Background
COVID-19 has caused a pandemic of respiratory illness that is likely to lead to complexities in treating thoracic malignancies. However, little is known about the host tissue and cellular responses associated with SARS-CoV-2 infection, symptoms, and disease severity.
Methods
Here, we use the Nanostring GeoMX Digital Spatial Profiler (DSP) and CoxMX Spatial Molecular Imager (SMI) technology to determine tissue signatures, and spatially resolved quantitative single-cell proteogenomic changes driven by SARS-CoV-2 infection. For GeoMx, n=10 COVID-19, n=10 pH1N1 and n=5 normal control tissues were compared. For CosMx, n=19 COVID-19 cores in technical replicates, and n=20 normal control tissues were compared. Tissue-based gene signatures were subsequently tested in the peripheral samples from COVID-19 patients.
Results
SARS-CoV-2 viral presence was confirmed by RNAscope and integrated to inform region of interest and cell types involved in infection. Analysis of the Nanostring GeoMx data revealed tissue signatures associated with SARS-CoV-2 infection, including Type 1 IFN, blood coagulation, hypoxia and angiogenesis. Analysis of the Nanostring CosMx data enabled single cell typing and mapping of tissue-specific signatures to cellular compartments of interest (e.g. macrophages, fibroblasts) and investigation of complex cell population heterogeneity and interactions. All these while preserving spatial context and highlighted differential cell type distribution in the lungs of COVID-19 patients compared to non-infected controls. Our tissue-based Type 1 IFN signatures, when tested in the blood, were found to be predictive of disease severity in COVID-19 patients when measured within the first few days of symptom onset.
Conclusion
Here, we’ve used innovative, cutting-edge spatial transcriptomics approaches to delineate tissue signatures and cellular profiles unique to COVID-19 and common across acute respiratory distress syndrome. These data will aid in understanding the proteogenomic landscape of SARS-CoV-2 infected lung tissues. Moreover, the study demonstrates how tissue-based findings can be rapidly developed into signatures tested in non-invasive samples.