Drosophila miR-124 regulates neuroblast proliferation through its target anachronism

MicroRNAs (miRNAs) are short non-coding RNAs that act as post-transcriptional regulators of gene expression. Animal miRNAs interact with target mRNAs via partial complementary base pairing to target sites. In most cases this results in repression of mRNA expression through destabilization of the target mRNA and/or translational repression (reviewed by Bushati and Cohen, 2007; Filipowicz et al., 2008; Bartel, 2009). miRNAs have been implicated in many biological phenomena, particularly those associated with dynamic cellular and/or developmental processes such as embryonic development and stem cell differentiation (Karp and Ambros, 2005; Shcherbata et al., 2006; Stadler and Ruohola-Baker, 2008). Given the complexity of the nervous system and the vast number of genes and signaling molecules required for its proper architecture and wiring, miRNAs are likely to play a significant role in central nervous system (CNS) development and function (Kosik, 2006; Christensen and Schratt, 2009).

Among the animal miRNAs implicated in CNS development, miR-124 has been the focus of considerable interest. miR-124 is deeply conserved and shows CNS-enriched expression in all animals examined (Stark et al., 2005; Cao et al., 2007; Kapsimali et al., 2007; Makeyev et al., 2007; Visvanathan et al., 2007; Rajasethupathy et al., 2009; Shkumatava et al., 2009). Over the past few years, there have been several reports on the roles of miR-124 in different model organisms. These studies were in large part based on miRNA overexpression or depletion using antisense oligonucleotides and reached different conclusions on whether miR-124 inhibits neuronal precursor proliferation and/or promotes neuronal differentiation (Lim et al., 2005; Cao et al., 2007; Makeyev et al., 2007; Visvanathan et al., 2007; Cheng et al., 2009; Maiorano and Mallamaci, 2009; Qiu et al., 2009; Arvanitis et al., 2010; Clark et al., 2010; Liu et al., 2011). In vivo studies using genetic mutants that completely remove miRNA function might provide the means to resolve the role of miR-124 in CNS development.

Here we report the in vivo analysis of miR-124 function during the development of the CNS in Drosophila. Mutants lacking miR-124 show reduced proliferative activity of neuronal progenitor cells in the developing larval central brain, resulting in a decrease in the production of adult postmitotic neurons. We identify anachronism as a functionally important target through which miR-124 controls neuronal stem cell proliferation.

Drosophila stocks were obtained from the Bloomington Stock Center. The UAS-ana-RNAi line is stock #27515. EP-ana is stock #22173. Flies were maintained on standard yeast-cornmeal-agar medium at 25°C unless otherwise stated. Canton S flies were used as the wild-type control.

To generate positively labeled MARCM clones, hsFLP, elav-Gal4, UAS-mCD8::GFP; FRT40A, tubP-Gal80/Cyo flies were crossed to FRT40A or FRT40A, miR-124^Δ177,w+/Cyo or FRT40A, miR-124^Δ177,w+/Cyo; UAS-miR-124 or FRT40A, miR-124^Δ177,w+, ana¹/Cyo. Embryos were collected over a 4-hour time window and raised at 25°C for 21-25 hours before a 60-minute heat-shock treatment at 37°C. The treated L1 larvae were grown at 18°C until dissection. All larvae were dissected at late wandering third larval stage unless otherwise stated.

