All experiments were performed strictly in accordance with the European Community Council directive of 24 November 1986 (86–609/ECC) and the decree of 20 October 1987 (87–848/EEC). Mice were raised in our animal facilities and treated according to the principles of laboratory animal care, and experiments and mouse housing were approved by the Animal Welfare Committee of Université catholique de Louvain (Permit Numbers: 2017/UCL/MD/008 and 222801). Vsx1-CreER^T2 and Rosa26R-tdTomato mice have been previously described (Madisen et al., 2010; Baudouin et al., 2021). The Arid3c^−/− null line was generated using CRISPR-Cas9 mediated homologous recombination. Two ultramers-loxP sequences (5’-AAGAGCAGGCTGAGCCATGGAAATACTTCCCTGATTCTCCTCTCCCTGTATTTGCCACGTGCTAGCATAACTTCGTATAATGTATGCTATACGAAGTTATGCAAGGTCTGAGCTACCTAGAGGCACCACTAGCTAGGACAGGTCGCGTAGATAGATTAGA-3′ and 5’-CCTGTCCTGCAGCAGGAGGGAATACAGTTAGACTGAAGGAAGGCTTTCCTGAAGGATGGGGAATTCATAACTTCGTATAATGTATGCTATACGAAGTTATCCGGGGCCTTGTGAATTGGGTAGTGGAGAAGGGATTTCTCCCTAGAAAATTTCAGTATTG-3′), two Arid3c-crRNA (5’-GGTAGCTCAGACCTTGCACG-3′ and 5’-CTTTCCTGAAGGATGGGCCG-3′) and Cas9 were injected in male pronuclei to target the exon 2 of the Arid3c gene. This enabled the simultaneous obtention of Arid3^+/− mice and of Arid3c^+/flox conditional mice (Supplementary Figure S1A). The removal of the second exon of Arid3c leads to a frameshift in the ORF and the creation of an early STOP codon, which causes a loss of 74% of the C-terminal part of the protein, including the ARID and REKLES domains. The Arid3c^+/− knockout (223 bp) or wild type (961 bp) alleles were identified by PCR using primers 1 and 2: 5’-GGCTCTGAGTGTCCATCCTC-3′ and: 5’-CGGAAGTGATGGACTGCTGC-3′. The Arid3c^flox allele (858 bp) was identified by PCR using primers 3 and 4: 5’-TGTATTTGCCACGTGCTAGCATAACTTCGTATAAT-3′ and 5’-CCCAATTCACAAGGCCCCGGATAACTTCGTATAGC-3′ (Supplementary Figure S1B). Further details are available upon request. Arid3c^+/− mice were then crossed to generate Arid3c^−/− constitutive knockout embryos, using Arid3c^+/+ embryos as controls. The Arid3c protein was undetectable in the spinal cord of Arid3c^−/− embryos (Supplementary Figure S1C). The morning of the vaginal plug was considered embryonic day (E)0.5. The embryos were collected at embryonic day E 10.5, 11.5, 12.5, 14.5, 16.5, or 18.5. For crossings with Vsx1-CreER^T2 transgenic mice, pregnant females were injected intraperitoneally with tamoxifen (100 mg/kg) twice at E9.5 to activate CreER^T2 activity.

Six spinal cords from E14.5 Vsx1|tdTomato embryos were dissected and dissociated using a neural tissue dissociation kit (MACS; Miltenyi Biotec #130-092-628) according to the manufacturer’s instructions. Dissociated cells were sorted by FACS (BD FACSAria III) to separate tdTomato⁺ cells (V2 INs) from tdTomato- cells (all the other cells of the spinal cord). Sorted cells were collected in TRIzol reagent, and RNA was purified with the RNeasy micro kit (QIAGEN #74004). RNA concentration and quality were assessed using a Bioanalyzer (Agilent) and submitted to Genewiz to prepare an ultra-low input RNA-seq library before sequencing with an Illumina HiSeq. Preliminary data were analyzed by Genewiz using the standard RNA-seq data analysis package.

