key: cord-0977212-ejfftwmi
authors: Lee, Hye Kyung; Knabl, Ludwig; Pipperger, Lisa; Volland, Andre; Furth, Priscilla; Kang, Keunsoo; Smith, Harold; Knabl, Ludwig; Bellmann, Romuald; Bernhard, Christina; Kaiser, Norbert; Gänzer, Hannes; Ströhle, Mathias; Walser, Andreas; Laer, Dorothee Von; Hennighausen, Lothar
title: Immune transcriptomes of highly exposed SARS-CoV-2 asymptomatic seropositive versus seronegative individuals from the Ischgl community
date: 2020-09-23
journal: Res Sq
DOI: 10.21203/rs.3.rs-69657/v1
sha: db2b371302a9b93710fd1a6389ba3802a7c07bdc
doc_id: 977212
cord_uid: ejfftwmi

To investigate prevalence of ongoing activation of inflammation following asymptomatic SARS-CoV-2 infection we characterized immune cell transcriptomes from 43 asymptomatic seropositive and 52 highly exposed seronegative individuals with few underlying health issues following a community superspreading event. Four mildly symptomatic seropositive individuals examined three weeks after infection as positive controls demonstrated immunological activation. Approximately four to six weeks following the event, the two asymptomatic groups showed no significant differences. Two seropositive patients with underlying genetic disease impacting immunological activation were included (Cystic Fibrosis (CF), Nuclear factor-kappa B Essential Modulator (NEMO) deficiency). CF, but not NEMO, associated with significant immune transcriptome differences including some associated with severe SARS-CoV-2 infection (IL1B, IL17A, respective receptors). All subjects remained in their usual state of health from event through five-month follow-up. Here, asymptomatic infection resolved without evidence of prolonged immunological activation. Inclusion of subjects with underlying genetic disease illustrated the pathophysiological importance of context on impact of immunological response.

To investigate prevalence of ongoing activation of in ammation following asymptomatic SARS-CoV-2 infection we characterized immune cell transcriptomes from 43 asymptomatic seropositive and 52 highly exposed seronegative individuals with few underlying health issues following a community superspreading event. Four mildly symptomatic seropositive individuals examined three weeks after infection as positive controls demonstrated immunological activation. Approximately four to six weeks following the event, the two asymptomatic groups showed no signi cant differences. Two seropositive patients with underlying genetic disease impacting immunological activation were included (Cystic Fibrosis (CF), Nuclear factor-kappa B Essential Modulator (NEMO) de ciency). CF, but not NEMO, associated with signi cant immune transcriptome differences including some associated with severe SARS-CoV-2 infection (IL1B, IL17A, respective receptors). All subjects remained in their usual state of health from event through ve-month follow-up. Here, asymptomatic infection resolved without evidence of prolonged immunological activation. Inclusion of subjects with underlying genetic disease illustrated the pathophysiological importance of context on impact of immunological response. On occasion COVID-19 patients can suffer from longer term sequelae. In one survey, thirty-ve percent of patients with mild outpatient-treated disease were reported as having not returned to their usual state of health by two to three weeks following infection 11 . Prolonged myocardial in ammation 12 and subacute thyroiditis 13 post resolution of acute infection have been reported. Although it is increasingly being recognized that up to 96% of infected individuals are asymptomatic 14, 15 , their immune responses and the prevalence of unrecognized ongoing in ammation have not been investigated and are not understood.

With the pandemic extending itself into populations with underlying genetic disease, there is an urgent need to understand their immune response to SARS-CoV-2 infections. Towards this end we determined the immune transcriptome of an asymptomatic seropositive patient with Cystic Fibrosis (CF) and one with Nuclear factor-kappa B Essential Modulator (NEMO) de ciency within our study population.

In an attempt to de ne the immune response in asymptomatic SARS-CoV-2 seropositive and highly exposed seronegative individuals within an isolated population, we conducted a study on residents from the ski resort of Ischgl that experienced a superspreading event in early March of 2020. This explosive local outbreak led to the spread of the virus throughout Austria and to many other European countries and worldwide 16 . Ischgl and the Paznaun valley were quarantined on March 13, 2020 and remained under lockdown for six weeks. An epidemiologic study targeting 79% (n=1473) of the population of Ischgl (n= 1867) was conducted between April 21 and 27 and revealed a seroprevalence of approximately 42% (Knabl et al., High SARS-CoV-2 Seroprevalence in Children and Adults in the Austrian Ski Resort Ischgl, submitted) with approximately 17% of these being asymptomatic. Our study encompassed 43 seropositive asymptomatic individuals (Group A) and 52 highly exposed seronegative individuals (Group B) with an equal gender and age distribution ( Fig. 1a and Table 1 ). Six households had both seropositive asymptomatic and highly exposed seronegative members (Supplementary Table 1 ). Only a few asymptomatic seropositive individuals had conditions that increased their risk of severe illness from COVID-19 17 , one patient with Cystic Fibrosis (CFTR G551D mutation) and one with Nuclear factor-kappa B Essential Modulator (NEMO) de ciency (Incontinentia pigmenti, IKBKG exon4_10del mutation) ( Table  1 ). All infected individuals in the study remained asymptomatic throughout their infection and were con rmed in in person interviews as having their usual state of health at the time of the study (4-6 weeks after the infection) by telephone and remained so at the second phone call ve months following the outbreak.

