|
|
Proc Natl Acad Sci U S A. 2009 April 21; 106(16): 6778–6783. Published online 2009 April 3. doi: 10.1073/pnas.0808899106. | PMCID: PMC2672493 |
Neuroscience Gcm protein degradation suppresses proliferation of glial progenitors Margaret Su-chun Ho,a Hungwen Chen,b Minghan Chen,a Cécile Jacques,c Angela Giangrande,c1 and Cheng-Ting Chiena2 Institutes of aMolecular Biology and bBiological Chemistry, Academia Sinica, Nankang, Taipei 115, Taiwan; and cInstitut de Génétique et de Biologie Moléculaire et Cellulaire, Institut National de la Santé et de la Recherche Médicale, Centre National de la Recherche Scientifique, Université Louis Pasteur de Strasbourg, BP10142, 67404 Illkirch Cedex, Communauté Urbaine de Strasbourg, France Received September 9, 2008. Gliogenesis in animal development is spatiotemporally regulated so that correct numbers of glia are present to support various neuronal functions. During Drosophila embryonic development, the glial regulatory gene, glial cell missing/glial cell deficient (gcm/glide), promotes glial cell fate and differentiation. Here we describe the ubiquitin–proteasome regulation of the Gcm protein and the consequence in gliogenesis without timely degradation of Gcm. Gcm binds to 2 F-box proteins, Supernumerary limbs (Slimb) and Archipelago (Ago), adaptors of SCF E3 ubiquitin ligases. Ubiquitination and proteasomal degradation of Gcm depend on slimb and ago. In slimb and ago double mutants, Gcm protein levels are enhanced. Concomitantly, glial cell numbers increase owing to proliferation, which can be phenocopied by Gcm overexpression only at the onset of glial differentiation. The glial lineage 5–6A in slimb ago mutants displays excess glial progenies and enhanced Gcm protein levels. We propose that downregulation of Gcm protein levels by Slimb and Ago is required for glial progenitors to exit the cell cycle for differentiation. Keywords: Drosophila, E3 ligase, F-box, supernumerary limbs (slimb), archipelago (ago) The generation and the maintenance of appropriate glial cell numbers are vital for neuronal functions and disease preventions. One aspect on how the glial cell number is regulated can be learned from the proliferation of intermediate progenitors in mammalian neurogenesis. The intermediate progenitor is generated from the asymmetric divisions of the neuroepithelial progenitors in the ventricular zone ( 1, 2). These intermediate progenitors proliferate in the subventricular zone by symmetric cell division, which provides the sufficient number of progenies for subsequent neuronal differentiation in the brain. The presence of intermediate progenitors in neurogenesis is conserved for specific larval brain neurogenic lineages in the invertebrate organism Drosophila ( 3). It would be interesting to understand the existence of similar proliferation modes to generate glial cells. In development of the Drosophila embryonic central nervous system (CNS), the glial cell fate is specified by the master regulatory gene, glial cell missing/glial cell deficient ( gcm/glide, henceforth gcm). Drosophila gcm is transiently expressed in cells destined to differentiate into glia and required for the gliogenesis of most glia in the CNS and all glia in the peripheral nervous system (PNS). Being considered as a binary switch between the neuronal and the glial cell fates, almost all glia are transformed into neurons in gcm mutant embryos, while ectopic expression of gcm converts neurons into glia ( 4– 6). Gcm promotes differentiation of glial cells by activating downstream target genes such as reversed polarity ( repo) and pointed ( pnt) ( 7, 8), and represses the neuronal fate through the activation of tramtrack ( ttk) ( 9, 10). These target genes encode respective transcription factors Repo, PntP1, and Ttk69 whose cooperative activities regulate expressions of further downstream genes in the differentiation of diverse glial fates. The mammalian homolog of gcm, GCMa, encodes an essential transcription factor for placental development, during which GCMa controls the differentiation of the syncytiotrophoblast layer ( 11). A second GCM gene, GCMb, is mutated in the hypoparathyroidism disease, a type of human disorder characterized with decreased levels of parathyroid hormone (PTH) and hypocalcemia ( 12). One of the intriguing issues in regard to cell type-specific transcription factors concerns the complexity in the regulation of their activities. As a transcription factor, the transient and pan-glial expression of gcm in a variety of spatiotemporally