To generate Ctnnal1 mice, ES cell line RRJ603 containing the gene trap vector pGT2Lxf in intron 1–2 of Ctnnal1 gene (BayGenomics) was used for blastocyst microinjection and founder breeding following standard procedures. See Supplementary Material and Methods for genotyping primers. All experiments were preapproved by The University of Southern California Institutional Animal Care and Use Committee.

α-catulin E10.5 WT and KO embryos were collected in TRIzol (Invitrogen) and total RNA extracted using the RNeasy Micro Kit (QIAGEN). To perform semi-quantitative RT-PCR, RNA was reverse-transcribed to cDNA with SuperScript II RT (Invitrogen) using oligo dTs. For reverse-transcription, 2 μl of α-catulin 10.5E WT and KO embryo RNA was used. Primers are provided in Supplementary Material and Methods.

α-catulin embryos were isolated in cold PBS and fixed in 4% paraformaldehyde (PFA) on ice for 30 min. Embryos were then washed well 3x in PBS. Embryos were next allowed to sink in 20% sucrose overnight at 4°C. The following day, embryos were incubated in 30% sucrose: OCT for 2 h at RT on a gentle rocker, then stored at 4°C overnight. Embryos were then embedded in OCT and sectioned at 10 μM for indirect detection of various markers. Samples were fixed in 4% PFA for 10 min and subsequently permeabilized in 0.1% Triton X-100 in PBS (PBS-T) for 10 min. Next, samples were blocked in 0.1% BSA, 2.5% HI-GS, 2.5% HI-DS in 0.1% PBS-T for 30 min at RT. Primary antibodies were diluted in 0.1% BSA in 0.1% PBS-T and incubated overnight at 4°C. Alexa Fluor 488 or 594 secondary antibodies were diluted 1:500 in blocking solution and incubated 1 h at RT. Photographs were taken using AxioImager Z1 (Zeiss). Primary antibody descriptions and dilutions are described in Supplementary Material and Methods.

Embryos were fixed in 4% PFA for 10 min, washed well in PBS, ethanol dehydrated, embedded in paraffin and sectioned 6μM thick. Samples were than deparaffinized and pretreated using antigen retrieval 2100 Retriever (Proteogenix). Endogenous peroxidase was blocked in 0.03% hydrogen peroxide for 5 min, washed in 0.3% PBS-T, blocked in 0.1% gelatin, 2.5% HI-GS, 2.5% HI-DS, 0.1% BSA in 0.3% PBS-T for 1 h and incubated with primary antibody in 0.1% BSA in 0.1% PBS-T overnight at 4°C. After washing well in 0.3% PBS-T, biotin-conjugated secondary antibodies (Vector Laboratories) were diluted 1:100 in blocking solution and incubated 1 h at RT. The ABC Kit was then used following manufacturer’s instructions (Vector Laboratories). Staining was detected using DAB Peroxidase Substrate Kit (Vector Laboratories) following manufacturer’s instructions. Scanning electron microscopy was performed in the Cell and Tissue imaging Core of USC, according to standard procedures. Antibodies are described in Supplementary Material and Methods.

For whole mount β-galactosidase detection, embryos were isolated in cold PBS and immediately fixed in cold 0.2% glutaraldehyde for 20 min on ice. Embryos were then washed well in cold PBS and stained in x-gal staining solution [5 mM EGTA (pH 8), 2 mM MgCl₂, 0.2% NP-40, 0.1% sodium deoxycholate, 2 mM CaCl₂ – before use, 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, 1 mg/mL x-gal was freshly added] overnight at 37°C with gentle shaking. Staining was stopped by washing in PBS until solution was clear. For counterstaining, embryos were sectioned 8 μM thick and then counterstained with nuclear fast red.

