Figures and data
Actomyosin forces increase during cranial neural fold elevation
(A) Schematic of the presumptive midbrain and hindbrain of mouse embryos during early elevation (4-6 somites, E8.25), late elevation (7-9 somites, E8.5), and apposition (10-12 somites, E8.75). En face views (top), transverse views (bottom). Dotted lines, locations of transverse views. Arrows represent mechanical forces. (B) Localization of the myosin IIB heavy chain (Myo IIB) and the phosphorylated myosin II regulatory light chain (pMRLC) in early and late elevation. (C and D) pMRLC localization is planar polarized in late (D) but not early (C) elevation. Plots show mean fluorescence intensity at mediolateral (ML) edges (0-15° relative to the ML axis) divided by the mean intensity at anterior-posterior (AP) edges (75-90° relative to the ML axis). (E-G) Schematic (E) and kymographs (F and G) of ML edges before and 2-8 s after ablation in early (F) and late (G) elevation embryos expressing myosin IIB-GFP. (H) ML edge recoil velocity. (I-K) Schematic (I) and kymographs (J and K) of AP edges before and 2-8 s after ablation in early (J) and late (K) elevation embryos expressing myosin IIB-GFP. (L) AP edge recoil velocity. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. 8 regions in 4 embryos in (C) and (D), 16-19 ablations in 6-9 embryos in (H and L). *p<0.04, ***p=0.0002 (Welch’s t-test). See Supplementary File 1 for for a summary of all data and statistical analyses. Maximum intensity projections, anterior up, edges oriented vertically in kymographs. Bars, 10 µm.
ESC-derived embryos recapitulate neural tube closure defects of Shroom3 mutants
(A) Schematic of embryo generation from mouse ESCs. ESCs engineered using CRISPR/Cas9 gene editing were injected into E2.5 host embryos, ESC-injected embryos were transferred to surrogate females, and developing embryos were recovered for analysis. (B) A majority of embryos generated by ESC injection were derived primarily from GFP-negative ESCs (259/288 embryos), whereas some contained a significant contribution of Histone-H2B-GFP-positive host cells (29/288 embryos) and were excluded from further analysis. (C) Light micrographs of E9.5 ControlESC and Shroom3ESC embryos (0/41 ControlESC embryos and 79/79 Shroom3ESC embryos displayed exencephaly). Lateral views, dotted lines indicate the lateral edges of the cranial neural plate. (D) Lateral midbrain cells stained for ZO-1 (top) and color-coded by apical cell area (bottom). (E, F) Average apical cell area (E) and apical cell area distributions (F) of lateral midbrain cells in ControlESC and Shroom3ESC embryos in mid-elevation (5-7 somites). ControlESC embryos were derived from unedited (HK3i) or Tyr mutant ESCs. A single value was obtained for each embryo and the mean±SEM between embryos is shown (1572-2744 cells in 3 embryos/genotype). *p<0.02 (Welch’s t-test). Maximum intensity projections, anterior up. Bars, 100 µm (B), 500 µm (C), 10 µm (D).
Vinculin is required for cranial neural fold elevation
(A) Transverse sections of the midbrain in ControlESC and VinculinESC embryos in early elevation (4-6 somites), late elevation (7-9 somites), and apposition (10-12 somites). F-actin (phalloidin) labels the apical surface of the neuroepithelium and laminin labels the basal surface. (B) Schematic of apical and basal span measurements. (C) Apical-to-basal span ratios. (D) Lateral midbrain cells stained for ZO-1 (pre-elevation) or N-cadherin (early and late elevation) and color-coded by apical cell area. (E-H) Average apical cell areas (E) and apical area distributions (F-H) of lateral midbrain cells in ControlESC and VinculinESC embryos before (F), early (G), and late (H) in elevation. A single value was obtained for each embryo and the mean±SEM between embryos is shown (1335-3955 cells in 3-4 embryos/genotype). *p<0.04, **p<0.004, ***p<0.001 (Welch’s t-test). Maximum intensity projections, apical up in A and B. Maximum intensity projections, anterior up in D. Bars, 100 µm (A), 10 µm (D).
