The zinc finger transcription factor RP58 (also known as ZNF238) regulates neurogenesis of the mouse neocortex and cerebellum (Okado et al. 2009; Xiang et al. 2011; Baubet et al. 2012; Ohtaka-Maruyama et al. 2013), but its mechanism of action remains unclear. In this study, we report a cell-autonomous function for RP58 during the differentiation of embryonic cortical projection neurons via its activities as a transcriptional repressor. Disruption of RP58 expression alters the differentiation of immature neurons and impairs their migration and positioning within the mouse cerebral cortex. Loss of RP58 within the embryonic cortex also leads to elevated mRNA for Rnd2, a member of the Rnd family of atypical RhoA-like GTPase proteins important for cortical neuron migration (Heng et al. 2008). Mechanistically, RP58 represses transcription of Rnd2 via binding to a 3′-regulatory enhancer in a sequence-specific fashion. Using reporter assays, we found that RP58 repression of Rnd2 is competed by proneural basic helix–loop–helix transcriptional activators. Finally, our rescue experiments revealed that negative regulation of Rnd2 by RP58 was important for cortical cell migration in vivo. Taken together, these studies demonstrate that RP58 is a key player in the transcriptional control of cell migration in the developing cerebral cortex.
During cortical development, projection neurons are born in the germinal ventricular zone (VZ) and migrate through a transitory intermediate zone (IZ) before settling within the cortical plate (CP) where they undergo terminal differentiation (Marin and Rubenstein 2003; Merot et al. 2009). The molecular control of these events involves cell proliferation, cell cycle exit, and neural differentiation (Ge et al. 2006; Nguyen et al. 2006). However, to date, relatively few studies have explored the transition from cell cycle exit to cell migration at the molecular level; or enquired whether or not these 2 processes are coupled (Ge et al. 2006; Kawauchi et al. 2006; Nguyen et al. 2006; Buchman et al. 2010; Singh et al. 2010).
In contrast, the molecular control of proteins driving cortical neuron migration and maturation are better understood. For instance, we know how transcription factors such as Dlx1/2 (Cobos et al. 2007), as well as members of the basic helix–loop–helix (bHLH) family, including Mash1 and Neurog2 (Hand et al. 2005; Ge et al. 2006; Heng et al. 2008; Dixit et al. 2011; Pacary et al. 2011), drive the maturation of embryonic cortical neurons. From these studies, we understand that corticogenesis is carefully regulated at the transcriptional level, right from neuronal migration (Hand et al. 2005; Pacary et al. 2011) to neurodifferentiation (Cobos et al. 2007) and acquisition of glutamatergic (Parras et al. 2002; Schuurmans et al. 2004; Molyneaux et al. 2007) or γ-aminobutyric acidergic (Marin and Rubenstein 2001; Cobos et al. 2007) neuronal subidentities. Such studies demonstrate that timing, as well as gene dosage, regulates neuronal migration and terminal differentiation.
In the present work, we focus on the zinc finger repressor RP58 (also known as Znf238 or Zfp238), which has recently been found to be important for the development of multiple cell lineages within the mouse embryo (Okado et al. 2009; Yokoyama et al. 2009; Xiang et al. 2011; Baubet et al. 2012). Within the nervous system, loss of RP58 leads to abnormal neurodifferentiation and progenitor proliferation within the neocortex (Okado et al. 2009; Xiang et al. 2011) as well as the cerebellum (Okado et al. 2009; Baubet et al. 2012). Recently, it was also reported that RP58 controls cortical neurogenesis (Hirai et al. 2012) and the early steps of cell migration within the embryonic mouse cortex through direct regulation of Neurog2, a proneural bHLH protein and a key regulator of cortical projection neuron identity within the cortex (Ohtaka-Maruyama et al. 2013). However, it is likely that RP58 has multiple downstream targets to direct the migration and differentiation of cortical neurons. In this study, we report that RP58 cell autonomously influences the migration and long-term positioning of cortical neurons and directly regulates the expression of Rnd2, a member of the Rnd family of atypical RhoA-like GTPases that influences the migration of newborn cortical neurons (Hand et al. 2005; Pacary et al. 2011). We demonstrate that RP58 mediates the expression of Rnd2 through its transcriptional repressor activity on a previously characterized 3′-regulatory enhancer (Heng et al. 2008). Finally, we perform rescue experiments to demonstrate that Rnd2 lies downstream of RP58 in the regulation of cell migration, including their multipolar-to-bipolar transition within the embryonic cortex. Thus, RP58 directly influences the positioning of cerebral cortical neurons through a transcriptional regulatory pathway involving the downstream target gene Rnd2.
