Reproducible morphological and molecular evolution of iPSC differentiation into neuroepithelial cells.

iPSC1 is HO193b, iPSC2 is SFC086-03-01 and iPSC3 is 802-30F. a, Schematic representation of the neuroepithelial induction protocol. Sub-confluent human iPSCs (in grey) are passaged and seeded at full confluency from the start of differentiation and cultured for eight days, until reaching neuroepithelial maturity. Further passage at this stage forms neural rosettes and allows further differentiation into post-mitotic neurons56. b, Summary of control lines used to validate reproducibility of differentiation into neuroepithelial cells. c,d, Representative high-resolution confocal top-view (c) and cross-section (d) images of day 2-8 cultures fluorescently labelled to show F-actin with inverted gray LUT. P values from one-way ANOVA with post-hoc Bonferroni. e, Dot plot representing cell’s apical area across timepoints of neuroepithelial differentiation, in three iPSC lines. Black bars represent the median apical area across three independent culture plates of each line. f, Heatmaps of hiPSC2 representing the apical area size changes during differentiation, in a contiguous area of 1,200 μm2 per timepoint. g, Quantification of epithelial thickness across neuroepithelial differentiation of control cell lines. Three culture replicates were measured for each time point and cell line. h, Dot plot representing the cell volume decrease across timepoints of differentiation with representative 3D volumetric images at day 2 and day 8 (h’). i, Representative western blot panel of NANOG, OCT4, CDH1, CDH2, PAX6, and LMNB protein levels across timepoints of neuroepithelial differentiation, in the iPSC2 line. k,j, Dot plots representing normalised band intensity of NANOG (j), and PAX6 (k), across differentiation for three control cell lines. NANOG expression is normalised to day 2 per line, whereas PAX6 is normalised to day 8 as it is not detected at day 2. P values from one-way ANOVA (e,g) or Wilcoxon paired test (j,k) with Bonferroni correction (***: p ≤ 0.001, *: p ≤ 0.005). Scale bar = 10 μm.

iPSC-derived neuroepithelial cells undergo ROCK-dependent apical constrictions and up-regulate expression of the PCP co-receptor VANGL2.

a, Live imaging of day 8 neuroepithelial cultures stained with CellMask Actin dye. Arrows indicate the apical surface of a constricting cell (not its neighbours do not constrict). b, Representative F-actin-stained images of cultures treated vehicle or 10 µM ROCK inhibitor (Y27632) for 4 hr. c, Dot plot representing the median apical area of day 8 SFC086 neuroepithelial cultures. d, Live-imaged membrane-stained day 8 neuroepithelial cultures before and after a circular laser ablation, shown by red dashed lines. Red ring indicates neuroepithelial deformation before and after the ablation, respectively. e, Median apical cluster constriction of neuroepithelial cells, across three independent replicates in one cell line. Each point represents the median displacement quantified in each plate after at least three laser ablations. f,g, Representative confocal images of SCRIB and VANGL2 immunolabelled cultures at day 2 and day 8, in top-view (f) and cross-section (g). Dashed lines indicate the basal side. h, Representative western blot panel of VANGL2, phosphorylated VANGL2 (pVANGL2), and GAPDH protein levels across timepoints of neuroepithelial differentiation. All analyses shown are in iPSC2. P values from Wilcoxon paired test with Bonferroni correction (**: p ≤ 0.01, *: p ≤ 0.005). Scale bars = 10 μm.

VANGL2-R353C mutation selectively impairs neuroepithelial apical constriction.

a, Schematic representation of PCP pathway, starting from non-canonical Wnt ligands (e.g. Wnt5a/11) interacting with membrane co-receptors and activating cytoskeletal remodeling responses. The transmembrane protein VANGL2 interacts with CELSR, which results in downstream interaction with cytoplasmic proteins such as PRICKLE and SCRIB. The VANGL2-R353C mutation is localised at the C-terminal region of the protein, in the prickle binding domain (Pk-D). b, Representative western blot against VANGL2 and pVANGL2 on day 8 of neuroepithelial differentiation in the isogenic control cell line (WT) and two knock in clones with the R353C mutation (KI-1 and KI-2). c, Dot plots representing VANGL2 and pVANGL2 relative band intensity levels after normalisation to GAPDH, respectively, across three replicates. Intensity levels across the three lines were normalised to the WT levels of each replicate. d, Representative high resolution confocal images of ZO-1 immunolabelled neuroepithelial cultures on day 8, across the WT and KI cell lines. e, Frequency plot showing the distribution of apical area sizes of three independent day 8 neuroepithelial cultures, across the WT and both knock-in (KIs) lines. *** p < 0.001 by Kolmogorov Smirnov test, lines are the aggregate of three independent differentiations per line. f, Representative confocal images of neuroepithelial cultures immunolabelled for phospho-myosin light chain (pMLC)II, comparing the WT and KI-1 and KI-2 cell lines. Top view shows the distribution of protein of interest across the apical areas and cross-section illustrates localisation along the apicobasal axis. Scale bar = 10 μm. g, Dot plot showing normalised paired relative apical pMLCII intensity, across the WT and KI-1 and KI-2 cell lines, on day 8 of differentiation. Relative apical intensity was normalised to WT and set at 1, for each paired analysis, across four independent plates. h, Live-imaged membrane-stained neuroepithelial sheet before and after a circular laser ablation, with overlayed image of post-ablation timepoint and particle image velocimetry (PIV) analysis. The size and colour of arrows in PIV analysis represent the direction and magnitude of recoil. Scale = 20 μm. i, Median apical cluster constriction of neuroepithelial cells, across seven independent replicates. Each point represents three individual quantifications in each plate and connected lines indicate paired cultures. P values from Wilcoxon paired test with Bonferroni correction (***: p ≤ 0.001, *: p ≤ 0.005, n.s.: not significant).

