A complex cellular choreography directs the development of the mammalian central nervous system; early in embryonic development, neural progenitor cells become neurons, whereas later in development they become the supporting glial cells. A step forward in understanding the molecular machinery of this change in cell fate is the recent identification of two transcription factors that serve as the key switches between neuronal and glial fates.

Hideyuki Okano and his colleagues at Keio University School of Medicine in Tokyo cultured mouse embryonic stem (ES) cells in conditions that led to the formation of neurospheres, which are groups of cultured cells that differentiate into neural cell types1. The primary neurospheres — ones that are first generated from the ES cells — develop into neurons, whereas subsequent neurospheres differentiate into glia. To find transcription factors that might be acting as the timed switch between neuronal and glial fates, the team screened the differentiating ES cells for transcription factor genes that were expressed at high levels in the first generation of neurospheres but at lower levels in a second generation of neurospheres.

They then knocked down the expression of these genes by RNA interference using small hairpin RNAs, and found that only those cells lacking expression of two transcription factors called Coup-tfI and Coup-tfII continued to produce neurons from successive generations of progenitor cells. This means that these transcription factors are probably part of the machinery required to switch neural progenitor cells to a glial fate. Their absence prevents the progenitor cells from executing their inbuilt program for producing glial cells.

In agreement with this, cells lacking Coup-tfI and II also continued to express Islet-1, a gene that is normally only expressed in the early stages of neuronal development. Knocking down Coup-tf expression in later-stage neural progenitors also allowed some of those cells to regain the ability to produce Islet-1-positive neurons. Cells lacking both Coup-tfs could not override the silencing of Gfap, a gene that, when active, leads neural progenitors to develop into glia. These cells were also less responsive to extracellular signal proteins that prompt glial cell development. Overall, Coup-tf expression may aid the transition from early to late neuronal development and may prepare progenitor cells to produce glial cells, the researchers say. Double knockdown experiments of Coup-tfs in developing mouse forebrain supported the role of these proteins in the timing of neural development.

Coup-tfs were already known to have a variety of roles in mammalian brain development, including the differentiation of oligodendrocytes, a type of glial cell, explains Takuya Shimazaki, a member of the Keio team. These experiments suggest that the Coup-tfs are working as gene repressors in this context, but a molecular mechanism for how they regulate the timing of neural cell specification is still unclear. “Given the many roles of Coup-tfs in mammalian brain development, it will be important to sort out how the same transcription factors seem to control, in a coordinated fashion, so many different aspects,” says Pierre Vanderhaeghen of the University of Brussels.

Okano and colleagues point to the known role of Seven-up, the fruitfly version of the Coup-tfs, in the temporal specification of neural cells in fly embryogenesis. Although more information on mechanism will be needed to bolster the presumed link between Coup-tfs and Seven-up, “this may provide a first hint of a common molecular logic between temporal patterning of neural specification in invertebrates and vertebrates,” says Vanderhaeghen.