Genetic regulation of enterocyte differentiation

Partly digested food entering the small intestine is finally broken down into a number of simple components by a variety of hydrolytic enzymes present on the surface of intestinal villi. These products of hydrolysis are then transported across the intestine by different carrier molecules present on the microvillus surface of villus enterocytes. Microvilli expand considerably the surface area over which nutrient digestion and absorption can take place. Genetic defects affecting any one of these three important components of enterocyte differentiation can lead to increasing morbidity and death (Milne, 1974; Freiburghaus et al. 1976; Phillips & Schmitz, 1992). Other factors affecting differentiation include interactions taking place between enterocytes and the underlying mesodermal tissue (Haffen et al. 1989) and between enterocytes and the gut-associated immune system (Smith, 1991). Change of diet and the proliferative state of the gut further affect the way in which enterocytes express microvillus structure and digestive and absorptive functions (Smith, 1992). All these factors should be considered by anyone wishing to understand how diet and health affect intestinal function. Use of appropriate experimental models has been important in allowing specific aspects of enterocyte differentiation to be studied separately. Tissue recombination experiments were needed, for instance, to positively identify mesodermal effects on enterocyte development (Kedinger et al. 1981); immunocyte effects on enterocyte differentiation are best studied in germ-free animals challenged by a single organism (Savidge et al. 1991), and sugar induction of disaccharidase expression is best interpreted in the absence of any effect on crypt cell proliferation (Collins et al. 1989). However, studying normal enterocyte differentiation in culture has not proved successful as yet and the use of transgenic animals has only recently begun to muddy the waters (Hauft et al. 1992). Dealing intelligently with known variables seems on the whole to be preferable to ignoring them or artificially increasing their number.