Glial cells constitute the second component of the nervous system and are important during neuronal development. In this paper we describe a gene, glial cell deficient, (glide), that is necessary for glial cell fate commitment in Drosophila melanogaster. Mutations at the glide locus prevent glial cell determination in the embryonic central and peripheral nervous system. Moreover, we show that the absence of glial cells is the consequence of a cell fate switch from glia to neurones. This suggests the existence of a multipotent precursor cells in the nervous system. glide mutants also display defects in axonal navigation, which confirms and extends previous results indicating a role for glial cells in these processes.

Glial cells display different features depending on their origin, morphology, position and gene expression. Although these cells play fundamental roles during neuronal development and functioning, as well as during axonal regeneration, little is known about the molecular mechanisms leading to their determination. The engrailed gene is required for the early neuronal to glial fate transition taking place in the grasshopper median neuroblast (MNB) lineage (Condron et al., 1994), whereas activation of protein kinase A is responsible for the glial to neuronal fate transition occurring in the late phases of the lineage (Condron and Zinn, 1995). Cell-cell communication involving the activity of neurotrophic factors is required for vertebrate glial cell proliferation and glial precursor cell survival (Marchionni et al., 1993; Jessen et al., 1994; Shah et al., 1994). Despite the progresses achieved, we are still missing most of the molecular actors involved in glial differentiation. Moreover, the glial promoting factors identified so far concern subgroups of glial cells or have been obtained in cell culture systems.

In recent years, the availability of genetic tools and specific markers have made it possible to explore in vivo the cellular and molecular basis of gliogenesis in Drosophila melanogaster. A number of genes have been found to be expressed in subsets of glial cells; these genes are required for some phases of glial differentiation but not for the determination of the glial fate. For example, the two products of the pointed locus are expressed in longitudinal and midline glia (Klämbt, 1993) and are required for the differentiation of these cells (Klaes et al., 1994). prospero is expressed in longitudinal glial cells; when the gene is mutated, these cells are spatially disorganised (Doe et al., 1991). More recently, a homeobox-containing gene, reversed polarity (repo), located on the third chromosome and expressed in all but midline glia, has been found to be necessary for the late steps of glial differentiation and for the maintenance of the glial fate (Campbell et al., 1994; Xiong et al., 1994; Halter et al., 1995).

In this paper we describe a new gene, glide (glial cell deficient), involved in the early phases of glial cell differentiation. The pattern of expression of β-galactosidase ?β-gal) in the original glial-specific enhancer trap line, rA87, is identical to that of the reproduct: all but midline and perineural cells are labelled from early stages of nervous system development. The mutation is embryonic lethal and causes a lack of repo expression throughout development, indicating that glial cells do not differentiate. All the glial cell types that express β-gal are affected by the mutation, suggesting that glide is a general glial promoting factor. In addition, most of the cells that would have taken the glial fate are transformed into neurones, suggesting that the glide product induces a choice between the glial and neuronal fate. Other phenotypes associated with this mutant are axonal misrouting and growth.

Stocks

The rA87 enhancer trap line that carries a transposon on the second chromosome was kindly provided by V. Auld and C. Goodman, who had mapped the insert to the position 30B. The wild-type stock was Sevelen. The ‘blue balancer’ CyO twi-lacZ carries a transposon containing the twist promoter sequences fused with the E. coli β-galactosidase coding sequences (Thisse et al., 1991). The strains, Df 30A-C, Df 166, DB1, DA2 and N7-4 were balanced over CyO (kindly provided by D. Kalderon). N7-4 was induced with DEB, DA2 and DB1 with EMS (for the origin of all mutations at 30A-C see Lane and Kalderon, 1993). Strains for mutagenesis were: w1118; Sb Δ2-3/TM6; CyO/cn bw sp; kar ry; vgU/CyO. Embryonic lethality was assessed using the strain In(2L) Gla Bc/SM1 Cy.

P mutagenesis

Before inducing excisions, the third chromosome was isogenised by a single cross of a rA87 male with CyO/M(2); kar ry females. rA87; kar ry males were then crossed with w1118/+; CyO/+; Sb Δ2-3/TM6 females. Excisions were recovered on the basis of the eye phenotype (kar ry): 100 crosses were established, of which 98 survived. For all the lines, viability and fertility were assessed in homozygous conditions. 14 excisions were homozygous lethal. Lethal lines were then crossed with deficiencies and with complementation groups in the insert region, N7-4, DA2 and DB1, to assess whether lethality was associated with the excisions. The results of these crosses showed that in two cases only N7-4 did not complement the lethality, which strongly suggests that N7-4 is the only locus affected by the mutation. The formal confirmation of this result will require the saturation of the region for lethal mutations. In the case of the other lethal excisions, lethality did not map to the 30A-C region, thus it was not associated with the excision. The lines carried a ‘blue balancer’, CyO twi-lacZ, which led to mesodermal expression of the β-gal. The embryos homozygous for the excision were thus identified as being devoid of mesodermal β-gal labelling. The two lines in which the lethality was associated with the excision were also crossed with a strain carrying the Bc larval marker to eliminate the possibility that lethality occurred at postembryonic stages.

