Activation of Hippo signaling pathway mediates mitochondria dysfunction and dilated cardiomyopathy in mice

Rationale: Mitochondrial dysfunction facilitates heart failure development forming a therapeutic target, but the mechanism involved remains unclear. We studied whether the Hippo signaling pathway mediates mitochondrial abnormalities that results in onset of dilated cardiomyopathy (DCM). Methods: Mice with DCM due to overexpression of Hippo pathway kinase Mst1 were studied. DCM phenotype was evident in adult animals but contractile dysfunction was identified as an early sign of DCM at 3 weeks postnatal. Electron microscopy, multi-omics and biochemical assays were employed. Results: In 3-week and adult DCM mouse hearts, cardiomyocyte mitochondria exhibited overt structural abnormalities, smaller size and greater number. RNA sequencing revealed comprehensive suppression of nuclear-DNA (nDNA) encoded gene-sets involved in mitochondria turnover and all aspects of metabolism. Changes in cardiotranscriptome were confirmed by lower protein levels of multiple mitochondrial proteins in DCM heart of both ages. Mitochondrial DNA-encoded genes were also downregulated; due apparently to repression of nDNA-encoded transcriptional factors. Lipidomics identified remodeling in cardiolipin acyl-chains, increased acylcarnitine content but lower coenzyme Q10 level. Mitochondrial dysfunction was featured by lower ATP content and elevated levels of lactate, branched-chain amino acids and reactive oxidative species. Mechanistically, inhibitory YAP-phosphorylation was enhanced, which was associated with attenuated binding of transcription factor TEAD1. Numerous suppressed mitochondrial genes were identified as YAP-targets. Conclusion: Hippo signaling activation mediates mitochondrial damage by repressing mitochondrial genes, which causally promotes the development of DCM. The Hippo pathway therefore represents a therapeutic target against mitochondrial dysfunction in cardiomyopathy.


SUPPLEMENTARY MATERIALS
Expanded Methods Table S1. Source and characteristics of antibodies used for immunoblotting Table S2. Oligonucleotide sequences of mouse primers for gene expression by RT-PCR Table S3. Full list of abbreviations use in figures

Transmission electron microscopy (EM) and quantification
Mitochondrial ultrastructure was studied in freshly collected LV specimens by EM (3 LV samples per age per genotype). Samples were dissected into 1 mm 3 blocks, immediately fixed with 2.5% glutaraldehyde buffer (2 h at 37 o C), and then post-fixed in 1% OsO4 (2 h at 4 o C). The samples were dehydrated in a graded series of alcohol and then exposed to propylene oxide for infiltration of the embedding medium, Epon 812 resin in 0.1 mol/L sodium phosphate buffer. Using LKB-V/Leica UC7 ultramicrotome (LKB-V/NOVA, Sweden and Leica, Germany), the embedded blocks were cut into sections (50-70 nm). Sections were stained with acidified uranyl acetate, followed by a modification of Sato's triple lead stain, and viewed with a transmission electron microscope (TEM; H-7650; Hitachi, Tokyo, Japan). For each sample, images from randomly chosen fields were obtained at magnifications of ×4,000, ×10,000 and ×30,000, respectively (6-10 images at each magnification per heart sample). The images (10,000 magnification) were analysed, in a blinded fashion, using ImageJ software on at least 8 images per heart. Area of individual mitochondria was manually measured and average used. The sum of all measurements of mitochondria size per image represented the total area by mitochondria per image field. Mitochondria with complete or incomplete outer membrane were individually observed and counted to obtain percentage of mitochondria with incomplete membrane. Sarcomere length was determined in 8-10 images (10,000 magnification) by drawing a line parallel to myofibril orientation across over 5-10 sarcomeres. Approximately 30 sarcomeres were measured from one image and the averages from all images calculated for comparison.

