Introduction

Development of the central nervous system occurs by proliferation of neural progenitor cells, migration and differentiation of their progeny over substantial distance. These different steps require extensive membrane remodelling and cytoskeleton dynamics¹. Neuronal migration involves the coordinated extension and adhesion of the leading process along the radial glial scaffold with the forward translocation of the nucleus, which requires regulation of centrosome and microtubule dynamics by different proteins^1–3. However, little is known about the molecular mechanisms underlying membrane dynamics during neuronal migration and morphogenesis.

Cell membrane curvature is a micro-morphological change involved in many important cellular processes including endocytosis, exocytosis, and migration⁴. Recent studies demonstrated that members of an extended protein family, characterized by the presence of a membrane binding and deforming BAR (Bin–Amphiphysin–Rvs) domain function at the interface between the actin cytoskeleton and plasma membrane during the formation of membrane protrusions or invaginations^5–7. These proteins have been shown to be important for proper neuronal morphogenesis⁸, and can either generate positive membrane curvature to facilitate the formation of plasma membrane invaginations (e.g. BAR, N-BAR and F-BAR domain proteins), or induce negative membrane curvature to promote the formation of plasma membrane protrusions (I-BAR and IF-BAR domain proteins)^9,10. An important part of the membrane deforming mechanism include the interaction with phosphatidylinositol-4,5-biphosphate PI(4,5)P2-rich membranes through their BAR domain and also with actin cytoskeleton through Wasp Homology-2 domain (WH2).

Regulation of the actin cytoskeleton plays a crucial role in several developmentally important cellular processes including neuronal migration, axonal and dendritic extension and guidance, and dendritic spines formation¹¹. We previously discovered that MIM (Missing-In-Metastasis; Mtss1), a member of the I-BAR protein family, has an important role in bending the dendritic plasma membrane in order to form dendritic spines and sculpt their morphology¹². In a more recent study, a closely related protein Abba (Mtss1L) was found to be important for exercise induced increase synaptic potentiation in the hippocampus¹³. Although an important role of MIM and Abba in synapse formation and plasticity starts to emerge, information about the role of the early expression of this protein family during development is poor.

Mutations in BAR domain proteins are increasingly found in mental retardation disorders. For instance, altered expression and mutation of F-BAR family proteins, SrGAP2 & 3, are known to be associated to neurological disorders (mental retardation, the 3p syndrome, schizophrenia, early infantile epileptic encephalopathy and severe psychomotor delay^14–18). Furthermore, previous whole-exome sequencing of pre-screened multiplex consanguineous families identified ABBA as a likely candidate linked to neurogenic disorders¹⁹, suggesting that some BAR containing proteins might play important functions during brain development. Despite Abba’s potential role in neuronal development, very little is known about how this protein may lead to neurological pathology.

Regulation of the actin cytoskeleton and plasma membrane dynamics in glial cells of the developing brain has remained poorly understood. Interestingly, the I-BAR protein Abba is the only glia-enriched regulator that has been reported so far²⁰.

We sought to determine the consequences of altered expression of Abba in radial glial progenitor (RGP) cells on cell cycle progression and neurogenesis and to test for subsequent neuronal migration. To address these issues, we used in utero electroporation to silence Abba with shRNA in embryonic mouse brains. Downregulation of Abba’s expression had a striking effect on RGP cells progression, resulting in a blocking of cytokinesis and a subsequent apoptosis. Biochemical analysis demonstrated that the cytokinesis block was due to an absence of interaction with the adhesion docking protein named Nedd9 (aka CAS-L or HEF1) that is well known to be a positive regulator of RhoA signaling in mitosis²¹. Using FRET based monitoring of RhoA activity, we also found a significant role of Abba on RhoA activity. Moreover, we observed a defect in neuronal migration associated with an increase in neuronal apoptosis. Importantly, we identified one patient harboring a missense variant in ABBA sequence and showing specific features such as: small head circumference, mental retardation, autism spectrum disorder (ASD) and craniofacial dysmorphism. We provide evidence that the (2011C>T(R671W)) patient mutation likely represents a dominant loss-of-function allele.

Overall, our data highlights the critical role of Abba in proper cortical development in neurogenesis and migration, and provides the first evidence that ABBA is a new causative gene for cortical malformation associated with ASD.

Materials and Methods

Plasmids construction

We utilized a single short hairpin RNA (shRNA) that targeted the coding sequence for Abba20. It is worth noting that the rat and mouse Abba sequences share a high degree of similarity, with a 95% homology. The specific region targeted by the shRNA displayed 100% homology. Both Abba and Nedd9 shRNA were designed using the siRNA Wizard Software (InvivoGen ©). As a negative control, we employed a non-targeting shRNA.

These shRNAs were subcloned into an mU6pro vector²². Additionally, a pCAG-RFP vector was co-injected to facilitate the visualization of fluorescent cells. Abba mouse full length with silent mutations²⁰, ABBA human full length and human ABBA-R671W carrying vectors were made by GeneCust© (France) and subcloned into the pCAGIG-IRES-GFP vector (Addgene, Cambridge, MA, USA).

Bioinformatic methods

ABBA coding sequences were obtained from NCBI. The multiple sequence comparison was done by Log-Expectation (MUSCLE) method and visualized with DNASTAR MegAlign Pro. The impact of ABBA R671W on the structure and function of a human ABBA protein was predicted using PolyPhen-2 and HumVar model²³.

