During embryonic development of the mammalian cerebral cortex, postmitotic cortical neurons migrate radially from the ventricular zone to the cortical plate. Proper migration involves the correct orientation of migrating neurons and the transition from a multipolar to a mature bipolar morphology. Herein, we report that the 2 isoforms of Myosin-10 (Myo10) play distinct roles in the regulation of radial migration in the mouse cortex. We show that the full-length Myo10 (fMyo10) isoform is located in deeper layers of the cortex and is involved in establishing proper migration orientation. We also demonstrate that fMyo10-dependent orientation of radial migration is mediated at least in part by the netrin-1 receptor deleted in colorectal cancer. Moreover, we show that the headless Myo10 (hMyo10) isoform is required for the transition from multipolar to bipolar morphologies in the intermediate zone. Our study reveals divergent functions for the 2 Myo10 isoforms in controlling both the direction of migration and neuronal morphogenesis during radial cortical neuronal migration.
During embryonic development of the mammalian cortex, cortical neurons originating from the ventricular zone (VZ) and subventricular zone (SVZ) migrate radially along the process fibers of radial glia through the intermediate zone (IZ) to their appropriate location within the developing cortical plate (CP; Rakic et al. 1974). Impairment or failure of cortical neuron migration due to genetic and environmental factors that disturb motility could possibly lead to a variety of cortical abnormalities in human such as periventricular heterotopia, mental retardation, seizure, and schizophrenia (Gleeson and Walsh 2000; McManus and Golden 2005; Sarkisian et al. 2006; Ayala et al. 2007; Ferland et al. 2009). Postmitotic neurons possess a multipolar morphology as they migrate through the SVZ and lower IZ (loIZ; Kriegstein and Noctor 2004; Liu et al. 2012). Upon reaching the upper IZ (upIZ), they adopt their mature bipolar morphology and start their radial migration (Tabata and Nakajima 2003; LoTurco and Bai 2006). The different neuronal migratory steps involved in cortical development entail significant cell shape remodeling changes, which are largely dependent on cytoskeleton and contractile proteins, cytoskeleton-associated proteins, and membrane-bound proteins (Rivas and Hatten 1995; Rakic et al. 1996; da Silva and Dotti 2002; Gupta et al. 2002; LoTurco and Bai 2006). The specific molecular machinery conferring motility to migrating neurons, however, remains poorly understood.
Myosins are involved in numerous neural development processes, including neurite outgrowth and axon pathfinding (Brown and Bridgman 2004). Myosin-10 (Myo10), an untraditional myosin family member, is an important cytoskeletal regulator, which interacts directly with both microfilaments and microtubules to regulate phagocytosis, mitosis, and endothelial cell migration (Cox et al. 2002; Pi et al. 2007; Woolner et al. 2008). There are 2 isoforms of Myo10 in the brain: full-length Myo10 (fMyo10) and headless Myo10 (hMyo10; Sousa et al. 2006). hMyo10, which lacks most of the myosin motor domain, is proposed to act as an endogenous dominant negative form by inhibiting the function of fMyo10 (Sousa et al. 2006; Wang et al. 2009; Raines et al. 2012). Immunoblots of brain lysates reveal that fMyo10 and hMyo10 are both highly expressed as early as embryonic day 7 (E7) and persist throughout the following developmental period, but are rapidly down-regulated after birth (Berg et al. 2000; Yonezawa et al. 2000). Recent studies have also shown the involvement of Myo10 in the migration of cranial neural crest cells and gonadotropin-releasing hormone (GnRH)-expressing neurons (Hwang et al. 2009; Nie et al. 2009; Wang et al. 2009). Several proteins such as drosophila enabled/vasodilator-stimulated phosphoprotein and deleted in colorectal cancer (DCC), which are important for the initiation of neuritogenesis and axon guidance, have been reported to be transported by Myo10 in cortical neurons (Tokuo and Ikebe 2004; Withee et al. 2004; Dent et al. 2007; Zhu et al. 2007; Bradford et al. 2009). In addition, Myo10 mRNA is 7-fold up-regulated in the dorsal root ganglia after sciatic axotomy (Tanabe et al. 2003), and variable deletions of the short arm of chromosome 5, where Myo10 gene is located in human, could be involved in the genetic Cri du chat syndrome (Overhauser et al. 1994). However, little information is available to show the role of Myo10 in mammalian brain development.
