We previously reported Scrib expression throughout the embryonic cerebral cortex both in neuronal progenitors and in the radial glia and also in the axonal fibers of the corpus callosum and midline glia at birth (Ezan et al., 2021). To extend these observations, we decided to query Scrib gene expression using publicly available cell-specific RNA sequencing data sets (Keil et al., 2018). In order to retrieve more specific information about the cell-type-specific expression patterns of Scrib, we reanalyzed the data of an RNA sequencing database (GEO data set GSE52564) of isolated cell types from young adult mouse brains (www.brainrnaseq.org) (Zhang et al., 2014). This data set allowed us to visualize the expression levels of Scrib that are expressed in a large variety of cortical cell types, ranging from neurons to astrocytes, oligodendrocyte precursor cells (OPCs), myelinating oligodendrocytes, microglia, and macrophages and also endothelial cells (Figure 1A). As compared with Nex, the fairly broad expression of Emx1 notably in neurons, astrocytes, and oligodendrocytes highlighted the validity of our approach and the necessity to use the Nex-Cre strain to specifically delete Scrib in neurons (Figure 1A). To specifically evaluate the effects of Scrib invalidation in post-mitotic neurons, we generated a novel cKO mouse strain: Scrib^fl/fl floxed mice were crossed with Nex-Cre mice to obtain a compound strain called Nex-Scrib ^−/− cKO mutant mice. Nex-Cre mice express the Cre recombinase at E11.5 in the hippocampus and precursor cells in the sub-ventricular and ventricular zones (SVZ and VZ, respectively), which eventually generate excitatory pyramidal neurons that disperse throughout the cortex (Goebbels et al., 2006; Belvindrah et al., 2007). Most importantly for our study, Cre-mediated recombination in Nex-Cre mice occurs neither in cortical progenitors, glial cells, and oligodendrocytes nor in GABAergic inhibitory interneurons. First, the efficiency of the excision of Scrib was validated by polymerase chain reaction (PCR) using DNA samples isolated from the cortex of Nex-Scrib cKOs conditional and control littermates (Figure 1B). Second, using a homemade Scrib-specific antibody (Montcouquiol et al., 2006), we performed a western blot analysis to examine to which degree the Scrib product was reduced in the cortex of the Nex-Scrib ^−/− cKO. As shown by western blot analysis in Figure 1C, Nex-Cre-mediated knockout of Scrib resulted in a reduction of Scrib protein in the P0 cerebral cortex. General observations of the Nex-Scrib ^−/− cKO mice indicated that these mice were globally indistinguishable from control littermates and there was no difference in body weight between them (Figure 1D).

We next investigated whether the cerebral cortex of Nex-Scrib ^−/− cKO mutants was impaired. We observed that the overall brain structure and particularly the dorsal surface areas of the cerebral hemispheres of the Nex-Scrib ^−/− cKO mutants appeared normal when compared to their control littermates (Figure 2A). Furthermore, an histological analysis showed no obvious reduction of cortical thickness throughout the cingulate (Cg), motor (M), primary somatosensory (S1), and secondary somatosensory (S2) in Nex-Scrib ^−/− cKO mutants versus the control (Figure 2B). We next performed an immunohistological analysis in Nex-Scrib ^−/− mutant mice motor cortices to assess if the cortical layering is impaired in these mutants. We selected antibodies directed against specific markers of callosal projection neurons (CPNs) such as CuxI and Satb2 (late-born neurons from layer II/III) and Ctip2 (earlier-born neurons from layer V) as some of these markers are essential to the formation of the corpus callosum (Alcamo et al., 2008; Britanova et al., 2008). This analysis revealed no major CPN production defects (see central histograms in insets in Figures 2C–E). However, the distribution of both early- and late-born neurons was affected with many CPNs mislocalized in deeper layers of the cortex at P0 (see arrowheads in Figures 2C–E). These defects in layering are exemplified by the quantification of markers in bin 10 (deepest bin), showing an increase in CuxI (ctrl, 0.6% ± 0.1; mutant, 17.3% ± 10.8; p = 0.046) (Figure 2C), Satb2 (ctrl, 20.3% ± 4.6; mutant, 48.6% ± 17.3; p = 0.04) (Figure 2D), and Ctip2 percentages (ctrl, 12.2% ± 5.1; mutant, 43.7% ± 17.3; p = 0.04) (Figure 2E). The fact that both early-born Ctip2-positive CPNs and also late-born CuxI- and Satb2-positive CPNs cannot reach their proper layer indicates that neuronal Scrib is essential at a cell-autonomous level during radial cortical migration.

