key: cord-0867654-o2k8pquk
authors: Chen, Y.-M.; Zheng, Y.; Yu, Y.; Wang, Y.; Huang, Q.; Qian, F.; Sun, L.; Song, Z.-G.; Chen, Z.; Feng, J.; An, Y.; Yang, J.; Su, Z.; Sun, S.; Dai, F.; Chen, Q.; Lu, Q.; Li, P.; Ling, Y.; Yang, Z.; Tang, H.; Shi, L.; Jin, L.; Holmes, E. C.; Ding, C.; Zhu, T.-Y.; Zhang, Y.-Z.
title: COVID-19 severity is associated with immunopathology and multi-organ damage
date: 2020-06-22
journal: nan
DOI: 10.1101/2020.06.19.20134379
sha: 7b21c42b570e25e1de81a7c86caf993b5d669388
doc_id: 867654
cord_uid: o2k8pquk

COVID-19 is characterised by dysregulated immune responses, metabolic dysfunction and adverse effects on the function of multiple organs. To understand how host responses contribute to COVID-19 pathophysiology, we used a multi-omics approach to identify molecular markers in peripheral blood and plasma samples that distinguish COVID-19 patients experiencing a range of disease severities. A large number of expressed genes, proteins, metabolites and extracellular RNAs (exRNAs) were identified that exhibited strong associations with various clinical parameters. Multiple sets of tissue-specific proteins and exRNAs varied significantly in both mild and severe patients, indicative of multi-organ damage. The continuous activation of IFN-I signalling and neutrophils, as well as a high level of inflammatory cytokines, were observed in severe disease patients. In contrast, COVID-19 in mild patients was characterised by robust T cell responses. Finally, we show that some of expressed genes, proteins and exRNAs can be used as biomarkers to predict the clinical outcomes of SARS-CoV-2 infection. These data refine our understanding of the pathophysiology and clinical progress of COVID-19 and will help guide future studies in this area.

