key: cord-0744633-ri0tijtx
authors: Robles, Juan Pablo; Zamora, Magdalena; Adan-Castro, Elva; Siqueiros-Marquez, Lourdes; Martinez de la Escalera, Gonzalo; Clapp, Carmen
title: The spike protein of SARS-CoV-2 induces endothelial inflammation through integrin α5β1 and NF-κB signaling
date: 2022-02-07
journal: J Biol Chem
DOI: 10.1016/j.jbc.2022.101695
sha: a2293ffc7a167096d13a8f17ee9e6efdda46afc4
doc_id: 744633
cord_uid: ri0tijtx

Vascular endothelial cells (ECs) form a critical interface between blood and tissues that maintains whole-body homeostasis. In COVID-19, disruption of the EC barrier results in edema, vascular inflammation, and coagulation, hallmarks of this severe disease. However, the mechanisms by which ECs are dysregulated in COVID-19 are unclear. Here, we show that the spike protein of SARS-CoV-2 alone activates the EC inflammatory phenotype in a manner dependent on integrin ⍺5β1 signaling. Incubation of human umbilical vein ECs with whole spike protein, its receptor-binding domain, or the integrin-binding tripeptide RGD induced the nuclear translocation of NF-κB and subsequent expression of leukocyte adhesion molecules (VCAM1 and ICAM1), coagulation factors (TF and FVIII), proinflammatory cytokines (TNF⍺, IL-1β, and IL-6), and ACE2, as well as the adhesion of peripheral blood leukocytes and hyperpermeability of the EC monolayer. In addition, inhibitors of integrin ⍺5β1 activation prevented these effects. Furthermore, these vascular effects occur in vivo, as revealed by the intravenous administration of spike, which increased expression of ICAM1, VCAM1, CD45, TNFα, IL-1β, and IL-6 in the lung, liver, kidney, and eye, and the intravitreal injection of spike, which disrupted the barrier function of retinal capillaries. We suggest that the spike protein, through its RGD motif in the receptor-binding domain, binds to integrin ⍺5β1 in ECs to activate the NF-κB target gene expression programs responsible for vascular leakage and leukocyte adhesion. These findings uncover a new direct action of SARS-CoV-2 on EC dysfunction and introduce integrin ⍺5β1 as a promising target for treating vascular inflammation in COVID-19.

adhesion in response to spike, the spike receptor-binding domain, and the RGD tripeptide. The effect of TNFα was not modified, confirming that it is integrin independent (21) .

Because integrin α5β1 activates NF-κB in ECs to elicit inflammation (18) , we asked whether the mechanism by which spike promotes leukocyte adhesion involved the activation of NF-κB. NF-κB is a transcriptional factor for numerous genes involved in inflammation that is held inactive in the cytoplasm by the interaction with proteins known as inhibitors of kappa B (IκB). The phosphorylation/degradation of IκB is needed for the nuclear translocation of NF-κB and its binding to promoter and enhancer regions of target genes (22) . The NF-κB cellular distribution was studied in HUVEC using fluorescence cytochemistry and a monoclonal antibody against the p65 subunit of NF-κB (Figure 3a ). In the absence of treatment, p65 was homogeneously distributed throughout the cytoplasm of cells. Spike induced the accumulation of p65 in the cell nucleus, and this redistribution was like the one induced by TNFα. Anti-α5 antibodies blocked the spike-induced nuclear localization of p65, but not in response to TNFα; whereas the NF-κB inhibitor BAY11-7085 blocked the nuclear translocation of p65 induced by both spike and TNFα (Figure 3a ). Consistent with these observations, spike and TNFα induced the degradation of the isoform ⍺ of the inhibitor IκB (IκB⍺) in HUVEC and anti-α5 antibodies prevented IκB⍺ degradation in response to spike but not in response to TNFα, whereas BAY11-7085 prevented IκB⍺ degradation by both spike-and TNFα (Figure 3b ).