Our previous results in global brain Scrib cKO showed a severe caudal agenesis of the corpus callosum and dorsal hippocampal commissure associated with Probst bundles which we attributed mostly to midline glial defects (Ezan et al., 2021). To assess if this phenotype could also have a neuronal component, we performed histological analysis on newborn Nex-Scrib ^−/− mutants (Figures 3A–B’). Coronal histological sections show a marked reduction in the size of the corpus callosum (CC hypoplasia), that is, ∼35% thinner in newborn Nex-Scrib ^−/− cKO mutants relative to their control littermate (Figure 3C). As expected, immunostaining of P0 coronal sections using antibodies against GFAP (marker of midline glia) and the axonal marker L1-CAM revealed that late deletion of Scrib resulted in a normal organization of the midline glia structures, with GFAP-positive cells positioned in proper locations, surrounding the CC, as in control littermate coronal sections (Figures 3D–E’). These findings indicate that Scrib deletion of in the dorsal telencephalon of Nex-Scrib ^−/− cKOs (including the cerebral cortex and the hippocampus) has a neuronal-specific contribution on the formation of commissures such as the corpus callosum. Taken as a whole, we show that neuronal Scrib is essential for CPN migration and proper cerebral cortex layering and thus in corpus callosum formation.

To assess behavioral alterations in adult Nex-Scrib ^−/− mice and their littermate controls (10–20 weeks), we performed tests of anxiety-like behavior and locomotor and exploratory activity. We showed that Nex-Scrib ^−/− mice exhibited significantly reduced exploration characterized by a reduction of the followed distance in the open field (t-test; t25 = 2.13, p < 0.05; Figure 4A) in the elevated plus-maze (t-test; t24 = 2.55, p < 0.01; Figure 4B) and in the Y-maze (t-test; t24 = 3.09, p < 0.01; Figure 4C). The Nex-Scrib ^−/− mice do not display changes in anxiety-like behavior as reflected by the time spent at the center of the open field (t-test; t25 = 0.41, ns; Figure 4A) and the time spent in the open arms of the elevated plus-maze (t-test; t24 = 1.38, ns; Figure 4B). In addition, the measure of spontaneous activity in a novel home cage showed Nex-Scrib ^−/− mice that were significantly less active than their controls (genotype effect: F1,143 = 6.36, p < 0.05; Figure 4D). The same hypoactivity was observed during the nycthemeral cycle in their home cages as determined by 24-h continuous monitoring of locomotor activity (genotype effect: F1,13 = 7,579, p < 0.05; Figure 4E). Since motor coordination and balance are essential to normal motor functions, we have submitted mice to the beam walking, grid handling, and rotarod tests. No defects were observed in Nex-Scrib ^−/− mice in terms of balance measured by the time spent on the beam (t-test; t14 = 0.33, ns; Figure 4F) and the number of paw slips during walking (Mann–Whitney test, ns; Figure 4F). Also, the motor coordination is not impaired as mice take longer to fall in the grid handling test (t-test; t15 = 2.19, p < 0.05; Figure 4G) and the rotarod (genotype effect: F1,81 = 12.88, p < 0.001; Figure 4H). These findings demonstrate that Nex-Scrib ^−/− mutant mice are hypoactive in novel and familiar environments, in contrast to Emx1-Scrib ^−/− mutant mice (Ezan et al., 2021). Taken together, our results point toward an important role for Scrib in the telencephalic glutamatergic circuitry in mouse locomotor behavior.

