The enteric system is formed by cells that migrate to the bowel from the neural crest. In order to gain insight into intraenteric factors that influence this migration, the colonization of the bowel of the ls/ls mouse was investigated. The terminal 2 mm of ls/ls intestine fails to become colonized by crest cells and thus remains aganglionic. The entire bowel of control mice and ls/ls mice was explanted before the appearance in situ of recognizable neurons and grown in organotypic tissue culture. Neurons, detected by the histochemical demonstration of acetylcholinesterase activity, developed throughout the length of the control gut, but, even in vitro, were excluded from the terminal segment of the ls/ls intestine. Co-culture experiments were done, in which primary and secondary sources of crest cells were combined with recipient segments of bowel, to test the ability of the recipient tissue to become colonized by neural precursors. The primary source was murine crest cells migrating away from an explant of the neuraxis. Secondary sources included avian and murine foregut (control and ls/ls) containing migratory crest cells as well as the quail ganglion of Remak. Recipient segments of bowel included control avian and murine hindgut, explanted before the tissue had become colonized by crest cells in situ, as well as the presumptive aganglionic bowel of ls/ls mice. Both primary and secondary sources of crest cells proved to be able to contribute neurons to the control segments of recipient hindgut. Species differences were no barrier to the colonization of the bowel in vitro. Moreover, the ls/ls foregut was as good a source of neural precursors for a normal recipient bowel, as was control avian or murine foregut. In contrast, none of the sources of crest cells that were utilized contributed neurons to the presumptive aganglionic gut of ls/ls mice. Both cells and processes of enteric neurons developing in vitro (detected by demonstrating neurofilament immunoreactivity) tended to be excluded from the presumptive aganglionic tissue. On the other hand, neurites, but not cell bodies, of dorsal root ganglia co-cultured with presumptive aganglionic ls/ls bowel did enter the abnormal zone. These data are consistent with the hypothesis that nonneuronal elements of the wall of the presumptive aganglionic region of the ls/ls gut are abnormal and prevent the colonization of this segment of the gut with viable neural precursors from the neural crest.(ABSTRACT TRUNCATED AT 400 WORDS)