MDCK (Madin-Darby Canine Kidney) cell line was purchased from the American Type Culture Collection (ATCC CRL-2936). MDCK cells were cultured in Dulbecco’s modified Eagles’s medium (DMEM) high glucose (Biowest #L0102-500) containing 10% fetal bovine serum (FBS) (Biowest, #S181S-500) and 1% antibiotics: penicillin (100 U/ml) and streptomycin (100 μg/ml) (Biowest #L0018-100). Cell line was grown at 37°C in humidified 5% CO2/95% air atmosphere. The number of living cells was calculated by Trypan Blue staining using EVE^TM Automatic Cell Counter (Nano EnTek, South Korea). Cell line was regularly tested for mycoplasma contamination using a PCR-based method (Young et al., 2010).

MDCK cells were seeded on 6-well plate in density of 70,000 cells and the next day were transfected with specific small interfering (si)RNA targeting α-catulin gene (siCTNNAL1) (HSS112822, Thermo Fisher Scientific) or with universal negative control siRNA (siNeg) (#12935200, Thermo Fisher Scientific) using Opti-MEM medium (Thermo Fisher Scientific, #11058021) and Lipofectamine RNAiMAX Reagent (Thermo Fisher Scientific, #13778100) according to manufacturer’s instructions. Before further experiments cells were cultured for 48 h. Effectiveness of silencing of α-catulin gene was confirmed by RT-qPCR. Independent MDCK cell lines were generated using lentiviral shRNA specific to canine α-catulin (KD1 and KD2) and non-effective shRNA scrambled cassette (control). Lentiviral cassettes were designed and cloned by OriGene (pGFP-C-shLentiviral vector, TR30023 and pGFP-C-shLenti-scrambled TR30021). Lentivirus was produced using Lenti-vpak Lentiviral Packaging Kit (Origene) according to the manufacturer protocol. GFP positive cells for stable cell lines were selected by FACS sorting.

To visualize active RhoA, MDCK cells were transfected with GFP-AHPH-DM (Addgene #71368) construct using Lipofectamine 3000 Transfection Reagent (Thermo Fisher Scientific #L3000001) according to manufacturer’s instructions. To visualize α-catulin protein, MDCK cells were transfected with construct of pEGFPC1 vector with cloned mouse α-catulin cDNA upstream of GFP using Lipofectamine 3000 Transfection Reagent (Thermo Fisher Scientific #L3000001) according to manufacturer’s instructions. Both constructs were verified through sequencing. Fluorescence was examined under confocal microscope LSM 700 (Zeiss) 24 h after transfection.

Total RNA was isolated from MDCK cell culture 2, 4, and 6 days after siRNA transfection using the RNeasy Mini Kit (Qiagen, #74106) according to manufacturer’s instructions. The concentration and purity of RNA samples were determined using DeNovix DS-11 Spectrophotometer. cDNA was synthetized from 1 μg of total RNA using the PrimeScript RT Master Mix (TAKARA BIO INC. #RR036A) in accordance with the manufacturer’s protocol. Level of expression of α-catulin gene was quantified by RT-qPCR. Each reaction mixture contained 1 μl of 25-times diluted cDNA, 6.25 μl of Fast SG qPCR Master Mix (2x) (EurX, #E0411) and 0,5 μM specific oligonucleotide primers (listed in section Supplementary Material and Methods) in a total volume of 12.5 μl. Reactions were performed in triplicates on the LightCycler 480 II (Roche) in the following conditions: 30 s at 95°C, followed by 45 amplification cycles (95°C for 10 s, 60°C for 10 s, 72°C for 10 s). Levels of expression of α-catulin gene were normalized to GAPDH. Primers are provided in Supplementary Material and Methods.

Single cell suspensions of 10,000 siRNA-treated MDCK cells were seeded on 8-well glass Millicell EZ slides (Merck Millipore, #PEZGS0816) pre-coated (37°C, 30 min) with collagen I solution (70%): matrigel (30%) (BD Matrigel Matrix, Growth Factor Reduced, BD Biosciences, #356231) mixture in the presence of 2% matrigel in culture media and cultured for 4 days. Collagen I solution was prepared as follows: 2 mg/ml collagen type I (Rat Tail High Concentration, Corning, #354249), 24 mM L-glutamine (Gibco Life Technologies, #25030-081), 20 mM HEPES (Alfa Aesar ThermoFisher, #J61047), 2.35 mg/ml Sodium Bicarbonate (Biowest, #L0680-500) and DMEM 1x.