Vinculin is not necessary to generate force but is required for actomyosin organization at tricellular and multicellular junctions
(A) Localization of myosin IIB and F-actin (phalloidin) in late elevation wild-type and VinculinΔEpi embryos. (B, C) Close-ups of tricellular and multicellular junctions in wild-type (B) and VinculinΔEpi(C) embryos. (D-G) Laser ablation experiments in ControlESC and VinculinESC embryos. (D, G) ML edges before and 2-8 s after ablation in early (D) and late (G) elevation embryos expressing GFP-Plekha7. (E, F) Peak recoil velocity after laser ablation of ML edges in early (E) and late (F) elevation. (H, I) GFP-Vinculin TCJ ratios (tricellular junction intensity divided by the mean intensity of the three connected bicellular junctions) in early (H) and late (I) elevation. (J, K) GFP-Vinculin localization in live wild-type embryos in early and late elevation. (L) Treatment of embryos for 2 h with 200 μm Rho-kinase inhibitor (Y-27632) in early elevation abolishes GFP-Vinculin localization at cell junctions. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. 14-26 ablations in 5-7 embryos/genotype in E and F, 60-80 tricellular junctions in 3-4 embryos in H, I, and K. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. ****p<0.0001, Welch’s t-test. Maximum intensity projections, anterior up, edges oriented vertically in kymographs. Bars, 10 µm (A, D, G, J, and L), 2 µm (B and C).
Vinculin is required for N-cadherin localization at tricellular and multicellular junctions
(A, B) Localization of N-cadherin and F-actin (phalloidin) in ControlESC and VinculinESC embryos in early (A) and late elevation (B). Arrowheads indicate examples of multicellular junctions with gaps in N-cadherin localization. (C, D) Close-ups of bicellular, tricellular, and multicellular junctions in ControlESC (C) and VinculinESC (D) embryos. (E, F) Number of gaps in N-cadherin localization in a 50 µm x 50 µm region of the lateral midbrain in early (E) and late (F) elevation. (G-I) Percentage of tricellular junctions (G), 4-cell junctions (H), and 5+-cell junctions (I) with gaps in GFP-Plekha7 localization. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. 5-6 regions in 3 embryos in E and F, 81-175 3-cell junctions, 21-57 4-cell junctions, and 7-19 5+ cell junctions/region in 8-12 regions from 4-6 embryos in G-I. **p<0.002, ****p<0.0001 (Welch’s t-test). Maximum intensity projections, anterior up. Bars, 10 µm (A and B), 2 µm (C and D).
Loss of Vinculin disrupts tricellular and multicellular adherens junctions, whereas ZO-1 localization is generally maintained
(A) Localization of the tight junction protein ZO-1 and the adherens junction protein GFP-Plekha7 in late elevation ControlESC and VinculinESC embryos. Bottom, merged images of the indicated regions showing a wild-type multicellular junction (magenta), a junction with a gap in GFP-Plekha7 localization (green), and a junction with a gap in GFP-Plekha7 and ZO-1 localization (yellow). Circles indicate gaps in GFP-Plekha7 (cyan) or ZO-1 (red) signal. (B, C) Number (left) and areas (right) of adherens junction gaps (B) and tight junction gaps (C) in a 50 µm x 50 µm region in late elevation ControlESC and VinculinESC embryos. Note the differences in scale between the adherens junction and tight junction plots. (D, E) Area distributions of adherens junctions gaps detected with GFP-Plekha7 (blue) and tight junction gaps detected with ZO-1 (red) in ControlESC (D) and VInculinESC (E) embryos. Five adherens junction gaps in (E) are outside of the x-axis range. (F, G) En face (XY) views and optically reconstructed (XZ) cross-sections of Airyscan z-stacks of GFP-Plekha7, ZO-1, and myosin IIB localization at multicellular junctions in late elevation wild type (F) and VinculinΔEpi (G) embryos. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. Mean±SD between regions in B and C (right plots), D, and E. 150-498 adherens junction gaps in 8-12 regions from 4-6 embryos. ***p=0.0004, ****p<0.0001 (Welch’s t-test). Maximum intensity projections, anterior up in A, F, and G (top panels). Maximum intensity projections, apical up in F and G (bottom panels). Bars, 10 µm (A, top panels), 2 µm (A, bottom panels, F, and G).