To validate that RNA sequencing (RNA-seq) performed on RNA extracted from peripheral blood mononuclear cells (PBMCs) could be used to identify gene expression changes in patients following SARS-CoV-2 infection, we conducted an unbiased RNA-seq analysis on PBMCs from four patients with mild symptoms (Group D), about three weeks after they had tested PCR-positive. These were compared to four highly exposed seronegative individuals (Group E) ( Fig. 1b and Supplementary Table 2) , who cohabitated with one of the symptomatic patients. Among the 175 genes whose expression was signi cantly elevated in PBMCs from symptomatic patients, genes were signi cantly enriched in 16

Hallmark gene sets (Supplementary Table 2 ), ve of which have de ned roles in immune regulation: TNFa/NFkB, mTORC1 signaling, IL2-STAT5 signaling, TGFb signaling and in ammatory response (Fig. 2 and  Supplementary Table 2 ). This demonstrates that the RNA-seq approach utilizing buffy-coat-isolated PBMCs identi ed statistically signi cant differences in in ammatory gene expression between infected and non-infected individuals. Notably, we observed signi cantly elevated expression of IL10 in the symptomatic patients, a cytokine whose elevated expression has been associated with disease severity 18 .

We then conducted an unbiased RNA-seq transcriptome analysis (Supplementary Table 3 ) and plasma cytokine pro ling ( Fig. 3 and Supplementary Table 4 ) from the asymptomatic seropositive (Group A) and highly exposed seronegative (Group B) Ischgl resident cohorts. Very few statistically signi cant changes in gene expression were found (11 induced, 7 down-regulated) (Supplementary Table 3 ). Quanti cation of in ammatory cytokines and chemokines revealed no signi cant differences in protein levels between Group A and B ( Fig. 3 and Supplementary Table 4 ).

The PBMC transcriptome data from the CF (Supplementary Table 5 ) and NEMO (Supplementary Table 6 ) patients in Group A were analyzed separately. While the NEMO patient showed no statistically signi cant differences as compared to Group A, the CF patient demonstrated statistically signi cant differences in expression for approximately 4670 genes in comparison with Group A. Overall, expression levels of 3020 genes were signi cantly higher and expression levels of 1648 genes were signi cantly lower (Supplementary Table 5 ). Genes were enriched in 32 Hallmark gene sets, 11 of which have de ned roles in immune regulation (Fig. 4a) . Expression of key immune signature genes, including interferon response genes, IL1B, IL17A and their respective receptors, and JAK-STAT pathway genes, were signi cantly induced ( Fig. 4b-d ).

The ability to assess the immune response of both symptomatic and asymptomatic SARS-CoV-2 infected individuals might provide critical information on dysregulated immune-response signatures that could foretell disease trajectories. While longitudinal studies on hospitalized patients demonstrate elevated levels of pro-in ammatory cytokines and signatures associated with ongoing in ammation 3 , there is a parallel need to pinpoint the immune response of infected, yet asymptomatic individuals. With the exception of a CF patient, we did not detect an aberrant immune transcriptome in 41 asymptomatic seropositive individuals that were infected during a super spreading event as compared to 52 highly exposed seronegative individuals from the same community. Several households housed both seropositive and seronegative individuals. In contrast to the seropositive asymptomatic individuals, a proin ammatory immune signature was detected in patients with mild symptoms. These results demonstrate that development of an antibody response to COVID-19 following viral exposure and seroconversion in asymptomatic cases is not necessarily associated with sustained alterations in the immune system transcriptome.

While other studies have focused on single cells RNA-seq (scRNA-seq) pro ling of small numbers of individuals with COVID-19 disease [4] [5] [6] [7] 9, 10 , the use of RNA-seq from mononuclear cells permitted us to analyze larger cohorts at a great depth. This approach was validated through the detection of in ammatory immune signatures in COVID-19 patients exhibiting mild symptoms and one asymptomatic seropositive CF patient.