Single cell suspensions of MDCK cells were seeded in density of 40,000 cells on 12 mm Nunc^TM Polycarbonate Cell Culture Inserts with 0,4 micron pore size in 12-well plate (ThermoFisher Scientific, #1406562) and grown for 4 days.

Cells for immunofluorescent staining were fixed in 4% PFA (Paraformaldehyde) for 10 min in room temperature and subsequently permeabilized in 0.1% Triton X-100 in PBS (PBS-T) for 10 min room temperature. Next, non-specific sites were blocked in 0.1% BSA, 2.5% NGS (Normal Goat Serum), 2.5% NDS (Normal Donkey Serum) in 0.1% PBS-T for 30 min at room temperature. Specific primary antibodies were diluted in 0.1% BSA in PBS-T and incubated overnight at 4°C. Secondary fluorochrome-conjugated antibodies were diluted 1:500 in blocking solution and incubated 1 h at room temperature in the dark. Primary and secondary antibodies descriptions are listed in section Supplementary Material and Methods. For actin visualization samples were incubated with phalloidin-TRITC (Fluka #77418) in dilution 1:200 for 1 h in room temperature in the dark. The cells were counterstained with nuclear dye 4’6-diamidino-2-phenylindole (DAPI; Calbiochem, # 268298) in dilution 1:1000 for 2 min in room temperature in the dark. Samples were mounted in FluorSave (#345789, Calbiochem). Images were acquired by a laser scanning confocal microscope LSM 700 (Zeiss) and analyzed with Zen2 (Blue and Black edition) Program (Zeiss). Primary antibody descriptions and dilutions are described in Supplementary Material and Methods.

To gain insight into the physiological function of α-catulin, we generated straight α-catulin knockout mice using embryonic stem (ES) cells with a gene trap vector, where β-galactosidase is inserted into the α-catulin locus (Bay-Genomic cell line RRJ603) (Figure 1A). We confirmed the correct insertion of the trapped vector in the α-catulin locus by sequencing cDNA isolated from ES cells (Supplementary Figure S1E) and genotyping (Figures 1B,C). α-catulin knockout mice die very early embryonically at day E10.5, with different severity of phenotype, which is likely due to the hypomorphic nature of the gene trap, as confirmed by sqRT-PCR using RNA isolated from the embryos (Figure 1D). However, those animals consistently show defects in NT closure (Figures 1F,G and as summarized in Figure 1E). Littermates lacking α-catulin exhibit massive malformation of the developing brain. The first obvious phenotype, a failure in neural tube closure, is clearly visible already at E10.5. In mice, the neural tube closure is initiated at the hindbrain/cervical boundary, at the midbrain/forebrain boundary, and at the most anterior part of the telencephalon from where closure spreads into the remaining areas of the neural tube (Nikolopoulou et al., 2017). Based on our observations, we suggest that α-catulin embryos might have a failure of the initiation point at the hindbrain/cervical boundary closure. NT closure defect is accompanied by a general increase in deformability at the tissue-level, especially in the hindbrain area (asterisk in Figure 1F). Loosely organized pattern of cells in open neural tube in the hindbrain area of α-catulin knockdown (KD) mice as compared to wild type (WT) littermate is observed in scanning electron microscopic images (Figures 1H,I,H’,I’). Utilizing the β-galactosidase gene trap vector, α-catulin expression can be studied by staining with x-gal since heterozygous α-catulin ± mice appeared normal. Surprisingly, the pattern of expression of α-catulin at early stages of the mouse development was very dynamic. Staining for x-gal, at E8.5, α-catulin is expressed in the neuroepithelium, at the level of specific rhombomeres in the hindbrain and along the neural tube as indicated by arrows (Figures 1J,J’). At E9.5, α-catulin is expressed in the hindbrain, the ectoderm of the branchial arches and nasal pits, as well as in the limb buds as indicated by arrows (Figures 1K,K’). At E10.5, α-catulin expression is also visible in the dermomyotome, developing eye, nasal pits and limb buds (Supplementary Figures S1A,A’,A”). Around day E14.5 α-catulin expression appears strongly around the eye, otic vesicle (arrows in Supplementary Figure S1B) and in the dorsal root ganglia (arrows in Supplementary Figures S1B,B’) where it persists after the birth (arrows in Supplementary Figures S1C,C’). In newborn (NB) mice, α-catulin is also expressed in the neural crest derived innervation (arrows in Supplementary Figure S1D). Taken together, this data reveal, for the first time, a crucial role for α-catulin during mouse NT closure, and highlights the dynamic expression pattern of α-catulin that can correspond to its importance in the plasticity of epithelial tissues during morphogenesis.