We conclude that spike activates NF-κB through its interaction with integrin α5β1.

To evaluate whether the spike-induced leukocyte adhesion to ECs is dependent on NF-κB activation, we tested the 

and NF-κB cytokines (TNF⍺, IL-1β, and IL-6) (Figure 3g ), and chemokines (CXCL8/IL8 and CCL2) (Figure 3h ) in HUVEC.

Upregulation of coagulation factors and proinflammatory cytokines were prevented by integrin ⍺5 antibodies and the inhibitor of NF-κB, whereas the spike-induced expression of chemokines was only prevented by the inhibition of NF-κB and not integrin ⍺5 immunoneutralization (Figure 3f-h) . These findings imply that spike activates NF-κB signaling pathways that are dependent and independent of α5β1 integrin to elicit the vascular proinflammatory and procoagulant state characteristic of severe COVID-19.

Spike binding to ACE2 is an alternative mechanism producing EC damage. Spike protein of SARS-CoV-2 downregulates ACE2 protein levels in ECs via the ubiquitin-proteasome system, which in turn leads to mitochondrial fragmentation, impaired eNOS activity, dysregulated renin-angiotensin system, and severity of COVID-19 (9) . Here, we show that spike upregulated (~2.5-fold) the expression of ACE2 mRNA levels in HUVEC and that this effect was blocked by anti-α5 antibodies and the inhibitor of NF-κB ( Figure. 3i). These findings imply that spike exerts dual actions on ACE2 levels in ECs depending on whether it binds to ACE2 or activates α5β1-NF-κB signaling pathways. Since ACE2 may be anti-inflammatory (24) , its induction by spike could be part of a protective mechanism against EC injury. Alternatively, spike-induced ACE2 expression may worsen EC viral infection by providing more SARS-CoV-2 receptor.

To assess whether the spike protein alone could promote EC dysfunction in vivo, 2.7 μg of spike were injected intravenously (i.v.) to reach an estimated ~10 nM concentration in serum and, after 2 h, mice were perfused and tissues (lung, liver, kidney, and eye) collected to measure mRNA expression of ICAM1, VCAM1, leukocyte marker (CD45), and proinflammatory cytokines (TNF, IL-1, and IL-6). The underlying rationale being that i.v. delivery and short-term (2 h) analysis in thoroughly perfused animals would reflect a direct effect of spike on EC mRNA expression of proinflammatory genes and leukocyte attachment to ECs in the various tissues. A single spike administration was enough to increase the expression levels of ICAM1 and VCAM1 in all tissues ( 

Besides leukocyte recruitment, β1 integrins promote vascular permeability, another critical aspect of inflammation (26) . Treatment with spike, the receptor-binding domain of spike, and the RGD tripeptide induced a drop in the trans-endothelial electrical resistance (TEER) of the HUVEC monolayer, indicative of hyperpermeability ( Figure   5a ). The drop was rapid, maximal at 30 minutes, and sustained thereafter. Furthermore, immunofluorescence showed the formation of actin stress fibers, EC retraction, and inter-endothelial gaps after 30 minutes of treatment with spike ( Figure 5b) . Likewise, spike interfered with the peripheral distribution of CD31, an adhesion protein that maintains the junctional integrity of ECs (27) (Figure 5b ). These observations provided direct evidence of spikeinduced increase in EC permeability.

RhoA, Rac1, and Cdc42 are Rho small guanosine triphosphatases (GTPases) that control the actin cytoskeleton and regulate EC barrier function (28) . Redistribution of actin stress fibers causing ECs contraction involves RhoA activation (29) , whereas Rac1 and Cdc42 induce cell spreading (30) . Western blot analysis of HUVEC lysates incubated with spike for 30 min showed increases in active RhoA and active Cdc42 and a decrease in active Rac1 that were blocked by the immunoneutralization of α5 ( Figure 5c ). Furthermore, endothelial nitric oxide synthase (eNOS) derived NO stimulates vasopermeability (30) , and spike promoted the phosphorylation/activation of eNOS in HUVEC in an α5-dependent manner (Figure 5c ). In contrast to the active isoforms, the total protein levels of these proteins were similar among treatments. Finally, immunoneutralization of integrin α5β1 prevented spike-, spike receptor-binding domain-, and RGD tripeptide-induced reduction in TEER (Figure 5d ). Taken together, these studies show that spike binding to integrin α5β1 regulates Rho GTPases and eNOS phosphorylation to promote EC hyperpermeability.