Vinculin is required to regulate force sensitive morphogenetic behaviors during elevation
(A) Stills from time-lapse movies of ControlESC and VinculinESCembryos expressing GFP-Plekha7. Arrows, dividing cells. (B, C) Stills from time lapse movies showing rosettes (B) and dividing cells (C) in ControlESC and VinculinESC embryos. t=0 in A and B is the start of the movie, t=0 in C is the time point immediately before the start of cleavage furrow ingression. (D) Left, percentage of rosettes with no gaps in GFP-Plekha7 signal or central gaps that repair, persist without changing size, or persist and increase in size in time-lapse movies of ControlESC and VinculinESC embryos. Right, percentage of rosettes with gaps in ControlESC and VinculinESC embryos. (E) Left, percentage of divisions with gaps in GFP-Plekha7 signal at interfaces between the dividing cell and neighboring cells that were not restored within 90 minutes. Right, percentage of divisions in which GFP-Plekha7 was not detected at a new vertex or interface within 90 minutes after the onset of cleavage furrow ingression. (F) Cumulative percentage of completed divisions that formed a new GFP-Plekha7-positive vertex or interface by the indicated times in ControlESC and VinculinESC embryos. Embryos were imaged immediately before and during early elevation (2-5 somites). Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean (in D right panel and E both panels). Mean±SEM between embryos in (D left panel and F). 47-48 rosettes in (D), 51-52 divisions in (E), and 28-50 divisions in (F) in 4-5 movies/genotype. *p<0.02, ***p<0.001 (Welch’s t-test). Maximum intensity projections, anterior up in A-C. Bars, 10 µm.
The response to edge ablations in the cranial neural plate requires actomyosin contractility
(A) Localization of the tight junction protein ZO-1 in the lateral midbrain of early and late elevation wild-type embryos (see Figure 1C and D for quantification). (B and C) ML edges before and 2-8 s after ablation in late elevation embryos expressing GFP-Plekha7 and treated for 2 h with water (B) or 200 µM of the Rho-kinase inhibitor Y-27632 (C). (D) Peak recoil velocity after laser ablation of ML edges. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. 15 ablations in 4 embryos/condition. ****p<0.0001 (Welch’s t-test). Maximum intensity projections, anterior up, edges oriented vertically in kymographs. Bars, 10 µm.
Generation and validation of Shroom3 mutant and GFP-Plekha7 ESCs
(A and B) ESC clones homozygous for a null mutation in Shroom3 (A) were identified by PCR screening and deep sequencing (B). (A) Red box, deleted region within exon 5 of the Shroom3-203 isoform. (B) Deletion breakpoints of four homozygous mutant clones. Top, 5’ and 3’ gRNA target sites. Red, PAM site. Dotted lines, predicted CRISPR cut sites. Bottom, deletion breakpoints. (C) Shroom3 protein is not detected by immunofluorescence in the lateral midbrain in Shroom3ESC embryos. (D and E) Embryos generated by injection of GFP-Plekha7-expressing ESCs displayed GFP-Plekha7 expression throughout the cranial neural plate (D) (black regions, dividing cells) that colocalized with N-cadherin (E). Maximum intensity projections, anterior up. Bars, 100 μm (D), 10 µm (C and E). (F) Pups generated from the GFP-Plekha7-expressing ESC clone R26-PHA7-EGFP A3. 25/27 animals had 100% ESC coat color and 2/2 adults tested displayed germline transmission.
