YAP Overcomes Mechanical Barriers to Induce Mitotic Rounding and Adult Cardiomyocyte Division

Abstract

Background: Many specialized cells in adult organs acquire a state of cell cycle arrest and quiescence through unknown mechanisms. Our limited understanding of mammalian cell cycle arrest is derived primarily from cell culture models. Adult mammalian cardiomyocytes, a classic example of cell cycle arrested cells, exit the cell cycle postnatally and remain in an arrested state for the life of the organism. Cardiomyocytes can be induced to re-enter the cell cycle by YAP5SA, an active form of the Hippo signaling pathway effector YAP.

Methods: We performed clonal analyses to determine the cell cycle kinetics of YAP5SA cardiomyocytes. We also performed single-cell RNA sequencing, marker gene analysis, and functional studies to examine how YAP5SA cardiomyocytes progress through the cell cycle.

Results: We discovered that YAP5SA-expressing cardiomyocytes divided efficiently, with >20% of YAP5SA cardiomyocyte clones containing ≥2 cardiomyocytes. YAP5SA cardiomyocytes re-entered cell cycle at the G1/S transition and had an S phase lasting ≈48 hours. Sarcomere disassembly is required for cardiomyocyte progression from S to G2 phase and the induction of mitotic rounding. Although oscillatory Cdk expression was induced in YAP5SA cardiomyocytes, these cells inefficiently progressed through G2 phase. This is improved by inhibiting P21 function, implicating checkpoint activity as an additional barrier to YAP5SA-induced cardiomyocyte division.

Conclusions: Our data reveal that YAP5SA overcomes the mechanically constrained myocardial microenvironment to induce mitotic rounding with cardiomyocyte division, thus providing new insights into the in vivo mechanisms that maintain cell cycle quiescence in adult mammals.

Keywords: Hippo pathway; P21; YAP; cell cycle; mitotic rounding; myocyte proliferation; sarcomere disassembly.

Fig 1.. Analysis of YAP5SA-induced cell cycle.

A-C: Analysis of cytokinesis using R26F-Confetti line. A: (Left) Timing of the injections and sample harvesting. Low dose tamoxifen (Tm) was injected into MCM; R26R-Confetti mice, then AAV9-YAP5SA were injected 2 days after. Saline was injected as control. Hearts were harvested in 2 weeks after virus injection. (Right) Schematic presentation of clonal analysis using R26R-Confetti mouse line. Low dose tamoxifen results in single cell labeling of fluorescent proteins. Cell proliferation results in clusters of 2 or more cells while non-proliferative cell remain as a clone of single cell. B: Representative images of control and YAP5SA hearts for YFP, WGA, and merged. Sections were stained with WGA to delineate cell boundary. Arrow show YFP-expressing CMs. C: Quantification of clonal formation in the control and YAP5SA hearts. Groups were compared using Chi-square test for distribution. Double and 3+ clones within treatments were compared using Fisher’s exact test (n=5 for control, n=6 for YAP5SA). **p<0.01, ****P<0.0001. D-I: Expression of cell cycle markers in adult heart were determined by immunohistochemistry. D: Timing of tamoxifen (Tm) injection and sample harvesting for panels E-I. YAP5SA and control mouse lines were used. E: Representative images of PCNA and cTNT expression in control and YAP5SA expressing hearts in transverse sections. Yellow arrowheads are PCNA positive nuclei in CMs. F: Quantification of PCNA positive nuclei. Groups were compared using the Mann-Whitney U test (n=5 each). **p<0.01. G: Timing of cyclin D1 (CCND1), CDK2, CYCLIN A2 (CCNA2), CDK1, and PHH3 expression during cell cycle. H: Quantification of CCND1, CDK2, CCNA2, CDK1, and PHH3 positive CM nuclei (n=5). I: Representative images of CCND1, CDK2, CCNA2, and CDK1 in YAP5SA expressing hearts. Representative images of PHH3 staining are in Fig. S2H, I.

Fig 2.. Cell cycle re-entry in adult cardiomyocytes.

