key: cord-0759440-alv7fkug
authors: Ahmetaj-Shala, Blerina; Peacock, Thomas P.; Baillon, Laury; Swann, Olivia C.; Gashaw, Hime; Barclay, Wendy S.; Mitchell, Jane A.
title: Resistance of endothelial cells to SARS-CoV-2 infection in vitro
date: 2020-11-09
journal: bioRxiv
DOI: 10.1101/2020.11.08.372581
sha: 4a64b96f43399a244f7dcb197cd17cff2f4bc34f
doc_id: 759440
cord_uid: alv7fkug

Rationale The secondary thrombotic/vascular clinical syndrome of COVID-19 suggests that SARS-CoV-2 infects not only respiratory epithelium but also the endothelium activating thrombotic pathways, disrupting barrier function and allowing access of the virus to other organs of the body. However, a direct test of susceptibility to SARS-CoV-2 of authentic endothelial cell lines has not been performed. Objective To determine infectibility of primary endothelial cell lines with live SARS-CoV-2 and pseudoviruses expressing SARS-CoV-2 spike protein. Methods and Results Expression of ACE2 and BSG pathways genes was determined in three types of endothelial cells; blood outgrowth, lung microvascular and aortic endothelial cells. For comparison nasal epithelial cells, Vero E6 cells (primate kidney fibroblast cell line) and HEK 293T cells (human embryonic kidney cells) transfected with either ACE2 or BSG were used as controls. Endothelial and Vero E6 cells were treated with live SARS-CoV-2 virus for 1 hour and imaged at 24 and 72 hours post infection. Pseudoviruses containing SARS-CoV-2, Ebola and Vesicular Stomatis Virus glycoproteins were generated and added to endothelial cells and HEK 239Ts for 2 hours and infection measured using luminescence at 48 hours post infection. Compared to nasal epithelial cells, endothelial cells expressed low or undetectable levels of ACE2 and TMPRSS2 but comparable levels of BSG, PPIA and PPIB. Endothelial cells showed no susceptibility to live SARS-CoV-2 or SARS-CoV-2 pseudovirus (but showed susceptibility to Ebola and Vesicular Stomatitis Virus). Overexpression of ACE2 but not BSG in HEK 239T cells conferred SARS-CoV-2 pseudovirus entry. Endothelial cells primed with IL-1ß remained resistant to SARS-CoV-2. Conclusion Endothelial cells are resistant to infection with SARS-CoV-2 virus, in line with relatively low levels of ACE2 and TMPRSS2, suggesting that the vascular dysfunction and thrombosis seen in severe COVID-19 is a result of factors released by adjacent infected cells (e.g. epithelial cells) and/or circulating, systemic inflammatory mediators.

COVID-19 represents one of the most important clinical challenges the scientific community has faced in recent memory. In the absence of an effective vaccine and no clear evidence that prior infection confers long-term immunity 1 , there is an urgent unmet need to understand disease pathology and to establish therapeutics that mitigate COVID-19 severity and reduce associated lethality 2 . SARS-CoV-2 spike protein binds to host cells via ACE2 3 and viral entry is facilitated by the cell-surface protease TMPRSS2 3 or lysosomal cysteine proteases cathepsin B/L (CTSB, CTSL) 3 . It has also been suggested that, as an alternative pathway, SARS-CoV-2 binds to cells via BSG (Basigin; also known as CD147 or EMMPRIN) 4, 5 , although firm evidence for BSG as a standalone receptor for SARS-CoV-2 remains the subject of investigation with a recent study noting no 'direct' binding of SARS-CoV-2 spike protein to BSG 6 .

Initial infection with SARS-CoV-2 occurs via the respiratory epithelium 7 . In most people symptoms are mild but in a significant minority COVID-19 progresses to severe disease and in those that 'recover' symptoms can persist leading to a syndrome recently defined as 'long-COVID' 8 . In severe COVID-19 and in long-COVID multiple organs including the cardiovascular system are affected 9, 10 . This secondary thrombotic/vascular clinical syndrome of COVID-19 suggests that SARS-CoV-2 infects not only respiratory epithelium but also the endothelium disrupting barrier function and allowing dissemination to other organs of the body 11 . This notion is supported by early reports showing that SARS-CoV-2 infects endothelial cells in vitro 12 and in vivo 13 . However, for the in vitro studies performed to date, nonphysiological cell lines were used and for in vivo studies serious concerns have been raised regarding the validity of the interpretation of data and the resulting conclusions 14 . Currently, therefore, it is not clear if endothelial cells are permissive to SARS-CoV-2 or not. Only by addressing this can biomarkers, mechanisms and therapeutic targets directed at the vasculopathy and thrombosis associated with COVID-19 be established.

