key: cord-0992394-ya8qpqy9 authors: Daly, James L.; Simonetti, Boris; Antón-Plágaro, Carlos; Kavanagh Williamson, Maia; Shoemark, Deborah K.; Simón-Gracia, Lorena; Klein, Katja; Bauer, Michael; Hollandi, Reka; Greber, Urs F.; Horvath, Peter; Sessions, Richard B.; Helenius, Ari; Hiscox, Julian A.; Teesalu, Tambet; Matthews, David A.; Davidson, Andrew D.; Cullen, Peter J.; Yamauchi, Yohei title: Neuropilin-1 is a host factor for SARS-CoV-2 infection date: 2020-06-05 journal: bioRxiv DOI: 10.1101/2020.06.05.134114 sha: 4530a08febf1b5fcd9aefc9275dba9d8009fc438 doc_id: 992394 cord_uid: ya8qpqy9 SARS-CoV-2 is the causative agent of COVID-19, a coronavirus disease that has infected more than 6.6 million people and caused over 390,000 deaths worldwide1,2. The Spike (S) protein of the virus forms projections on the virion surface responsible for host cell attachment and penetration. This viral glycoprotein is synthesized as a precursor in infected cells and, to be active, must be cleaved to two associated polypeptides: S1 and S2(3,4). For SARS-CoV-2 the cleavage is catalysed by furin, a host cell protease, which cleaves the S protein precursor at a specific sequence motif that generates a polybasic Arg-Arg-Ala-Arg (RRAR) C-terminal sequence on S1. This sequence motif conforms to the C-end rule (CendR), which means that the C-terminal sequence may allow the protein to associate with cell surface neuropilin-1 (NRP1) and neuropilin-2 (NRP2) receptors5. Here we demonstrate using immunoprecipitation, site-specific mutagenesis, structural modelling, and antibody blockade that, in addition to engaging the known receptor ACE2, S1 can bind to NRP1 through the canonical CendR mechanism. This interaction enhances infection by SARS-CoV-2 in cell culture. NRP1 thus serves as a host factor for SARS-CoV-2 infection, and provides a therapeutic target for COVID-19. (NRP1-GFP). Following a GFP-nanotrap, the isolates were probed with an antibody raised against S1. It was found that NRP1-GFP was associated with a number of proteins some of which could be S1-processed versions of the S protein ( Figure 1C ). To test this possibility, we transiently co-expressed HEK293T cells with GFP-tagged S1 (GFP-S1) and mCherry-tagged mouse Nrp1 (Nrp1-mCherry). mCherry-nanotrap established that Nrp1 associated with the S1 protein ( Figure 1D -comparable binding was also observed with NRP2-mCherry (Extended Figure 1A) . That deletion of the terminal 682 RRAR 685 residues reduced S protein binding indicated the association largely depended on the CendR motif ( Figure 1D ). Residual binding was observed with the ΔRRAR mutant indicating an additional CendR-independent association between Nrp1 and the S1 protein. To establish the functional relevance of this interaction, we first generated HeLa wild type and NRP1 knock out cell lines stably expressing ACE2, designated as HeLa WT +ACE2 and HeLa NRP1KO +ACE2, respectively. The two cell lines were then infected with SARS-CoV-2 (isolate SARS-CoV-2/human/Liverpool/REMRQ001/2020) and the cells fixed 16 hours post infection (h.p.i.) ( Figure 1E ). The level of ACE2 protein expression was comparable between these lines ( Figure 1F) . Infection was visualized and quantified by staining with an anti-SARS nucleocapsid (N) polyclonal antibody ( Figure 1G ). The percentage of infected cells was 73% lower in the HeLa NRP1KO +ACE2 cells compared to HeLa WT +ACE2 cells ( Figure 1G) . We also noted that SARS-CoV-2-infected HeLa WT +ACE2 cells displayed a distinct multi-nucleated syncytia cell pattern, as reported by others 5 . We developed an image analysis algorithm using supervised machine learning to distinguish and quantify the multi-nucleated and single-cell infections of SARS-CoV-2 (Extended Figure 2) 11 . Of infected HeLa WT +ACE2 cells, the majority (83%) formed syncytia while in HeLa NRP1KO +ACE2 cells this multi-nucleated cell phenotype was less prevalent ( Figure 1H ). Together these data establish that NRP1 promotes SARS-CoV-2 infection. The extracellular regions of NRP1 and NRP2 are composed of two complementary binding CUB domains (a1 and a2), two coagulation factor domains (b1 and b2), and a MAM domain (c1). Of these, particularly important is the b1 domain that contains the specific binding site for CendR motifs, with the b2 domain further stabilising this interaction (Figure 2A ) 12 . When we co-expressed the mCherry-b1 domain of human NRP1 together with GFP-S1, the single domain alone was sufficient for immunoprecipitation by the viral protein ( Figure 2B) . A construct with both b1 and b2 (mCherry-b1b2) displayed elevated association compared to b1 alone and full-length Nrp1-mCherry ( Figure 2B) . The S1 C-terminal residues 493-685 immunoprecipitated mCherry-b1b2 to a greater extent than full length S1, further narrowing the binding interface between the S1 C-terminus and NRP1 b1b2 domains (Extended Figure 1B ). Based on the crystal structure of NRP1 bound to the CendR motif of VEGF-A164 (12) , we generated molecular models of the SARS-CoV-2 S1 C-terminus ( 678 TNSPRRAR 685 ) bound to the b1 domain of NRP1 ( Figure 2C) . The S1 Cterminus model closely resembled that of endogenous CendR motifs bound to NRP1, consistent with the guanidino group of S1 R685 establishing a salt-bridge with the NRP1 carboxylate of D320 (12) . In the NRP1 b1 domain, T316 sits at the base of the CendR binding pocket, which in the case of VEGF-A164 controls ligand occupancy 12 . The S1 C-terminal carboxylate was predicted to interact mainly with the backbone amides of NRP1, namely K347 and E348 ( Figure 2C ). Based on this model, we investigated whether two key residues (R685 of S1 and T316 of NRP1) contributed to the interaction between S1 and NRP1. Site-directed mutagenesis of R685 of S1 493-685 to aspartic acid reduced S1 association with the NRP1 b1 domain by greater than 75 % ( Figure 2D ). Site-directed mutagenesis of T316 to arginine within mCherry-b1 reduced association with GFP-S1 493-685 , more than 80%, consistent with its inhibitory impact on VEGF-A164 binding 12 ( Figure 2E) . Finally, to evaluate the interaction of NRP1 with trimeric S on the surface of virions, we incubated vesicular stomatitis virus (VSV) pseudoviral particles harbouring SARS-CoV-2 S protein with wild type mCherry-b1 or mCherry-b1(T316R) immobilised on mCherry-nanotrap beads ( Figure 2F ). Wild type mCherry-b1 immunoprecipitated proteins that could correspond to processed forms of S1, whereas the T316R mutant did not ( Figure 2F ). To test the functional relevance of this association, we transiently expressed GFP, NRP1-GFP or NRP1(T316R)-GFP in HeLa NRP1KO +ACE2 cells. Wild type and NRP1(T316R) mutant expression and ACE2 expression levels were comparable ( Figure 3A ) and like wild type NRP1, NRP1(T316R)-GFP retained a localisation to the cell surface ( Figure 3B ). When these cells were infected with SARS-CoV-2, viral infection was significantly enhanced in cells expressing NRP1-GFP compared to GFP control, whereas cells expressing the T316R mutant failed to rescue infection ( Figure 3C ). This established that the SARS-CoV-2 S1 CendR ( 682 RRAR 685 ) and NRP1 interaction promotes infection. To evaluate the biological importance of the SARS-CoV-2 S1 interaction with NRP1, we turned to analysing infection of Caco-2 (human colon adenocarcinoma) and Calu-3 (human lung cancer) cells. Incubation of these cells with recombinant ACE2 inhibited SARS-CoV-2 infection by greater than 75%, presumably by binding to S1 and inhibiting interaction with ACE2 on the cell surface ( Figure 3D ). To establish the importance of the interaction with