Generation and validation of Vinculin mutant ESCs
(A and B) ESC clones homozygous or heterozygous for a null mutation in Vinculin (A) were identified by PCR screening and deep sequencing (B). (A) Red box, deleted region including exon 3 of the Vinculin-201 isoform. (B) Deletion breakpoints of Vinculin homozygous mutant clone 58 and heterozygous control clone 15. Top, 5’ and 3’ gRNA target sites. Red, PAM site. Dotted lines, predicted CRISPR cut sites. Bottom, deletion breakpoints. (C) Vinculin protein is absent in VinculinESC embryos (generated from clone 58) and present in ControlESC embryos (generated from clone 15) and +/+ embryos (FVB/N) (one E9.5 embryo/lane). (D) Light micrographs of E9.5 ControlESC and VinculinESC embryos (23/23 VinculinESC embryos and 0/15 ControlESC embryos displayed exencephaly). (E) Light micrographs of E9.5 wild-type and VinculinΔEpi embryos (11/11 VinculinΔEpi embryos and 0/30 wild-type and heterozygous littermate controls displayed exencephaly). Lateral views, dotted lines indicate the lateral edges of the cranial neural plate. Bars, 500 µm.
Apical-basal elongation, proliferation, and apoptosis are unaffected in Vinculin mutants
(A) Apical-basal cell height in transverse sections of the lateral midbrain of ControlESC and VinculinESCembryos in early elevation (4-6 somites), late elevation (7-9 somites), and apposition (8-10 somites). A single value was obtained for each embryo and the mean±SEM between embryos is shown (3-4 embryos/genotype). *p<0.03 (Welch’s t-test). (B) Dividing cells detected with phospho-histone H3 in the lateral midbrain of wild-type and VinculinΔEpiembryos in late elevation. Cell outlines are visualized with ZO-1. (C) Percentage of cells positive for phospho-histone H3 in wild-type and VinculinΔEpi embryos. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean, 6-7 regions in 3-4 embryos/genotype. (D, E) Detection of the apoptotic cell marker cleaved caspase 3 is similar in wild-type and VinculinΔEpi embryos in early (D) and late (E) elevation. Maximum intensity projections, anterior up. Bars, 10 µm (B), 100 µm (D, E).
Actomyosin localization in VinculinESC and VinculinΔEpi embryos
(A) Localization of myosin IIB in ControlESC and VinculinESC embryos in late elevation. Bottom panels, examples of multicellular junctions with wild-type myosin IIB localization (left) and moderate (cyan) or severe (red) gaps in myosin II localization (right). (B) Localization of F-actin (phalloidin) in ControlESC and VinculinESC embryos in late elevation. (C and D) Normalized intensity profiles of myosin IIB (C) and F-actin (phalloidin) (D) along 3 μm lines perpendicular to bicellular junctions. An average value was obtained for 10 bicellular junctions/neural fold and the mean±SEM between neural folds is shown (8 neural folds in 4 embryos/genotype). Maximum intensity projections, anterior up. Bars, 10 µm (A, top panels, B), 2 µm (A, bottom panels).
Generation and validation of Vinculin mutant ESCs expressing GFP-Plekha7
(A) Deletion breakpoints of Vinculin homozygous mutant clone 41 expressing GFP-Plekha7. Top, 5’ and 3’ gRNA target sites. Red, PAM site. Dotted lines, predicted CRISPR cut sites. Bottom, deletion breakpoints based on deep sequencing of amplicons around the 5’ and 3’ cut sites. (B) Vinculin protein is absent in VinculinESCGFP-Plekha7 embryos (generated from clone 41) and present in ControlESC GFP-Plekha7 embryos (generated from unedited wild-type GFP-Plekha7 ESCs) (one E9.5 embryo/lane). (C) Light micrographs of E9.5 ControlESC GFP-Plekha7 and VinculinESC GFP-Plekha7 embryos (28/28 VinculinESC GFP-Plekha7 embryos and 0/30 ControlESC GFP-Plekha7 embryos displayed exencephaly). Lateral views, dotted lines indicate the lateral edges of the cranial neural plate. Bar, 500 µm.
