Functional specification of the Caenorhabditis elegans nervous system by homeodomain transcription factors

Nervous systems are made of a large diversity of neuron types, that are characterized by specific molecular markers, their unique morphology and connectivity patterns, or their electrophysiological profiles. Those terminal differentiation features are specified by combination of transcription factors, defining genetic programs that controls neuronal identity. Some transcription factors act as "master regulators" or "terminal selectors", binding directly and continuously to many effector genes of a specific neuron type. Neuronal identity is specified by the code that the combinations of those transcription factors define. It is hard to assess to which extent those mechanisms shape more complex nervous systems, while C. elegans has a small number of well- defined neuron classes that make it more experimentally tractable. Thanks to CRISPR tagged alleles, a novel multicolor C. elegans reporter transgene strain and recent scRNA-Seq datasets, it has been possible to determine the full expression pattern of the gene family that is most likely to control neuronal identity, the highly conserved homeodomain transcription factors. From this, it is now possible to study their functional effects on neuronal identity on the full complement of C. elegans neuron types. I show that previously understudied homeodomains can indeed act as terminal selectors. I find that multiple types of neurons that had no previous known master regulators can be attributed a homeodomain regulator. I focus on the SIX homeodomain subfamily and identify a new type of subordinate transcription factor, "subtype terminal selectors". I find that such a "subtype terminal selector" is sufficient and necessary to specify a subset of features separating two very closely related neuron types, under the control of a shared terminal selector that control both shared and subtype specific features in those classes. Homeodomains remain front and center in neuronal identity control and could plausibly contribute to the specification of every single neuron class in C. elegans. Moreover, they have very conserved roles starting even in primitive nervous systems and likely played a major contribution to the evolution of cellular diversity in the nervous system. With my subtype terminal selector examples, I show possible mechanisms through which an ancestrally shared neuron type could progressively diverge towards two distinct neuron types