ASPP2 maintains the integrity of mechanically stressed pseudostratified epithelia during morphogenesis

Abstract

During development, pseudostratified epithelia undergo large scale morphogenetic events associated with increased mechanical stress. Using a variety of genetic and imaging approaches, we uncover that in the mouse E6.5 epiblast, where apical tension is highest, ASPP2 safeguards tissue integrity. It achieves this by preventing the most apical daughter cells from delaminating apically following division events. In this context, ASPP2 maintains the integrity and organisation of the filamentous actin cytoskeleton at apical junctions. ASPP2 is also essential during gastrulation in the primitive streak, in somites and in the head fold region, suggesting that it is required across a wide range of pseudostratified epithelia during morphogenetic events that are accompanied by intense tissue remodelling. Finally, our study also suggests that the interaction between ASPP2 and PP1 is essential to the tumour suppressor function of ASPP2, which may be particularly relevant in the context of tissues that are subject to increased mechanical stress.

Fig. 1. The ASPP2/PP1 complex is not required during preimplantation development.

a ASPP2 was detected by indirect immunofluorescence in E3.5 embryos to analyse its localisation pattern. A cross-section through the equatorial plane of a representative embryo is shown (top row), as well as a 3D opacity rendering of the same embryo (bottom row). The F-actin cytoskeleton and nuclei were visualised using Phalloidin and DAPI, respectively. A magnified image of the dashed area is shown on the right. Note how ASPP2 colocalises with F-actin at the apical junctions in cells of the trophectoderm (white arrowheads). The juxtaposed graph shows ASPP2 and F-actin signal intensity along the apical-basal axis of five cell-cell junctions in the TE. Error bars represent ±SD. AJ apical junction, B base of the trophectoderm. Scale bars: 20 and 5 μm (for the magnification). b The localisation pattern of YAP and Par3 was analysed in wild type and ASPP2^RAKA/RAKA embryos by indirect immunofluorescence. A cross-section of representative embryos through the equatorial plane shows the localisation of YAP in the nuclei of the trophectoderm in both wild type and ASPP2^RAKA/RAKA embryos. Maximum intensity projections of these embryos show the localisation of Par3 at the level of apical junctions in the trophectoderm (representative images from six wild type and eight ASPP2^RAKA/RAKA embryos). Scale bar: 20 μm. c ASPP2 knockdown in E3.5 embryos using siRNA against ASPP2 mRNA. ASPP2 knockdown was confirmed by indirect immunofluorescence. Note how signal at the apical junctions is specific to ASPP2 and how YAP is normally localised to the nuclei of TE cells in ASPP2-depleted embryos. Representative images from n = 19 control siRNA-injected embryos and n = 24 ASPP2 siRNA-injected embryos across two independent experiments. Scale bar: 20 μm. Source data are provided as a Source Data file.

Fig. 2. Absence of ASPP2 expression leads to structural defects in the epiblast.

