key: cord-0866515-dpusb2tk
authors: Nguyen, Linh; McCord, Kelli A.; Bui, Duong T.; Bouwman, Kim M.; Kitova, Elena N.; Kumawat, Dhanraj; Daskhan, Gour C.; Tomris, Ilhan; Han, Ling; Chopra, Pradeep; Yang, Tzu-Jing; Willows, Steven D.; Mason, Andrew L.; Lowary, Todd L.; West, Lori J.; Hsu, Shang-Te Danny; Tompkins, S. Mark; Boons, Geert-Jan; de Vries, Robert P.; Macauley, Matthew S.; Klassen, John S.
title: Sialic acid-Dependent Binding and Viral Entry of SARS-CoV-2
date: 2021-03-08
journal: bioRxiv
DOI: 10.1101/2021.03.08.434228
sha: c482652be1c4f3984849c1971309ddeba8fbd9ee
doc_id: 866515
cord_uid: dpusb2tk

Emerging evidence suggests that host glycans influence infection by SARS-CoV-2. Here, we reveal that the receptor-binding domain (RBD) of the spike (S)-protein on SARS-CoV-2 recognizes oligosaccharides containing sialic acid (SA), with preference for the oligosaccharide of monosialylated gangliosides. Gangliosides embedded within an artificial membrane also bind the RBD. The monomeric affinities (Kd = 100-200 μM) of gangliosides for the RBD are similar to heparan sulfate, another negatively charged glycan ligand of the RBD proposed as a viral coreceptor. RBD binding and infection of SARS-CoV-2 pseudotyped lentivirus to ACE2-expressing cells is decreased upon depleting cell surface SA level using three approaches: sialyltransferase inhibition, genetic knock-out of SA biosynthesis, or neuraminidase treatment. These effects on RBD binding and pseudotyped viral entry are recapitulated with pharmacological or genetic disruption of glycolipid biosynthesis. Together, these results suggest that sialylated glycans, specifically glycolipids, facilitate viral entry of SARS-CoV-2.

Many viruses exploit carbohydrates attached to protein and lipid carriers, generally described as glycans, appended to host epithelial cells for viral entry 1 . Sialoglycans, acidic, sialic acid (SA)containing glycans (e.g., gangliosides, mucin-type O-glycan, and complex N-glycans) densely displayed on the surface of mammalian cells 2 , act as co-receptors for a wide variety of viruses, including: orthomyxoviruses, paramyxoviruses, picornaviruses, reoviruses, polyomaviruses, adenoviruses, calicivirus, and parvoviruses 3 . SARS-CoV-2, which is responsible for the global outbreak of Coronavirus Disease 2019 (COVID-19) and a member of the coronavirus family, is believed to rely on a combination of angiotensin-converting enzyme 2 (ACE2) and glycans to bind and infect the lungs, as well as other tissues and organs [4] [5] [6] . Human coronaviruses generally rely on glycans to assist in cell entry 7 . For example, Middle East Respiratory Syndrome Coronavirus (MERS-CoV) binds sialoglycans to facilitate cellular entry 8 , the human betacoronaviruses OC43 and HKU1 engage sialoglycans with 9-O-acetylated SA as key receptors 9 , while Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV-1) and CoV-NL63 exploit acidic heparan sulfate (HS) polysaccharides 10, 11 .

There is emerging evidence that acidic glycans serve as co-receptors for SARS-CoV-2.

Electrospray ionization mass spectrometry (ESI-MS) analysis revealed that binding of oligosaccharide fragments of heparin, a highly sulfated form of heparin sulfate (HS), to the receptor binding domain (RBD) of the transmembrane spike (S) glycoproteins of SARS-CoV-2 inhibits ACE2 binding 12 . Consistent with this finding, Kwon et al. showed that free HS inhibits SARS-CoV-2 infection of Vero cells 13 . Esko and coworkers reported that HS enhances the affinity of the SARS-CoV-2 RBD for ACE2, suggestive of HS acting as a more classical co-receptor 14 .