Quantification of GFP-Vinculin expression from the R26 locus in ESC-derived embryos.
(A) Western blot showing the presence of GFP-Vinculin in addition to the endogenous Vinculin protein detected with the anti-Vinculin antibody (one embryo/lane). (B, C) Vinculin (B) and GFP-Vinculin (C) proteins showed no significant difference in total protein levels between 0-4 somite and 7-12 somite stages. Protein intensity was normalized to the intensity of the β-catenin loading control. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean.
VinculinΔEpi embryos display defects in F-actin and adherens junction localization.
(A and B) Number of gaps in N-cadherin localization at bicellular junctions in a 50 µm x 50 µm region of the lateral midbrain in ControlESC and VinculinESCembryos in early (A) and late (B) elevation. (C, D) Localization of N-cadherin in ControlESC and VinculinESC embryos in early and late elevation (reproduced from Figure 5A and B). Yellow circles show all tricellular and multicellular junctions scored as defective. (E) Localization of N-cadherin and F-actin (phalloidin) in wild-type and VinculinΔEpi embryos in late elevation. (F and G) Localization of GFP-Plekha7 and F-actin (phalloidin) (F) and number of gaps in GFP-Pha7 localization (G) in a 50 µm x 50 µm region of the lateral midbrain in wild-type and VinculinΔEpi embryos in late elevation. Arrowheads show examples of multicellular junctions scored as defective. 6 regions in 3 embryos/genotype in A, B, and G. Maximum intensity projections, anterior up. Bars, 10 µm.
Tight junction and adherens junction protein localization in VinculinESC and VinculinΔEpi embryos
(A) Localization of the tight junction protein ZO-1 and the adherens junction protein GFP-Plekha7 in early elevation ControlESC and VinculinESC embryos. (B, C) Number (left) and areas (right) of adherens junction gaps (B) and tight junction gaps (C) in a 50 µm x 50 µm region in early elevation ControlESC and VinculinESC embryos. Note the differences in scale between the adherens junction and tight junction plots. Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. 38-106 gaps from 6 regions in 3 embryos. *p<0.03 (Welch’s t-test). (D and E) Normalized intensity profiles of ZO-1, GFP-Plekha7, and myosin IIB measured along 3 μm lines parallel to the apical-basal axis of bicellular junctions in XZ reconstructions of AiryScan z-stacks in late elevation wild-type and VinculinΔEpi embryos. An average value was obtained for 10 bicellular junctions/neural fold and the mean±SEM between neural folds is shown (4 neural folds in 2 embryos/genotype). Maximum intensity projections, anterior up. Bar, 10 µm.
Apical constriction, high-order junctions, and cell division initiate correctly in VinculinESC embryos
(A) Stills from time-lapse movies of ControlESC (top) and VinculinESC(bottom) embryos expressing GFP-Plekha7, color coded by apical cell area. (B) The average apical cell area decreases similarly in ControlESC embryos and in VinculinESC embryos that did not display severe defects in cell adhesion. n=4 ControlESC embryos/time point, 5 VinculinESC embryos at 0 h, 4 VinculinESC embryos at 1.5 h, and 2 VinculinESC embryos at 3.0 h. (C, D) The numbers of rosettes (C) and dividing cells (D) were not significantly different in ControlESCand VinculinESC embryos. (E) Division times (time elapsed between the onset of cleavage furrow ingression and the appearance of GFP-Plekha7 at the new vertex or interface) in ControlESC (top) and VinculinESC (bottom) embryos. A single value was obtained for each embryo and the mean±SEM between embryos is shown in (B). Boxes, 25th-75th percentile; whiskers, 5th-95th percentile; horizontal line, median; +, mean. 4-5 embryos/genotype in (C-E). Maximum intensity projections, anterior up. Bars, 10 µm.