Immunohystochemistry

For immunohystochemistry of wild-type and mutant embryos, overnight or staged embryos were collected, dechorionated in bleach, rinsed in water and fixed in 50% heptane/50% PEM-formaldehyde for 15-30 minutes. After devitellinisation in methanol/heptane for 1 minute, embryos were treated with methanol and then with PBS-0.3% Triton X-100. Devitellinised embryos were incubated with blocking solution for 30 minutes at room temperature and then with primary antibody for 1 hour at room temperature, or overnight at 4°C. The blocking solution contained PBS, 5% NGS (Vector), 0.3% Triton X-100, 0.01% sodium azide. Primary and secondary antibodies were in the same solution. All primaries were preabsorbed with embryos overnight at 4°C at 1/10-1:/50 in blocking solution. The following primary antibodies were used in order to characterise the rA87 line and the glide mutants: rabbit anti-β-gal at 1:4000 dilution (Cappel); mouse anti-β-gal at 1:2000 (Promega); the rat antibody produced against the repo product, RK2, at 1:1000 (kindly provided by A. Tomlinson); rabbit anti-HRP at 1:4000 (USB); mouse anti-elav at 1:100 (kindly provided by G. Rubin); mouse anti-fas II at 1:5 (generous gift from J. Urban); mouse mAb 22c10 at 1:1 (generous gift from S. Benzer). After primary incubation, embryos were washed in PBS three times for 10 minutes each, at room temperature and then incubated with secondary antibodies which were conjugated with FITC or with Cy3 (Jackson) and used at concentrations varying between 1:400 to 1:600. Secondary incubation was for 2 hours at room temperature. Embryos were then washed as above, mounted in Vectashield and analysed by conventional (Zeiss Axiophot) or confocal microscopy (Leica DMRE).

rA87, a glial specific enhancer trap line

rA87 (kindly provided by V. Auld and C. Goodman) is an enhancer trap line in which β-gal is expressed in glial cells (Giangrande et al., 1993; Klaes et al., 1994; Halter et al., 1995). Both glial cells in the adult and in the embryonic peripheral and central nervous system (PNS, CNS) express β-gal (Fig. 1; data not shown and Giangrande et al., 1993). Different types of glial cells have been identified in the embryonic CNS (Wigglesworth, 1959; Meyer et al., 1987; Jacobs and Goodman, 1989; Klämbt and Goodman, 1991; Ito et al., 1995). For the sake of simplicity they have been classified into three categories according to their position and morphology (Ito et al., 1995): surface-associated-glia, cortex-associated-glia and neuropile-associated-glia. Surface-associated glia include the most superficial type of glial cells: perineural and subperineural glia; neuropile-associated glia include glial cells at the nerve roots, glial cells along the longitudinal fibres and midline glia. Peripheral glial cells (for examples, see Murray et al., 1984; Fredieu and Mahowald, 1989; Klämbt and Goodman, 1991; Giangrande et al., 1993; Giangrande, 1994; Choi and Benzer, 1994) wrap the axons of motor and sensory fibres. All types of glial cells so far described are labelled in rA87 with the exception of midline and perineural glia. In double labelling experiments with anti-β-gal and anti-repo, the two markers colocalise perfectly (Fig. 1).

Fig. 1.

β-gal in the rA87 line labels embryonic glial cells. In this and in the next figures, anterior is to the left, dorsal to the top. Ventral view of a stage 15-16 rA87 embryo labelled with RK2, an antibody that recognises the repo product (Campbell et al., 1994) revealed with Cy3-conjugated secondary (B,E) and with anti-βgal, revealed with a FITC-conjugated secondary (A,D). (C,F) Double exposures of the views shown in A,B and D,E respectively. A,B,C are taken at the level of longitudinal (lg) and peripheral glia (pg). D,E,F are taken at a more superficial level to show channel (cg), subperineural (sg) and exit (eg) glia. (es) indicates the β-gal-expressing stripe of ectodermal cells. Perfect colocalisation can be observed in C and F (see the orange nuclei). Bar 25 μm.

Fig. 1.

β-gal in the rA87 line labels embryonic glial cells. In this and in the next figures, anterior is to the left, dorsal to the top. Ventral view of a stage 15-16 rA87 embryo labelled with RK2, an antibody that recognises the repo product (Campbell et al., 1994) revealed with Cy3-conjugated secondary (B,E) and with anti-βgal, revealed with a FITC-conjugated secondary (A,D). (C,F) Double exposures of the views shown in A,B and D,E respectively. A,B,C are taken at the level of longitudinal (lg) and peripheral glia (pg). D,E,F are taken at a more superficial level to show channel (cg), subperineural (sg) and exit (eg) glia. (es) indicates the β-gal-expressing stripe of ectodermal cells. Perfect colocalisation can be observed in C and F (see the orange nuclei). Bar 25 μm.

β-gal expression appears at very early stages of glial development (Fig. 2 and Giangrande et al., 1993), concomitantly with the expression of repo, which is one of the earliest glial markers (Campbell et al., 1994; Xiong et al., 1994; Halter et al., 1995). Expression is first detectable at stage 11, in the longitudinal glioblast. During germ band contraction, the labelled cells increase in number and start migrating according to the pathways already described (Jacobs et al., 1989; Klämbt et al., 1991; Ito et al., 1995). Throughout development, the rA87 glial labelling seems identical to that described for repo, even though at the earliest stages some cells seem to be labelled only with anti-β-gal. In addition to glial cells, a stripe of ectodermal cells (Fig. 1) located just posterior to the stripe of engrailed-expressing cells (data not shown) is also labelled from stage 12. Other scattered β-gal-expressing cells are present in the dorsal region, at the border with the amnioserosa and between the epidermis and the central nervous system. The identity of these cells has not been assessed.

Fig. 2.

rA87 labelling profile at early stages of embryonic development. Embryos stained with anti-β-gal (A,C) and with the RK2 glial marker (B,D). (A,B) Stage 11 embryo, dorsal view: the arrows indicate the longitudinal glioblasts.

(C,D) Mid stage 12 embryo, lateral view. The expression pattern is very similar, even though the relative intensity of the labelling is not identical for the two markers and some nuclei seem labelled only with anti-β-gal (arrowheads). Bars, 50 μm (A,B); 25 μm (C,D).

Fig. 2.

rA87 labelling profile at early stages of embryonic development. Embryos stained with anti-β-gal (A,C) and with the RK2 glial marker (B,D). (A,B) Stage 11 embryo, dorsal view: the arrows indicate the longitudinal glioblasts.

