YAP-TEAD1 control of cytoskeleton dynamics and intracellular tension guides human pluripotent stem cell mesoderm specification

Abstract

The tight regulation of cytoskeleton dynamics is required for a number of cellular processes, including migration, division and differentiation. YAP-TEAD respond to cell-cell interaction and to substrate mechanics and, among their downstream effects, prompt focal adhesion (FA) gene transcription, thus contributing to FA-cytoskeleton stability. This activity is key to the definition of adult cell mechanical properties and function. Its regulation and role in pluripotent stem cells are poorly understood. Human PSCs display a sustained basal YAP-driven transcriptional activity despite they grow in very dense colonies, indicating these cells are insensitive to contact inhibition. PSC inability to perceive cell-cell interactions can be restored by tampering with Tankyrase enzyme, thus favouring AMOT inhibition of YAP function. YAP-TEAD complex is promptly inactivated when germ layers are specified, and this event is needed to adjust PSC mechanical properties in response to physiological substrate stiffness. By providing evidence that YAP-TEAD1 complex targets key genes encoding for proteins involved in cytoskeleton dynamics, we suggest that substrate mechanics can direct PSC specification by influencing cytoskeleton arrangement and intracellular tension. We propose an aberrant activation of YAP-TEAD1 axis alters PSC potency by inhibiting cytoskeleton dynamics, thus paralyzing the changes in shape requested for the acquisition of the given phenotype.

Fig. 1. YAP–TEAD1 transcriptional activity controls pluripotent stem cell mechanical properties regardless of contact inhibition.

a Barplot representation of the quantification of cell density in iPSCs grown onto micropatterned surfaces with the indicated diameters (μm), high density hMSCs and hNDFs. Values are expressed as means ± SD (n = 6, *P < 0.05, one-way ANOVA followed by Holm-Sidak’s multiple comparisons test). b Representative confocal images (n = 10) depicting YAP (green) and NANOG (red) expression in iPSCs grown onto micropatterns of the given diameters (μm). c Quantification of YAP distribution within the micropatterned colonies of controlled diameter as quantified by image analysis and expressed as the Pearson’s product moment correlation coefficient (n_{(140 ⌠m)} = 19; n_{(225 ⌠m)} = 9; n_{(500 ⌠m)} = 9; n_{(1000 ⌠m)} = 6). d Barplot representation of the quantification of YAP nucleus/cytoplasm intensity ratio in iPSCs grown onto micropatterned surfaces with the indicated diameters (μm), high-density hMSCs and hNDFs. Values are expressed as means ± SD (n = 6, *P < 0.05, one-way ANOVA followed by Holm-Sidak’s multiple comparisons test). e Representative confocal images of YAP–TEAD-mCherry reporter hESC line cultured onto micropatterns with the indicated diameters (μm) (n = 3). f wordcloud representation of the most significantly represented transcription factors known to bind the sequences identified as YAP targets by ChIP-seq analysis. Font size correlates inversely to −log10(P value). g Left: motif analysis identification of enriched YAP ChIP-seq peaks with the relative statistical significance. Right: barplot representation of enriched YAP ChIP-seq peaks with the relative statistical significance. h Left: graphical representation of TEAD-binding motif density within a 500-bp distance from YAP ChIP-seq peak. Right: western blot analysis for anti-YAP and -panTEAD antibodies in iPSCs immunoprecipitated for YAP endogenous protein. Input and IgG were used as positive and negative controls, respectively (n = 3). i Boxplot representation of the Elastic Modulus (or Young’s Modulus) of single iPSCs transfected with GFP (day 0) or co-transfected with GFP and either YAP-S127A or YAP-5SA/S94A mutants, as obtained by Atomic Force Microscope (AFM) analysis. Values are shown as median ± min/max (n = 12, **P < 0.01, Kruskal–Wallis test followed by post hoc Dunn’s test for multiple comparison). j Boxplot representation of the data obtained by analyzing the Elastic Modulus of single CTR or YAP−/− hESCs transfected with GFP (−) or co-transfected with GFP and either TEAD1 or TEAD4. Values are shown as median ± min/max (n = 12, **P < 0.01, Kruskal–Wallis test followed by post hoc Dunn’s test for multiple comparison). k Top: representative confocal images of NANOG (red) and YAP (green) expression at the centre or at the edge of hESC colonies (n = 3). Bottom: bright-field images of AFM cantilever contacting cells at hESC colony centre or edge and the respective Elastic modulus maps obtained from the measurement.