We have recently shown that blood outgrowth endothelial cells express relatively low levels of ACE2 and TMPRSS2 but high levels of BSG and speculated that if endothelial cells are infected by SARS-CoV-2, BSG, rather than ACE2, would act as the receptor for viral entry 15 .

However, in our previous work and that of others, a direct test of susceptibility to SARS-CoV-2 of authentic endothelial cell lines has not been performed. 

Primary human endothelial cell lines including blood outgrowth endothelial cells (obtained in house 16 ), lung microvascular endothelial cells and aortic endothelial cells (obtained from Lonza; UK) were used in this study. In each case cells from 3 separate donors/ per endothelial cell type were used except for qPCR studies using blood outgrowth endothelial cells where cells from 2 donors were used. Nasal epithelial cells were obtained from Promocell (Germany) and grown in submersed culture and differentiated nasal epithelial cells (MucilAir™) were obtained from Epithelix (Switzerland) and grown in air-liquid interface culture.

African green monkey (Vero E6) cells (ATCC) and human embryonic kidney cells (293T) -293T(ATCC) were cultured in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% foetal bovine serum (FBS), 1% non-essential amino acids (NEAA) and 1% penicillin-streptomycin (P/S; Gibco). HEK 293T-ACE2 cells were produced as previous appropriate BulletKits, 10% FBS (Biosera, UK) and 1% P/S (Gibco, UK). Treatment protocols (except for those using MucilAir™ cells) were conducted using 'treatment media' (EGM-2 medium (Promocell, Germany) supplemented with 2% FBS (Promocell, Germany) and 1% P/S (Gibco, UK). For all experiments, endothelial and nasal epithelial cells were used between passage 4-6. All cells were maintained at 37°C, 5% CO2.

Endothelial and nasal epithelial cells were plated in duplicate wells on uncoated 6 well plates in their cell-specific media (see above) and grown to confluence. The day before RNA extraction media was replaced with the treatment media. After 24 hours cells were washed with phosphate-buffered saline (PBS), duplicate wells combined and RNA extracted using RNeasy Extraction kit (Qiagen, UK). RNA was converted to cDNA using the iScript cDNA synthesis kit (BioRad, CA, USA). Gene expression levels were determined using a TaqMan expression assay, with the following primers (ThermoScientific, UK); ACE2 (Hs01085333_m1), TMPRSS2 (Hs00237175_m1), BSG (Hs00936295_m1), PPIA (Hs04194521_s1) and PPIB (Hs00168719_m1). Genes were quantified relative to housekeeping genes (GAPDH and 18S) by the comparative Ct method.

For infection studies with Mucilair human airway epithelial cells, SARS-CoV-2/England/IC19/202 (IC19) 18 was diluted in serum-free DMEM, 1% NEAA, 1% P/S to a multiplicity of infection (MOI) of 0.1. Inoculum was added to the exposed apical face of the cells and incubated at 37°C for 1 hour. Inoculum was then removed and cells maintained as described above. At each timepoint in the infection, virus was collected from the apical surface of the cultures by washing with pre-warmed PBS for 10 minutes at 37°C and quantified by plaque assay on Vero E6 cells. Briefly, cells were washed with PBS then serial dilutions of inoculum, diluted in serum-free DMEM, 1% NEAA, 1% P/S, were overlayed onto cells for one hour at 37°C. Inoculum was then removed and replaced with SARS-CoV-2 overlay media (1x minimal essential media (MEM), 0.2% w/v bovine serum albumin, 0.16% w/v NaHCO3, 10mM Hepes, 2mM L-Glutamine, 1x P/S, 0.6% w/v agarose). Plates were incubated for 3 days at 37°C before overlay was removed and cells were stained for 1 hour at RT in crystal violet solution.

