Potential interplay between luminal growth factors and increased tight junction permeability in epithelial carcinogenesis.

Abstract

It has been generally accepted for many years that carcinogenesis proceeds in at least two distinct stages, the first being initiation, the second being tumor promotion (Boutwell, ’74). Initiation has been shown to be a heritable change imposed on DNA itself, whereas promotion is regarded as an extranuclear event, non heritable, and requiring chronic exposure to a tumor promoting agent (Diamond et al., ’80). One may think then of carcinogenesis as a series of dominoes which must fall in sequence in order to finally produce the condition which we term cancer. One can likewise conceive of it as a series of necessary and finally sufficient events, leading in the end to a tumor. The majority of lethal cancers are epithelial in origin (Fraumeni et al., ’89). The fact that epithelia line all of the major body fluid compartments places epithelial cells specifically in more pronounced contact with environmental carcinogenic agents than other somatic cells. In other words, if the process of carcinogenesis requires a tumor promotion phase entailing chronic uninterrupted exposure to an exogenous tumor promoting substance, epithelial cells in tissues such as lung and colon would be repetitive targets. There are two key characteristics of epithelia, two traits which are common among all of the various epithelial tissues. First, all epithelial cells are intrinsically polar, having an apical cell surface facing the luminal fluid compartment, and a basal-lateral membrane facing the interstitial fluid compartment. These two membrane domains contain not only different classes of membrane proteins, but distinct lipid compositions as well (Schneeberger and Lynch, ’92). The second characteristic common through all epithelial tissues is the most basic function of any epithelium, namely its ability to form a barrier between the two fluid compartments which the epithelium separates. The common thread between these two most fundamental characteristics is the tight junction or zonula occludens. This protein (or protein and lipid) circumferential band around the apical pole of each epithelial cell is first and foremost the selectively permeable diffusion barrier for the paracellular pathway. It prevents free flow of solutes from one fluid compartment to the other along the paracellular route between epithelial cells (Farquhar and Palade, ’65). The tight junction also has a role in maintaining the apical/ basal-lateral polarity of individual epithelial cells (Cereijido et al., ’89). When considering the scope of epithelial cancers across such a variety of tissues, and the fact that according to the two-stage carcinogenesis model an extranuclear change is necessary for epithelial cancer, the primacy of epithelial polarity and barrier function makes them key areas on which to focus attention. Evidence for altered permeability of tight junctions in transformed epithelia has been known for almost 30 years (Martinez-Palomo, ’70). A decrease in the number of tight junctional strands has been described in transitional carcinoma of the urinary bladder (Saito, ’84). Decreased transepithelial impedance has been recorded across the colons of mice treated with chemical carcinogens (Davies et al., ’89). Inflammatory bowel disease linked with increased cancer risk has itself been linked with increased tight junction permeability, not only in affected patients but in unaffected relatives as well (Hollander, ’88). On a molecular level, the tight junctional protein, ZO-1 (Stevenson et al., ’89), has been shown to possess significant sequence homology to a septate tumor suppressor protein of Drosophila, whose mutation leads to neoplastic growth of epithelia within larvae (Willott et al., ’93; Woods and Bryant, ’91). Finally, the interaction of the normal APC (adenomatous polyposis coli) colon cancer susceptibility gene product with the cell adhesion protein, β-catenin