Hygor Marcos Ribeiro de Souza¹, Gabriela D. A. Pinto¹, Albert Koulman², Alexandre Guedes Torres¹, Tatiana El-Bacha¹

Pregnancy is characterized by profound physiological and metabolic adaptations, including significant modifications in lipid metabolism. SARS-CoV-2 infection has been associated with marked dysregulation of lipid pathways, contributing to systemic inflammation and disease severity. Despite increasing evidence on lipidomic alterations in COVID-19 patients, data regarding pregnant women remain scarce. Comprehensive lipid profiling in this specific population is essential to elucidate potential biomarkers related to disease progression and maternal–fetal outcomes. Here, liquid chromatography-high resolution mass spectrometry (LC-HRMS)-based lipidomic was used to identify lipids associated with COVID-19 pathophysiology and disease outcome in pregnant women.

Lipids were extracted from the plasma of 110 pregnant women who underwent RT-PCR testing for COVID-19 using nasal swabs. The cohort included 16 healthy controls, 23 symptomatic COVID-19–positive patients, 58 symptomatic COVID-19–negative patients with other respiratory infections, and 13 severe COVID-19 cases admitted to the intensive care unit. Samples were prepared by diluting 50 μL of plasma to 650 μL of chloroform into a centrifuge tube followed by vortexing for 30 s. Subsequently, 100 µL of SPLASH Lipidomix solution, 250 µL of methanol and 400 µL of acetone were added to each sample for lipid extraction by a modified Folch method, mixed for 30 s and centrifugated for 10 min at 20,000 ×g. The supernatant (700 µL) was collected and dried using a SpeedVac vacuum concentrator. The dried extracts were reconstituted in 100 µL of a mixture of acetonitrile:water:propan-2-ol (1:1:2, v/v/v) and vortex-homogenized. The resulting supernatant was used for LC-HRMS analysis.

In this study, 133 lipid species in plasma were identified (level 1 of identification) according to MSI protocol, which were included in univariate and multivariate data analyses. Identified lipid classes included glycerolipids (36 triacylglycerols (TG) and 1 diacylglycerol (DG)); glycerophospholipids (28 phosphatidylcholines (PC), 3 PC-plasmalogens (PC-P)); 5 lysophosphatidylcholines (LPC), 10 phosphatidylethanolamines (PE), 11 PE-plasmalogens (PE-P), 3 ether-linked phosphatidylethanolamines (PE-O), 2 lysophosphatidylethanolamines (LPE), 1 phosphatidylglycerol (PG), 10 phosphatidylinositols (PI), and 1 lysophosphatidylinositol (LPI)), and sphingolipids (18 sphingomyelins (SM), 4 hexosylceramides (Hex-Cer), and 3 lactosylceramides (Lac-Cer)). Principal components analysis (PCA) score plots showed a clear separation between of several COVID-19 cases versus all other groups, in the negative and positive analysis modes, respectively. A total of ten discriminant lipid classes were identified based on PCA loading plots. Notably, the lipid classes of SM, HexCer, PI, Lac-Cer, LPC, LPE, and PE-P were found at lower levels in the severe cases’ group, whereas PE, DG, and TG were elevated when compared to the other groups. In summary, our lipidomic analysis revealed a distinct lipid alteration in pregnant women with severe COVID-19, marked by reduced membrane lipids and increased energy-storage lipids, suggesting metabolic imbalance and potential biomarkers for disease severity in this vulnerable group.

Agradecimentos: Financial support: FAPERJ, CAPES and CNPq