YAP mediates compensatory cardiac hypertrophy through aerobic glycolysis in response to pressure overload

Abstract

The heart utilizes multiple adaptive mechanisms to maintain pump function. Compensatory cardiac hypertrophy reduces wall stress and oxygen consumption, thereby protecting the heart against acute blood pressure elevation. The nuclear effector of the Hippo pathway, Yes-associated protein 1 (YAP), is activated and mediates compensatory cardiac hypertrophy in response to acute pressure overload (PO). In this study, YAP promoted glycolysis by upregulating glucose transporter 1 (GLUT1), which in turn caused accumulation of intermediates and metabolites of the glycolytic, auxiliary, and anaplerotic pathways during acute PO. Cardiac hypertrophy was inhibited and heart failure was exacerbated in mice with YAP haploinsufficiency in the presence of acute PO. However, normalization of GLUT1 rescued the detrimental phenotype. PO induced the accumulation of glycolytic metabolites, including l-serine, l-aspartate, and malate, in a YAP-dependent manner, thereby promoting cardiac hypertrophy. YAP upregulated the GLUT1 gene through interaction with TEA domain family member 1 (TEAD1) and HIF-1α in cardiomyocytes. Thus, YAP induces compensatory cardiac hypertrophy through activation of the Warburg effect.

Keywords: Cardiology; Cardiovascular disease.

Figure 1. YAP promotes CM glycolysis in response to PO.

(A–D) Time-course analysis of glycolytic flux after PO in freshly isolated AMVMs from control mice. Glycolysis was evaluated by measuring the ECAR. Overall ECAR in Seahorse experiments (A), relative basal glycolysis (B), relative glycolytic capacity (C), and relative glycolytic reserve capacity (D). n = 11 –20 wells from 3 mice at each time point. *P < 0.05 versus sham, by 1-way ANOVA with Dunnett’s test (B–D). (E and F) Haploinsufficiency of YAP in AMVMs attenuated PO-induced glycolysis. Overall Seahorse experiment (ECAR) (E) and summary of glycolytic function (F). n = 16 to 18 wells from 3 mice. *P < 0.05, by 1-way ANOVA with Tukey’s test (F). Data represent the mean ± SEM. Glu, glucose; Oli, oligomycin.

Figure 2. Metabolomics analysis of glucose metabolism in response to PO.

(A and B) Summary of the metabolomics analysis of glucose metabolism after sham operation or 2 days of TAC in control and YAPch-KO mice. Black and gray indicate detectable and undetectable, respectively (A). Summary of intermediates of glucose metabolism, shown as box plots (B). n = 4–5 mice. *P < 0.05, by 1-way ANOVA with Tukey’s test. Data represent the mean ± SEM. See also Supplemental Table 2.

Figure 3. YAP facilitates glycolysis in CMs.

(A–D) Seahorse glycolytic flux analysis was performed in NRVMs transduced with Ad-LacZ or Ad-FLAG-YAP for 6 days in serum-free DMEM/F12 medium. Overall ECAR in Seahorse experiments (A), relative basal glycolysis (B), relative glycolytic capacity (C), and relative glycolytic reserve capacity (D). n = 12 wells from 3 independent experiments. *P < 0.05 versus LacZ, by 1-way ANOVA with Dunnett’s test (B–D). (E and F) Confirmation of YAP expression (E) and a summary of glycolytic function in NRVMs transduced with 1 MOI Ad-FLAG-YAP (F). An α-tubulin blot, serving as a loading control, was run in parallel and contemporaneously with the other blot (E). n = 25 wells from 3 independent experiments. *P < 0.05 by 2-tailed, unpaired Student’s t test (F). Data represent the mean ± SEM.

Figure 4. YAP increases glucose oxidation in NRVMs.

(A–D) A Seahorse Mito Stress Test was performed in NRVMs transduced with Ad-LacZ or Ad-FLAG-YAP in 10 mM glucose medium in the absence (A and B) or presence (C and D) of 1 mM pyruvate. Overall Seahorse experiment (OCR) (A and C) and summary of mitochondrial function (B and D). n = 14 (A and B) and n = 31 (C and D) wells from 3 independent experiments. *P < 0.05, by 2-tailed, unpaired Student’s t test. Data represent the mean ± SEM.

Figure 5. Glucose metabolism is regulated by YAP in CMs.

(A and B) Metabolomics analysis of glucose metabolism was performed in NRVMs transduced with Ad-LacZ or Ad-FLAG-YAP for 6 days in serum-free DMEM/F12 medium. (A) Heatmap represents the expression profile of intermediates of glucose metabolism. (B and C) Summary of the metabolomics analysis of glucose metabolism. Black and gray represent detectable and undetectable, respectively (B). Box plots show a summary of intermediates of glucose metabolism (C). n = 6 dishes from 3 independent experiments. See also Supplemental Table 4. (D and E) YAP increased Tyr105 phosphorylated Pkm2 in NRVMs transduced with Ad-LacZ or Ad-FLAG-YAP. Representative immunoblots (D) and a summary of quantification (E) are shown. α-Tubulin and PKM2 blots, serving as loading controls, were run in parallel and contemporaneously with other blots (D). n = 5 dishes from 5 independent experiments. *P < 0.05, by 2-tailed, unpaired Student’s t test (C and E). Data represent the mean ± SEM.

