Endocrine Abstracts

Introduction: Glucocorticoids (GCs) play a crucial role in various physiological processes, with adipose tissue (AT) being an important target organ for their metabolic actions. Besides lipolytic action, the mechanistic regulation of GCs in human AT during physiological exposure is still unclear. We integrated transcriptomic and metabolomic data within the AT to identify mechanistic markers of GC action.

Methods: In a randomised, cross-over trial, ten patients with primary adrenal insufficiency received saline i.v. (GC withdrawal) or circadian infusion of hydrocortisone (GC exposure), with > two weeks in-between. At the end of each 26 hour period (0800 h), an abdominal subcutaneous microdialysis (membrane Asahi Kasei Medical, 3MDa) was performed, with dialysate analysed for metabolomics by LC–MS. Abdominal subcutaneous AT biopsy was collected for transcriptomics by Affymetrix Array. Hypergraph network models were used to identify changes in the metabolome between GC exposure and withdrawal, and integrate metabolomic and transcriptomic data.

Results: In subcutaneous dialysate, 22 metabolites were significantly different between GC exposure and withdrawal. The top five metabolites were lysophosphatidylcholine (lysoPC) (16:0(OH)/0:0), octanoylcarnitine, lysoPC(0:0/16:0(OH)), decanoylcarnitine, and hexanoylcarnitine. In near physiological GC exposure, we observed increased concentrations of five acylcarnitines, three fatty acids (FA), proteolytic metabolites such as gamma-glutamylleucine, and three lysoPCs with longer carbon chains (≥20C). Four lysoPCs with shorter carbon chains (<20C) were found at decreased concentration in GC exposure. We identified a relationship between a group of four acylcarnitines and two lysoPCs during physiological GC exposure. This relationship was disrupted in the withdrawal state, suggesting GC-dependent regulation. DEGs (n=2048 P<0.05) between the interventions were enriched in regulation of lipid metabolic processes by gene ontology. Hypergraphs highlighted a cluster of nine genes, the expression of which was linked to concentrations of acylcarnitines, FA and cortisol. These included several well-known GC-response genes, such as KLF9, PDK4, and ZBTB16, which share positive relationships with these lipolytic and proteolytic metabolites.

Conclusions: Near physiological GC exposure increased lipolysis pathways, with increased concentrations of acylcarnitines, FA, and lysoPCs. A GC-dependent relationship between acylcarnitines and lysoPCs was identified. Acylcarnitines act as transporters of FA across the mitochondrial membrane for beta-oxidation. We observed a shift from shorter chain (C14-16) to longer chain (C20-22) lysoPCs, suggesting increased lipolysis of shorter chain FA. Integration of metabolomics and transcriptomics showed relationships between expression of GC-response genes, and key metabolites in the dialysate, supporting methodology and suggesting an important role of lipolysis, beta-oxidation, and proteolysis in physiological GC action in AT.