Targeting intracellular cholesterol imbalance rescues sarcomere–ER contact site signaling and ER remodeling in dilated cardiomyopathy

Figures

• 

  Fig. 1. Sarcomere protein mutations cause impaired lipid signaling in induced pluripotent stem cell–derived cardiomyocytes (iPSC-CMs) from patients with dilated cardiomyopathy (DCM). a–g Quantitative lipidomics analyses in DCM (PAT1, MUT1, PAT2, MUT2) iPSC-CMs compared with WT (WT1, WT2) iPSC-CMs. a Heatmap of relative lipid amounts. b Volcano plot for highly abundant lipids. The top five dysregulated lipids in DCM iPSC-CMs vs WT iPSC-CMs are marked in red. c Cholesterol levels. ***P < 0.001 for PAT1 vs WT1, MUT1 vs WT1, PAT2 vs WT2, and MUT2 vs WT2 (one-way analysis of variance (ANOVA) with Šídák’s multiple-comparisons test). d Lysophosphatidylcholine levels. ***P < 0.001 for PAT1 vs WT1 and MUT2 vs WT2, **P < 0.01 for MUT1 vs WT1, *P < 0.05 for PAT2 vs WT2 (one-way analysis of variance (ANOVA) with Šídák’s multiple-comparisons test). e Sphingomyelin levels. ***P < 0.001 for PAT1 vs WT1 and MUT1 vs WT1, *P < 0.05 for PAT2 vs WT2 (one-way analysis of variance (ANOVA) with Šídák’s multiple-comparisons test). f Diacyl-phosphatidylethanolamine levels. ***P < 0.001 for PAT1 vs WT1, PAT2 vs WT2, MUT1 vs WT1, and MUT2 vs WT2 (one-way analysis of variance (ANOVA) with Šídák’s…

• 

  Fig. 2. Sarcomere protein mutations cause disrupted ER network architecture in iPSC–derived cardiomyocytes from patients with DCM. a–c Decreased ER-positive intracellular area in DCM (PAT1, MUT1, PAT2, MUT2) iPSC-CMs compared with WT (WT1, WT2) iPSC-CMs. a Representative images of cells immunostained with an anti-Sec61β antibody. Scale bar, 20 µm. Selected regions are magnified 8.3×. b, c Quantification of (a). ***P < 0.001 for PAT1 vs WT1, MUT1 vs WT1, PAT2 vs WT2; **P < 0.01 for MUT2 vs WT2 (Kruskal‒Wallis test, Dunn’s post-hoc comparisons). Per group, n = 2 experiments except for WT1 and PAT1, which have n = 3 experiments; n = 54 cells (WT1), 68 cells (PAT1), 54 cells (MUT1), 31 cells (WT2), 49 cells (PAT2), 37 cells (MUT2). d Impaired ER distribution in DCM (MUT1, MUT2) iPSC-CMs compared with WT (WT1, WT2) iPSC-CMs. Per group, n = 2 experiments; n = 20 cells. e–i Quantitative analyses of ER organization in DCM (PAT1, MUT1, PAT2, MUT2) iPSC-CMs compared with WT (WT1, WT2) iPSC-CMs using images shown in (a). e Reduced number of ER junctions in the presence of DCM mutations. ***P < 0.001 for PAT1 vs WT1, MUT1 vs WT1, PAT2 vs WT2, MUT2 vs WT2 (Kruskal‒Wallis test, Dunn’s post-hoc…

