SARS-CoV2 Endotheliopathy: Insights from Single Cell RNAseq

To the Editor:

We read with interest the study by Joffre and colleagues, “COVID-19-associated Lung Microvascular Endotheliopathy: A “From the Bench” Perspective” (1). The authors propose a model of endotheliopathy due to both endothelial cell severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection as well as inflammatory cytokines that likely contribute to endothelial dysfunction. In support of this model, we have previously documented that both SARS-CoV2 endothelial cell infection and the infection-associated immune activation contribute to endothelial injury and dysfunction in SARS-CoV-2–infected K18-hACE2 mice, as well as our non-human primate model and patients with severe coronavirus disease (COVID-19) (2, 3). Here, we further mined our single-cell RNAseq data from the K18-hACE2 SARS-CoV2 infected mice and independently validated viral RNA in pulmonary microvascular endothelial cells in vivo.

Importantly, endothelial cells with viral RNA (defined by ORF10 expression) showed increased Ifi27l2a, Irf7, and reduced Cldn5 (Figure 1) compared with viral RNA negative endothelial cells in the same animals. Interestingly, in the K18-hACE2 model, these cells express very low levels of hACE2 compared with lung epithelial cells such as club and alveolar type II (AT2) cells (2). Increased Ifi2712a (or interferon, α-inducible protein 27 like 2A gene) is indicative for a type I interferon (IFN-1) response (4). Increased Irf7 (or interferon regulation factor 7) is implicative of the upregulation of IFN-1 dependent immune responses (5). Downregulation of Cldn5 (a critical component of endothelial tight junctions) is indicative of endothelial cell dysfunction. These data suggest that the transcriptional changes are due to the presence of viral RNA in the endothelial cells and the presence of the viral RNA is directly associated with endothelial dysfunction. We did not detect sub-genomic RNA in the endothelial cluster, which is a splice variant seen in cells that are actively replicating virus (2). Taken together these data, along with a very low level of hACE2 expression, support a model of viral transduction of endothelial cells, perhaps by exosomes or non-receptor mediated uptake of viral RNA. Further, our in vivo data support the in vitro model reported by Joffre and colleagues.

Figure 1.

TSNE plots of whole lung single cell RNAseq data in SARS-CoV-2 infected K18-hACE2 mice (Day 4 post infection). Endothelial cells were reclustered and then separated by the presence or absence of viral RNA, defined by orf10 expression. Irf7 and Cldn5 expression were differentially expressed in orf10 + endothelial cells. Quantitative expression data are in the table above. adj = adjusted; SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; TSNE = t-distributed stochastic neighbor embedding.