For infection studies using endothelial cells, cells were plated on sterile round 16mm diameter coverslips in 12 well plates (5x10 4 cells/well) without coating. Their usual media (see above)

was added and cells allowed to settle overnight. Control Vero E6 cells were plated to achieve approximately 80% confluency. The following day, media was removed and cells washed in PBS. For endothelial cells, treatment media was added either alone (untreated/ control) or with IL-1b (10ng/ml) for 3 hours. Meanwhile IC19 18 was diluted in serum free DMEM, 1%NEAA, 1% P/S, to a multiplicity of infection (MOI) of 0.1. After 3 hours treatment with either media alone or IL-1b, media was replaced with IC19 containing inoculum and incubated at 37°C for 1 hour. Inoculum was then removed and replaced with treatment media and cells were maintained until 24, 48 or 72 hours post-infection. At the appropriate timepoint, treatment media was removed and cells were fixed in 4% paraformaldehyde (PFA) for 30 minutes. PFA was removed with three washes of PBS. Cover slips were dehydrated in an ethanol series and stored in 100% ethanol at -20°C until further processing.

For fluorescent imaging, infected cells on cover slips were first rehydrated in an ethanol series.

Cells were permeabilised in PBS with 0.5% triton-X for 10 minutes, washed 3x in PBS, then blocked in PBS with 2% bovine serum albumin (BSA) and 0.1% tween (blocking buffer).

Primary antibodies were diluted 1:1000 in blocking buffer and incubated on cells at room temperature for 1 hour. Primary antibodies used were against spike protein (S) (Mouse monoclonal, Gene tex (1A9)) or nucleoprotein (N) (Rabbit monoclonal, Sino Biological).

Following 3x PBS washes, cells were incubated with secondary antibodies (anti-rabbit 488, anti-mouse 594) diluted 1:500 and DAPI (diluted 1 in 1000) in PBS with 2% BSA at room temperature for 1 hour in the dark. Stained cover slips were mounted on glass slides in ProLong gold antifade mounting medium (Invitrogen). Images were acquired using either a Zeiss Axiovert 135 TV microscope ( Figure 2B ) or a Zeiss Cell Observer widefield microscope ( Figure 2C ). All images were analysed and prepared using FIJI software 19 . For each cell type/condition slides were first reviewed by eye before representative fields of view were captured. Images were captured and processed in an identical manner across each experiment to ensure fair comparison either as a single plane of focus ( Figure 2B ) or as Z stacks presented as maximum intensity projections ( Figure 2C ). For quantitative analysis, all images were blinded and independently scored between 1-5, where 1= 0-2, 2= 3-5, 3=6-8, 4= 9-10 and 5= >10 nucleocapsid stained cells.

Lentiviral pseudotypes were generated as previous described 17 . Briefly, dishes of HEK 293T cells were co-transfected with 1 µg of HIV packaging plasmid, pCAGGs-GAGPOL, 1.5 µg of luciferase reporter genome construct, pCSLW, and 1 µg of the named envelope proteins in pcDNA3.1. Media was replaced at 24 hours to remove transfection reagents and plasmids and then pseudovirus-containing supernatants were harvested at 48 and 72 hours post-transfection. Membranes were probed with the primary antibodies: mouse anti-FLAG (Sigma: F1804); mouse anti-tubulin (abcam; ab7291); and rabbit anti-CD147 (abcam; ab108308). Near infrared (NIR) secondary antibodies, IRDye® 680RD Goat anti-mouse (abcam; ab216776) and

IRDye® 800CW Goat anti-rabbit (abcam; ab216773)) were subsequently used to detect primary antibodies. Western blots were visualised using an Odyssey Imaging System (LI-COR Biosciences).

IL-6 and IL-8 was measured using duo set ELISAs from R&D Systems according to manufactures instructions.

All data were analysed on GraphPad Prism v8 and are shown as individual data points and/or mean +/-standard error of the mean (SEM) for samples as described in the figure legends.

In our recent study, where we performed a systematic analysis of online transcriptomic databases in endothelial cells and nasal and bronchial epithelium, we found endothelial cells express relatively low levels of ACE2 and TMPRSS2 but high levels of BSG and PPIA/PPIB (also known as cyclophilin A/B; CypA/B) 15 Figure 4A and B).

There is mounting speculation that the vascular and thrombotic sequela associated with severe COVID-19 is a result of endothelial cell infection with SARS-CoV-2. However, there is currently no direct evidence to support this idea. In response to this, we have used standard While we have worked to the highest standards with empirical virology protocols, there are limitations in our approach, and we cannot definitively conclude that endothelial cells are nonsusceptible to infection by SARS-CoV-2 in some individuals or in some highly specific conditions in vivo. This is because the assay systems we used did not take account of critical factors present in at risk populations and/or at the site of inflammation. In an attempt to take account of basic inflammatory conditions we performed experiments in cells primed with IL-1ß, which did not confer infectability to any of our endothelial lines. However, as we find more about the complex mix of inflammatory mediators present in the lung and circulation in COVID-19 and the specific biological factors that predispose certain groups of individuals to severe disease, these can be recapitulated within in vitro assay systems. Nonetheless what our study does prove is that if endothelial cells are susceptible to SARS-CoV-2 at some distant point in the natural history of COVID-19, the pathways of viral entry are more complex than for airway epithelium.