(C,D) Mid stage 12 embryo, lateral view. The expression pattern is very similar, even though the relative intensity of the labelling is not identical for the two markers and some nuclei seem labelled only with anti-β-gal (arrowheads). Bars, 50 μm (A,B); 25 μm (C,D).

Glial cells are absent in flies that are mutant at the site of the insertion

We have performed a P element mutagenesis on the rA87 line in order to disrupt the activity of the gene adjacent to the insert. The original line carries an abnormal and variable number of ectopic sensory organs in the region of the twin sensillum on the wing margin (TSM) when homozygous for the insert (data not shown). This phenotype is increased when rA87 is in trans with deficiencies in the chromosome IIL that map in the region of the insert. The phenotype is associated with the N7-4 complementation group (see Materials and methods). Two lethal excisions that do not complement N7-4 lethality have been recovered after mutagenesis and analysed with several markers. These two excisions complement the lethality of DB1 and DA2, the complementation groups flanking N7-4 on the right (DB1) and on the left (DA2). Therefore, the lethality observed in the two lines is associated with the excision and the locus affected by the excision is N7-4. The most intriguing result is that in embryos from the two lethal lines and from N7-4 the number of repopositive cells is drastically reduced when the mutation is in homozygous conditions (Fig. 3). The average number of repopositive nuclei in a wild-type embryo is around 27-29 per hemineuromere of the ventral cord (Halter et al., 1995). In our mutants, the total number throughout the entire ventral cord varies from 0 to 24. Similar results were obtained using another marker, the prospero antibody, which recognises longitudinal glial cells (data not shown). Because of the phenotype, we have named this gene glide, which stands for glial cell deficient. All types of glial cells expressing the β-gal in the enhancer trap line are affected in the mutants. The few repo-positive cells are often organised in clusters containing variable numbers of cells, with the average being between three and four (data not shown). Phenotype strength in the two excisions is not identical: glide34 is slightly weaker than glide26. Both excision mutants are weaker compared to glideN7-4 (Fig. 3 and data not shown). Although gliderA87 displays a wing mutant phenotype, homozygous flies are perfectly viable and glial cells do not seem to be affected (data not shown).

Fig. 3.

Embryonic phenotype of glide. Stage 15-16 embryos, ventral views: wild type (A), glide26/glide26(B), glideN7-4/glideN7-4(C). A shows a double labelling with anti-HRP (green) and RK2 (orange). B shows the RK2 labelling in a mutant embryo: only very few repopositive nuclei (arrowheads) can be observed compared to the wild type. The repo-positive nucleus (asterisk) present in all the segments corresponds to the lateral bipolar dendritic cell (Halter et al., 1995).

(C) A glideN7-4 mutant embryo labelled with RK2. In this stronger allele almost no glial labelling is present. Bar 50 μm.

Fig. 3.

Embryonic phenotype of glide. Stage 15-16 embryos, ventral views: wild type (A), glide26/glide26(B), glideN7-4/glideN7-4(C). A shows a double labelling with anti-HRP (green) and RK2 (orange). B shows the RK2 labelling in a mutant embryo: only very few repopositive nuclei (arrowheads) can be observed compared to the wild type. The repo-positive nucleus (asterisk) present in all the segments corresponds to the lateral bipolar dendritic cell (Halter et al., 1995).

(C) A glideN7-4 mutant embryo labelled with RK2. In this stronger allele almost no glial labelling is present. Bar 50 μm.

Glial development is affected at early stages

The defect observed at the end of embryonic development could be due to different mechanisms: glial cells may have differentiated and then degenerated at some stages during embryogenesis, or they may have failed to differentiate. To distinguish between these two possibilities, we have analysed the mutants at different stages with respect to glial cell organisation and found that the number of repo labelled cells is drastically reduced throughout development. The defect is already present at the first stage at which repo labelling becomes detectable: compared to the wild type, only a few labelled cells are present in glide26 and glide34 at stage 12 (Fig. 4) and 11 (data not shown). As already seen in the fully differentiated embryo, the glideN7-4 mutation displays the strongest phenotype (data not shown).

Fig. 4.

Glial labelling in glide embryos at early stages of development. Confocal images of CyO/CyO early stage 12 (A) and glide34/glide34 mid stage 12 (B) embryo labelled with RK2. These lateral views show projections of optical sections. The projection mode consists of superimposing several optical sections which produces an image similar to that obtained by conventional microscopy, with the advantage that all planes are in focus. Note the low number of repo-labelled nuclei in the mutant brain and ventral cord. Bar 62.5 μm.

Fig. 4.

Glial labelling in glide embryos at early stages of development. Confocal images of CyO/CyO early stage 12 (A) and glide34/glide34 mid stage 12 (B) embryo labelled with RK2. These lateral views show projections of optical sections. The projection mode consists of superimposing several optical sections which produces an image similar to that obtained by conventional microscopy, with the advantage that all planes are in focus. Note the low number of repo-labelled nuclei in the mutant brain and ventral cord. Bar 62.5 μm.

Role of glide in glial differentiation

Lack of glial labelling indicates the absence of glial cells. To explore further the mutant phenotype and determine the fate of the cells that would have become glia, we have taken advantage of the finding that one of the excision mutants, glide34, still expresses the β-gal, displaying in heterozygous conditions a labelling profile identical to that found in gliderA87 (Fig. 5A). In glide34 homozygous embryos many cells are labelled, even though they display abnormal organisation (Fig. 5B). Some of the β-gal-positive cells are glial cells, as shown by double labelling with anti-β-gal and anti-repo (data not shown). The β-gal labelling profile is altered even more in the heteroallelic combination glide34/glideN7-4: β-gal-positive cells are fewer and more disorganised than in glide34/glide34(Fig. 5C). These results suggest that, in the mutant, most of the cells that would have normally chosen the glial fate are present but cannot undertake the normal differentiation pathway.