Actin and many of its associated proteins accompanied by apical accumulation of Rho and contractile actomyosin network, regulating apical constriction, were shown to play an important role in neurulation (Kinoshita et al., 2008). To determine the distribution of actin and keratin filaments in the neuroepithelium of α-catulin deficient mice, we stained coronal sections with phalloidin, nestin and p75 neurotrophin receptor specific antibodies. The neuroepithelium of α-catulin deficient mice showed lack of apical accumulation of actin (Figures 2E,E’,F,F’) and nestin filaments (Figures 2C,C’,D,D’). As seen by hematoxylin and eosin (H&E) staining, neuroepithelium of α-catulin deficient mice showed extra bending (arrow in Figure 2B) as compared to WT littermates (Figure 2A). As it was described previously, that α-catulin can act as a scaffold for Rho complex distribution (Park et al., 2002), we wanted to check if this observation is also relevant in vivo by staining hindbrain sections of α-catulin E10.5 WT and KD embryos with antibody for downstream target of active RhoA – P-Mlc2. The lack of apical accumulation of cytoskeleton and defects in neural tube bending and closure correlated with the lack of proper apical localization of P-Mlc2 in α-catulin-deficient neuroepithelium. P-Mlc2 appeared as a diffused signal extending through the broad area of α-catulin-deficient neuroepithelium (arrows in Figures 2H,H’) in comparison to WT (arrows in Figures 2G,G’). Since abnormal actomyosin accumulation might also affect the structure and/or function of apical junctional complexes, which are implicated in neuroepithelial morphogenesis (Nishimura and Takeichi, 2009) we also checked the localization of cell-cell junction component, β-catenin in those embryos. Immunostaining revealed proper localization of β-catenin, specifically localized to the lateral membranes of neuroepithelial cells with tight cell-cell junctions (Figures 2I,I’) in WT embryos. In contrast, α-catulin deficient embryos, despite strong β-catenin staining at the membranes, exhibited a slightly disorganized cell-cell junctions with less elongated cells and more rounded nuclei, suggesting defects in polarization of the neuroepithelium (Figures 2J,J’). In addition, some the cells were loosely separating from the neuroepithelium of α-catulin deficient embryos (arrows in Figures 2J,J’). This phenotype was quite common as it is also noticeable in neuroepithelium of α-catulin deficient embryos in Figures 2D’,N’ (asterisks), indicating instability of cell-cell junctions. High level of tissue disorganization in α-catulin-deficient mice also affected integrin-mediated basement membrane assembly of the neuroepithelium, as judged by the staining of extracellular matrix components, fibronectin and laminin, which were weakly expressed in α-catulin-deficient mice (Figures 2L,N, arrows in Figures 2L’,N’) as compared to WT littermates (Figures 2K,M, arrows in Figures 2K’,M’). Surprisingly, in some areas non-neuronal keratin 8 staining was observed in α-catulin-deficient mice thinner layer of neuroepithelium (Figure 2N, asterisks in Figure 2N’) which in WT littermates is only present in developing epidermis (Figure 2M, asterisk in Figure 2M’). Our initial observations of neuroepithelium architecture in α-catulin-deficient mice suggest that lack of proper apical actin accumulation with P-Mlc2 distribution might be essential in enabling apical constriction and adherens junctions remodeling as the neuroepithelium undergoes morphogenesis.

