Tissue banking from patients withSARS-CoV-2 (COVID-19) infection

ABSTRACT

The clinical spectrum of SARS-CoV-2 (COVID-19) infection ranges from asymptomatic infection to fatal pneumonia, but the determinants of outcome are not well understood. To characterize the immune response to COVID-19, weestablished a protocol to collect biologic specimens from patients with confirmed or suspected COVID-19. BetweenApril 9th and June 8th, 2020, we enrolled 146 inpatients and 169 outpatients at the University of Chicago. Wehypothesized that the complex interplay of viral, environmental, and host genetic factors may influence diseaseseverity in patients with COVID-19. To probe for genetic predispositions that may influence outcomes, we collectedgermline DNA from 140 patients spanning the breadth of clinical severity, which will be sequenced for SNPs ingenes previously implicated in immune responsiveness and ARDS. To determine whether a pattern of commensalbacteria correlates with disease severity, we will analyze the composition of airway microbiota from 226nasopharyngeal swabs, using viral quantification and 16S sequencing. Longitudinal serum samples from 156patients were obtained to probe for the presence of antibodies using an ELISA against the spike protein of SARS-CoV-2. In tandem, 36-color flow cytometry on PBMCs, from the same patients, will characterize immune cellphenotypes influenced by infection. We also hypothesized that by characterizing mechanisms of immune-hyperresponsiveness, we may elucidate key biologic pathways that inform the development of novel therapeutics.To determine if severity of disease and response to therapy correlates with soluble factors, we are performing 44-plex cytokine Luminex assays on serum samples. We will probe the adaptive immune response using an ELISAagainst the SARS-CoV-2 RBD domain, and by performing IFN-g ELISPOT analysis against peptide pools fromSARS-CoV-2 proteins. We developed a bioinformatic pipeline to integrate clinical data with the results from thediverse data types and will adopt a machine learning approach to identify parameters contributing to diseaseseverity, response to therapies, and outcomes. In establishing this protocol, there were significant biosafetyconsiderations. To limit potential exposure and virus transmission, research coordinators contacted inpatients byphone for an informed consent discussion, and patients completed the consent form electronically using REDCap(n=61). Inpatients who were unable to navigate the electronic consent were visited with a paper consent (n= 85).Samples were processed in a BSL2 laboratory with enhanced biosafety precautions. Where feasible, samples werecollected into reagents such as Zymo DNA/RNA shield to immediately inactivate the virus. Other safety measuresincluded heat inactivation of some samples and use of a laminar flow washer to minimize aerosolization duringFACS staining. In summary, we have established a biorepository of specimens from patients with COVID-19, including a subset with active cancer or a history of the disease (n=22).