Materials and Methods
Animals were maintained within the animal facilities at the Howard Florey Neuroscience Institute and Monash University, with all animal procedures approved by the Animal Ethics Committees serving both institutions, as well as guidelines provided by the National Health and Medical Research Council of Australia. Genetically modified RP58 null-mutant mice are described previously (Okado et al. 2009). Mouse embryos of either sex were used for all experiments, including in utero electroporation.
In Utero Electroporation
In utero electroporation (Nguyen et al. 2006; Pacary et al. 2011) was performed as described previously, but with the following control. Given the potential variability in electroporation experiments carried out by different operators, and within different laboratories, we imposed a restriction on the experimenter for our studies (see Authors’ Contributions) and did not interchange baseline (control) conditions between experiments. Low-endotoxin plasmids were injected at 1 μg/μL, whereas siRNA duplexes (Ambion) were injected at 10 μM. Following surgery, the mice were sacrificed by cervical dislocation, and the embryonic brains dissected in cold phosphate-buffered saline (PBS) then visualized for green fluorescent protein (GFP) expression. GFP-electroporated brains were fixed overnight (4% paraformaldehyde solution in the PBS diluent) before they were subjected to 3 washes in cold PBS, permeabilized in sucrose (20%)/PBS, and then embedded in optimum cutting temperature compount (Tissue-Tek) for sectioning (16 μm) with a cryostat (Leica) for immunostaining experiments (see below). The methods for the partitioning of subregions within the embryonic cortex for cell counting have been previously described and involve a combination of immunostaining for the neuronal marker βIII-tubulin together with staining of cell nuclei with 4′,6-diamidino-2-phenylindole (DAPI; Nguyen et al. 2006). The VZ/SVZ (sub-ventricular zone) is identified as a region adjacent to the ventricular lumen of the cortex comprising densely packed nuclei and weak βIII-tubulin staining, while the IZ exhibits strong βIII-tubulin staining and sparse nuclei distribution. The CP is readily identified by strong βIII-tubulin staining and densely packed nuclei. All cell counts were performed blinded to the condition.
DNA Cloning and siRNA
The pCIG2 vector (Hand et al. 2005) was used to label cells with GFP in this study, unless otherwise stated. A human RP58 expression construct pCIG-huRP58 was cloned by inserting an EcoRI/PmeI RP58 cDNA fragment into pCIG2 digested with EcoRI and SmaI, whereas a cre recombinase expression vector (pCIG-cre) has been previously described (Pacary et al. 2011). A mouse RP58 expression construct (pCA-muRP58) was cloned into a modified adenovirus vector (pAxCAwtit), which lacks virus synthesis genes (Kanegae et al. 1995). All shRNA vectors employ the pCA-b-EGFPm5 silencer 3 backbone (Bron et al. 2007), and the vectors used in this study (Control shRNA, Rnd2shRNA#1, and Rnd2shRNA#2) have all been previously described (Heng et al. 2008; Pacary et al. 2011), with the exception of RP58shRNA which has been cloned with the following oligomers:
5′-CTTCAAGTTGTTCGGATAATTCAAGAGATTATCCGAACAACTTGAAGTTTTT-3′ and 5′-AAAAACTTCAAGTTGTTCGGATAATCTCTTGAATTATCCGAACAACTTGAAGGGCC-3′. The following pool of RP58 targeting siRNAs (Thermo Fisher) was used (sense-strand sequences):
5′-CTTCAAGTTGTTCGGATAA-3′, 5′-GGTCAAAAGTGAAGCGCTG-3′ 5′-GACAATAGGTCATAGGTCA-3′, and 5′-GTATAACTTTAACGAGTGA-3′. For conditional knockdown of Rnd2 in RP58KO cortex, LoxP-stop-LoxP sequences interrupted the U6 promoter region (upstream of the shRNA sequences that target Rnd2). The cre expression plasmid driven by the RP58 promoter (RP58-cre) has recently been described (Ohtaka-Maruyama et al. 2012). All siRNA knockdown experiments were performed in parallel with a pool of 4 nontargeting siRNAs as the negative control. The forward and reverse primers for the analysis of Rnd2 mRNA are 5′-GTGTTTGAGAACTACACTGC-3′ and 5′-GTGTTTCTGGCCGGCTAATG-3′, respectively. The forward and reverse primers for the analysis of GAPDH mRNA are 5′-AAATGGTGAAGGTCGGTGTG-3′ and 5′-TGAAGGGGTCGTTGATGG-3′, respectively.
