Search

We review the profound effects that components of diets commonly consumed in western societies and linked through population studies to risk for colon cancer have on the development of intestinal cancer in humans and in mouse models. Focus is particularly on levels of vitamin D, interactive with calcium and fat, in establishing probability of tumor development even in mouse genetic models in which there is high penetrance of the disease. These dietary factors have also been used to develop a mouse model of dietary-induced sporadic colon cancer which exhibits similar lag, incidence, and frequency of tumor development, and relative incidence of carcinomas and adenomas, as seen for >90% of colon tumors that arise in the general population later in life. Potential mechanisms influenced by diet that alter probability of tumor development are outlined, including altered patterns of intermediary metabolism, differentiation, and inflammation in the intestinal mucosa, all apparent in the histopathologically normal intestinal mucosa well before neoplastic changes become detectable. This includes pathways by which macrophages signal to intestinal epithelial cells, revealing a new paradigm for how vitamin D may influence tumor development.

Homeostasis in the colonic epithelium is established by maintaining a balance among proliferating, differentiating, and apoptotic cells. This balance is regulated by a complex interaction of intrinsic genetic signals with external stimuli including growth and dietary factors. Moreover, the interaction of signals maintains not only the quantitative balance between dividing and terminally differentiated cells, but a spatial and temporal coordination of cell proliferation, differentiation, and apoptosis within the crypt architecture as well.

The molecular genetics of tumor initiation and progression is understood better for colonic cancer than for any other human solid tumor. While much remains to be learned regarding the mechanisms that contribute to the altered growth and metastatic properties of colonic epithelial cells, and in particular, how multiple alterations in gene structure and function interact in affecting cell and tissue homeostasis, the advances thus far are being moved into translational research and direct clinical application for diagnosis and prognosis of colon cancer, and new treatment modalities will certainly develop from these areas of investigation. Moreover, the approaches — both conceptual and methodological — that have driven the research in this disease now serve as models for understanding the fundamental process of tumorigenesis in other organs. Thus, the work in colon cancer has established a new paradigm regarding the causes and clinical progression of human cancer, and has already formulated new approaches to clinical management of populations at risk for disease development and of patients with established tumors. Therefore, an understanding of the molecular genetics of colonic tumorigenesis is essential for the practicing oncologist.

There has been enormous progress in defining structural alterations of genes common in colonic cancer. Accumulation of mutations and deletions in APC, p53, DCC, and Ki-ras are the sine qua non of the disease, although the roles and interactions of these genetic alterations in the biological and clinical heterogeneity of the disease are not understood. Of even greater interest from the point of view of disease prevention are inherited mutations found in the APC gene and in genes which encode mismatch repair functions, which lead to the high frequency of colonic cancer in familial polyposis (FAP) and hereditary non-polyposis colon cancer (HNPCC) families, respectively1–6. However, unlike inherited childhood cancers, such as retinoblastoma and Wilm’s tumor, inherited colon cancer takes decades to develop. This can be partially understood in the context of what has been learned regarding genetic alterations in sporadic cancers. For example, one may assume that the inherited mutation supplies the first of a series of genetic alterations, and by so doing reduces the overall time and elevates the probability of accumulating the necessary number of alterations in genes such as those cited above. In the case of FAP, the inheritance of a mutant allele for APC provides one of the key events directly, since mutations in the gene occur somatically in over 70% of human colon tumors, and the mutation is sufficient to initiate development of intestinal tumors in mice7,8. In HNPCC, the situation is more complicated, in that the inherited mutations in genes responsible for DNA mismatch repair fail to repair errors which arise during DNA synthesis at tens of thousands of loci throughout the genome, and it must be presumed that among this plethora of changes there are key genes, again perhaps coincident with some of those mentioned above, which are eventual targets that lead to tumor formation.

In recent years, new approaches to the detection of early lesions associated with large bowel neoplasia have been explored. These approaches have led to the identification of a series of progressive stages of abnormal cell development, both in rodents exposed to carcinogens and in humans with hereditary predisposition to colorectal cancer.