A: Overview of the proteins expressed in G1 and G1/S phase analyzed in this figure. # are Yap targets. B: Timing of tamoxifen (Tm) injection and sample harvesting for panels C-G. YAP5SA mouse line was used and hearts were harvested 1–5 days post tamoxifen. C-E: Expression of cyclin D1 (CCND1) and PCNA was analyzed 1–5 days after YAP5SA induction. C: Representative image of YAP5SA heart at day3 show non-myocyte CCND1 expression (white arrows). PCNA expressed in both CMs (yellow arrowheads), and non-myocyte (white arrowheads). D, E: Quantification of CCND1 (D) and PCNA (E) expressing nuclei in CMs. Groups were compared using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each time point). *p<0.05, ***p<0.001, ****p<0.0001. F, G: Expression of CDK2, CCNE1 (cyclin E1) and P-RB were analyzed at 2 and 3 days after tamoxifen injection. F: Representative images of CDK2, CCNE1, and P-RB in YAP5SA expressing hearts at day 3. Arrows show positive CM nuclei. G: Quantification of CDK2, CCNE, and P-RB expressing nuclei in CMs. Groups were compared using Mann-Whitney analysis. **p<0.01. H-J: Flow analysis of DNA synthesis. H: Timing of injections and harvesting (Left), and steps of flow analysis (Right). I: Representative FACS image of PCM1 positive nuclei (top) and EdU-stained nuclei in PCM1 population. J: Quantification of EdU incorporation in PCM1-positive CM nuclei of control (n=5 experiments) and YAP5SA expressing hearts (n=6 experiments). Groups were compared using Mann-Whitney test. **p<0.01.

Fig 3.. Cell cycle analysis of adult cardiomyocytes by EdU pulse and cell cycle markers.

A: Timing of tamoxifen (Tm) and EdU injection, and sample harvesting. EdU was injected 3 days after tamoxifen injection and hearts were harvested 24, 48, and 72 hours after EdU injection. B: Markers used to determine cell cycle length in pulse-labelled cells. CCNA2 (late S to early G2 phase), CDK1 (late G2 to M phase), and CCND (G1 phase) were co-labelled with EdU. C, D: Representative images and quantification of double positive nuclei for EdU and cell cycle markers CCNA2 (C) and CDK1(D) in YAP5SA-expressing CMs. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each time point). ****p<0.0001. E-H: Nucleation and ploidy analysis. E: Timing of tamoxifen (Tm) and EdU administration, and sample harvesting. EdU was given by Alzet pump at 3 days for 24 hr and hearts were harvested at day 4 and 6 of tamoxifen injection. F: Representative images of isolated CMs from day 4 and 6. CMs were stained for DAPI, EdU and cTNT. G: Nucleation of EdU+ CMs. Proportion of mono-nucleated (MonoNuc) and binucleated (BiNuc) CMs were compared between CMs from day 4 and day 6. Chi-square test was used to compare two groups (n=3 each group). **** p<0.0001. H: Ploidy of EdU+ CM nuclei. Day 4 MonoNuc (47 nuclei), Day 4 BiNuc (288 nuclei), Day 6 MonoNuc (136 nuclei), and Day 6 BiNuc (181 nuclei) were analyzed. Percentage of ploidy in individual samples were shown in Fig. S3I. Chi-square test was used to compare groups. *p<0.05, *** p<0.001, **** p<0.0001. I: Representative images and quantification of double positive nuclei for EdU and CCND1 in YAP5SA-expressing CMs. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each time point). ****p<0.0001. J, K: Analysis of EdU twin spots. J: Representative images of the EdU and CCND1 double positive twin spot (arrows) at 72 hours post EdU injection. K: Number of EdU twin spots, and CCND1 positive nuclei among EdU twin spots (n=5). Chi-square test was used. *p<0.05. Percentage of twin spots in each time point, and percentage of CCND1 positive EdU twin spot per sample are shown in Fig. S3J and K.

Fig 4.. Single cell analysis in YAP5SA heart.