a Immunofluorescence of wild type and ASPP2^ΔE4/ΔE4 E6.5 embryos using an anti-Par6 antibody. The phenotypic variability of ASPP2^ΔE4/ΔE4 embryos is illustrated, with embryos either lacking cavities (middle row, five out of nine embryos) or exhibiting smaller cavities (bottom row, four out of nine embryos). The green dashed line highlights the ectopic accumulation of cells in the epiblast of ASPP2^ΔE4/ΔE4 embryos. b Magnification of the corresponding regions shown in panel a. Blue arrowheads highlight the enrichment of F-actin at the apical junctions in the epiblast. Note how F-actin is not enriched at the apical junctions but is instead more homogenously distributed across the apical surface of epiblast cells in ASPP2^ΔE4/ΔE4 embryos (orange arrowhead). The insets within images are 2x magnifications of the corresponding dashed areas. c Quantification of F-actin signal intensity along the apical surface of epiblast cells of wild type (n = 3 embryos, five measurements per embryo) and ASPP2^ΔE4/ΔE4 embryos (n = 3 embryos, five measurements per embryo). Measurements were made on cross-sections along the apical domain of individual epiblast cells from apical junction to an apical junction (represented with a blue background in the graph). The 95% confidence interval is represented by the grey area. See material and methods for details. d Immunofluorescence of wild type (representative images from eight embryos) and ASPP2^ΔE4/ΔE4 E5.5 embryos (representative images from five embryos) using an anti-Laminin antibody. e Magnification of the corresponding dashed areas in panel d. f Immunofluorescence of wild type (representative images from seven embryos) and ASPP2^ΔE4/ΔE4 (representative images from two embryos) E6.5 embryos using an anti-SCRIB antibody. g Magnification of the corresponding dashed areas in panel f. Green arrowheads highlight basolateral SCRIB. Note the enrichment of SCRIB at the apical junctions in the epiblast of wild type embryos (blue arrowhead) and its absence in the corresponding localisation in ASPP2^ΔE4/ΔE4 embryos (orange arrowhead). Nuclei and the F-actin cytoskeleton were visualised with DAPI and Phalloidin, respectively. Scale bars: 20 μm. Source data are provided as a Source Data file.

Fig. 3. ASPP2^EpiΔE4/ΔE4 and ASPP2^RAKA/RAKA embryos phenocopy ASPP2^ΔE4/ΔE4 embryos.

a The expression of ASPP2 was conditionally ablated in the epiblast to test for its epiblast-specific requirement (ASPP2^EpiΔE4/ΔE4 embryos). The ASPP2 expression pattern was analysed by indirect immunofluorescence in ASPP2^EpiWT/ΔE4 (representative images from four embryos) and ASPP2^EpiΔE4/ΔE4 (representative images from four embryos) embryos. ASPP2 proteins were completely absent at the apical junction of epiblast cells in ASPP2^EpiΔE4/ΔE4 embryos. Note that the ASPP2 antibody results in a nonspecific nuclear signal (also seen in Fig. 1c when depleting ASPP2 by siRNA). The dashed area highlights the ectopic accumulation of cells in the epiblast. b Immunofluorescence of wild type (representative images from seven embryos) and ASPP2^RAKA/RAKA (representative images from three embryos) E6.5 embryos using an anti-Par6 antibody. The green dashed line highlights the ectopic accumulation of cells in the epiblast of ASPP2^RAKA/RAKA embryos. c Magnification of the corresponding dashed regions in b. Note the reduced amount of Par6 along the apical domain of epiblast cells in ASPP2^RAKA/RAKA embryos (orange arrowhead). Nuclei and the F-actin cytoskeleton were visualised with DAPI and Phalloidin, respectively. Scale bars: 20 μm.

Fig. 4. ASPP2 is required for apical daughter cell reincorporation into the epiblast following cell division events.