To investigate whether the spike protein alone could promote EC hyperpermeability in vivo, 2 μl of vehicle (PBS) alone or containing 0.5 μg of spike were injected into the vitreous of rats and the retinal vasculature evaluated in flat-mounted retinas 24 h after intravitreal injection. Spike treatment caused multiple retinal hemorrhagic areas that were absent in control retinas injected with PBS ( Figure 6a ). In other experiments, the accumulation of Evans blue- 

Inhibitors of integrin α5β1 have been developed as promising therapeutics. For example, volociximab, a chimeric anti-integrin α5β1 monoclonal antibody, is under clinical evaluation for the treatment of cancer (32) ; and preclinical studies have evaluated the efficacy of the integrin α5β1 binding peptide, ATN-161, to inhibit beta-coronavirus (33) and SARS-CoV-2 virus infections (16, 17) . Here, we show that both volociximab and ATN-161 block the binding of 

Accumulating evidence has defined COVID-19 as a vascular disease (1) (2) (3) 34) . Blood vessel injury causes progressive lung damage and multi-organ failure in severe COVID-19, owing to edema, intravascular coagulation, vascular inflammation, and deregulated inflammatory cell infiltration. Multiple mechanisms have been proposed for vascular dysfunction in COVID-19 (1, 2, 34) ; however, little is known regarding the direct action of SARS-CoV-2 on ECs (9, 11) .

ACE2 is the best-established host receptor for spike (35) (36) (37) , although other spike cell surface receptors have been described, i.e., neuropilin-1 (38) , toll-like receptors (10, 39) , and RGD-binding integrins (15, 16) . In particular, integrin 51 is an RGD-binding integrin that, upon spike binding, mediates SARS-CoV-2 entry and infection of epithelial cells and monocytes in vitro (16) and increases lung viral load and inflammation in vivo (17) . Ligation of integrin 51 by fibronectin RGD motif activates the expression of proinflammatory genes in ECs (18) , but the binding of spike to 51 in ECs and its impact on the EC inflammatory response has not been addressed.

In this work, we show that spike binding to integrin α5β1 activates the inflammatory program of EC. Spike stimulated the expression of adhesion molecules ICAM1 and VCAM1 and the attachment of leukocytes to EC monolayers ( Figure 8 ) like TNF, a well-known inducer of sustained EC inflammatory responses (19) . Although the effect of spike on the protein levels of ICAM1 and VCAM1 was not measured, a good correlation between mRNA and protein levels has already been established for both adhesion molecules in human vascular endothelial cells treated with different inflammatory mediators (40, 41) . Also, there is a good correlation between the dose-response curve of spike on VCAM1 and ICAM1 mRNA levels and that on leukocyte adhesion. This correlation suggests that spike upregulates the production of functional VCAM1 and ICAM1 in EC. Consistent with this notion, the intravenous administration of spike upregulated the expression of ICAM1 and VCAM1 and that of the leukocyte marker CD45 in different tissues to imply that spike stimulates leukocyte vascular attachment throughout the different vascular beds, as confirmed in retinal capillaries.