NRP1, we screened a series of monoclonal antibodies (mAb#1, mAb#2, mAb#3) raised against the NRP1 b1b2 ectodomain. All three monoclonal antibodies bound to the NRP1 b1b2 domain but only mAb#3 bound to the CendR-binding pocket ( Figure 3E, Extended Figure 3A) , as defined by a reduced ability to bind to a b1b2 mutant that targets residues (S346, E348, T349) at the opening of the binding pocket 12 ( Figure 2C ). All three monoclonal antibodies displayed staining by immunofluorescence in NRP1-expressing PPC-1 human primary prostate cancer cells but not in M21 human melanoma cells that do not express NRP1 (8) (Extended Figure 3B) . Using mAb#3 and a control monoclonal antibody targeting avian influenza A virus (H11N3) hemagglutinin we found that mAb#3 inhibited SARS-CoV-2 infection of Caco-2 and Calu-3 cells by 38% ( Figure 3F ). The molecular and cell biological mechanisms of SARS-CoV-2 infection remain largely unexplored. Cell entry of SARS-CoV-2 depends on priming by host cell proteases, including at the S1/S2 site, and the S2' site which drives fusion with cellular membranes 5, 6, 14 . Our data indicate that the SARS-CoV-2 S-protein binds to cell surface receptor NRP1 via the S1 CendR motif generated by the furin cleavage of S1/S2. This interaction promotes infection by SARS-CoV-2 in physiologically relevant cell lines widely used in the study of COVID-19. The molecular basis for the effect is unclear, but neuropilins are known to mediate the internalisation of CendR ligands through an endocytic process resembling macropinocytosis, and additionally serve as activators of receptor tyrosine kinases to mediate cell signalling 8, 15 . S1 binding to NRP1 may therefore contribute to viral entry and survival within the host cell. Interstingly, gene expression analysis of lung tissue from COVID-19 patients recently revealed an up-regulation of NRP1 and NRP2 (16) . A SARS-CoV-2 virus with a natural deletion of the S1/S2 furin cleavage site demonstrated attenuated pathogenicity in hamster models, causing less alveolar damage 17 . NRP1 binding to the CendR-motif in S1 is thus likely to play a role in the increased pathogenicity of SARS-CoV-2 compared with SARS-CoV. The interaction between S1 and NRP1 can be reduced by monoclonal antibodies that bind to the CendR-binding pocket of NRP1 b1. Disrupting CendR-peptide binding to neuropilins by antibodies and oligopeptide/peptidomimetics has been recognised as a potential anti-cancer strategy [18] [19] [20] [21] [22] [23] [24] [25] , and the same approach may provide a route to COVID-19 therapies. It is noteworthy that many other viruses including highly pathogenic influenza, alpha-, flavi-and retroviruses undergo furin cleavages and acquire a C-terminus that conforms to the CendR, which make them potential ligands for neuropilin binding 26-28 . They also include certain human β-coronaviruses of separate lineages such as Middle Eastern respiratory syndrome coronavirus (MERS)-CoV, and human coronavirus HKU1 (HCoV-HKU1) (28) . Taken together, we conclude that neuropilins are emerging targets in the urgent research and treatment of COVID-19 and, more broadly, important host factors to be considered in the wider context of viral entry mechanisms. The following antibodies were used in this study: mouse anti-β actin (Sigma-Aldrich, Calu-3, Caco-2 (a kind gift from Dr Darryl Hill), HeLa, HEK293T and Vero E6 cell lines were originally sourced from the American Type Culture Collection. Authentication was from the American Type Culture Collection. We did not independently authenticate the cell lines. Cells were grown in DMEM medium A clinical specimen in viral transport medium, confirmed SARS-CoV-2 positive by qRT-PCR (kindly proved by Dr Lance Turtle, University of Liverpool), was adjusted to 2 ml with OptiMEM (Gibco™, ThermoFisher), filtered