The raw data were already normalized and were further log2 transformed with a pseudo count of 1. Since the focus of the analysis was to compare cells expressing Arid3c versus cells expressing Gata3 or Chx10, the 33 clusters grouping from Delile et al. (2019) were subjected to filtering with the following rules: a cluster was included if it contained at least either 5 cells expressing Arid3c, or 10 cells expressing Gata3 or Chx10. Cells expressing both Gata3 and Chx10 were excluded and cells expressing Arid3c were grouped as a new additional cluster. Then, pairwise cluster comparisons were performed with the scoreMarkers function from the R (version 4.3.3)¹ package scran [version 1.30 (Lun et al., 2016)], with a blocking effect of the developmental stage of the cell. The value of the “log Fold Change detected” (logFC.detected) for all the pairwise comparisons of Arid3c⁺ cells versus each of the other clusters of cells was measured to rank candidate markers as it measures the log-fold change in the proportion of cells with detected expression between clusters (Amezquita et al., 2020). A positive value indicates that a greater proportion of Arid3c⁺ cells express the gene of interest compared to the other cluster. The minimum value of the logFC.detected (min.logFC.detected) across the pairwise comparisons was used to rank candidate markers as it corresponds to the most stringent value to find upregulated genes in the Arid3c⁺ cluster versus all of the other clusters (Amezquita et al., 2020).

In ovo electroporation was performed at Hamburger–Hamilton (HH) stage 15-16 and embryos were collected at HH26 (72 h after electroporation). The electroporated plasmid was pCIG-Arid3c-IRES-EGFP (1 μg/μl). Details about plasmids are available upon request.

Collected embryos were fixed in ice-cold PBS/4% paraformaldehyde (PFA) overnight. After washes in PBS, the fixed embryos were incubated in PBS/30% sucrose overnight at 4°C, embedded and frozen in PBS/15% sucrose/7.5% gelatin in isopentane at −55°C. In situ hybridization was performed on 14–20 μm serial cryosections. On the first day, slides were post-fixed in PBS/4% paraformaldehyde for 10 min and then acetylated in a 0.1 M triethanolamine–HCl–acetic anhydride (pH8) solution for 10 min. After blocking non-specific hybridization for 2 h at RT in hybridization buffer (50% formamide/10% dextran sulfate/10% salt 10x (NaCl 2 M/Tris–HCl 0.1 M pH7.5/NaH₂PO₄, 2H₂O 50 mM/Na₂HP O₄ 20 mM/EDTA 50 mM)/2% Denhardt’s 50x/1 mg/mL of tRNA), slides were incubated with 300 ng of DIG-labeled probe diluted in hybridization buffer overnight at 65°C. The next day, slides were washed four times for 45 min in a 1xSSC/50% formamide/0.1% Tween 20 washing buffer at 65°C and three times for 30 min in a 20% MAB5x (maleic acid 100 mM/NaCl 150 mM pH7.5)/0.1% Tween 20 buffer.

Slides were then subjected to an RNase treatment with 1:2000 diluted RNase T1 (Roche #10109193001) in an RNase buffer (Tris–HCl 10 mM pH7.5/400 mM NaCl/5 mM EDTA) for 30 min at 37°C. After a last 30 min wash in MABT buffer, slides were incubated with a 1:4000 anti-digoxigenin antibody (Roche #11093274910) in a blocking solution (1x MAB/0.1% Tween 20/2% Boehringer reagent/20% Horse serum) overnight at 4°C after a blocking step of 1 h in the same solution. On the last day, slides were washed three times for 30 min in MABT buffer and two times for 15 min in an alkaline phosphatase buffer (100 mM Tris–HCl pH9.5/100 mM NaCl/50 mM MgCl₂/0.1% Tween 2). For the detection of the probe, slides were incubated with 1:2000 diluted NBT (Roche #11383213001) and 1:350 diluted BCIP (Roche #11383221001) in alkaline phosphatase buffer. This solution was changed every 2 h until the staining was complete. Finally, the staining was stopped by washes in 1x PBS/0.1% Triton. Slides were washed in 1x PBS and mounted with Dako Fluorescence Mounting Medium (Dako #S3023). The probes used were DIG-conjugated Arid3c antisense (primer pair: 5′-CCCCAAGAGGAAGGAGTTTC-3′ and 5′-GCGTCGAGCAAAGAGGATAC-3′), Arid3a (primer pair: 5′-CTGTCTTAGCCGCACAATCA-3′ and 5′-AGGGTGACCAGGAAGGTTCT-3′) and Arid3b [primer pair: 5′-CAAGAACCCAGAGCAACACC-3′ and 5′-GAGTAGAGCCCGCAATGAGG-3′(Casanova et al., 2011)] RNA probes.