Fig. 5.

Cell fate analysis in glide mutants. (A-C) Double labelling with mouse anti-β-gal (red) and rabbit anti-HRP (green) in glide34/CyO (A), glide34/glide34 (B), and glide34/glideN7-4 (C) embryos at stage 15-16, projections of optic sections. The overall organisation of β-gal labelling is disrupted in the mutants; in addition, the number of labelled nuclei is drastically reduced in the heteroallelic combination (C). Note the defects at the longitudinal fibres (open arrows indicate the breaks), strongest in the heteroallelic combination. Bar 16 μm.

Fig. 5.

Cell fate analysis in glide mutants. (A-C) Double labelling with mouse anti-β-gal (red) and rabbit anti-HRP (green) in glide34/CyO (A), glide34/glide34 (B), and glide34/glideN7-4 (C) embryos at stage 15-16, projections of optic sections. The overall organisation of β-gal labelling is disrupted in the mutants; in addition, the number of labelled nuclei is drastically reduced in the heteroallelic combination (C). Note the defects at the longitudinal fibres (open arrows indicate the breaks), strongest in the heteroallelic combination. Bar 16 μm.

One possible explanation is that cells that normally take the glial fate are transformed into neurones. We have assessed the fate of the cells that maintain β-gal expression in the mutant by performing a double labelling with anti-β-gal and anti-elav, a marker specific for neuronal nuclei (Robinow and White, 1988). Most cells are labelled with both markers in glide34/glideN7-4 embryos (Fig. 6B), while this is never observed in the wild type (Fig. 6A). Colocalisation occurs at different layers in the ventral cord, indicating that most or all glial cell types are transformed into neurones. Similar results have been obtained with glide34/glide34 embryos but fewer cells are transformed (data not shown). The cell fate transformation has been confirmed by using mAb22c10, a marker that labels a subset of neuronal membranes (data not shown). The total number of neuronal nuclei has not been assessed due to their extremely compact organisation. However, we have not observed a dramatic increase of elav labelling in mutant embryos and in same cases we have actually found a decrease in the number of elav-positive nuclei (data not shown).

Fig. 6.

Cell fate transformation in glide mutants. (A,B) Embryos labelled with anti-elav (green) and anti-β-gal (red). Twelve optical sections taken through the ventral cord of a glide34/CyO (A) and twelve sections of a glide34/glideN7-4 (B) embryo. The sections were approx. 1 μm apart. The sections go from ventral (top) to dorsal (bottom). In the wild type, β-gal-positive cells (red) are always located at positions devoid of neurones, and no colocalisation can be observed between the two markers. Different glial cells can be recognised: subperineural and cortical glial cells at superficial positions (short arrows), and longitudinal glial cells (long arrows), located dorsal to the longitudinal fibres. In the mutant (B), most β-gal-expressing cells also express elav (orange nuclei) (open arrows), the transformation being more extensive dorsally than ventrally. Bar 16 μm.

Fig. 6.

Cell fate transformation in glide mutants. (A,B) Embryos labelled with anti-elav (green) and anti-β-gal (red). Twelve optical sections taken through the ventral cord of a glide34/CyO (A) and twelve sections of a glide34/glideN7-4 (B) embryo. The sections were approx. 1 μm apart. The sections go from ventral (top) to dorsal (bottom). In the wild type, β-gal-positive cells (red) are always located at positions devoid of neurones, and no colocalisation can be observed between the two markers. Different glial cells can be recognised: subperineural and cortical glial cells at superficial positions (short arrows), and longitudinal glial cells (long arrows), located dorsal to the longitudinal fibres. In the mutant (B), most β-gal-expressing cells also express elav (orange nuclei) (open arrows), the transformation being more extensive dorsally than ventrally. Bar 16 μm.

The glide mutation affects the axonal scaffold

Since it is known that glial cells are required during axono-genesis (Jacobs and Goodman, 1989; Bastiani and Goodman, 1986), we have examined the axonal scaffold in the glide mutants using markers that recognise neuronal membranes in the peripheral and in the central nervous system. Anti-HRP has allowed us to analyse the commissure organisation (Fig. 5). Anti-fas II labels the longitudinal fibres and the motor-fibres (Grenningloh et al., 1991) (Fig. 7). mAb22c10 recognises peripheral nerves and specific central axons (Fig. 8). One of the most striking phenotypes is the disruption of the longitudinal fibres in several segments, resulting in breaks in the ventral cord (Fig. 7) and in thickenings of longitudinal fibres at the positions preceding the breaks (data not shown). Other defects in axonal growth and pathfinding have been observed in the PNS and in the CNS: axonal fibres are present at ectopic positions (Fig. 7) and axons forming segmental and intersegmental nerves are located at incorrect positions or they are absent (Fig. 8). As for the glial phenotype, glideN7-4 displays the strongest phenotype (data not shown).

Fig. 7.

glide induces breaks in the longitudinal fibres and defects in the peripheral nerves. Embryos labelled with anti-fas II to reveal the longitudinal fibres (lf). (A) wild-type embryo at stage 16; (B) glide26 and (C) glideN7-4 homozygous embryo (stage 16-17). Longitudinal fibres are partially or completely interrupted in several segments (open arrows). Ectopic fibres exiting the central nervous system are indicated by arrowheads. (pn) indicates the peripheral nerves. Bar 25 μm.

Fig. 7.

glide induces breaks in the longitudinal fibres and defects in the peripheral nerves. Embryos labelled with anti-fas II to reveal the longitudinal fibres (lf). (A) wild-type embryo at stage 16; (B) glide26 and (C) glideN7-4 homozygous embryo (stage 16-17). Longitudinal fibres are partially or completely interrupted in several segments (open arrows). Ectopic fibres exiting the central nervous system are indicated by arrowheads. (pn) indicates the peripheral nerves. Bar 25 μm.