a Time-lapse imaging of wild type and ASPP2^ΔE4/ΔE4 embryos. mT/mG-labelled cell membranes were used to manually track cell movement. Yellow dots highlight mother cells at the apical surface of the epiblast immediately prior to a cell division event. Green and magenta dots identify the resulting daughter cells. Note how both daughters reintegrate the epiblast in the wild type whereas one of the two daughters fails to do so in the absence of ASPP2 even after a prolonged period of time (t = 82.5’). b Diagram illustrating the method used to quantify daughter cell movement following cell divisions. Daughter cell movement was characterised by both the distance travelled (d) and the direction of travel (θ) expressed as the angle between the reference vector (the green vector starting from the initial position of the mother cell prior to the division event to the centre of the embryonic region) and the vector characterising absolute daughter cell movement (the red vector starting from the initial position of the mother cell prior to the division event to the final position of the daughter cell). The left panel illustrates the case of a daughter moving basally to reincorporate the epiblast and the right panel describes abnormal daughter cell movement towards the centre of the embryonic region such as seen in ASPP2^ΔE4/ΔE4 embryos. c Graph quantifying cell movement in wild type (n = 3 embryos, 56 cells) and ASPP2^ΔE4/ΔE4 embryos (n = 3 embryos, 66 cells). For a given pair of daughter cells, each daughter was defined as ‘apical’ or ‘basal’ depending on their respective position relative to the centre of the embryonic region immediately after a cell division event. d Proportion of daughter cells with an overall apical or basal movement in wild type (n = 3 embryos, 56 cells) and ASPP2^ΔE4/ΔE4 embryos (n = 3 embryos, 66 cells). Left panel: Quantification of the proportion of daughter cells with an overall apical (θ from 0° to 90°) or basal movement (θ from 90° to 180°) in wild type and ASPP2^ΔE4/ΔE4 embryos. Right panel: quantification of the proportion of apical and basal daughters with an overall apical (θ from 0° to 90°) or basal movement (θ from 90° to 180°) in wild type and ASPP2^ΔE4/ΔE4 embryos. ****p < 0.0001, NS non-significant (two-sided Fisher’s exact test of independence. The Bonferroni method was used to adjust p values for multiple comparisons. P values from left to right: p = 6.16e-07, p = 1.62e-08, p = 2.83e-09, p = 7.08e-01). Source data are provided as a Source Data file.

Fig. 5. ASPP2 is required for epithelial integrity in the primitive streak.

a Posterior thickening in E7.5 ASPP2^RAKA/RAKA embryos in a BALB/C background. Left panel: the anteroposterior axis was defined using AMOT localisation pattern. Right panel: comparison of tissue thickness in the anterior (three measurements per embryo) and the posterior (three measurements per embryo) of wild type (n = 5 embryos) and ASPP2^RAKA/RAKA embryos (n = 5 embryos). For the box plots, the top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The whiskers show the minima to the maxima values and the central line indicates the median. *p < 0.05, ****p < 0.0001 (nested ANOVA, p values from left to right: p = 0.047, p = 3.95e-5). b Cells accumulate in the primitive streak region of ASPP2^RAKA/RAKA embryos. Immunofluorescence of E7.5 wild type (representative images from 55 embryos) and ASPP2^RAKA/RAKA (representative images from 23 embryos) embryos using a T (Brachyury) antibody. c Cells ectopically accumulating in the primitive streak region are unable to apically constrict and do not have enriched F-actin at the apical junctions (area delineated by the dotted orange line) in comparison to wild type (blue arrowheads and magenta dotted lines). Green dotted ROI: the apical surface of the epiblast in the lateral region of the embryo. Lower panel: Magnified regions highlighted in green and orange, respectively. Right panel: quantification of F-actin signal intensity along the apical surface of epiblast cells in the primitive streak region of wild type (n = 3 embryos, five cells per embryo) and ASPP2^RAKA/RAKA embryos (n = 3 embryos, five cells per embryo). The 95% confidence interval is represented by the grey area. d–f Airyscan imaging reveals the extent of F-actin disorganisation at the surface of cells accumulating ectopically in the primitive streak region of ASPP2^RAKA/RAKA embryos. d 3D opacity rendering of embryo optical halves, enabling visualisation of the apical surface of epiblast cells in the proamniotic cavity. Note the absence of the typical F-actin mesh pattern at the apical surface of cells in the posterior of ASPP2^RAKA/RAKA embryos (green dotted line). e Cross-section through the primitive streak region, showing enriched F-actin at the apical junctions of wild type (representative image from three embryos) embryos (blue arrowheads) and the formation of F-actin spike-like structures at the contact-free surface of ASPP2^RAKA/RAKA (representative image from three embryos) embryos. f En face view of the epiblast’s apical surface in the posterior of an ASPP2^RAKA/RAKA embryo. Green dotted lines demarcate the disorganised apical region of the posterior and the more organised lateral regions of the epiblast. Right panel: magnification of the epiblast’s apical surface in the posterior of an ASPP2^RAKA/RAKA embryo showing F-actin forming spike-like structures. Nuclei and the F-actin cytoskeleton were visualised with DAPI and Phalloidin, respectively. Scale bars: 50 μm (a–c), 20 μm (e). Source data are provided as a Source Data file.