Human subjects

Families were identified through our clinical and research programs, personal communication, as well as the MatchMaker Exchange (MME) including GeneMatcher²⁴ (http://www.genematcher.org). Informed consent for publication and analysis of photos, imaging and clinical data was obtained from the patients’ legal guardians. Brain magnetic resonance imaging (MRI) studies were performed on the subject and reviewed by the investigators in accordance with Department Ethical permission number (MEC-2012387) to conduct human exome studies.

RhoA activity assay

Fluorescence from Raichu-Rac1 was imaged using a confocal microscope (LSM 710/Axio; Carl Zeiss) controlled by ZEN 2011 software (Carl Zeiss) and equipped with water immersion objective (63×/1.0 NA; Carl Zeiss). Transfected cells were illuminated with an argon laser (Lasos; Lasertechnik GmbH) at 458 nm (CFP and FRET) and 514 nm (YFP) using 0.5–10% of full laser power. Emission was collected at 461–519 nm (CFP) and 519–621 nm (FRET and YFP). The pinhole was fully opened. Scanning was performed in XY mode using 6× digital zoom that resulted in a pixel XY size of 68 × 68 nm. FRET signal was calculated as reported elsewhere²³. In brief, using ImageJ 1.48v software images where background subtracted and FRET image generated using three images: MDonor (CFP excitation and CFP emission filters), MIndirectAcceptor (CFP excitation and YFP emission filters), and MDirectAcceptor (YFP excitation and YFP emission filters), using the following equation: FRET = (MIndirectAcceptor − MDonor × β − MDirectAcceptor × (γ − αβ))/(1 − βδ). Coefficients α, β, δ, and γ were obtained by independent control experiments analyzing cells expressing CFP or YFP as described previously²⁵. Coefficients α, δ, and γ were determined in cells expressing only YFP (acceptor) and calculated according to equations: α = MDonor/MDirectAcceptor; γ = MIndirectAcceptor/MDirectAcceptor; δ = MDonor/MIndirectAcceptor. Coefficient β was determined in cells expressing only CFP (donor) and calculated according to the equation: β = MIndirectAcceptor/MDonor. After generation of FRET image, intensity of FRET signal was calculated using a mask obtained from a corresponding YFP image and covering neuronal cell bodies. The images were also analysed using Pix-Fret²⁶

In Utero Electroporation

Animal experiments were performed in agreement with European directive 2010/63/UE and received approval N°: APAFIS#2797 from the French Ministry for Research and in agreement with the National Animal Experiment Board, Finland (license number ESAVI/18276/2018).Plasmids were transfected using intraventricular injection followed by in utero electroporation^27,28. Timed pregnant C57BL6N mice (Janvier Labs; E14; E0 was defined as the day of confirmation of sperm-positive vaginal plug) received buprenorphine (Buprecare,0.03 mg/kg) and were anesthetized with sevoflurane/isoflurane (4%-4.5% induction, 2%-anaesthesia maintenance) 30 minutes later. For pain management, bupivacaine (2 mg/kg) was administered via a subcutaneous injection at the site of the future incision. Uterine horns were exposed, the lateral ventricle of each embryo was injected using pulled glass capillaries and a microinjector (Picospritzer II; General Valve Corporation, Fairfield, NJ, USA) with Fast Green (2 mg/ml; Sigma, St Louis, MO, USA) combined with the different DNA constructs at 1.5 μg/μl or with 1.5 μg/μl shRNA constructs together with 0.5 μg/μl pCAGGS-RFP. Plasmids were further electroporated by discharging a 4000 mF capacitor charged to 50V with a BTX ECM 830 electroporator (BTX Harvard Apparatus, Holliston, MA, USA). The voltage was discharged in five electrical pulses at 950ms intervals via 5mm tweezer-type electrodes laterally pinching the head of each embryo through the uterus. The mice were followed-up post-surgically and given buprenorphine (Buprecare 0.03 mg/kg) if needed. The number of animals has been calculated on the basis of the requirement for adequate numbers of brain slices and sections for sufficient imaging to provide statistically significant data on the effects of RNAi, small molecules, and other reagents used in the proposed analysis, plus controls.

Organotypic cultures and Live Imaging

To assess the impact of altered ABBA expression in the RGP cells on neurogenesis and cell cycle progression, we utilized a combination of in utero electroporation and time-lapse imaging of organotypic cultures. The organotypic cultures were prepared as described²⁸ with the following exceptions. After surrounding the brains (E17) with 4.5% UltraPure Low Melting Point Agarose (Invitrogen, 16520-100), 350 μm thick slices were cut with Vibratome 1000 Plus Sectioning System (No. 054018). Before placing the slices on a coated (83 µg/ml Laminin, Sigma L2020; 8,3 µg/ml Poly-L-Lysine, Sigma P4707) cell culture insert (Millicell, PICMORG50), sections were carefully cleaned from agarose to ensure adequate oxygenation. The inserts with slice cultures were transferred into a 6-well plate containing 1 ml of warmed slice culture medium. For live-imaging we used Andor Dragonfly 550 high speed Spinning Disc confocal equipped with Nikon Perfect Focus System, motorized XY stage, and environmental chamber. The system uses Fusion 2.0 software (version 2.1.0.48). The image acquisition was executed using Andor iXon 888 U3 EMCCD camera, 20x magnification with voxel size 0.65 x 0.65 x 0.9524 micron^3 (channel/excitation: RFP/561 nm, GFP/488 nm). Time lapse-imaging of organotypic cultures was started at E17+DIV0 with intervals of 10 minutes and total duration of 15 hours. The data was collected from 6 pups (2 scramble, 4 shRNA), 17 slices (5 scramble, 12 shRNA) and 24 areas (9 scramble, 15 shRNA; some slices were recorded from two different areas).

