Although proneural genes are known to regulate the differentiation, position and identity of neurons during the early development of a wide range of organisms — from flies to humans — it is unclear which genes lie immediately downstream of them. A series of experiments using transgenic mice, reported in Development, show, for the first time, a conclusive link between a proneural gene and its direct target, and the authors link this interaction to a specific neuronal fate.

Proneural genes encode basic helix–loop–helix (bHLH) proteins, which activate transcription by binding to the E-box regulatory sequence of their target gene. In the developing spinal cord of the mouse, cells adjacent to the roof plate express the proneural gene Math1 and also a Bar-class homeobox gene Mbh1. Previous work has shown that Math1 is involved in controlling the number of dI1 cells and commissural neurons, that dI1 cells express Mbh1, and that Mbh1-positive cells give rise to commissural neurons. However, until now, the functional relationship between the two genes and their influence on neuronal cell fate was unclear. Now Saba and colleagues have shown that Mbh1 is a direct downstream target of MATH1 and that MBH1 confers commissural neural identity in the spinal cord.

Using a series of deletions and a LacZ reporter gene, the authors discovered that the Mbh1 enhancer contains an E-box that is well conserved among mice, rats and humans, and showed that a mutation known to disrupt the binding of MATH1 to this site abolishes Mbh1 expression. This proved that Mbh1 expression in the dorsal spinal cord requires an E-box, as might be expected if MATH1 binds directly to regulate Mbh1 expression. The relationship between Mbh1 and MATH1, which was indicated by other studies, was then confirmed by knockout and overexpression studies. Mbh1 expression was absent from Math1-knockout embryos, and expressing MATH1 on the sides of the dorsal cord led to the ectopic expression of Mbh1 in these regions.

Biochemical assays reinforced the conclusion drawn from the in vivo experiments. An anti-MATH1 antibody specifically immunoprecipitated Mbh1 enhancer DNA fragments that contain the E-box, thereby showing that endogenous MATH1 bound the enhancer at the binding site. Similar results were obtained for misexpressed MATH1. But what does Mbh1 do? Evidence obtained from chimaeric proteins of MBH1 showed that the protein requires a transcriptional-activator domain to inhibit the generation of commissural neurons by MATH1, which indicates that MBH1 is a transcriptional repressor.

This is the first demonstration that an aspect of neuronal identity is determined immediately downstream of a bHLH protein, and reveals a chain of events leading to specific commissural neural identity in the spinal cord.