Using Longitudinal Multiomics to Characterize Immune Correlates of COVID-19 Disease Severity and Clinical Outcomes

The Challenge: Unraveling the Complex Interplay Between Immunology, Virology, and Genetics in COVID-19

One of the biggest challenges to curbing the COVID-19 pandemic—which has claimed the lives of nearly 7 million people since it began¹—has been gaining a full understanding of how and why the virus causes so many different clinical outcomes, from asymptomatic disease to death. While studies assessing the relationship between different variants and immune responses to them have aided in our understanding of COVID-19 disease pathogenesis, no studies have connected specific immune response correlates to variation in clinical outcome, which is necessary to develop better diagnostics and therapeutics.

To address this knowledge gap, the Immunophenotyping Assessment in COVID-19 Cohort (IMPACC)—a geographically diverse US consortium of 15 centers and 20 hospital recruitment sites—was established. Researchers performed a combination of longitudinal serology, proteomics, metabolomics, CyTOF (leukocyte frequency and phenotype), gene expression, genomics, transcriptomics, and microbiome analysis on this cohort to gain a thorough, unbiased understanding of different COVID-19 disease trajectories.²

The Metabolon Insight: Contextualizing Metabolites with Disease Trajectories

To identify metabolites associated with COVID-19 clinical outcomes and other physiological measurements, the researchers analyzed the metabolomes of 1275 plasma samples from 486 participants using Metabolon’s Global Discovery Panel.

The Solution: Differentially Dysregulated Plasma Metabolomes are Associated with COVID-19 Disease Trajectories

All biological states associated with the five different COVID-19 disease trajectories as defined by IMPACC were identified by combining six different data types, including metabolomics. Weighted Correlation Network Analysis (WCNA) of the 1,017 plasma metabolites detected using global untargeted metabolomics identified 42 total modules present in patients upon hospital admission, 18 of which were significantly associated with clinical outcome. Notably, seven of these were enriched in individuals with mild COVID-19 disease while 11—including branched-chain amino acid (BCAA), urea cycle, phenylalanine metabolism, tyrosine metabolism, and monoacylglycerol metabolism metabolites—were associated with more severe disease. Previous studies have shown that reactive oxygen species (ROS) production and inflammation in endothelial cells are enhanced by increased BCAAs.³ 

Longitudinal analyses also revealed an additional 26 metabolite modules associated with disease trajectory. Phospholipid metabolites were enriched in those with mild disease at admission and increased over time in all individuals, with the exception of those who eventually succumbed to their infections. In these individuals, phospholipid metabolites eventually decreased over time. A similar decreasing pattern was observed in histidine metabolism and glycerophospholipids metabolites, an interesting result given that histidine residues are often present on viral envelope proteins and aid viral entry into host cells.

Contextualizing metabolomic data with other analyses revealed important insights into severe COVID-19 disease. In addition to the findings discussed above, the researchers identified decreases in NK cell and phospholipid metabolite activators, increased blood neutrophils, increased circulating myosins (which could indicate muscle damage), changes in the cells lining the airways, and an increased abundance of anaerobic bacteria in the airway of patients with fatal SARS-CoV-2 infection. These results paint a picture of heightened levels of viremia driving local and systemic inflammatory responses that impair innate and adaptive immunity and lead to metabolic pathway dysregulation.

Importantly, the data collected and described in this study confirm previous findings from smaller cross-sectional cohorts.

The Outcome: Connecting the Dots Across the Cellular and Molecular Signatures Driving Diverse COVID-19 Clinical Outcomes

This study is a comprehensive profile of COVID-19 disease trajectory and clinical outcomes made possible by the diversity of participants in IMPACC. Not only were metabolic data combined with other genomic, transcriptomic, and physiological measures to hone in on the complexities behind COVID-19 immune responses, but these data were combined with extensive clinical data to characterize five unique COVID-19 disease trajectories and to provide a more complete picture of the cellular and molecular signatures of severe and fatal COVID-19. Importantly, this study highlights the importance of a multiomics approach to studying infectious diseases like COVID-19, as the authors emphasize that while many signatures were observed across multiple analyses, most were identified using only one omic assay.