Immunostaining of Brain Slices

Mice brains were fixed (E17) in Antigen Fix (Diapath), Brain slices were sectioned coronally (80μm) on a vibratome (Leica microsystems). Brain slices were washed with PBS (Phosphate-buffered saline pH 7.4) and stained in PBS 0.3% Triton X-100 supplemented with 5% of donkey serum. Primary antibodies were incubated overnight at 4°C, sections were then washed with PBS and incubated in secondary antibodies for 2 hours at room temperature. Antibodies used in this study were: Abba (homemade), Vimentin (Millipore, MAB3400), Ki67 (Millipore, AB9260), phospho-histone H3 (Abcam, ab14955), Pax6 (biolegend, PRB-278P).

Cells culture and transfection

Rat C6 glioma cells were cultured at 37°C under a humidified atmosphere with 5% CO2 with a complete medium DMEM supplemented with 10% foetal bovine serum (FBS, Sigma) and 100 units/mL antibiotics/antimycotics. Cells were transfected using the Neon® Transfection System (Life Technologies) according to the manufacturer’s protocol. Briefly, cells were trypsinized and counted with the cell counter SCEPTER (Millipore). Electroporation was performed with 500 000 cells and a total amount of 5 µg of DNA containing a reporter plasmid encoding RFP (1:5) and shRNAs or mismatch constructs, with the following configuration: 1860 V, 1 pulse, 20 ms. Cells were then cultured on a 6 wells plate during 2 days before RNA extraction.

Immunocytochemistry

HEK293T cells were fixed in Antigen Fix (Diapath), for 15 min and blocked at room temperature for 1 h with 5% normal goat serum,0.3% Triton X-100 in PBS and incubated overnight at 4°C with a set of antibodies against Abba (homemade), Phalloidin (Thermo Fisher, A12379), Nedd9 (Abcam, ab18056), Hoechst (Thermo Fisher, 62249). Cells were then washed with PBS and incubated in secondary antibodies for 2 hours at room temperature.

RT-PCR and qPCR

Total RNA was isolated from C6 using RNeasy Plus Mini kit and cDNA was synthesized using the Quantitect Reverse Transcription kit, according to the manufacturer’s protocol (QIAGEN). Quantitative PCR (qPCR) was performed on a Light cycler 480 using SYBR-Green chemistry (Roche) and specific primers for Abba (F: CGAGACTCGCTGCAGTATTCC and R: CCATTCACAGAGTAGCAGTCG, 113 bp), Nedd9 (Qiagen, QT01610385) and CyclophilinA (Qiagen, QT00177394) as a control probe. qPCR was performed with 5 µL of diluted cDNA template, specific primers (0.6 µM) and SYBR Green I Master Mix (7.5 µL) at a final volume of 15 µL. Each reaction was performed at an annealing temperature of 60°C and for 50 cycles. Reactions were performed in duplicate and melting-curve analysis was performed to assess the specificity of each amplification. A standard curve was performed for each gene with a control cDNA diluted at different concentrations. Relative expression was assessed with the calculated concentration in respect to the standard. All experiments were performed in triplicate.

Microscopy and Image Analysis

All images were collected with an IX80 laser scanning confocal microscope (LSM 710/Axio; Carl Zeiss) controlled by ZEN 2011 software (Carl Zeiss)). Cells & Brain sections were imaged using a 60x 1.42 N.A. oil objective or a 10x 0.40 N.A. air objective. All images were analyzed using ImageJ (NIH, Bethesda, MD). To quantify the total number of C6 cells alive and in mitosis on 12mm diameter coverslips 2 & 3 days after Abba shRNA & control transfections, we have arbitrary designed 4 identical regions comprised of 9 squares using ZEN software acquisition system and count manually the total number of alive and mitotic cells in each condition. All cell quantifications in brain slices were initially normalized to electroporation efficacy.

Statistical data

Statistical analyses were performed with Prism (GraphPad Software, La Jolla, CA, USA). A two-sample Student’s t-test was used to compare means of two independent groups if the distribution of the data was normal. If the values come from a Gaussian distribution (D’agostino-Pearson omnibus normality test) the parametric unpaired t-test with Welch’s correction was used. But, when the normality test failed, the non-parametric Mann–Whitney test was used. Significance was accepted at the level of p < 0.05.

No statistical methods were used to predetermine sample sizes, but our sample sizes are similar to those generally employed in the field. No randomization was used to collect all the data, but they were quantified blindly.

Yeast two-hybrid screen

Yeast two-hybrid screening was performed by Hybrigenics Services (Paris, France). The Abba coding sequence containing amino acids (aa) 1-715 was cloned into pB27 (N-LexA-bait-C fusion) and pB66 (N-GAL4-bait-C fusion) and sequence verified. These constructs were used as baits to screen the mouse embryo brain RP2 prey library. Total of 91 million (pB27) and 31 (pB66) million interactions were analyzed and the detected interactions were assigned with statistical confidence score, the Predicted Biological Score (PBS) (www.hybrigenics-services.com).