Figure 6. Overexpressed YAP increases PEP and malate through glycolysis.

NRVMs were transduced with Ad-LacZ or Ad-FLAG-YAP for 5 days in serum-free DMEM/F12 medium, and then the medium was replaced with DMEM containing U-¹³C-glucose. (A–C) Schematic of metabolism of U-¹³C-glucose (A), mass isotopomer distributions of 2-PG/3-PG, PEP, citrate, and malate (B), and relative levels of 2-PG/3-PG, PEP, citrate, and malate (C). n = 4 dishes from 2 independent experiments. *P < 0.05, by 1-way ANOVA with Tukey’s test. Data represent the mean ± SEM.

Figure 7. Increased glycolytic flux is essential for YAP-mediated cardiac hypertrophy.

Cardiac hypertrophy was evaluated with WGA staining. (A) Representative images of NRVMs transduced with Ad-LacZ or Ad-FLAG-YAP in glucose-free or normal DMEM medium for 5 days are shown. Scale bars: 50 μm. (B) Effect of knockdown of glucose metabolic genes on YAP-mediated cardiac hypertrophy. n = 63–113 cells from 3 independent experiments. *P < 0.05, by 1-way ANOVA with Tukey’s test. Data represent the mean ± SEM. See also Supplemental Figure 10, which shows knockdown of glucose metabolic genes confirmed by immunoblotting. siCont, siControl. (C) Glycolytic pathways, indicated by red boxes, are required for YAP-mediated cardiac hypertrophy.

Figure 8. YAP induces cardiac hypertrophy through GLUT1-mediated glycolysis in adult CMs.

(A–I) Overexpressed YAP induced hypertrophy in AMVMs, accompanied by the upregulation of GLUT1 expression and glycolysis. AMVMs were transduced with 0.1 MOI Ad-LacZ or Ad-FLAG-YAP for 2 or 4 days. Representative immunoblots (A) and quantification results (B) after 4 days of culturing. An α-tubulin blot, serving as a loading control, was run in parallel and contemporaneously with the other blots (A). n = 5 dishes from 3 mice. Overall ECAR (C) and summary of basal glycolysis (D) after 2 days of culturing. n = 10 wells from 3 mice. Representative images of WGA staining (E) and summary of relative CM sizes (F) after 4 days of culturing. Scale bars: 100 μm. n = 101 cells from 3 mice. *P < 0.05, by 2-tailed, unpaired Student’s t test (B, D, and F). (G–I) YAP-induced glycolysis and hypertrophy were inhibited by 20 nM BAY-876, a selective GLUT1 inhibitor, in cultured AMVMs. Glucose consumption (G) and lactate production (H) in media after 48 hours of culturing were measured in AMVMs transduced with Ad-LacZ or Ad-FLAG-YAP in the presence or absence of BAY-876. n = 6 wells from 6 mice. (I) CM size was evaluated with WGA staining after 4 days of culturing. n = 120 cells from 6 mice. *P < 0.05 versus LacZ plus vehicle and ^#P < 0.05 versus YAP plus vehicle, by 1-way ANOVA with Tukey’s test (G–I). Data represent the mean ± SEM (B–D, G, and H). Results in F and I are shown as box plots.

Figure 9. Overexpression of GLUT1 ameliorates cardiac dysfunction in YAPch-KO mice during acute PO.

(A–D) AAV-GLUT1 injection increased GLUT1 expression in mouse hearts. LVs from control and YAPch-KO mice 2 weeks after injection of AAV-control or AAV-GLUT1 (n = 4 mice) (A and B) and sham operation or 1 week of TAC after injection of AAV-control or AAV-GLUT1 (n = 4–5 mice) (C and D) were homogenized and subjected to immunoblotting. α-Tubulin blots, serving as loading controls, were run in parallel and contemporaneously with other blots (A and C). (E–G) Cardiac function was evaluated by echocardiography 1 week after TAC. Representative M-mode echocardiographic images (E), FS (F), and LVIDd (G). Vertical scale bars: 3.5 mm; horizontal scale bars: 200 ms. n = 6–9 mice. (H) LungW/TL. n = 5–8 mice. *P < 0.05, by 2-tailed, unpaired Student’s t test (B) or 1-way ANOVA with Tukey’s test (D and F–H). Data represent the mean ± SEM.

Figure 10. Overexpression of GLUT1 rescues compensatory cardiac hypertrophy in YAPch-KO mice during acute PO.

