Specification of Neuronal Identities by Feedforward Combinatorial Coding

Open Access
Publisher Website
Google Scholar
Abstract
Neuronal specification is often seen as a multistep process: earlier regulators confer broad neuronal identity and are followed by combinatorial codes specifying neuronal properties unique to specific subtypes. However, it is still unclear whether early regulators are re-deployed in subtype-specific combinatorial codes, and whether early patterning events act to restrict the developmental potential of postmitotic cells. Here, we use the differential peptidergic fate of two lineage-related peptidergic neurons in the Drosophila ventral nerve cord to show how, in a feedforward mechanism, earlier determinants become critical players in later combinatorial codes. Amongst the progeny of neuroblast 5–6 are two peptidergic neurons: one expresses FMRFamide and the other one expresses Nplp1 and the dopamine receptor DopR. We show the HLH gene collier functions at three different levels to progressively restrict neuronal identity in the 5–6 lineage. At the final step, collier is the critical combinatorial factor that differentiates two partially overlapping combinatorial codes that define FMRFamide versus Nplp1/DopR identity. Misexpression experiments reveal that both codes can activate neuropeptide gene expression in vast numbers of neurons. Despite their partially overlapping composition, we find that the codes are remarkably specific, with each code activating only the proper neuropeptide gene. These results indicate that a limited number of regulators may constitute a potent combinatorial code that dictates unique neuronal cell fate, and that such codes show a surprising disregard for many global instructive cues. The nervous system contains a daunting number of different cell types, perhaps as many as 10,000 in mammals, far outnumbering regulatory genes in many animal species. Studies of the determinants of cell fate in many systems during the last decade have supported the conclusion that cell fate is not determined by any one regulatory gene, but results from the combinatorial action of several regulators. Many questions about the nature of such codes, however, remain. It is not known, for example, how complex such codes are or how they are established. It is also unclear whether they are confined in their action or if they act outside of their normal context. To address these outstanding issues, we have used two unique subsets of Drosophila neurons, identifiable by their specific expression of two different neuropeptide genes. We have identified two partially overlapping and relatively simple codes, consisting of four to seven regulators that act to specify these two cell types. Intriguingly, specification is achieved in a feedforward manner such that A activates B, followed by A/B activating C, and A/B/C activating D. Each code is surprisingly potent, and can ectopically activate neuropeptide gene expression in a variety of neurons, with a surprising disregard for many early patterning events.