Collected embryos were fixed in ice-cold PBS/4% paraformaldehyde (PFA) for 15–25 min, depending on the developmental stage. After washes in PBS, the fixed embryos were incubated in PBS/30% sucrose overnight at 4°C, embedded and frozen in PBS/15% sucrose/7.5% gelatin in isopentane at −55°C. Immunolabeling was performed on 14 μm serial cryosections as previously described (Espana and Clotman, 2012). Primary antibodies against the following proteins were used: rabbit anti-Arid3c (Tidwell et al., 2011) at 1:1000, sheep anti-Chx10 (Exalpha Biologicals #X1179P) at 1:200, goat anti-Foxp1 (R&D #AF4534) at 1:500, rat anti-Gata3 (Absea Biotechnology #111214D02) at 1:15 or rabbit anti-Gata3 (Cell signaling #5852) at 1:200, chicken anti-GFP (Aves lab #GFP-1020) at 1:2000, guinea pig anti-OC1/HNF6 (Espana and Clotman, 2012) at 1:6000, goat anti-Isl1 (R&D #AF1837) at 1:1000, rat anti-OC2 (Clotman et al., 2005) at 1:40, guinea pig anti-OC3 (Pierreux et al., 2004) at 1:6000, goat anti-Sox1 (R&D # AF3369) at 1:500 and mouse anti-Sox2 (Invitrogen # 20G5) at 1:250. Following secondary antibodies were used: donkey anti-chicken/Alexa Fluor 488, donkey anti-goat/Alexa Fluor 594, donkey anti-goat/Alexa Fluor 488, donkey anti-goat/Alexa Fluor 647, donkey anti-mouse/Alexa Fluor 594, donkey anti-mouse/Alexa Fluor 488, donkey anti-rabbit/Alexa Fluor 594, donkey anti-rabbit/Alexa Fluor 488, donkey anti-rabbit/Alexa Fluor 647, donkey anti-rat/Alexa Fluor 594, donkey anti-rat/Alexa Fluor 488, donkey anti-guinea pig/Alexa fluor 488 purchased from Thermo Fisher Scientific or Jackson Laboratories and used at dilution 1:500.

Immunofluorescence or in situ hybridization images of cryosections were acquired on an epifluorescence microscope EVOS FL Auto Imaging System (Thermo Fisher Scientific), on a Zeiss AXIO Observer Z1 Inverted LED Fluorescence Motorized Microscope using ZEN Blue Zeiss software, on a Leica DM2500 microscope with a Leica DFC420C camera, on a confocal laser Scanning biological microscope FV1000 Fluoview with the FV10-ASW 01.02 software (Olympus) or on a confocal microscope Leica Stellaris 8 Falcon using Las X software. The images were treated with Fiji-ImageJ, Adobe Photoshop CC, or Las X office software to match the brightness and contrast with the observations.

