Search

Compelling evidence has clearly linked the metabolism of arachidonic acid (AA) to intestinal tumorigenesis. Nonsteroidal anti-inflammatory drugs (NSAIDs), inhibitors of cyclooxygenase (COX)-1 COX-2 or both, are effective chemopreventive agents in the prevention of intestinal tumors and in the regression of preexisting tumors in humans and animal models. Because NSAIDs inhibit the committed step in prostaglandin biosynthesis, some researchers have suggested that unmetabolized free AA levels induce ceramide-mediated apoptosis whereas others attribute the antitumorigenic effects of NSAIDs to reductions in prostaglandins, in particularly prostaglandin (PC) [E.sub.2]. This study set out to establish, with a greater degree of certainty, that the AA cascade was important in intestinal tumorigenesis by using [Apc.sub.Min/+] mice. Elevating intracellular concentrations of AA in tumors of [Apc.sub.Min/+] mice above basal levels had no effect on tumor load in the presence or absence of NSAIDs. However, reducing tissue AA content (with n-3 fatty acids) was correlated with a reduction in tumor number, and these effects were reversed in add-back experiments using dietary AA. Furthermore, inhibition of de novo biosynthesis of AA by using SC26196, a selective inhibitor of [DELTA]-6 desaturase, repeatedly reduced tumor number by 35-40%, and by-passing this inhibition with concomitant treatment of dietary AA rescued the tumors. To establish whether these effects could be related to prostaglandins, we set out to regress preexisting tumors with the NSAIDs piroxicam and sulindac and rescue these tumors (in vivo) with stable analogues of PG[E.sub.2]. We established that piroxicam and sulindac could eliminate 80-95% of preexisting tumors within 6 d. Immunohistochemical analysis of the tumors demonstrated that COX-2 expression was localized in the stoma of the tumors and protein levels progressively declined with length of NSAID exposure. These changes were associated with decreases in [beta]-catenin, Bcl-2 and mitosis and increases in apoptosis in tumor epithelial cells. When EP receptor agonists (16,16-dimethyl-PG[E.sub.2] and 17-phenyl-trinor-PG[E.sub.2]) were provided to the NSAID-treated animals, tumor regression was significantly attenuated (rescuing the tumors) and the apoptosis-mitosis index normalized to control levels. In a follow-up study, we systemically sequestered PG[E.sub.2] after intraperitoneal administration of a monoclonal antibody against PG[E.sub.2]. This treatment reduced the number of preexisting tumors by 35% (P < 0.05) within 4 d. Collectively, these data demonstrate for the first time that AA is important in intestinal tumorigenesis in vivo but there appears to be a threshold for this effect, NSAIDs regress preexisting tumors in part by inhibiting PG biosynthesis and PG[E.sub.2] is important in maintaining intestinal tumor integrity in vivo. [Supported in part by grants from the American Institute for Cancer Research, Pharmacia Corporation and the Tennessee Agricultural Experiment Station.]

Although high-calcium diets have been reported to reduce the risk of colorectal cancer, our preliminary data with the adenomatous polyposis coli (Apc) Min mutation (Min/+;ApcMin/+) mouse shows a paradoxical increase in intestinal tumor loads (> 65%) with high calcium diets. Since we previously demonstrated that increasing dietary calcium reduces adiposity, and ApcMin/+ mice on high calcium diets exhibited profound loss of adipose tissue, we hypothesized that loss of an adipose tissue-derived tumor suppressor factor(s) resulted in increased tumor susceptibility in animals on the high calcium diet. Accordingly, tumor prone ApcMin/+ mice were crossed with obesity prone lethal yellow agouti (Ay/a) mice to generate obese Ay/ApcMin/+ mice. Low (0.2%), normal (0.5%), and high (1.2%) calcium diets were fed to both Ay/ApcMin/+ mice and ApcMin/+ mice from 35—40 days until 90 days of age (n=21/strain, n=7/diet group). The high calcium diet reduced weight gain in both strains (P < 0.01) and reduced fat pad mass by 46—57% in Ay/ApcMin/+(P < 0.004) and by 65—82% in ApcMin/+(P < 0.03).ApcMin/+ mice on the high calcium diet exhibited an increase in tumor number (76 vs. 29, P=0.009), but this effect was not seen in the Ay/ApcMin/+ mice.β-Catenin and cyclin D1 gene expression were significantly induced with high calcium diet in intestinal tumor tissue of ApcMin/+ mice but not in Ay/ApcMin/+ mice. We conclude that the differential effect of dietary calcium on intestinal tumorigenesis in lean vs. obese ApcMin/+ may result from the loss of adipose-derived protective factor(s) due to the substantial loss of body fat in ApcMin/+ mice fed a high calcium dairy diet, increasingβ-catenin and cyclin D1 in tumors. [ABSTRACT FROM AUTHOR]