We thank the FILM Facility at Imperial College London for the use of the microscopes and their technical support. 

The authors have no disclosures to declare. 

Expression levels for the genes ACE2, TMPRSS2, BSG, PPIA and PPIB were obtained from aortic (AoEC), microvascular (HMVEC) and blood outgrowth (BOEC) endothelial cells and nasal epithelial cells (NEC). Data for each donor were normalised using the average of the housekeepers (18S and Gapdh) and analysed using a comparative Ct method (2DDCt). Data are shown as the mean +/-SEM fold change compared to nasal epithelium (n=3 wells using cells from 2 donors) for AoEC (n=3 wells using cells of 3 separate donors), HMVEC ((n=3 wells using cells of 3 separate donors and BOECs (n=2 wells using cells of 2 separate donors). 

of antibody positivity to SARS-CoV-2: a community study of 365

SARS-CoV-2 and the pathophysiology of coronavirus disease 2019 (COVID-19)

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

CD147 as a Target for COVID-19 Treatment: Suggested Effects of Azithromycin and Stem Cell Engagement

SARS-CoV-2 invades host cells via a novel route: CD147-spike protein

No evidence for basigin/CD147 as a direct SARS-CoV-2 spike binding receptor

SARS-CoV-2 entry factors are highly expressed in nasal epithelial cells together with innate immune genes

The lasting misery of coronavirus long-haulers

Extrapulmonary manifestations of COVID-19

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

Endothelial cell dysfunction: a major player in SARS-CoV-2 infection (COVID-19)?

Inhibition of SARS-CoV-2 Infections in Engineered Human Tissues Using Clinical-Grade Soluble Human ACE2

Endothelial cell infection and endotheliitis in COVID-19

Caution in Identifying Coronaviruses by Electron Microscopy

Cardiorenal tissues express SARS-CoV-2 entry genes and basigin (BSG/CD147) increases with age in endothelial cells

A bioassay system of autologous human endothelial, smooth muscle cells, and leukocytes for use in drug discovery, phenotyping, and tissue engineering

The furin cleavage site of SARS-CoV-2 spike protein is a key determinant for transmission due to enhanced replication in airway cells

Selfamplifying RNA SARS-CoV-2 lipid nanoparticle vaccine candidate induces high neutralizing antibody titers in mice

Fiji: an open-source platform for biological-image analysis

Ebola virus inhibits induction of genes by double-stranded RNA in endothelial cells

Understanding SARS-CoV-2-Mediated Inflammatory Responses: From Mechanisms to Potential Therapeutic Tools

72 (C) hours post infection with SARS-CoV-2 (MOI 0.1) were determined using florescent imaging. Mock controls (media only) experiments were run simultaneously using each endothelial cell line. Data is shown as n=3 (pooled donors) for Mucilair cells (A) and n=3 (separate donors) for human aortic (AoEC), lung microvascular (HMVEC) and blood outgrowth endothelial cells (BOEC)

. Data for each donor were corrected using the average of the housekeepers (18S and GAPDH) and analysed using a comparative Ct method (2DDCt). Data are shown as the mean +/-from n=3 wells using cells from 3 separate donors for AoEC and HMVEC and n=3 wells using cells from 2 separate donors for NEC and n=2 wells from 2 separate for BOECs.

Levels of SARS-CoV-2 nucleocapsid and spike protein in Vero E6 and endothelial cells at 24 and 72 hours post infection with SARS-CoV-2 (MOI 0.1) in untreated (A-B) and IL-1ß (10ng/ml; 3 hours) (C-D) were determined using florescent imaging. One representative image was taken and blinded images were scored between 1-5, where 1= 0-2, 2= 3-5, 3= 6-8, 4= 9-10 and 5= >10 virus nucleocapsin/spike protein staining. Data are shown as individual scores for n=3 (separate donors) for human aortic (AoEC), lung microvascular (HMVEC) and blood outgrowth endothelial cells (BOEC) and n=1 for Vero E6 cells (untreated). 

Endothelial cells