Motor behavior tests were performed on Arid3c^+/+ control mice (two males and nine females) or Arid3c^−/− null mice (four males and five females) at approximately 2 months of age. All mice had a week of acclimation before starting experiments and were weighed each day. For physiocage testing, each animal was housed for 48 h in a metabolic cage in which its motor activity (horizontal and rearing) and its food and water intake were measured by weight or infrared sensors. For the open-field test, each animal was placed in a new environment (Plexiglass box, 60 cm square), and its general activity and motor behavior were analyzed for 20 min. The grip strength test, balance beam test, rotarod, and catwalk test were performed as previously described (Audouard et al., 2012).

For the quantifications of in ovo electroporation experiments, labeled cells were counted on 8–12 sections on each side of the spinal cord in three independent embryos using the count analysis tool of Adobe Photoshop CC software. Quantifications on the “electroporated” (targeted) side of the spinal cord were compared to quantifications on the “non-electroporated” (contralateral) side of the same embryo. For quantifications in mouse embryos, the different levels of the spinal cord (brachial, thoracic, and lumbar) were determined using immunolabeling for Foxp1, which enables the visualization of the lateral motor columns in brachial or lumbar regions. At E12.5, labeled cells were counted on both sides of the spinal cord on 4–11 sections at the brachial level, 6–12 sections at the thoracic level, and 6–13 sections at the lumbar level for each of the 5–6 embryos analyzed per genotype using the count analysis tool of Fiji-ImageJ software. At E14.5, labeled cells were counted on both sides of the spinal cord on 4–11 sections at the brachial level, 3–18 sections at the thoracic level, and 4–10 sections at the lumbar level for each of the 3–6 embryos analyzed per genotype using the count analysis tool of Fiji-ImageJ software. Quantifications in the Arid3c^−/− null mutant embryos were compared with quantifications of Arid3c^+/+ control embryos from the same litters. Statistical analyses and graph production were conducted using JMP Pro 17 software. For mouse embryo analyses, differences in cell numbers between two different groups were evaluated using either a Wilcoxon–Mann–Whitney’s non-parametrical test, a Welch’s t-test, or a Student’s t-test, depending on the normality and the homoscedasticity of the data. For motor behavior analyses, tests were performed on 9 Arid3c^−/− null mice (4 males, 5 females) and 11 Arid3c^+/+ control mice (2 males, 9 females). Statistical analyses and graph production were conducted using JMP Pro 17 Software. For the open-field, physiocage, balance beam, rotarod, grip strength test, and catwalk, differences between the two experimental groups were evaluated using either a Wilcoxon–Mann–Whitney’s non-parametrical test, a Welch’s t-test, or a Student’s t-test, depending on the normality and the homoscedasticity of the data. For the catwalk test, the calculated p-values were adjusted using the Benjamini–Hochberg p-value correction, and some parameters (print maximum area, print intensity, print width, and print length) were adjusted according to the weight of each animal in order to avoid any weight-related bias. In all statistical analyses, a p < 0.05 was defined as significant.

Given that our and other studies evidenced that approximately 25% of V2 INs remain to be characterized (Li et al., 2010; Baudouin et al., 2021), we performed an RNA-seq experiment to identify genes enriched in V2 INs during spinal cord development using embryos wherein tdTomato is specifically produced in all the V2 populations [Vsx1|tdTomato embryos; Figure 1 (Baudouin et al., 2021)]. Spinal cords of six E14.5 Vsx1|tdTomato embryos were dissected and dissociated. Dissociated cells were sorted by FACS to separate the tdTomato⁺ V2 INs from the tdTomato- cells to compare the transcriptome of V2 INs with that of all the other cells of the spinal cord, thereby identifying V2-enriched transcripts (Figures 1A,B). The sequencing results showed a significant enrichment in known V2 markers, including markers of V2 precursors (Sox14, Gata2) (Zhou et al., 2000; Karunaratne et al., 2002; Francius et al., 2014; Clovis et al., 2016), of V2a (Vsx2/Chx10, Lhx3 and Lhx4) (Ericson et al., 1997; Briscoe et al., 2000; Renaux et al., 2024) and of V2b (Gata3, Tal1, also known as SCL) (Karunaratne et al., 2002; Smith et al., 2002; Muroyama et al., 2005) INs with a fold-change superior to 67.8 (adj p ≤ 0.0001), which confirms their relevance (Figure 1C). Interestingly, we could also identify genes expressed in very small V2 IN populations such as Sox1 (V2c INs and dorsal V2 population) (Panayi et al., 2010; Baudouin et al., 2021) and Pax6 (V2-Pax6 INs) (Panayiotou et al., 2013), or present in V2 subpopulations like Onecut factors (Francius et al., 2013; Harris et al., 2019), which further validates our analysis (Figure 1C). Here, we decided to focus our study on Arid3c, which displayed a 22.6-fold enrichment in V2 INs versus the other cells of the spinal cord (Figure 1C, adj. p-value of 0.000301), and which had never been described in V2 INs yet.