Fig. 6. ASPP2^RAKA/RAKA embryo are more susceptible to mechanical stress.

a, b FLIM measurements of the FLIPPER-TR tension probe in E6.5 embryos. a Representative FLIM image of an E6.5 embryo. Note that Lifetime smaller than 3.75 and higher than 4.75 are blue and red, respectively. b Mean lifetime values at the apical surface of the epiblast, exVE and emVE (n = 9 embryos). For the box plots, the top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The whiskers show the minima to the maxima values and the central line indicates the median. ****p < 0.0001, ***p < 0.001, **p < 0.01 (ANOVA, followed by Tukey’s test. P values from left to right: p = 0, p = 1.52e-4, p = 1.41e-3). c Localisation pattern of SHROOM2 in E6.5 embryos. Blue arrowhead highlights the accumulation of SHROOM2 at the apical junctions in the epiblast. d wild type (n = 4) and ASPP2^RAKA/RAKA (n = 2) embryos were grown for 30′ in cylindrical cavities made of biocompatible hydrogels. The localisation pattern of GATA6 and Myosin was then analysed by immunofluorescence. e Magnification of the embryos shown in b. The green dotted line highlights the ectopic accumulation of cells seen in ASPP2^RAKA/RAKA embryos. Note how Myosin is enriched at the apical junctions of wild type epiblast cells (blue arrowheads). The orange arrowhead points to the abnormal distribution of Myosin at the apical surface of these cells. Nuclei and the F-actin cytoskeleton were visualised with DAPI and Phalloidin, respectively. Scale bars: 20 μm. Source data are provided as a Source Data file.

Fig. 7. ASPP2 controls the localisation of apical F-actin and tensions in the epiblast.

a 3D opacity rendering showing the localisation of ASPP2 in E5.5 wild type embryos at the apical junctions of the visceral endoderm where it colocalises with F-actin. b Cross-section (top row) and 3D opacity rendering (bottom row) of the proamniotic cavity showing the localisation pattern of ASPP2 and F-actin at the apical junctions (white arrowhead). c The outer surface of the VE and apical surface of the epiblast were computationally ‘unwrapped’, revealing the enrichment of ASPP2 at specific locations along the apical junctions, often at F-actin-rich tricellular junctions (green arrowheads). a–c Representative images from six embryos. d The interaction between endogenous ASPP2 and the F-actin-binding protein Afadin was examined in Caco-2 cells by co-immunoprecipitation (representative images from three independent experiments). Molecular weights are indicated in kilodaltons. e The localisation pattern of endogenous ASPP2 and Afadin in Caco-2 cells was examined by immunofluorescence (representative images from five independent experiments). The bottom row represents the magnified region highlighted by a dotted box and shows the enrichment of ASPP2 and Afadin at tricellular junctions. ASPP2, Afadin and F-actin signal intensity was quantified across tricellular junctions (graph on the right). f The localisation pattern of Afadin in the proamniotic cavity was analysed by immunofluorescence in E6.5 wild type embryos. The blue arrowhead highlights the colocalisation of Afadin with F-actin at a tricellular junction. g The localisation pattern of F-actin was analysed by time-lapse microscopy in wild type (representative images from ten embryos) and ASPP2^RAKA/RAKA (representative images from six embryos) LifeAct-GFP positive embryos. Note how apical F-actin is disrupted in ASPP2^RAKA/RAKA LifeAct-GFP positive embryos (orange arrowhead). The colour scale represents pixel intensity (grey levels). h At later time points, the ectopic accumulation of cells in the epiblast of ASPP2^RAKA/RAKA LifeAct-GFP positive embryos was evident (dotted line). i Representative FLIM images of ASPP2^EpiWT/ΔE4 (n = 9 embryos) and ASPP2^EpiΔE4/ΔE4 (n = 7 embryos) embryos (left) and comparison of mean lifetime values in the epiblast tissue, including delaminating cells (right). The dotted line highlights epiblast cells. For the box plots, the top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The whiskers show the minima to the maxima values and the central line indicates the median. Outliers are represented with black dots. **p < 0.01 (unpaired two-sided Student’s t-test, p = 4.84e-3). Nuclei and the F-actin cytoskeleton were visualised with DAPI and Phalloidin respectively. Scale bars: 20 μm. Source data are provided as a Source Data file.

