Reply to Khatun et al.

From the Authors:

We thank Khatun and colleagues for their insightful comments on our recent manuscript describing the in vitro effects of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and of coronavirus disease (COVID-19) patient sera on human endothelial cells (1). We are pleased that the authors found our proposed model of COVID-19–associated endotheliopathy to be consistent with their published data (2, 3). We apologize for not citing these important publications. In our study, we detected SARS-CoV-2 RNA in lysates of primary human lung microvascular endothelial cells (HMVEC) treated with live SARS-CoV-2 and observed that exposure to live virus induced an increase in permeability and activation of HMVEC. We also found increased permeability of HMVEC treated with sera from COVID-19 patients that correlated with disease severity. We postulated that COVID-19–associated endotheliopathy results from a combination of the actions of systemic and localized inflammatory mediators and cells on the endothelium, and by SARS-CoV-2 infection of the endothelium. Using a model of severe COVID-19 in K18-hACE2 mice, the authors previously reported SARS-CoV-2 in lung capillary endothelial cells in both the early and delayed stages of severe COVID-19, as well as lung edema, perivascular inflammation, upregulated adhesion molecule expression, and decreased vascular endothelial–cadherin expression (2, 3). Similarly, they observed co-localization of SARS-CoV-2 viral proteins with the endothelial marker CD31 in nonhuman primates with severe COVID-19, and also in autopsy samples from a patient that died of COVID-19. In their comment, the authors present additional single-cell RNAseq data on samples from SARS-CoV-2–infected mice. Their data confirm the presence of viral RNA within endothelial cells and demonstrate increased expression of Ifi27l2a and Irf7, and decreased expression of Cldn5 by infected endothelial cells. These new data further support the concept that SARS-CoV-2 infection of endothelial cells may directly promote endothelial barrier dysfunction, and they suggest that viral infection of endothelial cells contributes to the COVID-19–associated endotheliopathy. An interesting aspect of the authors’ studies is that K18-hACE2 mice have very low ACE2 expression by endothelial cells. This suggests that ACE2 may not be required for SARS-CoV-2 entry into endothelial cells and supports a model of viral transduction of endothelial cells described by the authors.

We believe that the results of our clinical-translational study and the authors’ basic-translational work support our proposed multicomponent model of COVID-19–associated endotheliopathy. Furthermore, there is encouraging consistency in the effects of SARS-CoV-2 on cultured human endothelial cells (1) and in vivo on mouse endothelial cells (2, 3). The concordance in endothelial infection and dysfunction between the humans and mice suggests shared mechanisms of COVID-19–associated endotheliopathy. It also indicates that mouse models may lead to insights into processes that cannot be studied longitudinally in humans, such as how COVID-19 affects fixed cells in organs and tissue, including endothelial cells, tissue leukocytes and neurons, which are inaccessible in real-time in humans. Also, the apparent concordance between human and mouse endothelial cells may facilitate complementary preclinical testing of potential therapies using cultured human endothelial cells and mouse models, and even facilitate studies on effects of COVID-19 on endothelium that might predispose to the chronic effects of SARS-CoV-2 infection on the lung and other organs.

The available data suggest that the etiology of COVID-19–associated endotheliopathy involves viral infection of the endothelium, as well as the actions of systemic and localized mediators and cells on the endothelium. More work is needed to understand the mechanisms, and the relative importance of endothelial infection and of systemic and localized inflammatory cells and mediators in COVID-19. Although the specific roles of these factors in driving COVID-19–associated endotheliopathy and respiratory failure remain to be determined, collectively the data suggest that the endothelium may represent an important but understudied therapeutic target in COVID-19.