Ceramide-Protein Interactions Modulate Ceramide-Associated Lipotoxic Cardiomyopathy

Abstract

Lipotoxic cardiomyopathy (LCM) is characterized by abnormal myocardial accumulation of lipids, including ceramide; however, the contribution of ceramide to the etiology of LCM is unclear. Here, we investigated the association of ceramide metabolism and ceramide-interacting proteins (CIPs) in LCM in the Drosophila heart model. We find that ceramide feeding or ceramide-elevating genetic manipulations are strongly associated with cardiac dilation and defects in contractility. High ceramide-associated LCM is prevented by inhibiting ceramide synthesis, establishing a robust model of direct ceramide-associated LCM, corroborating previous indirect evidence in mammals. We identified several CIPs from mouse heart and Drosophila extracts, including caspase activator Annexin-X, myosin chaperone Unc-45, and lipogenic enzyme FASN1, and remarkably, their cardiac-specific manipulation can prevent LCM. Collectively, these data suggest that high ceramide-associated lipotoxicity is mediated, in part, through altering caspase activation, sarcomeric maintenance, and lipogenesis, thus providing evidence for conserved mechanisms in LCM pathogenesis in mammals.

Keywords: Annexin; Drosophila; FASN; Unc-45; apoptosis; diabetic cardiac disease; heart; lipogenesis; myriocin; sphingolipids.

Figure 1.. Ceramide Accumulation Induces LCM

(A) Ceramide metabolism in the sphingolipid pathway. (B–G) Analysis of 3-week-old flies with global RNAi-mediated knock down of Cdase, Sk1, or Sk2 (using the ubiquitous Act5 driver [4414-Gal4]). (B) Whole-fly ceramide content (C14:1 sphingoid backbone with C20, C22, or C24 secondary fatty acyl chains), (C) representative M-modes (DD, diastolic diameter; SD, systolic diameter), (D) diastolic diameter, (E) systolic diameter, (F) fractional shortening, and (G) contractile events. (H–K) Wild-type w1118 flies were fed either 100μM C14:1/C2:0, C14:1/C16:0, or C16:1/C16:1 ceramides from day 1 and analyzed at 3 weeks of age relative to 0 μM (vehicle-fed) controls. (H) Diastolic diameter, (I) systolic diameter, (J) fractional shortening, and (K) contractile events. G/K: *p < 0.05 (χ² test); for all other panels, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 (Student’s t test). Error bars represent SEM.

Figure 2.. Inhibition of Ceramide Synthesis Confers a Constricted Phenotype

(A) De novo ceramide synthesis was blocked genetically via knock down of lace (SPT) or schlank (CerS) or pharmacologically via inhibition of SPT by myriocin. (B–E) Analysis of flies with global knock down of lace or schlank. (B) Diastolic diameter, (C) systolic diameter, (D) fractional shortening, and (E) contractile events. (F–I) Wild-type w¹¹¹⁸ flies were fed from day 1 of adulthood with myriocin (100 μM) plus 0 (vehicle), 10, or 100 μM C_14:1/C2:0 ceramide and analyzed 3 weeks later. (F) Diastolic diameter, (G) systolic diameter, (H) fractional shortening, and (I) contractile events. For (E) and (I) *p < 0.05 (χ² test); for all other panels, *p ≤ 0.05, (**p ≤ 0.01, ***p ≤ 0.001 (Student’s t test). Error bars represent SEM.

Figure 3.. Heart-Specific Modulation of Ceramide Metabolism Induces Cardiac Defects

(A–D) Analysis of flies with heart-specific knock down of Cdase, Sk1, or Sk2 (using by the Hand2 driver). (A) Diastolic diameter, (B) systolic diameter, (C) fractional shortening, and (D) contractile defects. (E–H) Analysis of flies with heart-specific knock down of lace or schlank. (E) Diastolic diameter, (F) systolic diameter, (G) fractional shortening, and (H) contractile events. (I) F-actin staining (green) of the flies analyzed in (A)–(H) showing differences in diameters and irregularities in circumferential myofibril structure and arrangement. Gaps between myofibrils are indicated by arrowheads. Scale bar, 10 μm. White arrows denote gaps. (D) and (H) *p < 0.05 (χ² test); all other panels, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 (Student’s t test). Error bars represent SEM.