Mitochondrial lipid and amino acid profiling
Tissue homogenization and lipid extraction. Tissue samples were homogenized using a sonicator probe in ice cold PBS. Lipids were extracted from homogenized tissue as previously described. 1 In brief, tissue or plasma was mixed with 10 volume of butanol:methanol (1:1) with 10 mM ammonium formate containing a mixture of internal standards. Samples were vortexed thoroughly and set in a sonicator bath for 1 hour at room temperature. Samples were then centrifuged (14,000g, 10 min, 20 o C) and the supernatant was transferred into sample vials with glass inserts for mass spectrometry analysis.
Liquid chromatography mass spectrometry (LC/MS/MS). The detailed lipidomics and assay conditions have been reported previously. 2 Briefly, lipidomic analysis was performed by LC ESI-MS/MS using an Agilent 1290 liquid chromatography system and an Agilent 6490 QQQ mass spectrometer. We employed source conditions identical to those previously described. 2 Liquid chromatography was performed on a Zorbax C18, 1.8 m, 50× 2.1 mm column (Agilent Technologies). Solvents A and B consisted of water, acetonitrile and isopropanol in the ratio (50:30:20) and (90:9:1) respectively, both containing 10 mM ammonium formate with solvents B also containing 5 M medronic acid. Gradient conditions were as previously described [1,2]. Columns were heated to 45 °C and the auto-sampler regulated to 25 °C.
Analysis of mitochondria-rich lipid species. Quantification of lipid species was determined by comparison to the relevant internal standards. In the present study that was focused on mitochondria, only 3 lipids that are largely or entirely localized in mitochondria were analysed, i.e. acyl-carnitine (AC), cardiolipin (CL) and ubiquinone (or coenzyme Q10, CoQ10). Lipid characterization and quantification were conducted as we previously described in detail [1,2]. Alterations in the species of the three lipids were expressed as relative change to that of nTG control.
Amino acid analysis of heart and plasma. The extracted myocardial samples were further analysed for specific amino acids using a separate targeted HILIC-MS/MS method. An Acquity UPLC BEH Amide 1.7 µm 2.1  100 mm column (Waters) was used. In this assay Solvent A comprises of 50% acetonitrile in water, whereas Solvent B comprises of 95% acetonitrile in water, both with 10 mM ammonium formate. Chromatography was used to separate out analytes prior to mass spectrometetry analysis. Starting at 100% B, this was held for 1 minute before a linear gradient to 0% B over 5 minutes. 1 minute was then spent holding 0% B before switching to 100% B over 0.1 minute, then held at 100% B for an additional 2.9 minutes for equilibration. Total run time was 10 minutes per sample. Results were normalised to a spike internal standard (L-Leucine-5,5,5-d3, Sigma Aldrich).
High performance liquid chromatography (HPLC) assay of ubiquinone: Tissue content of ubiquinone in LVs of 3-wk-old mice was determined using HPLC. LV tissues (20 mg) were homogenized in 2 ml extract (N-hexane: ethanol = 8:1), centrifuged (3,000 rpm for 5 min), and supernatant was harvested. This process was repeated 3 times and the final supernatant was evaporated to dryness in a vacuum centrifuged concentrator (1500 rpm, 37℃, 50 min). Sample was redissolved in pure ethanol (100 L) and assayed using SHIMADZU HPLC system (LC-2030C 3D, mobile phase: 3:7 methanol:ethanol 1 ml/min). Ubiquinone was identified by an internal standard and the amount quantified using ubidecarenone (Sigma, Lot# LRAC3727) derived standard curve.