Since neural tube closure defects were consistent and predominant in α-catulin-deficient mice, we focused further analyses on the neuroepithelium. The neuroepithelium of normal E10.5 embryos consist of the ventricular zone (VZ) layer, containing proliferating neural precursors which express Sox2, and the mantle containing post-mitotic neurons which start to express neurofilament (NF). α-catulin-deficient mice show some degree of localized expression of Sox2 (arrows in Figures 3C,C’,F,F’) as compared to its wild type (WT) littermate (arrows in Figures 3A,A’,E,E’). Post-mitotic neurons which start to express neurofilament are also visible in α-catulin-deficient neuroepithelium (arrows in Figure 3D) as compared to WT littermates (arrow in Figure 3B). Despite high level of disorganization, which is even more obvious in the coronal plane section stained with Sox2 (Figure 3F), as compared to WT (Figure 3E), Sox2 staining still localizes in the ventricular layer with decreased expression in mantle layer (asterisks in Figures 3E,E’,F,F’).

The subcellular organization and dynamics of actomyosin network are challenging to characterize in developing embryos. Therefore, to get better insight into the role of α-catulin in the actomyosin coordination in epithelial cells, in parallel to our study on mice, we used MDCK canine kidney epithelial cells, that are widely used for studying cell polarity, actin organization and adhesion. The effect of suppression of α-catulin on the organization of epithelial cells in 3D organotypic culture was assessed in cysts formed from MDCK II cells transfected with siRNA targeting α-catulin gene (siCTNNAL1) and control (siNEG). The efficiency and stability of knockdown was verified by RT-qPCR, on days 2, 4, and 6 after siRNA transfection. The level of α-catulin in siCTNNAL1 MDCK cells was at least 95% lower in siCTNNAL1 than in control (siNEG) cells at all time points (Figure 4A). Deficiency of α-catulin resulted in impaired cyst formation (Figure 4C) with high level of cells disorganization, often forming multiple lumens (arrows in Figures 4C, 5C) or no visible lumens at all. To perform quantification, we counted cysts in both groups (siCTNNAL1 and siNEG) that contained 5 or more cells, naming “good,” cysts that contained regular, single layer and single lumen and “bad,” cysts with irregular or multiple layers, without lumen or with multiple lumens. In control siNEG cells “good” cysts constituted 88% and “bad” cysts only 12%, whereas in siCTNNAL1 cells proportion between “good” and “bad” cysts were 26% and 74% respectively (n = 100 for each group) (Figure 4B and Supplementary Figures S2A,B). Cell-cell contacts were assessed by the staining with E-cadherin specific antibody. In control cells, E-cadherin signal was present essentially at cell–cell contacts of cysts with a spherical monolayer (Figures 4C,D). After depletion of α-catulin, E-cadherin was still present at cell-cell contacts, however, staining was diffused in some areas and disorganized (Figure 4D). We further analyzed polarity status in these MDCK cysts by staining actin to visualize the apical domain. As expected, control MDCK cysts formed a central single lumen represented with Phalloidin staining, whereas α-catulin depleted MDCK cells formed multilumen cysts with less abundant apical actin that partially spread to the lateral domains (Figure 4D). Consistent data were obtained using independent shRNA stably expressed in MDCK cells using lentiviral transduction (Supplementary Figures S3, S4). To further investigate possible disruption in polarization in cysts formed by α-catulin deficient MDCK cells, we immunostained them with antibody specific for well-established apical marker, podocalyxin. We noticed proper apical accumulation of podocalyxin in cysts formed by control (siNEG) cells and limited accumulation of this protein in cysts formed by siCTNNAL1 MDCK cells (Figure 5C). In addition, MDCK cells transfected with siRNA targeting α-catulin gene (siCTNNAL1) and control (siNEG) were seeded on filters and 4 days after transfection podocalyxin was visualized localizing on the apical membrane of control (siNEG) monolayers (Figure 5A upper panel and Figure 5B left image). Removal of α-catulin resulted in diffuse localization of podocalyxin and lack of proper apical accumulation of this marker, visible in projection from above and in the side view (Figure 5A lower panel, Figure 5B right image). Consistent data were obtained using independent shRNA stably expressed in MDCK cells using lentiviral transduction (Supplementary Figure S5B). These findings indicate that depletion of α-catulin has a strong effect on cell polarity, with high number of cysts presenting multilumens. This is also in line with our observations of the neuroepithelium of α-catulin deficient mice that lacked apical accumulation of actin and despite β-catenin staining at the membranes, exhibited a markedly disorganized cell-cell junctions with some cells loosely separating from the neuroepithelium.