through a 0.2 µm filter and used to infect Vero E6 cells. After 1 h the inoculum was diluted 1:3 (vol/vol) with MEM supplemented with 2% FCS and incubated at 37 °C in a 5% CO2 incubator for 5 days. The culture supernatant was passaged twice more on Vero E6 cells until cytopathic effect was observed and then once on Caco-2 cells to produce the stock The VSVΔG system was used to generate pseudovirus particles decorated with SARS-2-S as described previously 5 hours before the supernatant was harvested and clarified by centrifugation at 2,000 x g for 10 minutes. For immunoprecipitation experiments, VSV pseudoviral particles were concentrated 10-fold using 100KDa Amicon ® Ultra centrifugal filter units. The genes of interest were subcloned into the lentiviral vector pLVX for the generation of lentiviral particles. Lentiviral particles were produced and harvested in HEK293T cells. HeLa cells were transduced with lentiviral particles to produce stably expressing cell lines. Transduced HeLa and HEK293T cells were grown in DMEM supplemented with 10% (vol/vol) FCS and penicillin/streptomycin and grown at 37 °C in a 5% CO2 incubator. Following transduction, pLVX-expressing cells were selected with puromycin or blasticidin accordingly. Cells were lysed in PBS with 1% Triton X-100 and protease inhibitor cocktail for western blotting. The protein concentration was determined using a BCA assay kit (Thermo Fisher Scientific) and equal amounts were resolved on NuPAGE 4-12% precast gels (Invitrogen). Blotting was performed onto polyvinylidene fluoride membranes (Immobilon-FL, EMD Millipore), followed by detection using the Odyssey infrared scanning system (LI-COR Biosciences). When using the Odyssey, we routinely performed western blot analysis where a single blot was simultaneously probed with antibodies against two proteins of interest (distinct antibody species), followed by visualization with the corresponding secondary antibodies conjugated to distinct spectral dyes. Kit #1721067, Bio-Rad) was added as described in the manufacturer instructions. The absorbance of the samples was read at 655 nm using Tecan Sunrise microplate reader (Tecan, Switzerland). The SARS-CoV-2 S gene was cloned into pLVX vectors using a commercially synthesized EGFP-S gene fusion plasmid (the S gene sequence was that of SARS-CoV-2 isolate Wuhan-Hu-1; GenBank: MN908947.3; GeneArt, ThermoFischer) as a template by Gibson assembly (NEB). For the untagged version, in brief, the S gene was amplified using overlapping primers and cloned into a 'pLVX-MCS-T2A-Puro' vector previously digested with EcoRI/BamHI. The isolated S1 constructs and S1 truncations were amplified from commercially synthesized plasmids (GeneArt, The 4DEQ.pdb crystal structure of the VEGF-A peptide attached to the type C1 domain of NRP1 via a linker was used to model the potential for the furin-cleaved S protein from SARS-CoV-2 to bind. A model of the S protein with loops intact had been previously assembled from the cryo-EM structures pdbs 6VXX and 6VSB. From this model, the TNSPRRAR residues pertaining to the post-furin cleaved, Cterminus of the S1 domain of the S protein from SARS-CoV2 were placed into the NRP1 binding site, guided by the position of the VEGF-A PRR residues in 4DEQ.pdb. All authors read and approved the final manuscript. The authors declare no competing interests. Alignment of the S protein sequence of SARS-CoV and SARS-CoV-2; SARS-CoV-2 S possesses a furin cleavage site at the S1/S2 boundary. (B). Illustration depicting the CendR motif binding to NRPs, the box highlights the similarity between wellestablished NRP1 ligands and the C-terminal -RRAR motif of SARS-CoV-2 S1. (C). Co-immunoprecipitation of the SARS-CoV-2 S protein by GFP-tagged NRP1 in HEK293T cells. HEK293T cells lentivirally