Temporal effects of safflower oil diet-based linoleic acid supplementation on Barth syndrome cardiomyopathy

Barth syndrome (BTHS) is an X-linked mitochondrial lipid disorder caused by mutations in TAFAZZIN (TAZ)¹. Cardiomyopathy is a major clinical feature in BTHS¹. Thus far, there is no curative therapy for BTHS¹.

TAZ is an acyltransferase essential for remodeling/maturation of cardiolipin (CL), a signature phospholipid of mitochondria. Nascent CL contains a mixture of fatty acyl chains. TAZ remodels CL to mature CL containing more polyunsaturated fatty acids. In mammalian heart, the major mature CL species is tetralinoleoyl-CL, containing four linoleic acid (LA) side chains¹. The acyl composition of CL determines its biophysical properties and functions. TAZ deficiency causes inefficient transacylation from linoleoyl-phospholipid to monolysocardiolipin (MLCL), resulting in decreased mature CL, accumulation of MLCL and nascent CL, and concomitantly elevated MLCL to total CL ratio (MLCL/CL)¹.

Concentrations of specific free fatty acids determines the acyl composition of nascent CL. LA supplementation increases mature CL levels by increasing incorporation of linoleoyl groups into de novo synthesized CL in BTHS fibroblasts² and mitigates contractile defects of BTHS-iPSC derived cardiomyocytes³, suggesting that LA supplementation could ameliorate BTHS by increasing incorporation of LA into nascent CL towards mature CL, thereby bypassing TAZ remodeling. LAis one of two essential fatty acids that must be obtained through diet. Over 70% ofsaffloweroil is composed ofLA. Here, we investigate potential therapeutic effects of safflower oil based LA supplementation on BTHS cardiomyopathy in vivo by utilizing Taz global (gKO) and cardiomyocyte-specific knockout (cKO)⁴ mouse models. All mouse protocols were approved by the Institutional Animal Care and Use Committee.

Our high-LA safflower oil diet (LAD) contained 10% LA (LA weight/chow weight) and the low-LA palm oil control diet (CD) contained 1% LA. Pregnant Taz^+/− females were fed with LAD or CD starting at the day of detecting copulatory plugs from crossing with wildtype males. Male offspring, which were Taz gKO (Taz^−/Y) or wildtype, were maintained on specific diets throughout their life (Figure A). LAD did not prevent early perinatal lethality in gKO mice (Figure B). However, body weights of gKOs were significantly increased in LAD versus CD mice (Figure C–D). Long-term survival of Taz gKO mice was comparable in LAD and CD groups (Figure E).

A. Schematic of experimental protocol for dietary linoleic acid supplementation in Tafazzin (Taz) global knockout (gKO) and wildtype (WT) mice. In this study, dietary LA supplementation was achieved by using safflower oil as an ingredient in rodent diet. Palm oil, which contains low LA, was used in control diets. In compliance with dietary recommendations to prevent cardiovascular disease, total fat content of all diets was 28% of the total kcal intake. Our custom high-LA safflower oil diet (LAD) contained 10% LA (LA weight/chow weight; law/cw) and the low-LA palm oil control diet (CD) contained 1% LA (law/cw). Pregnant Taz^+/− females were fed with LAD or CD starting at the day of detecting copulatory plugs from crossing with WT (Taz^+/Y) males. Male offspring, which were Taz gKO (Taz^−/Y) or WT, were maintained on LAD or CD throughout their life. Echocardiography analysis was performed at 2, 4, 6, 8 months of age. B. Percentage of WT and gKO mice in LAD and CD groups. According to Mendelian ratios, the expected percentage of each genotype is 50%. n = 222 for CD group; n=154 for LAD group. C. Representative image illustrating the improvement of body size in LAD treated gKO mice at 4 months of age. Scale bar: 1 cm. D. Body weight of Taz gKO and WT mice under LAD or CD treatments at 2, 4, 6 and 8 months of age. n = 9 for CD-treated WT mice; n = 10 for LAD-treated WT mice; n = 10 for CD-treated gKO mice; n = 12 for LAD treated-gKO mice. E. Kaplan-Meier survival curves for Taz gKO and WT mice under LAD or CD treatments. n = 17 for CD-treated WT mice; n = 15 for LAD-treated WT mice; n = 11 for CD-treated gKO mice; n = 12 for LAD-treated gKO mice. F-H. Echocardiographic measurements of left ventricular percentage of fractional shortening (% FS) (F), end-systolic LV internal diameter (LVIDs) (G), and end-diastolic LV internal diameter (LVIDd) (H) for WT and gKO mice in LAD and CD groups at 2, 4, 6, 8 months of age. n = 5 for CD-treated WT mice; n = 5 for LAD treated-WT mice; n = 5 for CD-treated gKO mice; n = 7 for LAD-treated gKO mice. I. Schematic of experimental protocol for dietary LA supplementation in Taz cardiomyocyte-specific knockout (cKO) and Cre negative control (Ctrl) mice. Taz cKO and Ctrl mice were fed with LAD or CD starting at 1 month of age, and continued until 8 months of age. Echocardiography analysis was performed at baseline (one month of age) and followed up at 4, 6, 8 months of age. We did not observe any lethality in Taz cKO mice by 8 months of age. J-K. Echocardiographic measurements of left ventricular percentage of fractional shortening (% FS) (J) and end-systolic LV internal diameter (LVIDs) (K) for WT and cKO mice in LAD and CD groups at 1, 4, 6, 8 months of age. n = 7 for CD-treated Ctrl mice; n = 4 for LAD-treated Ctrl mice; n = 10 for CD-treated cKO mice; n = 9 for LAD-treated cKO mice. L. Representative electron micrographs of the hearts of cKO mice in LAD and CD groups at 4 months of age. Scale bar: 2 μm. M. Quantification of the percentage of mitochondria with abnormal cristae at 4 months of age. n=3 mice per group. N. Quantified fluorescence intensity of MitoSOX (Mitochondrial Superoxide Indicator) staining on the cardiac mitochondria isolated from Taz cKO and Ctrl mice in LAD and CD groups at 4 months of age. n=3 mice per group. O-P. Ratios of the amount of each cardiolipin (CL) molecule in Taz cKO hearts over its amount in Ctrl hearts under either CD (red or orange) or LAD (green or blue) treatments at 4 (O) and 8 (P) months of age. Values were converted to log2. CL molecules which displayed statistically significant differences between cKOs vs Ctrls were selected for this analysis. Q. Ratios of the amount of each CL molecule in CD-treated cKO hearts over its amount in LAD-treated cKO hearts at 4 (green) and 8 (blue) months of age. Data are represented as the mean ± SEM. α: P< 0.05 CD-treated WT or Ctrl vs CD-treated gKO or cKO. β: P< 0.05 CD-treated WT or Ctrl vs LAD-treated WT or Ctrl. ε: P< 0.05 LA-treated WT or Ctrl vs LAD-treated gKO or cKO. δ: P< 0.05 CD-treated gKO or cKO vs LAD-treated gKO or cKO. *: P < 0.05 between indicated groups, by mixed-effects ANOVA (four group comparison with repeated observations for Figure D, F, G, H, J, K), two-way ANOVA (four group comparison for Figures M, N), or 2-tailed Student’s t test (two group comparison for Figures O, P, Q). Tukey’s multiple comparison test was used for multiple pairwise comparisons following ANOVA analysis. Survival data in Figure E is analyzed by Kaplan-Meier survival analysis with a log-rank method of statistical analyses.