The fact that both the RGD tripeptide and the receptor-binding domain of spike elicited almost identical responses to those of spike, that the RGD tripeptide itself blocked spike-and spike receptor-binding domain-binding to α5β1, and that α5β1-neutralizing antibodies prevented the spike-induced proinflammatory effect in ECs indicated that spike, through its RGD motif in its receptor-binding domain, binds to integrin ⍺5β1 in ECs to promote inflammation ( Figure 8 ). Furthermore, we show that the transcription factor NF-κB is a primary contributor to α5β1 signaling in J o u r n a l P r e -p r o o f ECs in response to spike. Spike acted on ECs to stimulate IκB⍺ degradation, nuclear translocation of NF-κB, expression of adhesion molecules (ICAM1 and VCAM1), coagulation factors (TF and FVIII), proinflammatory cytokines (TNF, IL-1β, and IL-6), and leukocyte adhesion, and treatment with the inhibitor of NF-κB and anti-α5 antibodies prevented all actions ( Figure 8 ). These findings are consistent with a previous study showing that NF-κB is the signaling pathway by which fibronectin ligation to α5β1 upregulates proinflammatory genes in ECs (18) .

Indeed, NF-κB is a predominant signaling molecule activated in ECs by proinflammatory cytokines, such as TNF via an integrin-independent route (19) .

Neutrophils, macrophages, and lung epithelial cells react to spike (10, 39) and to the envelope protein (42) via the activation of toll-like receptor-induced cytokine production, a mechanism underlying the cytokine storm observed in severe COVID-19 (25) . Integrin ⍺5β1 is widely expressed in immune cells (18) , and it may cooperate with toll-like receptors to boost their signaling pathways (43) . However, human ECs express few or no toll-like receptors on their surface (44) , implying that interactions between toll receptors and integrin ⍺5β1 do not contribute to the proinflammatory effect of spike in the endothelium. Another mechanism by which the activation of ⍺5β1-NF-κB signaling in ECs by spike can influence inflammation is the upregulation of ACE2 ( Figure 8 ). A previous report showed that by binding to the ACE2 receptor in EC, spike causes the downregulation of ACE2 protein levels and the subsequent inhibition of mitochondrial function and eNOS activity leading to EC damage (9) . Because ACE2 is anti-inflammatory (24) , and its overexpression protects against severe COVID-19 symptoms (45), spike-induced downregulation of ACE2 would cause inflammation to thrive. Paradoxically, we found that the spike activation of ⍺5β1-NF-κB signaling in ECs increases ACE2 mRNA levels. Provided that this effect translates into more ACE2 protein, these findings suggest two contrasting possibilities: (a), upregulation of ACE2 expression is as a protective mechanism against endothelial cell injury; (b), upregulation of ACE2 expression worsens viral infection of ECs by providing more SARS-CoV-2 receptor for infectivity.

Excessive vasopermeability leading to edema is another hallmark of severe COVID-19. Here, we showed that spike binding to integrin α5β1, through its RGD motif, stimulated the hyperpermeability of EC monolayers ( Figure 8 ).

Spike-induced hyperpermeability of monolayers was mimicked by its receptor-binding domain and the RGD tripeptide and prevented by neutralizing antibodies against the α5β1 integrin and α5-integrin subunit. Furthermore, spike-induced stimulation of EC hyperpermeability occurs in vivo as demonstrated by the intravitreal delivery of spike resulting in multiple hemorrhagic areas and enhanced extravasation of Evans-blue linked albumin in retinal capillaries. Spike-induced disruption of the barrier function of retinal capillaries is like the one following intravitreal administration of vascular endothelial growth factor (VEGF), a major vascular hyperpermeability factor in diabetic retinopathy, an inflammatory eye disease (46, 47) .

1-integrins are known to stabilize EC-cell junctions during development (48) , but also to promote their disruption under the context of inflammation (26) . Several inflammatory agents (LPS, IL-1, and thrombin) signal via 1integrin to promote EC permeability and contractility in vitro and vascular leakage in vivo (49) . Consistent with this observation, we showed that the spike signals through α5β1 to promote the formation of stress fibers, endothelial Altogether, we have provided in vitro and in vivo evidence to show that by binding to α5β1, spike changes the EC phenotype to promote vascular inflammation ( Figure 8 ). Upon spike-mediated α5β1 activation, ECs lose their ability to control permeability or quiesce leukocytes, both of which are characteristics of EC dysfunction (19) . These findings provide insights into the mechanisms making COVID-19 a vascular disease and encourage the development of therapeutic approaches that directly focus on α5β1-mediated vascular changes.