• 

  Fig. 3. The ER architecture of DCM (MUT) iPSC-CMs is dysfunctional due to loss of membrane curvature. a–e Visualization of the ER membrane curvature in TPM1-L185F (MUT1) iPSC-CMs compared with isogenic control iPSC-CMs by cryo-electron microscopy. a Vitrified healthy control (WT1) iPSC-CM attached to the electron microscopy grid. Scale bar, 20 µm. b Cross-sectional lamella preparation using cryo-focused ion beam scanning electron microscopy (FIB-SEM). Scale bar, 10 µm. c, d Cryo-electron tomography (cryo-ET) reconstruction resolves ER membrane curvature in DCM (MUT1) iPSC-CMs compared with WT (WT1) iPSC-CMs. Representative z-slices of the three-dimensional tomograms showing the ER of WT1 iPSC-CMs (c) and MUT1 iPSC-CMs (d). Scale bar, 0.2 µm. e, f Enlarged sections (zoomed in 2.7×) of the different z-slices through the reconstructed tomograms shown in c, d, respectively, and the corresponding segmentation for the ER. The numbers in the left bottom corners of the enlarged sections indicate the serial numbers of z-slices. Scale bars, 0.2 µm. Per group, n = 1 experiment; n = 68 tilt series (WT1), n = 17 tilt series (MUT1). g Average ER membrane curvature is reduced in DCM (MUT1) iPSC-CMs…

• 

  Fig. 4. DCM mutations in sarcomere proteins disrupt sarcomere/ER connections in DCM (MUT) iPSC-CMs. a, b DCM (PAT1, MUT1, PAT2) iPSC-CMs are disorganized compared with WT (WT1, WT2) iPSC-CMs and PAT-COR (PAT1-COR) iPSC-CMs. a Representative images of cells immunostained with anti-TPM1 and anti-ACTN2 antibodies. Scale bar, 20 µm. Selected regions are magnified 2.3×. b Quantification of (a). Unbiased assessment was generated by three individuals independently assessing de-identified data. Data are presented as the mean ± SD. Per group, n = 2 experiments, except for WT1, which has n = 3 experiments; n = 63 cells (WT1), n = 30 cells (WT2), n = 27 cells (PAT1), n = 44 cells (MUT1), n = 28 cells (PAT2). ***P < 0.001 for PAT1 vs WT1, MUT1 vs WT1, and PAT2 vs WT2. **P < 0.01 for PAT1-COR vs PAT1 (ordinary one-way ANOVA followed by Tukey’s multiple comparison test). c–m Co-immunoprecipitation of sarcomere proteins with CKAP4 using cell lysates from DCM (MUT1) iPSC-CMs vs WT (WT1). Beads decorated with CKAP4-specific antibody were used. c Representative membrane scans for input and eluted fractions for individual proteins are shown. As a negative control, pooled lysates from MUT1 iPSC-CMs and WT1…

• 

  Fig. 5. An increase in cholesterol levels phenocopies disturbed lipid signaling and disrupted ER architecture in WT control iPSC-CMs. a, b Increased plasma membrane (PM) cholesterol in healthy control iPSC-CMs treated with 2 µM water soluble cholesterol (WT1-chol) compared with control vehicle (WT1-cntrl). a Representative images of cells stained with His-GFP-D4. Scale bar, 20 µm. b Quantification of (a). ***P < 0.001 for WT1-chol vs WT1-cntrl (unpaired t-test). Per group, n = 2 experiments; n = 32 cells (WT1-cntrl), 33 cells (WT1-chol). c, d Decreased ER-positive area in healthy control iPSC-CMs treated with 10 µM water soluble cholesterol (WT1-chol) compared with control vehicle (WT1-cntrl). c Representative images of cells immunostained with an anti-KDEL antibody. Scale bar, 20 µm. Selected regions are magnified 10×. d Quantification of (c). ***P < 0.001 for WT1-chol vs WT1-cntrl (Mann‒Whitney test). Per group, n = 2 experiments; n = 32 cells (WT1-cntrl), 38 cells (WT1-chol). e, f Decreased ER-positive area in healthy control iPSC-CMs treated with 1 mM mevalonolactone (WT1-mev) compared with control vehicle (WT1-cntrl). e Representative images of cells immunostained with an anti-KDEL…