Immunostaining and RNA In Situ Hybridization
Immunostaining on mouse brain sections was performed as described previously (Nguyen et al. 2006; Heng et al. 2008; Pacary et al. 2011) with the following primary antibodies: Mouse monoclonal (clone B01P; 1:250, Abnova) and rabbit polyclonal (1:400, Proteintech Group, Inc.) anti-RP58, rabbit anti-Ki67 (1:1000, Leica), rabbit anti-pHH3 (1:500, Millipore), mouse anti-βIII-tubulin (1:1000, Covance), rabbit (Invitrogen) and chicken (Chemicon) polyclonal anti-GFP antibodies (1:1000 and 1:700, respectively), and mouse monoclonal anti-MAP2 (1:500, Sigma). Following overnight incubation with the primary antibody, appropriate fluorescent secondary antibodies (1:800, Invitrogen) were applied before sections were counterstained with DAPI solution (in PBS) to visualize cell nuclei. In situ hybridization was performed for Rnd2 as described previously (Heng et al. 2008). Images were acquired with the following epifluorescent microscopes [Axioplan 2 and Axioimage (Zeiss), confocal microscopes (Radiance 2100 (BioRad), and Eclipse C1 (Nikon)]. Data from all experiments are expressed as mean ± standard error of the mean (SEM). An unpaired t-test was applied for statistical analyses between 2 treatments, while a 1-way analysis of variance (ANOVA) was employed to test multiple conditions, followed by an appropriate post hoc test for multiple comparisons. A detailed description of statistics and sample sizes for each experiment have been generated for this study (Supplementary Table 1).
Chromatin immunoprecipitation (ChIP) experiments with mouse embryonic forebrain tissues were performed as previously described (Heng et al. 2008) using mouse monoclonal RP58 antibody (clone M04, Novus Biologicals) and analyzed via conventional PCR.
Cell Culture, Western Blotting, and Luciferase Expression Assays
Cell lines (human embryonic kidney HEK293T and mouse embryocarcinoma P19) were maintained in Dulbecco's modified eagle medium containing 10% fetal bovine serum (heat inactivated), penicillin/streptomycin, and glutamate. Western blotting was performed using standard protocols (Heng et al. 2008), and also using the HEK293T cell line. The following primary antibodies were used for immunodetection: Mouse monoclonal anti-FLAG (1:1000, Sigma), mouse anti-actin C2 (1:1000, Santa Cruz Biotechnology); along with appropriate secondary antibodies: Goat anti-rabbit IgG (H + L) horseradish peroxidase (HRP) conjugate or goat anti-mouse IgG (H + L) HRP conjugate (1:5000, BioRad). Signal was revealed using ECL western blotting detection reagents according to the manufacturer's instructions (Amersham Biosciences). Mouse embryocarcinoma P19 cells were used for luciferase reporter assays using an expression construct harboring the 3′-Rnd2 enhancer sequence characterized in a previous study (Heng et al. 2008). The quantity of plasmid DNA was normalized for transfections with combinations of transcription factor construct in the presence of the luciferase reporter construct.
Defective Cell Proliferation and Cell Positioning Within the RP58(−/−) Mutant Cortex
RP58 is known to be widely expressed in the developing cortex, suggesting important roles during neurogenesis (Fig. 1A and Okado et al. 2009). We assessed its importance for forebrain development by analyzing the cortices of embryonic (E14.5) RP58(−/−) null-mutant mice. We found a significant increase in the expression of the proliferation markers Ki67 and pHH3, as well as a decrease in the expression of the neuronal differentiation marker βIII-tubulin compared with wild-type littermates (Fig. 1B–G). While these observations are consistent with the requirement for RP58 in neurogenesis and neuron positioning (Okado et al. 2009; Xiang et al. 2011), we uncovered further evidence that RP58 loss causes abnormalities in neuronal migration and maturation. First, we observed aberrant positioning of Ki67-expressing cells, which were located nearer to the pial surface within the RP58(−/−) null-mutant cortex compared with cells within the cortices of wild-type littermates (Fig. 1H). In addition, there was a significant increase in the fraction of Ki67-expressing cells that coexpress βIII-tubulin within the RP58(−/−) mutant cortex (Fig. 1I). These observations indicate that proliferation, neuronal differentiation, and cell positioning are disrupted within the RP58(−/−) mutant cortex.