Fig. 8. ASPP2 is not required for cell fate specification during gastrulation.

a The primitive streak expands comparatively in E7.5 wild type and ASPP2^ΔE4/ΔE4 embryos. Mesoderm cells were labelled by immunofluorescence using an antibody against Brachyury (T). b Patterning proceeds normally in the absence of ASPP2. The ectoderm and cardiac progenitors were labelled in E8.5 wild type and ASPP2^ΔE4/ΔE4 embryos with antibodies against SOX2 and NKX2.5, respectively. ys yolk sack, al allantois, s somites, hf head fold, am amnion, pc proamniotic cavity. c Cardiac progenitors can differentiate into cardiomyocytes in E9.5 ASPP2^ΔE4/ΔE4 embryos. The presence of the contractile machinery (magenta arrowheads) was assessed in wild type and ASPP2^ΔE4/ΔE4 embryos using an antibody against sarcomeric α-actinin. Nuclei and the F-actin cytoskeleton were visualised with DAPI and Phalloidin, respectively. Scale bars: 50 μm (a, c), 100 μm (b).

Fig. 9. ASPP2 is required for tissue integrity across a variety of pseudostratified epithelia.

a Somite architecture is disrupted in ASPP2^ΔE4/ΔE4 embryos. The dotted line highlights the contour of a somite in an ASPP2^ΔE4/ΔE4 embryo. The star indicates the ectopic accumulation of cells in the centre of this somite. Arrowheads point to mitotic figures. b–e Quantification of somite characteristics in wild type (n = 10 embryos, 58 somites) and ASPP2^ΔE4/ΔE4 (n = 6 embryos, 35 somites) embryos at E8.5. For the box plots, the top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The whiskers show the minima to the maxima values and the central line indicates the median. Outliers are represented with black dots. *p < 0.05, ****p < 0.0001 (unpaired two-sided Student’s t-test; p = 0 in b, p = 3.093e-05 in c, p = 0.025 in d, p = 1.116e-08 in e). f Apical-basal polarity is defective in the somites of ASPP2^ΔE4/ΔE4 embryos. Par6 localised apically in wild type somites (arrowhead) whereas it was absent in ASPP2^ΔE4/ΔE4 embryos (star). de definitive endoderm. g Head fold formation is defective in ASPP2^RAKA/RAKA embryos. The organisation of apical F-actin was disorganised locally in the anterior ectoderm of ASPP2^RAKA/RAKA embryos (orange dotted line). F-actin signal intensity along the apical surface of ectoderm cells in disrupted areas in ASPP2^RAKA/RAKA embryos (n = 3 embryos, five cells per embryo) was compared to wild type cells (n = 3 embryos, five cells per embryo). Measurements were made on cross-sections along the apical domain of individual ectoderm cells from apical junction to the apical junction (represented with a blue background in the graph). The 95% confidence interval is represented by the grey area. Nuclei and the F-actin cytoskeleton were visualised with DAPI and Phalloidin, respectively. Scale bars: 20 μm (a, f), 100 μm (g). Source data are provided as a Source Data file.