BSCL2/Seipin deficiency in hearts causes cardiac energy deficit and dysfunction via inducing excessive lipid catabolism

Abstract

Background: Heart failure (HF) is one of the leading causes of death worldwide and is associated with cardiac metabolic perturbations. Human Type 2 Berardinelli-Seip Congenital Lipodystrophy (BSCL2) disease is caused by mutations in the BSCL2 gene. Global lipodystrophic Bscl2^-/- mice exhibit hypertrophic cardiomyopathy with reduced cardiac steatosis. Whether BSCL2 plays a direct role in regulating cardiac substrate metabolism and/or contractile function remains unknown.

Methods: We generated mice with cardiomyocyte-specific deletion of Bscl2 (Bscl2^cKO ) and studied their cardiac substrate utilisation, bioenergetics, lipidomics and contractile function under baseline or after either a treatment regimen using fatty acid oxidation (FAO) inhibitor trimetazidine (TMZ) or a prevention regimen with high-fat diet (HFD) feeding. Mice with partial ATGL deletion and cardiac-specific deletion of Bscl2 were also generated followed by cardiac phenotyping.

Results: Different from hypertrophic cardiomyopathy in Bscl2^-/- mice, mice with cardiac-specific deletion of Bscl2 developed systolic dysfunction with dilation. Myocardial BSCL2 deletion led to elevated ATGL expression and FAO along with reduced cardiac lipid contents. Cardiac dysfunction in Bscl2^cKO mice was independent of mitochondrial dysfunction and oxidative stress, but associated with decreased metabolic reserve and ATP levels. Importantly, cardiac dysfunction in Bscl2^cKO mice could be partially reversed by FAO inhibitor TMZ, or prevented by genetic abolishment of one ATGL allele or HFD feeding. Lipidomic analysis further identified markedly reduced glycerolipids, glycerophospholipids, NEFA and acylcarnitines in Bscl2^cKO hearts, which were partially normalised by TMZ or HFD.

Conclusions: We identified a new form of cardiac dysfunction with excessive lipid utilisation which ultimately causes cardiac substrate depletion and bioenergetics failure. Our findings also uncover a crucial role of BSCL2 in controlling cardiac lipid catabolism and contractile function and provide novel insights into metabolically treating energy-starved HF using FAO inhibitor or HFD.

Keywords: BSCL2/Seipin; heart failure; lipid metabolism; lipidomics.

FIGURE 1

Mice with cardiac‐specific deletion of BSCL2 develop systolic heart dysfunction with dilation. (A) Ventricle weight (VW) normalised to body weight (BW) in 3‐month‐old (3 M), 5‐month‐old (5 M) and 6‐month‐old (6 M) male Cre‐;Bscl2^f/f (Ctrl), Cre+;Bscl2^w/w , and Cre+;Bscl2^f/f (cKO) mice. 3 M old: Ctrl, n = 6; Cre+;Bscl2^w/w , n = 5; cKO, n = 6. 5 M old: Ctrl, n = 6; Cre+;Bscl2^w/w , n = 6; cKO, n = 6. 6 M old: Ctrl, n = 6; Cre+;Bscl2^w/w , n = 6; cKO, n = 9. (B and C) left ventricle post wall thickness at end systole (LVPWs, mm) and at end diastole (LVPWd, mm), respectively; (D) left ventricle anterior wall thickness at end systole (LVAWs, mm); (E and F) left ventricle internal diameter at end systole (LVIDs, mm) or end diastole (LVIDd, mm); (G) ejection fraction (%) and (H) fractional shortening (%) in 3 M, 5 M and 6 M old male mice. 3 M old: Ctrl, n = 8; Cre+;Bscl2^w/w , n = 8; cKO, n = 12. 5 M old: Ctrl, n = 6; Cre+;Bscl2^w/w , n = 8; cKO, n = 10. 6 M old: Ctrl, n = 11; Cre+;Bscl2^w/w , n = 8; cKO, n = 9. (I) Representative echocardiography, and (J) qRT‐PCR analysis of cardiac stress genes in ventricles of 6‐month‐old male mice. n = 6 per group. *, p < .05; **, p < .005. Two‐way ANOVA followed by Tukey's post‐hoc tests