A-E: Analysis of YAP5SA and control CM drop-seq 6 days after TAM injection. A: UMAP of re-clustered CMs. B: Cell composition graph of 5 clusters. C: Top 10 Differentially expressed genes, ranked by descending fold change from each cluster. D: Representative GO for 5 clusters. GO analysis from each cluster is shown in the Fig. S6C–G. E: Cell trajectory analysis of clusters 2, 4, and 5 colored by pseudotime (top). Cellular density of each cluster along Comp1 axis (middle). Patterning of trajectory determining genes along pseudotime (bottom). F-H: Comparison between adult and P2 CM datasets. F: Volcano plots of differentially expressed genes between adult control and P2 sham CMs from the CM cluster shown in Fig. S6. G: UMAP of CM clusters from the drop-seq analysis. Same image as in Fig. S5D. H: P2 CM score. P2 data from the Fig. 4F was mapped on the Fig. 4G.

Fig 5.. Verification of single cell analysis by immunohistochemistry.

A: Timing of tamoxifen (Tm) injection and sample harvesting for panels B-E. B-C: Sarcomere disassembly was determined by cTNT staining. Representative images of the control, and YAP5SA day 2, day 4, and day 5 after tamoxifen are shown. Asterisks show CMs with sarcomere disassembled. C: Representative images of sarcomere structure imaged by SIM microscopy. D: Quantification of sarcomere disassembly. Disassembly score is a percentage of the area devoid of cTNT staining. Score of each CMs were measured. Groups were compared by using nested one-way ANOVA with the Tukey post-hoc test for multiple comparisons (n=5 each time point, 20 CMs each time point). **p<0.01, ***p<0.001. E: Quantification of CDK1 positive CMs from day 2 to 5. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each time point). ***p<0.001. F: Timing of tamoxifen (Tm) injection and sample harvesting for panels G-J. G: Representative image of cTNT staining in CDK1 positive CM at Day 6. H: Quantification of disassembly score in Flag-positive CMs and CDK1-positive CMs. ****p<0.0001. Groups were compared by using nested t-test (n=5 each group). I, J: Sarcomere disassembly and SPTAN1 expression. I: Representative images of control and YAP5SA hearts stained with SPTAN1, cTNT, and DAPI. J: Quantification of SPTAN1 intensity and sarcomere disassembly score. Each group was analyzed with simple linear regression (n=5 each). Only YAP5SA group is significant and R is shown. ****P<0.0001. K-M: Effect of SPTAN1 haploinsufficiency in YAP5SA-induced CMs. K: Timing of tamoxifen (Tm) injection and sample harvesting for panels L, M. L: Representative images of YAP5SA and YAP5SA; Sptan1 het for cTNT, WGA, and DAPI. M: Sarcomere disassembly and CDK1 expression of Sptan1 het (control), YAP5SA, and YAP5SA; Sptan1 het CMs were quantified. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=3 each genotype). **p<0.01, ****p<0.0001.

Fig 6.. Mitotic Rounding in YAP5SA CMs.

A-E: Expression of mitotic rounding genes in control and YAP5SA CMs. A-Expression of 45 mitotic rounding genes were examined in CM re-cluster 1–5. B: UMAP of re-clustered CMs. Same image as in Fig. 4A. C: Mitotic rounding score. D: Feature plots of representative genes, Myh9, Actn4, and F2r. E: Quantification of rounding score. Groups were compared by one-way ANOVA. *p<0.05, ***p<0.001, ****p<0.0001. F, G: Solidity of the control and YAP5SA CMs. Random CMs from control and FLAG-positive YAP5SA CMs were analyzed for solidity. Groups were compared by using nested t-test (n=5 each group, 20 CMs each heart). ****P<0.0001. H: Solidity of the CMs from YAP5SA hearts. CMs of Flag(−), Sptan1(+), and CDK1 (+), were analyzed for solidity. Groups were compared by using one-way Anova (n=5). *P<0.05, **P<0.01, ***P<0.001. ****P<0.0001. I, J: Solidity of YAP5SA; Sptan1 het CMs. Same samples as Fig. 5K–M were used. I: Representative images of YAP5SA and YAP5SA; sptan1 het CMs. J: Solidity of control, YAP5SA, Yap5SA; Sptan1 het CMs. Groups were compared by using one-way Anova (n=3 each genotype). ***P<0.001, ****P<0.0001.

Fig 7.. Phospho-mimetic YAP blocks cell cycle progression in YAP5SA-expressing cardiomyocytes.