Fig. 2. Contact inhibition of YAP nuclear localization in PSCs is restored by Tankyrase-p130-AMOT.

a Venn diagram representation of the common pool of YAP-interacting proteins in iPSCs (day 0, n = 3) and iPSC-derived cardiomyocytes (iPSC-CMs, day 15, n = 3) as obtained by mass spectrometry analysis of the endogenous YAP protein. b Graphical representation of YAP interactome in day 15 iPSC-CMs (blue) versus day 0 iPSCs (red). Mass spectrometry results were fed to Cytoscape and analyzed by KEGG database. The size of the origin of the nodes is proportional to the P value (P < 0.01; Kappa score = 0.3). The fractions of the colours are weighted on the number of proteins belonging to the given node at day 0 or day 15. c Western blot analysis of the indicated proteins in iPSCs at the indicated days of cardiac differentiation. Alpha sarcomeric actinin (α-ACTININ) was used as a marker of differentiated cardiomyocytes. GAPDH was used for total protein loading normalization. The blots are representative of three independent experiments. d Left: barplot representation of AMOT RNA fold regulation in H9 hESCs transduced with AMOT-p130 (AMOT), AMOT-p130-Y242/287A or empty vector (mock). The data are indicated as average ± SD n = 2. Right: representative fluorescence-brightfield superimposed image and relative quantification of YAP–TEAD-mCherry hESCs transduced with either AMOT-p130 (AMOT) or AMOT-p130-Y242/287A vectors. Image analysis of mCherry fluorescence within PSC colony is shown. e Representative confocal images depicting AMOT (red) and YAP (green) expression in H9 hESCs treated or not with XAV939 for 48 h (n = 3). (f) Top: western blot analysis of the indicated proteins in cytoplasm (cyto) or nucleus (nu) of iPSCs treated or not with XAV939 for 48 h. GAPDH and LAMIN A/C were used to normalize cytoplasmic and nuclear proteins, respectively. Bottom: quantification of YAP protein levels in cytoplasm (cyto) or nucleus (nu) of iPSCs treated or not with XAV939 for 48 h. The blots are representative of three independent experiments. g Barplot representation of the Elastic Modulus of hESCs (CTR) transfected with either AMOT-p130 (AMOT) or AMOT-p130-Y242/287A, or treated with XAV939 for 48 h as obtained by AFM analysis (n = 12, ****P < 0.0001, Kruskal–Wallis test followed by post hoc Dunn’s test for multiple comparisons).

Fig. 3. YAP–TEAD1 acts downstream of substrate stiffness to transcriptionally control cytoskeleton-related genes and PSC mechanics.

a Graphical representation of the experimental setup used to assess pluripotent stem cell (PSC) response to changes in physiological substrate stiffness. YAP–TEAD-mCherry hESC reporter cells were cultured onto soft surface (0.5 kPa) and then moved to surfaces with increasing stiffness (2, 20 and 64 kPa). b Representative FACS plots depicting mCherry fluorescence in YAP–TEAD-mCherry hESC reporter cells cultured for 48 h on substrates with physiological stiffness (n = 3). c Boxplot representation of the Elastic Modulus of CTR and YAP−/− hESCs grown onto substrates with increasing stiffness (2, 20 and 60 kPa). The values were obtained by AFM and are expressed as Pascal (Pa) (n = 12, *P < 0.05, one-way ANOVA test followed by Holm-Sidak’s test for multiple comparison). d Left: representative confocal images depicting F-actin (green) cytoskeleton arrangement in isogenic (CTR) and YAP−/− hESCs (YAP−/−) grown onto substrates with increasing physiological stiffness. Right: barplot representation of the intensity of green channel (F-actin) in isogenic (CTR) and YAP−/− hESCs (YAP−/−) grown onto substrates with increasing physiological stiffness (n = 3). e Left: orthogonal sections from Z-stack confocal images showing the basal (left) and apical (right) distribution of F-actin in CTR, YAP−/− and YAP −/− hESCs in which YAP has been re-expressed (RESCUE). F-actin is stained with Phalloidin (green) and nuclei counterstained with DAPI (blue). Side views show sagittal sections of the monolayered cells. Right: 3D Z-stack reconstruction and cross-sectional view of the perinuclear actin of CTR, YAP−/− and RESCUE hESCs (n = 3). The images were obtained by IMARIS software after staining with Phalloidin (F-actin, green) and DAPI (nuclei, blue). f Orthogonal sections from Z-stack confocal images showing F-actin organization (green) in CTR and XAV939-treated hESCs for 48 h. Side views show sagittal sections of the monolayered cells. F-actin was stained with Phalloidin (green) and nuclei were counterstained with DAPI. g Boxplot representation of the elastic modulus of CTR, YAP−/− and RESCUE hESCs (n = 12, ****P < 0.0001; *P < 0.05, Kruskal–Wallis test followed by post hoc Dunn’s test for multiple comparisons). h Schematic representation of the strategy followed to discover proteins involved in cytoskeleton organization which are regulated by substrate stiffness through YAP in pluripotent stem cells (PSCs). i Volcano plot representation of differentially regulated genes in CTR versus YAP−/− hESCs grown on substrates with physiological (0.5 and 64 kPa) and tissue culture polystyrene (TCPS). (n = 3, P < 0.05, log2 Fc < |0.58|). j Venn diagram representation of differentially regulated genes in CTR versus YAP−/− hESCs grown on substrates with physiological (0.5 and 64 kPa) and tissue culture polystyrene (TCPS). k Venn diagram representation of PSC YAP bona fide targets that have an annotation for cytoskeleton organization (GO: 0007010) and found dysregulated onto substrate with controlled stiffness (0.5 and 64 kPa) and Tissue culture polystyrene (TCPS). l Volcano plot representation of cytoskeleton-bound proteins significantly regulated in YAP−/− compared to CTR hESCs, as identified by TMT Mass Spectrometry. (n = 5, P < 0.05, log2 Fc > |0.58|). l Left: barplot representation of cytoskeleton-bound proteins significantly regulated in YAP−/− compared to CTR hESCs, as identified by TMT mass spectrometry, that were defined as YAP bona fide targets with cytoskeleton annotation (GO:0007010). Right: identification of TEAD1-binding sites in selected YAP targets with cytoskeleton organization annotation. m Dotplot representation of elastic modulus in CTR and YAP−/− hESCs treated or not with F-actin polymerizing agent jasplakinolide for 24 h as obtained by AFM analysis. (n = 12, ****P < 0.0001, Kruskal–Wallis test followed by post hoc Dunn’s test for multiple comparisons).