Because the spike protein is the main immunogen in COVID-19 vaccines, the notion that immunization with spike could cause EC dysfunction should be discussed. In contrast to the spike protein levels [~30 ng mL -1 ] found in severe COVID-19 (53) , the circulating levels of the spike protein after an mRNA vaccine are minute [~30 pg mL -1 ] (54), much lower than the EC50 value of spike determined in this study [~300 ng mL -1 ]. Furthermore, most of the spike remains attached to the cell surface and does not disperse much from the injection site (54) . Also, spike is no longer detected in circulation after the second dose of vaccine, presumably because antibodies generated by the first immunization quickly and effectively remove the minute amounts of spike reaching the circulation (54) . Therefore, vaccination should help prevent, and not worsen, EC damage in COVID-19.

Finally, we showed that two inhibitors of α5β1, volociximab and ATN-161, developed as promising therapeutics, blocked spike-induced leukocyte adhesion and hyperpermeability of EC. Indeed, ATN-161 has already been shown to inhibit SARS-CoV-2 virus infection in vivo (17) . Our findings, add ECs as a target for these drugs to reduce vascular inflammation in COVID-19, suggest the therapeutic potential of agents commonly used to treat vascular diseases, and provide tools for guiding research into the molecular mechanisms mediating the pathophysiology of COVID-19. 

Human umbilical vein endothelial cells (HUVEC) were obtained as described (56) . HUVEC were maintained in F12K media supplemented with 20% fetal bovine serum (FBS), 100 μg mL −1 heparin (Sigma-Aldrich), 25 μg mL −1 endothelial cell growth supplement (ECGS, Corning, NY, USA), and 100 U mL −1 penicillin-streptomycin.

HUVEC were seeded on a 96 well plate and grown to confluency. HUVEC monolayers were treated for 16 hours with TNFα, spike, the spike receptor-binding domain, and the RGD tripeptide alone or in combination with anti-α5β1 antibodies (5 µg mL -1 ), anti-α5 antibodies (5 µg mL -1 ), volociximab (5 µg mL -1 ), or ATN-161 (500 nM) in 20% FBS F12K without heparin or ECGS. For inhibition of NF-κB, cells were pretreated for 30 minutes with 5 µM of BAY 11-7085. Leukocytes were prepared from whole blood collected into heparin tubes. Briefly, blood was centrifuged (300 x g for 5 minutes) and the plasma layer discarded. The remaining cell pack (~5 mL) was diluted in 45 mL of red blood cells lysis buffer containing150 mM of NH4Cl, 10 mM NaHCO3, and 1.3 mM EDTA (disodium) and gently rotated for 10 minutes at RT. Leukocytes were collected by centrifugation (300 x g for 5 min), the pellet washed with cold PBS followed by centrifugation (300 x g for 5 min), and resuspended into 5 mL of warm PBS 

HUVEC were seeded on 18 mm-coverslips coated with fibronectin (1 µg cm -1 ) and placed in a 12-well plate. Cells were grown in complete media to 80% confluence, and on the day of the assay, the medium was replaced with 0.5% 

