SARS-CoV-2 Spike Protein Destabilizes Microvascular Homeostasis

Abstract

SARS-CoV-2 infection can cause compromised respiratory function and thrombotic events. SARS-CoV-2 binds to and mediates downregulation of angiotensin converting enzyme 2 (ACE2) on cells that it infects. Theoretically, diminished enzymatic activity of ACE2 may result in increased concentrations of pro-inflammatory molecules, angiotensin II, and Bradykinin, contributing to SARS-CoV-2 pathology. Using immunofluorescence microscopy of lung tissues from uninfected, and SARS-CoV-2 infected individuals, we find evidence that ACE2 is highly expressed in human pulmonary alveolar epithelial cells and significantly reduced along the alveolar lining of SARS-CoV-2 infected lungs. Ex vivo analyses of primary human cells, indicated that ACE2 is readily detected in pulmonary alveolar epithelial and aortic endothelial cells. Exposure of these cells to spike protein of SARS-CoV-2 was sufficient to reduce ACE2 expression. Moreover, exposure of endothelial cells to spike protein-induced dysfunction, caspase activation, and apoptosis. Exposure of endothelial cells to bradykinin caused calcium signaling and endothelial dysfunction (increased expression of von Willibrand Factor and decreased expression of Krüppel-like Factor 2) but did not adversely affect viability in primary human aortic endothelial cells. Computer-assisted analyses of molecules with potential to bind bradykinin receptor B2 (BKRB2), suggested a potential role for aspirin as a BK antagonist. When tested in our in vitro model, we found evidence that aspirin can blunt cell signaling and endothelial dysfunction caused by bradykinin in these cells. Interference with interactions of spike protein or bradykinin with endothelial cells may serve as an important strategy to stabilize microvascular homeostasis in COVID-19 disease. IMPORTANCE SARS-CoV-2 causes complex effects on microvascular homeostasis that potentially contribute to organ dysfunction and coagulopathies. SARS-CoV-2 binds to, and causes downregulation of angiotensin converting enzyme 2 (ACE2) on cells that it infects. It is thought that reduced ACE2 enzymatic activity can contribute to inflammation and pathology in the lung. Our studies add to this understanding by providing evidence that spike protein alone can mediate adverse effects on vascular cells. Understanding these mechanisms of pathogenesis may provide rationale for interventions that could limit microvascular events associated with SARS-CoV-2 infection.

Keywords: COVID-19; endothelial cells; spike protein.

FIG 1

Surface expression of ACE2 in normal and COVID-19 lung alveoli. (A): Representative DIC (BW), ACE2/DAPI (red/blue) epi-fluorescence microscopy images and mean fluorescence intensity (MFI) of ACE2 from indicated numbers of healthy (A-i), and SARS-CoV-2 (A-ii) lung FFPE sections shown in the upper and lower rows respectively. (B): Quantified MFI data of ACE2 levels in normal and COVID-19 lung alveoli (n = 10 + 4; ***, p < 0.001).

FIG 2

Effects of SARS-CoV-2 spike protein (SpkPr) on primary human aortic endothelial cells (HAoEC) and alveolar epithelial cells (HAEpC). (A-i, ii) Representative images and relative MFI/DAPI data of surface ACE2 on HAoEC and HAEpC, after 24h exposure to SpkPr (10 ng/ml) and respective controls. (B-i) Krüppel-like Factor 2 (KLF2) levels presented as normalized values of KLF2/DAPI fluorescence following 24h exposure of SpkPr (10 ng/ml) and B-ii surface vWF levels following 24h exposure of SpkPr (10 ng/ml) and respective controls—untreated or Den-SpkPr. (B-iii) Control experiment showing nominal surface vWF expression following SpkPr treatment. (n = 300+ cells; NS, not significant; *, P < 0.05; **, P < 0.005; ***, P < 0.001).

FIG 3

SARS-CoV-2 spike protein (SpkPr) induced pan-caspase activation in endothelial cells. (A-i) Representative SR-FLICA (red), DIC, and DAPI (blue) composite images of individual untreated, staurosporin (1 μM positive control), and SpkPr (1, 10, and 100 ng/ml −72h) treated endothelial cells, and (A-ii) respective normalized mean fluorescence intensity values. (B-i) Representative SR-FLICA (red), DIC, and DAPI (blue) composite images of individual untreated, staurosporin (1 μM −1 h positive control), and SpkPr 10 ng/ml (1, 4, and 24h) treated endothelial cells, and (B-ii) respective normalized mean fluorescence intensity values. (n > 100 cells, **, P < 0.005; ***, P < 0.001).

FIG 4

SARS-CoV-2 spike protein (SpkPr) and PARP-1 cleavage in endothelial cells. (A) Representative SR-FLICA (red), actin, and DAPI (blue) composite images of individual untreated, staurosporin (1 μM positive control), and SpkPr (1, 10, and 100 ng/ml) treated endothelial cells for 72 h (Inset: actin disassembly). (B) Representative Western blot data showing cleaved PARP-1(asp214) after 72 h of exposure of recombinant spike protein-Wuhan, & recombinant spike Pr. Active trimer (100 ng/ml, n = 3). (C-i–iv), Representative DIC and DAPI (blue) composite images of individual EC following indicated treatments. Percentage of morphologically apoptotic ECs after indicated treatments. (n > 100 cells, NS, not significant; *, P < 0.05).

FIG 5

Effects of aspirin (Asp) on the binding affinity of BK to BKRB2, and on BK induced calcium signaling in HAoEC. (A i-ii), Ribbon view and binding affinity (ΔG in kcal/mol) of ED3 and ED4 domain of human BKRB2 with aspirin and with BK in the presence of aspirin respectively as computed using AutoDock Vina molecular docking software. (B-i), High resolution DIC, Fluo-3 (green), Mito Deep Red (MitoDR, red) and co-localization images of HAoEC without or with BK exposure for 15 s. (B i-ii)-Fraction of cytoplasmic Fluo-3 (Ca⁺⁺) overlapping MitoDR fluorescence in HAoEC following BK, Asp-BK, Ica, Ica-BK treatment, or respective control conditions (data from 3 independent experiments, *, P < 0.05). (C i-ii) Representative images of HAoEC (left) and MFI values of cytoplasmic vWF in HAoEC and following indicated treatments. (n > 300 cells, *, P < 0.05).

FIG 6

Schematic diagram showing triggering of contact activation system by a negatively charged surface leads to activation of the coagulation cascade and bradykinin (BK) generation. BK binds to the BKRB2 on endothelial cells, increases plasma coagulation parameters secondary to increased alveolar capillary permeability in and extravascular coagulation.