A key question in developmental neurobiology is to reveal the gene regulatory mechanisms that control neuronal diversity. Although remarkable progress has been made in understanding how transcription factors (TFs) generate diversity, the role of epigenetic regulators in this process remains largely unknown. The motor circuit of Caenorhabditis elegans provides an ideal model system to probe the genetic and epigenetic mechanisms of neuronal diversity. We have previously shown that the evolutionarily conserved COE (Collier, Olf, Ebf) type TF UNC-3 acts as a terminal selector in the majority of ventral nerve cord (VNC) cholinergic motor neuron (MN) classes. Apart from controlling shared features among all MN classes (e.g. the expression of enzymes and transporters in the acetylcholine biosynthetic pathway), UNC-3 is also required for MN diversity by directly activating the expression of terminal identity genes (e.g. ion channels, neurotransmitter receptors) that are selectively expressed in specific MN classes (SAB, DA, DB, VA, VB, AS). However,
unc-3 is expressed in all these MN classes, leading us to hypothesize the existence of repressor proteins that restrict the ability of UNC-3 to activate these terminal identity genes in a broad, not class-specific manner. To test this hypothesis, we performed a forward genetic screen using a fluorescent reporter for the terminal identity gene
glr-4, which encodes a glutamate receptor subunit selectively expressed in the SAB class and DA9 MNs. We discovered that
pbrm-1, the sole C.elegans ortholog of the evolutionarily conserved chromatin regulator BAF180, acts as a repressor of the
unc-3-dependent gene
glr-4. Despite its ubiquitous and continuous expression,
pbrm-1 selectively controls AS terminal identity features by repressing
glr-4 only in the AS class of MNs. We further found that all other VNC MN classes are normally generated in
pbrm-1 mutants and the expression of several known MN class-specific,
unc-3-dependent genes is unaffected by PBRM-1. Since PBRM-1/BAF180 is a subunit of the PBAF, a chromatin remodeling complex of the SWI/SNIF family, we reasoned that animals lacking gene activity for other PBAF subunits may display a similar AS phenotype. We indeed found that SWSN-9, the sole C.elegans ortholog of human BRD7 and BRD9, is also required for
glr-4 repression in AS MNs. Neuron-specific rescue further demonstrated that PBRM-1 and SWSN-9 act cell-autonomously. Altogether, we provide novel insights on the epigenetic mechanisms that generate neuronal diversity by uncovering a previously unrecognized, neuron-specific role for the PBAF chromatin remodeling complex in the selective repression of terminal selector target genes.