What the fly's glia tell the fly's brain

Introduction The brain is an astonishingly complex organ that records, responds and adapts to experience-all the result of simple interactions between neurons and glia. Whereas the neurons interconnect to form electrically active circuits, the role of glia is a mystery (Figure 1). Do glial cells just provide a passive framework that supports, nourishes, and insulates neurons, or, in addition, do glial cells play more active roles in signaling and plasticity? The importance of glia is suggested by their increase in number during evolution; glial cells constitute 25%, 65%, and 90% of cells in the Drosophila, rodent, and human brain, respectively. How can the unknown, and perhaps unimagined, functions of a cell type be determined? Studies of gliausing histology, immunostaining, tissue culture, and electrophysiology have led to a number of hypotheses about what glial cells do (Barres, 1991). For instance, glial cells may guide developing neurons and their axons to their targets, may promote the survival and differentiation of neurons, may form the blood-brain barrier, and may help regulate the extracellular concentrations of ions and neurotransmitters. The problem has been how to test these hypotheses definitively. Would it matter if there were no glia? In two landmark studies, the labs of Corey Goodman and of Yoshiki Hotta have selectively eliminated glial cells in vivo to explore how the brain develops and functions without them (Hosoya et al., 1995; Jones et al., 1995). They have identified a Drosophila gene, glial cells missing (gem), that encodes a novel nuclear protein that is required for glial cell fate determination and is transiently expressed early during the development of nearly all glia. In the absence of functional GCM protein, nearly all glial cells differentiate into neurons, whereas ectopic expression of gem in immature neurons or their precursor cells transforms them into glia (Figure 2). Thus, gem functions as a binary genetic switch that determines whether developing neural cells will become neurons or glia. Although loss-of-function homozygousgcm mutations are ultimately lethal to thedeveloping embryos, much of neural development is completed prior to their death, allowing a direct examination of the role of glia in neural development. In addition, mutation of another Drosophila glia-specific homeodomain protein, REPO, has also been found to eliminate many glia (Campbell et al., 1994; Halter et al., 1995; Xiong and Montell, 1995). In this minireview, we will summarize the insights into glial function that fly genetics have recently provided; the implications of the gem genetic switch for understanding how neural cell fate is controlled are discussed elsewhere (Anderson, 1995). Minireview