Co-immunoprecipitation

C6 cells or cortical tissue were washed with ice cold PBS and lysed using lysis buffer (20 mM Tris-HCl [pH 8,0], 137 mM NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA) with 50 µg/ml PMSF (Roche) and a protease inhibitor cocktail (Roche). Cell lysate was agitated for 30 min at 4°C and centrifuged at 12000 rpm for 20 min at 4°C. The supernatant was collected and protein concentrations were determined with a Pierce™ BCA Protein Assay kit, according to the manufacturer’s protocol (Thermo Scientific).

Before Co-IP, lysates were precleared with washed Pierce Protein A/G Magnetic Beads (Thermo Scientific) overnight in gentle shaking at 4°C, to remove nonspecific binding. Beads used for Co-IP were washed and pre-blocked with 3% BSA in PBS overnight in gentle shaking at 4 °C. Washing steps were done using a wash buffer (1% BSA, 150 mM NaCl, 0,1% Tween-20 in TBS).

For Co-IP, 1000 µg of protein from precleared lysate was incubated with 20 µl of specific primary antibodies (Abba: homemade; Nedd9: Abcam, ab18056) overnight in gentle shaking at 4°C. Negative controls were incubated without primary antibodies under the same conditions. Immunocomplexes were pulled down by incubating the antigen-antibody mixture with pre-blocked beads for 1 h in gentle shaking at room temperature. Beads were washed 3 times with wash buffer and once with purified water before elution. Bound proteins were eluted from the beads by incubating them in 1 x Laemmli buffer containing 150 mM NaCl for 10 min at room temperature.

Original C6 cell lysate and immunoprecipitated proteins were resolved by SDS-PAGE (4-20% Mini-PROTEAN TGX Gel; Bio-Rad) and transferred using Trans-Blot Turbo Mini PVDF Transfer Packs and Transfer System (Bio-Rad). Membranes were blocked with TBS containing 0,1% Tween-20 and 5% milk and blotted with primary antibodies overnight in shaking at 4°C. HRP-conjugated secondary antibodies used in this study were Rabbit anti-mouse IgG and Goat anti-rabbit IgG (Invitrogen). Protein bands were visualized using the ChemiDoc™ MP imaging system (Bio-Rad).

Flow-cytometry

To measure cell cycle length, 4×105 cells were seeded in 6-well plates, transfected and cultured for 24h. Plasmid transfection was performed with jetPEI by following the Batch HTS protocol suggested in the data sheet of the product. Briefly, transfection complexes, which consist of both DNA and jetPEI reagent diluted in NaCl, is mixed with the suspension of cells obtained after trypsinization and distributed in the wells for culture over 24h. Then, cell cycle synchronization was induced by 24h serum starvation and re-started using fetal calf serum supplemented DMEM. The cells were detached with tripsin+EDTA and fixed with 3.7% formaldehyde after 24 hours in culture. 5×105 fixed cells from each stop point were stained in 200ul of Hoechst 33342 (1:2000 dilution) for 30min, washed with PBS 1X and diluted in flow buffer. Acquisition of 100 000 total events was performed in NovoCyte Quanteon 4025 and the analysis performed in (FlowJo v10.8 or NovoExpress 1.6.1) software. Graphs and statistical analysis were performed in GraphPad Prism 5.

Results

Abba knockdown disrupted radial glia morphology

Previous studies of the expression pattern of Abba in the developing brain has disclosed a conspicuous expression in radial glia. In order to investigate the function of Abba in the developing brain, we used a mouse in utero RNAi approach to knockdown Abba at E14. For this purpose, we opted for shRNA knockdown using a previously validated targeting sequence²⁰. Revalidation of the knockdown efficiency of this target with three different Abba-shRNA (Abba-shRNA1-3) in rat C6 glioma cells showed a 70– 80% reduction of Abba mRNA and protein levels with Abba-shRNA3 (Supplemental Figure 1, Abba-shRNA3).

To determine whether knockdown of Abba expression alters neuronal migration, Abba-shRNA3 combined with a red fluorescent protein (RFP) construct was introduced into neural progenitor cells of mouse neocortex by in utero electroporation at E14 and evaluated at E16-18 (Figure 1A). Electroporation of E14 targets the progenitors that will migrate and populate cortical layers II/III neurons. In E17 brain sections, we observed that some neurons previously electroporated at 3 days earlier with the scrambled construct reached the cortical plate as expected (Figure 1B). However, in utero expression of Abba-shRNA3 induced a significant arrest of cells within the ventricular zone (VZ) (Figure 1A-C; scramble: 24,3 ± 6.4%; n=9, Abba-shRNA3: 63,7 ± 2,9%; p<0.0001; n=8). Surprisingly, we also observed a striking decrease in the total number of RFP positive cells in brain sections electroporated with Abba-shRNA3 at E17 and E18 (Figure 1D-E; scramble: 969,2±109.0 cells/mm²; n=9, Abba-shRNA3: 311,3±28,61 cells/mm²; p<0.0001; n=8). These results indicate an important role of Abba in the proliferation and migration of cortical neuronal progenitors.

Figure 1.