distinct glial precursors serves as an excellent model to study this developmental complexity. Not only how gcm is turned on initially in cells adopting the glial fate is unresolved, but also how the expressed Gcm protein is downregulated once glial cells undergo differentiation is undetermined. In addition, the consequence of upregulated Gcm protein stability in gliogenesis is not understood. Recent advances have implicated the ubiquitin–proteasome degradation pathway as the major means to control protein levels ( 13, 14). Among the involved components, the F-box proteins are characterized as substrate-specific regulators of SCF E3 ubiquitin ligases, in protein degradation ( 15– 17). Three Drosophila F-box proteins of the FBW subfamily, Supernumerary limbs (Slimb), Archipelago (Ago), and the uncharacterized CG9144, carry WD repeats for substrate binding. In many cases, a single F-box protein can recognize distinct substrates in a cell-context-dependent manner. For example, Slimb controls the protein levels of Cubitus interruptus to mediate the transduction of Hedgehog signaling and Period in circadian clock neurons ( 18– 20). Ago downregulates the protein levels of Cyclin E during cell cycle progression and dMyc in cell growth ( 21, 22). Here we report the downregulation of Gcm by Slimb and Ago in glial differentiation during Drosophila embryonic development. Gcm interacts with F-box proteins Slimb and Ago that promote the ubiquitination and proteasomal degradation of Gcm. While single mutants of slimb or ago display normal gliogenesis, the number of CNS glia increases in the double slimb ago mutant, which can be phenocopied by overexpression of Gcm and suppressed by the reduction of the gcm gene dosage. The increase of the glial cell population is not caused by the conversion of neurons to glia, but rather by enhanced proliferation after glial induction. Finally, we analyzed the 5–6A lineage in the slimb ago mutant to correlate the generation of ectopic glia and enhanced Gcm expression. Ubiquitination of Gcm Requires F-Box Proteins Slimb and Ago in S2 Cells. The ubiquitination and degradation of GCMa is mediated by the SCF FBW2 E3 ubiquitin ligase complex in which the F-box protein FBW2 is the substrate adaptor ( 23). Sequence comparison indicates that FBW2 shares limited identity (≈18%) to all 3 Drosophila FBWs ( 16). We then tested whether any of the 3 FBWs interacts with Gcm. Cell extracts from Drosophila S2 cells transfected with a Flag-tagged gcm plasmid (p Flag-gcm) alone or with a plasmid carrying a Hemagglutinin (HA)-tagged F-box gene (p slimb-HA, p HA-agoWD, or p HA-CG9144) were analyzed by co-immunoprecipitation ( A). The substrate-binding WD domain of Ago (AgoWD) was used instead of the full-length HA-Ago protein due to its low levels of expression. With the Flag antibody, Flag-Gcm proteins were detected with a molecular weight of ≈72 kDa ( A Middle). The Flag antibody also precipitated Slimb or AgoWD but not CG9144 ( A, lanes 2–4). Flag-GFP proteins in the parallel control failed to pull down any of the F-box proteins ( A, lanes 5–7). Furthermore, as shown in supporting information (SI) Fig. S1A, a Slimb truncation without the substrate-binding WD domain (SlimbΔWD) was not pulled down by Flag-Gcm. Together, these assays indicate specific interactions of Gcm with substrate-binding WD domains of Slimb and Ago. To show that Gcm undergoes ubiquitination, S2 cells transfected with p Flag-gcm and the plasmid expressing HA-ubiquitin (p HA-Ub) were analyzed by immunoprecipitation with the HA antibody and Western blots with the Flag antibody. Ubiquitinated Gcm (Ub-Gcm) species were detected in cells cotransfected with p Flag-gcm and p HA-Ub but not either one alone, indicating that Gcm undergoes ubiquitination in vivo ( B, lanes 2–4). No signals were seen in cells cotransfected with p Flag-gcm and p HA-UbKo, the plasmid with mutations in 7 lysines (p HA-UbKo) that produces the nonconjugatable form of HA-Ub ( B, lane 5), suggesting that the signals seen in lane 4 reflect polyubiquitinated species. In another set of experiments, co-immunoprecipitation with the Flag antibody and Western blot by the HA antibody produced the same results ( Fig. S1B). To examine if Slimb and Ago are involved in the ubiquitination of Gcm, the expressions of slimb and ago were knocked down by respective dsRNA in S2 cells ( Fig. S1C) and ubiquitination of Gcm was analyzed. While the addition of slimb dsRNA suppressed some Gcm ubiquitination and the addition of ago dsRNA had no effect, the