key: cord-1031155-vh5ut34u
authors: Lee, Katherine S.; Russ, Brynnan P.; Wong, Ting Y.; Horspool, Alexander M.; Winters, Michael T.; Barbier, Mariette; Bevere, Justin R.; Martinez, Ivan; Damron, F. Heath; Cyphert, Holly A.
title: Diet induced obesity and type 2 diabetes drives exacerbated sex-associated disease profiles in K18-hACE2-mice challenged with SARS-CoV-2
date: 2022-04-28
journal: bioRxiv
DOI: 10.1101/2022.04.26.489580
sha: 3b4ce4453028eb6ce3638b0656859d4e0f6ee72c
doc_id: 1031155
cord_uid: vh5ut34u

SARS-CoV-2 infection results in wide-ranging disease manifestation from asymptomatic to potentially lethal. Infection poses an increased threat of severity to at-risk populations including those with hypertension, diabetes, and obesity. Type 2 Diabetes (T2DM), is characterized, in part, by insulin insensitivity and impaired glucose regulation. T2DM patients have increased disease severity and poorer outcomes with COVID-19. We utilized the diet-induced obesity (DIO) model of Type 2 Diabetes in SARS-CoV-2-susceptible K18-hACE2 transgenic mice to better understand the obesity co-morbidity. Female DIO, but not male DIO mice challenged with SARS-CoV-2 were observed to have shortened time to morbidity compared to normal diet mice. Increase in susceptibility to SARS-CoV2 in female DIO was associated with increased total viral RNA burden compared to male mice. RNAseq analysis was performed on the lungs of non-challenged, challenged, females, males, of either normal diet or DIO cohorts to determine the disease specific transcriptional profiles. DIO female mice had more total activated genes than normal diet mice after challenge; however, male mice experienced a decrease. GO term analysis revealed the DIO condition increased interferon response signatures and interferon gamma production following challenge. Male challenged mice had robust expression of antibody-related genes suggesting antibody producing cell localization in the lung. DIO reduced antibody gene expression in challenged males. Collectively this study establishes a preclinical T2DM/obesity co-morbidity model of COVID-19 in mice where we observed sex and diet specific responses that begin to explain the effects of obesity and diabetes on COVID-19 disease.

Introduction distribution) is accompanied by enlargement of individual adipocytes which become stressed and 78 hypoxic at the cellular level. Chronic exposure to stress signals, hypoxic conditions, and oxidative 79 stress causes adipocytes to produce cytokines like CRP, TNF-alpha, and IL-6 in addition to their 80 healthy secretions intended to maintain homeostasis (20) . The resulting recruitment to and 81 activation of proinflammatory-type macrophages cells within the adipose tissue raises basal 82 inflammation systemically in a phenomenon known as "metabolic inflammation" (21-23). This 

T2DM is estimated to be the second-most common comorbidity in patients with severe COVID-95 19, resulting in a 2-3 times greater likelihood to succumb compared to healthy persons (28). Over 96 460 million people worldwide have been diagnosed with diabetes mellitus (either T1DM or T2DM) 97 and greater than 60% of type 2 diabetics are also clinically characterized as obese (29). SARS-

CoV-2 infection combined with the metabolic dysfunction in T2DM is associated with an increased 99 risk of pneumonia requiring ventilation, ICU admission, and "long COVID" (30). While COVID-19 100 vaccine implementation around the world has been a positive effort for protecting vulnerable 101 populations, T2DM has been linked to reduced COVID-19 vaccine efficacy, with lower IgG and 102 neutralizing antibody development (14, 31) . Because of their predisposed risk to severe outcomes, defining the immunological profile of type 2 diabetics is a necessary step towards solving vaccine-104 established protection discrepancies.

Most of our knowledge regarding the positive correlation between metabolic dysfunction and 106 SARS-CoV-2 severity comes from retrospective clinical studies, where it becomes impossible to 107 discern the molecular mechanisms that governed severe outcomes (32) (33) (34) (35) (36) (37) (38) (39) (40) . In order to identify 108 and characterize the mechanisms behind increased infection and severity we developed a 109 preclinical model of disease comorbidities using the K18-hACE2 transgenic mouse model and antibody-related gene diversity compared to normal diet K18-hACE2 mice in addition to 120 differential gene expression profiles. Our data illustrate how metabolic dysfunction can enhance 121 COVID-19 disease and suggest a synergism between hyperglycemia and gene expression profile 122 changes. This data helps to link molecular alterations with infection severity, thus constructing a 123 profile of potential therapeutic targets for the treatment and prevention of death by COVID-19 124 illness.