Our present characterization of the Nex-Scrib ^−/− cKO line revealed that Scrib is essential to cortical layering. While Scrib is inactivated in cortical progenitors, in radial glial cells, in Emx1-Scrib ^−/− mutant mice, the ablation of Scrib solely in postmitotic cortical neurons, during cortical development stages, led to significant neuronal migration defects. Indeed, Scrib loss in neurons did affect their ability to migrate into adequate cortical layers, consistent with our observations in Emx1-Scrib ^−/− cKO (Ezan et al., 2021). More specifically, we observed cortical layering defects in Nex-Scrib ^−/− cKO brains with mispositioned neurons from layers II/III and V. Both early-born Ctip2-positive CPNs and also late-born CuxI- and Satb2-positive CPNs are impacted, merely by their numbers, but rather by their positioning, implying a prominent role for Scrib throughout every cortical layer. Notably, the extent of these layering defects is similar to the one observed in Emx1-Scrib ^−/− cKO mutants (Ezan et al., 2021), indicating that Scrib contribution to the neuronal radial migration relies primarily, if not totally, on its cell-autonomous neuronal contribution. Notably, none of the Nex-Scrib ^−/− cKO mice suffered from microcephaly, and this was expected considering that the main characteristic of this cKO line is to bypass the negative impact of Scrib loss on proliferation and neurogenesis observed in the Emx1-Scrib ^−/− mutant mice (Ezan et al., 2021). At present, the precise molecular mechanisms at play in Scrib-dependent neuronal migration remain elusive (Jossin, 2020) and are challenging to define owing to the complex interplay of neurons with radial glia in the cerebral cortex (Marin et al., 2010; Hansen and Hippenmeyer, 2020). However, coordinated regulation of adhesion forces between neurons and radial glia is required for proper neuronal migration, and we suspect that Scrib may be central to this process as shown in another context (Qin et al., 2005).

In the present study, we show that the neuronal-specific invalidation of Scrib leads to a partial but marked corpus callosum hypoplasia (reduction of CC in thickness)/dysgenesis. We can infer that these commissural defects stem from the CPN migration defects observed in the Nex-Scrib ^−/− cKO mice: while some CPNs do reach their proper cortical layer, the others are unable to correctly send their axonal projections toward the contralateral side as observed in many other models with partial ACC (Edwards et al., 2014). As expected, Scrib loss solely in neurons does spare midline glial structures that are essential to promote interhemispheric tissue fusion and remodeling (Yu et al., 2010; Gobius et al., 2016), and the few callosal fibers reaching the midline can ultimately cross it. Tissue fusion is a central process during neural tube closure (Pai et al., 2012) and during interhemispheric remodeling in CC formation (Wahlsten et al., 2006; Paul et al., 2007), and we suspect that Scrib is involved in this process in other tissues as well (see below). Our previous study indicated that the broad loss of Scrib expression in the midline glia leads to the disorganization of that structure (Ezan et al., 2021). The formation of Probst bundles (Shu et al., 2003) in Emx1-Scrib ^−/− mutant mice suggested that a non-autonomous Scrib-dependent mechanism affects the guidance and the crossing of the commissural axons (Ezan et al., 2021). Supporting this hypothesis, the partial corpus callosum dysgenesis and the proper midline glia architecture we observe in the Nex-Scrib ^−/− cKO mice are in favor of an additional requisite contribution of Scrib in glial cells for a complete CC formation and hence the strong ACC phenotype we observe in our global Scrib cKO models. Interestingly, Scrib is enriched in oligodendrocytes (where Emx1 is expressed) and is involved in the myelination initiation (Jarjour et al., 2015). However, oligodendrocyte-specific deletion of Scrib does not appear to impact CC formation. Also, further experiments using conditional approaches to inactivate Scrib in glial subpopulations by different means would help clarify this issue.

