Transcriptional regulation of glial cell specification

Available online at www.sciencedirect.com R

Developmental Biology 255 (2003) 138 –150

www.elsevier.com/locate/ydbio

Transcriptional regulation of glial cell specification Gianluca Ragone,1 Ve´ronique Van De Bor,1 Sandro Sorrentino, Martial Kammerer, Anne Galy, Annette Schenck, Roberto Bernardoni, Alita A. Miller, Nivedita Roy, and Angela Giangrande* Institut de Ge´ne´tique et Biologie Mole´culaire et Cellulaire, IGBMC/CNRS/ULP/INSERM - BP 10142 67404 Illkirch, c.u. de Strasbourg, France Received for publication 18 July 2002, revised 24 September 2002, accepted 13 November 2002

Abstract Neuronal differentiation relies on proneural factors that also integrate positional information and contribute to the specification of the neuronal type. The molecular pathway triggering glial specification is not understood yet. In Drosophila, all lateral glial precursors and glial-promoting activity have been identified, which provides us with a unique opportunity to dissect the regulatory pathways controlling glial differentiation and specification. Although glial lineages are very heterogeneous with respect to position, time of differentiation, and lineage tree, they all express and require two homologous genes, glial cell deficient/glial cell missing (glide/gcm) and glide2, that act in concert, with glide/gcm constituting the major glial-promoting factor (Hosoya et al., 1995; Jones et al., 1995; Kammerer and Giangrande, 2001; Vincent et al., 1996). Here, we show that glial specification resides in glide/gcm transcriptional regulation. The glide/gcm promoter contains lineage-specific elements as well as quantitative and turmoil elements scattered throughout several kilobases. Interestingly, there is no correlation between a specific regulatory element and the type of glial lineage. Thus, the glial-promoting factor acts as a naive switch-on button that triggers gliogenesis in response to multiple pathways converging onto its promoter. Both negative and positive regulation are required to control glide/gcm expression, indicating that gliogenesis is actively repressed in some neural lineages. © 2003 Elsevier Science (USA). All rights reserved. Keywords: glide/gcm; Glia; Specification; Nervous system; Fly; Promoter

Introduction Nervous system development depends on transcription factors that promote either glial or neuronal differentiation in multipotent precursors. While it has been shown that the same instructive factor influences neuronal differentiation and specification, it is not known whether this also applies to glia (Guillemot, 1999; Isshiki et al., 2001; Skeath and Doe, 1996). Fly glial cells differentiate at specific times and positions from two types of precursors: glioblasts (GBs), which only produce glia, and neuroglioblasts (NGBs), which produce glia and neurons. Thus, the control of glial specification responds to precise positional information. Gliogenesis de-

* Corresponding author. Fax: ⫹33-388653201. E-mail address: [email protected] (A. Giangrande). 1 G. Ragone and V. Van De Bor contributed equally to this work.

pends on Glial cell deficient/Glial cell missing (Glide/Gcm), a transcription factor that controls the differentiation of all glia of the peripheral nervous system (PNS) and that of most lateral glia of the central nervous system (CNS) (Hosoya et al., 1995; Jones et al., 1995; Vincent et al., 1996). A homologous gene, glide2, is expressed at a lower level and plays a somewhat less important role in gliogenesis than glide/gcm (which will be referred to as glide, for the sake of simplicity) (Kammerer and Giangrande, 2001). glide and glide2 are both necessary and sufficient for glial differentiation, but do not seem to dictate lineage specificity (Akiyama-Oda et al., 1998; Bernardoni et al., 1998; Hosoya et al., 1995; Jones et al., 1995; Kammerer and Giangrande, 2001; Van De Bor and Giangrande, 2002). Indeed, the findings that the profile of glide expression resembles that of gliogenesis and that a single type of gene accounts for all lateral glial promoting activity strongly suggest that the major step controlling both gliogenesis and the type of glial

0012-1606/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0012-1606(02)00081-7

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

cell resides at the level of glide transcription. Thus, understanding the regulation of the glial promoting factor is fundamental to unravel the mechanisms that control cell specification in the nervous system. Here, we present the functional characterisation of the glide promoter by germ line rescue of the mutant phenotype. By using progressive promoter deletions, we show that several upstream regulatory sequences are necessary to control glide expression in all glial lineages. Rescue of identified lineages depends on distinct fragments, strongly suggesting that glide is directly regulated by genes that assign lineage identity in the nervous system. Interestingly, glial cells rescued by a given fragment do not share common features concerning their lineage tree, position, or time of differentiation. We also show that high levels of glide expression are a prerequisite for glial differentiation. Finally, we provide evidence that glide must be repressed in neuronal lineages, indicating that correct glial differentiation requires both positive and negative regulation of the glial promoting factor. This first report on a glial-specific promoter indicates that the control of glial differentiation and specificity resides on the convergence of several pathways onto a single locus at the crossroads between glia and neurons. Such pathways involve factors that assign cell identity along the main body axes as well as factors that control neural vs epidermal fate choice in the fly embryonic neurogenic region.

Materials and methods

139

26], respectively. For all rescue experiments, series of transgenes showing comparable levels of glide expression were used. Each transgene was analysed in homozygous conditions. In situ hybridisation and immunolabelling Embryo preparation, antibody incubation, and in situ hybridisation were performed as in Bernardoni et al. (1997). Digoxigenin-labelled glide riboprobe and Fluorescein-labelled huckebein riboprobe were obtained by using fulllength cDNAs. ladybird early riboprobe was obtained by using an 800-bp cDNA fragment. Embryos were mounted in Vectashield medium (Vector). The following primary antibodies were used: rabbit anti-Repo (1:1000) (provided by A. Travers), mouse anti-22c10 (1:10) (provided by S. Benzer), rat anti-RK2 (1:1000) (provided by A. Tomlinson), rabbit anti-␤Gal 55976 (1:1000) (Cappel), rabbit anti-Eagle (1:1000) (provided by J. Urban), mouse anti-Invected (1: 10), mouse anti-BP102 (1:100) (from DSHB), mouse antiFasII (1:50) (provided by J. Urban), rat anti-Krupple (1:200) (gift from D. Kosman), mouse anti-digoxigenin (1:1000) (Boehringer), and mouse anti-fluorescein (1:1000) (Molecular Probes). Anti-Invected is currently used to follow the expression of engrailed (Patel et al., 1989). Secondary antibodies coupled to Oregon Green (Molecular Probes), Cy3, Cy5, and FITC (Jackson), were used at 1:400. Preparations were analysed by using a confocal microscope (DMRE, Leica).