Fig. 8.

glide induces defects in axonal growth. Stage 15 embryos, ventral view: wild-type (A,C), glideN7-4(D), glide34 (B) embryos labelled with mAb22c10. In A and B anterior is to the left, in D and E anterior is to the top. Note the defect in the mutant background at the level of the intersegmental (IS) and segmental (S) nerves. In the wild type (A), the two nerves are separated from one another as they exit the central nervous system then, towards the periphery, they run parallel. In the mutant (B), the two nerves run separately throughout their length. so, indicates the ventral sensory organs. (C,D) Projections of optical sections obtained by confocal microscopy. The intersegmental nerve is formed by the aCC and VUM axons (aCC, VUM). In the mutant (D), errors are present at the position of the segmental and intersegmental nerves: in some cases, axon fibres are at incorrect positions (open arrow), in other cases they seem absent (asterisks). Bars, 25 μm (A,B); 10 μm (C); 16 μm (D).

Fig. 8.

glide induces defects in axonal growth. Stage 15 embryos, ventral view: wild-type (A,C), glideN7-4(D), glide34 (B) embryos labelled with mAb22c10. In A and B anterior is to the left, in D and E anterior is to the top. Note the defect in the mutant background at the level of the intersegmental (IS) and segmental (S) nerves. In the wild type (A), the two nerves are separated from one another as they exit the central nervous system then, towards the periphery, they run parallel. In the mutant (B), the two nerves run separately throughout their length. so, indicates the ventral sensory organs. (C,D) Projections of optical sections obtained by confocal microscopy. The intersegmental nerve is formed by the aCC and VUM axons (aCC, VUM). In the mutant (D), errors are present at the position of the segmental and intersegmental nerves: in some cases, axon fibres are at incorrect positions (open arrow), in other cases they seem absent (asterisks). Bars, 25 μm (A,B); 10 μm (C); 16 μm (D).

We describe a mutation that affects glial cells of the peripheral and central nervous system. In the original enhancer trap line, gliderA87, β-gal is expressed in all glial cells except midline and perineural glial cells. Mutations at the glide locus lead to lack of glial cell differentiation and to a fate switch: cells that would have taken the glial fate are transformed into neurones. This indicates that the glide product is a glial promoting factor. In addition, mutations at the glide locus lead to defects in axonal guidance.

glide is necessary for glial cell determination

In gliderA87 embryonic and pupal nervous system, β-gal labelling appears very early on during glial differentiation. The early and general glial expression of β-gal suggests that the gene adjacent to the insert plays an important role in the first steps of glial differentiation, which has prompted us to perform a mutagenesis at that locus.

In the mutant, glial labelling is missing throughout development. All glial cell types in the CNS and in the PNS are affected, except for the perineural and midline glia. The classification of the outermost layer of the CNS, the perineural glia, is still a matter of debate because the ultra-structural features of these cells are rather different from the other glial cells. It has recently been proposed that these cells are still immature in late embryonic stages and that some cells classified as subperineural glial cells are actually perineural cells (see Ito et al., 1995 for a review). As for midline glial cells, they have been shown to be very different from all the other types of glial cells in terms of origin, morphology and patterns of gene expression (Poulson, 1950; Thomas et al., 1984; Crews et al., 1988; Klämbt et al., 1991).

In the weak glide34 allele, cells are born and express β-gal. Most of them, however, do not express repo, one of the earliest glial markers. The few cells that do differentiate into glia are often present in clusters of variable cell number, suggesting defects in glial cell proliferation and migration. glide34 represents most likely a hypomorph in which cells express enough product to enter the differentiation pathway, but not to express the glial phenotype (see lack of repo expression). Indeed, in the strongest combinations in which the β-gal profile can be followed, glideN7-4/glide34, labelling is much lower in terms of intensity and number of cells. This suggests that in glideN7-4, the strongest allele, even the very first step of differentiation cannot be undertaken.

All these results call for glide to have a role in the early steps of glial differentiation. A number of genes affecting glial differentiation have already been identified in the fly (prospero: Doe et al., 1991; pointed: Klämbt, 1993 and repo: Campbell et al., 1994; Xiong et al., 1994; Halter et al., 1995). None of them, however, has such a generalised and early effect on glial development: glide represents the first example of a general glial differentiation promoting factor.

Molecular analysis of the glide gene in wild-type and in mutant flies will enable us to determine whether glide expression reflects the β-gal profile observed in rA87, as has been shown to be the case of many other enhancer trap lines, and to characterise the molecular defect present in the excisions. It will also be important to explore the molecular relationships between glide and other genes expressed in glial cells and to define the cascade of events that lead to gliogenesis. For example, the finding that repo and prospero are not expressed in glide embryos, and that glide is expressed in repo and pointed mutants (Klaes et al., 1994; Halter et al., 1995), already places glide upstream in the hierarchy and suggests that repo, pointed and prospero constitute glide targets. Thus, the glide gene and mutations could be important tools for identifying the other molecular players of glial differentiation.

Role of glide in the neuronal-glial fate choice

Glial cells are missing in glide embryos. In addition, glide mutants display a cell fate switch: cells that would have taken the glial fate are transformed into neurones. This suggests that glia and neurones originate from multipotent precursors. Such precursors may give rise to two types of lineages: mixed neuronal and glial lineages or lineages of only one cell type. Mixed lineages indicating the existence of a common precursor have already been observed in fly and grasshopper embryonic CNS (Udolph et al., 1993; Condron and Zinn, 1994). In grasshopper, the median neuroblast first divides asymmetrically three or four times to give purely neuronal progeny, then, due to a fate switch that requires engrailed activity, divides approximately eight times to give rise to the midline glial cells. Finally, following protein kinase A activation, the median neuroblast switches fate again and produces more neurones (Condron and Zinn, 1994, 1995; Condron et al., 1994). In Drosophila, the neuroglioblast 1-1 delaminates from the ectoderm and gives rise to neurones and glial cells (Udolph et al., 1993). Genetic analyses have also suggested that glial and neuronal differentiation are tightly associated: mutants inducing neuronal hypo-or hyperplasia have parallel effects on glial cell development (Hartenstein et al., 1992; Nelson and Laughon, 1994; Giangrande, 1995; Giangrande, unpublished results). In the future, it will be important to determine whether multipotent precursors always give rise to mixed neuronal and glial lineages or whether, in physiological conditions, some of these precursors only differentiate into one type of cell, glia or neurones, due to the activity of cell fate determining genes like glide.