Masters and Perlman, 2013). In addition, they cause hepatic and neurological diseases of 53 varying severity (Masters and Perlman, 2013) . 54 The world is currently experiencing a disease pandemic (COVID-19) caused by a newly 55 identified coronavirus called SARS-CoV-2 (Wu et al., 2020a). At the time of writing, there 56 have been more than 6 million cases of SARS-CoV-2 and over 387,000 deaths globally 57 (WHO, 2020). The disease leads to both mild and severe respiratory manifestations, with the 58 latter prominent in the elderly and those with underlying medical conditions such as 59 cardiovascular and chronic respiratory disease, diabetes, and cancer (Guan et al., 2020) . In 60 addition to respiratory syndrome, mild gastrointestinal and/or cardiovascular symptoms as 61 well as neurological manifestations have been documented in hospitalized COVID-19 patients 62 (Mao et al., 2020) . Combined, these data point to multiple organ failures, and hence that 63 COVID-19 pathogenesis is complex, especially in patients experiencing severe disease. 64 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint 4 It is believed that SARS-COV-2 is able to use angiotensin-converting enzyme 2 (ACE 2) 65 as a receptor for cell entry (Zheng et al., 2020; Zhou et al., 2020b) . ACE2 is attached to the 66 outer surface (cell membranes) of cells in the lungs, arteries, heart, kidney, and intestines 67 (Hamming et al., 2004) . Additionally, ACE2 is expressed in Leydig cells in the testes (Jiang et 68 al., 2014) and neurological tissue . As such, it is possible that these organs 69 might also be infected by SARS-CoV-2. The host immune response to SARS-CoV-2 may also 70 impact pathogenicity, resulting in severe tissue damage and, occasionally, death. Indeed, 71 several studies have reported lymphopenia, exhausted lymphocytes and cytokine storms in (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint 5 We studied 66 clinically diagnosed and laboratory confirmed COVID-19 patients hospitalized 86 at the Shanghai Public Health Clinical Center, Shanghai, China between January 31st and 87 April 7th, 2020 (Fig. 1A , Tables S1 and S2). At the time of writing, 55 (49 mild and 6 severe) of 88 the 66 patients have recovered and been discharged following treatment, while five patients 89 (1 mild and 4 severe) remain in the hospital and are receiving ongoing treatment. 90 Unfortunately, six patients (all severe) died. 91 Molecular variation associated with COVID-19 pathophysiology 92 Serial blood and throat swab samples were collected from all patients, as well as from 17 93 healthy volunteers. To determine whether COVID-19 pathophysiology was associated with 94 particular molecular changes, a total of 23,373 expressed genes, 9,439 proteins, 327 95 metabolites and 769 exRNAs were examined using a multi-omics approach combining 96 transcriptomics, proteomics, and metabolomics (Fig. 1B) . Compared with healthy controls, 97 mild and severe patients had significantly different expression patterns (higher or lower) in 98 6.79% and 26.0% of expressed genes, 52.1% and 51.7% of proteins, 7.34% and 15.6% of 99 metabolites and 39.9% and 20.5% of exRNAs, respectively (Fig. 1C, Tables S3-S6) . 100 Significant differences in the principal component 1 (PC1), PC2 and/or PC3 between healthy 101 controls, mild and severe COVID-19 patients were observed (Figs. 2A and S1A). Remarkably, 102 there were significant correlations between multi-omics data and classical blood and 103 biochemical parameters (Fig. 2B) , suggesting that the molecular changes identified directly 104 impact the pathophysiology of COVID-19. 105 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The correlation between molecular variation and COVID-19 pathophysiology was best 106 reflected in the proteomic analysis (Fig. 2) . Specifically, there was significant downregulation 107 in the tricarboxylic acid cycle (TCA) and glycolytic pathways in both mild and severe patients 108 compared to healthy controls (Figs. 2C and S1B). However, the hypoxia-inducible factors 109 (HIF-1) signaling pathways and well-known host defense pathways (e.g., T cell receptor 110 signaling pathway, ISG15 antiviral signaling pathway) were elevated in these patients (Figs. 111 2C and S1B). Additionally, we identified 14 co-expression groups ("modules") of proteins that 112 were highly correlated to clinical parameters ( (Fig. 2F) . Notably, correlations between the proteins in these modules were also identified, 117 suggesting that proteins may interact in defining clinical outcome (Figs. S2C and S2D). In 118 addition to proteins, lipoprotein variation was also significantly correlated with immune 119 changes including IgG, monocytes and procalcitonin (Fig. 2B) . Combined, these data reveal a 120 significant association between specific molecular variations and COVID-19 pathophysiology. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Notably, the majority of proteins related to organ function were downregulated in COVID-19 127 patients (Fig. 3B ). As expected, lung-enhanced proteins varied significantly in the plasma of 128 both mild and severe patients. Likely because lung-enhanced proteins are not rich in the 129 human protein atlas, the lung-enhanced proteins in either mild or severe patients did not 130 achieve the top rank. Nevertheless, lung abnormality was reflected in the activation of the 131 HIF-1 signaling pathway and reactive oxygen species metabolic processes in all patients (Fig.   132 3D). Liver-and brain-enhanced proteins also varied significantly, followed by those from the 133 testis, intestine and other organs, suggesting that these organs might also be seriously 134 affected (Fig. 3A ). Severe brain dysfunction was reflected in the significant decline of 135 brain-enhanced proteins regulating neurotransmitter synthesis, neurotransmitter transport, 136 and the numbers of neurotransmitter receptors, as well as a significant decrease in proteins 137 including ENO1, MBP and NEFM that are known biomarkers to reflect brain dysfunction (Figs. 138 3C and 3F). Liver-enhanced proteins, that regulate the transportation of sterol and cholesterol, 139 were downregulated, while those involved in acute inflammatory response were elevated in 140 both mild and severe patients (Fig. 3E ). Testis-enhanced proteins involved in the cell cycle and 141 cell proliferation were upregulated in all male patients, although proteins (e.g. YBX2) 142 associated with reproduction were significantly downregulated. Heart specific proteins related 143 to cardia muscle contraction and oxidative reduction were reduced in COVID-19 patients (Fig.   144 3H). Finally, variation in tissue-enhanced proteins was also associated with COVID-19 severity. 145 For example, brain-enhanced proteins enriched in tubulin accumulation were upregulated in 146 mild disease patients, indicating multiple neuron cell apoptosis in these patients. However, 147 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Organ dysfunction was also reflected in the relative proportion of different cell 150 populations. We identified 16 cell types whose abundance changed significantly following 151 virus infection (Fig. 3F ). For example, the set of proteins expressed by alveolar type 1/2 152 epithelial cells (AT1 and AT2) were significantly downregulated in all patients (Fig. 3G) . In Table S7 ). Notably, IFN 165 signaling was continuously activated in severe patients during the entire period of 166 hospitalization (Fig. 4A) , while negative regulators of innate immune signaling (e.g. TRIM59, 167 USP21 and NLRC3) were downregulated (Fig. S3A) . Additionally, significant increases of IL-6, 168 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Higher neutrophil counts were observed in severe patients but not in mild patients during 172 hospitalization (Fig. S3C) . Examination of the neutrophil transcriptomic signatures revealed 173 that excessive neutrophil activation was associated with severe rather than mild disease. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint 10 exhaustion markers (Fig. S3F) . Strikingly, the severe disease group had a greater abundance 190 of ARG1 (Fig. S3G) . Finally, the mild group had higher TCR diversity than the severe group 191 (Fig. S3H ). In sum, our data suggest that the T cell response is indispensable to successful 192 host defense against SARS-CoV-2. 193 Finally, we investigated the immune signatures associated with poor COVID-19 194 prognosis. Notably, KEGG functional analysis revealed that gene sets of the "IL-17 signaling 195 pathway" were significantly enriched in the severe-fatality group. Further analysis of the 196 signature components revealed that p38 MAPK activation was dominant in fatal cases, while 197 higher levels of IL13 and IFNG were present in survivors (Fig. 4D ). These gene signatures 198 might contribute to greater neutrophil influx (CXCL2 and CXCL6) and inflammation (S100A8), 199 and could be detrimental in the severe disease group (Fig. 4E ). (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. 285 We used a multi-omics approach to identify numerous expressed genes, proteins, 286 metabolites and exRNAs from COVID-19 patients with a range of disease severities, and that 287 were significantly correlated with key clinical features as well as to classic blood and 288 biochemical parameters (Fig. 2) . These data therefore provide a comprehensive molecular 289 view of the pathophysiology of COVID-19. Finally, based on our multi-omics data (Fig. S6) , 290 mild and severe COVID-19 cases may need different therapeutic strategies. 291 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. lung, liver, brain, heart in COVID-19 patients, and also identify damage in other organs 302 including the testis (Fig. 3) . In the case of brain and testis damage, a key issue is how 303 SARS-CoV-2 is able to cross the blood-brain or the blood-testis barriers? One possibility 304 might be that heparin was prescribed for coaggregation problems commonly observed in (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (Fig. 3F) , which in turn will lead to the accumulation alveolar fluid and 317 hence cause hypoxia (Vadász and Sznajder, 2017). In addition, we also noted that HIF1a 318 signaling was modified which may further worsen ALI (Dada et al., 2003) . Thus, our molecular 319 data suggest that removal of excess alveolar fluid and the restoration of alveolar structure will 320 be of major clinical importance. 321 Our data also identified immune pathophysiology a factor that greatly impacted (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. 