addition of both slimb and ago dsRNA suppressed Gcm polyubiquitination more than the addition of either one alone ( B Right, quantifications in Fig. S1B). Therefore, Slimb and Ago share a partially redundant function in the ubiquitination of Gcm in S2 cells. To test whether Gcm protein is regulated by the 26S proteasome degradation pathway, S2 cells were transfected with pFlag-gcm and treated with cycloheximide (CHX) to block protein synthesis. As shown in C and D, the Flag-Gcm protein level decreased over a time course of 7 h, with a half-life of 4 h. When blocking the proteasomal machinery with the inhibitor MG132, the Flag-Gcm protein became stabilized with a half-life of more than 7 h. This proteasomal degradation of Gcm was also partially blocked by the addition of slimb and ago dsRNAs ( C and D). Nonetheless, silencing of slimb and ago expressions did not completely recover the Gcm protein level to that treated with MG132, indicating the involvement of pathways other than the 2 FBWs Slimb and Ago in the proteasomal regulation of Gcm in the S2 cell system. Double Mutants of slimb ago Exhibit Abnormal Glial Patterns. To assess the role of slimb and ago during glial development, the embryonic CNS pattern was revealed by immunostaining for Repo, a glial-specific transcription factor induced by Gcm in all glia except in the midline glia, and Fasciclin II (FasII) to depict the axonal patterns ( C). While embryos homozygous for the null allele slimbP1493 or ago3 ( 19, 21) exhibited normal glial patterns ( Fig. S2), the double slimb ago mutant showed a distinct glial pattern to that of wild-type animals (compare A′ and B′). One prominent phenotype was the increase of Repo-positive glial populations. The number of glia was calculated for the region between the midline (arrow in C) and the 5 lateral chordotonal (LCH) organ-associated glia (arrowhead in C). In the heterozygous slimb ago/twi-GFP control, each hemisegment of A1–A4 included 31.3 ± 0.4 glia, very similar to the reported number of 30 glia ( 24, 25). In the homozygous slimb ago CNS, nevertheless, the number of glia increased to 39.9 ± 0.6 (black bar in D), indicating that ectopic glia are produced in the slimb ago CNS. To test whether the increase of glial cells was caused by the lack of F-box gene activity, a slimb transgene under the control of the ubiquitous tubulin promoter ( tub-slimb) ( 20) was introduced into the slimb ago double mutant. The expression of slimb significantly suppressed the generation of ectopic glial cells (from 39.9 ± 0.6 to 34.1 ± 0.5, gray bar in D), indicating that Slimb controls the glial cell number during gliogenesis. Reduced Gcm Gene Dosage Suppresses the slimb ago Glial Phenotype. To investigate whether the glial phenotype in slimb ago double mutants is mediated by gcm, a loss-of-function allele, gcmN7–4, was used ( 4– 6). Homozygous gcmN7–4 embryos have been shown to lack glia almost completely. The gcmN7–4 allele was combined with slimbP1493 and ago3 mutations to generate triple mutants. The absence of almost all glia in the triple mutant, identical to what had been observed in the gcmN7–4 single mutant, indicates that the generation of ectopic glia in the slimb ago mutant is dependent on gcm activity ( Fig. S3B). Also the presence of one copy of gcmN7–4 suppressed the slimb ago glial phenotype ( Fig. S3C). In the gcmN7–4/+; slimb ago CNS, the number of glia was reduced significantly (31.5 ± 0.5, D). The presence of the gcm null allele, gcm26, in the slimb ago genetic background also suppressed the double mutant phenotype (29.8 ± 0.7). Taken together, these results suggest that the E3 ligase components Slimb and Ago function through gcm and downregulate gcm activity in controlling the glial population. Endogenous Gcm Expression Is Elevated in slimb ago Mutants. To examine Gcm expression in wild-type and slimb ago mutants during glial development, we generated antibodies against the DNA binding region of Gcm (see Materials and Methods). These antibodies were used in immunolabeling for embryos that were costained for Repo during gliogenesis. Similar to the gcm mRNA and protein expression pattern shown previously ( 4, 5), Gcm protein was initially detected in one lateral cell, presumably the longitudinal glioblast (LGB) that also expresses Repo in stage 10 wild-type embryos (arrows in Fig. S4A). Moreover, adjacent to the Gcm- and Repo-positive cell, one Gcm-positive cell was devoid of Repo expression, indicating that the expression of Gcm precedes the expression of Repo during