Mixed lineages have also been observed in vertebrates. When in vivo lineage analyses or in vitro clonal experiments are performed on cells deriving from the neural crest, mixed as well as pure clones have been observed, indicating that a common precursor exists for glial and neuronal cells (see for reviews, Le Douarin and Ziller, 1993; Stemple and Anderson, 1993). In addition, the fate of clonal cultures of crest cells can be manipulated by treatment with glial growth factor (GGF; Shah et al., 1994). Under basal growth conditions, most clones, after culture for 16 days, contain both neurones and Schwann cells, whereas few contain Schwann cells only. Addition of GGF reverses the clone phenotypes: the large majority only contains Schwann cells, with few being mixed.

Most but not all glial cells are transformed into neurones in allelic combinations of intermediate strength. This suggests that in the strongest allele all the glial cells identified by the enhancer trap line display the cell fate switch phenotype and that glide most likely acts as a general promoter factor. In this context, glide would be one of the players in the neuronal-glial fate choice. Given the profile of expression in the enhancer trap line, the glide product is most likely required in the precursors of glial cells, however, the mechanisms by which glide acts are still unknown. It may play an instructive role, dictating the glial fate in cells that express it or repressing a default neuronal state in those cells. In this case, glide would be specifically expressed in glial precursors. Alternatively, glide may play a permissive role, interacting with, or responding to, glialspecific genes. In this case, it is not required that glide is expressed specifically in the glial lineage. Further studies will be necessary to explore these two possibilities.

glide affects central and peripheral axonal pathways

In glide mutants, axons do not behave normally. The overall organisation of the axonal scaffold is deranged, with the most striking phenotype being the breaks in the longitudinal fibres. This phenotype is consistent with previous observations showing that longitudinal glial cells form a pattern that prefigures the first axon pathway and that altered development of these cells induces defects in the longitudinal fibres (Jacobs and Goodman, 1989; Doe et al., 1991; Jacobs, 1993; Campbell et al., 1994; Klaes et al., 1994). In addition, the roots of the intersegmental and segmental nerves are abnormal, and the peripheral nerves display navigation defects, confirming and extending the results obtained in grasshopper. Laser ablation of the glial cell called Segment Boundary Cell induces errors in the grasshopper aCC axons, which normally form the intersegmental nerve (Bastiani and Goodman, 1986). Interestingly, the anterior and posterior commissures are not affected by glide mutation. This agrees with the fact that these structures depend on the presence of midline cells (Thomas et al., 1988; Klämbt et al., 1991; Menne and Klämbt, 1994), which do not express β-gal in the original enhancer trap line, nor in the mutant.

Altogether, these results strongly suggest that glial cells play crucial roles during axonogenesis. This is in agreement with the results that pioneer axons are almost dispensable for the navigation of late axons (Lin et al., 1995). The ablation of specific glial cell types, using transformed flies carrying toxins or cell death genes under the control of cell specific promoters, will enable us to explore further the mechanisms underlying axonal navigation.

glide affects neuronal organisation

Although mutations at the glide locus induce cells to take the neuronal instead of the glial fate we have not observed a severe increase in elav labelling in glide embryos. Indeed, the overall number of neuronal cells seems reduced, compared to wild type, even though this is not detectable at early stages of embryonic development (data not shown). There are at least two possible explanations for the observed phenotype: glial cell differentiation promoted by glide may be necessary for neuroblast proliferation or for neuronal fate maintenance. Studies in flies and in vertebrates have already shown that glial cells do provide support for neuronal survival and for the division neuronal precursors (Pixley, 1992; Buchanan and Benzer, 1993; Ebens et al., 1993; Xiong and Montell, 1995). A better understanding of the glide product function will allow us to study cell-cell interactions in the differentiation and maintenance of the nervous system.

We would like to thank C. Goodman, S. Bradley and V. Auld for providing the rA87 line and for communicating unpublished results. The antibodies 22c10, RK2, anti-repo, anti-elav, anti-fas II, antiprospero, were kindly provided by S. Benzer, A. Tomlinson, A. Travers, G. Rubin, J. Urban and H. Vaessin, respectively. The deficiencies and the complementation groups at 30A-C were a generous gift from D. Kalderon. The twi-lacZ CyO balancer was constructed by M. Haenlin. We would like to thank C. Carteret and C. Ackerman for excellent technical help and P. Heitzler for many discussions and advise. Pictures were realised with the help of people in the photographic facility. We are indebted to E. Borrelli and P. Sassone-Corsi for constant support. Many thanks to P. Lawrence, M. Schubiger, P. Simpson, J. Reed, G. Technau and J. Urban for many thoughtful and helpful comments on the manuscript. This work was supported by funds from the Institut National de la Santé et de la Recherche Médicale, the Centre National de la Recherche Scientifique, the Centre Hospitalier Universitaire Régionale and the Association pour la Recherche contre le Cancer.