To confirm that RP58 directly influences the expression of cell proliferation markers Ki67 and pH3 within cortical cells, we performed in utero electroporation experiments to study the consequences of RP58 knockdown in E14.5 cortical cells electroporated with RP58 siRNA and a GFP construct, and then analyzed the cortices 36 h later (Supplementary Fig. 1). As shown, knockdown of RP58 leads to a significant increase in the expression of Ki67 and pH3, and codelivery of an RP58 expression construct abrogates this effect in GFP-labeled cortical cells (Supplementary Fig. 1C–J). In addition, we observed that RP58 siRNA-treated cells were defective in their migration from the germinal VZ to the SVZ and IZ (Supplementary Fig. 1K), and were reminiscent of the defective cell positioning of cortical cells within the RP58(−/−) mutant cortex. The migration defect of RP58 siRNA-treated cells could also be corrected when an RP58 expression construct is cotransfected (Supplementary Fig. 1K). Thus, our in utero electroporation experiments show that RP58 cell autonomously mediates the expression of pH3 and Ki67 as well as mediates the migration of cortical cells from the VZ.
We next investigated whether suppression of RP58 leads to a defect or a delay in neuronal migration within the developing cortex. To study this, we performed in utero electroporation studies on E14.5 cortices. The results showed that RP58 deficiency led to the impairment of neuronal migration into the CP at E17.5 (Fig. 2A–D). Similar defects in migration were observed in RP58 siRNA-treated postmitotic neurons, which were labeled with GFP under the control of a NeuroD promoter construct (Yokota et al, 2007) (Supplementary Fig. 2C,D), and confirm an important function for RP58 in the regulation of neuronal migration. We performed immunostaining to confirm that there was no effect on the morphology of radial glia, or excessive cell death (as evaluated by immunostaining for activated Caspase 3) as a result of RP58 knockdown (Supplementary Fig. 3A,B). There was also no evidence for the ectopic expression of the marker glial fibrillary acidic protein (GFAP) following RP58 siRNA treatment (Supplementary Fig. 3D). Thus, we performed in utero electroporation with an RP58 shRNA construct to find that knockdown of RP58 resulted in a defect in the positioning of cortical neurons within the postnatal day P17 cortex (Fig. 2E,F). These results demonstrate that knockdown of RP58 leads to defective cortical cell migration, and disrupt the long-term positioning of cortical neurons within the P17 cortex.
RP58 Directly Regulates Rnd2 Expression in the Developing Cortex
The precise mechanisms for RP58 control of cortical development are not fully understood, although a recent study suggests that RP58 could influence the activity of the proneural transcription factor Neurog2 (Ohtaka-Maruyama et al. 2013). Given that Neurog2 has been shown to activate cell migration by controlling Rnd2 (Heng et al. 2008), we performed in situ hybridization experiments to explore the possibility that loss of RP58 could disrupt cortical development and, in turn, lead to altered Rnd2 expression within the cortex. Rnd2 normally exhibits a dynamic expression pattern within the developing cortex (Fig. 3A) with strong expression in cells of the IZ undergoing active migration, followed by a sharp down-regulation in expression within CP cells where they cease migration prior to terminal neurodifferentiation. In contrast, Rnd2 expression was significantly increased within the RP58(−/−) mutant cortex (Fig. 3C), including cells of the CP marked by expression of the microtubule-associated protein (Map2) (Fig. 3B). This suggests that Rnd2 expression within the cortex might be regulated by RP58.
Previous studies have reported that RP58 is a DNA-binding transcriptional repressor (Aoki et al. 1998; Xiang et al. 2011), which raised the possibility that RP58 could directly control Rnd2 expression through a putative regulatory enhancer within Rnd2. We had previously characterized an enhancer element (known as the Rnd2 3′enhancer), which was important for regulating Rnd2 expression in the cortex (Heng et al. 2008), and was also found to comprise an RP58-binding site (Quandt et al. 1995) (data not shown). We performed ChIP experiments with embryonic E14.5 cortical tissue and confirmed that RP58 binds to this enhancer element in vivo (Fig. 3D). In control experiments, we did not detect the binding of RP58 to a nonrelevant sequence within the open-reading frame of the Rnd2 gene, while the binding of a previously characterized RP58-binding site on the Id2 gene (Yokoyama et al. 2009) was detected in our assay.