the development of glial cells (arrowheads in Fig. S4A). In stage 11 and 13 wild-type embryos, Gcm is expressed in a number of CNS precursor cells ( Fig. S4 B and C). Gcm expression in the LGB is most prominent among the Repo-positive cells in stage 11 embryos (arrows in Fig. S4B), even though a subset of other cells also expresses both Gcm and Repo. Later on, by stage 13, signals of Gcm and Repo overlap substantially in many nuclei, indicating that Gcm is expressed in glial cells after their induction ( Fig. S4C). By stage 15, Repo accumulates in all glial cells. In contrast, the mature profile of Gcm expression is different from that observed at earlier stages ( Fig. S4D). Gcm protein was detected as scattered puncta throughout the CNS and the intensity of signals was not as strong as what we had observed for earlier-stage embryos. Nonetheless, subsets of the Gcm puncta-containing cells were still labeled with Repo (arrows in Fig. S4D). The significance of these punctate signals is currently unclear. To confirm that these staining patterns reflect Gcm expression, we performed immunostaining in homozygous gcm26 embryos. In these embryos, the Gcm and Repo antibodies detect only unspecific signals ( Fig. S4E). In toto, these results suggest that Gcm is most prominently expressed during embryonic development, when Repo has yet to be initiated, and the Gcm expression disappears by the time when most Repo signals are still persistent at stage 15. Nonetheless, there are cells that express both Gcm and Repo. Thus, the temporal sequence of Gcm and Repo expressions is consistent with Gcm activating repo gene expression. We then investigated the profile of Gcm expression in slimb ago mutants. Similar to what was observed in the wild type, Gcm was detected in a subset of Repo-positive cells in heterozygous slimb ago/+ embryos in stage 13 (A). In homozygous slimb ago embryos of the same stage, however, Gcm expression was elevated in Repo-positive cells (arrows in B). Not only were Gcm levels elevated in the absence of Slimb and Ago (from 10.3 ± 1.4 to 23.0 ± 1.9 in E), but also the number of Gcm- and Repo-positive cells was increased (from 1.5 ± 0.2 to 4.6 ± 0.3 in F, S13). The punctate Gcm pattern in late stage 15 was still observed in slimb ago mutant embryos ( C and D), with the number of Repo-positive cells expressing punctuate Gcm increased compared to that of the heterozygous slimb ago/+ embryos (3.2 ± 0.3–7.3 ± 0.5 in F, S15). These results suggest that Slimb and Ago are required to downregulate Gcm protein levels in Repo-positive differentiating glial cells. Gcm Overexpression in Glial Lineages Phenocopies slimb ago Mutants. We next determined whether the elevation of the Gcm protein level observed in the slimb ago mutant affects glial development. Gcm was ectopically expressed using tissue-specific GAL4 drivers and CNS glial patterns were examined. Expressions of Gcm using a neural-specific sca-GAL4 driver virtually converted all neurons into glia ( Fig. S5B), consistent with previous studies ( 4, 5). This massive increase in the glial cell population, however, was not seen in the slimb ago mutant. Interestingly, gcm-GAL4-driven Gcm expression (denoted as gcm> gcm) in specified glial cells displayed an enlarged glial cell population, similar to what was observed in the slimb ago mutant ( Fig. S5C). The number of Repo-positive cells was increased from the gcm-GAL4 control of 30.9 ± 0.5 to 36.5 ± 0.6 ( Fig. S5E). Furthermore, when Gcm expression was driven by repo-GAL4 at a later stage than gcm-GAL4, the number of glial cells remained similar to that in the gcm-GAL4 control ( Fig. S5 D and E), suggesting that the increase of Gcm protein levels later during glial differentiation fails to recapitulate the slimb ago mutant phenotype. These results suggest that Gcm is effective at inducing ectopic glia during a short time period right after glial induction by gcm. Gcm Is Implicated in Cell Proliferation. The increase in the glial population in the slimb ago mutant could be due to a rise in cell proliferation or to an inhibition of apoptosis. To discriminate between these 2 possibilities, we first examined the mitotic cell population in the slimb ago CNS using the phospho-Histone H3 marker (PH3). Interestingly, the number of PH3-positive cells dramatically increased in the slimb ago mutant ( Fig. S6 A and B). In contrast, the number of apoptotic cells remained unaltered as indicated by the expression of activated Caspase-3 (Cas-3) ( Fig. S6 C and D). These results indicate