In this study, we show that Myo10 is expressed in the developing cortex, and its 2 isoforms have distinct expression patterns. Knockdown of Myo10 expression via in utero electroporation caused migration defects in the embryonic cortex. Furthermore, we demonstrate that fMyo10 is involved in establishing the proper orientation of migrating neurons, possibly via the regulation of DCC, whereas hMyo10 is critical for the morphological transition of developing cortical neurons from their immature multipolar to their mature bipolar morphologies. Taken together, our data suggest that Myo10 acts as a regulator of multiple steps of cortical neuron migration and morphogenesis in the developing cortex.
Materials and Methods
C57/B6 background mice were handled in accordance with the guidelines from the Institutional Animal Care and Use Committees of Northeast Normal University, China. Females maintained on a 12-h light/dark cycle were bred overnight with males. Noon following breeding is considered as E0.5.
Primary antibodies used included rabbit anti-Myo10 (Zhu et al. 2007), mouse anti-Tuj1 (Abcam), mouse anti-microtubule-associated protein 2 (Sigma), mouse anti-neuronal nuclei (Abcam), rabbit and mouse anti-green fluorescent protein (GFP) (Invitrogen), rabbit anti-Ki67 (Abcam), mouse anti-Nestin (Abcam), rabbit anti-Pax6 (Convance), mouse anti-beta actin (Abcam), and rabbit anti-DCC (Santa Cruz). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (Roche).
To construct short hairpin RNA (shRNA) vectors, oligonucleotides targeting the mouse fMyo10 (ENSMUST00000110457) 5′ untranslated region (UTR; 5′-CTTGGAGGAAGGAGAGACA-3′), the mouse hMyo10 (ENSMUST00000022882) 5′UTR (5′-GGCACCTATGGTCAATGTA- 3′), and the mouse DCC (ENSMUST00000114943; shRNA #1: 5′-GCAATTTGCTCATCTCTAA-3′, shRNA #2: 5′-GCATACCAATTATCCATAA-3′) were generated using BLOCK-iT™ RNAi Designer (Invitrogen) and inserted into the pSuper vector. The nonsilencing control shRNA used in the experiments contains no homology to the known mammalian genes.
Brains were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) overnight, then placed in 35% sucrose in phosphate buffer saline (pH 7.4) overnight before being embedded in O.C.T. compound (Tissue-Tek), and frozen for sectioning with a Leica cryostat. Immunostaining was performed with standard protocols involving overnight incubation with primary antibodies at 4°C and incubation with appropriate fluorescent secondary antibodies for 1.5 h at room temperature. Citrate antigen retrieval was performed prior to the blocking step.
In Situ Hybridization
In situ hybridization was performed with digoxigenin-labeled riboprobes, as described previously (Ding et al. 2003). Antisense and sense probes targeting the motor region of mouse fMyo10 (∼700 bp) and the 5′UTR region of hMyo10 (∼400 bp) were generated from cDNA, cloned into pGEM-T vector (Promega), and synthesized by T7 and Sp6 polymerase.
In Utero Electroporation
Pregnant mice were anesthetized and their uterine horns were exposed with a midline laparotomy incision. Embryos were removed and carefully placed on humidified gauze pads. Plasmid DNA plus 0.01% Fast Green (Fluka) was injected into the lateral ventricles of the embryonic brain with a glass micropipette. A volume of 2 μL of shRNA plasmids (2 μg/μL) or expression constructs (2 μg/μL) were coinjected with the cytomegalovirus (CMV) early enhancer element and chicken beta-actin promoter-enhanced green fluorescent protein (EGFP)-expressing plasmids. For rescue experiments, expression constructs were coinjected with shRNA and CAG-EGFP plasmids. For electroporation, 5 × 50 ms, 37 V square pulses separated by 950 ms intervals were delivered with forceps-type electrodes connected to an ECM 830 electroporator (BTX Harvard Apparatus). The uterus was then replaced into the abdominal cavity, and the abdomen wall and skin were sutured using the surgical needle and thread. The whole procedure was completed within 40 min. The pregnant mouse was warmed in an incubator until it became conscious, and embryos were allowed developed in utero for the time indicated.
