Dynamic modulations of sphingolipids and glycerophospholipids in COVID‐19

BACKGROUND: A heterogeneous clinical phenotype is a characteristic of coronavirus disease 2019 (COVID‐19). Therefore, investigating biomarkers associated with disease severity is important for understanding the mechanisms responsible for this heterogeneity and for developing novel agents to prevent critical conditions. This study aimed to elucidate the modulations of sphingolipids and glycerophospholipids, which have been shown to possess potent biological properties. METHODS: We measured the serum sphingolipid and glycerophospholipid levels in a total of 887 samples from 215 COVID‐19 subjects, plus 115 control subjects without infectious diseases and 109 subjects with infectious diseases other than COVID‐19. RESULTS: We observed the dynamic modulations of sphingolipids and glycerophospholipids in the serum of COVID‐19 subjects, depending on the time course and severity. The elevation of C16:0 ceramide and lysophosphatidylinositol and decreases in C18:1 ceramide, dihydrosphingosine, lysophosphatidylglycerol, phosphatidylglycerol and phosphatidylinositol were specific to COVID‐19. Regarding the association with maximum severity, phosphatidylinositol and phosphatidylcholine species with long unsaturated acyl chains were negatively associated, while lysophosphatidylethanolamine and phosphatidylethanolamine were positively associated with maximum severity during the early phase. Lysophosphatidylcholine and phosphatidylcholine had strong negative correlations with CRP, while phosphatidylethanolamine had strong positive ones. C16:0 ceramide, lysophosphatidylcholine, phosphatidylcholine and phosphatidylethanolamine species with long unsaturated acyl chains had negative correlations with D‐dimer, while phosphatidylethanolamine species with short acyl chains and phosphatidylinositol had positive ones. Several species of phosphatidylcholine, phosphatidylethanolamine and sphingomyelin might serve as better biomarkers for predicting severe COVID‐19 during the early phase than CRP and D‐dimer. Compared with the lipid modulations seen in mice treated with lipopolysaccharide, tissue factor, or histone, the lipid modulations observed in severe COVID‐19 were most akin to those in mice administered lipopolysaccharide. CONCLUSION: A better understanding of the disturbances in sphingolipids and glycerophospholipids observed in this study will prompt further investigation to develop laboratory testing for predicting maximum severity and/or novel agents to suppress the aggravation of COVID‐19.