Fly strains

Cloning

The wild-type strain was Sevelen. The sevenup-lacZ /TM3 (kindly provided by G. Flores) and huckbeinlacZ lines were used to label subsets of glia. The w1118 strain was used to generate transgenic lines. glideN7-4/CyO twilacZ and glide26/CyO twi-lacZ were the glide null mutant strains (Kammerer and Giangrande, 2001; Vincent et al., 1996). While glideN7-4 contains a mutation that abolishes DNA binding, glide26 carries a deficiency of the transcribed sequences. The following strains were used to characterise glide regulatory sequences: [w; P(2kb-glideM25B,w⫹), N74/CyO twi-lacZ], [w; P(2kb-glideM25B,w⫹), 26/CyO twilacZ], [w, P(4kb-glideF5,w⫹); N7-4/CyO twi-lacZ], [w, P(4kb-glideF5,w⫹); 26/CyO twi-lacZ], [w; P(6kbglide8M6,w⫹), N7-4/CyO twi-lacZ], [w, P(6kb-glide8M6,w⫹), N7-4/CyO twi-lacZ; P(6kb-glide4M6,w⫹), mshlacZ/TM3], [w; P(6kb- glide8M6,w⫹), 26/CyO twi-lacZ], [w; P(9kbglide4F6,w⫹), N7-4/CyO twi-lacZ], [w; P(9kb- glide4F6,w⫹), N7-4/CyO twi-lacZ; P(9kb-glide6F5,w⫹), hkblacZ/TM3], [w; P(9kb-glide4F6,w⫹), 26/CyO twi-lacZ]. For the sake of simplicity, these lines have been called [2kb-glide, N7-4], [2kb-glide, 26], [4kb-glide; N7-4], [4kb-glide; 26], [6kbglide, N7-4], [6kb-glide, N7-4; 6kb-glide], [6kb-glide, 26], [9kb-glide, N7-4], [9kb-glide, N7-4; 9kb-glide], [9kb-glide,

Genomic DNA (13 kb) containing the glide coding sequences as well as 3⬘ and 5⬘ sequences was isolated from a recombinant phage (J. Reed and A. G., unpublished data). A genomic HindIII–SalI fragment of 5.6 kb containing the glide coding sequences, 2 kb of upstream, and 1.7 kb of downstream sequences was subcloned in the CASPER4 injection vector. This construct was called 2kb-glide. Three other constructs were produced, containing progressively longer fragments of 5⬘ sequences: 4kb-glide, 6kb-glide, and 9kb-glide, corresponding to EcoRI–SalI, XbaI–SalI, and EcoRI–SalI fragments, respectively. These constructs were injected to w1118 flies. Different transgenic lines were obtained for each construct and crossed into glide null mutant backgrounds (glideN7-4 or glide26) in order to establish the extent of rescue.

AE127

DNA-binding assay Glide binding sites (GBS) were searched by using the MatInspector algorithm (Heinemeyer et al., 1999). The GBS matrix is based on the consensus sequence defined by Akiyama et al. (1996). Search was performed at high (core sequence ⫽ 1.00 and matrix similarity ⫽ 0.85) and low

140

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

Fig. 1. Organisation of the CNS precursors. Schematic representation of CNS precursors from a thoracic and an abdominal hemisegment. Anterior is to the top; vertical line on the left indicates the midline. Neuroblasts, glioblasts, and neuroglioblasts are represented by different patterns: coloured ovals (NBs), vertical stripes (NGBs), and crossed stripes (GBs). LGB indicates the longitudinal glioblast, and MNB the median neuroblast. S1–S5 represent the five waves of precursor delamination.

(core sequence ⫽ 0.70 and matrix similarity ⫽ 0.70) stringencies. A GST fusion protein containing the first 202 amino acids of Glide, GST- glideDBD, was produced as described in Miller et al. (1998). As probes, 30 mers containing the different GBSs were prepared as described in Miller et al. (1998). Five-hundred nanograms of purified GST-glideDBD were incubated with 5000 c.p.m. of the 32P-labeled probe as described in Akiyama et al. (1996). The reaction mixtures were loaded on an 8% polyacrylamide gel and run at room temperature.

Results Glial cell organisation in the fly embryo Glial cells of the central nervous system (CNS) are subdivided into two classes (Van De Bor and Giangrande, 2002). Midline glia are necessary for the establishment of the commissures (Jocobs, 2000). Lateral glia, which we will

refer to as glia throughout the text, are located along the neuropile and within the cortex to insure neuronal function and survival. Some lateral glial cells (also called peripheral glia) migrate to the periphery and line sensory/motor axon bundles (Granderath and Klambt, 1999; Ito et al., 1995; Van De Bor and Giangrande, 2002). Glia constitute a rather heterogeneous population of cells with regard to their origin, morphology, time, and position of precursor delamination, suggesting quite a complex regulation. According to their lineage tree, glial precursors can be subdivided into two classes: pure precursors, called glioblasts (GBs) (longitudinal glioblast or LGB and 6-4abdominal or 6-4A), and mixed precursors, called neuroglioblasts (NGBs) (1-1abdominal or 1-1A, 2-2thoracic or 2-2T, 2-1, 1-3, 2-1, 2-5, 3-5abdominal or 3-5A, 5-6, 6-4thoracic or 6-4T, 7-1, 7-4) (Fig. 1). Among mixed precursors, 6-4T constitutes an example of type I NGB, in the sense that it produces glia and neurons as it delaminates. Type II NGBs, which produce neurons before producing glial cells, include 1-1A, 2-2T, and 5-6 (Bernardoni et al., 1999; Bossing et al., 1996b; Schmidt et al., 1999; Schmidt et al., 1997; Udolph et al., 2001). Glial differentiation is controlled by positional cues, since distinct glial lineages differentiate within each segment according to precise D/V and A/P coordinates. Moreover, glial specification is under the control of homeotic genes: some precursors are gliogenic only in the thoracic or in the abdominal segments (1-1A, 2-2T, 3-5A, and 6-4A). Finally, glial differentiation is tightly temporally regulated: (1) glial cells do not arise all at the same time; and (2) most glial cells are produced by stem cells that delaminate in the first waves. Since the glide locus contains all glial promoting activity, we used glide (1) to follow gliogenesis; and (2) to dissect the process that controls glial differentiation and specification in time and space. Definition of the glide promoter by germ line transformation To identify the sequences that control glial differentiation and specification, we analysed the glide promoter in vivo by rescue of the mutant phenotype. Two lines of evidence strongly suggested that the glide regulatory region is rather complex. First, a GAL4 line that contains 11 repeats of the Glide Binding Site (GBS) does not recapitulate glide expression in the sense that it shows mosaic expression, and only 1 or 2 glioblasts and a few other glia express GAL4 in the nervous system. In addition, it also shows neuronal expression (Booth et al., 2000). Second, glide is part of a gene complex (Kammerer and Giangrande, 2001). We therefore produced a construct carrying 9 kb of 5⬘ sequences, the glide coding sequences, and 1.7 kb of 3⬘ sequences. We then performed a terminal deletion analysis and produced lines containing 6, 4, or 2 kb of upstream

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

Fig. 2. glide locus, transgenic lines and GBSs. (A) Organisation of glide locus and transgenic constructs. Line at the top represents sequences at 30B containing glide and glide2. Below the map are indicated the constructs used for transgenesis, containing from 2142 bp (2 kb) to 8920 bp (9 kb) of the glide 5⬘ sequences. EcoRI (E), HindIII (H), SalI (S), XhoI (X). Ovals represent the putative GBSs found using MatInspector against the GBS matrix. Black ovals indicate the 10 GBSs found between glide and glide2 (sites 1–10) when performing a high-stringency search. White ovals indicate sites found when performing a low-stringency search (sites I–VI). Sites in opposite orientation are labelled with an “r.” (B) Relative abilities of different sites to bind a purified GlideDBD fusion protein; nonspecific DNA (NS). (C) Sequence of putative GBSs used in the gel shift assay.