We conducted luciferase reporter assays using P19 embryocarcinoma cells to evaluate the transcriptional activity for RP58 through this enhancer element (Fig. 3E). Interestingly, RP58 is known to bind sequences comprising E-boxes (Aoki et al. 1998), which are hexanucleotide-binding sites recognized by bHLH transcription factors such as Neurog2, NeuroD1, and NeuroD2 (Bertrand et al. 2002). Since 2 E-box sequences are found on this enhancer, we carried out assays to determine the importance of the E-boxes for RP58-mediated reporter activity. As shown, while RP58 can suppress reporter activity through this enhancer construct, mutations to both E-boxes abolished the repressor function of RP58 (Fig. 3E). In addition, while the proneural bHLH protein Neurog2, as well as the bHLH neurodifferentiation factors such as NeuroD1 and NeuroD2, can bind and activate this enhancer (Heng et al. 2008), we found that RP58 antagonizes the stimulatory function of all 3 bHLH proteins in our assay (Fig. 3F). In summary, our experiments point to the transcriptional repression of Rnd2 by RP58, most likely by binding to the E-boxes present within this 3′-regulatory enhancer.
RP58 Mediates the Migration of Cortical Neurons, Including Their Morphological Transition Within the Embryonic E17.5 Cortex
Several transcription factors, including Neurog2 and COUP-TFI, are known to regulate the morphology and migration of neurons (Hand et al. 2005; Heng et al. 2008; Alfano et al. 2011) and so we investigated if RP58 was also an important transcription factor regulating these features of immature cortical neurons. To this end, we electroporated RP58 siRNAs together with a GFP construct in the E14.5 cortex, then dissociated cortical cells, and analyzed their morphologies (revealed by GFP activity) following 3 days of culture in vitro (Fig. 4A–E). Upon analysis, we observed a significant increase in the proportion of multipolar-shaped neurons compared with control treatment (Fig. 4D). Furthermore, these neurons extended much shorter neurites compared with control (Fig. 4E). These observations indicate that RP58 influences the morphological differentiation of embryonic cortical neurons.
As projection neurons are born within the embryonic cortex, they depart the germinal VZ/SVZ as bipolar-shaped cells before they transit within the IZ as multipolar-shaped neurons and then finally adopt a bipolar shape and enter the CP (Tabata and Nakajima 2003; Noctor et al. 2004). Since several genes, including Rnd2, are known to control the morphology of immature cortical neurons in vitro and in vivo (LoTurco and Bai 2006; Heng et al. 2008), we next analyzed the morphologies of RP58-deficient neurons within electroporated E17.5 brains to determine whether there was an underlying cell morphological basis to their migration defect (Fig. 4F–L). In our electroporation experiments, we can chart the multipolar-to-bipolar migration of neurons from the IZ to the CP by performing confocal microscopy to analyze their morphologies within each cortical subcompartment (Heng et al. 2008; Alfano et al. 2011; Pacary et al. 2011). In control-treated brain sections, we observed a decreasing proportion of multipolar-shaped neurons from the lower IZ through the lower CP, and a concomitant increase in the proportion of bipolar-shaped neurons within these compartments (Fig. 4J). Using this morphological assay, we then investigated the consequences of RP58 siRNA treatment on the morphology of GFP-labeled cells within these subcompartments. Compared with control samples, a higher proportion of multipolar neurons lodged within the upper IZ (uIZ) of RP58 siRNA-treated samples (31.8 ± 0.3% treated vs. 19.3 ± 0.3% controls) (Fig. 4J,K). In addition, treated neurons had more processes that were also significantly shorter in length compared with controls (Fig. 4F,G,I). These findings strongly suggest that RP58 siRNA-treated cells were impaired in their ability to undergo a multipolar-to-bipolar conversion during their migration from the IZ to the CP.
