Saturated cardiolipins are potent disruptors of inner mitochondrial membrane structure and function

Figure 1. Cardiac-specific knockouts of Tafazzin alter the acyl chain profile of cardiolipin.A, a conserved remodeling process ensures homogeneously unsaturated acyl chains on CL molecules. Unsaturated CL molecules exhibit a large magnitude of intrinsic curvature and are therefore depicted as having an inverted conical structure. The curvatures of MLCL and CLsat species have not been experimentally determined. B, schematic representation of lipidomics analysis of cardiomyocyte-specific Taz knockouts (Taz-cKO) from whole mouse hearts. Lipidomics data was originally collected and described in (41). C, Taz-cKO hearts exhibit accumulation of MLCL and an increase in MLCL with more saturated acyl chains. Lipidomic analysis was performed on n = 3 to 4 mice per group. ∗∗∗∗p < 0.0001 and ∗∗∗p = 0.0001 by an unpaired two-tailed t test. Error bars indicate standard deviation (SD). D, Taz-cKO hearts lose unsaturated CL species and accumulate CLsat species. ∗∗∗∗p < 0.0001, ∗∗∗p = 0.0002 and ∗∗p = 0.0013 by an unpaired two-tailed t test. Lipidomic analysis was performed on n = 3 to 4 mice per group. Error bars indicate SD.

Figure 2. Manipulation of saturated CL abundance in C2C12 cells by fatty acid supplementation.A, exogenous addition of unsaturated and saturated fatty acids changes the double bond profile of the total PL pool in C2C12 myoblasts. Palm treatment decreases the double bond index (DBI) of the PL pool. DBI is defined as the mean number of double bonds in a given PL acyl chain. Cells were fed with either 100 μM of OA or Palm for 24 h and subjected to lipidomics in biological triplicates (N = 3). ∗∗p = 0.0020 unpaired t test of OA vs. Palm. Error bars indicate SD. B, treatment of TAZ-KO C2C12 myoblasts with 100 μM Palm results in a major increase in total CL saturation. TAZ-KO C2C12 myoblasts were generated and validated in (38). Lipidomics analysis was performed on C2C12 cells in N = 3 biological triplicates. ∗∗p = 0.0010 WT OA vs. WT Palm. ∗∗∗∗p < 0.0001 for TAZ-KO OA vs. TAZ-KO Palm. ∗∗∗∗p < 0.0001 for POCL quantification for WT OA vs. WT Palm and for TAZ-KO OA vs. TAZ-KO Palm. p-values were obtained by two-way ANOVA analysis with Tukey’s correction for multiple comparisons. Error bars indicate SD. C, Palm-treated TAZ-KO cells exhibit an increase in di-saturated CL species. ∗∗∗∗p < 0.0001, ∗p…

Figure 3. Accumulation of saturated CL species drives mitochondrial dysfunction.A, TAZ-KO cells treated with Palm exhibit fragmented mitochondrial morphologies. Mitochondrial morphology was visualized by staining cells with 100 nM Mitotracker red CMX-ROS. Scale bars, 10 μm. The mitochondrial form factor was quantified using the Mitochondria Analyzer plugin (81) from n = 10 cells in N = 3 biological replicates. Form factor values closer to one indicate spherical, fragmented mitochondrial networks while larger values indicate more complex topologies. ∗∗∗p = 0.0008 for WT Palm vs. TAZ-KO Palm, and ∗∗∗∗p < 0.0001 for WT OA and TAZ-KO OA vs. TAZ-KO Palm. p-values were obtained by two-way ANOVA analysis with Tukey’s correction for multiple comparisons. Error bars indicate SD. B, Palm treatment in TAZ-KO cells results in a mild alteration to the abundance of OXPHOS complexes. 30 μg of whole cell extracts were subjected to SDS-PAGE and subsequently analyzed by Western blot against an OXPHOS antibody cocktail. Actin served as a loading control. C, Palm treatment of TAZ-KO C2C12 cells exacerbates defects in mitochondrial respiration. Seahorse respirometry of WT and TAZ-KO C2C12 cells treated with…

Figure 4. Saturated CLs increase membrane ordering of the IMM.A, Mito-Laurdan (2.5 μM) fluorescence microscopy of WT and TAZ-KO C2C12 cells treated with 100 μM of OA, Palm, or a BSA control. Heatmap images are shown of Mito-Laurdan generalized polarization (GP) calculated from ordered and disordered channels. More positive GP indicates an increase in membrane order. Scale bars, 10 μm. Histograms of GP distribution by number of pixels of WT and TAZ-KO C2C12 cells treated with 100 μM of OA, Palm, or a BSA control. Higher GP indicates an increase in membrane order. The dashed line at GP = 0 is shown as an arbitrary reference point for clarity. B, Mito-Laurdan mean GP of individual cells (light grey) and replicate means (black) are shown of WT and TAZ-KO C2C12 cells treated with a 100 μM of OA, Palm, or a BSA control. Values obtained from n > 20 cells of N = 3 replicates. p-value between WT + BSA and TAZ-KO + Palm obtained by one-way ANOVA with multiple comparison correction (Dunnett’s test). Other comparisons against WT + BSA were not significant (p > 0.05). C, POCL liposomes exhibit lower membrane fluidity compared to TOCL and MLCL. GP was determined after staining of liposomes with 2.5 μM…

Figure 5. Altered mechanical properties of saturated cardiolipin.A, POCL is a low curvature CL species. Spontaneous curvature values at 35 °C were determined using small angle X-ray scattering (SAXS) on hydrated lipid suspensions that are diagramed to the left. Representative SAXS profiles of the HII phase of 100% DOPE, and hosted TOCL (10%), POCL (10%) and POPC (10%) in DOPE (90%) are shown. The dashed line indicates the scattering angle of the first HII Bragg peak in pure DOPE. Shifts to the left indicate widening of HII tubules, corresponding to increases in c0. All samples were relaxed in 12% (w/w) tricosene. Extrapolated curvatures (c0) are plotted to the right. B, Martini3 models for CLsat species in comparison to previously employed TOCL and MLCL components. C, extrapolated curvatures (c0) from CL bilayers. The pure systems reflect unicomponent membranes of CL and MLCL species with increasing saturation shown in B. The mixtures reflect combinations of CL species based on lipidomic profiles of C2C12 WT and TAZ-KO cells supplemented with either Palm or OA. D, corresponding bending moduli (Kc) for the systems shown in C. Like for c0, Kc is most affected by saturated CL species like…