sequences (Fig. 2). For the sake of simplicity, these transgenes will be referred to as 9, 6, 4, and 2 kb, respectively. Gain-of-function experiments have shown that high levels of Glide are necessary to induce gliogenesis, suggesting that quantitative regulation plays an important role during development. Using the GBS identified by selection assays (Akiyama et al., 1996; Schreiber et al., 1997) and the MatInspector logarithm [(Heinemeyer et al., 1999); see Materials and methods], 10 putative GBSs were found in the 27 kb between glide and glide2 (GBS 1 to 10 in Fig. 2A) (Kammerer and Giangrande, 2001). Using less stringent parameters, we found more putative sites and tested them in gel shift binding assays (sites I to VI in Fig. 2A). As expected, GBS 1 to 3, which were found at high stringency, bind Glide very efficiently. Interestingly, some sites identi-

141

fied under low-stringency conditions and containing mismatches are also able to bind Glide efficiently (Fig. 2B and C), indicating that the 2 kb contains at least one strong binding site (I), the 4 kb two (I and 1), the 6 kb three (I, 1 and IV), and the 9 kb five strong (I, 1, IV 2, 3) and one weak sites (VI) (Fig. 2). Each transgenic line was introduced into the mutant background and analysed in homozygous conditions. For each construct, different lines were obtained, each expressing the transgene at levels that depend on the site of insertion. Accordingly, the eye color of these lines varied from pale yellow to bright red or wild type (white being the marker contained in the Casper injection vector). To take position effects into account, we analysed rescue levels in a pair of lines (from the same construct) showing a comparable white⫹ eye phenotype and found similar results. Because the results are consistent, we only show data from one line per construct. The series of 2-, 4-, 6- and 9-kb transgenic lines used and shown in the rescue analysis displays the same eye color, in order to compare transgenes showing similar levels of glide expression. glide null embryos (glideN7-4) lack most glial cells. In the ventral cord, these cells are completely absent (Fig. 3B) or very rarely present 3 cells in average throughout the whole ventral cord against 60 cells present in each neuromere of a wild type ventral cord). This dramatic and invariable glial phenotype leads to major defects in axonal guidance and fasciculation (Fig. 4B, H, and K). Such defects result in an overall lack of nerve cord condensation (Hosoya et al., 1995; Jones et al., 1995; Vincent et al., 1996). The rescue potential of the different constructs was assessed with respect to four parameters: (1) glial differentiation, (2) axonal phenotype, (3) ventral cord condensation, and (4) profile of glide expression. Glial rescue was assessed by analysing the expression of Reverse Polarity (Repo), a nuclear protein expressed in all lateral glial cells (Campbell et al., 1994; Halter et al., 1995; Xiong et al., 1994) (Fig. 3). Glial lineages were identified by position, time of differentiation, and expression of lineage-specific markers (see below). Introduction of the 2-kb transgene leads to rescue of subperineural glia (SPG), intersegmental/segmental nerve glia (ISNG, SNG), and peripheral glia (PG) (Fig. 3C, and data not shown). These glial cells arise from 1-1A, 1-3, and 2-2T, suggesting that: (1) glide expression has been rescued in those lineages, and (2) the 2 kb contains lineage-specific elements. Thus, rescue is obtained at ventral and dorsal positions, from early and late delaminating neuroglioblasts. The 4 kb induces more cells to express repo (Fig. 3D). In addition to the lineages rescued by the 2 kb, central body glia (CBG), channel glia (CG), and some longitudinal glia (LG) are also rescued (Fig. 3D, and data not shown). These glia arise from the 5-6, the 7-4, and the LGB, respectively. A higher number of glial cells was observed upon rescue with the 6 kb (Fig. 3E) and even higher with the 9 kb, even though gaps are still present along the longitudinal connec-

142

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

tives (Fig. 3F). For most glia, rescue penetrance and expressivity increase with the size of the transgene. Finally, better rescue of the glial phenotype was observed with two 4-kb, 6-kb, or 9-kb transgenes than with single transgenes, suggesting an important role for quantitative regulation (Fig. 3, compare E and G for the 6-kb, and F and H for the 9 kb). Indeed, more glial cells are rescued within the lineage and/or the lineage is rescued more frequently with two doses than with one. Nevertheless, two 6-kb copies do not show the same extent of rescue as that shown by one 9-kb copy (Fig. 3G and F), supporting the idea that qualitative elements are present between 6 and 9 kb upstream to glide. Rescue of axonal defects Genetic and toxin-mediated ablation of lateral glial cells results in axon defasciculation and defects in pathfinding, indicating that glial cells are necessary for axonal navigation (Hidalgo and Booth, 2000; Hidalgo et al., 1995; Hosoya et al., 1995; Kinrade et al., 2001; Takizawa and Hotta, 2001; Vincent et al., 1996). To assess the rescue of glial function, we tested whether Repo-positive cells are able to provide cues to navigating axons. The Fas II antibody labels motoneurons and three longitudinal fascicles per hemisegment (Grenningloh et al., 1991) (Fig. 4A). In glide null embryos, only one or two fascicles, the most medial ones, are detectable in several segments (Fig. 4B). Moreover, some axons abnormally cross the midline (Kinrade et al., 2001) (see arrowhead in Fig. 4B). The fact that lateral longitudinal fascicles are not lost in all segments suggests that cells that should have become glia can provide some guidance cues to axons, even in the absence of Glide. This is likely due to the presence of some glide2 expression. Indeed, the expressivity of the phenotype increases when all glial-promoting activity is lost using a deficiency that removes both glide and glide2. mAb22c10 recognises the Futsch microtubule associated protein, which labels peripheral and central axons (Hummel et al., 2000; Roos et al., 2000). In wild-type embryos, peripheral nerves navigate separately as they exit the CNS and then meet and migrate together toward the periphery (Fig. 4J). In mutant embryos, the two nerves meet more distally or never do so (compare arrowheads in Fig. 4J and K) (Takizawa and Hotta, 2001; Vincent et al., 1996). Finally, the BP102 serum labels many CNS axons running along the longitudinal connectives and the commissures (Hummel et al., 1999) (Fig. 4G). In glide mutants, connectives are thin or interrupted, commissures appear fused, and the ventral cord fails to condense (Fig. 4H, and data not shown). Using the three markers, we found that rescue of Repo labelling nicely correlates with rescue of axonal defects in the CNS and in the PNS (Fig. 4C–F, I, and L, and data not shown). Defects are progressively rescued as transgene length increases. Indeed, rescue of longitudinal connectives