Time-lapse Imaging in Cortical Slices
In utero electroporation was performed as described above at E15.5. Two days after electroporation, embryonic brains were dissected out in cold artificial cerebrospinal fluid. Brain slices (300 μm thick) were sectioned with the Leica Vibratome VT1000. To visualize neuronal migration, slices were transferred onto Millicell inserts (Millipore) in Neurobasal medium (Invitrogen) containing 2% B-27 supplement, 2 mm l-glutamine, and penicillin/streptomycin (50 U/50 μg/mL). The glass-bottomed dish was then fitted into a temperature-controlled chamber on the microscope stage for 15 h at 37°C under 5% CO2 air atmosphere. Live-cell imaging was done using the Olympus FV1000 Viewer laser scanning confocal microscope.
GnRH-expressing neuronal cell line was cultured in Dulbecco's modified Eagle medium supplemented with newborn calf serum and transfected with shRNA plasmids using the Amaxa electroporation protocol.
Both pGL3-fM-5′UTR and pGL3-hM-5′UTR were constructed by cloning the 5′UTR region of fMyo10 and hMyo10 into pGL3-basic vector, respectively. Assays were performed using the instructions provided with the Dual-Luciferase® Reporter Assay System (Promega).
Microscope Image Acquisition and Statistical Analysis
Brain sections through the somatosensory cortex were observed under a Nikon Eclipse TE2000-U and Olympus FV1000 Viewer laser scanning confocal microscope with FV10-ASW 1.7 software (Olympus), where only the brightness, contrast, and color balance were optimized. Cortical subregions were identified based on cell density using DAPI staining. The numbers of misorientated, multipolar or bipolar cells were counted in the upIZ– lower CP (loCP) using the ImageJ software. Misorientated cells were defined as cells with a leading process oriented at least 60 degree angle off the normal pial surface-directed radial direction (measured with ImageJ software); multipolar cells were defined as cells with more than 3 processes. Statistical analysis were performed using either unpaired 2-tailed Student's t-test or 1-way analysis of variance (ANOVA) followed by a protected least significant difference Fisher post hoc test for multiple comparisons (OriginPro 8 SR1 and MS Office). Data presented as mean ± standard error of the mean (SEM).
Myo10 Is Expressed in the Developing Cortex
To obtain information about the role of Myo10 in cortical development, we stained embryonic mouse brain sections with the antibody that recognizes both fMyo10 and hMyo10 (Zhu et al. 2007), as well as with an antibody that labels Tuj1, a pan-neuronal marker. Myo10 was expressed widely in E15.5 cerebral cortex with a lower expression level at the VZ/SVZ and a higher expression level in the IZ where it overlapped with Tuj1 (Fig. 1a). By using in utero electroporation, we labeled cortical pyramidal neurons generated from the dorsolateral VZ of the forebrain with GFP (Zhang et al. 2012) and found that Myo10 was expressed in these neurons (Fig. 1b). Western blotting of embryonic mouse brain lysates revealed that hMyo10 is expressed at a higher level than fMyo10 at E16.5 (Fig. 1c). To analyze cellular expression patterns of fMyo10 and hMyo10 in the developing cortex, we first mapped out fMyo10 and hMyo10 mRNA sequences to identify differences. As previously noted in a study of human Myo10 (Sousa et al. 2006), besides the fMyo10-unique motor domain sequence, the 2 mouse isoforms have conserved tail domains with completely different 5′ untranslated regions (5′UTRs). For this reason, motor domain and 5′UTR-targeting digoxigenin-labeled RNA probes were designed to target fMyo10 and hMyo10 mRNAs (Fig. 1d). Being consistent with Myo10 immunoreactivity, in situ hybridization of E16.5 mouse cerebral cortices showed that fMyo10 mRNA was located in the region covering the VZ/SVZ and the lower part of the IZ, whereas hMyo10 mRNA was expressed from the VZ to the entire IZ (Fig. 1e). These results indicate that the 2 Myo10 isoforms are differentially expressed in the developing cortex.