To confirm the potential expression of Arid3c in V2 INs, we first performed in situ hybridization experiments. Using Vsx1|tdTomato embryos at E12.5 or E14.5, Arid3c expression could be detected in tdTomato⁺ cells, i.e., V2 INs (Baudouin et al., 2021). Arid3c seemed expressed in a ventrolateral subpopulation of V2 INs, located in the vicinity of the motor columns (Figure 2A). Arid3c expression could also be observed at E10.5 and E16.5 in the same location, although it seemed more restricted at later stages (Figure 2B). Coimmunofluorescence for Arid3c and tdTomato confirmed that all the Arid3c-producing cells also contained tdTomato, i.e., are V2 INs (Figure 3A). Although this is the first report of Arid3c expression in the CNS, published data suggested the expression of its two paralogs, Arid3a and Arid3b, in the embryonic spinal cord. Therefore, we also assessed the presence of Arid3a or Arid3b transcripts in this tissue. Arid3a was broadly expressed in the developing ventral spinal cord at E10.5 and E11.5 but its expression steadily decreased with time and became undetectable at E14.5, which indicates a more transient expression than Arid3c (Figure 2C). Arid3a also seemed to be transiently expressed in the dorsal root ganglia at E10.5 and E11.5 (Figure 2C). Regarding Arid3b, its expression at E10.5 and E11.5 seemed restricted to a population corresponding to MNs in the ventral spinal cord. It was also detected in the dorsal root ganglia, the notochord, and the presumptive sympathetic chain (Figure 2D). Starting from E12.5, its expression became weaker and more diffuse and was completely undetectable at E14.5 (Figure 2D). As for Arid3a, these results indicate a transient and less restricted expression of Arid3b than Arid3c. Altogether, these results show that the 3 Arid3 genes are expressed in the embryonic spinal cord, but only Arid3c is specific to V2 INs.

To assess in which V2 population Arid3c is produced, we performed immunofluorescence experiments for Arid3c and different V2 IN population markers. We never observed any colocalization of Arid3c with Chx10 (V2a), Shox2 (V2a/d), or Gata3 (V2b). Furthermore, despite their distribution in the vicinity of the motor columns, Arid3c⁺ cells did not contain Isl1, confirming that these are not MNs (data not shown). In contrast, Arid3c was detected from E12.5 onward in cells containing Sox1 located ventrally and laterally, corresponding to V2c INs (Panayi et al., 2010) (Figure 3B). However, the distribution of those two proteins was not entirely overlapping. Indeed, at E12.5, cells that contained only Arid3c (mean value of 7.2 cells/section, 45.4%), cells showing a co-labeling for Arid3c and Sox1 (4.7 cells/section, 29.3%) and cells producing only Sox1 (4.0 cells/section, 25.2%) could be identified. At E14.5, the proportion of cells containing only Arid3c decreased (2.9 cells/section, 29.0%) at the benefit of Sox1⁺ only cells (3.9 cells/section, 38.4%), while the proportion of Arid3c⁺Sox1⁺ cells remained similar (3.2 cells/section, 32.6%) (Figures 3B,C). At both stages, the proportion between these cell categories was similar between the different levels of the spinal cord (Figure 3C). Similarly, the total number of Arid3c⁺ cells remained constant between the different spinal cord levels, varying between mean values of 9.65, 11.65, and 13.57 cells/section at E12.5 at brachial, thoracic, or lumbar levels respectively, and between 6.88, 5.95, and 6.41 cells/section at E14.5.