FIGURE 2

Cardiac‐specific deletion of BSCL2 induces cardiac triglyceride turnover and excessive fatty acid oxidation. (A) Ventricular TG content as normalised to wet tissue weight. n = 9 per group. (B) Representative Western blotting and quantification of lipolytic proteins in heart homogenates. n = 9 per group. (C and D) Isolated adult cardiomyocytes, representative Western blotting in duplicates and quantification from three independent experiments. (E) Acid soluble metabolites (ASM) and CO₂ production after incubating heart crude mitochondrial fraction with C‐palmitate. Ctrl, n = 8; cKO, n = 9. (F) Oleate oxidation rate; (G) glucose oxidation rate in ex vivo perfused working hearts. n = 6 per group. (H) Glycogen content. Ctrl, n = 10; cKO, n = 9. (I) qRT‐PCR analysis of genes involved in fatty acid metabolism, mitochondrial biogenesis and glucose metabolism. Ctrl, n = 8; cKO, n = 10. All experiments used 3‐month‐old male Cre−;Bscl2^f/f (Ctrl) and Bscl2^cKO (cKO) mice. *, p < .05; **, p < .005, unpaired t‐tests (parametric)

FIGURE 3

Bscl2^cKO mice develop massive cardiac lipid remodelling and exhibit reduced metabolic reserve. (A) The total lipid ion abundance normalised to tissue weight. (B) Heatmap of lipid species including glycerophospholipids [i.e., phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylglycerol (PG), cardiolipin (CL), phosphatidic acid (PA) and acylcarnitines (AC)], non‐esterified free fatty acids (NEFA), sphingolipids, sterols and glycerolipids based on Z‐scores calculated from the summed ion abundances normalised to tissue weight. (C) Pie chart representing the proportional (%) distribution of summed ion abundances of glycerolipid, glycerophospholipid, sphingolipid, NEFA and sterol lipid classes. (D–F) Comparison of the total normalised ion abundances for (D) glycerolipids including TG, DAG and MAG, (E) NEFA and ACs, and (F) specific AC and hydroxyl acylcarnitines (OHAC) species. Global lipidomic analysis of ventricles by shotgun mass spectrometry was performed in non‐fasting 6‐month‐old male Cre−;Bscl2^f/f (Ctrl) and Bscl2^cKO (cKO) mice. n = 4 per group with each pooled from three animals. *, p < .05; **, p < .005, unpaired t‐tests (nonparametric)

FIGURE 4

Cardiac dysfunction in Bscl2^cKO mice is associated with bioenergetics failure independent of mitochondrial respiratory dysfunction. (A) Representative transmission electron micrographs of ventricles from 6‐month‐old Cre−;Bscl2^f/f (Ctrl) and Bscl2^cKO (cKO) mice. Red arrowheads: lipid droplets (LDs). Upper panels: scale bar = 2 μm; lower panels: scale bar = 0.5 μm. n = 3 per group. (B) Succinate/rotenone‐driven mitochondrial oxygen consumption rates (OCR) were measured in isolated mitochondria from hearts of 6‐month‐old Ctrl and cKO mice by Seahorse XF24 analyser with sequential addition of ADP, oligomycin (OG), FCCP and antimycin A (AA). Basal, coupled, uncoupled and maximal mitochondrial respiration were shown in (C). n = 4 in triplicates per group. Two‐way ANOVA followed by Tukey's post‐hoc tests. (D) Electron flow assays measuring complex I–IV activities (n = 3 in triplicates) and (E) representative Western blotting and quantification in isolated mitochondria from ventricles of 6‐month‐old Ctrl, Bscl2^cKO , Cre+;Bscl2^w/w mice. n = 4 per group. Mitochondrial HSP60 was used as loading control. (F) Acid‐soluble metabolites (ASM) after incubating heart crude mitochondrial fraction with C‐palmitate. Ctrl, n = 4 in duplicates; cKO, n = 6 in duplicates. (G) Glucose oxidation rate after incubating heart crude mitochondrial fraction with C‐glucose. n = 6 per group. (H) Ventricular glycogen contents, n = 7 per group. Panels F and G were performed in 6‐month‐old Ctrl and cKO mice. Unpaired t‐tests (parametric). (I) ventricular ATP contents in non‐fasting 3 and 6‐month‐old male Ctrl and Bscl2^cKO mice. 3 M: Ctrl, n = 6; cKO, n = 7, 6 M: Ctrl, n = 6; cKO, n = 8. *, p < .05, multiple unpaired t‐tests with Holm–Sidak method

