The E cell is the progenitor of the entire C. elegans intestine. During embryogenesis, the E daughters, Ea and Ep, ingress, marking the onset of gastrulation. E undergoes up to five rounds of division to generate the 20 gut cells. The E cell is specified by the
end-1 and
end-3 genes, which are activated by maternal SKN-1 and its zygotic targets, MED-1 and MED-2 (Bowerman et al., 1992; Zhu et al., 1997; Maduro et al., 2001). In a fraction of
skn-1 and med mutant embryos, differentiated gut is made, but the E daughters do not ingress normally. This observation implies that gastrulation and production of intestinal cells from the E lineage are separable processes (Bowerman et al., 1994; Maduro et al., 2001). We have observed that many of the
skn-1(-) and
med-1,2(-) terminal embryos that make gut contain more than the wild-type number of gut cells, as revealed using gut-specific nuclear GFP markers. For example, among the ~50% of
med-1,2(RNAi) embryos that still make gut, 20% contain >20 gut nuclei. In both
skn-1 and med mutants, expression of
end-1 and
end-3 reporters is reduced, but not eliminated (Kasmir et al, WCWM 2000; Maduro et al., 2001). Moreover, we have observed extra gut cells in embryos in which
end-3 function is debilitated by a point mutation or by RNAi. In contrast, overproduction of
end-1 causes the E lineage to arrest prematurely with too few divisions. These data suggest that the
end-1,3 expression levels determine the number of cell division rounds that occur in the E lineage: elevated end levels apparently inhibit cell division, while subnormal levels allow for extra divisions. This observation may explain why the end genes are turned off shortly after they are first activated during normal development. We have found that mutants in which the E daughter cells fail to ingress (
gad-1 mutants; Knight and Wood, 1998, and
emb-5 mutants, Nishiwaki et al., 1993) also produce an excess of gut cells: e.g., over 60% of
gad-1(RNAi) embryos contain 25-30 gut nuclei. We found that
gad-1 embryos show delayed onset of
med-1 expression, suggesting that
end-1,3 expression may also be delayed or reduced. Therefore, GAD-1 and EMB-5 might indirectly affect gastrulation, perhaps by regulating the timely expression of
med-1,2 . Many maternal mutants are blocked in gastrulation but nonetheless produce intestine (e.g. cited in Knight and Wood, 1998), suggesting that there may be many components to this process. We propose that proper regulation of
end-1,3 expression is critical for (a) specification of E fate, (b) gastrulation, and (c) correct cell division patterns within the E lineage. Our observations also suggest that additional regulators of gastrulation, end gene expression, and E fate, might be identified by screening for mutants with elevated gut cell numbers.