AAV-GLUT1 injection rescued cardiac hypertrophy in YAPch-KO mice 1 week after TAC. (A) LVW/TL. n = 5–8 mice. (B and C) Representative images of WGA staining of LVs (B) and quantification of CM CSA (C). Scale bars: 50 μm. n = 5–8 mice. (D and E) AAV-GLUT1 injection did not alter the cell proliferation index of CMs or non-CMs in LVs. Heart sections were stained with anti–Ki-67 and anti–cardiac troponin T antibodies and DAPI. Percentage of Ki-67 labeling in CMs (n = 5–8 mice) (D) and non-CMs (n = 5–8 mice) (E). *P < 0.05 by 1-way ANOVA with Tukey’s test. Data represent the mean ± SEM.

Figure 11. TEAD1 and HIF-1α are required for YAP-induced GLUT1 expression in CMs.

(A–I) The effect of knockdown of TEAD1 (A–C), HIF-1α (D–F), or c-Myc (G–I) on YAP-induced GLUT1 expression in NRVMs. n = 6 dishes from 3 independent experiments. *P < 0.05, by 1-way ANOVA with Tukey’s test (A–I). (K–M) Representative immunoblots (J) and expression levels of TEAD1 (K), HIF-1α (L), and c-Myc (M) in NRVMs. n = 6 dishes from 3 independent experiments. *P < 0.05, by 2-tailed, unpaired Student’s t test (J–M). (N–Q) NRVMs were transduced with Ad-LacZ, Ad-FLAG-YAP-WT, or Ad-FLAG-YAP-S94A. Representative immunoblots (N) and expression levels of GLUT1 (O) are shown. n = 4 dishes from 4 independent experiments. Overall ECAR (P) and summary (Q). n = 36–40 wells from 3 independent experiments. *P < 0.05 versus LacZ and ^#P < 0.05 versus YAP-WT, by 1-way ANOVA with Tukey’s test (O and Q). α-Tubulin blots, serving as loading controls, were run in parallel and contemporaneously with other blots (A, B, E, G, H, J, and N). Data represent the mean ± SEM.

Figure 12. YAP, TEAD1, and HIF-1α bind to the Glut1 promoter in the heart.

(A) Schematic representation of HREs and the predicted TRE in the mouse Glut1 promoter. The predicted TRE differs in 2 base pairs (green) from the consensus TRE. Point mutations are indicated in red. (B and C) Effects of siRNA against Tead1 or Hif1a (B) and Glut1 promoter mutations (C) on YAP-induced activation of the Glut1 promoter. n = 6–10 wells from 3–4 independent experiments. *P < 0.05 versus each mock; ^#P < 0.05 versus each control YAP; and †P < 0.05 versus control mock, by 1-way ANOVA with Tukey’s test. (D and E) ChIP assays of the Glut1 promoter were performed in NRVMs using the indicated antibodies. n = 4 dishes from 2 independent experiments. *P < 0.05, by 1-way ANOVA with Tukey’s test. (F–H) ChIP assays of the Glut1 promoter were performed using pooled hearts from WT mice after 2 days of sham operation or TAC, with antibodies against YAP (F), TEAD1 (G), and HIF-1α (H). (I–L) ChIP assays of the Glut1 promoter were performed using hearts from control or YAPch-KO mice after 2 days of sham operation or TAC, with antibodies against HIF-1α (I and J) and TEAD1 (K and L). n = 4–6 mice. *P < 0.05, by 1-way ANOVA with Tukey’s test. (M) ChIP-Seq was performed using pooled hearts from WT mice subjected to sham operation or 4 days of TAC with anti-YAP antibody. Schematic representation of the Glut1 promoter is aligned with the results of ChIP-Seq in the Glut1 promoter. Triangles indicate new peaks after PO. ATG, start codon. Data represent the mean ± SEM.

Figure 13. YAP physically interacts with both TEAD1 and HIF-1α.

(A–C and G) HEK293 cells were transfected with plasmids encoding FLAG-YAP, Myc-TEAD1, and HA–HIF-1α P402A/P564A, a stable HIF-1α mutant. (D–F) NRVMs were transduced with Ad-FLAG-YAP, Ad-Myc-TEAD1, and Ad-HA–HIF-1α P402A/P564A. The cells were treated with 3 μM MG-132 twenty-four hours prior to being harvested. Immunoblotting (IB) and immunoprecipitation (IP) were performed using the indicated antibodies. Data are representative of 3 independent blots. The lanes were run on the same gel but were noncontiguous (A–G). (H–K) YAP directly bound to both TEAD1 and HIF-1α, while TEAD1 and HIF-1α only interacted with one another through YAP. The indicated GST-fused proteins and recombinant proteins were incubated, followed by pulldown with glutathione-sepharose beads. Data are representative of 3 independent blots. (L) Schematic illustration of the role of YAP in regulating CM glycolysis in the heart during acute PO.