Multi-omic Profiling Reveals Early Immunological Indicators for Identifying COVID-19 Progressors

Abstract

The pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has led to a rapid response by the scientific community to further understand and combat its associated pathologic etiology. A focal point has been on the immune responses mounted during the acute and post-acute phases of infection, but the immediate post-diagnosis phase remains relatively understudied. We sought to better understand the immediate post-diagnosis phase by collecting blood from study participants soon after a positive test and identifying molecular associations with longitudinal disease outcomes. Multi-omic analyses identified differences in immune cell composition, cytokine levels, and cell subset-specific transcriptomic and epigenomic signatures between individuals on a more serious disease trajectory (Progressors) as compared to those on a milder course (Non-progressors). Higher levels of multiple cytokines were observed in Progressors, with IL-6 showing the largest difference. Blood monocyte cell subsets were also skewed, showing a comparative decrease in non-classical CD14^-CD16⁺ and intermediate CD14⁺CD16⁺ monocytes. Additionally, in the lymphocyte compartment, CD8⁺ T effector memory cells displayed a gene expression signature consistent with stronger T cell activation in Progressors. Importantly, the identification of these cellular and molecular immune changes occurred at the early stages of COVID-19 disease. These observations could serve as the basis for the development of prognostic biomarkers of disease risk and interventional strategies to improve the management of severe COVID-19.

Fig. 1.. Immune Profiler cell frequency and subset-resolved multi-omic data overview.

(A) Box plots of cell subset frequencies grouped by Non-progressor (NP) versus Progressor (P) pairwise outcome groups. (B) Identified number of significant differentially expressed genes (DEGs) between Non-progressors and Progressors by cell subset from RNA-seq data. (C) Identified number of differentially expressed proteins (DEPs) between Non-progressors and Progressors by cell subset from TaPE-seq data. (D) Identified number of differentially accessible regions (DARs) between Non-progressors and Progressors by cell subset from ATAC-seq data; For bar graphs (B to D), FDR cutoffs for each data type were set and graphed as FDR ≤ 0.05 (black bar) or FDR ≤ 0.1 (gray bar); Abbreviations for cell subsets are defined in table S3.

Fig. 2.. Elevated cytokine and chemokine levels in Progressors.

(A) Concentration in pg/ml of cytokines and chemokines elevated in plasma of Non-progressor (NP) versus Progressor (P) at a FDR ≤ 0.05 as measured by Luminex assay. (B) Cytokine concentration ratios for NP vs P. (C) Spearman’s correlations between immune cell subset frequencies (as a percentage of PBMCs) and plasma IL-6 concentrations with a ρ ≥ 0.2 or ρ ≤ −0.2. Abbreviations for cell subsets are defined in table S3.

Fig. 3.. Perturbations in Progressor monocyte cell subset frequencies.

(A) Ratio of monocyte subsets to total frequency of monocytes in Non-progressors (NP) versus Progressors (P). Significance assessed by p-values for each individual test. (B) Principal Component Analysis of all participant monocyte subsets’ ATAC-seq data, showing a spectrum ranging from classical monocytes (MoCl; red), to intermediate monocytes (MoIn; green), to non-classical monocytes (MoNC; blue) (left panel). The intermediate monocytes of NP (blue) or P (red) are colored (right panel). (C) Box plot of NP and P’s intermediate monocytes on the ATAC-Seq epigenetic spectrum. A Wilcoxon signed-rank test was performed to measure significant differences in Progressor versus Non-progressor intermediate monocytes based on their distances between the MoCl centroid and the MoNC centroid. (D) Typical biaxial CD14/CD16 plot showing defined gates for monocytes that delineate classical, intermediate, and non-classical monocyte subsets (left panel). Finer partitions along the biaxial CD14/CD16 plot, from G0-G10, allow for comparative measurements of frequency across the classical, intermediate, non-classical monocyte transitional path (right panel). (E) Cell frequencies of the G0-G10 gating partitions to total monocytes in NP versus P. Gating partitions are typically ascribed to classical (G0-G1), intermediate (G6-G7), and non-classical (G9-G10) monocyte CD14/CD16 gates; *p≤0.05; **p≤0.01; ***p≤0.001

Fig. 4.. Monocyte Progressor versus Non-progressor DEGs.

(A) Log2 fold-change (log2FC) of DEGs in classical monocytes (MoCl), intermediate monocytes (MoIn), and non-classical monocytes (MoNC) between Non-progressors (NP) and Progressors (P) at FDR ≤ 0.1. (B) DEGs with proximally associated DARs comparing all Progressor to Non-progressor cell subsets at FDR ≤ 0.1. Distance indicates the distance between the ATAC peak from the corresponding gene’s TSS. (C) Chromatin contact via HiChIP between regions with eQTL SNPs (rs6083011, rs6113954, and rs6515302) and the THBD TSS in monocyte subsets, with gene track and literature-reported SNP rs1042580 for context. The black vertical line represents the transcriptional start site (TSS), and colored vertical lines represent variant positions, with the lines for rs6083011, rs6113954, and rs6515302 overlapping in the figure.

Fig. 5.. Highly activated phenotype in effector memory CD8+ T cells of Progressors.

(A) Log2 fold-change (log2FC) of hyperactivation-related genes by RNA-seq in effector memory CD8⁺ T cells between Non-progressors (NP) and Progressors (P). (B) Log2FC in chromatin accessibility by ATAC-seq of hyperactivation-related genes in effector memory CD8⁺ T cells between Non-progressors and Progressors. Multiple bars indicate multiple identified peaks within +/− 1kb of gene TSS. (C) Log2FC of T cell dysfunction or activation-related genes by RNA-seq in effector memory CD8⁺ T cells between NP and P. (D) Log2FC in chromatin accessibility by ATAC-seq of dysfunction or activation-related genes in effector memory CD8⁺ T cells between Non-progressors and Progressors. Multiple bars indicate multiple identified peaks within +/− 1kb of gene TSS. (E) Log2FC of cell surface protein levels of regulatory and activation markers by TaPE-seq in effector memory CD8⁺ T cells between NP and P.