We also evaluated the impact of Abba knockdown on radial glial morphology by imaging vimentin-stained Abba knockdown cells at 3 days post-electroporation. All E17 Abba knockdown brains, but not in age-matched controls, displayed disruption of radial glial fibers in the VZ/SVZ (Figure 1D and F). Control vimentin-positive cells displayed the characteristic features and polarized morphology of radial glia with processes extending from the VZ to the pial surface. In contrast, in Abba knockdown brains, vimentin-positive cells exhibited misoriented apical processes, detached end feet in the ventricular surface (Figure 1D) and decreased number of radial glial progenitor cells in the VZ (Figure 1F; scramble: 126,5 ± 13,4 cells, n=8; Abba-shRNA3: 65,1 ± 26,2 cells, n=7; p=0.0033). All together these data show that the loss of Abba can disrupt radial glial cells morphology thus affecting their progenitor function.

These data suggest that Abba is required for radial glial organization, neuronal migration, regulation of neuronal proliferation and cell survival. Importantly, these results are in accord with the reported association of ABBA with developmental neurological diseases including intellectual disability and microcephaly¹⁹.

ABBA is required for cell cycle progression of radial glia progenitors

Development of the cerebral cortex occurs through a series of stages, beginning with RGPs. These stem cells exhibit an unusual form of cell-cycle-dependent nuclear oscillation between the apical and the basal regions of the ventricular zone, known as interkinetic nuclear migration (INM). RGP cells undergo mitosis only when they reach the SVZ^30,31. It has been shown that inhibition of INM disrupts RGP cell cycle progression^32–34.

To test the effect of altered radial glial cell morphology on cell cycle progression, we have quantified the nuclear distance from the VS for RGP cells after electroporation of scrambled or Abba RNAi (Figure 2A). We observed an accumulation of Abba-shRNA3 RGP cells close to the ventricular surface (0-10 mm; scramble: 30.2 ± 3.2%, n=7; Abba-shRNA3: 38.3 ± 2.3%, n=7; p= 0,000156) which can correspond to mitotic cell, but also further away (>30 mm; scramble 12,0 ± 3.3%, n=7; Abba-shRNA3: 21,9 ± 2.4%, n=7; p= 0,000023) which support our previous observations on cells detaching from the VS (Figure 1D and F). Furthermore, we observed that Abba-shRNA3 decrease the percentage of the RGP cells marker Pax6 (Figure 2B and C; scramble: 31,9 ± 0,6%, n=8; Abba-shRNA3: 21,1 ± 0,6%, n=5; p=<0.0001). We have then tested for cell cycle effects, and found a decrease on the percentage of Abba shRNA3-expressing RGP cells positive for the cell cycle marker Ki67 (Figure 2D and E; scramble: 46,5 ± 0,8%, n=7; Abba-shRNA3: 23,7 ± 1,369%, n=5; p= <0.0001) and also for the late G2/M phases marker phospho-histone 3 (PH3, Figure 2 F and G; scramble: 4,5 ± 0,1764%, n=9; Abba-shRNA3: 1,264 ± 0,5191%, n=5; p= <0.0001). More detailed examination of the impact of Abba-shRNA3 on cell cycle progression using flow cytometry showed an accumulation in S phase. Interestingly, these effect was rescued by overexpression of Abba-FL (Supplemental fig 2 G1: scramble: 45,2 ± 1,4 %, Abba-shRNA3: 37,4 ± 1,0 %, Abba-shRNA3+Abba-FL: 42,8 ± 1.3 %; S: scramble: 34,3 ± 2,8 %, Abba-shRNA3: 48,2 ± 2.7 %, Abba-shRNA3+Abba-FL: 38,9 ± 3.4; G2: scramble: 17,1 ± 2,1 %, Abba-shRNA3: 14,4 ± 2,7%, Abba-shRNA3+Abba-FL: 15,9 ± 1,9 %; n=10; p=<0.001). These results indicate that Abba plays a major role in coordinating cell cycle progression in radial glia.

Figure 2.

Abba localizes to the cleavage furrow and is important for cytokinesis

To further characterize the role of Abba in mitotic progression, we synchronized C6 rat glioma cell line and stained with Abba antibody. We found that Abba localized to the putative apical membrane initiation site and cytokinetic bridge, proximal to the contractile actin ring during late telophase and cytokinesis (Figure 3A and B respectively). The actomyosin contractile ring is formed at the plasma membrane in response to accumulation of phosphoinositide 4,5 bisphosphate,PI(4,5)P2, to generate constricting force to separate the two daughter cells³⁵. Abba is known to interact with PI(4,5)P2 via its membrane deforming I-BAR domain²⁰ that has the capacity to deform and recognise the membrane curvature.

Figure 3.

To determine the impact of the loss of Abba on cytokinesis, we quantified the number of C6 cells in cytokinesis 3 days after transfection with Abba-shRNA3 or scramble shRNA. Importantly we found a two-and-a-half-fold increase in the percentage of Abba-shRNA3-expressing cells in cytokinesis (Figure 3B and D; scramble 0,8285 ± 0,1161%, n=12; Abba-shRNA3 2,078 ± 0,2682%, n=12; p=0,0358), suggesting a prominent role of Abba in cytokinesis. To validate this in vivo, we performed live imaging in cortical organotypic slices from animals that were in utero electroporated with either Abba-shRNA3 or scramble vectors. Analysis of live imaging recording showed a strong mitotic block at the ventricular surface (Figure 3C). We then examined the impact of cytokinesis block on cell survival and found a lower total number of cells 4 days after transfection with Abba-shRNA3 compared with control scramble (Figure 3E; scramble: 380,3 ± 48,16%, n=12; Abba-shRNA3: 155,8 ± 22,32%, n=12; p=<0.0001). Taken together, these data provide evidence that Abba is required for the completion of mitosis, and its absence results in a reduced number of cell progeny.