Gene targeting by ends-out homologous recombination was as previously described (Chen et al., 2011). Two independent knockout alleles of miR-124, namely Δ4 and Δ177, were made using modified ends-out gene targeting vectors (Weng et al., 2009). Loss of the miR-124 hairpin sequence in these two alleles was verified by PCR and TaqMan miRNA qPCR assay (Invitrogen). The UAS-miR-124 lines were made by cloning a 250 bp genomic fragment containing the miRNA hairpin into the 3′UTR of dsRed in pUAST as described (Brennecke et al., 2005). The ana 3′UTR luciferase reporters were made by cloning the 900 bp ana 3′UTR downstream of luciferase, under control of the tubulin promoter (Brennecke et al., 2003). ana 3′UTR reporters with mutated miR-124 sites were generated by PCR using primers designed to change the seed region from GTGCCTT into GTACATG. PCR products were sequence verified. The piB-miR-124 plasmid was generated by replacing the GFP reporter in piB-GFP (Bateman et al., 2006) with a 430 bp genomic fragment containing the miRNA hairpin in the center using SalI and BamHI sites. The miR-124 RMCE-hairpin-rescued strain was generated using the piB-miR-124 construct as described (Weng et al., 2009). miRNA quantitative real-time PCR (qRT-PCR) was used to verify that the rescued allele produced a normal level of mature miR-124 miRNA.

Drosophila S2 cells were transfected in 24-well plates with 250 ng miRNA expression plasmid or empty vector, 25 ng firefly luciferase reporter plasmid, and 25 ng Renilla luciferase DNA as a transfection control. Transfections were performed in triplicate in at least three independent experiments. Sixty hours after transfection, Dual-Luciferase Reporter Assays (Promega) were performed according to the manufacturer’s instruction. For luciferase assays on larval CNS, tissues were dissected and immediately lysed in Passive Lysis Buffer (Promega). Luciferase activity was normalized to total protein content measured in the same sample using the Bradford method (Bio-Rad).

For miRNA qRT-PCR, primer sets designed to amplify mature miR-124 were obtained from Applied Biosystems. Reverse transcription was performed on 20 ng total RNA. miRNA levels were calculated relative to reference genes snoR442 and 2SrRNA (Applied Biosystems), after having confirmed that the level of these two small RNAs remains constant in the relevant fly strains. For mRNA qRT-PCR, total RNA was purified using the RNeasy Kit and treated with on-column DNase for 60 minutes (Qiagen) to eliminate DNA contamination. First-strand synthesis used oligo(dT) primers and SuperScript RT-III (Invitrogen). Measurements were normalized to rp49 (RpL32 – FlyBase) and Actin 42A controls.

TU tagging was performed as described (Miller et al., 2009) with the following modification. Larvae of the indicated genotypes were collected in groups of 20 at 72 hours after larval hatching and transferred to food vials with 4-TU-containing yeast paste at 29°C for 16 hours before CNS dissection.

Larval CNS tissues were dissected and fixed in 4% formaldehyde in PBS with 0.1% Triton X-100 for 20 minutes on ice. The following primary antibodies were used: rat anti-Elav at 1:50, mouse anti-Prospero at 1:100 (Developmental Studies Hybridoma Bank); rabbit anti-phospho-histone H3 at 1:200 (Cell Signaling); chicken anti-GFP at 1:2000 (Abcam); and guinea-pig anti-Deadpan (a gift from James Skeath, Washington University, St Louis, MO, USA). Secondary antibodies used were conjugated to Alexa Fluor 555, 633 or 488 (Invitrogen) and used at 1:500, 1:300 or 1:500, respectively. The DNA stain was DAPI (Sigma). Samples were mounted in Vectashield (Vector Laboratories). Quantification of mitotic neuroblasts was performed using projections of confocal sections.

The miR-124 locked nucleic acid probe was from Exiqon. Anti-DIG-POD primary antibody (1:200) was detected using the Tyramide Signal Amplification Kit (cyanine 3) from Perkin Elmer following standard protocols.

