low in calcium and vitamin D) formed tumors whereas knock-out mice fed with their regular chow (which was enriched with soybean oil) did not. And p27 knock-out mice given a Western diet, full of fat and lacking calcium and vitamin D, formed the most number of tumors.
The researchers then looked at the role that diet plays when expression of two tumor suppressor genes is reduced. The researchers mated mice that had only one APC gene to p27 positive or negative mice, producing variants with no p27 genes or one or two. The APC gene is responsible for an inherited colorectal cancer condition known as FAP (familial adenomatous polyposis), in which patients develop hundreds of potentially precancerous polyps.
They found that on the AIN-76A diet, loss of one APC gene but retention of the "wild-type" (normal) p27 genes was enough to induce tumor formation in 22 percent of the mice, but that number increased to 72 percent when one allele of the p27 gene was lost, and 93 percent when both alleles of the p27 gene were lost. But all of the knock-out mice variants developed tumors in both their small and large intestines when fed a Western-style diet.
"This showed that the diet overwhelmed any protective effect the p27 genes conferred," says Yang. "It not only increased tumor size and number, but also pushed the cancer into an invasive state."
Finally, they examined what happened when the knock-out mice were fed a Western diet enhanced with sulindac, an agent they had previously found help suppress tumor formation. To their surprise, none of the knock-out mice developed cancer.
"To us this means that as long as an animal can activate the p21 gene, this can have an overriding affect on suppressing tumor formation caused by either loss of other tumor suppressor genes or diet," says Yang. "We are testing this concept further, including exactly how the p21 gene gets activated."
Warren R. Froelich | EurekAlert!