Animal, Ethics, Biosafety statement. administration of 25μL dose per nare (50μL total).

Disease scoring of SARS-CoV-2 challenged mice.

Challenged K18-hACE2 mice were evaluated daily through in-person health assessments in the 157 BSL3 facility as well as surveillance using SwifTAG Systems video monitoring. Health 158 assessments of the mice were scored based on the following criteria: weight loss (scale 0-5 (up 159 to 20% weight loss)), appearance (scale 0-2), activity (scale 0-3), eye closure (scale 0-2), and 160 respiration (scale 0-2). All five criteria were scored based off a scaling system where 0 represents 161 no symptoms and the highest number on the scale denotes the most severe phenotype as 162 previously described by our lab (54). Additive health scores of the criteria listed above were 163 assigned to each mouse after evaluation and assessed so that mice scoring 5 or above, 20% 164 weight loss, or significant drops in temperature received immediate euthanasia. Cumulative 165 disease scoring was calculated by adding the disease scores of each mouse within the group on 166 each day. Morbid mice that were euthanized during the study before day 14, retained their disease 167 score for the remainder of the experiment for reporting purposes.

Euthanasia and tissue collection.

Mice were euthanized either due to disease scores or at the end of the experiment with an IP 170 injection of Euthasol (390mg/kg) (Pentobarbital) followed by cardiac puncture as a secondary 171 measure of euthanasia. Each animal was the dissected to collect tissues for pathological analysis 172 as previously described (55)Cardiac blood was collected in BD Microtainer gold serum separator 173 tubes and centrifuged at 15,000 x g for 5 minutes to separate the serum for downstream analysis.

PBS (1mL) was pushed by catheter through the nasal pharynx and collected in a 1.5mL Eppendorf 175 tube for Nasal Wash. For future RNA purification, 500μL of nasal wash was added to 500μL of 176 TRIzol reagent (Zymo Research Catalog No R2050-1-200) and the remainder of the nasal wash 177 was stored for serological analysis. Lung and brain tissues were dissected from each animal. The 178 right lobe of the lung was homogenized in 1mL PBS in gentleMACS C tubes (order number: 130-096-334) using the m_lung_02 program on the gentleMACS Dissociator. For RNA purification 180 from the tissues, lung homogenate (300μL) was added to 1000μL of TRIzol Reagent. For serology 181 and cytokine analysis, 300μL of lung homogenate was centrifuged at 15,000 x g for 5 minutes to 182 separate and collect the supernatant. Brain tissue was also homogenized in 1mL PBS using 183 gentleMACS C tubes and the same setting as lung on the gentleMACS Dissociator. From the 184 homogenate, 500 μL was combined with 1000μL of TRI Reagent for RNA purification.

qPCR SARS-CoV-2 viral copy number analysis of lung, brain, and nasal wash.

From the lung, brain and nasal wash, RNA was purified using the Direct-zol RNA miniprep kit 187 (Zymo Research R2053) and the manufacturer's protocol. qPCR using the Applied Biosystems (Table S1 ). Statistical analysis was performed with the Differential Expression for RNA 229 software compared to the reference set "affy mg u74a" to determine GO terms from gene ontology and biological process databases (FDR ≤ 0.05) (60). GO Term heat maps were generated using 231 Morpheus (61). In order to analyze the expression of the hACE2 transgene, the RNA reads were 232 mapped to the human ACE2 gene (GRCh38). SARS-CoV-2 reads were analyzed by mapping the 233 reads to the SARS-CoV-2 WA-1 reference genome. hACE2 and viral reads were normalized by 234 dividing counts per 50M total reads in each sample. Raw read data is available at NCBI SRA: 235 SUBXXXXXX (submission complete, pending processing).

RNAseq fold change gene expression data was submitted to Ingenuity Pathway analysis using a 238 cut off of P= 0.05. Pathways that were statistically enriched were exported and plotted into heat 239 maps using Morpheus as described above.

Tests to determine statistical significance were performed using GraphPad Prism version 9. In all 242 DIO K18-hACE2 mouse studies n = 5 per group. In challenge dose determination studies, n ≥ 3.

Statistically significant differences between Kaplan-Meyer curves were analyzed using Mantel-

Cox log-rank tests. Student's t-tests were used for comparisons made between two groups. When 245 three or four groups were being compared, statistical differences were assessed using one-way 246 ANOVA with Dunnett's multiple comparisons test or Two-Way ANOVA with Tukey's multiple 247 comparisons test for parametric data. For any non-parametric data, Kruskal-Wallis tests with 248 Dunn's multiple comparisons tests were used.