Having observed that the morphology of cortical cells within the IZ was affected by RP58 siRNA treatment within the brains of electroporated E17.5 embryos, we investigated whether bipolar-shaped neurons of the CP were also affected. We found that RP58 siRNA-treated CP neurons had shorter leading processes compared with control-treated cells (Fig. 4I; 18.6 ± 0.9 μm on RP58 siRNA-treated CP neurons vs. 27 ± 0.9 μm in control CP neurons), and with almost twice the fraction of RP58 siRNA-treated CP neurons comprising a leading process which measured <20 μm in length. Interestingly, the abnormal morphology of RP58 siRNA-treated CP neurons was reminiscent of the morphology of CP neurons, which overexpress Rnd2 (Heng et al. 2008) and also harbor short-leading processes compared with control treatment (Fig. 4I; 16.2 ± 0.5 μm on Rnd2-overexpressing CP neurons vs. 27 ± 0.9 μm in control CP neurons). Therefore, RP58 is important for regulating the in vivo morphology of embryonic cortical neurons within the embryonic E17.5 cortex.
Rnd2 Lies Downstream of RP58 for the Regulation of Cell Migration In Vivo
RP58(−/−) mutant cortical cells are defective in their migration (Fig. 2C,D) and have elevated Rnd2 expression, and our experiments, so far, suggest that this is likely to be responsible for their abnormal migration (Fig. 3). To test this hypothesis, we set out to rescue the migration defect in RP58-deficient neurons by suppressing Rnd2 expression in vivo. Suppression was performed using RP58 siRNAs together with expression constructs encoding Rnd2 shRNAs introduced at E14.5 and analyzed at E17.5 (Fig. 5). To achieve dosage-related knockdown, we performed 2 independent rescue experiments using 2 Rnd2 shRNA constructs with different levels of Rnd2 knockdown (Supplementary Fig. 4A), and different capacities for suppressing cortical cell migration (Supplementary Fig. 4C–F). We coelectroporated RP58 siRNA-treated cells with a “weak” Rnd2 shRNA (named as Rnd2shRNA1), or a “strong” Rnd2 shRNA (named as Rnd2shRNA2), and examined the effect on cell migration (Fig. 5). As shown, while RP58 siRNA-treated cells exhibited a migration defect in the CP compared with control treatment, codelivery of the Rnd2shRNA1 construct capable of “weak” RNAi activity was able to rescue their migration, albeit not to wild-type levels within the CP. In contrast, codelivery of a “strong” Rnd2shRNA2 vector in RP58 siRNA-treated cells not only failed to rescue their migration, but rather increased the severity of the CP migration defect within the E17.5 cortex (Fig. 5F and Supplementary Fig. 5). These results would suggest that severe deficiencies in both RP58 and Rnd2 produced accretion effects on the migration of cortical neurons. Hence, these data demonstrate that proper migration of neurons requires both appropriate and balanced levels of Rnd2.
To further test the above contention in RP58 knockout (KO) mice, we evaluated the capacity for Rnd2 RNAi to restore the migration of RP58(−/−) neurons (Supplementary Fig. 6). We performed rescue experiments using conditional shRNA vectors encoding Rnd2shRNA1 and Rnd2shRNA2 sequences, which express these hairpin sequences for suppression of Rnd2 expression only in the presence of cre recombinase (Supplementary Fig. 4B). These constructs were coelectroporated with a cre recombinase construct under the control of the RP58 enhancer sequence (Ohtaka-Maruyama et al. 2012) to observe the migration of RP58(−/−) mutant cortical cells at E14.5, and sampled 3 days post manipulation (E17.5). As shown, RP58(−/−) mutant cells exhibit a severe defect in their migration compared with wild-type littermates (Supplementary Fig. 6B,E). Surprisingly, treatment with Rnd2shRNA1 increased the migration of RP58 KO neurons into the CP, although their positioning within the CP was not restored to wild-type levels (Supplementary Fig. 6A–F). In contrast, strong knockdown of Rnd2 in RP58(−/−) mutant cells electroporated with the Rnd2shRNA2 construct failed to evoke a migration rescue, which is consistent with our rescue experiments with RP58 siRNA-treated cells (Fig. 5). Hence, these results strongly support our hypothesis that Rnd2 levels are elevated in RP58-deficient cells, and that RP58 balances the appropriate levels of Rnd2 expression that must be achieved for the proper migration of immature cortical neurons.