Loss of Myo10 Impairs Cortical Neuron Migration
To understand how Myo10 contributes to cortical development, developing somatosensory cortical neurons in E15.5 mouse embryos were electroporated with Myo10 shRNA; its efficacy and specificity in knocking down Myo10 expression have been verified in GnRH-expressing neurons and cortical neurons (Zhu et al. 2007; Yu et al. 2012). A scrambled version of the Myo10 shRNA sequence was used as a control, and an EGFP expression plasmid was cotransfected with both shRNA constructs to clearly identify shRNA-transfected neurons. The distribution of GFP+ cells was analyzed 3 days after electroporation at E18.5, and the cortex was regionally divided into the VZ/SVZ, IZ, loCP, and upper CP (upCP). Intensity of Myo10 immunofluorescence was reduced in cortical neurons expressing Myo10 shRNA compared with control neurons (Fig. 1b). There was an obvious reduction of GFP-labeled neurons in the CP and a concomitant increase in the number of neurons in the VZ/SVZ–IZ of hemispheres transfected with Myo10 shRNA compared with brains electroporated with control plasmids (Fig. 2a,b), suggesting that migration was hindered by Myo10 knockdown. In support of this, we observed fewer GFP-labeled MAP2+ neurons in the Myo10-deficient CP (Fig. 2c,d). In addition, when cells were electroporated at E14.5, a similar migration defect was observed in cortices examined at E18.5, and neurons that reached the outermost region of the CP were unable to form a normal apical dendritic tree, as shown by the reductions in dendrite length and branch numbers (Supplementary Fig. S1). Although the migratory capacity of Myo10-deficient cortical neurons appeared to be severely compromised, no significant effects were observed on proliferating Ki67+ cells, Pax6+ progenitor cells, or Nestin+ radial glia fibers in E18.5 cortices (Supplementary Fig. S2). Taken together, these results indicate that Myo10 is required for cortical neuronal migration.
fMyo10 and hMyo10 Are Required for Different Phases of Migration
Having found that knocking down Myo10 impairs neuronal migration in the developing cortex, we set out to examine the contribution of individual Myo10 isoforms. Silencing plasmids were designed to specifically knock down fMyo10 or hMyo10 by targeting the unique sequences in the 5′UTR of each mRNA (Supplementary Fig. S3). fMyo10 and hMyo10 shRNAs were then transfected separately via in utero electroporation into the E15.5 cortex. Similar to what we found after silencing Myo10, we observed a reduction in the number of GFP+ cells in the E18.5 CP of cortices transfected with either fMyo10 or hMyo10 shRNAs (Fig. 3). Interestingly, however, we also observed a striking increase in the number of GFP+ neurons in the VZ/SVZ of fMyo10 shRNA-transfected, but not hMyo10 shRNA-transfected, cortices (Fig. 3a,b). In addition, there was an obvious increase in the fraction of cells remaining in the IZ after expressing hMyo10 shRNA, but not after expressing fMyo10 shRNA (Fig. 3a,c). We therefore reasoned that the 2 Myo10 isoforms have divergent functions in cortical neuronal migration.
The fMyo10-derived transgene Myo10ΔMo lacks the motor domain, but retains all the same functional domains as hMyo10 (Supplementary Fig. S3a). It has been reported that Myo10ΔMo is capable of antagonizing fMyo10 function, by acting as a dominant negative (Zhu et al. 2007), and can also serve as an hMyo10 analog (Yu et al. 2012). We thus investigated whether coexpressing Myo10ΔMo could affect the migratory defects caused by either fMyo10 shRNA or hMyo10 shRNA. In cortices cotransfected with fMyo10 shRNA and Myo10ΔMo, the proportion of neurons retarded in the VZ/SVZ was reduced compared with cortices transfected with fMyo10 shRNA alone (Fig. 3a,b). In contrast, hMyo10 shRNA plus Myo10ΔMo cotransfected cortices showed a significant increase in the number of neurons in the VZ/SVZ relative to cortices transfected with hMyo10 shRNA alone, and the proportion of neurons in the IZ was reduced near control levels (Fig. 3a,c). Taken together, these results suggest that both fMyo10 and hMyo10 play different roles in regulating cortical neuron migration.