K18-hACE2 mice develop obesity, metabolic dysfunction, and type two diabetes due to 252 high fat diet.

The COVID-19 pandemic has illustrated that humans respond to infection with a great deal of 

Obese female mice experience greater viral RNA burden in the lungs.

To begin identifying the factors that may contribute to the changes in survival that were observed 292 in DIO mice, we next investigated differences in viral burden measured by qRT-PCR analysis of 293 lung tissue for nucleocapsid transcript copy number. Total lung RNA was isolated from mice at 294 euthanasia at their respective humane endpoints (Fig. 2CD ). Viral RNA burden was found to be 295 higher in the lungs of DIO female mice than in female normal diet controls and no difference was 

Transcriptomic analysis of viral RNA confirms females have increased viral RNA due to 302 DIO condition.

As a secondary method of evaluating viral burden at, total lung RNA was used to perform bulk 304 RNAseq analysis to measure the number of virus gene transcripts per total tissue RNA. Viral

RNA reads were mapped to the SARS-CoV-2 reference genome and represented per 50M 306 illumina reads obtained per sample. Total viral reads perfectly mirrored the nucleocapsid qRT-307 PCR analysis (Fig. 3B) . A 100-fold increase in viral RNA was also observed for DIO females compared to normal females (Fig. 3B) . To identify correlations between viral burden and 309 morbidity, total viral reads were plotted against the day post-challenge that humane euthanasia 310 occurred (Fig. 3C ) or against disease score (Fig. 3D) . High viral reads corresponding to the 311 mortality window 6 days post-challenge were observed for DIO mice where normal diet mice 312 which survived longer underwent euthanasia (due to disease score or planned experimental 313 endpoint) at lower viral burdens. Disease scoring comparisons had greater variance, but DIO 314 mice trended towards having higher viral reads in the lung with higher disease scores (Fig. 3D) . predictor of outcome as well as the memory response that is protective against reinfection (76,77).

Dysregulation of the T cell response early on due to comorbidities may be partially responsible 

To further refine our RNAseq data analysis, Geno Ontology "GO Term" analysis was performed 384 to identify biological processes that were affected in our experiment based on the lung tissues' transcriptional profiles. Although DIO induction and sex resulted in different lists of suggested GO terms, a conservative list of terms was present in each analysis that was related to activation 387 of the immune response in viral infection. The terms were graphed with their corresponding 388 enrichment ratio to compare their relevance in the genetic profiles of each experimental condition 389 (Fig. 7A) . The highest enrichment ratios were seen for the GO terms related to interferon response 390 and were increased in the female DIO challenge sets compared to others. GO terms related to 391 antigen processing and presentation pathways appeared to be absent in both DIO groups. To 392 measure the inflammatory response to virus, of the levels of innate cytokines were measured in 393 serum collected at euthanasia (data not shown). The most striking differences in cytokine 394 production were noted for IFN-γ (Fig. 7B ). In females, DIO induction caused an increase in 395 production of IFN-γ compared to minimal production in female challenge control mice (Fig. 7B ).

DIO induction in males did not result in a significant change in IFN-γ production. This is important 397 as IFN-γ production is triggered during SARS-CoV2 infection and is essential for viral clearance 398 and resolution of the infection(85). Together, these data suggest that changes in the immune 399 response triggered by T2DM/obesity may hinder positive disease outcomes. As shown above, 400 challenged DIO females had 100-fold higher viral RNA burden in the lung (Fig. 3A) and we 401 speculate that the presence of higher levels of viral particles in the lung of DIO females is 402 associated with the increased interferon response observed here.

Sex and metabolic dysfunction influence antibody production

In addition to T cell and interferon response, antibodies also play an important role in the immune on the previously implicated model of diet induced obesity and T2DM (Fig. 1 ). Utilizing our model,

we designed experiments to evaluate sex and DIO as central variables. We utilized the Alpha 435 variant of SARS-CoV-2 which we previously identified as having high virulence compared to 436 ancestral strains (55). K18-hACE2 mice with DIO showed that disease as well as morbidity and mortality were affected very little by the addition of DIO in males; however, the DIO condition 438 greatly affected females (Fig. 2) and resulted in 100-fold higher viral RNA burden in the lung (Fig.   439 3). In order to characterize the augmented host response to viral challenge we utilized RNAseq 440 analysis of the lung tissue to analyze the pathogen-specific airway responses to the presence of 441 virus (Fig. 4) . Pathway analysis using GO term and Ingenuity Pathway analyses illuminated more Pathway, while the addition of the DIO condition enhances many of the pathway's genes (Fig. 6) .

