Hippo Deficiency Leads to Cardiac Dysfunction Accompanied by Cardiomyocyte Dedifferentiation During Pressure Overload

Abstract

Rationale: The Hippo pathway plays an important role in determining organ size through regulation of cell proliferation and apoptosis. Hippo inactivation and consequent activation of YAP (Yes-associated protein), a transcription cofactor, have been proposed as a strategy to promote myocardial regeneration after myocardial infarction. However, the long-term effects of Hippo deficiency on cardiac function under stress remain unknown.

Objective: We investigated the long-term effect of Hippo deficiency on cardiac function in the presence of pressure overload (PO).

Methods and results: We used mice with cardiac-specific homozygous knockout of WW45 (WW45cKO), in which activation of Mst1 (Mammalian sterile 20-like 1) and Lats2 (large tumor suppressor kinase 2), the upstream kinases of the Hippo pathway, is effectively suppressed because of the absence of the scaffolding protein. We used male mice at 3 to 4 month of age in all animal experiments. We subjected WW45cKO mice to transverse aortic constriction for up to 12 weeks. WW45cKO mice exhibited higher levels of nuclear YAP in cardiomyocytes during PO. Unexpectedly, the progression of cardiac dysfunction induced by PO was exacerbated in WW45cKO mice, despite decreased apoptosis and activated cardiomyocyte cell cycle reentry. WW45cKO mice exhibited cardiomyocyte sarcomere disarray and upregulation of TEAD1 (transcriptional enhancer factor) target genes involved in cardiomyocyte dedifferentiation during PO. Genetic and pharmacological inactivation of the YAP-TEAD1 pathway reduced the PO-induced cardiac dysfunction in WW45cKO mice and attenuated cardiomyocyte dedifferentiation. Furthermore, the YAP-TEAD1 pathway upregulated OSM (oncostatin M) and OSM receptors, which played an essential role in mediating cardiomyocyte dedifferentiation. OSM also upregulated YAP and TEAD1 and promoted cardiomyocyte dedifferentiation, indicating the existence of a positive feedback mechanism consisting of YAP, TEAD1, and OSM.

Conclusions: Although activation of YAP promotes cardiomyocyte regeneration after cardiac injury, it induces cardiomyocyte dedifferentiation and heart failure in the long-term in the presence of PO through activation of the YAP-TEAD1-OSM positive feedback mechanism.

Keywords: apoptosis; cell cycle; cell proliferation; heart failure; mice.

Figure 1:. PO-induced YAP activation was transient in control mice but sustained in WW45cKO mice

(A, B) Representative gel pictures and quantitative analyses of immunoblot in hearts of Ctr and WW45cKO mice after operation, at the indicated time points (n=4, each). *P<0.05, **P<0.01 by ANOVA, compared with Ctr-TAC. #P<0.05 compared with Ctr-sham. (C) Representative immunostaining and quantitative analysis of YAP (YAP, red; DAPI, blue) in Ctr and WW45cKO 1 and 4 weeks after TAC. Nuclear accumulation of YAP is indicated by arrow heads (n=4, each). p-YAP indicates phospho-YAP.

Figure 2:. WW45cKO mice show severe cardiac dysfunction in response to PO, with higher level of markers of CM cycle re-entry.

(A) Representative gross morphology and longitudinal heart sections from Ctr and WW45cKO, stained with HE 4 weeks after TAC. Scale bars, 5.0 mm. (B) LV weight to TL ratio (n=7, each). (C) Representative WGA staining to assess CSA 4 weeks after operation (sham: n=6, each; TAC: n=8, each). (D) Representative immunostaining and quantitative analysis of pHH3 in the hearts of Ctr and WW45cKO mice 1 week after operation (pHH3, red; Sarc actinin, green; DAPI, blue) (n=6, each). (E) Representative echocardiographic tracings of the hearts of Ctr and WW45cKO mice 4 weeks after operation. (F) Kaplan-Meier survival curves after TAC. All results are expressed as mean ± SEM. *P<0.05, **P <0.01 by ANOVA. p-YAP indicates phosphorylated YAP.

Figure 3:. TEAD1 plays an essential role in mediating the exacerbation of heart failure in WW45 cKO mice in response to PO by facilitating CM de-differentiation.

(A) ECDF showing the difference in regulation between WW45cKO-TAC and Ctr-TAC for genes oppositely regulated in cardiac development and disease. Red line : genes DN in embryonic development and UP in TAC (n=438); Blue line : genes UP in embryonic development and DN in TAC (n=365); Black line : other genes. P-values are based on the K-S test comparing red or blue line genes with black line genes. (B) Heatmap showing relative expression of genes regarding cardiac differentiation. Gene set derived from the association of GO:0055007. The normalized read counts were subject to median centering before visualization. (C) TFBS analysis is based on UP or DN genes in WW45cKO vs Ctr (Arrow : TEAD1). Top five TFBS gene sets are shown with the q-values for false discovery rate control. (D) Representative gel pictures of immunoblot in the cytosolic and nuclear fractions of Ctr and WW45cKO mice 4 weeks after operation (n=4, each). (E) Representative gel pictures of immunoblot in the hearts of Ctr, WW45cKO, WW45cKO + TEAD1 +/− and TEAD1 +/− mice 4 weeks after operation. (F) Kaplan-Meier survival curves after TAC. Results are expressed as mean ± SEM. *P<0.05, **P<0.01 by ANOVA. NS indicates not significant.