In Drosophila, miR-124 is expressed in the CNS from early embryonic stages to adulthood (Stark et al., 2005; Ruby et al., 2006). To further characterize the neuronal expression pattern of miR-124, a 4.7 kb genomic region upstream of the miR-124 locus was used to drive expression of a GFP reporter using the GAL4 system (Brand and Perrimon, 1993). This reporter showed GFP expression in the developing CNS of the embryo. In mature third instar larvae, GFP expression was high in the central brain and ventral nerve cord, and detectable but somewhat lower in the optic lobes (Fig. 1A). miR-124-GFP was co-expressed with neuroblast (NB) markers such as the transcription factor Deadpan (Dpn) and with the postmitotic neuronal marker Embryonic lethal abnormal vision (Elav) (Fig. 1B). Projection from a series of optical sections showed miR-124-GFP expression in most Dpn-positive cells (Fig. 1C), suggesting that miR-124 is expressed in most of the neuronal precursors of the developing larval brain.

We next examined miR-124 expression using fluorescent in situ hybridization and made use of the MARCM strategy (Lee and Luo, 1999; Wu and Luo, 2006) to compare clones of genetically marked miR-124 mutant cells with control clones. Mature miR-124 was abundant in postmitotic neurons and present at low levels in NBs (Fig. 1D), but no signal was detected in the miR-124 mutant tissue, indicating the specificity of the probe (Fig. 1E). These observations indicate that miR-124 is expressed in proliferating neuronal progenitors as well as in differentiating postmitotic neurons.

In order to explore miR-124 function in the neuronal progenitor population, a mutant was generated using an ends-out gene-targeting vector modified to support recombinase-mediated cassette exchange (RMCE) (Weng et al., 2009). The mutant deleted 304 bp encompassing the miR-124 hairpin sequence and introduced the mini-white reporter flanked by inverted attP sites to permit subsequent genetic rescue by RMCE (Fig. 2A). Reintroduction of the miR-124 hairpin sequence into the targeted locus by RMCE restored the expression of mature miR-124 transcript to normal levels (Fig. 2B). Homozygous miR-124 mutants showed reduced viability when reared under normal conditions, together with their heterozygous siblings. However, when the mutant individuals were isolated as embryos and raised under non-competitive conditions (in groups of 20-30 mutant larvae per vial) they were fully viable and showed no obvious morphological defects. RMCE-rescued mutants were fully viable under normal conditions.

Drosophila central brain NBs undergo repeated rounds of asymmetric division to maintain the stem cell-like NB and to generate intermediate progenitors called ganglion mother cells (GMCs). Multiple rounds of NB division produce clonally related populations of neurons and glia. After an initial phase of proliferation in the embryo, most NBs undergo a period of quiescence but then resume proliferation during larval development. This second phase of NB proliferation generates the neurons that make up the bulk of the adult CNS (Hartenstein et al., 2008). Larval brain NBs are classified as type I or type II depending on their proliferative characteristics. Type I NBs divide repeatedly to produce a series of GMCs that each divide once more to produce two postmitotic progeny. Type II NB lineages differ in that the intermediate GMC progenitors undergo multiple rounds of proliferation. This transit amplification process produces larger clones from type II NBs. Most of the subsequent analysis will focus on type I NBs because these constitute the majority of larval brain NBs.

In light of the finding that miR-124 is expressed in larval NBs, we sought to determine its role in the development of the central brain NB lineages by comparing genetically marked NB clones that were normal or mutant for miR-124. MARCM clones were induced randomly in early first instar larvae and analyzed in late third instar larval brains. All labeled control and miR-124 mutant type I NB lineages contained one large cell, the NB, which expressed Dpn (Fig. 3A,B), as well as many smaller cells. GMCs were identified by expression of the differentiation marker Prospero (Pros) (Fig. 3C,D) and by Partner of numb (Pon) (Fig. 3E,F), which is expressed in the NB and segregated to the GMCs during asymmetric NB division. Pros and Pon were detected in wild-type and mutant GMCs. GMCs were found clustered directly adjacent to the large apical NB in control and mutant clones (Fig. 3C-F). In addition to the NB and GMC precursors, control and miR-124 clones contained many smaller cells that expressed Elav, representing the neuronal progeny of the GMCs (Fig. 3A-F). The presence of the labeled NB and GMC together with many postmitotic progeny in each clone reflects the successive rounds of asymmetric NB division to produce many GMCs.