• 

  Fig. 6. Pathological ER/cholesterol signaling is reversed by genetic or chemical rescue in DCM (MUT) iPSC-CMs. a, b Decreased plasma membrane (PM) cholesterol in patient-derived iPSC-CMs with subsequent CRISPR/Cas9-mediated correction of the mutation TPM1-L185F (PAT1-COR) relative to DCM patient-specific (PAT1) iPSC-CMs and wild-type control iPSC-CMs (WT1). a Representative images of cells stained with His-GFP-D4. Scale bar, 20 µm. b Quantification of (a). ***P < 0.001 for PAT1-COR vs PAT1 and for PAT1-COR vs WT1; **P < 0.01 for PAT1 vs WT1 (Kruskal‒Wallis test, Dunn’s post-hoc comparisons). Per group, n = 2 experiments, except for WT1 iPSC-CMs, which had n = 3 experiments; n = 125 cells (WT1), n = 67 cells (PAT1), 74 cells (PAT1-COR). Data for PAT1 and WT1 iPSC-CMs are the same as presented in Fig. 1i. c, d Increased ER-positive area in DCM patient-specific iPSC-CMs with corrected mutation TPM1-L185F (PAT1-COR) compared with DCM patient-specific (PAT1) iPSC-CMs. c, Representative images of cells immunostained with an anti-Sec61β antibody. Scale bar, 20 µm. Selected regions are magnified 15×. d Quantification of (c). ***P < 0.001 for PAT1-COR vs PAT1 and for PAT1 vs WT1; ns not…

• 

  Fig. 7. Myocardial tissue from patients with DCM and end-stage heart failure recapitulate lipid imbalance and defective ER organization. Left ventricular (LV) heart tissues from DCM patients with end-stage heart failure (DCM-HF) were analyzed compared to LV heart tissues from donor hearts (cntrl) without heart disease that failed transplantation.40 a Lipidomics results are presented as a volcano plot. The top five dysregulated lipids in DCM-HF vs control are marked in red. b Cholesterol levels. **P < 0.01 for DCM-HF vs cntrl (Mann‒Whitney test). c Sphingomyelin levels. ***P < 0.001 for DCM-HF vs cntrl (Mann‒Whitney test). d Diacyl-phosphatidylserine levels. *P < 0.05 for DCM-HF vs cntrl (Mann‒Whitney test). e Phosphatidylethanolamine with plasmalogen 18:1 moiety levels. ***P < 0.001 for DCM-HF vs cntrl (Mann‒Whitney test). f Diacyl-phosphatidylethanolamine levels. **P < 0.01 for DCM-HF vs cntrl (Mann‒Whitney test). Samples from 4 patients (DCM-HF2, DCM-HF3, DCM-HF7, DCM-HF11) and 4 control donors (cntrl) were analyzed. Per group, n = 4 technical replicates. g–i Analysis of ER organization in adult human heart tissue was performed by immunostaining and confocal imaging. LV tissues from…

• 

  Fig. 8. Analysis of structural and curvature-defining ER proteins in DCM heart tissue and analysis of the ER following cholesterol-overload in adult cardiomyocytes. a–i High-content immunoblot analysis of ER structural and curvature-defining proteins in left ventricular (LV) heart tissues from DCM patients with end-stage heart failure (DCM-HF) compared to LV heart tissues from donor hearts (cntrl) without heart disease that failed transplantation.40 Patient DCM-HF8 received lipid-lowering therapy (simvastatin). a Heatmap of relative protein levels. Data are normalized to GAPDH and to respective protein values in the cntrl1 group. b REEP5 protein levels. ***P < 0.001 for DCM-HF vs cntrl (Mann‒Whitney test). c RTN4 protein levels. ns, not significant for DCM-HF vs cntrl (unpaired t-test). d ATL3 protein levels. ns, not significant for DCM-HF vs cntrl (Mann‒Whitney test). e LNP protein levels. *P < 0.05 for DCM-HF vs cntrl (unpaired t-test). f ZFYVE27 protein levels. *P < 0.05 for DCM-HF vs cntrl (Mann‒Whitney test). g KTN1 protein levels. ns not significant for DCM-HF vs cntrl (Mann‒Whitney test). h RRBP1 protein levels. ns, not significant for DCM-HF vs cntrl (Mann‒Whitney test). i CKAP4…