Isolated crypts form spheres prior to full intestinal differentiation when grown as xenografts: an in vivo model for the study of intestinal differentiation and crypt neogenesis, and for the abnormal crypt architecture of juvenile polyposis coli

We describe a model system in which single crypts, isolated from newborn rats, were embedded in a type I collagen gel and subcutaneously grafted to the flanks of nude mice, whereupon they underwent full intestinal morphogenesis. Small fragments of small intestine and colon were incubated with the divalent cation chelator EDTA, resulting in the release of crypts and villi. Released crypts were then suspended sparsely in type I collagen gel. Segments of gel containing a single crypt were grafted subcutaneously into a nude mouse. Grafts were harvested at weekly intervals. By 2 days, the mouth of the crypts had joined to seal the crypt and, within 1 week, the structure ballooned to form a spherical cystic structure lined by flattened epithelial cells showing no evidence of cytodifferentiation. After 2 weeks, host stromal cells had invaded the collagen and settled around this spherical crypt. At points where stromal cells appeared in contact with the crypt, the epithelium exhibited a more columnar phenotype. By 4 weeks, the 'crypt sphere' was surrounded by stroma expressing alpha-smooth muscle actin and, at this time, multiple buds appeared that gave rise to new crypts. By 5 weeks, villi had formed and cell lineages associated with the small intestine and colon were present; the original single crypt had transformed into a functional intestinal unit. Therefore, we have shown that a single crypt has the potential to grow, give rise to other crypts and dependent structures such as villi. This model has considerable potential for use in gene transfer experiments in the study of intestinal differentiation, and for the analysis of crypt neogenesis via crypt fission. Moreover, the appearances showed a close resemblance to those seen in juvenile polyposis syndrome (JPS), where the budding and fission of single crypts isolated by stromal overgrowth offers an alternative explanation for the histogenesis of JPS.