Our data suggest that female K18-hACE2-mice on normal diets have lower viral RNA burden and 451 decreased inflammatory responses, and DIO impairs the clearance of virus in the lung which 452 results in their enhanced morbidity (Fig. 2-3 ). Collectively these data begin to shed light on the 453 effects of DIO on COVID-19.

To the best of our knowledge this is the first study in K18-hACE2-mice to evaluate DIO and sex 455 utilizing transcriptomic analysis to better understand SARS-CoV-2 VOC-specific host responses.

A previous MedRxiv study utilized a mouse-adapted strain of SARS-CoV-2 to challenge DIO 457 C57Bl6 mice to measure the protective efficacy of human convalescent serum treatment (87).

One correspondence describes a small study where DIO mice were challenged with SARS-CoV- were to be studied, we would expect to have either shorter or longer time to morbidity during 472 which host response profiles may further develop or remain hidden due to the disease timeline.

Finally, our study focused on defining transcriptomic responses to characterize the altered host 474 responses to SARS-CoV-2 challenge. We did not analyze specific cell populations through cell 475 isolation and flow cytometry, nor did we evaluate potential mechanisms responsible for this co- (92). We also saw that T cell populations were augmented across our study 493 groups with male DIO challenged mice demonstrating the most noticeable decrease in T cell 494 subtypes. In severe human COVID-19 cases, it has been demonstrated that lymphocytopenia 495 (reduced T cell counts) is probably associated with CD4+ and CD8+ T cell exhaustion (93) (94) (95) (96) (97) (98) (99) .

This may be another contributor towards SARS-CoV-2's severity in T2DM. Lastly, our RNA-Seq 497 data in normal diet challenge mice compared with normal diet infected patient samples also 498 demonstrates overlap in gene expression profiles, highlighting the relevance and use of the k18-499 hACE2 model (100, 101) . We believe that our model demonstrates some of the pathology seen in 500 human cohorts (summarized in Fig. 9 ). This overlap in human and mouse data is encouraging as 501 modeling SAR-CoV-2 in the laboratory setting is critical for understanding the molecular 502 mechanisms at work.

The T2DM-COVID-19 mouse model has allowed us to observe responses to SARS-CoV-2 504 challenge augmented by both obesity and sex. It is apparent that DIO affects female mice and 505 enhances viral virulence; however, it is not clear how to best ameliorate this issue using 506 therapeutic interventions. In a prior study (data unpublished), we evaluated treatment with 507 Baricitinib, a JAK inhibitor that could dampen inflammation caused by SARS-CoV-2. We 508 hypothesized that decreasing inflammation would improve survival outcomes of SARS-CoV-2-509 challenged mice; however, the drug did not improve survival of SARS-CoV-2 challenged mice 510 (data not shown). This data leads us to believe that there are still molecular events that underpin 511 overall inflammation as well as a dynamic viral clearance timeline that need to be adjusted to 512 improve protection. Comparing that study with our DIO model, it is clear we need to profile the 513 cell populations that respond to each phase of viral infiltration and identify if they are ineffective 514 at stopping disease progression. Furthermore, we propose that this DIO model can be used to evaluate differences in vaccine-induced immunity among comorbid groups which is rarely 516 evaluated. We plan to continue our utilization of the novel DIO-COVID-19 mouse model to 517 uncover therapeutic strategies to improve disease and survival outcomes in diverse persons 518 infected with SARS-CoV-2. 

Comorbidity and 609 its Impact on Patients with COVID-19

Comorbidities 612 and mortality in COVID-19 patients

Comorbidities in SARS-CoV-2 patients: A systematic review and meta-analysis. MBio

Control Prev

Sex differences in immune responses

Sex differences 625 in immune responses that underlie COVID-19 disease outcomes

Glucocorticoids, sex hormones, and immunity

Covid-19 kills more men than women: An overview of 631 possible reasons. Front Cardiovasc Med

Adult Obesity Facts | Overweight & Obesity | CDC

The State of Obesity 2018: Better Policies for a Healthier America

Mortality and critical 637 care unit admission associated with the SARS-CoV-2 lineage B.1.1.7 in England: an 638 observational cohort study