A: Timing of AAV9-YAP127D, tamoxifen (Tm), and EdU injection, and sample harvesting for the panels B-F. B, C: Sarcomere disassembly was determined by cTNT staining in Flag-positive CMs. B: Representative images of the YAP5SA + AAV9-GFP, and YAP5SA + AAV9-YAPS127D are shown. Asterisks show flag-positive CMs. C: Quantification of sarcomere disassembly in the control (MCM + AAV9-S127S) and YAP5SA plus GFP or YAPS127D. Groups were compared by nested one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each). *p<0.05, ***p<0.001, ****p<0.0001. D: Solidity of control, YAP5SA, and YAP5SA; YAPs127D CMs. Groups were compared by nested one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each). ****p<0.0001. E: Cell cycle stages were determined in control, YAP5SA, and YAP5SA; YAPs127D CMs. CDK2, CCNA2, CDK1, PHH3, and CCND1 expressing CMs were quantified. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. F: EdU pulse labelled CMs at 72 hours after EdU injection. Double positives for EdU and CDK2 or CCND1 were examined. Groups were compared using the Mann-Whitney U test (n=5 each). **p<0.01. G-I: Confetti analysis with YAP5SA and YAP5SA + YAPS127D hearts. G: Timing of AAV9-YAP127D and Tm injection, and sample harvesting for the panels H, I. H: Representative images of YAP5SA; confetti and YAP5SA; confetti with AAV-YAP S127D. I: Quantification of clonal formation in the YAP5SA and YAP5SA+YAPS127D hearts. Groups were compared using Chi-square test for distribution (n=4). ****p<0.0001. J-O: Inhibition of sarcomere disassembly by YAP-S127D during endogenous regeneration. J: Timing of injections, surgery, echo analysis and sample harvest for panels K-O. K: Representative images of trichrome stained hearts for sham and MI hearts at 28 dpmi. L: Quantification of scar size at 28 dpmi. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=12 for both sham combined, n=6 for saline MI, n=6 for S127D MI). **p<0.01, ****p<0.0001. M: Ejection fraction and fractional shortening at 8 and 28 dpmi. Groups were compared by using two-way ANOVA (n=6 each treatment). ***p<0.001, ****p<0.0001. N-O: Sarcomere disassembly and PHH3 expression at 4 dpmi. N: Representative images of MI heart with saline and S127D injection. O: (Left) Quantification of sarcomere disassembly. Groups were compared by nested t-test (n=6 each). (Right)Quantification of PHH3 expressing CMs. Groups were compared using the Mann-Whitney U test (n=6 each). **p<0.01.

Fig 8.. Function of P21 in YAP5SA-expressing CMs.

A-C: Timing of P21 nuclear accumulation. A: Timing of tamoxifen (Tm) injection and sample harvesting for panels B, C. B, B: Representative images of P21 immunostaining in YAP5SA hearts. Arrows are P21 positive CM nuclei. C: Quantification of P21-expressing CM nuclei of the control and YAP5SA from day2 to 5. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each time point).: **p<0.01, ****p<0.0001. E-N: Inhibition of P21 in YAP5SA-expressing animals. D: P21 inhibitor UC2288 inhibits gene expression of cdkn1a. E: Injection protocol of tamoxifen (Tm) and P21 inhibitor, and sample harvesting timing for panels F-H. F: Survival of vehicle (n=10) and P21 inhibitor (n=12)-treated YAP5SA animals at day 5. Groups were compared with qui square test. *p<0.05. G: Representative images of P21 and CDK1 immunostaining in YAP5SA+vehicle and YAP5SA+P21 inhibitor are shown. Arrows are P21 positive or CDK1 positive CM nuclei. H: Quantification of P21, CDK1, and PHH3 positive CM nuclei. Groups were compared by using one-way ANOVA with the Tukey post-hoc test for pairwise comparisons (n=5 each treatment). **p<0.01, ****p<0.0001. I-K: Confetti analysis of YAP5SA vehicle and P21 inhibitor (P21 inh) treated hearts. I: Timing of Tamoxifen and viral injection, and sample collection for panels J, K. J: Representative images of vehicle and P21 inh treated YAP5SA hearts. K: Quantification of clonal formation in the vehicle and P21 inh treated YAP5SA hearts. Groups were compared using Chi-square test for distribution (n=5). ****p<0.0001. L: Model for YAP5SA-induced adult CM cell cycle.