is best in 6 kb and 9 kb, which also show the best longitudinal glia rescue (Fig. 3). Finally, nerve condensation is also progressively rescued as transgene length increases (data not shown). Thus, glial cells present in the rescued background have recovered their function. glide expression and glial differentiation The promoter analysis shows that glide cis-regulatory elements control glial specification. Since glide is the earliest indicator of glial differentiation, we used double labelling with glide riboprobe and lineage-specific markers to identify the different types of glial cells in wild-type embryos. Since lineage position and birthdate are very stereotyped, these parameters also helped in defining the identify of glial cells. huckebein (hkb) was used to recognise 1-1A, 2-2T, and 1-3 lineages, Ultrabithorax (Ubx) the 1-1, Invected the 6-4 and the 7-4, Eagle and Krupple the 6- 4, sevenup, and ladybird early (lbe) the 5-6 (Bossing et al., 1996a; Broadus et al., 1995; Higashijima et al., 1996; Jagla et al., 1998; Romani et al., 1996; Schmid et al., 1999; Schmidt et al., 1997) (J. Urban, personal communication) (Fig. 5, supplementary data, and data not shown). Most markers have a very dynamic pattern during development. Some of them recognise the lineage at early stages, their expression preceding that of glide but rapidly fading away; some recognize late stages, and some can be used throughout the lineage. In addition, some markers recognise only part of the lineage. For the sake of simplicity, we describe the profile of one of them, eagle, which labels the whole 6-4 lineage throughout development. The profiles of the others are available upon request. The eagle gene has been extensively used to label the 6-4 abdominal and thoracic lineages (6-4A and T, respectively) (Akiyama-Oda et al., 1999, 2000a, b; Bernardoni et al., 1999; Dittrich et al., 1997; Ragone et al., 2001). In the Thoracic lineage, eagle is expressed in the Neuroglioblast, like glide, and does not depend on glide expression (see Fig. 6A and B), which makes it possible to use it as an independent marker. Upon division, eagle expression becomes less prominent in one daughter, the Glioblast, which moves medially and produces three glial cells, than in the other, the Neuroblast, which stays lateral and produces neurons (Bernardoni et al., 1999) (Fig. 5A and C). Only the glial component expresses glide. Position and intensity of Eagle labelling can thus be used to identify the neural and glial component of the lineage. Colocalization of glide and Eagle in 6-4T is shown in Fig. 6A. Colocalization in the glial component of the 6-4T progeny is shown in Fig. 7A. 6-4A, a pure glioblast that divides once and produces two glial cells (Fig.5A) expresses Eagle in a more transient fashion. On average, Eagle can be detected slightly later and at lower levels than in the 6-4T (Figs. 6A and 7A). Eagle and glide colocalise in glioblast 6-4A and in its progeny (Figs. 6A and 7A, and data not shown).

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

143

Fig. 3. In vivo rescue of the glial phenotype. Ventral views of stage 15 embryos labelled with anti-Repo. Anterior is to the left. (A) Wild-type (WT), (B) N7-4, (C) 2kb-glide, N7-4, (D) 4kb-glide, N7-4, (E) 6kb-glide, N7-4, (F) 9kb-glide, N7-4, (G) 6kb-glide, N7-4; 6kb-glide, (H) 9kb-glide, N7-4; 9kb-glide. Arrows, arrowheads, and concave arrowheads indicate longitudinal, peripheral glia, and intersegmental/segmental nerve glia, respectively. (T3) and (A1) indicate the third thoracic and the first abdominal segments, respectively. Scale bar, 50 ␮m.

glide starts being expressed at stage 10 at the position of the LGB. Soon after, the LGB migrates ventrally, and a second lineage becomes labelled at a dorsal position, the 1-3. By the end of stage 10, the LGB is flanked by the 7-4. The number of labelled cells in each lineage increases as development proceeds, due to cell proliferation. Other lineages start expressing glide by stage 11: first the 6-4T, then the 6-4A, the 5-6, and the abdominal-specific lineage 3-5, which is located very dorsally in the ventral cord. Another cell starts expressing glide; by its proximity to the 1-3

lineage, it may be part of it or it might belong to the 2-5 lineage. By mid- stage 11, the early lineages continue to migrate and proliferate. At early stage 12, glide is expressed in 1-1A, and soon after in 2-2T, two precursors that produce similar types of glial cells in the abdomen and in the thorax, respectively. By stage 12, all but 2-1 and 7-1 lineages, which produce glia very late (Schmid et al., 1999), have differentiated. Extensive migration and proliferation do not allow a comprehensive characterisation of glial types at later

144

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

stages, therefore 2-1 and 7-1 have not been taken into consideration. Positive, negative, and temporal glide regulation control glial specification To identify the glide cis-regulatory regions that control glial specification, we compared the profile of glide expression between wild type and rescued animals. The analysis was realised at stages 11, 12, and 13 in an RNA null background, using the glide26 allele. Stage 11 allowed the analysis of most lineages, even though proliferation and/or migration are still occurring (Fig. 6A), and stages 12 and 13 were used to follow the 1-1A and the 2-2T lineage (Fig. 7A, and data not shown). As mentioned above, transgenes showing comparable levels of glide expression were used in all rescue experiments in order to avoid differences due to dosage effects. The 2 kb contains information for 1-1A, 2-2T, and 1-3 lineages, which express glide in all segments of all embryos (Figs. 6C and 7C, and Table 1), in agreement with the observation that this transgene rescues subperineural glia (SPG), intersegmental/segmental nerve glia (ISNG, SNG), and peripheral glia (PG) (Fig. 3, and data not shown). The 4 kb contains additional information for 5-6 (100% penetrance and expressivity) as well as for 3-5, 6-4A, and 7-4 (partial rescue) (Fig. 6D, and Table 1), as confirmed by the presence of central body glia (CBG), channel glia (CG), and longitudinal glia (LG), which originate from those lineages (Fig. 3, and data not shown). The 6 and 9 kb contain some information for 6-4T and 6-4A (Fig. 6E and F, and Table 1). Indeed, glide is not expressed in the 6-4T of 2 and 4 kb lines and is only expressed in few lineages in the 6 and 9 kb lines (3% of lineages, see Table 1) (Figs. 6C–F and 7C–F). As expected, in the mutant and in the transgenes that do not rescue the 6-4T, Eagle is expressed at high levels and is often located at a more lateral position. This further indicates that these Eagle-expressing cells do not display gliaspecific behaviors. As for the 6-4A, we find Eagle but no glide expression in the mutant or in the transgenic lines at stage 11 (Fig. 6B and D). At later stages, Eagle is expressed at low levels in the progeny of the 6-4A, where it transiently colocalizes with glide in wild type embryos (Fig. 7A, and data not shown). We do not observe such colocalization in the mutant (Fig. 7B). In transgenic embryos, the 6-4A is partially rescued (Fig. 7F, and Table 1). As for the thorax, 6-4A lineages that are not rescued express Eagle at higher levels than the wild type. Again, marker intensity and position of glide expression have been used to trace lineage rescue. Fig. 4. In vivo rescue of the axonal phenotype. Ventral cord of stage 15 embryos labelled with anti-Fas II (A–F), BP102 (G–I), and mAb22c10 (J–L). Anterior to the left. (A, G, J) Wild-type (WT), (B, H, K) N7-4, (C) 2kb-glide, N7-4, (D) 4kb-glide, N7-4, (E) 6kb-glide, N7-4, (F, I, L) 9kbglide, N7-4. Asterisks and arrows indicate defects in connectives (absent or

thinner, respectively). Arrowhead in (B) shows the abnormal crossing at the midline. Arrowheads in (J, K) indicate peripheral (intersegmental and segmental) nerve roots. Notice that the extent of rescue increases with the length of the promoter. Scale bars, 25 ␮m (A–I) and 50 ␮m (J–L).