that cell proliferation is enhanced in the slimb ago mutant. To assess ectopic proliferation in the Gcm-expressing glial precursors, embryos of the gcm> GFP control and gcm> GFP+ gcm that overexpresses Gcm were immunostained for PH3 and the nuclear marker Hoechst ( A and B, Fig. S6 E and F). Whereas 4.0 ± 0.2 PH3-, Hoechst-, and GFP-positive cells were scored in the gcm> GFP+ gcm embryos for each hemisegment, there were only 1.4 ± 0.1 such cells in gcm> GFP embryos ( E). Examples of cells marked positively for PH3, Hoechst, and GFP are shown in Fig. S6I. To examine the persistence of proliferation in later stages of glial development, gcm> GFP and gcm> GFP+ gcm embryos were labeled with PH3 and Repo ( C and D, Fig. S6 G and H). Expression of Repo indicates an entering into glial differentiation, and concomitantly the termination of cell proliferation, as shown by 0.2 ± 0.1 PH3- and Repo-positive cells in each hemisegment. In gcm> GFP+ gcm embryos the number of PH3- and Repo-positive cells per hemisegment was increased to 0.8 ± 0.1 ( E). Taken together, Gcm overexpression triggered by gcm-GAL4, likely before glial differentiation, promotes glial cell proliferation, leading to an increase in the glial cell population. To examine how Gcm promotes glial proliferation, expressions of Gcm target genes in gcm> GFP and gcm> GFP+ gcm embryos in stage 13 were compared by quantitative PCR (Q-PCR) analysis ( F). These targets include the glial differentiation gene repo and the neuronal repressor gene ttk that have been previously described ( 7). When Gcm expression was triggered by the gcm-GAL4 driver, expression of ttk remained constant whereas repo expression decreased, indicating a possible delay in glial differentiation ( F). Expressions of cell cycle regulators Cyclin B ( Cyc B) and Cyclin E ( Cyc E) were also analyzed. In gcm> GFP+ gcm embryos, ectopic Gcm induced Cyc B, suggesting that Cyc B may be a downstream gene for the Gcm activator. In contrast, Cyc E expression was down-regulated, a reduction in the number of cells in the G1 phase. These results, taken together with the increase in the number of PH3-positive cells, are consistent with the assumption that glial progenitors stay in the G1 phase and express Cyc E before the onset of glial differentiation. Forced expression of Gcm promotes progenitors to progress through the G2 phase and mitosis, leading to further proliferation in glial lineages. Analysis of Neuroglial Lineage 5–6A in slimb ago Mutants. We then investigated in the slimb ago mutant how glial proliferation was enhanced in a specific lineage, the abdominal lineage 5–6 or 5–6A, which is derived from a lateral neuroglioblast (NGB) and contains 10–14 local interneurons and 1 subperineural and 1 exit glial cell ( 26, 27). The enhancer trap line sevenup-lacZ ( svp-lacZ) that labels the progenies of 5–6A was used to examine the slimb ago CNS of early stage 12. In svp-lacZ, slimb ago/twi-GFP embryos, only 1.4 ± 0.1 Repo- and LacZ-positive glial cells per hemisegment were scored (arrows in A). In contrast, extra Repo- and LacZ-positive glial cells were seen in the slimb ago CNS (3.0 ± 0.2, B and E). Scoring Elav- and LacZ-positive cells in both genotypes reveals no alternation in the number of neurons in the 5–6A lineage ( E). The enhancer trap line, K-lacZ, was used to examine 5–6A-derived glia near the CNS/PNS exit zone at stage 15 ( 26). Indeed, ectopic Repo- and LacZ-positive glial cells were also observed in the slimb ago CNS (arrows in C and D). Altogether, these results demonstrate that the lineage composition of 5–6A is affected and these extra 5–6A glia can account for part of the increased glial population in the slimb ago mutant. We then addressed whether these supernumerary glial cells in slimb ago embryos express higher or persistent levels of Gcm during proliferative phases ( Fig. S7 A and B). In slimb ago/twi-GFP CNS carrying the 5–6A marker svp-lacZ, an average of 0.9 ± 0.1 Gcm and LacZ double positive cells were detected per hemisegment in stage 12, in comparison to 2.9 ± 0.1 cells in homozygous slimb ago CNS ( E). Also, higher levels of Gcm expression were detected in these double mutant cells ( F). Thus, the Gcm protein level is elevated and the higher-level GCM protein is present more persistently in the presumptive 5–6A-derived glial progenies when they continue proliferation in the slimb ago mutant. To further examine the proliferation status of the extra 5–6A-derived progenies expressing higher levels of Gcm, the Cyc B antibodies were used