It was previously shown that apical accumulation of Rho in developing chick embryos undergoing folding of the neural plate during neural tube formation correlated with similar accumulation of activated myosin II (Kinoshita et al., 2008). The timing of accumulation and biochemical activation of both RhoA and myosin II coincided with the dynamics of neural tube formation. As our observation in mice and 3D cultures strongly suggested that α-catulin is an important mediator of apical accumulation of actin and signaling to actin/myosin cytoskeleton, we wanted to check if α-catulin could be acting as a scaffold for RhoA distribution. This could be potentially true since α-catulin was previously described as a scaffold for Lbc-RhoGEF affecting RhoA signaling (Park et al., 2002). First we tested intercellular localization of α-catulin in polarized cells. We used MDCK cells grown for 3 days on collagen-coated glasses, transfected with α-catulin-GFP expressing plasmid (efficiency of transfection shown in Supplementary Figure S6B), previously verified by Western Blot (Supplementary Figure S6A) since canine specific α-catulin antibody was not available. Cells were grown for additional day, followed by staining with anti-GFP antibody and phalloidin to visualize actin. Analysis of immunofluorescent staining of cell monolayers by confocal microscopy revealed strong co-localization of α-catulin and actin at the apical membrane spanning till the cell-cell junctions (arrowheads in Figure 6A). Z-stack analysis of the apical membrane clearly demonstrated co-localization of α-catulin with apical actin (arrowheads in Figure 6B-1) and some enrichment at the cell-cell junctions (asterisks in Figures 6B-2,B-3). Because apical localization and activity of RhoA is a central feature of establishing efficient contractile actinmyosin network, we next investigated if depletion of α-catulin would affect this activity. In order to analyze the distribution of active RhoA in α-catulin depleted polarized MDCK cells, 2 days after transfection with siRNA targeting α-catulin gene (siCTNNAL1) and control (siNEG), cells were trypsynized, counted and seeded on filters and grown for 3 days, then they were transfected with previously described GFP-AHPH sensor known to bind only to active RhoA (RhoA-GTP) and grown for additional day (Tse et al., 2012; Priya et al., 2015). After that, cells were stained with phalloidin to visualize actin. We examined GFP signal that indicated active RhoA localization in polarized α-catulin deficient and control MDCK cells and noticed specifically localized active RhoA on apical membrane in control cells (arrowheads in Figure 6C) and its random distribution (present also in basolateral membrane) in α-catulin depleted MDCK cells (arrowheads in Figure 6D). Z-stack analysis of the apical membranes of α-catulin control and deficient cells strongly demonstrated exclusive apical accumulation of active RhoA in control cells and random distribution throughout the α-catulin deficient cell (arrowheads in Figures 6E,F). Our data suggest that α-catulin might be a scaffold for active RhoA localization to the cell apex.

