Nuclear receptor corepressors Ncor1 and Ncor2 (Smrt) are required for retinoic acid-dependent repression of Fgf8 during somitogenesis

- Genetic loss of Ncor1 and Ncor2 results in ectopic Fgf8 expression and small somites.
- NCOR1/2 corepressors, but not coactivators, are recruited to the Fgf8 RARE by RA.
- Genomic deletion of Fgf8 RARE with CRISPR/Cas9 often results in small somite defect.
- Nuclear receptor corepressors NCOR1 and NCOR2 mediate RA-dependent Fgf8 repression.

Retinoic acid (RA) repression of Fgf8 is required for many different aspects of organogenesis, however relatively little is known about how endogenous RA controls gene repression as opposed to gene activation. Here, we show that nuclear receptor corepressors NCOR1 and NCOR2 (SMRT) redundantly mediate the ability of RA to repress Fgf8. Ncor1;Ncor2 double mutants generated by CRISPR/Cas9 gene editing exhibited a small somite and distended heart phenotype similar to that of RA-deficient Raldh2−/− embryos, associated with increased Fgf8 expression and FGF signaling in caudal progenitors and heart progenitors. Embryo chromatin immunoprecipitation studies revealed that NCOR1/2 but not coactivators are recruited to the Fgf8 RA response element (RARE) in an RA-dependent manner, whereas coactivators but not NCOR1/2 are recruited RA-dependently to a RARE near Rarb that is activated by RA. CRISPR/Cas9-mediated genomic deletion of the Fgf8 RARE in mouse embryos often resulted in a small somite defect with Fgf8 derepression caudally, but no defect was observed in heart development or heart Fgf8 expression. This suggests the existence of another DNA element whose function overlaps with the Fgf8 RARE to mediate Fgf8 repression by RA and NCOR1/2. Our studies support a model in which NCOR1/2 mediates direct RA-dependent repression of Fgf8 in caudal progenitors in order to control somitogenesis.