This led us to raise three hypotheses regarding the identity of the Arid3c⁺ cells, which could either represent an intermediate differentiation state between V2b and V2c INs, i.e., cells having lost the expression of Gata3 and transiently expressing Arid3c before reactivating Sox1, or cells that correspond to another uncharacterized V2 IN population, or complement V2c INs to constitute a single functional V2 subset. To address these hypotheses, we analyzed the colocalization of Arid3c and Sox1 at later developmental stages. Even though the number of cells was very low at E16.5 and E18.5, we were still able to detect the same three cell categories (Figure 3B). In contrast, Arid3c was not detectable in the adult spinal cord (data not shown). Altogether, these results suggest that Arid3c identifies an embryonic population of V2 INs that has not been characterized yet but partly overlaps with the V2c INs characterized by the presence of Sox1. Whether these combined populations constitute a functional subset of V2 INs remains to be investigated.

The diversity of cell populations in the embryonic or postnatal spinal cord has been recently investigated in different scRNAseq studies (Delile et al., 2019; Osseward et al., 2021; Russ et al., 2021). In an attempt to characterize spinal Arid3c⁺ cells, we re-analyzed an embryonic spinal cord scRNAseq dataset (E9.5–13.5) to identify genes that could be specifically co-expressed with Arid3c in V2 INs (Delile et al., 2019). V2 subclustering and cluster transcriptome analysis unveiled a list of genes specifically enriched in Arid3c⁺ cells as compared to V2b-Gata3⁺ INs, V2a-Chx10⁺ INs, and the other cells of the spinal cord, although their expression level was lower than that of Arid3c (Figure 4A). GO analysis did not significantly associate these genes with specific pathways or biological processes (data not shown). To validate this list of genes and assess if some of these factors could constitute additional markers of this population, we addressed the distribution of corresponding proteins using immunofluorescence. However, none of the tested proteins were exclusively present in Arid3c⁺ cells, although some were enriched in this population. Indeed, a majority of Arid3c⁺ cells and V2c INs were characterized by the presence of OC1, and all of them were producing OC2, OC3, and Sox2 (Figure 4B). Regarding calcium storage proteins, Arid3c⁺ cells were not producing calbindin or calretinin (Figure 4C). Thus, no additional markers of the Arid3c⁺ cells could be identified. However, markers shared with V2c INs added to their similar location reinforce the hypothesis that Arid3c⁺ cells and V2c could represent a single functional V2 population.

Arid3c was detected in part of the V2c INs and in a restricted number of cells located in their direct vicinity, but not in V2a or V2b populations, suggesting that it may stimulate the differentiation of small V2 populations at the expense of the major ones. To test this hypothesis, we analyzed V2 differentiation in HH26 chicken embryos electroporated with an expression vector for Arid3c. Ectopic Arid3c expression alongside the dorsoventral axis of the spinal cord had no significant impact on the production of V2a or V2b INs (Figure 5; p = 0.2071 and 0.0505, respectively). Unfortunately, we could not analyze V2c INs differentiation as we observed that the electroporation process seems to disrupt the organization of some Sox1⁺ progenitors that are displaced laterally from the ventricular zone to the mantle layer, making it impossible to distinguish the Sox1⁺ V2c INs from these artifactually located Sox1 progenitors (data not shown). Nevertheless, these observations indicate that Arid3c is not sufficient to impose an alternative identity on major V2 IN populations.

