Molecular Mechanisms of Covid-19 in the Lungs and Heart: Insights from Spatial Transcriptomics

Introduction: Spatial transcriptomics has become a powerful tool for interrogating a disease-specific transcriptome within the context of tissue architecture. In this study, we used spatial transcriptomics to investigate the molecular mechanisms underlying COVID-19 in the lungs and heart. Methods: We performed sequencing-based spatial transcriptomics (ST), using 10X Genomics’ Visium, on 6 paired fatal COVID-19 lung and heart tissue samples and 2 control samples per organ. With the gene by spot matrix, we performed dimensional reduction, clustering (Louvain), differential expression (Wilcoxon Rank sum test), and pathway analysis (GSEA). spacexr was used to estimate the cell type composition of ST spots. Results: By histology and ST, COVID-19 tissue was defined by a loss in parenchymal cells and increased inflammation ( fig. 1E ). In both the heart and lung, we found a cluster of ST spots specific to fatal COVID-19 ( fig. 1A, B ; P<0.001). Between the two organs, these clusters shared genes and pathways relating to tissue remodeling, B cell action, and complement pathway activation ( fig. 1C, D ). Response to wounding and blood vessel endothelial cell migration pathways were distinct to the heart cluster, also explaining the increased capillary cell weight in the COVID-19 heart compared to the control ( fig. 1D, E ; P=0.0167). Conclusion: Our results suggest that there is a shared spatial niche between the heart and lungs in COVID-19 infection and highlight the importance of studying multiple organs in understanding the disease. These findings provide new insights into the molecular basis of COVID-19 and have the potential to inform the development of novel therapies for this disease. Figure 1 (A) UMAP colored by (left) clusters and (right) disease. (B) Cluster proportions between COVID-19 and control tissue (C) Scatter plot of shared differentially expressed genes. (D) GSEA pathway score. (E) Spatial mapping of clusters and cell-type deconvolution.