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

As a general rule, rescue penetrance and expressivity increases with promoter length and transgene dose (see Table 1). The expression profile of Repo- positive cells follows that of glide, except for the 6-4 lineage, where glide expression is too low to allow differentiation of MM-CBGs and CBGs glial cells, which normally originate from that lineage (Table 1, and data not shown). The direct correlation between length of promoter sequences and rescue of glial lineages calls for mechanisms requiring positive glide regulation. However, we observed that proper glial differentiation is also regulated by glide repression (Fig. 6C). In the 2 kb, ectopic glide expression is present at the position of three lineages that correspond to 4-1, 5-2, and 5-3, as confirmed by using lineage-specific markers (gooseberry distal, invected, fushi tarazu, and slit (Broadus et al., 1995; Schmid et al., 1999; Schmidt et al., 1997; data not shown). This results in Repo labelling (data not shown). Thus, a negative control is exerted on glide expression during development. Repression resides distal to the 2 kb, since no ectopic labelling was observed in the longer transgenes. Finally, differentiation of some glial lineages takes place earlier or later than in the wild type (compare Figs. 2, 6, and 7). For example, in the wild type, glide is detected in 1-1A only at stage 12, whereas in the 2 kb, it is already expressed at high levels at stage 11 (Figs. 6A and C and 7A). glide expression in 7-4 is already present at stage 10 in the wild type but only at stage 13 in the rescued embryos (data not shown). Also, late but not early glide (and Repo) expression was detected in the LGB of rescued embryos (data not shown). Thus, the glide promoter contains specific temporal information.

Discussion Gliogenesis and positional cues Nervous system differentiation is controlled by the coordinated activity of secreted and nuclear factors that dictate neural competence as well as positional information. This allows generation of different types of neurons and glial cells following stereotyped spatiotemporal profiles. The questions that arise are: How does positional information signal to genes that induce cell differentiation? and, How is specificity achieved so that distinct types of neurons and glia differentiate? In flies and vertebrates, neuronal differentiation depends on transcription factors of the bHLH family, also called proneural factors (Guillemot, 1999; Modolell and Campuzano, 1998). Each gene activates a particular differentiation pathway by interacting with specific partners and/or by triggering specific targets. Thus, bHLH factors integrate positional information and contribute to the specification of the neuronal type (Campos-Ortega, 1993; Chitnis, 1999; Dambly- Chaudiere and Vervoort, 1998; Guillemot, 1999;

145

Jan and Jan, 1994; Kageyama et al., 1997; Modolell and Campuzano, 1998; Skeath, 1999). Indeed, fly external sensory organs depend on the achaete scute complex, whereas chordotonal organs depend on the Atonal bHLH protein (Dambly-Chaudiere and Vervoort, 1998). Vertebrate and invertebrate glial cells also originate from specific regions of the nervous system and arise at stereotyped times during development, suggesting a requirement for precise positional cues. The present study shows that the glide promoter contains lineage-specific elements as well as quantitative and temporal elements scattered throughout several kilobases. There is no correlation between a specific regulatory element and the mode of glia differentiation. For example, the shortest fragment rescues two NGBs, the 1-1A, which arises ventrally, delaminates early, and produces CNS glia, and the 1–3, which arises very dorsally, delaminates late, and produces peripheral glia. Moreover, 1-1A glial-derived cells appear late, whereas those derived from the 1–3 are the first ones. Interestingly, distinct cisregulatory elements responding to anteroposterior cues dictate the differentiation of glial precursors along the A/P axis. Indeed, all the lineages rescued by the 2-kb transgene are located anteriorly within each neuromere (Fig. 6 and 7), while longer promoters result in the rescue of more posteriorly located precursors. Spatial regulation overlaps with temporal regulation, since 9 kb of upstream sequences allow for glide expression in all lineages, but display delayed or precocious onset/ decay in some lineages. Our data show that temporal regulation does contribute to the normal profile of glide expression and glial differentiation. The importance of glide temporal regulation has also been demonstrated in the development of the adult nervous system (Van De Bor et al., 2002). These data suggest that fly gliogenesis is a process in which several developmental pathways converge onto a single gene, glide, which is at the crossroads between neurons and glia. Indeed, due to its uniform and delayed expression in all glial lineages, it is unlikely that glide2 is critical for the induction of specific glial fates. Thus, the key for glia specification resides in the glide promoter. Gliogenesis differs from neuronal development in the sense that the glial-promoting factor does not have a role in cell specification, whereas neurogenesis depends on factors that dictate both cell differentiation and specification. Whether these differences in developmental strategies reflect the fact that glial cells are more plastic in nature, as shown by the recent observation that glial cells may constitute a source of stem cells (Anderson, 2001, and reference therein), remains to be elucidated. Quantitative regulation of glide Spatial information is contained in the 9kb transgene, even though additional quantitative regulation may require the whole region between glide and glide2. This likely accounts for the lack of viability rescue even with the

146

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

Table 1 Rescue of glide expression in the embryonic glial lineages Lineages GENOTYPE 1-1 2kb-glide 4kb-glide 6kb-glide 9kb-glide 9kb 2 doses

100 100 100 100 100

1-3 38 (7) 18 (6) 52 (7) 104 (21) 16 (4)

100 100 100 100 100

2-2 66 (7) 48 (6) 72 (7) 181 (20) 24 (4)

100 100 100 100 100

3-5 22 (5) 6 (2) 40 (7) 80 (16) 16 (4)

0 31 59 59 66

5-6 22 (5) 16 (6) 17 (5) 41 (12) 6 (3)

0 100 100 100 100

38 (7) 48 (6) 60 (7) 177 (20) 24 (4)

6-4T

6-4A

0 0 2 3 6

0 6 13 26 25

30 (7) 32 (6) 46 (9) 98 (19) 18 (4)

7-4 38 (7) 15 (6) 24 (9) 61 (19) 8 (4)

0 12 12 60 60

LGB 66 (7) 50 (6) 26 (4) 126 (15) 25 (4)

0 ND ND 100 ND

6 (7)

*

Columns indicate the lineages, rows the different constructs. Three values are given for each lineage: the percentage of rescue relative to each lineage (bold), the number of lineages (right top), and the number of embryos (right bottom) analysed. Asterisk indicates that the percentage has been determined by following Repo expression, starting from stage 13. We were not able to follow the 2-1, 2-5, and 7-1 glial progeny (see second paragraph of Results). Stages considered: 11 and 12.