transduced to express untagged SARS-CoV-2 S protein were transiently transfected with GFP or NRP1-GFP and subjected to a GFP-trap based immunoprecipitation. (D). CendR motif dependent coimmunoprecipitation of the SARS-CoV-2 S1 protein by Nrp1 in HEK293T cells cotransfected to express GFP-tagged S1 or GFP-S1 ΔRRAR and mCherry-tagged I AYSNNT I A I PT NNSYECD I P I GAG I CASYQTQTNSPRRARSVASQS I I AYTMSLGAENSVAYSNNS I A I PT Figure 1 . a. SARS-CoV-2 S1 protein also interacts with NRP2. Coimmunoprecipitation of the SARS-CoV-2 S1 protein by Nrp1 and NRP2 in HEK293T cells. HEK293T cells were co-transfected with combinations of GFP-tagged S1 and mCherry-tagged Nrp1 or NRP2 and subjected to a mCherry-trap based immunoprecipitation. Summary of the relative levels of binding. The band intensities were measured from n= 3 independent experiments using Odyssey software. The band intensities, normalized to mCherry expression, are presented as the average fraction relative to the amount of S1 immunoprecipitated by Nrp1-mCherry. two-tailed unpaired t-test; P= 0.2421. b. The 493-685 portion of the SARS-CoV-2 S1 protein has enhanced interaction with NRP1 b1b2. HEK293T cells were co-transfected to express GFP-tagged S1 or GFP-tagged S1 493-685 and mCherry-tagged NRP1 b1b2. Extendend Figure 1 +mCh-b1b2 G F P G F P -S 1 G F P -S 1 4 9 3 -6 8 5 G F P G F P -S 1 G F P -S 1 4 Figure 3 WHO Coronavirus disease An interactive web-based dashboard to track COVID-19 in real time Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein A multibasic cleavage site in the spike protein of SARS-CoV-2 is essential for infection of human lung cells SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor Cell entry mechanisms of SARS-CoV-2 C-end rule peptides mediate neuropilin-1-dependent cell, vascular, and tissue penetration Neuropilin functions as an essential cell surface receptor Neuropilin regulation of angiogenesis, arteriogenesis, and vascular permeability nucleAIzer: A parameter-free deep learning framework for nucleus segmentation using image style transfer Structural basis for selective vascular endothelial growth factor-A (VEGF-A) binding to neuropilin-1 Structural basis for ligand and heparin binding to neuropilin B domains Physiological and molecular triggers for SARS-CoV membrane fusion and entry into host cells Mechanisms and regulation of endothelial VEGF receptor signalling Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in Covid-19 Attenuated SARS-CoV-2 variants with deletions at the S1/S2 junction Blocking neuropilin-1 function has an additive effect with anti-VEGF to inhibit tumor growth A phase I study of the human monoclonal anti-NRP1 antibody MNRP1685A in patients with advanced solid tumours A monoclonal antibody targeting neuropilin-1 inhibits adhesion of MCF7 breast cancer cells to fibronectin by suppressing the FAK/p130Cas signalling pathway Blocking neuropilin-2 function inhibits tumor cell metastasis Inhibition of SARS-CoV-2 infection by treatment with recombinant ACE2 in Caco-2 and Calu-3 cells. Cells were pre-treated with recombinant At 16 h.p.i. the cells were fixed and stained for SARS-CoV-2-N, and infection was quantified ELISA of anti-NRP1 monoclonal antibodies (mAb#1, mAb#2, mAb#3) diluted 1/10 using plates coated with NRP1 b1b2 wild type, b1b2 mutant (S346A, E348A, T349A) or BSA, used as control. No mAb was also used as a control. Binding is represented as arbitrary units of absorbance at 655 nm. (F) Calu-3 cells. Cells were pretreated with 50 µg/mL of anti-avian influenza A virus hemagglutinin (H11N3) (Ctrl) mAb or mAb#3 for 1 h followed by infection with SARS Cells were fixed at 16 h.p.i. and stained for SARS-CoV-2-N and Hoechst to stain nuclei. p=0.002 (Caco-2) The bars, error bars and circles represent the mean, s.e.m. and individual data points