Spina bifida patient amniocyte-derived iPSC lines can identify line-specific deficiencies in neuroepithelial cell function.

a, Experimental timeline of patient amniocyte-derived iPSC lines, following microfluidic reprogramming. Whole exome sequencing (WES) was performed before and after reprogramming to identify line-specific deficiencies at genomic level and pair them to neuroepithelial and/or neural function level. Initial iPSC cell expansion is not shown, but adds a few weeks to the timelines, still within the remaining period of normal gestation. b, Representative high-resolution top-view and cross-section images of day 8 neuroepithelial cultures visualizing F-actin. Two control cell lines (GOC1 and GOC2) were compared with two SB patient-derived hiPSCs cell lines (GOSB1 and GOSB2). c, Dot plot representing cell apical areas across the GOC1, GOC2, GOSB1, and GOSB2 cell lines at day 8 of neuroepithelial differentiation. Three independent plates were quantified for each line and the median apical area for each plate is shown with a black bar. d, Quantification of epithelial thickness on day 2 and day 8 samples, across four cell lines, GOC1, GOC2, GOSB1, and GOSB2. Points represent independent plate quantification for each line. Three paired replicates were measured for each time point and each cell line. Statistical test used for median values across quantifications was Wilcoxon paired test values with post-hoc Bonferroni. e, Representative confocal images of LMN A/C, CDH2, SOX2, and LMNB immunolabelling in GOC2, and GOSB2, on day 8 of neuroepithelial differentiation. f, Venn diagrams of GOSB1 and GOSB2 (f’) gene lists, after exome-sequencing, with at least one genetic variation observed in other datasets. Other datasets include previously associated mutations identified in humans under after search for the term “Neural Tube Defect” and “Spina Bifida” in ClinVar dataset, Lee and Geeson5. g, Filtering of variants in GOSB1 and GOSB2 to identify NTD-associated variants predicted to have a high impact on the transcript and which have a frequency <0.1 in gnomAD exome datasets. h, RNA structure of the MED24 major allele and rare variant (gnomAD reference 17-40019203-C-CACAT, allele exome frequency 0.018) predicted by RNAFold69. i, Bar plots representing normalised band intensity of MED24 across four cell lines. GOC2’s intensity levels were set as 1 in each of three replicates. j, Western blot panel of iPSCs from four cell lines, GOC1, GOC2, GOSB2, and SFC086-06 showing MED24 protein and housekeeping GAPDH levels. Statistical test used for values was one-way ANOVA with post-hoc Bonferroni (***: p ≤ 0.001, **: p ≤ 0.001). Scale = 20 μm.

Diminished neuronal maturation suggests post-neurulation deficit of an iPSC line derived from a patient with spina bifida.

a, Representative confocal images of GOC1, and GOSB1 neuronal rosette cultures (day 20) immunolabelled against PAX6 and F-actin. Scale = 100 μm. b, Representative immunofluorescence panel of SOX2 and TUJ1 immunofluorescence of day 30 cultures, showing GOC1, GOC2, and GOSB1 lines. Scale = 50 μm. c, d, Dot plot representing TUJ1/DAPI staining intensity (c) and the percentage of nuclei positive for SOX2 (d). Values were calculated in three independent replicates shown as individual dots and the average is shown with the bars. e, Widefield images of day 45 GOC1 and GOSB1 neuronal cultures immunolabelled for TUJ1. Scale = 200 μm. f, Box plot representing the TUJ1 positive area quantified across three independent cultures of GOC1 and GOSB1 cell lines, on day 45 of differentiation. P values from one-way ANOVA with post-hoc Bonferroni (***: p ≤ 0.001, **: p ≤ 0.001, n.s.: not significant). g, Summary of patient-associated hiPSCs lines used in this study along with the stage of failure to form a functional neuroepithelium or undergo mature neuronal formation in vitro.

hiPSCs line information.

Composition of neural maintenance medium.