longest construct. Misexpression data suggested that high Glide levels are required for glial differentiation. Mesodermal Glide leads to ectopic glia in that layer, however, the number and type of mesodermal cells that adopt the glial fate strictly depending on the levels of expression (Akiyama-Oda et al., 1998; Bernardoni et al., 1998). These data might also indicate that cells outside the nervous system have different potentials to produce glia. The present study, however, clearly shows that high Glide levels are necessary to produce glia. First, double dose of the same construct shows better rescue than single dose. Second, transgenic lines carrying the same construct display different rescue potentials depending on the position of the insert (data not shown). Dosage effect, however, cannot account for all rescue potential. Indeed, while large constructs show better rescue (higher penetrance and expressivity) when expressed in double dose as compared to single dose (two vs. one insert; homo- vs. heterozygous conditions for the same insert), this is not true for the shortest, 2kb transgene. Thus, qualitative information assigns lineage identity and mostly resides in the proximal region, whereas quantitative information strengthens the gliogenic potential and is scattered throughout many kilobases. An important need for qualitative regulation is also stressed by the fact that the glial lineages do not show major differences in the levels of glide expression and yet they show different requirements with respect to the glide promoter. High levels of glide expression likely depend on the presence of several GBSs in the region between glide and glide2 (Kammerer and Giangrande, 2001; Miller et al., 1998) (present study). They also depend on lineage-specific partners of glide, as shown for the Prospero transcription factor in the 6-4 and 7-4 lineages (Akiyama-Oda et al., 2000b; Freeman and Doe, 2001; Ragone et al., 2001). We propose that autoregulation and cofactor-mediated maintenance of Glide play an important role in gliogenesis. Indeed, high and protracted expression may be necessary for late steps, such as glial proliferation, as previously suggested by the phenotype of glide hypomorphic mutations (Hosoya et al., 1995; Vincent et al., 1996). The use of glide (rather than reporter gene)-containing

constructs has allowed us to carry out a comprehensive analysis of the rescued glial phenotype and to perform a functional assay (axonal phenotype). Moreover, the spatiotemporal profile of transgene expression has been assessed. This is specifically important for genes that, as glide, are transiently expressed (Bernardoni et al., 1997; Hosoya et al., 1995; Jones et al., 1995). Finally, possible involvement of coding and untranslated sequences can be taken into account, which is important in the case of glide, a gene that autoregulates and displays asymmetric RNA distribution (Akiyama-Oda et al., 1999; Bernardoni et al., 1999; Miller et al., 1998; Ragone et al., 2001). All these conditions have allowed us to define the spatial, temporal, and quantitative requirements of glide. Neural stem cell specification requires both negative and positive regulation of glide expression glide overexpression during embryogenesis produces glial cells in all the tissues in which it has been tested (AkiyamaOda et al., 1998; Bernardoni et al., 1998; Hosoya et al., 1995; Jones et al., 1995). This indicates that glide is necessary and sufficient to induce the glial fate. It also suggests that gliogenesis requires glide- positive regulation. Such mechanisms are indeed at work, since adding back wild- type copies of glide into a mutant background restores both glide expression and glial differentiation. Nevertheless, glide also needs to be repressed in nongliogenic lineages, as seen in the 2-kb transgene. This first promoter analysis opens the perspective that glial differentiation occurs through negative regulation of the glialpromoting activity. This finding also calls for a provocative view of a default glial state in some lineages of the nervous system. It is possible that a situation similar to the one observed in fly glia also takes place in vertebrate gliogenesis. Indeed, vertebrate proneural genes, which code for transcriptional activators, seem to inhibit the glial fate (Hojo et al., 2000). This suggests that their role is to activate a repressor, which in turn shuts off the expression of yet unknown glial-promoting factors. The identification of such factor(s) will shed light on the mechanisms that drive glial differentiation in vertebrates. The present study has allowed us to define both mode of

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

147

Fig. 5. Gliogenic lineages and glide expression. Schematic drawing of glide (A) and Repo (B) expression in the gliogenic lineages. Anterior is to the top. Vertical and horizontal lines indicate midline and segment border, respectively. T3 and A1 indicate third thoracic and first abdominal segments, respectively. Glial subsets were identified by using lineage- specific markers [(Ubx) Ultrabithorax, (inv) Invected, (kr) Krupple, (lbe) ladybird early, huckebein-lacZ (hkb-lacZ) and sevenup-lacZ] and labelled with different colours. For the sake of simplicity, only Eagle and huckebein labelling are documented in this study. All the data concerning the other markers are available upon request. LGB and 3-5 were identified by the position. Red dots indicate nonidentified lineages. (C) Identification of the 6-4T glial-derived cells at stage 13, modified from Ragone et al. (2001). In green, the Eagle lineage-specific marker; in red, the Repo glial-specific marker. The 6-4T lineage has been highlighted in blue. 7-3, 2-4, and 3-3 indicate the other lineages identified by Eagle. T2 and T3 indicate second and third thoracic segments, respectively. Lateral cells express Eagle at high levels and constitute the neural component (arrow), and medial cells express Eagle at low levels and constitute the glial component of the lineage (arrowheads) (two MM-CBGs and one M-CBG per hemisegment). Bar, 25 ␮m.

glide regulation and glide requirements in the different types of glial cells. It has shown that Glide constitutes a naive glial-promoting factor that acts as a switch in response to positive and negative pathways. We have now set the bases to identify such pathways: in vitro and in vivo studies are in progress to identify the factors that integrate differentiation and specification.

Acknowledgments We thank S. Benzer, G. Flores, A. Tomlinson, A. Travers, and J. Urban for flies and antibodies. We also wish

to thank Rachel Walther for excellent technical help, and J.L. Vonesch and D. Hentsch for their precious help with the confocal microscope and imaging. Confocal microscopy was developed with the aid of a subvention from the French MESR (95.V.0015). Work in the lab was supported by the Institut National de la Sante´ et de la Recherche Me´ dicale, the Centre National de la Recherche Scientifique, the Hoˆ pital Universitaire de Strasbourg, the Human Frontier Science Program, the Association pour la Recherche contre le Cancer, and by EEC grant (Contract QLG3-CT-200001224). V.V.D.B. was supported by MRT, ARC fellowships, and by the “Socie´ te´ de secours des amis de la science”; M.K. by Ligue contre le Cancer, BDI (CNRS), and

Fig. 6. Rescue of glide expression at stage 11. Ventral views of late stage 11 embryos simultaneously labelled with the glide-specific riboprobe (red) and either Eagle antibody (A, B, D–F) (green) or huckebein-specific riboprobe (C) (green). Eagle-expressing lineages are, from anterior to posterior, 2-4, 3-3, 6-4 (thoracic or abdominal), and 7-3. Some of them are indicated as landmarks. In (D–F), 2-4 and 3-3 are very closely located and cannot be distinguished at this magnification. (A) Wild-type (WT), (B) glide26. (C–F) Rescue obtained by introducing different constructs into the glide26 background: (C) 2kb-glide, (D) 4kb-glide, (E) 6kb-glide, (F) 9kb-glide. Anterior is to the left; white bars indicate the position of the midline. All panels show third thoracic (T3) and first abdominal segments (A1). (A) At late stage 11, glide is expressed in the following lineages: 1-3, LGB, 5-6, 6-4T, 7-4, 3-5A, and 6-4A. Nomenclature of neural precursors is as in Bossing et al. (1996b), Schmid et al. (1999) and Schmidt et al. (1997). (B) No glide expression is detectable in glide26 embryos. (C) Note the presence of ectopic labelling in the abdominal segment, close to the midline (arrows). Also note that huckebein is not yet expressed in the 1-3 lineage. Arrowheads and short arrows in (E) and (F) indicate LGB labelling and the position of 6-4A, respectively. Note that 1-1A labelling is already present in the rescued lines (C–F). Bar, 25 ␮m.