Patients with CF manifest cytokine dysfunction and hyperin ammation that overlaps with the pathophysiology of COVID-19 19 . While there are limited data on the immune response of CF patients to COVID-19 infection, preliminary information suggests that the course of disease may be milder than expected 20, 21 . The immune transcriptome of the asymptomatic seropositive CF patient provided evidence of highly activated cytokine signaling pathways. Expression of key components of interferon, interleukin and JAK-STAT pathways is highly elevated. Many of these genes, such as IL1B and its receptors are also highly activated in COVID-19 patients 9 . In contrast, the immune transcriptome of a SRAS-CoV-2 infected asymptomatic patient with a NEMO de ciency syndrome, a rare primary immunode ciency, was indistinguishable from asymptomatic seropositive controls. It remains to be understood how elevated cytokine signaling, documented here in an asymptomatic CF patient, contributes to disease progression in non-CF patients and why CF patients, with chronic high expression, do not invariably experience disease progression. A critical question for development of targeted therapies is to discern between direct pathogenic immune determinants of severe disease and correlates of in ammation.

Limitations of our study include the translatability of our ndings to other populations. There were few underlying health issues in this rural alpine population living at an altitude of 1,400 meters. For example, obesity, which is associated with an in ammatory state and is recognized as risk factor for severe COVID-19 disease 22, 23 , was less than 10% in our study population, differing greatly from higher prevalence rate in other infected populations. Similarly, other de ned risk factors for severe disease such as diabetes and chronic kidney disease, were also comparatively low. Extraction of the buffy coat and puri cation of RNA. Extraction of the buffy coat and subsequent RNA puri cation will be performed as described 10 . In short, the drawn blood is centrifuged at 1,600g for 10 min at 4°C. After vacuming off the plasma layer, the buffy coat layer is carefully collected. The obtained buffy coat is mixed with 1 mL RBC lysis buffer and incubated for 10 min at room temperature. After a centrifugation step, supernatants were discarded, and the pellets mixed with 1 mL RBC lysis buffer. The pellet is washed with PBS buffer and then mixed with 1 mL TRIzol ® . For extraction of the RNA the TRIzol reagent single-step method will be used 11 . After addition of 2 M sodium acetate (pH 4), the tube is mixed thoroughly by invertion before adding chloroform/isoamyl alcohol (49:1) and another mixing step. The sample gets incubated on ice for 15 min and subsequently centrifuged for 20 min at 10,000g and 4°C. The obtained aqueous phase is transferred to a new Eppendorf tube, 1 mL isopropanol is then added, which is followed by an incubation at -20°C for 1 h to precipitate RNA. The nal RNA precipitate is won by centrifugation at 10,000g at 4°C and discarding of the supernatant. Pellets are stored at -80°C until dispatch.

mRNA sequencing (mRNA-seq) and data analysis. Nanophotometer (Implen) was used to analyze each sample for concentration and the RNA quality was assessed by an Agilent Bioanalyzer 2100 (Agilent Technologies). The Poly-A containing mRNA is puri ed by poly-T oligo hybridization from 1 μg of total RNAs and cDNA was synthesized using SuperScript III (Invitrogen). Libraries for sequencing were prepared according to the manufacturer's instructions with TruSeq Stranded mRNA Library Prep Kit (Illumina, RS-20020595) and paired-end sequencing was done with a NovaSeq 6000 instrument (Illumina). mRNA-seq read quality control was done using Trimmomatic 24 (version 0.36) and STAR RNA-seq 25 (version STAR 2.5.4a) using 150bp paired-end mode was used to align the reads (hg19). HTSeq 26 was to retrieve the raw counts and subsequently, R (https://www.R-project.org/), Bioconductor 27 and DESeq2 28 were used. Additionally, the RUVSeq 29 package was applied to remove confounding factors. The data were pre-ltered keeping only those genes, which have at least ten reads in total. Genes were categorized as signi cantly differentially expressed with an adjusted p-value (pAdj) below 0.05 and a fold change > 2 for up-regulated genes and a fold change of < -2 for down-regulated ones. The visualization was done using dplyr (https://CRAN.R-project.org/package=dplyr) and ggplot2 30 . The genes were cutoff in the standard, less than 5 value and less than 1 log 2 fold change and then conducted gene enrichment analysis (https://www.gsea-msigdb.org/gsea/msigdb).

Statistical analysis. For comparison of protein levels, single sample pairs were evaluated with the unpaired two-tailed t-test to compare the distributions of two groups with the Welch's t test to compare the two distributions (GraphPad PRISM version 8.2.0). For comparison of RNA expression levels between CF patient and asymptomatic seropositive cohort, a two-way ANOVA followed by Tukey's multiple comparisons test was used (GraphPad PRISM). P<0.05 was considered statistically signi cant.