Endocrine and metabolic 640 link to coronavirus infection

Available from: 646 /pmc/articles/PMC6824443/ 647 14. Pal R, Bhadada SK, Misra A. COVID-19 vaccination in patients with diabetes mellitus: 648 Current concepts, uncertainties and challenges

Type 2 diabetes 652 complications and comorbidity in Sub-Saharan Africans. EClinicalMedicine

The 655 comorbidity burden of type 2 diabetes mellitus: patterns, clusters and predictions from a 656 large English primary care cohort

Dynamic interplay between metabolic syndrome 660 and immunity

A comprehensive definition for metabolic syndrome

The metabolic syndrome

Hypoxia and adipose tissue function and dysfunction in obesity

Adipokines in inflammation and metabolic 671 disease

Alternative activation of macrophages

Dysfunction: Role of Exercise

Obesity impairs 680 resistance to Leishmania major infection in C57BL/6 mice

Disease Severity in Female Mice Following West Nile Virus Infection. Front Immunol

Diet-Induced Obese Mice Have Increased 687

Mortality and Altered Immune Responses When Infected with Influenza Virus 1

Increased morbidity of 691 obese mice infected with mouse-adapted SARS-CoV-2

Landstra CP, de Koning EJP. COVID-19 and Diabetes: Understanding the 17

Commission on diabetes: using data to transform diabetes care and patient lives

Diabetes and COVID19: a bidirectional relationship

Levels in Both Diabetic and Non-Diabetic Individuals After BNT162b2 mRNA COVID-19

Age and Hospitalization Risk in People With Type 1 Diabetes

Data From the T1D Exchange Surveillance Study

Associations of type 1 712 and type 2 diabetes with COVID-19-related mortality in England: a whole-population 713 study

COVID-19 pandemic, coronaviruses, and diabetes mellitus

Phenotypic 717 characteristics and prognosis of inpatients with COVID-19 and diabetes: the 718 CORONADO study

COVID-19-related mortality in people with type 1 and type 2 diabetes in England: a 721 population-based cohort study

Association of blood glucose control 723 and outcomes in patients with COVID-19 and pre-existing type 2 diabetes

Risk 726 factors, mortality, and cardiovascular outcomes in patients with type 2 diabetes

Clinical characteristics and 729 outcomes of patients with diabetes and COVID-19 in association with glucose-lowering 730 medication

Severe 732 COVID-19 in people with type 1 and type 2 diabetes in Sweden: A nationwide 733 retrospective cohort study. Lancet Reg Heal -Eur

Evaluation of K18-736 hACE2 Mice as a Model of SARS-CoV-2 Infection

Animal models for COVID-19

K18-743 hACE2 mice develop respiratory disease resembling severe COVID-19

SARS-CoV-2 infection 746 of human ACE2-transgenic mice causes severe lung inflammation and impaired function

Transgenic Mouse Model Recapitulates Non-Severe and Severe COVID-1

Response to Infectious Dose of SARS-CoV-2 Virus 2 3. 2021

Human 755 angiotensin-converting enzyme 2 transgenic mice infected with SARS-CoV-2 develop 756 severe and fatal respiratory disease

CoV-2 in Transgenic Mice Expressing Human Angiotensin-Converting Enzyme 2

A Mouse Model of SARS-761

CoV-2 Infection and Pathogenesis

Mice transgenic for human 763 angiotensin-converting enzyme 2 provide a model for SARS coronavirus infection

Characterization of 766 inflammation and insulin resistance in high-fat diet-induced male C57BL/6J mouse model 767 of obesity. Anim Model Exp Med

A Mouse Model of Diet-Induced Obesity Resembling Most Features of 771

Evaluating antibody 774 mediated protection against Alpha, Beta, and Delta SARS-CoV-2 variants

K18-hACE2 transgenic mice

Mediated Protection against Alpha, Beta, and Delta SARS-CoV-2 Variants of Concern in 779

Mediated Protection against Alpha, Beta, and Delta SARS-CoV-2 Variants of Concern in 783

SARS-CoV-2 infection 786 of human ACE2-transgenic mice causes severe lung inflammation and impaired function

Intranasal 789 administration of BReC-CoV-2 COVID-19 vaccine protects K18-hACE2 mice against 790 lethal SARS-CoV-2 challenge. NPJ vaccines

An interactive tool for comparing lists with Venn's diagrams

BioVenn -A web application for the comparison and 795 visualization of biological lists using area-proportional Venn diagrams