Overall, our observations using complementary Scrib cKO models indicate that it has a major role in cortex development and commissure formations on a large spatio-temporal scale. While Scrib loss in Emx1-Scrib ^−/− mutant mice displays major microcephaly and complete CC agenesis, Nex-Scrib ^−/− mutant mice display “only” cortical layering defects and partial CC dysgenesis. An explanation is that the time course and the extent of the genetic modification in the latest model were insufficient to fully impact these processes. In our previous investigation, Scrib ^fl/fl floxed mice were crossed with FoxG1-Cre or Emx1-Cre mice to obtain FoxG1-Scrib ^−/− and Emx1-Scrib ^−/− cKO mutant mice, respectively, with an early Scrib excision in different cell types in the brain (see Table 1). Cre-mediated recombination in FoxG1-Cre mice starts as early as E8.5 and occurs mainly in all cells from both dorsal and ventral telencephalons (Hébert and McConnell, 2000). This includes neuroepithelial progenitors that specify various neuronal and glial cells from the neocortex and hippocampus and also ganglionic eminence progenitors giving rise to GABAergic interneurons. In Emx1-Cre mice, the Cre recombinase is expressed as early as E10.5 and only in the dorsal telencephalon (Gorski et al., 2002). Emx1-Cre expression led to Scrib excision of most neurons of the neocortex and hippocampus in glial cells of the pallium but not in the GABAergic interneurons. While these global Scrib cKOs clearly define a role for Scrib in normal forebrain formation, they do not distinguish between a neuronal and non-neuronal site of action. Also, it could be argued that the partial phenotype observed in Nex-Scrib ^−/− cKO mice is presently due to a subspecific invalidation of Scrib. The absence of Probst bundles in Nex-Scrib ^−/− cKO mice suggests that neuronal Scrib is merely implicated in axonal guidance during CC formation and Scrib impact on this process may rely on other cell types (e.g., glial cells). A requirement for Scrib in axonal guidance has indeed been shown at the optic chiasm (Rachel et al., 2000) and during spinal commissure formation in zebrafish (Sun et al., 2016). Alternatively, given its essential role in many cellular processes, another consideration is that the lack of neuronal Scrib is functionally compensated by other LAP (leucine-rich repeats and PDZ) family members (Santoni et al., 2002), as recently evidenced during apicobasal polarity establishment in mammalian epithelial cells (Choi et al., 2019).

From the behavioral standpoint, we observed in the present study a hypoactive locomotor behavior phenotype in Nex-Scrib ^−/− cKO mice in every paradigm tested (Plus maze, Open field, Y-maze). This observation is quite remarkable considering that our previous analysis using the Scrib cKO mice revealed an opposite phenotype in locomotor behavior (Ezan et al., 2021) (see Table 1). Specifically, we observed in Emx1-Scrib ^−/− mutant mice a hyperactive locomotor behavior comparable to established models (Itohara et al., 2015). Surprisingly, Scrib signaling solely in glutamatergic neurons not only proved to convey an opposite effect on locomotion compared to Emx1-Scrib ^−/− mutant mice but also seems to be involved in the regulation of motor coordination. These results strongly support a critical role for Scrib in regulating locomotor behavior of the animal and also highlight a spatial and temporal dependence on Scrib regarding this regulation. To ensure direct comparison to our previous study, our present work is limited to male mice. A comparative study in female mice is warranted to evaluate if the deletion of Scrib in glutamatergic neurons would also result in reduced locomotion. Because of the early and broad expression of Emx1 in the brain, it is impossible to identify what specific circuit drives the hyperactive locomotor behavior. On the other hand, Nex expression is restricted to postmitotic glutamatergic neurons, supporting the hypothesis that the pathological origin for the hypoactive locomotor behavior has a strict neuronal origin. In addition, since Nex-Cre;Rosa26 reporter strains show some expression in the cerebellum (Goebbels et al., 2006; Belvindrah et al., 2007), where Scrib is expressed (Moreau et al., 2010), we cannot exclude the possibility that some effects that we observe in our tests are due to an impact of a loss of Scrib expression in the cerebellum. Importantly, locomotor behavior is unaffected in CamK-Scrib ^−/− mutant mice that target postnatal excitatory neurons of the cortex and the hippocampus (Hilal et al., 2017), demonstrating that neuronal Scrib signaling is pivotal for the locomotor output essentially during embryonic brain development. Thus, these findings support the relevance of this gene in glutamatergic circuits to control the excitatory drive on locomotor behavior, a previously unknown function of Scrib on behavioral performance. It is presently unclear how Scrib could select the cell-type type in which they exert these opposing effects, but we can hypothesize that this happens through the regulation of glutamatergic signaling (Piguel et al., 2014). Given the pronounced locomotor dysfunction in Scrib cKO mice, which is also altered in Nex-GluN2B cKO (Wang et al., 2011), deficiencies in Scrib expression could contribute to the impairment of the cortical excitatory/inhibitory balance (Nelson and Valakh, 2015). In the control situation, Scrib helps to maintain a stable glutamatergic activity at the neuronal circuit level, possibly through the regulation of GluN2B subunits as has been shown by us previously in other contexts (Piguel et al., 2014), thereby representing an important regulator of the motor cortex. An activity-related control mechanism of the glutamatergic signaling is removed when deleting Scrib from cortical/hippocampal glutamatergic neurons. Consequently, the excitability of the whole circuit is unbalanced in an opposite manner in Nex-Scrib ^−/− cKO and Emx1-Scrib ^−/− cKO mutant mice, but which both represent a deleterious state, leading to locomotor behavioral alterations. In the future, it would be of interest to survey VRJS patients in clinical studies that may report dysfunctions in locomotion and general activity. Altogether, these findings underline the complexity and multi-faceted nature of Scrib in the forebrain and emphasize the requirement for additional cell-specific investigations in parallel with global brain experimentations.