FIGURE 5

Inhibition of fatty acid oxidation partially improves metabolic reserve and cardiac function in Bscl2^cKO mice. (A) 6‐month‐old male Cre−;Bscl2^f/f (Ctrl) and Bscl2^cKO (cKO) mice received daily i.p. injection with PBS or trimetazidine (TMZ) at 15 mg/kg body weights (BW) for a total of 6 weeks. (B) Ratio of ventricle weight (VW) to tibia length (TL). Ctrl, n = 5; cKO, n = 9. (C) Left ventricular internal diameter in end systole (LVIDs); (D) ejection fraction and (E) fractional shortening. PBS‐Ctrl, n = 10; PBS‐cKO, n = 9. n = 9 per TMZ‐treated group. (F) Heatmap of major lipid species based on Z‐score calculated from the summed ion abundances normalised to milligram of wet tissue; (G and H) comparison of the total normalised ion abundances for (G) glycerolipids as well as (H) NEFA and ACs. n = 3 per PBS‐treated group. n = 4 per TMZ‐treated group. Each sample was pooled from three animals. (I) Ventricular glycogen. PBS‐Ctrl, n = 5; PBS‐cKO, n = 9. n = 9 per TMZ‐treated group. (J) Ventricular ATP contents normalised to gram of wet tissue. PBS‐Ctrl, n = 5; PBS‐cKO, n = 8. TMZ‐Ctrl, n = 5, TMZ‐cKO, n = 9. (K) Representative Western blotting and quantification of total cell extracts from ventricles of 6‐month‐old PBS and TMZ‐treated Ctrl and Bscl2^cKO mice. n = 6 per group. Data were normalised to GAPDH with PBS‐treated Ctrl set to 1. Two independent experiments. (L) qRT‐PCR analysis of lipid and glucose transport genes and cardiac stress genes in ventricles of 6‐month‐old PBS and TMZ‐treated male mice. n = 6 per group. *, p < .05; **, p < .005. ns: not significant. Two‐way ANOVA followed by Tukey's post‐hoc tests

FIGURE 6

High‐fat diet feeding increases metabolic reserve and prevents cardiac dysfunction in Bscl2^cKO mice. (A) 3‐month‐old Bscl2^f/f (Ctrl) and Bscl2^cKO (cKO) mice were fed with normal chow diet (NCD) or high‐fat diet (HFD, 60% fat calories) for a total of 3 months (M). (B) Ratio of ventricle weight (VW) to tibial length (TL). NCD‐Ctrl, n = 6; NCD‐cKO, n = 7; HFD‐Ctrl, n = 7; HFD‐cKO, n = 11. (C) Left ventricular internal diameter in end systole (LVIDs); and (D) fractional shortening. NCD‐Ctrl, n = 9; NCD‐cKO, n = 12; HFD‐Ctrl, n = 12; HFD‐cKO, n = 16. (E) Heatmap of major lipid species based on Z‐score calculated from summed ion abundances normalised to milligram of wet tissue; (F and G) comparison of the total normalised ion abundances for (F) glycerolipids and NEFA as well as (G) ACs. n = 3 per NCD group. n = 4 per HFD group. Each sample was pooled from three animals. (H) Myocardial ATP content as normalised to gram of wet tissue. NCD‐Ctrl, n = 8; NCD‐cKO, n = 8; HFD‐Ctrl, n = 7; HFD‐cKO, n = 11. (I) Representative Western blotting of total heart extracts from ventricles of Ctrl and Bscl2^cKO mice. n = 3 per group. Two independent experiments. *, p < .05; **, p < .005. Two‐way ANOVA followed by Tukey's post‐hoc tests

FIGURE 7

Schematic diagram of the indispensable role of BSCL2 in regulating cardiac lipid metabolism and function. Cardiac deletion of BSCL2 causes ATGL overexpression and excessive fatty acid oxidation (FAO) which exhaust intramyocellular triglyceride (TG) and induce drastic depletion of cardiac lipidome, ultimately resulting in energetic and contractile failure in mice. Partial deletion of ATGL, inhibiting FAO by trimetazidine or increasing lipid supply via high‐fat diet (HFD) feeding replenishes cardiac lipidome and alleviates cardiac dysfunction caused by loss of BSCL2