Bastiani
,
M.J.
and
Goodman
,
C.S.
(
1986
).
Guidance of neuronal growth cones in the grasshopper embryo III. Recognition of specific glial pathways
.
J. Neurosci
.
6
,
3542
3551
.
Buchanan
,
R. L.
and
Benzer
,
S.
(
1993
).
Defective glia in the Drosophila brain degeneration mutant drop-dead
.
Neuron
10
,
839
850
.
Choi
,
K.-W.
and
Benzer
,
S.
(
1994
).
Migration of glia along photoreceptor axons in the developing Drosophila eye
.
Neuron
12
,
423
431
.
Campbell
,
G.
,
Göring
,
H.
,
Lin
,
T.
,
Spana
,
E.
,
Anderson
,
S.
,
Doe
,
C. Q.
and
Tomlinson
,
A.
(
1994
).
RK2, a glial-specific homeodomain protein required for embryonic nerve cord condensation and viability in Drosophila
.
Development
120
,
2957
2966
.
Condron
,
B. G.
,
Patel
,
N. H.
and
Zinn
,
K.
(
1994
).
engrailed controls glial/neuronal cell fate decision at the midline of the central nervous system
.
Neuron
13
,
541
554
.
Condron
,
B. G.
and
Zinn
,
K.
(
1994
).
The grasshopper median neuroblast is a multipotent progenitor cell that generates glia and neurons in distinct temporal phases
.
J. Neurosci
.
14
,
5766
5777
.
Condron
,
B. G.
and
Zinn
,
K.
(
1995
).
Activation of cAMP-dependent protein kinase A triggers a glial-to-neuronal cell-fate switch in an insect neuroblast lineage
.
Curr. Biol
.
5
,
51
61
.
Crews
,
S. T.
,
Thomas
,
J. B.
and
Goodman
,
C. S
(
1988
).
The Drosophila single-minded gene encodes a nuclear protein with sequence similarity to the per product
.
Cell
52
,
143
151
.
Doe
,
C. Q.
,
Chu-La-Graff
,
Q.
,
Wrigth
,
D. M.
and
Scott
,
M. P.
(
1991
).
The prospero gene specifies cell fates in the Drosophila central nervous system
.
Cell
65
,
451
464
.
Ebens
,
A. J.
,
Garren
,
H.
,
Cheyette
,
B. N. R.
and
Zipursky
,
L.
(
1993
).
The Drosophila anachronism locus: a glycoprotein secreted by glia inhibits neuroblast proliferation
.
Cell
74
,
15
27
.
Fredieu
,
J. R.
and
Mahowald
,
A. P.
(
1989
).
Glial interactions with neurons during Drosophila embryogenesis
.
Development
106
,
739
748
.
Giangrande
,
A.
,
Murray
,
M. A.
and
Palka
,
P.
(
1993
).
Development and organization of glial cells in the peripheral nervous system of Drosophila melanogaster
.
Development
117
,
895
904
.
Giangrande
,
A.
(
1994
).
Glia in the fly wing are clonally related to epithelial cells and use the nerve as a pathway for migration
.
Development
120
,
523
534
.
Giangrande
,
A.
(
1995
).
Proneural genes influence gliogenesis in fly
.
Development
121
,
429
438
.
Grenningloh
,
G.
,
Rehm
,
E. J.
and
Goodman
,
C. S.
(
1991
).
Genetic analysis of growth cone guidance in Drosophila: fasciclin II functions as a neuronal recognition molecule
.
Cell
67
,
45
57
.
Halter
,
D. A.
,
Urban
,
J.
,
Rickert
,
C.
,
Ner
,
S. S.
,
Ito
,
K.
,
Travers
,
A. A.
and
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
.
Hartenstein
,
A. Y.
,
Rugendorff
,
A.
,
Tepass
,
U.
and
Hartenstein
,
V.
(
1992
).
The function of the neurogenic genes during epithelial development in the Drosophila embryo
.
Development
116
,
1203
1220
.
Ito
K.
,
Urban
J.
and
Technau
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.
and
Goodman
,
C. S.
(
1989
).
Embryonic development of axon pathways in the Drosophila CNS. I. A glial scaffold appears before the first growth cones
.
J. Neurosci
.
9
,
2402
2411
.
Jacobs
,
J. R.
,
Hiromi
,
Y.
,
Patel
,
N. H.
and
Goodman
,
C. S.
(
1989
).
Lineage, Migration, and morphogenesis of longitudinal glia in the Drosophila CNS as revealed by a molecular lineage marker
.
Neuron
2
,
1625
1631
.
Jacobs
,
J. R.
(
1993
).
Perturbed glial scaffold formation precedes axon tract malformation in Drosophila mutants
.
J. Neurobiol
.
24
,
611
626
.
Jessen
,
K. R.
,
Brennan
,
A.
,
Morgan
,
L.
,
Mirsky
,
R.
,
Kent
,
A.
,
Hashimoto
,
Y.
and
Gavrilovic
,
J.
(
1994
)
The Schwann cell precursor and its fate: a study of cell death and differentiation during gliogenesis in rat embryonic nerves
.
Neuron
12
,
509
527
.
Klaes
,
A.
,
Menne
,
T.
,
Stollewerk
,
A.
,
Scholz
,
H.
and
Klämbt
,
C.
(
1994
).
The ets transcription factors encoded by the Drosophila gene pointed direct glial cell differentiation in the embryonic CNS
.
Cell
78
,
149
160
.
Klämbt
,
C.
(
1993
).
The Drosophila gene pointed encodes two ets-like proteins which are involved in the development of the midline glial cells
.
Development
117
,
163
176
.
Klämbt
,
C.
and
Goodman
,
C. S.
(
1991
).
The diversity and pattern of glia during axon pathway formation in the Drosophila embryo
.
Glia
4
,
205
213
.
Klämbt
,
C.
,