Notably, destroying cellular HS enzymatically with heparanase, or competing with unfractionated heparin, both significantly reduced SARS-CoV-2 infection 14 . It was speculated early on that SARS-CoV-2 may exploit sialoglycans on cells 15 , and some evidence has emerged suggesting that SA can bind SARS-CoV-2. Results of biolayer interferometry showed that SA-conjugated gold nanoparticles exhibit high avidity for the SARS-CoV-2 S1 protein, which contains both the Nterminal domain (NTD) and RBD of the S protein 16 . Efforts to quantify these interactions on model sialylated species (e.g. 3¢-sialyllactose), however, were challenging and only revealed weak binding (Kd ~mM) 16 . Given this weak binding, it is, perhaps, not surprising that traditional glycan microarray screening found no significant interactions between the RBD and sialylated N-glycans or gangliosides 4 . Beyond acidic glycans, recent work also suggests that the RBD mimics a galectin scaffold and can bind blood group A antigen 17 . Moreover, it was also recently reported that the sialylated N-and O-glycans on ACE2 may play a role in S-protein cell binding 18, 19 .

In light of the suggested role for glycans in SARS-CoV-2 infection, we analyzed a large library of glycans as ligands for the RBD of SARS-CoV-2. Using catch-and-release electrospray ionization mass spectrometry (CaR-ESI-MS) [20] [21] [22] , a label-free method for quantifying weak, yet biologically-relevant, interactions within a complex mixture, we discovered that several classes of sialoglycans are bound by the RBD, with ganglioside oligosaccharides being the top hits. Cellbased studies reveal that RBD binding and SARS-CoV-2 pseudotyped virus entry of ACE2expressing cells is decreased upon decreasing SA levels on cells pharmacologically, genetically, or enzymatically. Blocking glycolipid biosynthesis produced similar decreases in RBD binding and viral infection, pointing to RBD-glycolipid interactions as being important for SARS-CoV-2 infection of cells.

The SARS-CoV-2 S protein possesses 26 potential glycosylation sites (22 N-and 4 O-glycosylation sites) 23 . This extensive degree of glycosylation, together with the appearance of multiple oligomeric states (monomer, dimer, trimer and hexamer) in ESI-MS analysis ( Supplementary Fig. 1a) , makes the application of CaR-ESI-MS screening to the S protein challenging. Consequently, the RBD, which contains only two N-(N331 and N343) and two O-glycosylation sites (T323 and S325), was selected for CaR-ESI-MS screening ( Supplementary Fig. 1b,c) [24] [25] [26] [27] . A total 140 glycans ( Supplementary Fig. 2 ), which consist mostly of mammalian glycans (1-133) but also several non-mammalian rhamnose-containing oligosaccharides as negative controls (134-140), were used for screening against the SARS-CoV-2 RBD. To ensure that all species were of different molecular weights, the library was divided into sixteen sub-libraries of glycans (Library A -P, Supplementary Table 1 and Supplementary Fig.   3 ). In all cases, a reference protein (Pref) was added to control for non-specific glycan-RBD interactions during ESI 28 .

Numerous glycan ligands were released from RBD in CaR-ESI-MS performed at both at 25 ºC ( Supplementary Fig. 4 ) and 37 ºC (Fig. 1a) . The low relative abundances of released ligands are indicative of generally low affinities. No binding was detected for the non-mammalian rhamnose-containing glycans (134-140) or any detectable glycans released from Pref,. These results indicate that the low abundance ligands released from the RBD are the result of specific interactions and not false positives.