HUVEC grown to 80% confluency on 6-well plates were incubated under starving conditions (0.5% FBS, F12K medium) for 30 minutes with the NF-κB activation inhibitor, BAY 11-7085 (5 µM) or anti-α5 antibodies (5 µg mL -1 ) followed by a 4-hour incubation with 100 nM spike. RNA was isolated using TRIzol reagent (Invitrogen, tgtatgggccaggagaaagg-3') and reverse (5'-cccactcctgcctttctaca-3'); F8/Factor VIII forward (5'-aaagactcacattgatggcc-3') and reverse (5'-tctggattttgtgcatctgg-3'); TNF⍺ forward (5'-accacttcgaaacctgggat-3') and reverse (5'tcttctcaagtcctgcagca-3'); IL-1β forward (5'-ggagaatgacctgagcacct-3') and reverse (5'-ggaggtggagagctttcagt-3'); IL-6 forward (5'-cctgatccagttcctgcaga-3') and reverse (5'-ctacatttgccgaagagccc-3'); CXCL8/IL8 forward (5'gcagagggttgtggagaagt-3') and reverse (5'-accctacaacagacccacac-3'); CCL2 forward (5'-gcaagtgtcccaaagaagct-3') and reverse (5'-gctgcagattcttgggttgt-3'); ACE2 forward (5'-ccgaaatacgtggaactcatcaa-3') and reverse (5'cacgagtcccctgcatctaca-3'); and GAPDH forward (5'-gaaggtcggagtcaacggatt-3') and reverse (5'tgacggtgccatggaatttg-3'). The amplification conditions were 10 seconds at 95°C, 30 seconds at each primer pairspecific annealing temperature, and 30 seconds at 72°C for 35 cycles. The mRNA expression levels were calculated by the 2 -ΔΔCT method normalized to the human GAPDH transcript.

HUVEC were grown to confluence on a 6.5 mm transwell with a 0.4 µm pore coated with 1 µg cm -1 fibronectin (Thermo Fisher Scientific). Trans-endothelial electrical resistance (TEER) was measured using the epithelial EVOM 2 Volt/Ohm meter (World Precision Instruments, Sarasota, FL, USA). The assay was performed once the TEER measurement was stable for at least two days. Cells were pretreated for 5 minutes with anti-α5 antibodies (5 µg mL -1 ), anti-α5β1 antibodies (5 µg mL -1 ), volociximab (5 µg mL -1 ), or 500 nM ATN-161 followed by the treatment with 100 nM spike, spike receptor-binding domain, or the RGD tripeptide. TEER measurements were made over 120 minutes.

HUVEC were seeded on 18 mm coverslips coated with fibronectin (1 µg cm -1 ) and placed in a 12-well plate. Cells were grown to confluence and, once the monolayer was completely formed, cells were starved (0.5% FBS) for 1 hour before the addition of 100 nM of spike for 30 minutes. Cells were washed, fixed with 4% PFA for 30 minutes, permeabilized with 0.5% Tx-100-PBS for 30 minutes, blocked with 0.1% Tx-100, 1% BSA, 5% normal goat serum- 

Female C57BL6 mice (8-12 weeks-old) were intravenously injected with 2.7 μg of spike in 50 μL of PBS. After two hours, animals were euthanized by cervical dislocation, and intracardially perfused with at least 10 mL of PBS to remove the blood and non-adhered leukocytes. Eyes, lung, liver, and kidney were placed immediately in TRIzol reagent (Invitrogen), retrotranscribed with the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA), and transcription products quantified using Maxima SYBR Green qPCR Master Mix J o u r n a l P r e -p r o o f (Thermo Fisher Scientific) in a final reaction of 10 μL containing 20 ng of cDNA, and 0.5 μM of each of the following mouse primers: ICAM1 forward (5'-gctgggattcacctcaagaa-3') and reverse (5'-tggggacaccttttagcatc-3'); VCAM1 forward (5'-attgggagagacaaagcaga-3') and reverse (5'-gaaaaagaaggggagtcaca-3'); TNF⍺ forward (5'catcttctcaaaattcgagtgacaa-3') and reverse (5'-tgggagtagacaaggtacaaccc-3'); IL-1β forward (5'-gttgattcaaggggacatta-3') and reverse (5'-agcttcaatgaaagacctca-3'); IL-6 forward (5'-gaggataccactcccaacagacc-3') and reverse (5'aagtgcatcatcgttgttcataca-3'); CD45 forward (5'-tatcgcggtgtaaaactcgtca-3') and reverse (5'-gctcaggccaagagactaacgt-3'); and GAPDH forward (5'-gaaggtcggtgtgaacggatt-3') and reverse (5'-tgactgtgccgttgaatttg-3'). The amplification conditions were 10 s at 95°C, 30 s at each primer pair-specific annealing temperature, and 30 s at 72°C for 35 cycles. The mRNA expression levels were calculated by the 2 -ΔΔCT method normalized to the mouse GAPDH transcript.