Figure 4.. Targeted Inhibition of DeNovo Ceramide Synthesis Prevents Ceramide-Associated LCM

(A) Compared to control Canton-S flies, lace transheterozygotes and Sk2 hypomorphic mutants exhibit cardiac defects. (B) Administration of myriocin (100 μM for 3 weeks) prevents LCM phenotypes in Sk2 hypomorphic mutants. (C) Heart-specific inhibition of de novo ceramide synthesis by lace RNAi prevents ceramide-associated LCM in Sk2 hypomorphic mutants. (D) Control flies fed with 100 μM ceramide from day 1 of adulthood for 3 weeks exhibit LCM, which is prevented by heart-specific knock down of lace or schlank. Far right panels: *p < 0.05 (χ² test); all other panels, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 (Student’s t test). Error bars represent SEM.

Figure 5.. The Ceramide-Protein Interactome

(A) Silver-stained SDS-PAGE gel of CIPs pulled down from wild-type flies or mouse heart using a biotinylated ceramide–streptavidin bead trap. Lanes are: LD,ladder; whole extract, mouse heart; PDSN, post-pull-down supernatant (unbound proteins); CIPs, proteins bound to beads; biotin + BSA, biotinylated SA control. (B) Venn diagram showing orthologous CIPs pulled down in whole-fly extracts and mouse heart extracts. (C) Uniprot names of orthologous fly and mouse CIPs. (D) Cytoscape BiNGO analysis showing enrichment of fly and mouse CIPs in overlapping gene ontology categories for biological process, molecular function, andcellular compartment.

Figure 6.. Ceramide-Associated Pathogenesis of LCM-like Phenotypes Is Caspase-Dependent

(A) The intrinsic apoptotic pathway in Drosophila, which is highly conserved in mammals. (B) Flies expressing GFP driven by the hid promoter were fed from day 1 of adulthood with vehicle or 100 μM ceramide and analyzed 3 weeks later. Red line indicates outline of cardiac chamber, arrows indicate GFP expression in the cardiomyocyte. (C–F) Analysis of flies with heart-specific knockdown of Anx (Annexin X) fed from day 1 of adulthood with vehicle or 100 μM ceramide and analyzed 3 weeks later. (C) Diastolic diameter, (D) systolic diameter, (E) fractional shortening, and (F) contractile events. (G–J) Analysis of flies with heart-specific overexpression of dIAP fed from day 1 of adulthood with vehicle or 100 μM ceramide and analyzed 3 weeks later. (G) Diastolic diameter, (H) systolic diameter, (I) fractional shortening, and (J) contractile events. (F) and (J) *p < 0.05 (χ² test); all other panels, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 (Student’s t test). Error bars represent SEM.

Figure 7.. Ceramide-Associated LCM Is CIPDependent

(A) Membranes were spotted with 1 μg, 0.5 μg, and 0.1 μg of either C6 biotin, C14:1, or C18:1 ceramide as indicated on the blot. Membranes were incu-bated with purified dUnc45 (top two) or hFASN (bottom two). (B) Bound CIPs were detected by incubation withrespective antibodies. (C) Relative cardiac dUnc45 and dFASN1 mRNAexpression under normal (white) and high ceramide (black) feeding conditions was measured by qPCR. (D–G) Analysis of cardiac parameters in flies with heart-specific overexpression of Unc-45 (D), heart-specific knock down of FASN1 (E), and in combination (F) fed from day 1 of adulthood with vehicle or 100 μM C14:1/C2:0 ceramide and analyzed 3 weeks later. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. (G) Contractile events in the flies described in (B)–(D), (χ² test, *p < 0.05). Error bars represent SEM. Sph, sphingosine; SM, sphingomyelin; Cer1P, ceramide 1-phosphate; C6/C16Cer, C6/C16 fatty acyl chain on a C18 sphingoid backbone; DH, dihydro derivatives.