At 37 ºC, the pentasaccharide (70) of the ganglioside GM1 was the preferred ligand. Other gangliosides were also recognized, with a preference for monosialylated (70 -73) over di-and trisialylated gangliosides. Acidic human milk oligosaccharides (HMOs) were consistently observed at high relative abundances and some of the ABH blood group antigens (in particular A and H types 2, 3 and 4) exhibit moderate CaR-ESI-MS signal. Interestingly, chondroitin sulfate (33, 34) and heparosan-derived oligosaccharides containing GlcA (35, 36) , which are acidic glycans, showed little to no binding to RBD. These results, when considered with the preference for monosialylated over di-and trisialylated ganglioside oligosaccharides, argue against electrostatics being the dominant force underpinning RBD-ligand binding. Overall, the screening results obtained at 25 ºC are similar to those obtained at 37 ºC in terms of the glycans that are Figure 1. Glycan library screening and affinities for RBD. a, Normalized abundances of released glycans from the SARS-CoV-2 RBD by CaR-ESI-MS at 37 o C. Summary of the chargenormalized (relative to 70) abundances of released ligands measured by CaR-ESI-MS screening of Library A -O against SARS-CoV-2 RBD. Measurements were performed in negative ion mode with a UHMR Orbitrap mass spectrometer at an HCD energy of 50 V on aqueous ammonium acetate (100 mM, pH 6.9) solutions of SARS-CoV-2 RBD (13 μM) and glycan library (containing 50 nM of each glycan). The different classes of oligosaccharides are distinguished by colour: mint green -Lewis antigens (1-14); light orange -blood group A antigens (15) (16) (17) (18) (19) (20) ; light pink -blood group B antigens (21) (22) (23) (24) (25) (26) ; tropical pink -blood group H antigens (27) (28) (29) (30) (31) (32) ; ice blue -sulfated compounds (34) (35) (36) (37) (38) and heparan sulfates (39) (40) (41) (42) , light violet -antigen-related glycans (43- Table 2 and Supplementary Fig. 5a ). According to reported glycomics studies, RBD produced from HEK293 cells has predominantly complex type N-glycans and core 1 and 2 mucin-type O-glycans (Supplementary Table 3 Table 4 and Supplementary Fig.   6 ). Based on the reported O-glycans and results of the current N-glycan analysis, the glycan composition of each RBD species was putatively assigned (Supplementary Table 5 ). It is found that the major glycoforms possess 2-8 Neu5Ac residues and are di-and trifucosylated Table 6 ). It is notable that RBD glycosylation has a minimal effect on the measured Kd ( Supplementary Fig. 7) , suggesting that the glycosylation sites are remote from the glycan binding site(s) and do not strongly influence the conformations of glycan binding motifs within RBD.

The measured affinities are all ≥100 µM, with ganglioside 70 (GM1) having the highest affinity (160 ± 40 μM) for RBD at 25 °C. For the HS 39 -42, the trend in affinities is 39 ≈ 40 > 41 ≈ 42, with Kd ranging from 200 μM to 400 μM. With the exception of 77 (which was undetectable at 37 °C), the temperature dependence of the measured Kd was relatively minor, with changes of less than 50%. Based on the difference in affinities, the average association enthalpy change is only -3.4 kcal mol -1 , which is modest for protein-glycan interactions; however, the entropy change is also favourable (with average entropy changes of 10-20 cal K -1 mol -1 ). The trend in affinities agree well with the trends in relative abundances measured by CaR-ESI-MS ( Supplementary Fig. 8 ), establishing the reliability of CaR-ESI-MS for identifying glycan ligands and differentiating the high affinity ligands from the low affinity ligands and non-binders. 

The CaR-ESI-MS screening revealed a preference for the oligosaccharide derived from gangliosides, with GM2 and GM1 being the top hits. As gangliosides are glycolipids normally embedded within a lipid bilayer, we screened the RBD against a library of six gangliosides (GM1, GM2, GM3, GD1a, GD2 and GT1b, each nominally 1% of total lipid), presented together in a nanodisc composed of 1,2-dimyristoyl-sn-glycero-3phosphocholine (DMPC) 30 . To positively identify the gangliosides bound to RBD, ions centered at m/z 3,540 that fell within a 100 m/z window, which encompasses ions at charge state -9 of any RBD-ganglioside complexes present ( Supplementary Fig. 9a) , were isolated and then collisionally heated to release bound gangliosides ( Supplementary Fig. 9b) . Notably, signals corresponding to deprotonated GM1, GM2, and GM3 ions were measured; no signals corresponding to the disialylated gangliosides -GD1a, GD2, or GT1b -were detected. As a negative control, these measurements were repeated using identical experimental conditions but in the absence of RBD and no ganglioside ions were detected ( Supplementary Fig. 9c ). Even though GM3 was the most abundant of the gangliosides detected, this may not represent an intrinsic affinity for the RBD, but rather a greater release efficiency from the ND as the RBD complex, as suggested from a previous study 30 .

Inhibition by SA suggests multiple glycan binding sites. The screening and affinity results reveal that the RBD binds preferentially to sialoglycans and HS oligomers and also binds to other structures, such as ABH antigens (Fig. 1a) . The diversity of recognized structures raises the possibility that the RBD possesses multiple glycan binding sites, with distinct binding properties.

Experimentally derived structural data has yet to be reported for RBD-glycan complexes and the only glycan binding site (for HS) was predicted from molecular docking 6 glycan ligands increased relative to those of the acidic ligands. These data suggest that the RBD possesses at least two glycan binding sites, with distinct preferences for neutral and acidic ligands.