The method previously described was used (58) . Briefly, female C57BL6 mice (8-12 weeks-old) were i.v. injected 

Male Wistar rats (250-300 g) anesthetized with 1 μL g -1 of a mix of 60% ketamine and 40% xylazine were intravitreally injected with 2 μL of vehicle (PBS) containing 0.5 μg or 0.25 μg of spike for flat-mounted retina or Evans blue method evaluation of retinal vasopermeability, respectively. Vasopermeability assays were carried out 24 h after intravitreal injections. Retinas were flat-mounted, fixed for 15 min in 4% paraformaldehyde at room temperature, washed, mounted on glass slides with 50% glycerol (Sigma-Aldrich), and observed under light-field microscopy. The Evans blue method was as previously described (59) . Briefly, anesthetized rats were injected intrajugularly with 45 mg kg -1 of the Evans blue tracer (Sigma-Aldrich) and, after 2 h, 1 mL of blood was obtained from the heart to quantify plasma Evans blue concentration. Rats were then perfused with 200 mL PBS (pH 3.5 at 37°C) at 70 mL min -1 via left ventricle and the retinas dissected, vacuum-dried for 4 h, and weighted. The Evans blue tracer was extracted in 250 μL of formamide (Sigma-Aldrich) at 72°C for 18 h. Absorbance was measured in the supernatant at 620 nm. The tracer concentration in retinal extracts was calculated through the interpolation of the absorbance (OD620) of the supernatants to those of a standard curve of Evans blue in formamide. Values were normalized by the plasma concentration and to the retina-body weight ratio.

Availability of data and material. All data generated or analyzed during this study are included in this manuscript. 

COVID-19: the vasculature unleashed

Endothelial activation and dysfunction in COVID-19: from basic mechanisms to potential therapeutic approaches

COVID-19 is, in the end, an endothelial disease

Endothelial Functions

Causes of death and comorbidities in hospitalized patients with COVID-19

Endothelialitis plays a central role in the pathophysiology of severe COVID-19 and its cardiovascular complications

Shared Features of Endothelial Dysfunction between Sepsis and Its Preceding Risk Factors (Aging and Chronic Disease)

Endothelial cell infection and endotheliitis in COVID-19

SARS-CoV-2 Spike Protein Impairs Endothelial Function via Downregulation of ACE 2

SARS-CoV-2 spike protein induces inflammation via TLR2-dependent activation of the NF-kB pathway. eLife

SARS-CoV-2 Spike Protein Induces Paracrine Senescence and Leukocyte Adhesion in Endothelial Cells

SARS-CoV-2 spike promotes inflammation and apoptosis through autophagy by ROS-suppressed PI3K/AKT/mTOR signaling

A potential role for integrins in host cell entry by SARS-CoV-2

SARS-COV-2 and infectivity: Possible increase in infectivity associated to integrin motif expression

The Spike Glycoprotein of SARS-CoV-2 Binds to β1 Integrins Expressed on the Surface of Lung Epithelial Cells

The Integrin Binding Peptide, ATN-161, as a Novel Therapy for SARS-CoV-2 Infection

In Vivo protection from SARS-CoV-2 infection by ATN-161 in k18-hACE2 transgenic mice

Alpha 5 beta 1 integrin activates an NF-kappa B-dependent program of gene expression important for angiogenesis and inflammation