Study design and patient cohort 373 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. analysis. Libraries that passed the following criteria were retained: (i) more than five million 432 reads; (ii) more than 90% of reads aligned to the human reference genome; (iii) over 10,000 433 genes were expressed (a gene with FPKM>0.5 was identified as an expressed gene). In 434 addition, to monitor data quality across batches, libraries of some heathy control samples 435 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint 23 value of q < 0.25 was considered statistically significant. The Normalized Enrichment Score 456 (NES) of significant immune modules from BTMs was used to denote enrichment levels. 457 The fraction of the cell subsets was calculated using the enrichment-score based (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Four data sets representing the (i) clinical tests, (ii) exRNA-seq, (iii) mRNA-seq, and (iv) 557 proteomics quantification analysis were used to develop prognostic models for the prediction 558 of patient outcomes (i.e. good or poor). Patients with a "good" outcome included those with 559 mild or severe syndrome but who were discharged after treatment; while patients with "poor" 560 outcomes included those who died or remained in ICU for more than two months. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint 29 more robust prognostic biomarkers, 50 runs of five-fold cross-validation process were 578 therefore applied to the whole data set. The variables used by the best performing machine 579 learning algorithm were identified as prognostic biomarkers for each data set. 580 581 Learning curve model comparison (LCMC) was performed using Predictive Modeling Review 582 as available in JMP Genomics 10 583 (https://www.jmp.com/en_us/software/genomics-data-analysis-software.html). LCMC reveals 584 the effects of sample size on the accuracy and variability of the predictive models using 10 585 runs of 4-fold cross-validation. 586 We performed LCMC with prognosis (good or poor) as target variables, and the clinical 587 variables, exRNA, mRNA, or proteomics measurements as predictors. Fig. S5C shows each 588 individual (RMSE) and (AUC) learning curve and the average for each of the eight partition 589 tree models for clinical endpoints, as well as exRNA using K-fold cross validation. The LCMC 590 suggested that with up to 15 samples, eight partition tree models reached AUC as 1 for 591 clinical variables. However, more than 23 and 30 samples were needed for one and three 592 models, respectively, to reach AUC of 1 for exRNA-seq data. The variability of RMSE and 593 AUC for the proteomic and mRNA-seq data (not shown) were between that observed for 594 clinical variables and the exRNA data.