Fig. 7. Rescue of glide expression at stage 12. Embryos and labelling as in Fig. 6. (A) By stage 12, glide is normally expressed in the following lineages: 1-3, LGB, 2-5, 5-6, 6-4T, 7-4, 3-5A, and 6-4A. In (A) and (B), 6-4T indicates the whole lineage, as shown by the dotted area: note that, in (A), two out of the three Eagle-expressing cells also express glide. These two medially migrating cells constitute the glial component. The lateral cell (arrowhead), which only expresses Eagle, constitutes the neural component. glide and, faintly, Eagle are also expressed in the two daughters of the 6-4A pure glioblast. In the mutant (B), notice the absence of glide expression in 6-T and 6-4A. Also notice that 6-4A Eagle labelling is stronger than in wild type and that 6-4T cells migrate less than in wild type. These features support the idea that 6-4- derived cells have lost the glial identity. (C) Notice that at least one cell per hemisegment, placed close to the midline, ectopically expresses glide in the abdominal segments in the 2kb (arrow). 2-2T labelling is already present in the rescued lines (C–F). (F) glide expression is rescued in one 6-4A (top), but very faintly in the other (bottom), which keeps high levels of Eagle expression (see regions defined by dotted lines). Bar, 25 ␮m (A–E) and 18 ␮m (F).

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

European Leucodystrophy Association; A.S. by the Ernst Schering Research Foundation; S.S. G.R., and R.B. were supported by EEC, FRM, and Ligue contre le Cancer; and A.G. and N.R. were supported by EEC and INSERM. References Akiyama, Y., Hosoya, T., Poole, A.M., Hotta, Y., 1996. The gem-motif: a novel DNA-binding motif conserved in Drosophila and mammals. Proc. Natl. Acad. Sci. USA 93, 14912–14916. Akiyama-Oda, Y., Hosoya, T., Hotta, Y., 1998. Alteration of cell fate by ectopic expression of Drosophila glial cells missing in non-neural cells. Dev. Genes Evol. 208, 578 –585. Akiyama-Oda, Y., Hosoya, T., Hotta, Y., 1999. Asymmetric cell division of thoracic neuroblast 6-4 to bifurcate glial and neuronal lineage in Drosophila. Development 126, 1967–1974. Akiyama-Oda, Y., Hotta, Y., Tsukita, S., Oda, H., 2000a. Distinct mechanisms triggering glial differentiation in Drosophila thoracic and abdominal neuroblasts 6-4. Dev. Biol. 222, 429 – 439. Akiyama-Oda, Y., Hotta, Y., Tsukita, S., Oda, H., 2000b. Mechanism of glia-neuron cell-fate switch in the Drosophila thoracic neuroblast 6-4 lineage. Development 127, 3513–3522. Anderson, D. J., 2001. Stem cells and pattern formation in the nervous system: the possible versus the actual. Neuron 30, 19 –35. Bernardoni, R., Kammerer, M., Vonesch, J.L., Giangrande, A., 1999. Gliogenesis depends on glide/gcm through asymmetric division of neuroglioblasts. Dev. Biol. 216, 265–275. Bernardoni, R., Miller, A.A., Giangrande, A., 1998. Glial differentiation does not require a neural ground state. Development 125, 3189 –3200. Bernardoni, R., Vivancos, B., Giangrande, A., 1997. glide/gcm is expressed and required in the scavenger cell lineage. Dev. Biol. 191, 118 –130. Booth, G.E., Kinrade, E.F., Hidalgo, A., 2000. Glia maintain follower neuron survival during Drosophila CNS development. Development 127, 237–244. Bossing, T., Technau, G.M., Doe, C.Q., 1996a. huckebein is required for glial development and axon pathfinding in the neuroblast 1-1 and neuroblast 2-2 lineages in the Drosophila central nervous system. Mech. Dev. 55, 53– 64. Bossing, T., Udolph, G., Doe, C.Q., Technau, G.M., 1996b. The embryonic central nervous system lineages of Drosophila melanogaster. I. Neuroblast lineages derived from the ventral half of the neuroectoderm. Dev. Biol. 179, 41– 64. Broadus, J., Skeath, J.B., Spana, E.P., Bossing, T., Technau, G., Doe, C.Q., 1995. New neuroblast markers and the origin of the aCC/pCC neurons in the Drosophila central nervous system. Mech. Dev. 53, 393– 402. Campbell, G., Goring, H., Lin, T., Spana, E., Andersson, S., Doe, C.Q., Tomlinson, A., 1994. RK2, a glial-specific homeodomain protein required for embryonic nerve cord condensation and viability in Drosophila. Development 120, 2957–2966. Campos-Ortega, J.A., 1993. Early neurogenesis in Drosophila melanogaster, in: Bate, M., Martinez Arias, A. (Eds.), The Development of Drosophila melanogaster, Vol. II. Cold Spring Harbor Laboratory Press, New York, pp. 1091–1129. Chitnis, A.B., 1999. Control of neurogenesis: lessons from frogs, fish and flies. Curr. Opin. Neurobiol. 9, 18 –25. Dambly-Chaudiere, C., Vervoort, M., 1998. The bHLH genes in neural development. Int. J. Dev. Biol. 42, 269 –273. Dittrich, R., Bossing, T., Gould, A.P., Technau, G.M., Urban, J., 1997. The differentiation of the serotonergic neurons in the Drosophila ventral nerve cord depends on the combined function of the zinc finger proteins Eagle and Huckebein. Development 124, 2515–2525. Freeman, M.R., Doe, C.Q., 2001. Asymmetric Prospero localization is required to generate mixed neuronal/glial lineages in the Drosophila CNS. Development 128, 4103– 4112.