Evolving functions of endothelial cells in inflammation

Endothelial FN (Fibronectin) Deposition by α5β1 Integrins Drives Atherogenic Inflammation

Signalling pathways of the TNF superfamily: a double-edged sword

How NF-κB is activated: the role of the IκB kinase (IKK) complex

Mechanisms of transendothelial migration of leukocytes

Angiotensin-converting enzyme 2 protects from lethal avian influenza A H5N1 infections

The SARS-CoV-2 Spike Protein Subunit 1 induces COVID-19-like acute lung injury in Κ18-hACE2 transgenic mice and barrier dysfunction in human endothelial cells

2021) β1-and β2-integrins: central players in regulating vascular permeability and leukocyte recruitment during acute inflammation

PECAM-1: regulator of endothelial junctional integrity

The actin cytoskeleton in endothelial cell phenotypes

Cytokine-induced F-actin reorganization in endothelial cells involves RhoA activation

Activation of Rac and Cdc42 by integrins mediates cell spreading

The NO cascade, eNOS location, and microvascular permeability

Phase Ib safety and pharmacokinetic study of volociximab, an anti-α5β1 integrin antibody, in combination with carboplatin and paclitaxel in advanced non-small-cell lung cancer

ATN-161 reduces virus proliferation in PHEV-infected mice by inhibiting the integrin α5β1-FAK signaling pathway

COVID-19 -A vascular disease

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronaviruses

Cell entry mechanisms of SARS-CoV-2

Neuropilin-1 is a host factor for SARS-CoV-2 infection

SARS-CoV-2 spike protein S1 subunit induces pro-inflammatory responses via toll-like receptor 4 signaling in murine and human macrophages

Direct reciprocal effects of resistin and adiponectin on vascular endothelial cells: a new insight into adipocytokine-endothelial cell interactions

Vascular cell adhesion molecule-1 (VCAM-1) gene transcription and expression are regulated through an antioxidant-sensitive mechanism in human vascular endothelial cells

TLR2 senses the SARS-CoV-2 envelope protein to produce inflammatory cytokines

The epithelial αvβ3-integrin boosts the MYD88-dependent TLR2 signaling in response to viral and bacterial components

Differential expression of Toll-like receptor 2 in human cells

Association between ACE2 and TMPRSS2 nasopharyngeal expression and COVID-19 respiratory distress

Vasoinhibin gene transfer by adenoassociated virus type 2 protects against VEGF-and diabetes-induced retinal vasopermeability

Diabetic Retinopathy: Vascular and Inflammatory Disease

Integrin β1 controls VE-cadherin localization and blood vessel stability

Targeting β1-integrin inhibits vascular leakage in endotoxemia

Regulation of the endothelial barrier function: a filum granum of cellular forces, Rho-GTPase signaling and microenvironment

SARS-CoV-2 Spike Protein Induces Degradation of Junctional Proteins That Maintain Endothelial Barrier Integrity

Role of prostaglandins and nitric oxide in acute inflammatory reactions in guinea-pig skin

The SARS-CoV-2 Spike protein disrupts human cardiac pericytes function through CD147 receptor-mediated signalling: a potential non-infective mechanism of COVID-19 microvascular disease

Circulating Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Vaccine Antigen Detected in the Plasma of mRNA-1273 Vaccine Recipients

A serological assay to detect SARS-CoV-2 seroconversion in humans

A protocol for isolation and culture of human umbilical vein endothelial cells

CellProfiler: image analysis software for identifying and quantifying cell phenotypes

Requirement of NOX2 Expression in Both Retina and Bone Marrow for Diabetes-Induced Retinal Vascular Injury

Sensitive Blood-Retinal Barrier Breakdown Quantitation Using Evans Blue

We thank Xarubet Ruíz Herrera, Fernando López Barrera, Adriana González Gallardo,

. J o u r n a l P r e -p r o o f