to immunostain svp-lacZ embryos either heterozygous or homozygous for slimb ago ( Fig. S7 C and D). More LacZ-positive cells were found to be Cyc B positive in the slimb ago homozygotes, indicating that these 5–6A-derived cells enter the G2 phase (1.1 ± 0.2–3.1 ± 0.2, E). In addition, Cyc B levels were also enhanced in these LacZ-positive cells ( F). Thus, in the absence of Slimb and Ago, enhanced proliferation is detected in the 5–6A lineage. Here we describe the ubiquitination and proteosomal degradation of Gcm through interaction with the F-box proteins Slimb and Ago. In double mutants for slimb and ago, the glial population is increased and the Gcm protein levels are enhanced. Genetic analyses indicate that Slimb and Ago regulate gliogenesis by downregulating Gcm activity. Persistent Gcm expression in the glial progenitor promotes further proliferation. Analyses of the glial lineage 5–6A in slimb ago mutants suggest that the generation of ectopic glia, the elevation of Gcm levels, and the proliferation of glial progenitors are highly correlated. Regulation of Gcm Protein Levels Independent of Glial Cell Induction. The Drosophila glial-specific transcription factor Gcm has been considered as a typical binary switch in glial cell induction. We suggested that the increased glial cell number in the slimb ago mutant is not because of fate transformation from neurons. The PNS lineages expressed normal Elav and Repo patterns in the slimb ago mutant ( Fig. S2 D–G). Also, the thoracic 6–4 lineage (6–4T) displayed similar numbers of neurons and glia in wild type and the slimb ago mutant ( Fig. S2 H and I). In addition, the number of 5–6A neurons (LacZ- and Elav-positive cells) was almost identical to that in wild type, albeit the glial number was increased in the slimb ago mutant ( E). 22C10 staining for subsets of CNS neurons revealed no alternations of cell fate ( Fig. S8). These observations also suggest that Slimb and Ago are necessary to downregulate Gcm levels for glial progenitors to differentiate in specific glial lineages. Mutant embryos of slimb ago failed to hatch and gcm> gcm animals were lethal at pupal stages, which are possibly caused by extra glial cells. To understand whether extra glial cells cause neural defects such as CNS structures, double mutant and Gcm overexpression embryos were labeled with the axonal marker 22C10 and the midline marker Slit. The axonal and midline structures remained intact in these mutant embryos ( Fig. S8). Interestingly, migration and morphology of glial cells were abnormal in slimb ago mutant embryos ( Fig. S9), but not in gcm> gcm embryos (data not shown). These defects, therefore, are likely caused by an upregulation of protein levels for factors other than Gcm. Our study also provides mechanistic insights that the ubiquitination and degradation of Gcm promoted by Slimb and Ago play a critical role in regulating the abundance of Gcm in glial development. Nonetheless, upregulated Gcm protein levels in slimb ago mutant embryos in stage 13 might be suppressed in stage 15 (compare B and D). In S2 cells, the effect on Gcm protein levels by knockdown of slimb and ago could not fully recapitulate the blockage of proteasomal activity (D). These results indicate the existence of other Gcm downregulation mechanisms. Gcm Promotes Proliferation of Glial Intermediate Progenitors. The evidence for glial proliferation in the slimb ago double mutant was first shown by the increased number of cells expressing the mitotic marker PH3 ( Fig. S6). This proliferation is also evident in Gcm overexpression, however, at a specific time window as revealed by using temporal-distinct GAL4 drivers, suggesting that the Gcm protein has a permissive rather than an instructive role in proliferation before the onset of glial differentiation. Expression of Gcm by gcm-GAL4 promotes cell cycle progression. In particular, Cyc B transcript levels were upregulated while Cyc E transcript levels were reduced upon Gcm overexpression ( F). During Drosophila retina development, Cyc E protein peaks in G1-arrested precluster cells that lack Cyc B, and Cyc E diminishes in cells that express Cyc B and undergo an extra round of cell division ( 28). Our result echoes the complementary expression of Cyc B and Cyc E in cell cycle progression in retina development. This result is consistent with a model whereby Gcm promotes the progression of G1-arrested cells for further rounds of cell divisions. These G1-arrested cells are otherwise ready to differentiate into terminal