Cardiac function was assessed in LAD-gKOs, revealing fully prevention of cardiomyopathy at 4 months of age, when CD-gKOs displayed dilated cardiomyopathy (DCM). Beneficial effects of LAD extended to 6 months of age, when DCM of gKOs was partially rescued. Cardiac function in LAD-gKOs trended towards improvement at 8 months of age, although this trend was not statistically significant compared to CD-gKOs (Figure F–H). The unsaturated fatty acid γ-linolenic acid drives cardiac metabolic maturation⁵. We did not observe an effect of linoleic acid supplementation on cardiomyocyte maturation. Moreover, potential benefits of LAD initiated after birth remain unknown.

To avoid systemic effects of global loss of Taz, Taz cKO and control (Ctrl) mice were fed LAD or CD from 1 to 8 months of age (Figure I). LAD fully protected cKOs from DCM (Figure J–K) at 4 months of age, but protective effects declined with aging. At 6 and 8 months, LAD partially ameliorated DCM in cKO mice (Figure J–K). We found improved cristae morphology and slightly attenuated mitochondrial superoxide in LAD-cKO hearts, compared with CD-cKOs at 4 months of age (Figure L–N).

Lipidomic analyses were performed on Taz cKO and control hearts following LAD or CD at 4 and 8 months of age. We observed “BTHS-like” CL profiles in cKO hearts both in LAD and CD groups (Figure O–Q). However, LAD ameliorated the severity of CL abnormalities in cKO hearts. At 4 months of age, we found that levels of total CL and mature CL (70:7, 72:6, 72:7, 72:8, 74:7, 74:10) declined less in LAD-cKO hearts compared with CD group. MLCL amounts, MLCL/CL ratios and nascent CL (66:1, 68:1, 66:2, 70:2, 72:2, 72:3) increased less in LAD-cKO hearts compared with the CD group (Figure O). Beneficial effects of LAD on CL profiles declined at 8 months of age, when there was no significant difference in the ratios of total CL, MLCL and CL(68:1) between LAD and CD groups (Figure P). However, increases in MLCL/CL ratios and levels of CL(66:1, 66:2, 70:2, 72:2, 72:3) in LAD-cKO hearts at 8 months of age were less severe relative to the CD group, albeit the extent of rescue was less than that observed at 4 months of age (Figure Q). These results may explain the observed decline in beneficial effects over the course of treatment. Preserved CL abnormalities in LAD-cKOs perhaps led to cumulative defects contributing to cardiomyopathy at later stages.

Our study demonstrated that dietary LA supplementation can delay development of BTHS cardiomyopathy at early stages, with declining effectiveness as the disease progresses, providing important insights into temporal effects of LA in BTHS therapy. Future studies will examine temporal specific effects of LA supplementation on CL profiles and how this contributes to mitochondrial and cardiac dysfunction in BTHS cardiomyopathy.

The data, analytical methods, and study materials that support the findings of this study will be available to other researchers from the corresponding authors on reasonable request.

Nonstandard Abbreviation and Acronyms

BTHS

Barth syndrome

TAZ

TAFAZZIN

CL

cardiolipin

LA

linoleic acid

gKO

Tafazzin global knockout

cKO

Tafazzin cardiomyocyte-specific knockout

CD

control diet

LAD

high linoleic acid diet