Original reports for VRJS described patients with dysmorphic facial features, cardiac ventricular septal defects, coloboma, and skeletal abnormalities together with neurological features including delayed psychomotor development, recurrent seizures, CC agenesis, and microcephaly (Verheij et al., 2009; Dauber et al., 2013). The minimal common deletion encompassed the SCRIB and PUF60 genes whose invalidation in zebrafish showed their contribution to VRJS clinical features (Dauber et al., 2013). Our previous study (Ezan et al., 2021) and the present one support a role for Scrib in some brain structural deficits observed in VRJS. A case with 8q24.3 deletion including SCRIB showed dysmorphic facial features, cardiac defects, and vertebrate anomalies but no microcephaly, highlighting the complexity of human genetics (Wells et al., 2016). Other data suggest that Scrib may also be implicated in extra-neurological features of VRJS. Scrib mutant Crc and myocardium-specific Scrib cKOs display ventricular septal defects and thinned myocardial walls that recapitulate cardiac features of VRJS patients (Phillips et al., 2007; Boczonadi et al., 2014) in line with a fetal case carrying SCRIB haploinsufficiency (Wells et al., 2016). Also, an exome analysis identified a rare SCRIB variant that resulted in coloboma (Hocking et al., 2018). A common denominator to the Scrib-associated defects such as NTDs, ACC, ventricular defects, or coloboma is the failure in the tissue fusion process (Yu et al., 2010; Hocking et al., 2018). Finally, in the light of the growing interest for an ontogenetic link between the nervous system and craniofacial development, it is worth mentioning the implication of PCP signaling to the latest (Topczewski et al., 2011). A role for SCRIB in the face, vertebrae, and digit development was suggested (Wells et al., 2016), and Scrib expression is increased in a craniosynostosis chick embryo model (Li et al., 2011) and also in mature osteoclast in vitro, hinting at a function in bone mass regulation (Orriss et al., 2018). At present, an implication of SCRIB in craniofacial defects may appear provocative, but these data further warrant investigation on its role in craniofacial and skeletal development and its relationship toward brain formation (Motch Perrine et al., 2017).

Taken as a whole, our results summarized in Table 1 show that 1) various neuroanatomical effects of Scrib are conducted by different cellular circuits, which genetic approaches can dissect; 2) cortical glutamatergic neurons mediate a large portion of the cell-autonomous deleterious effects of Scrib on neuronal migration and thus cortical layering; 3) glutamatergic neurons do partially mediate the effects of Scrib on global cerebral cortex formation and corpus callosum formation; 4) Scrib expression in cortical glutamatergic neurons plays a pivotal role in mediating the hypolocomotor effects observed in Nex-Scrib ^−/− cKO mice. Together with our previous study, our present work suggests a dysfunction in the telencephalic neuronal circuitry in behaving Scrib ^−/− cKO mice that ultimately points to a dysregulation of the excitation/inhibition balance as a potential mechanism for locomotor dysregulation in our models. This neuroanatomical and behavioral integrative study further strengthens our perception of the importance of Scrib in brain formation relevant to neurodevelopmental diseases such as the rare Verheij/8q24.3 deletion syndrome.

Mice were housed in the animal facility of the Neurocentre Magendie under controlled conditions (lights on from 7:00 a.m. to 7:00 p.m.) with ad libitum food and water. For timed pregnancy, detection of the vaginal plug assessed in the morning was designated as embryonic day E0.5. This study was carried out in compliance with the ARRIVE guidelines and according to the European Directive 2010/63/EU. It was approved by the Bordeaux University Animal Care and Use Committee (#5012016-A and #5012015-A).