Fig. 4. YAP–TEAD1-driven cell stiffening correlates with intracellular tension and determines cell contractile force.

a Boxplot representation of the elastic modulus of isogenic CAL51 and YAP−/− CAL51 cells as obtained by AFM. The values are expressed in Pascal (Pa). (n = 12, ****P < 0.0001, Mann-Whitney test). b Schematic representation of the genetically encoded Förster Resonance Energy Transfer (FRET) sensor based on vinculin tension. c Dotplot representation of FRET index in isogenic CAL51 and YAP−/− CAL51 cells as determined by FRET for vinculin tension sensor (n = 6, ****P < 0.0001, Mann-Whitney test). d Representative traction force maps for paxillin-GFP transfected CTR and YAP−/− CAL51 cells. Stress values are expressed in Pascal (Pa). e Barplot representation of traction forces exerted by CTR and YAP−/− CAL51 cells. The values are represented as median ± SD (n = 6, **P < 0.01, Mann–Whitney test).

Fig. 5. YAP–TEAD1 sustained activation hampers the remodelling of cytoskeleton required for PSC mesoderm specification.

a Representative FACS plots depicting mCherry fluorescence in YAP–TEAD-mCherry PSC reporter cells cultured for 48 h in control (hESCs) or in mesoderm, ectoderm or endoderm media. The data are presented as percentage ± SD (n = 3). b Representative confocal images for the indicated lineage-specific markers (mesoderm, ectoderm and endoderm, in red) as detected in CTR or YAP−/− hESCs after 3 days stimulation with lineage-specific differentiation medium. Nuclei were counterstained with DAPI (blue) and image analysis is shown to quantify the intensity of the fluorescent signals (n = 3). c Barplot representation of the expression of the indicated mesoderm genes in CTR or YAP−/− hESCs induced to mesoderm specification for 2 days (n = 4, *P < 0.05, one-way ANOVA test followed by post hoc Holm-Sidaks test for multiple comparisons). The data are shown as fold regulation ± SD in YAP−/− as compared to CTR hESCs. d Representative confocal images depicting F-actin organization (green) in CTR and YAP−/− cells in the undifferentiated state (day 0) or induced to mesoderm specification (day 2) (n = 3). e Barplot representation of the expression of the indicated mesoderm genes in CTR hESCs cultured in mesoderm differentiation medium, supplemented with jasplakinolide (24 h) for 2 days. The data are shown as fold regulation ± SD in treated as compared to untreated cells (n = 3, *P < 0.05, ANOVA test followed by post hoc Holm-Sidaks test for multiple comparisons). f Barplot representation of the expression of the indicated mesoderm genes in YAP−/− hESCs cultured in mesoderm differentiation medium, supplemented with jasplakinolide for 3 days. The data are shown as fold regulation ± SD in treated as compared to untreated cells (n = 3, no significance found after ANOVA test followed by post hoc Holm-Sidaks test for multiple comparisons). g Barplot representation of the expression of the indicated mesoderm genes in iPSCs transfected with either YAP-S127A or YAP-5SA-S94A, or treated with nucleus export blocker Leptomycin B and induced to mesoderm differentiation for 3 days. The data are shown as fold regulation ± SD in treated cells as compared to CTR (n = 4, *P < 0.05, one-way ANOVA test followed by post hoc Holm-Sidaks test for multiple comparisons). h Schematic representation of the model proposed for YAP–TEAD1 interference with cytoskeleton remodelling during PSC mesoderm specification.

Fig. 6. Proposed model for Tankyrase-mediated regulation of cytoskeleton stability and cell mechanics through AMOT and YAP–TEAD1 during PSC specification.

Left: pluripotent stem cells (PSCs) growing into high confluence colonies display high Tankyrase activity, in turn keeping the levels of AMOT low, independently of cell–cell interactions. Under such circumstances, YAP is free to shuttle to the nucleus and modulate, among the others, the expression of genes involved in actin stability, like TNIK, PAK1, THY1 and MID1. Right: during mesoderm specification, Tankyrase activity is low, AMOT protein expression increases, so that in confluent cells, YAP can be restricted to the cytoplasm in response to cell–cell contact. In these conditions, cytoskeleton remodelling can occur.