Univariate statistical analysis was performed using student's t-test, Mann-Whitney U tests or 597 ANOVA tests to compare continuous variables. Chi-square tests and Fisher's exact tests 598 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint

Neutrophil extracellular traps induce tissue-invasive monocytes in granulomatosis with 637 polyangiitis

Epidemiological description, case-fatality rate, and trends of Hantavirus Pulmonary 640 Syndrome: 9 years of surveillance in Argentina

xCell: digitally portraying the tissue cellular 642 heterogeneity landscape

A Review of MicroRNA Biomarkers in Traumatic Brain Injury

Neurological manifestations in COVID-19 caused by SARS-CoV-2. CNS

Evidence of the COVID-19 Virus 648 Targeting the CNS: Tissue Distribution, Host-Virus Interaction

The 654 pathogenesis of Ebola virus disease

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Severe respiratory illness caused by a novel coronavirus

Imbalanced host response to SARS-CoV-2 drives development of COVID-19

Contribution of cytokines to 663 tissue damage during human respiratory syncytial virus infection

Antigen receptor repertoire profiling from RNA-seq data

MiXCR: software for comprehensive adaptive immunity 670 profiling

Characteristics of Liver Tests 672 in COVID-19 Patients

COVID-19: immunopathology and its implications for therapy

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Specific ACE2 expression in 676 cholangiocytes may cause liver damage after 2019-nCoV infection

T cell-mediated immune response to 679 respiratory coronaviruses

Epidemiological and 681 clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, 682 China: a descriptive study. The Lancet

The spatial and cell-type 685 distribution of SARS-CoV-2 receptor ACE2 in human and mouse brain

fastp: an ultra-fast all-in-one FASTQ 688 preprocessor

Potential roles and targeted therapy of 690 the CXCLs/CXCR2 axis in cancer and inflammatory diseases

p38 mitogen-activated protein kinase pathways in asthma and COPD

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Dynamic balance of pro-and anti-inflammatory signals controls disease and limits 696 pathology