Nex-Cre (also called Neurod6-cre) mice were obtained from Klaus-Armin Nave (Max Plank Institute, Germany). Scrib ^fl/fl floxed mice were generated previously in collaboration with Neal Copeland, Nancy Jenkins, and Rivka Rachel at NIC/NIH (Yamben et al., 2013). The Scrib mouse gene contains 38 exons that are translated into a full-length 180 kDa protein. In Scrib ^fl/fl floxed mice, lox-P sequences are inserted before exon 2 and after exon 8 (Hilal et al., 2017). All the exons between 2 and 8 are excised after recombination, leading to the loss of the full-length protein (Ezan et al., 2021). Heterozygous rodents were intercrossed to generate homozygous and wild-type littermates as previously published (Ezan et al., 2013). Specifically, heterozygous male Nex^Cre/+; Scrib ^fl/fl mice were intercrossed with female Scrib ^fl/fl mice to produce offspring of two genotypes, with or without the Nex^Cre transgene: Nex^+/+; Scrib ^fl/fl (which is the control littermate, Ctrl) or Nex^Cre/+; Scrib ^fl/fl (referred to as Nex-Scrib ^−/−). Nex-Scrib ^−/− mutant mice are viable and fertile and do not display any gross physical abnormalities. Nex-Cre mice express the Cre recombinase at or before E12.5 in the hippocampus and in precursor cells in VZ/SVZ, which eventually generate pyramidal neurons that disperse throughout the cortex (Goebbels et al., 2006; Belvindrah et al., 2007). It is also expressed in a scattered pattern in the cerebellum (Goebbels et al., 2006; Belvindrah et al., 2007). Of note, Cre-mediated recombination in Nex-Cre mice occurs neither in glial cells and oligodendrocytes nor interneurons. Genotyping was performed on cortex samples by PCR using the following primers: F (5′-gca​cac​tgg​gta​tca​tgg​cta-3′), R1 (5′-gca​atc​tcc​aga​gcc​tta​cag​a-3′), and R2 (5′-ccc​ttg​gaa​acc​tac​atc​cca​a-3′) (Ezan et al., 2021). Wild-type (WT), floxed, deleted Scrib alleles were distinguished by the following amplified products: for the WT band (F + R1; 437 bp), flox band (F + R1; 541 bp), and cKO band (F + R2; 193 bp) (Figure 1B). Cre genotyping was performed using the following primers: F (5′-cgg​cat​ggt​gca​agt​tga​ata-3′) and R (5′-gcg​atc​gct​att​ttc​cat​gag-3′), resulting in a 300 bp band. PCR product analysis was done using a Labchip GX microfluidic electrophoresis system (Perkin-Elmer) using the DNA5k kit and visualized by a virtual gel image generated by the Labchip GX software. Cre minus littermates showed no detectable phenotype and were thus used as controls.

Cerebral cortices of P0 pups were homogenized in RIPA buffer [10 mM Tris-Cl (pH 8.0), 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl] supplemented with protease inhibitor cocktail (Complete™; Roche). Using the Pierce BCA Protein assay kit (Thermo Scientific), we determined the protein sample concentration and equal amounts of protein were diluted in the sample buffer, separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and visualized using the enhanced chemiluminescence (ECL) as previously described (Ezan et al., 2021). Protein extracts were separated on an 8% gel and transferred overnight to PVDF membranes (Millipore). After blocking with 5% non-fat milk in 1×Tris-buffered saline pH 7.4; 0.05% Tween-20 (TBS-T) for 30 min at room temperature (RT), membranes were first incubated with homemade rabbit anti-Scrib ab (MM468; 1:5,000) (Montcouquiol et al., 2006) and mouse anti-GAPDH (Millipore; 1:5,000) for 1 h at RT and then with secondary antibodies (Amersham; donkey anti-rabbit or anti-mouse IgG conjugated to horseradish peroxidase, 1:20,000) in 1% non-fat milk in TBS-T for 1 h at RT. Immunoreactive signals were detected using the Pierce ECL substrate (Thermo Scientific). Band intensity analysis was performed by densitometry with the ImageJ software (http://imagej.nih.gov), and band intensity was then quantified as a percentage of control band intensity using a representative of at least three independent experiments.

For histology, brains (P0 new-born) were harvested and fixed in Bouin’s fixative (Electron Microscopy Sciences) overnight as previously described (Ezan et al., 2021). Brains were dehydrated in ethanol, paraffin-embedded, and sectioned coronally (20 µm); then, sections were collected onto Superfrost plus Gold slides (Thermo Scientific) and stained with hematoxylin and mounted with Entellan (Millipore). The brains sections were examined under a Leica MZ-16 stereomicroscope, imaged using the NanoZoomer 2.0-HT slide scanner in the Bordeaux Imaging Center (http://www.bic.u-bordeaux.fr) and finally analyzed using the NDP viewer software (Hamamatsu).