fMyo10 Regulates Neuronal Orientation, Whereas hMyo10 Regulates Neuronal Morphology
Early postmitotic neurons in the SVZ and loIZ adopt a multipolar morphology, and during this stage their locomotion appears random and meandering. Thereafter, they develop a bipolar shape in the upIZ and initiate outward-directed radial migration along radial fibers that extend to the pial surface (Tabata and Nakajima 2003; LoTurco and Bai 2006). To explore the possible cellular mechanism underlying the migratory defects observed in Myo10-deficient cortices, we analyzed more closely the migratory behavior and morphological features of shRNA-transfected neurons in the upIZ and loCP. We found that the proportion of multipolar neurons at E18.5 was increased after knocking down Myo10, fMyo10, or hMyo10, when compared with controls (Fig. 4a,b,e). However, the effect was much more dramatic in the hMyo10 shRNA group, than in the other treatment groups (Fig. 4e). These results suggested that hMyo10 plays a dominant role in regulating the morphogenesis that normally occurs in cortical neurons in the upIZ.
To assess the migratory behavior of transfected neurons, we established the following criteria: Neurons with a leading process pointing toward the pia mater were considered to be migrating radially, whereas those with leading processes that deviated from the normal radial trajectory by more than 60° were considered to be nonradially migrating. Compared with controls, the population of neurons migrating radially in the upIZ and loCP was not different in the hMyo10 shRNA group, but was greatly increased in both the Myo10 and fMyo10 shRNA groups (Fig. 4a,b,d). To further investigate this phenotype, we coelectroporated the different Myo10 shRNAs with centrin-red fluorescent protein (RFP), which labels the centrosome to reveal the direction of migration (Sapir et al. 2008). In control neurons, centrosomes were positioned at the base of the leading process extending toward the pia mater, thus indicating that these neurons were migrating radially (Fig. 4c). In contrast, centrosome positioning in Myo10 or fMyo10 shRNA-expressing neurons was highly variable, with some neurons pointing toward the VZ, away from the direction of radial migration (Fig. 4b). Importantly, this migratory defect phenotype was not observed in hMyo10 shRNA-expressing neurons (Fig. 4b), suggesting that fMyo10 is involved in regulating the radially migrating direction, while hMyo10 is not.
To further ensure these findings, we conducted time-lapse imaging on cortical neurons 2 days after electroporation at E15.5. We traced GFP-labeled neurons in the upIZ for 15 h under culture conditions. As expected, neurons expressing control plasmid migrated at an average speed of 13 μm/h toward the upCP with a bipolar shape, and the leading process pointed to the pial matter (Fig. 4f, top panel). On the other hand, neuron expressing fMyo10 shRNA plasmids displayed a curved leading process, migrated with varied directions, and most of them remained near the original sites at the end of observation (Fig. 4f, middle panel). Furthermore, in the hMyo10 shRNA-electroporated group, a clearly visible leading process was not observed at the beginning of the observation. Among 5 neurons examined, 3 neurons migrated upwards for a short distance and 2 neurons remained at the original sites. At the end of observation, these 5 neurons grew out multiple processes (Fig. 4f, bottom panel).