To visualize mitotic cells, we made use of an antibody specific for the phosphorylated form of histone H3 (pH3). A subset of NBs and GMCs was labeled with anti-pH3 in control and miR-124 mutant clones (Fig. 3G,H), indicating that NBs and GMCs were mitotically active. Comparable observations were made for marked miR-124 mutant type II NB clones. Therefore, loss of miR-124 does not seem to alter neuronal cell fates in type I or type II larval brain NB lineages.

Although loss of miR-124 did not compromise the ability of type I NBs to undergo asymmetric division, a closer examination revealed a reduction in the number of postmitotic neurons per clone (Fig. 4A). The average number of GFP-labeled Elav-positive cells was determined for control and mutant type I NB clones. Control clones contained an average of 85 cells, versus 66 cells in miR-124 mutant clones (Fig. 4A; ^*P<0.05). To confirm that this difference was due to the absence of miR-124, a clonal rescue experiment was carried out. The MARCM system makes use of elav-Gal4 to direct UAS-GFP expression to positively label clones (note that the elav-Gal4 driver is expressed in all cells of the NB lineage, in contrast to Elav protein which is limited to postmitotic neurons). To perform genetic rescue, we co-expressed a UAS-miR-124 transgene along with UAS-GFP in control and miR-124 mutant MARCM clones. This restored mutant type I NB clone size to normal control levels (Fig. 4A; P=0.44 for control versus rescued mutant). Comparable results were obtained for type II NB lineages. These findings indicate that miR-124 is required in the NB lineage to generate the normal number of postmitotic neurons.

To determine whether this difference in clone cell number might be due to reduced proliferative activity in the miR-124 mutant clones, the number of MARCM clones containing type I NBs labeled with the mitotic marker pH3 was scored: 35% of miR-124 mutant clones had pH3-positive NBs compared with 46% of control clones (Fig. 4B; ^**P<0.01). The mitotic index was restored to normal levels in the UAS-miR-124 clonal rescue experiment (Fig. 4B; P=0.28 for control versus rescued mutant; ^*P<0.05 for mutant versus rescued mutant). A similar experiment using co-expression of UAS-p35 to block apoptosis in the miR-124 mutant MARCM clones did not restore the mitotic index. Together, these experiments suggest that reduced proliferative activity could account for the reduction in mutant clone size and suggest that miR-124 is necessary for maintaining proliferation of neuronal progenitors in the developing CNS.

Computational predictions have suggested that miR-124 targets mainly epithelial-specific or non-neuronal genes, including the glial-specific genes repo and Gliotactin (Stark et al., 2005). Intriguingly, the list of predicted targets of miR-124 includes anachronism (ana), which has been reported to encode a negative regulator of NB proliferation that is expressed specifically in a subset of glial cells (Ebens et al., 1993). As a first step to determine whether ana might be a functionally important target of miR-124, we examined ana transcript levels by qRT-PCR in RNA samples from dissected third instar brains of control and miR-124 mutants. ana transcript levels were elevated in the mutant brains and were restored to normal in the RMCE-rescued mutant (Fig. 5A; ^**P<0.01). This indicates that the endogenous ana transcript is affected in a manner consistent with regulation by miR-124.

The ana 3′ untranslated region (UTR) is predicted to contain two potential miR-124 binding sites (Fig. 5B). To address whether ana is a direct target of miR-124, we generated luciferase reporter constructs carrying the full-length ana 3′UTR or a mutant version in which three nucleotides of one of the predicted miR-124 sites were mutated to compromise pairing with the miRNA seed region (Fig. 5B, gray). Co-expression with miR-124 in S2 cells significantly reduced luciferase activity from the reporter carrying the intact target site but had little effect on the mutant form of the ana reporter (Fig. 5C; ^**P<0.01). This suggests that miR-124 can act directly via this target site to repress ana expression.