Since polarization of epithelial cells and distribution of apical proteins is dependent on proper cell-cell junction formation and stabilization, we wanted to get better insight into the organization of adherens junction in α-catulin deficient cells. MDCK cells transfected with siRNA targeting α-catulin gene (siCTNNAL1) and control (siNEG) were seeded on filters for 4 days, followed by immunostaining with antibody specific for E-cadherin and stained with phalloidin to visualize actin. Confocal microscopy analysis in xy axis (projection from above) and xz (side view) of stained cells monolayer revealed disrupted formation of monolayer in α-catulin deficient cells, despite preserved cell-cell contacts as judged by E-cadherin staining (Figures 7A,B). In control, MDCK cells formed single-cell monolayer, with equal cell sizes and even, flat apical surface of monolayer. In contrast, α-catulin deficient MDCK cells created irregular, “bumpy” layer, with cells often overlapping each other (arrowheads in Figure 7A lower panel, Figure 7B right panel and asterisks in Figure 6D), indicating lack of stabilization in cell-cell contacts. Consistent data were obtained using independent shRNA stably expressed in MDCK cells (Supplementary Figure S5A). In order to examine an intercellular localization of α-catulin, we used MDCK cells grown on collagen-coated glasses, transfected with α-catulin-GFP expressing plasmid (previously verified by Western Blot Supplementary Figure S6A) followed by phalloidin staining to visualize actin. α-catulin-GFP was enriched at the cell-cell junctions (Figure 7C and arrowheads in enlarged boxes marked by double star). However, more pronounced localization was observed along the filamentous actin (Figure 7C and arrowheads in enlarged boxes marked by single star). Accumulation of α-catulin-GFP was not so dramatic as other typical cell-cell junction proteins, e.g., αE-catenin. We next assessed the distribution of adherens junction components in flat cultures of α-catulin deficient and control cells. Co-staining of E-cadherin and αE-catenin revealed specific localization of both proteins to adherens junctions in control MDCK cells, showing equal distribution of junctional complexes between cells (arrowheads in Figure 8A upper panel). Actin cytoskeleton was properly organized in control cells, with enrichment at the membrane and actin cables spanning throughout the cells (arrowheads in Figure 8A upper panel). In the absence of α-catulin, we observed unequal distribution of E-cadherin and αE-catenin with some areas enriched with those proteins at the junctions (arrows in Figure 8A lower panel), some with slightly diffused pattern near the membrane (arrowheads in Figure 8A lower panel) and some lacking the staining at all (asterisks in Figure 8A lower panel). Actin cytoskeleton in α-catulin deficient MDCK cells was not properly organized with limited accumulation at the junctions as an actin belt, that would indicate proper junctions stabilization (arrowheads in Figure 8A lower panel). Analysis of active P-Mlc2, downstream target of RhoA, indicated that depletion of α-catulin in MDCK cells resulted in lack of proper localization of active P-Mlc2, which could explain defects in the stability of adherens junctions. In the α-catulin deficient cells, P-Mlc2 staining was strong predominantly in the areas where cell-cell junctions were formed, as indicated by E-cadherin localization (arrows in Figure 8B lower panel). In some areas, P-Mlc2 staining was slightly diffused at the membranes which also correlated well with E-cadherin staining (arrowheads in Figure 8B lower panel). Most of the staining appeared as a diffused pattern throughout the cells without noticeable actomyosin cables (asterisks in Figure 8B lower panel). In contrast, control cells showed well organized P-Mlc2 staining, enriched at the junctions (arrowheads in Figure 8B upper panel) and forming actomyosin network of cables (arrows in Figure 8B upper panel). Analysis of cell–cell contacts revealed, that although adherens junction components were mostly present at the junctions of α-catulin deficient cells; however, they were not properly organized, which is consistent with the observation in mouse embryos (Figures 2I,J). Because the linkage with dynamically regulated actomyosin network is required for strong hemophilic interactions of cadherins, the loss of α-catulin might affect the integrity of apical junctions. It is important to note that the lateral membrane contacts are still formed in α-catulin depleted cells although not stabilized. This indicated that the loss of α-catulin might affect the integrity of apical junctions not by influencing junction formation directly, but rather by regulating active RhoA distribution, followed by proper P-Mlc2 localization at the cytoskeleton. Importantly, the lack of apical accumulation of cytoskeleton and defects in neural tube bending and closure correlated with the lack of apically localized P-Mlc2 in α-catulin-deficient neuroepithelium (Figures 2H,H’) as compared to the WT littermate neuroepithelium (Figures 2G,G’).