Figure 4:. YAP-TEAD1 activity regulates OSM.

(A) Representative gel pictures of immunoblot in the hearts of Ctr and WW45cKO mice 4 weeks after operation (n=6, each). (B) Representative gel pictures of immunoblot in CMs isolated from adult WW45cKO mice 1 week after operation. (C) Relative mRNA expression in adult CMs isolated from the hearts 1 week after TAC (n=3, each). (D) Representative gel pictures of immunoblot in neonatal CMs transduced with adenovirus harboring LacZ or TEAD1. (E) ChIP assay of YAP binding to the Osm promoter (left) and the Osmr promoter (right) in neonatal CMs transduced with adenovirus harboring LacZ or TEAD1. PCR target regions and quantitative analyses are shown (n=4, each). (F) Representative pictures of ChIP-Seq in the Osmr promoter at the transcription start site (TSS) using YAP1, TFIIB, Pol II, or control IgG antibodies. Mice were subjected to TAC for 4 days (n=3). The fragment densities (y-axis) were aligned with the chromosomal coordinates (x-axis) using the Integrated Genome Browser (IGB). Shown are the binding sites of YAP1, TFIIB, and Pol II along the Osmr (left) and Actc1 (right) genes. The arrow defines the direction of the transcription start sites. (G) Reporter genes assays to evaluate the role of the TEAD1 binding site in the Osm promoter in neonatal CMs transduced with adenovirus harboring LacZ, TEAD1 or YAP (n=6, each). Results are expressed as mean ± SEM. *P<0.05, **P<0.01 by ANOVA. For box plots, whiskers show minima and maxima within 1.5 interquartile range. Ad-YAP indicates adenovirus harboring YAP; mutOSM, OSM promoter containing a mutated TEAD binding site; sh-TEAD1, short hairpin RNAs targeting TEAD; sh-Scr, scrambled short hairpin RNA; and sh-YAP, short hairpin RNAs targeting YAP.

Figure 5:. OSM induces CM de-differentiation through YAP-TEAD1.

(A) Representative gel pictures of immunoblot in CMs isolated from adult mice with or without OSM treatment. (B) Representative immunostaining and quantitative analysis of YAP in CMs isolated from adult mice with or without OSM treatment (n=3, each). (C) Reporter gene assays were conducted in order to evaluate the transcriptional activity of Tead1 in response to OSM in neonatal CMs transduced with Ad-Scr, Ad-sh-TEAD1 or Ad-sh-YAP, and then transfected with 8×GTIIC-luciferase plasmid in the presence or absence of OSM (n=6, each). (D) Representative double-immunostaining and quantitative analysis of YAP and ACTA2 in neonatal CMs with or without OSM treatment (YAP, green; ACTA2, red; DAPI, blue)(n=4, each). (E) Representative gel pictures of immunoblot in neonatal CMs transduced with Ad-Scr, Ad-sh-TEAD1 or Ad-sh-YAP, in the presence or absence of OSM (n=4, each). (F) Representative double-immunostaining and quantitative analysis of Sarcomeric actinin and ACTA2 in neonatal CMs transduced with adenovirus with or without OSM treatment (Sarc actinin, green; ACTA2, red; DAPI, blue) (n=4, each). Results are expressed as mean ± SEM. *P<0.05, **P<0.01 by ANOVA.

Figure 6:. Figure 6 anti-OSM blocking antibody improves cardiac dysfunction in WW45cKO during PO through suppression of CM de-differentiation

WW45cKO mice were subjected to TAC for 4 weeks in the presence of either Ctr-IgG or anti-OSM blocking antibody(anti-OSM). (B) %FS and LVEDD (n=4, each). (C) LVEDP (n=4, each). (D) Lung weight/TL (n=4, each). (E) Representative gel pictures in the hearts after TAC in the presence of either Ctr-IgG or anti-OSM. Results are expressed as mean ± SEM. *P<0.05, **P<0.01 by ANOVA.

Figure 7:. A schematic representation of the current hypothesis.

TAC suppresses the feedback loop of YAP-TEAD1-OSM through activation of the Hippo pathway and consequent suppression of YAP. In the absence of Hippo activation, TAC dramatically stimulates the feedback loop of YAP-TEAD1-OSM, thereby inducing de-differentiation of CMs and preventing the heart from maintaining cardiac output in the presence of PO.