Abba recruits Nedd9 to the cytokinetic bridge and is required for RhoA activation

To investigate the mechanism by which Abba operates in mitosis, we performed a yeast-two-hybrid (Y2H) screen in E10.5-E12.5 mouse brain embryo library using full length mouse Abba as a bait. The resulting analysis identified three high (PBS: B) / good (PBS: C) confidence interactors: the E3 ubiquitin ligase Beta-Trcp2 (PBS: B), scaffolding protein Nedd9 (PBS: C), the transcription factor Otx2 (PBS: C) (Fig 4A; Supplemental table 2). Among these candidate proteins, Nedd9 (aka HEF-1 or Cas-L) caught our interest as it was previously reported in breast cancer cells to localize to the cleavage furrow during cytokinesis where it activates RhoA, as well as abnormal expression of Nedd9 results in cytokinesis defect^21,36,37 (Figure 4B).

Figure 4.

We first tested whether Abba and Nedd9 physically interact, as suggested by the Y2H screen. To this end, we performed immunoprecipitation experiments from embryonic day 18 cortical tissue as well as C6 cells using anti-Nedd9 as well as anti-Abba antibodies. The precipitated Nedd9-Abba complex was separated on SDS-page and followed by western blot with anti-Abba and anti-Nedd9 antibodies respectively. Indeed, Abba precipitated with Nedd9 in both directions, but not using control beads from brain cortex homogenates (Figure 4C) and C6 cells (Supplemental figure 3). Nedd9 stability is not affected by the availability of Abba as western blots show no changes in Nedd9 expression levels in Abba-shRNA3 expressing cells (data not shown). Interestingly however, when we analyzed the localization of Nedd9, we found that Abba was required for targeting Nedd9 to the cleavage furrow/cytokinetic bridge (Figure 4D). These data suggest that Abba recruits Nedd9 to the cleavage furrow during RGP cells division, which fits to its role in sensing curved PI(4,5)P2 rich membranes. Nedd9 plays a role in RGP cytokinesis in vivo^21,36,37. To knockdown Nedd9, we used shRNA targeting Nedd9 which reduced mRNA levels by 75% in C6 cells (Supplemental Figure 3A). We used a previously characterised Nedd9-shRNA to determine whether knockdown of Nedd9 expression alters RGP cells morphology and function. To this end, Nedd9-shRNA combined with a red fluorescent protein (RFP) construct was introduced into neural progenitor cells of mouse neocortex by in utero electroporation at E14 and stained with the radial glial cell marker, vimentin. Interestingly, similarly to Abba knockdown, Nedd9 downregulation resulted in disruption of INM as a larger portion of progenitors accumulated close to the ventricular surface (Figure 4E-F; scramble: 25 ± 2.30% n=8, Nedd9-shRNA: 46.92±2.78%; n=7; p=<0.0001). Using the late G2/M phases marker, PH3, we observed similar results as with Abba shRNA and point mutation no difference in number of PH3 positive cells (Supplemental figure 3C; scramble: 4.50 ± 0.18%, n=9; Nedd9-shRNA: 5.07 ± 0.35%, n=6, P=0,1135). The role of Nedd9 in cell cycle progression is consistent with previous published results^21,36,37 and provide evidence that both Abba and Nedd9 function to ensure correct mitotic completion of RGP cells.

Finally, to determine whether Abba is important for Nedd9 activity and how Abba and Nedd9 converge to regulate the cell-cycle, we assessed the activity of RhoA, a downstream signalling molecule of Nedd9³⁸ which is important for cytokinetic actomyosin ring assembly and dynamics^38–40. To determine RhoA activity we used a GFP-based FRET construct for RhoA designated ‘‘Ras and interacting protein chimeric unit’’ Raichu-RhoA. In order to test the efficacy of the assay we used Calpeptin, a RhoA activator which, as expected, increased RhoA activity (% change compared to Raichu-RhoA: Calpeptin 132,2 ± 3,2 n=33; p=0.002) confirming the functionality of our assay (Figure 5). Importantly, when Abba was silenced by shRNA there was a significant 31,5% decrease in Raichu-RhoA activity as compared to the control shRNA treated cells (Figure 5B; scramble: 103,1.03 ± 5.3; n=102, Abba-shRNA3: 68,5 ± 2,3; n=46; p<0.0001). In addition, this effect was rescued by the co-expression of shRNA insensitive plasmid carrying the full-length sequence of Abba (Abba-shRNA + Abba-FL 139,8 ± 19,6 n=26; p<0.0001 compared to Abba-shRNA3). Abba-FL did not change RhoA activity significantly compare to scramble expressing cells (Abba-FL 105,2 ± 16,0; n=37). These data provide support for the function of Abba in spatial assembly of cytokinetic signaling through the Nedd9-RhoA axis.

Figure 5.