The Morphological Defects Observed in RP58-Deficient Neurons Is Corrected by Attenuating Rnd2 Expression
Having found that suppression of Rnd2 with Rnd2shRNA1 construct increases the migration of RP58-deficient neurons, we further analyzed the morphologies of the neurons in these rescue experiments in the E17.5 cortex (Fig. 6). The results demonstrate a significant increase in the proportion of multipolar-shaped neurons throughout the IZ and CP following RP58 siRNA treatment (Fig. 6A,B), and indicate that RP58-deficient neurons are defective in their multipolar-to-bipolar transition from the IZ to the CP. In contrast, treatment of RP58-deficient neurons with Rnd2shRNA1 (weak activity) ameliorated this morphological defect (Fig. 6A–C), which was accompanied by a significant increase in the proportion of bipolar-shaped cells in the medial and uIZ. In addition to a restoration of neuronal morphologies within the IZ and CP, the lengths of the leading process of RP58-deficient CP neurons were also increased on treatment with Rnd2shRNA1, and more closely resembling control neurons (Fig. 6E,F). In contrast, the proportion of multipolar (MP) neurons within the uIZ of RP58-deficient cells electroporated with Rnd2shRNA2 (strong activity) remained comparable with levels detected in RP58 siRNA-treated cortices (Fig. 6B–D). Within RP58 siRNA/Rnd2shRNA2-treated cortices, the leading process of GFP-labeled CP neurons was increased compared with control treatment, and these neurons also exhibited increased branching (Fig. 6H). Similar findings were made following attenuation of Rnd2 expression in RP58(−/−) cells within the cortices of KO mice (Supplementary Fig. 6G–O). In summary, these results demonstrate that Rnd2 lies downstream of RP58 for the regulation of morphological differentiation of neurons within the embryonic cortex which, in turn, directly impacts on their MP-to-BP transition as they reach the CP. These multiple strands of evidence indicate that the transcriptional control of Rnd2 by RP58 has strong and immediate consequences for the shapes and positions of migrating cortical neurons.
Transcription factors play critical roles in regulating gene expression for the orderly production and maturation of cortical neurons within the developing cerebral cortex (Schuurmans and Guillemot 2002; Molyneaux et al. 2007), and our present study provides a significant role for RP58 in these processes. Previous studies have reported that loss of RP58 leads to defective neurogenesis of the cortex and cerebellum, features which are associated with changes in the expression of proneural genes, including members of the Neurogenin and NeuroD families of bHLH transcription factors (Okado et al. 2009; Xiang et al. 2011; Baubet et al. 2012; Ohtaka-Maruyama et al. 2013). However, it is unclear whether or not RP58 can directly influence cortical neuron development, given its reported expression within neurons of the postnatal cortex (Ohtaka-Maruyama et al. 2007). The current study has provided renewed understanding in the following ways. First, we show that RP58 directly promotes the migration and positioning of cortical neurons during development. Secondly, we supply a mechanism by showing that RP58 regulates the expression of a downstream target gene, Rnd2 through direct binding of a recognition motif present in its 3′-enhancer. Thirdly, we find that the transcriptional repressor functions for RP58 antagonize its transcriptional activator functions (by proneural bHLH proteins such as Neurog2, NeuroD1, and NeuroD2). Fourthly, we show that RP58 directly controls Rnd2 levels for the morphological transition of neurons in the course of their migration within the developing cortex, including their MP-to-BP transition.
It has been reported that RP58 promotes cortical neurogenesis through direct regulation of the proneural bHLH gene Neurog2, and that a RP58–Neurog2 signaling pathway is important for the production of new neurons within the embryonic cortex, as well as their cell migration (Ohtaka-Maruyama et al. 2013). However, it is unclear if the RP58–Neurog2 axis is alone responsible for all the observed phenotypes, or if additional factors are at play. Our results implicate Rnd2 as a key downstream effector of RP58 in its regulation of cortical neuron differentiation, including their morphological differentiation and cell migration. Our studies provide evidence for a transcriptional role for RP58 in regulating Rnd2 levels for cortical neuron development. However, as previously reported, RP58 also regulates Neurog2 expression (Ohtaka-Maruyama et al. 2013), and Neurog2 stimulates Rnd2 expression (Heng et al. 2008), thus we propose that RP58 influences cell migration directly through an RP58-Rnd2 pathway, as well as through an indirect, RP58–Neurog2 pathway within the embryonic cortex (Supplementary Fig. 7A).