Data availability. The RNA-seq data for patients will be uploaded in GEO.

Declarations 27. Huber, W. et al. Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods 12, 115-21 (2015 Study Design. a In the Ischgl community, 43 seropositive asymptomatic and 52 highly exposed seronegative individuals underwent phlebotomy and health evaluation between April 21 and 27, 2020 with a follow up health evaluation mid-August, 2020. The SARS-CoV-2 superspreading event occurred in the community between the end of February and March 13 of 2020 when the town was quarantined. b

The non-Ischgl cohorts were recruited from other parts of Tyrol. Four SARS-CoV-2 PCR positive patients exhibiting mild symptoms who underwent phlebotomy and health evaluation approximately three weeks after being con rmed as PCR positive for SARS-CoV-2 and four highly exposed seronegative individuals, from the same household as one of the symptomatic patients were included. Group D, n=4) and non-infected (green dots, Group E, n=4) individuals. Mean indicated. *Padj < 0.04, **Padj < 0.001, ***Padj < 0.0001, ****Padj < 0.00001 (DESeq2).

Comparison of cytokine and chemokine levels between asymptomatic seropositive and highly exposed seronegative cohorts. No statistically signi cant differences in relative steady state protein levels of 28 cytokines and 10 chemokines between asymptomatic seropositive (red, Group A, n=41) and highly exposed seronegative (blue, Group B, n=52) were found. Relative protein levels measured using LEGENDplex assays. Boxplots show median (middle bar), interquartile range (IQR) (box), 1.5 X IQR (whiskers). 

SARS-CoV-2: a storm is raging

COVID-19 cytokine storm: the interplay between in ammation and coagulation

Longitudinal analyses reveal immunological mis ring in severe COVID-19

Heightened Innate Immune Responses in the Respiratory Tract of COVID-19 Patients

A single-cell atlas of the peripheral immune response in patients with severe COVID-19

Deep immune pro ling of COVID-19 patients reveals distinct immunotypes with therapeutic implications

Imbalanced Host Response to SARS-CoV-2 Drives Development of COVID-19

COVID-19 severity correlates with airway epithelium-immune cell interactions identi ed by single-cell analysis

Immunophenotyping of COVID-19 and in uenza highlights the role of type I interferons in development of severe COVID-19

Single-cell landscape of bronchoalveolar immune cells in patients with COVID-19

Symptom Duration and Risk Factors for Delayed Return to Usual Health Among Outpatients with COVID-19 in a Multistate Health Care Systems Network -United States

Outcomes of Cardiovascular Magnetic Resonance Imaging in Patients Recently Recovered From Coronavirus Disease 2019 (COVID-19)

Subacute thyroiditis in a patient infected with SARS-COV-2: an endocrine complication linked to the COVID-19 pandemic

Clinical Course and Molecular Viral Shedding Among Asymptomatic and Symptomatic Patients With SARS-CoV-2 Infection in a Community Treatment Center in the Republic of Korea

Proportion of asymptomatic coronavirus disease 2019 (COVID-19): a systematic review and meta-analysis

Longitudinal COVID-19 pro ling associates IL-1RA and IL-10 with disease severity and RANTES with mild disease

COVID-19 meets Cystic Fibrosis: for better or worse

A multinational report to characterise SARS-CoV-2 infection in people with cystic brosis

Impact of SARS-CoV-2 infection in patients with cystic brosis in Spain: Incidence and results of the national CF-COVID19-Spain survey

Obesity and impaired metabolic health in patients with COVID-19

Individuals with obesity and COVID-19: A global perspective on the epidemiology and biological relationships

Trimmomatic: a exible trimmer for Illumina sequence data

STAR: ultrafast universal RNA-seq aligner

HTSeq--a Python framework to work with high-throughput sequencing data

This is a list of supplementary les associated with this preprint. Click to download

Our gratitude goes to the individuals from the Ischgl community who contributed to this study to advance our understanding of COVID-19. This work utilized the computational resources of the NIH HPC Biowulf cluster (http://hpc.nih.gov). NGS was conducted in the NIH Intramural Sequencing Center, NISC (https://www.nisc.nih.gov/contact.htm).

This work was supported by the Intramural Research Program (IRP) of National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK).

Study design: LH, LK and HKL; patient recruitment and care: LK, LK Sr, RB, SB, NK, HG, MS, AW, DvL; RNA preparation: AV; cytokine assays: LP; RNA-seq analysis: HKL; Transcriptome analysis and data interpretation: HKL, LK, LH and PAF; manuscript generation; HKL, LK, PAF and LH. All authors approved the nal version.

The authors declare no competing nancial interests.