Quantitative real-time PCR
RNA was extracted from LV tissue using RNAiso PLUS (Takara Bio, Japan, Code No. 9108/9109). RNA (1 μg) was used for cDNA synthesis. Quantitative PCR was performed by SYBR-based RT-PCR assays using PrimeScript TM reagent kit with gDNA Eraser (Perfect Real-Time) (Takara, Code No: RR047A), and TB Gree TM Premix Ex Tag TM II Tli RNAaseH Plus (Takara, Code No. RR820A). Table S1 lists target genes examined and sequences of primers (Sangon Biotech, Shanghai, China). Quantitative real-time PCR reaction was performed using Bio-Rad CFX96 Real-time PCR Detection System with the following two step procedures: Step-1: 95 o C for 30 sec, and Step-2 (PCR reaction): 40 cycles of amplification at 95 o C for 5 sec and 60 o C for 30 sec. The expression of target genes was normalized to that of GAPDH using the method of 2 -ΔΔct , and presented relative to that of the nTG value.
YAP silencing by small interfering RNA in H9c2 cells. YAP gene knockdown was performed in rat cardiomyoblasts (H9c2) using small interfering RNA (siRNA, GenePharma, Shanghai), according to manufacturer's instructions. Cells were cultured in 6-well plates with standard antibacterial-free medium. To make siRNA delivery medium, 2.5 L lipofectamine 2000 (Invitrogen 11668019) was added in 250 L Opti-MEM (Gibco, 11058021) and mixed for 5 min, followed by addition of specific rat YAP-siRNA (F: 5-GGA GAA GUU UAC UAC AUA ATT-3; R: 5-UUA UGU AGU AAA CUU CUC CTT-3) or non-targeting control siRNA (10 nM each) with 20 min allowed for full mixing. After reaching 40-50% confluence, cells were incubated with the siRNA delivery medium (0.5 mL per well containing 1.5 mL standard medium). Six hours afterward, the delivery medium was replaced. After incubation for another 48 h, cells were harvested and protein was extracted for immunoblotting assay.    Figure S1.
Dilated cardiomyopathy phenotype of 3-wk Mst1-TG mice. Images of lungs and the liver from nTG and Mst1-TG mice at 3 weeks (A) in comparison to that of 6-mo-old counterparts (B). C, body weight-normalized organ weights of 3-wk-old male and female mice. Pulmonary and hepatic congestion in TG mice indicated presence whole heart failure. *P < 0.05, †P<0.01, #P < 0.001 vs. nTG group.
RNA sequencing data were collected from LV tissues of mice aged at 3-wk (n = 4/genotype) and 15-wk (n = 6 for nTG and n = 7 for TG). We generated 25 M (3-wk) or 38.6 M (15-wk) reads/sample, of which 83.3% of reads were uniquely mapped (SD 2.1%). An average of 24.0 M reads was assigned to genes (SD 3.2 M).
A, Venn diagrams showing number of genes detected in 3-wk and 15-wk TG hearts (left), and number of differentially expressed genes (DEGs, FDR<0.05) due to Mst1-overexpression in both age groups (right). B, Volcano plots of DEGs of 3-wk and 15-wk TG mice relative to respective nTGs. Red dots denote genes with FDR < 0.05). For clarity reason, only DEGs are presented. C, Venn diagram of number of genes showing significant changes due to Mst1 overexpression in both age groups. D, Filled contour plot of DEGs in TG hearts showing a positive correlation between 3-wk and 15-wk TG groups.
Downregulation of gene sets of mitochondrial biosynthesis, assembly and turnover in TG hearts.
Filled contour plots showing that 3-wk and 15-wk TG hearts exhibited similar changes in the down-regulation of gene sets related to mitochondria biosynthesis, assembly, turnover, cristae formation, protein import and metabolism of cofactors. Figure S4.
Downregulation of gene-sets of mitochondrial metabolism. Filled contour plots showing that 3-wk and 15wk TG mouse hearts exhibited consistent down-regulation of gene-sets related to mitochondria metabolism. Figure S5.
Downregulation of mitochondrial protein import genes in Mst1-TG hearts by RNAsequencing.
Heatmap for mitochondrial protein import genes (set size n = 57) in 3-wk and 15-wk Mst1-TG hearts. Among the gene-sets were numerous members of translocases of outer-membrane (TOMs) or translocases of innermembrane (TIM) that mediate inward transportation of cytosolic proteins into mitochondria. Group size: n = 4/group for 3-wk and n = 6-7/group for 15-wk group.

Figure S8.
Changes in YAP-target genes in TG heart by transcriptome.
A, Volcano plots showing up-and down-regulated YAP-target genes in 3-wk and 15-wk TG relative to respective nTG hearts. Red dots denote genes with FDR < 0.05. B, Venn diagram of downregulated genes in both TG groups. Bar graph depicting the number of downregulated mitochondrial genes as percentages of total downregulated YAP-target genes.