In rescue experiments, we found that Rnd2shRNA1 improved the migration of RP58-deficient cells, which we interpret as a restoration of Rnd2 inhibition in these cells. However, we did not observe a full restoration of their migration within the CP, despite improvement in certain aspects of their morphological differentiation, including the length of their leading process. One interpretation for this finding is that levels of Rnd2 must be finely balanced in order for cortical cells to migrate and differentiate within the E17.5 cortex. Indeed, Rnd2 is regulated by several transcription factors within the developing cortex, including COUP-TFI (Alfano et al. 2011), Neurog2 (Heng et al. 2008), and RP58. Nevertheless, we cannot rule out the possibility that other factors are also important for their “intracortical” positioning within the CP, and these may also be regulated by RP58 for their in vivo migration.
Interestingly, we found that the Neurog2-type E-box-binding sites were important for the regulatory activity of RP58 on a previously characterized Rnd2 3′-enhancer (Heng et al. 2008). We also found that RP58 competes with proneural bHLH transcriptional activators, such as Neurog2, NeuroD1, and NeuroD2, on the Rnd2 3′-enhancer. One possible scenario is that RP58 regulates Rnd2 levels in direct competition with bHLH transactivators (such as Neurogenins and NeuroD), both spatially and temporally, in the developing cortical landscape (Supplementary Fig. 7A,B). In addition, recent findings have revealed that RP58, NeuroD1, and Rnd2 are direct target genes of Neurog2 (Schuurmans et al. 2004; Gohlke et al. 2008; Heng et al. 2008), hence RP58 may form part of a signaling network through which Neurog2 specifies the migratory properties of newborn cortical projection neurons within the embryonic cerebral cortex. Indeed, RP58 may reciprocally regulate the expression of Neurog2 and Neurod1 as neurons mature within the cortex (Xiang et al. 2011; Ohtaka-Maruyama et al. 2013), and so it will be important to delineate the precise role for RP58 in both the generic properties of cortical neuron differentiation, as well as within distinct subprograms that control aspects of their differentiation, such as cell migration.
The present study has implications for our understanding of human cortical development. Failures in normal cortical development can result in severe neurological defects, including intellectual disability and epilepsy (Leventer et al. 2008). Interestingly, RP58 (also known as ZNF238) was identified as a gene that is deleted in patients diagnosed with Terminal 1q deletion syndrome, a genetic disorder characterized by microcephaly, and frequently associated with defects in the maturation of callosal projection neurons (van Bon et al. 2008; Perlman et al. 2013). In light of the current findings for RP58 in the regulation of cortical progenitor proliferation and neurodifferentiation (including their migration), it is likely that a similar role for RP58 exists in human corticogenesis.
J.I-T.H. and F.G. conceived this study, and all experiments were performed by J.I-T.H, Z.Q. (under guidance of S.-S.T.), C.O.-M. (under guidance of H.O.), and D.C. J.I.-T.H performed the migration studies using siRNAs on wild-type mice, whereas Z.Q. performed the primary neuronal culture experiments. C.O.-M. performed the migration studies with the RP58 null-mutant mouse line (generated by H.O. and M.K.). J.I.-T.H. wrote the manuscript with S.-S.T. and F.G. All authors consulted and commented on the manuscript.
This work was supported by grants from Project and Program grants from the National Health and Medical Research Council (NHMRC) and the Victorian Government through the Operational Infrastructure Scheme. The Australian Regenerative Medicine Institute is supported by grants from the State Government of Victoria and the Australian Government. J.I-T.H is a recipient of a Career Development Fellowship (ID: 1011505) from the NHMRC (Australia). This work was supported by research grants in the natural sciences from the Mitsubishi Foundation, grant-in-aid for Scientific Research (C) to C.O-M (#20570172), and grant-in-aid for Scientific Research on Innovative Areas “Neural Diversity and Neocortical Organization” from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan.
We thank Dr Hidenori Tabata (Keio University) for helpful discussions. Conflict of Interest: None declared.
- cerebral cortex
- cell motility
- guanosine triphosphate phosphohydrolases
- helix-loop-helix motifs
- rhoa gtp-binding protein
- rna, messenger
- zinc fingers
- transcription factor
- transcriptional activation
- transcriptional repression
- transcriptional control
- rna, small interfering
- neural development
- rheumatoid neutrophilic dermatosis
- enhancer of transcription
- neuron migration
- transcription coactivator
- projection neuron