WebGestalt 2019: gene set analysis toolkit 798 with revamped UIs and APIs

(alpha) and B.1.351 (beta) variants induce pathogenic patterns in K18-804 hACE2 transgenic mice distinct from early strains

The 806 alpha/b.1.1.7 sars-cov-2 variant exhibits significantly higher affinity for ace-2 and requires 807 lower inoculation doses to cause disease in k18-hace2 mice

(Alpha) variant exhibits increased respiratory replication and shedding in 811 nonhuman primates. Emerg Microbes Infect

Pathogenic 814 and transcriptomic differences of emerging SARS-CoV-2 variants in the Syrian golden 815 hamster model

Low dose inocula of SARS-818

CoV-2 Alpha variant transmits more efficiently than earlier variants in hamsters

Alpha variant exhibits comparable fitness to the D614G strain in a Syrian hamster 823 model

Delta variant induces enhanced pathology and inflammatory responses in K18-hACE2

Intranasal 830 administration of BReC-CoV-2 COVID-19 vaccine protects K18-hACE2 mice against 831 lethal SARS-CoV-2 challenge

npj Vaccines

Differences in Susceptibility to Respiratory and Systemic Pneumococcal Disease in Mice

Sex differences in immune responses

838 Available from: www

Host-Pathogen Transcriptomics by Dual RNA-Seq

Dual-seq transcriptomics 843 reveals the battle for iron during Pseudomonas aeruginosa acute murine pneumonia

In Vivo Transcriptome of Bordetella pertussis during Infection of Mice. mSphere

Characterizing the innate and adaptive responses of immunized mice to Bordetella pertussis infection 852 using in vivo imaging and transcriptomic analysis

Distinctive features 855 of SARS-CoV-2-specific T cells predict recovery from severe COVID-19. Cell Rep

Broad and strong memory CD4+ 858 and CD8+ T cells induced by SARS-CoV-2 in UK convalescent individuals following 859

The nf-κb rela 861 transcription factor is critical for regulatory t cell activation and stability. Front Immunol 862

Alternative p38 activation pathway mediated by T cell receptor-proximal tyrosine 866 kinases

The kinase pdk1 is critical for promoting t 869 follicular helper cell differentiation

Decreased T-cell 871 receptor signaling through CARD11 differentially compromises forkhead box protein 3-872 positive regulatory versus T(H)2 effector cells to cause allergy

STAT3 roles in viral infection: antiviral or proviral

C-C motif chemokine ligand 2 881 (CCL2) mediates acute lung injury induced by lethal influenza H7N9 virus

The dual nature of type I and type II interferons. Front Immunol

Sex Differences 887 in COVID-19: Candidate Pathways, Genetics of ACE2, and Sex Hormones 3 4

The N501Y mutation in SARS-CoV-2 spike leads to morbidity in obese and aged mice 890 and is neutralized by convalescent and post-vaccination human sera Contributed equally. 891

Western diet 893 increases COVID-19 disease severity in the Syrian hamster

Impact of overlapping risks of 896 type 2 diabetes and obesity on coronavirus disease severity in the United States

Oxidative stress and diabetic complications

Identifying molecular insight of synergistic complexities for SARS-CoV-2 infection with pre-existing type 2 904 diabetes

More Severe in Patients With Diabetes? The Role of Angiotensin-Converting Enzyme 908 2, Endothelial Dysfunction and the Immunoinflammatory System. Front Cardiovasc Med

Severely ill COVID-19 patients display impaired exhaustion features in SARS-CoV-2-912 reactive CD8+ T cells. Sci Immunol

Reduction and Functional 915 Exhaustion of T Cells in Patients With Coronavirus Disease 2019 (COVID-19). Front 916 Immunol

Activation or exhaustion of CD8+ T cells in patients with COVID-19

SARS-CoV-2 Induces 921 Lymphocytopenia by Promoting Inflammation and Decimates Secondary Lymphoid 922

Lymphopenia predicts 924 disease severity of COVID-19: a descriptive and predictive study

Impaired immune 928 cell cytotoxicity in severe COVID-19 is IL-6 dependent

Reduction and Functional 931 Exhaustion of T Cells in Patients With Coronavirus Disease

Host 935 transcriptomic profiling of COVID-19 patients with mild, moderate, and severe clinical 936 outcomes

Comprehensive transcriptomic analysis of COVID-19 blood, lung, and airway. Sci 939 Reports

This project was executed with support from the Vaccine Development Center at the West Virginia