glia. One possibility that glial progenitors are more proliferative in gcm> gcm embryos is that excess Gcm protein allows cells to stay in the progenitor-like state, causing a delay in entering differentiation. In the gcm> GFP control, 1 proliferating cell (labeled by PH3) per hemisegment was found, while among these cells, only 0.2 cells expressed the glial differentiation marker Repo. The number of proliferating cells was increased to 4.0 cells per hemisegment and only 0.8 cells were ready to enter differentiation in gcm> gcm+ GFP ( E). Therefore, most of the ectopic PH3-positive cells in Gcm overexpression are not yet ready to enter differentiation. Also, expression of repo mRNA was initially suppressed by Gcm overexpression in stage 12–13 ( F) but was enhanced in stage 15 (data not shown). Delay in entering differentiation may cause later defects such as migration to stereotyped positions for axonal wrapping ( Fig. S9). These lines of evidence support the notion that persistent Gcm expression in glial precursors promotes glial proliferation before differentiation. It is intriguing to speculate that elevated Gcm levels may promote the proliferation of intermediate progenitors, similar to those for generating radial glial cells ( 1, 29). The intermediate progenitors carry the potentials to self-renew and to amplify the number of cell progenies. Therefore, the Gcm transcriptional factor would be the critical regulator to promote continuous proliferation of intermediate progenitors, and down-regulation of Gcm by Slimb- and Ago-mediated protein degradation would be essential for entering glial differentiation. Fly Stocks. Mutant alleles in this study are as follows: slimbP1493 and tub-slimb ( 19, 20), ago3 ( 21), K-lacZ ( 26), svp-lacZ, gcmN7–4, gcm26 ( 27), gcm-GAL4 ( 30– 32), UAS-gcm ( 33), sca-GAL4, and repo-GAL4. Other stocks are Sp/ CyO, twi-GFP; Dr/TM6B, Sb, Dfd-YFP and Sp/CyO, twi-lacZ; Dr/TM3, Ser, twi-GFP (Bloomington Drosophila Stock Center, BDSC). Antibodies and Immunohistochemistry. The mouse Gcm polyclonal antibodies were raised against a region of Gcm (amino acids 34–186) by LTK BioLaboratories. For other primary and secondary antibodies, please see SI Materials and Methods. Embryos were collected from 9 to 10.5 h for stage 12 and from 12 to 14 h for stage 15 for immunostaining. Cell Transfection, dsRNA Knockdown, Immunoprecipitation, and Q-PCR. Plasmids p Flag-gcm, p HA-agoWD, p HA-CG9144, and p HA-Ub were constructed in this study (see SI Materials and Methods) and p slimb-HA was provided by I. Edery ( 20). S2 cell plasmid transfection and dsRNA treatment, Western blot, immunoprecipitation, and Q-PCR are described in detail in SI Materials and Methods. We thank J. Jiang, K. Moberg, I. Edery, C. Jagla, M. Bhat, J. Urban, C. Klambt, U. Tepass, the Bloomington Drosophila Stock Center, and the Developmental Studies Hybridoma Bank for reagents and L. Soustelle and C. Diebold for technical help. M.S.H. is supported by postdoctoral fellowships from Academia Sinica and the National Science Council of Taiwan and a European Molecular Biology Organization travel grant. C.J. was supported by Ministére de la Recherche et de l'Espace and Association pour la Recherche sur le Cancer fellowships. A.G. is supported by Institut National de la Santé et de la Recherche Médicale, Centre National de la Recherche Scientifique, Hôpital Universitaire de Strasbourg, Association pour la Recherche contre le Cancer, Ligue contre le cancer, Agence Nationale de la Recherche, and the European Economic Community. C.T.C. is supported by Academia Sinica and the National Service Center. 1. Merkle FT, Alvarez-Buylla A. Neural stem cells in mammalian development. Curr Opin Cell Biol. 2006;18(6):704–709. [PubMed] 2. Noctor SC, Martinez-Cerdeno V, Ivic L, Kriegstein AR. Cortical neurons arise in symmetric and asymmetric division zones and migrate through specific phases. Nat Neurosci. 2004;7(2):136–144. [PubMed] 3. Bello BC, Izergina N, Caussinus E, Reichert H. Amplification of neural stem cell proliferation by intermediate progenitor cells in Drosophila brain development. Neural Dev. 2008;3:5. [PubMed] 4. Hosoya T, Takizawa K, Nitta K, Hotta Y. Glial cells missing: a binary switch between neuronal and glial determination in Drosophila. Cell. 1995;82(6):1025–1036. [PubMed] 5. Jones BW, Fetter RD, Tear G, Goodman CS. Glial cells missing: A genetic switch that controls glial versus neuronal fate. Cell. 1995;82(6):1013–1023. [PubMed] 6. Vincent S, Vonesch JL, Giangrande A. Glide directs glial fate commitment and cell fate switch between neurons and glia. Development. 