For immunofluorescence staining (IF), experiments were performed as previously described in Ezan et al., 2021. P0 pups were anesthetized with pentobarbital and perfused transcardially with 4% paraformaldehyde (PFA) in PB buffer (0.1% PBS, 0.9% NaCl, PH = 7.4). After postfixation in 4% PFA overnight at 4°C, dissected brains were infused in 30% sucrose in PB overnight and embedded in O.C.T (Sakura Finetek). Brains were cryosectioned coronally (20 μm thickness) and mounted on Superfrost plus slides (VWR). Sections were PBS-hydrated, permeabilized with 0.2% Triton-X100/PBS (PBS-T), blocked using 10% Normal Goat Serum (NGS), and incubated overnight with the following primary antibodies in 5% NGS overnight at 4°C: rat Ctip2 (Abcam; #ab18465, 1:500), mouse Satb2 (Abcam; #ab51502, 1:100), rabbit Cux1 (Santa Cruz Biotechnology; #sc-13024; 1:100), rabbit anti-GFAP (DAKO; #Z0334; 1:1,000), and rat anti-L1-CAM (Millipore; #MAB5272; 1:1,000). After three washes in PBS-T, samples were incubated for 2 h with the secondary antibody Alexa Fluor^® 488 or 546 Goat Anti-Mouse/Rat/Rabbit IgG (Life Technologies; 1:200) and then with 4,6-diamidino-2-phenylindole (DAPI) (Life Technologies; 1:20,000) for 30 min. Finally, after three PBS washes, samples were mounted with the Prolong Gold anti-fading reagent (Invitrogen). Immunostained sections were imaged using a Zeiss Axio Imager Z1 microscope and Axio Vision (Version 4.7) imaging analysis software (Carl Zeiss). Images were processed with Photoshop CS5 software (Adobe) and ImageJ software (http://imagej.nih.gov).

Sections chosen for analysis were matched along the rostral-caudal axis and observed at the caudal level for every sample. After conversion to gray values and normalization to background staining, each region of interest (ROI) in the motor cortex was subdivided into 10 equal bins from the pia (bin 1) to the outer border of the intermediate zone (bin 10) to assess CPN distribution. Doing so, layers II–III (Cux1- and Satb2-positive) were corresponding to bins 2–4, while layer V (Ctip2-positive) was corresponding to bins 5–7. The number and distribution of CuxI-, Satb2-, Ctip2-, and DAPI-labeled cells in each zone were determined manually using the cell counter plugin for ImageJ (http://rsbweb.nih.gov/ij/plugins/cell-counter.html) and blindly to the genotype. Data are given as a ratio (in percent) of the total of cells positive for each marker to DAPI-positive cells in each bin (mean ± SEM). Analysis was performed using at least three independent experiments (three to four brains per genotype were quantified). The effect of the genotype on the distribution of cells within each bin was assessed using the Student test (t-test).

Behavioral experiments are briefly described below as they were performed as described previously (Moreau et al., 2010; Hilal et al., 2017; Ezan et al., 2021). All animal experiments were approved by the French Ministry of Research and local Ethics Committee and performed in accordance with the National Health and Medical Research Council Guidelines for the care and use of animals for scientific purposes. Male Nex-Scrib ^−/− cKOs littermate mice were 3 months old at the beginning of each study. Mice were individually housed 1 week before starting the behavioral experiment on a 12:12 h light–dark cycle and given ad libitum access to food and water. Behavioral studies were carried out during the light period. Behavior apparatuses were cleaned between mice with Phagospray.

Statistical analyses were carried out using the GraphPad Prism statistical package (GraphPad). Validation of normality of distribution and homogeneity of variance were performed, and unpaired Student’s two-tailed t-tests for two data sets were used to compare groups with similar variances. Details of n values are indicated along the p values in figure legends. p ≤ 0.05 was considered as statistically significant. Statistics was derived from at least three independent experiments and not from technical replicates. For behavior analyses, repeated measure ANOVA was used for the evaluation of the effect of genotype and time in actimetry and rotarod tests. The Bonferroni posthoc test was used when appropriate. Student’s t-test was used for comparing genotypes in other behavior tests.