The entire C. elegans intestine is derived from a single endodermal progenitor cell (E), the posterior daughter arising from the asymmetric division of the EMS blastomere. During early embryonic development, maternally provided SKN-1/Nrf2 activates the mesendoderm gene regulatory network (GRN) in both E and its sister, MS. A triply redundant Wnt/MAPK/Src signaling system from the neighboring P2 blastomere polarizes EMS, resulting in activation of E fate on the side contacting it. In MS, and in an unsignaled E cell, POP-1/Tcf represses expression of the redundant endoderm specifying factors, the END-1 and -3 GATA-type transcription factors. In a normal E cell, Wnt (initiated by the MOM-2/Wnt ligand) and MAPK signaling (through the MOM-4 MAPKKK) converge on POP-1 to convert it from a repressor to an activator of the end genes which, in collaboration with SKN-1, activates E cell fate (Thorpe et al. 1997; Maduro and Rothman 2002; McGhee 2007; Maduro 2017).
Basal transcription factor 3 (BTF3) facilitates transfer of nascent polypeptide chains into mitochondria and regulates transcription in plants and animals (Jamil et al. 2015). We previously found that the C. elegans BTF-3 orthologue, ICD-1, is required to prevent apoptosis: eliminating
icd-1 leads to increased cell death in embryos and larvae (Bloss et al. 2003). However, consistent with its potential function as a transcription factor, ICD-1 contains a putative nuclear localization signal in the N-terminus (Fig.1A) (Lange et al. 2007). Here, we report that ICD-1 performs a function in endoderm specification. We found that
icd-1 RNAi does not affect endoderm specification in a wild-type N2 background or the gut-less phenotype of
skn-1(-) embryos. However, knockdown of
icd-1 strongly suppresses the absence of gut in
mom-2/Wnt(-) embryos (
mom-2(
or42): 26.0% s.d. 4.5% with gut vs.
mom-2(
or42);
icd-1(RNAi): 75.9% s.d. 2.6%). Similarly, depleting ICD-1 rescues the gut-less phenotype of
mom-4/Tak1(-) embryos (
mom-4(
or39): 67.9% s.d. 6.3% with gut vs.
mom-4(
or39);
icd-1(RNAi): 95.5% s.d. 1.2%) (Fig. 1B). Our findings suggest a model in which ICD-1 antagonizes the SKN-1 input, perhaps by preventing SKN-1 from binding to
end-1/3promotors. This possibility is also consistent with the finding that competition between ICD-1 and SKN-1 is seen in the context of the unfolded protein response (UPR). SKN-1 binds to and activates
hsp-4, which codes for an endoplasmic reticulum chaperone BiP (Glover-Cutter et al. 2013). Depleting ICD-1 results in upregulation of
hsp-4 and activation of the UPR (Arsenovic et al. 2012), suggesting that ICD-1 and SKN-1 perform opposing functions in other contexts. In this hypothesized model, ICD-1 may act to fine-tune developmental signals, thereby ensuring proper specification and differentiation of endoderm (Fig. 1C).