149

Granderath, S., Klambt, C., 1999. Glia development in the embryonic CNS of Drosophila. Curr. Opin. Neurobiol. 9, 531–536. Grenningloh, G., Rehm, E.J., Goodman, C.S., 1991. Genetic analysis of growth cone guidance in Drosophila: fasciclin II functions as a neuronal recognition molecule. Cell 67, 45–57. Guillemot, F., 1999. Vertebrate bHLH genes and the determination of neuronal fates. Exp. Cell Res. 253, 357–364. Halter, D.A., Urban, J., Rickert, C., Ner, S.S., Ito, K., Travers, A. A., Technau, G. M., 1995. The homeobox gene repo is required for the differentiation and maintenance of glia function in the embryonic nervous system of Drosophila melanogaster. Development 121, 317– 332. Heinemeyer, T., Chen, X., Karas, H., Kel, A.E., Kel, O.V., Liebich, I., Meinhardt, T., Reuter, I., Schacherer, F., Wingender, E., 1999. Expanding the TRANSFAC database towards an expert system of regulatory molecular mechanisms. Nucleic Acids Res. 27, 318 –322. Hidalgo, A., Booth, G.E., 2000. Glia dictate pioneer axon trajectories in the Drosophila embryonic CNS. Development 127, 393– 402. Hidalgo, A., Urban, J., Brand, A.H., 1995. Targeted ablation of glia disrupts axon tract formation in the Drosophila CNS. Development 121, 3703–3712. Higashijima, S., Shishido, E., Matsuzaki, M., Saigo, K., 1996. eagle, a member of the steroid receptor gene superfamily, is expressed in a subset of neuroblasts and regulates the fate of their putative progeny in the Drosophila CNS. Development 122, 527–536. Hojo, M., Ohtsuka, T., Hashimoto, N., Gradwohl, G., Guillemot, F., Kageyama, R., 2000. Glial cell fate specification modulated by the bHLH gene Hes5 in mouse retina. Development 127, 2515–2522. Hosoya, T., Takizawa, K., Nitta, K., Hotta, Y., 1995. glial cells missing: a binary switch between neuronal and glial determination in Drosophila. Cell 82, 1025–1036. Hummel, T., Krukkert, K., Roos, J., Davis, G., Klambt, C., 2000. Drosophila Futsch/22C10 is a MAP1B-like protein required for dendritic and axonal development. Neuron 26, 357–370. Hummel, T., Schimmelpfeng, K., Klambt, C., 1999. Commissure formation in the embryonic CNS of Drosophila. Dev. Biol. 209, 381–398. Isshiki, T., Pearson, B., Holbrook, S., Doe, C.Q., 2001. Drosophila neuroblasts sequentially express transcription factors which specify the temporal identity of their neuronal progeny. Cell 106, 511–521. Ito, K., Urban, J., Techanu, G.M., 1995. Distribution, classification, and development of Drosophila glial cells in the late embryonic and early larval ventral nerve cord. Roux’s Arch. Dev. Biol. 204, 284 –307. Jacobs, J.R., 2000. The midline glia of Drosophila: a molecular genetic model for the developmental functions of glia. Prog. Neurobiol. 62, 475–508. Jagla, T., Bellard, F., Lutz, Y., Dretzen, G., Bellard, M., Jagla, K., 1998. ladybird determines cell fate decisions during diversification of Drosophila somatic muscles. Development 125, 3699 –3708. Jan, Y.N., Jan, L.Y., 1994. Neuronal cell fate specification in Drosophila. Curr. Opin. Neurobiol. 4, 8 –13. Jones, B.W., Fetter, R.D., Tear, G., Goodman, C.S., 1995. glial cells missing: a genetic switch that controls glial versus neuronal fate. Cell 82, 1013–1023. Kageyama, R., Ishibashi, M., Takebayashi, K., Tomita, K., 1997. bHLH transcription factors and mammalian neuronal differentiation. Int. J. Biochem. Cell Biol. 29, 1389 –1399. Kammerer, M., Giangrande, A., 2001. Glide2, a second glial promoting factor in Drosophila melanogaster. EMBO J. 20, 4664 – 4673. Kinrade, E.F., Brates, T., Tear, G., Hidalgo, A., 2001. Roundabout signalling, cellcontact and trophic support confine longitudinal glia and axons in the Drosophila CNS. Development 128, 207–216. Miller, A.A., Bernardoni, R., Giangrande, A., 1998. Positive autoregulation of the glial promoting factor glide/gcm. EMBO J. 17, 6316 – 6326. Modolell, J., Campuzano, S., 1998. The achaete-scute complex as an integrating device. Int. J. Dev. Biol. 42, 275–282.

150

G. Ragone et al. / Developmental Biology 255 (2003) 138 –150

Patel, N.H., Martin-Blanco, E., Coleman, K.G., Poole, S.J., Ellis, M.C., Kornberg, T. B., Goodman, C.S., 1989. Expression of engrailed proteins in arthropods, annelids, and chordates. Cell 58, 955–968. Ragone, G., Bernardoni, R., Giangrande, A., 2001. A novel mode of asymmetric division identifies the fly neuroglioblast 6-4t. Dev. Biol. 235, 74 – 85. Romani, S., Jimenez, F., Hoch, M., Patel, N.H., Taubert, H., Jackle, H., 1996. Kruppel, a Drosophila segmentation gene, participates in the specification of neurons and glial cells. Mech. Dev. 60, 95–107. Roos, J., Hummel, T., Ng, N., Klambt, C., Davis, G.W., 2000. Drosophila Futsch regulates synaptic microtubule organization and is necessary for synaptic growth. Neuron 26, 371–382. Schmid, A., Chiba, A., Doe, C.Q., 1999. Clonal analysis of Drosophila embryonic neuroblasts: neural cell types, axon projections and muscle targets. Development 126, 4653– 4689. Schmidt, H., Rickert, C., Bossing, T., Vef, O., Urban, J., Technau, G.M., 1997. The embryonic central nervous system lineages of Drosophila melanogaster. II. Neuroblast lineages derived from the dorsal part of the neuroectoderm. Dev. Biol. 189, 186 –204. Schreiber, J., Sock, E., Wegner, M., 1997. The regulator of early gliogenesis glial cells missing is a transcription factor with a novel type of DNA-binding domain. Proc. Natl. Acad. Sci. USA 94, 4739 – 4744.

Skeath, J.B., 1999. At the nexus between pattern formation and cell-type specification: the generation of individual neuroblast fates in the Drosophila embryonic central nervous system. Bioessays 21, 922–931. Skeath, J.B., Doe, C.Q., 1996. The achaete-scute complex proneural genes contribute to neural precursor specification in the Drosophila CNS. Curr. Biol. 6, 1146 –1152. Takizawa, K., Hotta, Y., 2001. Pathfinding analysis in a glia-less gcm mutant in Drosophila. Dev. Genes Evol. 211, 30 –36. Udolph, G., Rath, P., Chia, W., 2001. A requirement for Notch in the genesis of a subset of glial cells in the Drosophila embryonic central nervous system which arise through asymmetric divisions. Development 128, 1457–1466. Van De Bor, V., Giangrande, A., 2002. glide/gcm: at the crossroads between neurons and glia. Curr. Opin. Genet. Dev. 12, 465– 472. Van De Bor, V., Heitzler, P., Leger, S., Plessy, C., Giangrande, A., 2002. Precocious expression of the glide/gcm glial-promoting factor in Drosophila induces neurogenesis. Genetics 160, 1095–1106. Vincent, S., Vonesch, J.L., Giangrande, A., 1996. Glide directs glial fate commitment and cell fate switch between neurones and glia. Development 122, 131–139. Xiong, W.C., Okano, H., Patel, N.H., Blendy, J.A., Montell, C., 1994. repo encodes a glial-specific homeo domain protein required in the Drosophila nervous system. Genes Dev. 8, 981–994.