The C. elegans intestinal specification network consists of an interlinked set of feed-forward loops consisting pairs of homologous genes that have similar but not entirely redundant functions. Knocking out one of the genes in a pair usually yields a more severe phenotype than knocking out the other. For example, the
end-3 deletion allele
ok1448 has a 7% penetrance while the
end-1 deletion allele
ok558 has zero penetrance.
elt-7 and
elt-2 are even more divergent. While
elt-2 deletions are lethal,
elt-7 is non-essential. Perhaps as a consequence,
elt-7 remains the least studied gene within the C. elegans intestinal specification network.
Gene expression noise has previously been shown to lead to partial penetrance of
skn-1 mutations in this network. We use single-molecule fluorescence in situ hybridization (smFISH) to quantitatively measure the levels and variation in expression profiles of each gene within this network under perturbations such as RNAi treatment and gene deletions. This approach reveals a conditional regulatory role of
elt-7 and clarifies the network topology. In wild-type N2 worms,
elt-7 is expressed concurrently with
end-1, suggesting that it is predominantly activated by
end-3. We observe delayed onset of
elt-7 and an increase of
elt-2 noise in
end-3 -/- worms, revealing that
elt-7 regulation of
elt-2 expression is compromised when
end-3 is absent. While
elt-7 and
elt-2 expression are similar to wild-type in
end-1 -/- worms, we show that
elt-7 functions to maintain
elt-2 level when
end-1 is present. This role would only be relevant and successful when
elt-7 is expressed at the appropriate time. By knocking down
elt-7 in
end-3 and
end-1 deletion strains, we measure the direct effects of
end-3 and
end-1 on
elt-2. Finally, we use our high resolution data to construct a mathematical model to explain how these deletions and knockdowns lead to gene expression noise and hence to partial penetrance in this network.