We are using the C. elegans intestine to understand the robustness and flexibility of transcription factor networks and how these are influenced by the chromatin environment. The C. elegans endoderm develops under control of a series of partly redundant GATA-type transcription factors: END-3 > END-1 > ELT-7 > ELT-2. ELT-2 is the predominant transcription factor driving intestine differentiation and function, but is not normally involved in endoderm specification. We have recently shown (to our surprise) that ELT-2 can replace all other intestine GATA factors, i.e. can actually specify the intestine, if it is expressed under the
end-1 promoter in addition to its own promoter. However, multiple copies of transgenic
end-1p::
elt-2 are more efficient at specifying the endoderm (88% rescue) than a single copy MosSCI insertion (<1% rescue). In a genetic screen, we mutagenized approximately 14,000 genomes with EMS and identified 17 strains with enhanced ability of a single copy
end-1p::
elt-2 to specify endoderm. The starting
end-1(-)
end-3(-) strain, with a single copy insert of
end-1p::
elt-2, was rescued by an extra-chromosomal array carrying multiple copies of the
end-1p::
elt-2 together with a heat shock inducible
peel-1 construct. Thus we were able to kill all worms that still relied on the extra-chromosomal array for their survival after the EMS treatment. Whole genomic sequences of these independent strains provide a short list of potential candidate genes. Two such candidate genes that were independently mutated in multiple strains have been verified by RNAi or CRISPR: K01G5.9 (4 independent mutations), encoding a Taspase 1 homologue and
pqn-82 (2 independent mutations), encoding a homologue of a S. cervisiae SWI/SNF subunit. Taspase appears to be involved in the proteolytic processing of general transcription factors such as TFIIA and trithorax. The SWI/SNF complex appears to be involved in chromatin remodeling and transcriptional regulation. Disrupting the function of these two candidate genes shows a twofold enhancement of ELT-2 levels in the early endoderm relative to the unmutagenized strain, resulting in 50% rescue rate for
pqn-82 and 7% rescue rate for K01G5.9. We are currently working on a method to identify the mutated genes in the remaining 11 candidate strains that were hit only once. Thus, this genetic screen could be revealing a set of general chromatin proteins that control the levels of specific transcription factors, such as ELT-2, in the early embryo.