Identification of ABBA missense variant R671W in an individual with neurodevelopmental syndrome

Mutations in genes effecting mitotic progression of neuronal progenitors have been linked to neurodevelopmental disorders⁴¹. Importantly, we identified a patient carrying a novel heterozygous missense variant in ABBA coding MTSS1l/MTSS2 gene. This mutation (2011C>T(R671W)) affects residues that are conserved from Zebrafish to Human and reside in exon 15 (Figure 6A and B). Consistent with a de novo origin of the variant, it was not found in the parents of the affected individual or in normal controls. This patient presents some characteristic craniofacial dysmorphism as mild deep-set eyes, thickened helices, synophrys, narrow forehead, bitemporal narrowing, small bilateral epicanthal folds, upslanting palpebral fissures, medial flaring of eyebrows, flat midface, V-shaped palate, and dysmorphic ears. Brain magnetic resonance imaging sequences showed the presence of a microcephalic brain associated with a smaller head linked to mild IQ. The patient also suffered from Attention-Deficit/Hyperactivity Disorder (ADHD) (Figure 6C, Supplemental Table 1). Moreover, a recent publications found six additional patients carrying the same ABBA variant that shared similar brain malformation^42,43. This variant replaces an arginine to tryptophan that has a large aromatic group that could lead to destabilization of a tertiary structure. Indeed, bioinformatic prediction using AlphaFold predicts (pLDDT) with high 84.4 confidence score that R671 is located in an alpha helical fold close to the C-terminal end of the protein⁴⁴. Furthermore, analysis using Poly-Phen-2 structure and sequence-based algorithm to predict the impact of R671W substitution on the structure and function of ABBA, we found a high probability that the mutation is damaging for the protein function²³ (Naïve bayes probability score 0.952) (Figure 6D).

Figure 6.

Expression of R671W Human variant results in defects in mitosis and neuronal migration

To test the effects of the human missense variant in mouse brain development, we introduced the human ABBA cDNA encoding either wild type or patient mutation R671W into developing mouse brains by in utero electroporation at E14 followed by imaging and quantification of targeted cell at E17. Interestingly, whereas wild type ABBA (ABBA-FL) expression showed no defects in radial glial morphology or neuronal migration, the R671W mutant expression resulted in disorganized radial glial fibers and accumulation of multipolar neurons in the SVZ/lower IZ with very few bipolar neurons (Figure 6E). When analyzing the distribution of electroporated neuron progenitors, we observed clear accumulation in the VZ/SVZ (Figure 6F; ABBA-FL 18.67±1.44%; n=8, ABBA-R671W 39.77±2.73% n=6; P=0.0001). We also found a significant increase in RGP nuclei at the VS of the brain (Figure 6G, C ABBA-FL 32.56±2.58%; n=8, ABBA-R671W 50,04±2.04%; n =6; p=0.000001. Interestingly, no difference in accumulation of PH3+ cells was found (Supplemental Figure 3 B and C; Abba-human-WT 4.50 ± 0.17%; n=9, Abba-human-mutation 4.71 ± 1.98%; p=0.76; n=5), indicative of a dysfunction in neuronal maturation but a less prominent impact on cell division. In line with these results the effect of ABBA-R671W on RhoA activation was not significantly different to scramble expressing cells (142 ± 17,9 n=15). The data presented indicate that the abnormal expression of the R671W mutation, rather than the normal protein, leads to abnormal development of the developing cortex. These findings offer evidence suggesting a potential connection between the missense mutation in ABBA discovered in the patient and abnormal brain development. However, it is unlikely that the mechanism behind this phenotype is a result of ABBA-R671W acting in a dominant-negative manner on cell cycle progression.

Discussion

Abba was initially discovered as a novel regulator of actin and plasma membrane dynamics, specifically in radial glial C6 cells. However, its role in brain radial glial progenitor (RGP) cells in live organisms has remained unclear. In this study, we observed that decreased expression of Abba in radial glia within the cortical ventricular zone leads to impaired radial migration and disrupted organization of radial glia. Notably, knock-down of Abba specifically hindered cytokinesis in RGP cells, resulting in abnormal cell morphology, mitosis, and reduced numbers of glial and neuronal cells. The underlying mechanism behind this effect appears to involve impaired signaling through the actin regulatory protein RhoA, likely through its interaction with Nedd9 during the early and late stages of telophase. In accordance with these findings, we identified a missense variant in a patient with microcephaly and intellectual disability exhibiting similar clinical characteristics. Furthermore, overexpression of the mutant protein in cortical progenitors of mice resulted in phenotypes resembling, although not identical to, the down-regulation of Abba in the cortex. The results presented in this study highlight the critical importance of Abba in cortical development and contribute to elucidating the potential mechanism underlying abnormal brain development associated with human variants of ABBA.

Radial glial exiting from cytokinesis requires Abba

BAR proteins can interact with PI(4,5)P₂-rich membranes through their BAR domain, and several studies indicate that PI(4,5)P₂ play a central role in cytokinesis. Indeed, it has been reported that in mammalian cells PI(4,5)P₂ accumulates at the cleavage furrow and recruits membrane proteins required for stability of the furrow because of a role in adhesion between the contractile ring and the plasma membrane. Moreover, overexpression of proteins that bind to PI(4,5)P₂ perturbs cytokinesis completion, by interfering with adhesion of the plasma membrane to the contractile ring at the furrow, but not ingression of the cleavage furrow^45,46. Immunocytochemistry of Abba shows a strong accumulation at the cleavage furrow. In addition, in absence of Abba, we observed an accumulation of cells blocked in cytokinesis (Figure 3). these observations and previous findings showing that Abba can interact with PI(4,5)P₂-rich membranes through its BAR domain as well as with the actin cytoskeleton through Wasp Homology-2 domain (WH2), lead us suspect that Abba could link the actin contractile ring to the plasma membrane and participate in the mechanism for the invagination of the nuclear membrane during cytokinesis.

