GLUCOCORTICOIDS REGULATE MITOCHONDRIAL FATTY ACID OXIDATION IN FETAL CARDIOMYOCYTES

The late gestational rise in glucocorticoids contributes to the structural and functional maturation of the perinatal heart. Here, we hypothesised that glucocorticoid action contributes to the metabolic switch in perinatal cardiomyocytes from carbohydrate to fatty acid oxidation. In primary mouse fetal cardiomyocytes, dexamethasone treatment induced expression of genes involved in fatty acid oxidation and increased mitochondrial oxidation of palmitate, dependent upon glucocorticoid receptor (GR). Dexamethasone did not, however, induce mitophagy or alter the morphology of the mitochondrial network. In neonatal mice, dexamethasone treatment induced cardiac expression of fatty acid oxidation genes in vivo. However, dexamethasone treatment of pregnant C57Bl/6 mice at embryonic day (E)13.5 or E16.5 failed to induce fatty acid oxidation genes in fetal hearts assessed 24 hours later. Instead, at E17.5, fatty acid oxidation genes were down-regulated by dexamethasone, as was GR itself. PGC-1α, required for glucocorticoid-induced maturation of primary mouse fetal cardiomyocytes in vitro, was down-regulated in vivo in fetal hearts at E17.5, 24 hours after dexamethasone administration. Similarly, following a course of antenatal corticosteroids in a sheep model of preterm birth, both GR and PGC-1α were down-regulated in fetal heart. These data suggest endogenous glucocorticoids support the perinatal switch to fatty acid oxidation in cardiomyocytes through changes in gene expression rather than gross changes in mitochondrial volume or mitochondrial turnover. Moreover, our data suggest that treatment with exogenous glucocorticoids may interfere with normal fetal heart maturation, possibly by down-regulating GR. This has implications for clinical use of antenatal corticosteroids when preterm birth is considered a possibility.


98
The dramatic increase in fetal glucocorticoid hormone concentration in late gestation is 99 essential to support the transition from intrauterine to extrauterine life (Hillman et al., 2012;

129
During the transition to a higher oxygen environment and a greater cardiac workload at 130 birth, the cardiac preference for energy substrate switches. The fetal heart derives most of 131 its ATP from glucose and lactate oxidation, with only a minor contribution from fatty acids.

132
After birth, the increased demand for ATP is met primarily by oxidation of long chain fatty 133 acids (Lopaschuk & Jaswal, 2010). This is associated with increased mitochondrial 134 functional capacity. PGC-1a, a master transcriptional regulator of mitochondrial capacity, 135 is expressed in the late gestation fetal heart and expression increases markedly after birth 136 (Lehman et al., 2000). Mice with global knock-out of PGC-1a show 50% mortality before 137 weaning (Lin et al., 2004), suggesting it is important in the perinatal period. PGC-1a is a 138 glucocorticoid target gene and, in vivo in fetal heart, is induced 6 hours after glucocorticoid

165
Here, we hypothesised that glucocorticoids increase fetal cardiomyocyte capacity for fatty 166 acid oxidation. We also asked if any glucocorticoid-mediated increase in mitochondrial 167 fatty acid oxidation capacity involves mitochondrial remodelling by mitophagy.

199
Hearts were removed and frozen as above. Tissues were identified by animal ID (blinding 200 to genotype/treatment group) and stored at -80°C prior to analysis.

219
Two days after their initial steroid or saline treatment, ewes received an intravenous

276
For each experiment, treatment groups were randomised across the plate. Each well was 277 assigned a number (based on the number of treatment groups) so that numbers were 278 spread across the plate. Treatments were then randomised to numbers. On the day of the 9 assay, complete culture medium was gently aspirated from the cells and exchanged for 280 Seahorse assay medium (with supplements defined below). The cells were gently rinsed 3 281 times with Seahorse assay medium, leaving a final volume of 525µl for the assay. During 282 all assays, 3 measurements at 2.5 minute intervals, were recorded at baseline and after 283 each drug addition.

286
ECF assays were normalised to protein measured using sulforhodamine B (SRB) dye-287 based protein assay (Skehan et al., 1990). Initial experiments confirmed the linearity of

414
Internal controls were Tbp (for mouse), and PGK1 and SDHA for sheep fetal heart; these 415 did not differ across treatments. To normalise the spread, data were log10 transformed.

719
The data that support the findings of this study are available from the first and/or 720 corresponding author upon reasonable request.

996
Representative data are from an experiment performed on cardiomyocytes pooled from 997 several fetuses across 4-5 wells, mean ± SD, ns=not significant, *p<0.05, ***p<0.001 for 998 comparisons between Veh/Dex, # p<0.05 for comparisons between control and Eto by 999 two-way ANOVA followed by post hoc Sidak's tests. Three samples were excluded due to 1000 a technical failure (leakage of AR from the port during basal measurements).

1006
After 24 hours, medium was exchanged for Seahorse assay medium supplemented with 1007 5mM glucose, 1mM pyruvate and 0.5mM carnitine. Cells were treated with etomoxir (Eto,