To further assess the functionality of the miR-124 target site in vivo, transgenic flies expressing the ana 3′UTR luciferase reporter were generated. Luciferase activity levels were compared in brains dissected from late third instar control larvae and miR-124 mutants. We observed elevated luciferase activity in miR-124 mutant brains carrying the intact ana reporter transgene, as compared with the level of expression of this transgene in the control brains (Fig. 5D; ^*P<0.05). The UTR reporter carrying the mutated version of miR-124 site 1 did not show statistically significant upregulation in miR-124 mutant brains. Taken together, these experiments provide evidence that miR-124 can act directly via the site identified in the 3′UTR to regulate ana mRNA levels in vivo.

As a first step to assess whether elevated ana expression contributes to the miR-124 mutant phenotype, we made use of the MARCM system to generate clones of cells overexpressing ana in the neuronal lineage. This resulted in a reduced proliferative index (Fig. 6A; ^*P<0.05). This experiment shows that upregulation of ana is sufficient to mimic the effects of the miR-124 mutant on NB proliferation.

As a more rigorous functional test of whether ana overexpression contributes to the reduced NB proliferation observed in miR-124 mutants, we asked if limiting ana levels could suppress the mutant phenotype. As a first approach, we examined proliferation of miR-124 mutant clones in flies heterozygous for a null allele of ana (Park et al., 1997). Limiting the capacity to overexpress ana in this way restored the proliferative activity of miR-124 mutant NBs to near normal levels (Fig. 6B; ^*P<0.05). Together, these experiments suggest that the elevated level of ana transcript that we observed in the miR-124 mutant brain contributes to the mutant phenotype.

The data presented thus far provide evidence that ana is a functionally significant target of miR-124 in vivo. However, miR-124 is expressed in the neuronal lineage, whereas ana has been reported to be glial-specific in expression (Ebens et al., 1993). This apparent conundrum might be resolved if ana transcript were in fact present in the neuronal lineage, perhaps at low levels.

To explore this possibility we made use of the TU-tagging method (Miller et al., 2009) to selectively label and purify RNA from neuronal cells of the NB lineage. The method is based on cell type-specific expression of the uracil phosphoribosyltransferase (UPRT) enzyme, which permits incorporation of a 4-thiouracil base into newly synthesized mRNA. elav-Gal4 was used to drive expression of UAS-UPRT in all neuronal cells of the larval brain. As a positive control for neuronal mRNA, we performed RT-PCR for elav mRNA. The glial-specific transcript repo was used as a control for glial cell mRNA. elav mRNA was recovered in the elav-Gal4 samples, whereas repo transcript was not detectably recovered (Fig. 7A). ana mRNA was detected using primers specific for the mature spliced form of the transcript, indicating that ana was synthesized in the elav-Gal4-expressing cells of wild-type neuronal lineages (Fig. 7A).

Next, we used qRT-PCR to measure ana transcript levels by TU tagging RNA from miR-124-expressing cells, using miR-124-Gal4 to direct UAS-UPRT expression. Samples were prepared from control brains and miR-124 mutant brains and PCR was normalized to the level of actin mRNA. The level of elav transcript was unchanged, but ana transcript was increased ∼2-fold in the miR-124 mutant cells (Fig. 7B). Again, repo transcript was not detected, confirming the absence of glial cells in this population. These findings indicate that (1) ana transcript is expressed in neuronal cells and (2) the level of ana transcript increased in miR-124-expressing cells in the miR-124 mutant.

Having established that ana transcript levels were elevated in the miR-124-expressing cells in the miR-124 mutant, we used the MARCM system to selectively reduce ana levels in these cells. We first monitored the efficacy of depletion of ana transcript in miR-124-expressing cells. The level of ana transcript was compared by qRT-PCR of RNA from miR-124 mutant brains expressing miR-124-Gal4 alone or together with a UAS-ana-RNAi transgene. ana transcript levels in the mutant were reduced to near normal control levels by the RNAi treatment (Fig. 7C; ^**P<0.01). MARCM clones mutant for miR-124 were produced with or without expression of the ana RNAi transgene and assayed for proliferative capacity. We found that selectively lowering ana levels in the miR-124 mutant NB clones was sufficient to restore proliferation to near normal levels (Fig. 7D; ^*P<0.05).

