Sphingolipidome of plasma, liver, and adipose tissues and its association with insulin response to oral glucose testing in Icelandic horses

Insulin dysregulation (ID) is a determinant of equine metabolic syndrome. Among the sphingolipids, ceramides contribute to the development of ID; however, the cross talk between the liver and adipose tissue (AT) depots and the variation among AT depots in terms of ceramide metabolism are not well understood. We aimed to characterize the sphingolipidome of plasma, liver, and AT (nuchal, NUAT; subcutaneous, SCAT; omental, OMAT; retroperitoneal, RPAT) and their associations with insulin response to oral glucose testing (OGT) in normoinsulinemic and hyperinsulinemic horses. Plasma, liver, and AT samples were collected from 12 Icelandic horses upon euthanasia and analyzed by liquid chromatography-mass spectrometry. Eighty-four targeted compounds were effectively quantified. Comparing the AT depots, greater (false discovery rate, FDR < 0.05) ceramide, dihydroceramide, and sphingomyelin concentrations and lower glucosyl- and galactosyl-ceramides were found in RPAT and OMAT than in NUAT and SCAT. Hyperinsulinemic response to OGT was associated with sphingolipidome alterations primarily in the RPAT and OMAT, whereas the NUAT sphingolipidome did not show signs of ceramide accumulation, which was inconsistent with the previously proposed role of nuchal adiposity in ID. The plasma sphingolipidome was not significantly associated with the liver or AT sphingolipidomes, indicating that plasma profiles are determined by an interplay of various organs. Furthermore, hepatic sphingolipid profiles were not correlated with the profiles of AT depots. Finally, statistically valid partial least square regression models predicting insulin response were found in the plasma ( Q 2 = 0.58, R 2 = 0.98), liver ( Q 2 = 0.64, R 2 = 0.74), and RPAT ( Q 2 = 0.68, R 2 = 0.79) sphingolipidome, but not in the other adipose tissues.