We next analyzed neuronal morphology and migration after coexpressing Myo10ΔMo with either hMyo10 or fMyo10 shRNA. Interestingly, coexpression of Myo10ΔMo with hMyo10 shRNA led to a drastic decrease in the proportion of multipolar neurons in the upIZ and loCP, bringing it back to control levels (Fig. 4g,i). However, this manipulation also resulted in an increase in the proportion of nonradially migrating neurons (Fig. 4h). A possible explanation is that the expression of the Myo10ΔMo construct may lead to overloading of hMyo10 activity, which promotes the morphogenesis and the morphological transition from multipolar to bipolar shapes by inhibiting neurite morphogenesis and, thus, multipolar phenotype caused by hMyo10 shRNA is largely rescued. On the other hand, in the case of coexpressioning Myo10ΔMo with fMyo10 shRNA, premature morphological transition could be induced by the expression of Myo10ΔMo, which leads to an increase of nonradially migrating bipolar neurons compared with that caused by fMyo10 shRNA alone (Fig. 4h). Taken together, these findings suggest that the Myo10 has divergent functions in cortical neuron migration with both of its isoforms, whereby fMyo10 is mainly required to establish proper orientation and hMyo10 is predominately involved in regulating the morphological transition.
DCC Is Involved in fMyo10-Regulated Orientational Selection, but Not hMyo10-Controlled Morphological Transition
Previous studies have reported that Myo10 transports the netrin-1 receptor DCC as cargo in neurons during axonal pathfinding (Zhu et al. 2007), we therefore wondered whether this is also the case in radial cortical neuron migration. Immunohistochemical analysis revealed that Myo10 expression overlapped with that of DCC in E15.5 mouse cortices (Fig. 5a). To determine whether DCC plays a role in cortical neuron migration, we designed and tested DCC shRNA plasmids, the most effective of which was then electroporated into mouse E15.5 cortices (Supplementary Fig. S4). Three days after electroporation, cortical neurons expressing DCC shRNA were abnormally located, as evidenced by an accumulation of labeled neurons in the VZ/SVZ (Supplementary Fig. S5). The observed phenotype was reminiscent of what we found after knocking down fMyo10, leading us to ask whether the migratory defects associated with the loss of Myo10 are DCC dependent. To test this, a DCC expression vector (Zhu et al. 2007) was coelectroporated into the embryonic cortex together with Myo10 shRNAs. Unexpectedly, severe inhibition of CP entry was observed after over-expression of DCC (Supplementary Fig. S4). Nevertheless, over-expression of DCC partially rescued the migratory defects caused by fMyo10 shRNA, as shown by the increase of the proportions of neurons in the CP and by decrease of that in the VZ/SVZ (Fig. 5b,d), with the percentage of nonradially migrating neurons falling back to control levels in the upIZ–loCP (Fig. 5c,f). On the other hand, over-expression of DCC in hMyo10 shRNA-electroporated cortices appeared to further impede radial migration, as shown by the reduction of neuronal proportion in the CP and by the increase of neuronal population in the IZ and VZ/SVZ (Fig. 5b, e), with a higher percentage of multipolar migrating neurons in the upIZ and loCP (Fig. 5c,g). These findings suggest that the regulatory effects of fMyo10 on cortical neuron migration are mediated at least in part by DCC.
In this study, we present evidence showing that Myo10 is an important regulator of radial neuronal migration in the embryonic mouse cortex. Knocking down Myo10 led to migratory defects, as evidenced by the accumulation of cortical neurons in deep cortical regions.
The 2 Myo10 isoforms, fMyo10 and hMyo10, were differentially expressed within the embryonic mouse cortex from VZ/SVZ to IZ, and that knocking each of them down resulted in distinct migratory defects in the developing cerebral cortex. By examination of the morphology of these neurons, we found that the majority of neurons expressing fMyo10 shRNA showed bipolar shape but displayed varied orientations in the SVZ/IZ, suggesting that fMyo10 is implicated in the selection of radial migration when the migration is initiated. By contrast, most of the neurons expressing hMyo10 shRNA exhibited multipolar shape in the upIZ/loCP, implying that the morphological transition from multipolar to bipolar neuronal shapes is interrupted during the initiation of migration from the IZ to the CP.