Jacobs
,
J. R.
and
Goodman
,
C. S.
(
1991
).
The midline of the Drosophila central nervous system: a model for the genetic analysis of cell fate, cell migration and growth cone guidance
.
Cell
64
,
801
815
.
Lane
,
M. E.
and
Kalderon
,
D.
(
1993
).
Genetic investigation of cAMP-dependent protein kinase function in Drosophila development
.
Genes Dev
.
7
,
1229
1243
.
Le Douarin
,
N. M.
and
Ziller
,
C.
(
1993
).
Plasticity in neural crest cell differentiation
.
Curr. Opin. Cell Biol
.
5
,
1036
1043
.
Lin
,
D. M.
,
Auld
,
V. J.
and
Goodman
,
C. S.
(
1995
).
Targeted neuronal cell ablation in the Drosophila embryo: pathfinding by follower growth cones in the absence of pioneers
.
Neuron
,
14
,
707
715
.
Marchionni
,
M. A.
,
Goodearl
,
A. D. J.
,
Chen
,
M. S.
,
Bermingham-McDonogh
,
O.
,
Kirk
,
C.
,
Hendricks
,
M.
,
Danehy
,
F.
,
Misumi
,
D.
,
Sudhalter
,
J.
,
Kobayashi
,
K.
,
Wrobleski
,
D.
,
Lynch
,
C.
,
Baldassarre
.,
M.
,
Hiles
,
I.
,
Davis
,
J. B.
,
Hsuan
,
J. J.
,
Totty
,
N.F.
,
Otsu
,
M.
,
McBurney
,
R. N.
,
Waterfield
.,
M. D.
,
Stroobant
,
P.
and
Gwynne
,
D.
(
1993
).
Glial growth factors are alternatively spliced rebB2 ligands expressed in the nervous system
.
Nature
362
,
312
318
.
Menne
,
T. V.
and
Klämbt
,
C.
(
1994
).
The formation of commissures in the Drosophila CNS depends on the midline cells and on the Notch gene
.
Development
120
,
123
133
.
Meyer
,
M. R.
,
Reddy
,
G. R.
and
Edwards
,
J. S.
(
1987
).
Immunological probes reveal spatial and developmental diversity in insect neuroglia
.
J. Neurosci
.
7
,
512
521
.
Murray
,
M. A.
,
Schubiger
,
M.
and
Palka
,
J.
(
1984
).
Neuron differentiation and axon growth in the developing wing of Drosophila melanogaster
.
Dev Biol
104
,
259
273
.
Nelson
,
H. B.
and
Laughon
,
A.
(
1994
).
Drosophila glial development is regulated by genes involved in the control of neuronal cell fate
.
Roux’s Arch. Dev
.
204
,
118
125
.
Pixley
,
S. K.
(
1992
).
CNS glial cells support in vitro survival, division and differentiation of dissociated olfactory neuronal progenitor cells
.
Neuron
8
,
1191
1204
.
Poulson
,
D. F.
(
1950
).
Histogenesis, organogenesis, and differentiation in the embryo of Drosophila melanogaster
.
In Biology of Drosophila
(ed.
M
Demerec
) pp.
168
274
,
New York
:
J. Wiley and Sons, Inc
.
Robinow
,
S.
and
White
,
K.
(
1988
).
The locus elav of Drosophila melanogaster is expressed in neurons at all developmental stages
.
Dev. Biol
.
126
,
294
303
.
Shah
,
N. M.
,
Marchionni
,
M. A.
,
Isaacs
,
I.
,
Stroobant
,
P.
and
Anderson
,
D.
(
1994
).
Glial growth factor restricts mammalian neural crest stem cells to a glial fate
.
Cell
77
,
349
360
.
Stemple
,
D. L.
and
Anderson
,
D. J.
(
1993
).
Lineage diversification of the neural crest: in vitro investigation
.
Dev. Biol
.
159
,
12
23
.
Thisse
,
C.
,
Perrin-Schmitt
,
F.
,
Stoetzel
,
C.
and
Thisse
,
B.
(
1991
).
Sequence-specific transactivation of the Drosophila twist gene by the dorsal gene product
.
Cell
65
,
1191
1201
.
Thomas
,
J. B.
,
Bastiani
,
M. J.
,
Bate
,
C. M.
and
Goodman
,
C. S
(
1984
).
From grasshopper to Drosophila: a common plan for neuronal development
.
Nature
310
,
203
207
.
Thomas
,
J. B.
,
Crews
,
S. T.
and
Goodman
,
C. S.
(
1988
).
Molecular genetics of the single-minded locus: a gene involved in the development of the Drosophila nervous system
.
Cell
52
,
133
141
.
Udolph
,
G.
,
Prokop
,
A.
,
Bossing
,
T.
and
Technau
,
G. M.
(
1993
).
A common precursor for glia and neuron in the embryonic CNS of Drosophila gives rise to segment-specific lineage variants
.
Development
118
,
765
775
.
Wigglesworth
,
V. B.
(
1959
).
The histology of the nervous system of an insect Rhodnius prolixus (Hemiptera) I. The central ganglia
.
Quart. J. Microsc. Sci
.
100
,
299
313
.
Xiong
,
W.-C.
and
Montell
,
C.
(
1995
).
Defective glia induce neuronal apoptosis in the repo visual system of Drosophila
.
Neuron
14
,
581
590
.
Xiong
,
W.-C.
,
Okano
,
H.
,
Patel
,
N. H.
,
Blendy
,
J. A.
and
Montell
,
C.
(
1994
).
repo codes a glial-specific homeo domain protein required in the Drosophila nervous system
.
Genes Dev
.
8
,
981
994

While this manuscript was being reviewed two papers have been published on the same gene by Jones, B. W., Fetter, R. D., Tear, G. and Goodman, C. S. Cell82, 1013-1023 (1995) and by Hosoya, T., Takizawa, K., Nitta, K. and Hotta, Y. Cell82, 1025-1036 (1995). The results of their analysis on the glide gene, which they called gcm (glial cell missing), are in agreement with ours.