Given that ana and miR-124 are co-expressed in the NB lineage, we next asked whether the role of miR-124 was to tune ana expression to optimal levels in the NB. Tuning relationships have been shown previously: in miR-430 regulation of Nodal activity (Choi et al., 2007); in miR-8 regulation of Atrophin in the CNS (Karres et al., 2007); and in miR-14 regulation of Sugarbabe in insulin-producing neurosecretory cells (Varghese et al., 2010). In the latter two cases, it was shown that the level of target expression that remained after miRNA-mediated downregulation was functionally significant in the miRNA-expressing cells. To address this issue we performed an experiment similar to that in Fig. 7D, except that the MARCM system was used to express the UAS-ana-RNAi transgene along with the UAS-GFP marker in wild-type control NB clones that expressed miR-124 at normal levels. RNAi-mediated depletion of ana in these clones did not alter NB clone size (Fig. 7E). Thus, further reduction of ana levels below that achieved by miR-124 appears to be without consequence for NB proliferation. On this basis, we suggest that miR-124 reduces ana to functionally insignificant levels in the NB lineage.

Taken together, these experiments provide evidence that miR-124 acts to regulate ana expression in larval central brain NB lineages and that this regulation is important for the control of NB proliferation.

A previous report has provided evidence that ana acts on quiescent larval NBs to control their entry into the proliferative phase (Ebens et al., 1993). Loss of ana function resulted in premature onset of NB proliferation during larval development. That study reported that ana expression was specific to a subset of glial cells, which are developmentally unrelated to the glial progeny of the NB lineages. ana encodes a secreted glycoprotein, so its apparent glial-specific expression led to the proposal that Ana protein acts non-autonomously from glia to regulate NB proliferation. However, our experiments have demonstrated that ana is expressed in neuronal cells and that miR-124 acts to control ana expression in the NB lineage.

Our findings indicate that miR-124 reduces ana to functionally inconsequential levels in the central brain NB lineage, and that failure to do so impairs NB proliferation. Normal levels of ana expression in early larval brains are required to control the timing of onset of NB proliferation. It might be that the functional source of Ana protein at this stage is the previously described cortex glia. However, once the larval NBs have entered proliferation, ana is expressed in the NB lineage and its expression there must be limited by miR-124 to prevent impairment of proliferation. The levels of ana expression that are needed to impair proliferation of these cells might be much lower than those found in the cortex glia, perhaps because ana is being expressed by the cells upon which it acts within the NB lineage. Whether this reflects the non-autonomous activity of a secreted protein or an autocrine effect on the NB and GMC to affect their proliferation remains to be explored.

The finding that miR-124 activity is required in the NB lineage to support proliferation contrasts with findings from vertebrate systems that have suggested a role for miR-124 in limiting neuronal progenitor proliferation and in promoting neuronal differentiation. Several independent studies have reported that miR-124 promotes neuronal differentiation in mouse neural progenitor cells in culture by downregulating inhibitors of differentiation including the RNA splicing regulator PTB1, the SCP1 phosphatase, the transcription factor SOX9 and the ephrin B1 receptor (Makeyev et al., 2007; Visvanathan et al., 2007; Cheng et al., 2009; Arvanitis et al., 2010). Depletion of miR-124 using antisense oligonucleotides has been reported to promote proliferation and reduce differentiation of neuronal progenitors isolated from the subventricular zone (SVZ) of the mouse embryonic CNS (Cheng et al., 2009). Comparable results were obtained by injection of a pump to deliver antisense oligonucleotides to deplete miR-124 in vivo in the SVZ of the mouse brain (Cheng et al., 2009). In the chick spinal cord Visvanathan et al. found that depletion of miR-124 by antisense injection led to reduced neuronal differentiation and ectopic proliferation (Visvanathan et al., 2007). However, a contemporaneous study that undertook similar experiments did not observe these results (Cao et al., 2007). Taken together, the analysis of vertebrate miR-124 function mainly lends support to the idea that its expression in differentiating neurons acts to turn off negative regulators of differentiation and that loss of the miRNA supports proliferation of neuronal precursors.