Abba interacts with Nedd9 to regulate RhoA activity

Here we have found that Abba interacts directly with Nedd9 and down regulation of Abba leads to significant changes in RhoA activation. Nedd9 is a focal adhesion protein involved in mitotic entry and cleavage furrow ingression⁴⁷. Similar to the results found in this study, previous results have shown that cells with abnormal expression of Nedd9 remain in G1 subsequently leading to cell apoptosis³⁷. This is consistent with the view that Nedd9 triggers cells to enter mitosis. Nedd9 is known to be a positive regulator of RhoA signaling in mitosis. Indeed, activation of RhoA is required for cell rounding at early stages of mitosis and cleavage furrow ingression, and deactivation of RhoA is essential for cell abscission²¹. Moreover, it has been shown that Nedd9 overexpression blocks cells at abscission by maintaining levels of RhoA activation, and decreased Nedd9 expression inhibits cleavage furrow ingression by non-activated RhoA pathway. In light of these results, we hypothesized that during cytokinesis at the cleavage furrow the membrane composition change and is enriched in PI(4,5)P₂ leading to Abba recruitment and its interaction with Nedd9 which is essential to activate the RhoA pathway cleavage furrow ingression.

Role of ABBA in neuronal migration and microcephaly

The proper formation of the cerebral cortex relies on the sequential generation of cortical layers, which is highly dependent on the normal production, survival, and migration of neuronal progenitors. Abnormalities in these processes are associated with pathological conditions such as microcephaly, leading to impaired cognitive development. Interestingly, both downregulation of Abba and overexpression of the ABBA point mutation construct result in a similar accumulation of neuronal progenitors in the ventricular zone (VZ) and subventricular zone (SVZ). This may be attributed to abnormal proliferation of neuronal progenitors caused by disrupted interaction with critical proteins involved in cell cycle progression. This disruption aligns with the hypothesized impact on Nedd9 interaction and reduced RhoA activity observed in conditions of low Abba expression.

One possible explanation for these results could be the disruption of the predicted local tertiary structure. In this scenario, heterozygous expression of the human R671W variant would exert a dominant negative effect on ABBA’s role in brain development, leading to microcephaly and cognitive delay. This notion is supported by recent work disclosing additional patient carrying the R671W variant⁴². In the same study the significant neurological phenotypes were observed in a drosophila model where the ortholog of human MTSS2 and MTSS1 mim was deleted. However, from a clinical genetics’ standpoint, it is unlikely to find patients with the recurrent R671W mutation without any homozygous or compound heterozygous loss-of-function mutations elsewhere in the ABBA gene. This could also suggest a gain-of-function effect of the R671W mutation. Supporting this notion, overexpressing ABBA-R671W in cells expressing the wild-type Abba in this study did not result in a dominant-negative decrease in RhoA activation, nor did it affect the expression of PH3 in vivo. These findings make is plausible to suggest that a mechanism responsible for the phenotype associated with overexpression of the human variant may primarily involve post-cell division processes, such as cell migration.

Although previous research has indicated that Abba is not expressed in mature neurons²⁰, more recent studies have shown transient expression of Abba in neurons during conditions of increased neuronal plasticity¹³. While our study primarily focused on Abba’s role in radial glial cell proliferation, the data also indicate an additional role in the migration of immature neurons. This is supported by the observed accumulation of Abba-shRNA3 electroporated cells in the ventricular zone. Additionally, we observed a significant decrease in process length in migrating cortical neurons with downregulated Abba and expressing ABBA-R671W mutant but interestingly dendritic processes were increased in length in neurons overexpressing ABBA (Supplemental Figure 4). Given that disorganization of radial glial progenitor cells does not lead to neuronal cell death, these findings suggest that Abba may be involved in additional molecular mechanisms in differentiated neurons. Further experiments are needed to determine whether this phenotype is intrinsic to Abba’s role in migrating neurons or if it is influenced by the disruption of guiding radial glia.

In conclusion, this study provides compelling evidence for the significant role of Abba in brain development. We demonstrate that Abba’s involvement in the cell division of radial glial progenitors is a key component of this mechanism. Furthermore, these findings have translational relevance as they contribute to a better understanding of the putative mechanisms underlying human variants associated with microcephaly and developmental disabilities.

Data Availability

All data produced in the present study are available upon reasonable request to the authors

Acknowledgements

We thank Drs Lauren Briere and Gabrielle Lemire for discussion about the data. We thank Biomedicum Imaging Unit (Helsinki, Finland) for their technical assistance with the time-lapse imaging of organotypic cultures. This project was supported by Eranet Neuron III program project ACROBAT, ANR project GABGANG and Academy of Finland grant 341361 to C.R., and by Fondation pour la Recherche Medicale “aide au retour en France” to A.C. Academy of Finland (317038, 319907) and Sigrid Juselius foundation to J.S. and Finnish Cultural Foundation to E.E.

Author Contributions

A.C conceived the project. A.C. designed and performed most of the experiments, analyzed and interpreted all the data, and prepared figures. A.C. and C.R. wrote the original draft of the manuscript. J.S. designed experiments, interpreted data, edited the manuscript. E.E: performed and analyzed experiments and edited the manuscript. H.K. performed organotypic slices culture and live imaging experiments. All authors read and approved the final manuscript.

Conflict of interest statement

None declared