1996;122(1):131–139. [PubMed] 7. Campbell G, et al. RK2, a glial-specific homeodomain protein required for embryonic nerve cord condensation and viability in Drosophila. Development. 1994;120(10):2957–2966. [PubMed] 8. Klaes A, Menne T, Stollewerk A, Scholz H, Klambt C. The Ets transcription factors encoded by the Drosophila gene pointed direct glial cell differentiation in the embryonic CNS. Cell. 1994;78(1):149–160. [PubMed] 9. Giesen K, et al. Glial development in the Drosophila CNS requires concomitant activation of glial and repression of neuronal differentiation genes. Development. 1997;124(12):2307–2316. [PubMed] 10. Yuasa Y, et al. Drosophila homeodomain protein REPO controls glial differentiation by cooperating with ETS and BTB transcription factors. Development. 2003;130(11):2419–2428. [PubMed] 11. Yu C, et al. GCMa regulates the syncytin-mediated trophoblastic fusion. J Biol Chem. 2002;277(51):50062–50068. [PubMed] 12. Sticht H, Hashemolhosseini S. A common structural mechanism underlying GCMB mutations that cause hypoparathyroidism. Med Hypotheses. 2006;67(3):482–487. [PubMed] 13. Cardozo T, Pagano M. The SCF ubiquitin ligase: insights into a molecular machine. Nat Rev Mol Cell Biol. 2004;5(9):739–751. [PubMed] 14. Willems AR, Schwab M, Tyers M. A hitchhiker's guide to the cullin ubiquitin ligases: SCF and its kin. Biochim Biophys Acta. 2004;1695(1–3):133–170. [PubMed] 15. Ho MS, Ou C, Chan YR, Chien CT, Pi H. The utility F-box for protein destruction. Cell Mol Life Sci. 2008;65(13):1977–2000. [PubMed] 16. Ho MS, Tsai PI, Chien CT. F-box proteins: The key to protein degradation. J Biomed Sci. 2006;13(2):181–191. [PubMed] 17. Kipreos ET, Pagano M. The F-box protein family. Genome Biol. 2000;1(5):REVIEWS3002. [PubMed] 18. Grima B, et al. The F-box protein slimb controls the levels of clock proteins period and timeless. Nature. 2002;420(6912):178–182. [PubMed] 19. Jiang J, Struhl G. Regulation of the Hedgehog and Wingless signalling pathways by the F-box/WD40-repeat protein Slimb. Nature. 1998;391(6666):493–496. [PubMed] 20. Ko HW, Jiang J, Edery I. Role for Slimb in the degradation of Drosophila Period protein phosphorylated by Doubletime. Nature. 2002;420(6916):673–678. [PubMed] 21. Moberg KH, Bell DW, Wahrer DC, Haber DA, Hariharan IK. Archipelago regulates Cyclin E levels in Drosophila and is mutated in human cancer cell lines. Nature. 2001;413(6853):311–316. [PubMed] 22. Moberg KH, Mukherjee A, Veraksa A, Artavanis-Tsakonas S, Hariharan IK. The Drosophila F box protein archipelago regulates dMyc protein levels in vivo. Curr Biol. 2004;14(11):965–974. [PubMed] 23. Yang CS, et al. FBW2 targets GCMa to the ubiquitin-proteasome degradation system. J Biol Chem. 2005;280(11):10083–10090. [PubMed] 24. Bossing T, Udolph G, Doe CQ, Technau GM. The embryonic central nervous system lineages of Drosophila melanogaster. I. Neuroblast lineages derived from the ventral half of the neuroectoderm. Dev Biol. 1996;179(1):41–64. [PubMed] 25. Schmidt H, et al. The embryonic central nervous system lineages of Drosophila melanogaster. II. Neuroblast lineages derived from the dorsal part of the neuroectoderm. Dev Biol. 1997;189(2):186–204. [PubMed] 26. De Graeve F, et al. The ladybird homeobox genes are essential for the specification of a subpopulation of neural cells. Dev Biol. 2004;270(1):122–134. [PubMed] 27. Ragone G, et al. Transcriptional regulation of glial cell specification. Dev Biol. 2003;255(1):138–150. [PubMed] 28. Firth LC, Baker NE. Extracellular signals responsible for spatially regulated proliferation in the differentiating Drosophila eye. Dev Cell. 2005;8(4):541–551. [PubMed] 29. Kriegstein A, Noctor S, Martinez-Cerdeno V. Patterns of neural stem and progenitor cell division may underlie evolutionary cortical expansion. Nat Rev Neurosci. 2006;7(11):883–890. [PubMed] 30. Soustelle L, Giangrande A. Novel gcm-dependent lineages in the postembryonic nervous system of Drosophila melanogaster. Dev Dyn. 2007;236(8):2101–2108. [PubMed] 31. Paladi M, Tepass U. Function of Rho GTPases in embryonic blood cell migration in Drosophila. J Cell Sci. 2004;117(Pt 26):6313–6326. [PubMed] 32. Sepp KJ, Auld VJ. Conversion of lacZ enhancer trap lines to GAL4 lines using targeted transposition in Drosophila melanogaster. Genetics. 1999;151(3):1093–1101. [PubMed] 33. Bernardoni R, Miller AA, Giangrande A. Glial differentiation does not require a neural ground state. Development. 1998;125(16):3189–3200. [PubMed]
|