A possible basis for the difference between our findings and those in vertebrate systems is that the expression of miR-124, although broadly CNS specific, differs in detail between insects and vertebrates. We observed expression in neural progenitors, which has not been generally reported in the vertebrate systems (Wienholds et al., 2005; Cao et al., 2007; Makeyev et al., 2007; Visvanathan et al., 2007; Cheng et al., 2009), although the presence of the miR-124 primary transcript in the NB and GMC is consistent with the expression of the miR-124-GFP reporter. In situ hybridization does not reveal significant levels of mature miR-124 in vertebrate neural progenitors, but one report using an in vivo sensor for miRNA activity has indicated that miR-124 activity can be detected in mouse neural progenitor cells (De Pietri Tonelli et al., 2006). Taken together, these findings prompt the question of whether there might be a corresponding function in the neural progenitors of the vertebrate CNS that might have been overlooked owing to low-level expression of the mature miRNA.

We have also considered the possibility that the difference between our findings and those reported using vertebrate models reflect methodological differences, i.e. the use of a genetic null mutant versus miRNA depletion using injected or transfected antisense oligonucleotides to reduce miRNA activity. Previous analysis of genetic mutants (Bushati and Cohen, 2007) has not supported the conclusions of antisense injections to deplete miRNA function in Drosophila embryos (Leaman et al., 2005). Antisense methods allow partial reduction of function and might introduce a degree of experimental variability. It will be of interest to learn whether mouse knockouts of miR-124 support the findings reported using antisense methods.

A third and perhaps more interesting possibility is that the different effects of miR-124 on neuronal progenitor proliferation reflect evolutionary divergence of miR-124 function. miR-124 has hundreds of potential targets as identified by computational prediction, by expression profiling of miRNA overexpression and depletion and by immunopurification of miRNA-containing ribonucleoproteins (Lim et al., 2005; Ruby et al., 2006; Ruby et al., 2007; Clark et al., 2010; Chi et al., 2009; Friedman et al., 2009). There is little evidence of conservation of the identified or predicted targets between insects, nematodes and vertebrates. Furthermore, each of the reports on miR-124 function in vertebrates has attributed its role to the regulation of different targets. This might reflect subtly different roles for the miRNA in different neuronal progenitor cell models in vitro, in different regions of the developing CNS or at different stages of development. Evidence has been presented for different miR-124 functions at different stages in Xenopus eye development (Qiu et al., 2009; Liu et al., 2011). It is also possible that miR-124 acts via several functionally significant targets that each serve as repressors of neuronal differentiation in vivo.

In Drosophila, we have identified ana as a target of miR-124 in vivo and provided direct genetic evidence that downregulation of ana expression in neuronal progenitors is required to support a normal level of proliferation within the larval central brain. The ana gene is not conserved beyond the Drosophila family. This leads to the intriguing proposal that miR-124 might have acquired a novel target in Drosophila, which has led to an entirely distinct function in the control of CNS proliferation from that found in vertebrates. It might also suggest that an evolutionarily ancient and presumably conserved role of miR-124 awaits discovery.

We are very grateful to Natascha Bushati for contributions to the early stages of this project. We thank James Skeath for anti-Deadpan antibody; Kah-Junn Tan for technical support; and Phing Chian Chia, Marita Buescher, David Foronda, Juan Hector Herranz Munoz and Jishy Varghese for helpful discussion.