Alternatively, Arid3c may be necessary for the proper production of smaller V2 IN populations, including V2c (Figures 3, 4). To address the potential roles of Arid3c in V2 IN development and in the formation of the locomotor circuits, we analyzed the phenotype of V2 INs in Arid3c^−/− null mutant mouse embryos (Supplementary Figure S1) at E12.5 and E14.5. The constitutive loss of Arid3c had no impact on the number of V2c INs, whatever the embryonic stage or the level of the spinal cord analyzed (Figures 6A,B). Previous studies suggested that V2c derived from V2b INs (Panayi et al., 2010) and our observations suggest that Arid3c⁺ cells relate to V2c (Figures 3, 4). We assessed whether the loss of Arid3c could impact the amount of V2b INs. However, we could not detect any change in the number of V2b cells regardless of the embryonic stage or the spinal cord level (Figures 6A,B). Consistently, the production of V2a INs was not altered (Figures 6A,B). Taken together, these observations suggest that Arid3c is not necessary for proper production of the V2c INs or for further subdivision of V2b or V2a INs.

Although the production of the described V2 IN populations was not altered in the absence of Arid3c, later steps of development could be affected, a question difficult to address given the absence of additional specific markers of the V2c and of the Arid3c⁺ INs. To get around this limitation and assess the functional output of the spinal motor circuitry that these cells likely participate in, we characterized the general locomotor behavior of the Arid3c^−/− null mutants. First, a physiocage and an open-field test were performed to analyze the general activity of these mice. Although the loss of Arid3c did not impact any of the parameters analyzed in the open field (total distance, velocity, duration, and frequency in the peripheral or central zones, latency before the first center exploration; Figure 7A and data not shown), the Arid3c^−/− null mice displayed an increased horizontal activity as compared to Arid3c^+/+ control mice (Figure 7B; p = 0.0008) and a trend to increased rearing (Figure 7B; p = 0.0822) that is not explained by an extra intake of food or water (data not shown). However, these observations point to a possible mild anxiety in Arid3c^−/− null mice (Seibenhener and Wooten, 2015), rather than an intrinsic locomotor defect. Consistently, their general motor coordination was comparable to that of Arid3c^+/+ control mice in the balance beam (latency to cross the beam and number of foot slips; Figure 7C) or rotarod tests (time of latency before the fall of the animal; Figure 7D), with similar performance improvement with repetitions (Figures 7C,D). As alterations in motor circuit activity can result in reduced muscle strength, this parameter was evaluated in a grip-strength test. However, both Arid3c^−/− null mice and Arid3c^+/+ control mice exerted similar forelimb or combined forelimb and hindlimb grip strength (Figure 7E). Finally, possible gait defects of Arid3c^−/− null mice were assessed using a catwalk test. The general locomotion of these mice (regularity index, base of support of the paws, print positions, percentage of support on one, two, three, or four paws) was not altered by the loss of Arid3c (Figure 7F, and data not shown). Consistently, the different locomotion phases that were analyzed (diagonal, girdle, or ipsilateral phases) were similar between mutant or control mice (Figure 7G). However, when looking individually at the front or hind paws, the maximum area and the width of the hind paw prints were increased in Arid3c^−/− null mice (Figure 7H; adj. p = 0.0044 and 0.0144, respectively). Similarly, the duty cycle, which expresses stance duration as a percentage of the duration of the step cycle of the hind paws, was higher in Arid3c^−/− null mice (Figure 7H; adj. p = 0.0112). Differences in the stand index (measure for the speed at which the paw loses contact with the glass plate), stand (duration of contact of a paw with the glass plate in a step cycle), swing (duration of absence of contact with the glass plate in a step cycle), and swing speed (speed of the paw during swing) of the hind paws were also observed in mice lacking Arid3c (Figure 7H; adj. p = 0.0112, 0.0196, 0.0112, 0.0208, 0.04438, respectively). Altogether, those results indicate that even though the general motor behavior of the mice was not impacted by the loss of Arid3c, some subtle aspects of the locomotor execution seem to be altered.