Origin and evolution of pathogenic coronaviruses

Hypoxia-induced endocytosis of Na, K-ATPase in alveolar epithelial cells is 701 mediated by mitochondrial reactive oxygen species and PKC-ζ

DIA-NN: 704 neural networks and interference correction enable deep proteome coverage in high 705 throughput

Duration for Carrying SARS-CoV-2 707 in COVID-19 Patients

ACE2 expression in kidney and testis 709 may cause kidney and testis damage after 2019-nCoV infection

Firmiana: towards a 712 one-stop proteomic cloud platform for data processing and analysis

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Heparin-binding protein (HBP/CAP37): a missing link in 716 neutrophil-evoked alteration of vascular permeability

Large Differences in Small RNA Composition Between Human Biofluids

Identification of a microRNA signature of renal ischemia reperfusion injury

Clinical 724 characteristics of coronavirus disease 2019 in China

S100A8/A9: a mediator of severe asthma 727 pathogenesis and morbidity? Can

Tissue 729 distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first step 730 in understanding SARS pathogenesis

Clinical features of 732 patients infected with 2019 novel coronavirus in Wuhan

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Angiotensin-converting enzyme 2 and angiotensin 1-7: novel therapeutic targets

Eliminating the dication-induced 738 intersample chemical-shift variations for NMR-based biofluid metabonomic analysis

Quantitative Lipoprotein Subclass and Low Molecular Weight Metabolite Analysis in

Human Serum and Plasma by (1)H NMR Spectroscopy in a Multilaboratory Trial

MSFragger: ultrafast and comprehensive peptide identification in mass 746 spectrometry-based proteomics

Neutrophils: between host defence, immune modulation, and tissue injury

A 751 novel coronavirus associated with severe acute respiratory syndrome

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Molecular signatures of antibody responses derived from a systems biology 755 study of five human vaccines

Hypothesis for potential pathogenesis of 757 SARS-CoV-2 infection--a review of immune changes in patients with viral 758 pneumonia

Neurologic 760 Manifestations of Hospitalized Patients With Coronavirus Disease

Matrix

Asthma: Effect of Asthma Severity, Allergen 766 Challenge, and Inhaled Corticosteroids

Janeway's immunobiology

The host immune response in 769 respiratory virus infection: balancing virus clearance and immunopathology

Mast 772 cells increase vascular permeability by heparin-initiated bradykinin formation in vivo

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Transcript-level 775 expression analysis of RNA-seq experiments with HISAT, StringTie and Ballgown

Dysregulation of immune 778 response in patients with COVID-19 in Wuhan, China

Comprehensive Analytic Platform for Extracellular RNA Profiling

Unique human immune signature of Ebola virus disease in Guinea

MicroArray Quality Control (MAQC)-II study of common practices for the development 788 and validation of microarray-based predictive models

MicroArray Quality Control (MAQC) project shows inter-and intraplatform 791 reproducibility of gene expression measurements

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. novel potential biomarker for early diagnosis of acute myocardial infarction in humans. 836 Eur Heart J 31: 659-666. 837 All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint 2 (C) Summary of differentially expressed genes, proteins, metabolites and exRNAs between 7 uninfected controls and COVID-19 patients (mild and severe) in the multi-omics data.

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. 

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. 

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. 

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint 7 and severe patient groups. The fold changes in tissue-enhanced proteins between 57 mild/severe patient samples and control samples are shown on the right of heatmap.

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. 

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. 

(E) T cell and innate immune response elucidate immunopathology of COVID-19.

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. 

(C) COVID-19 severity is associated with significant changes in lipoprotein subclasses 82 All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted June 22, 2020. . https://doi.org/10.1101/2020.06.19.20134379 doi: medRxiv preprint