The idea that the 2 Myo10 isoforms play distinct roles in cortical neuron migration is further supported by the results obtained by coexpression of Myo10ΔMo, a construct that lacks the motor domain of Myo10 with the same domain as hMyo10 (Zhu et al. 2007). Coexpression of Myo10ΔMo and fMyo10 shRNA significantly increased the proportion of misorientated bipolar neurons compared with that caused by the expression of fMyo10 shRNA alone. As discussed above, hMyo10 promotes the morphological transition from multipolar to bipolar shapes by inhibiting neurite morphogenesis. Coexpression of Myo10ΔMo with fMyo10 shRNA may further increase the proportion of misorientated bipolar neurons by promoting premature morphological transition. Consistently, the phenotype (the increase of the proportion of multipolar neurons) caused by the expression of hMyo10 shRNA is largely rescued by coexpression of Myo10ΔMo possibly by replenishing hMyo10 activity. On the other hand, it is generally accepted that the newly generated postmitotic neurons at least in the VZ/SVZ are bipolar shape with short process, they achieve multiple processes in the IZ and then transform into bipolar neurons migrating into the CP. We found that fMyo10 mRNA is mainly expressed in the VZ/SVZ, but the intense Myo10 immunoreactivity is present in the IZ, suggesting fMyo10 protein may also be expressed by neurons in the IZ. Further studies are needed to explore if the orienting effect by fMyo10 occurs before or after the multiple processes stage. On the basis of the previous and present findings, we propose that the role of hMyo10 in the morphological transition in the IZ may be achieved by remodeling the multiple neurites via the rearrangement of cytoskeleton, contractile proteins, cytoskeleton-associated proteins, and membrane-bound proteins (da Silva and Dotti 2002; Brown and Bridgman 2003, 2004; Schaar and McConnell 2005), hereby transforming to the bipolar shape for radial migration.
DCC functions as a netrin-1 receptor, and its role in axon growth and guidance has been well characterized (Keino-Masu et al. 1996; Fazeli et al. 1997). Although DCC has been found to direct the several types of neuronal migration by receiving environmental signals (Yee et al. 1999; Schwarting et al. 2001; Murase and Horwitz 2002; Ding et al. 2005; Guijarro et al. 2006; Shi et al. 2008; Marcos et al. 2009), its involvement in cortical radial migration has not been reported. Using loss- and gain-of-function approaches, we demonstrated here that DCC is required for neuronal migration in the developing cortex. Both over-expressing and knocking down DCC led to cortical migration defects, implying that proper expression of DCC in appropriate cortical regions is necessary for normal cortical development. This speculation is supported by a recent in vivo study, showing that the balance between DCC and Unc5D (another receptor for Netrin-1) is required for morphological transition during the multipolar phase of migrating cortical neurons (Miyoshi and Fishell 2012).
Structural analysis has provided evidence showing that cargoes including DCC competitively bind to Myo10 (Hirano et al. 2011; Wei et al. 2011). As reported in our previous study in axonal pathfinding (Zhu et al. 2007), fMyo10 may also bind the intracellular domain of DCC and cargo it to the membrane, thereby regulating cortical neuron migration. It is possible that after being assembled to the membrane by fMyo10. DCC appears not to be involved in hMyo10-implicated morphological transition and, therefore when the 2 manipulations, expression of DCC and knocking down hMyo10, were applied simultaneously, an additive effect in impairing cortical neurons migration was observed.
In summary, we propose that Myo10 is involved in 2 key aspects of cortical radial migration: First, fMyo10 helps establish the proper migration orientation of postmitotic neurons by interacting with membrane components such as DCC; secondly, hMyo10 regulates the transition of cortical neurons from the multipolar to the bipolar morphology (Supplementary Fig. S5).
This work was supported by the Program of Introducing Talents of Discipline to Universities Project (B07017), the Fundamental Research Funds for the Central Universities (09SSXT129), the National Natural Science Foundation of China (30670689, 301170142, 31030034, and 31271486), and the Scientific Research Foundation from Northeast Normal University (09ZDQD04 and 10JCXK002).
We thank Dr W.C. Xiong (Georgia Health Sciences University, US) for providing Myo10 antibodies, plasmids and helpful advice, and Dr O. Reiner (Weizmann Institute of Science, Israel) for providing the centrin-RFP plasmid. Conflict of Interest: None declared.