Sialic acid-dependent RBD-binding and viral entry. The results above suggest that SAglycoconjugates are important for infection of cells by SARS-CoV-2. To test this hypothesis, we used ACE2-expressing HEK293 cells (Fig. 4a) . When employing a trimeric RBD with a Cterminal mPlum fusion 31 , we observed robust binding of the SARS-CoV-2 RBD to ACE2 + cells, but no binding to ACE2cells (Fig. 4b) . Pseudotyped SARS-CoV-2 lentivirus, encoding GFP, shows robust infection of ACE2 + 293 cells 24 hours after a one-hour incubation (Fig. 4c) .

To test the role of SA in RBD binding and viral entry, we employed pharmacological, genetic, and enzymatic approaches to decrease the SA levels. ACE + HEK293 Cells were first treated with the sialyltransferase (ST) inhibitor 3FNeu5Ac 32 , or its non-fluorinated analog as a control, for three days, which significantly decreased SA on the cells as measured by flow cytometry with the lectins SNA (Fig. 4d) and PNA (Fig. 4e) . In 3FNeu5Ac-treated cells, RBD binding was consistently decreased (Fig. 4f) and pseudotyped lentivirus showed less infection (Fig.   4g ). The decrease in infectivity was approximately 30-40%, which was consistent over numerous independent experiments. Using immunofluorescence staining of Vero-E6 cells, a decrease in RBD binding was also observed in cells treated with 3FNeu5Ac ( Supplementary Fig. 10) .

Importantly, ST inhibition did not alter ACE2 levels ( Fig. 4h and (Supplementary Fig. 10 ). Using a control lentivirus that does not encode the SARS-CoV-2 Spike protein, no differences were observed in viral entry upon 3FNeu5Ac treatment (Supplementary Fig. 11a ).

We also genetically-ablated SA through isolation of a number of CMP sialic acid synthetase (CMAS) positive (CMAS + ) and negative (CMAS -/-) clones by CRISPR/Cas9 within ACE2 + HEK293 cells, which were all tested in parallel to minimize concerns of clonal variability.

Compared to the twelve isolated CMAS + clones, the three CMAS -/clones exhibited greatly decreased SNA staining (Fig. 4i) and enhanced PNA staining (Fig. 4j) , which is consistent with the abrogation of CMP-SA biosynthesis. RBD binding was decreased (Fig. 4k) and pseudotyped viral entry was likewise suppressed in the CMAS -/clones (Fig. 4l) , for which there were no significant differences in ACE2 expression levels (Fig. 4m) . A single CMAS + and CMAS -/clone, with comparable ACE2 expression levels, were selected and fed with 3FNeu5Ac, with results showing that 3FNeu5Ac-treatment only decreased pseudotyped viral infectivity in the CMAS + clone (Fig. 4n) . The lack of an effect of 3FNeu5Ac in CMAS -/cells strongly suggests that the effects of 3FNeu5Ac on infectivity are due to the lower SA levels on the cells.

As a third method for reducing SA levels on cells, Vero-E6 cells (Fig. 4o) and Ferret lung tissue sections (Fig. 4p) were fixed and pre-treated with neuraminidase from Vibrio cholerae prior to assessing RBD binding by immunofluorescence staining. SARS-CoV-2 RBD binding was decreased by nearly 80% for both cells and tissue, despite no differences in ACE2 levels, with the expected abrogation of SA reported by SNA and MAA lectin staining (Fig. 5p) . These results were consistent with three different sources of neuraminidase and, required overnight treatment with the neuraminidases (Supplementary Fig. 12 ).

Glycolipids are critical for SARS-CoV-2 infection. We next investigated the potential for glycolipids to play a role in RBD binding and viral entry, given that gangliosides were the top hit in our binding assays. We used an inhibitor of UDP-glucose ceramide glycosyltransferase (UGCG), called GENZ-123346. ACE2 + HEK293 cells treated for two days with this inhibitor exhibited decreased Cholera toxin (CTX) staining by flow cytometry, which is indicative of decreased glycolipid levels on the cells (Fig. 5a,b) . SARS-CoV-2 RBD binding (Fig. 5c) and pseudotyped viral infection (Fig. 5d) Fig. 11b) .

As a complementary approach, glycolipids were also depleted in cells by genetic ablation of UGCG by CRISPR/Cas9. Three UGCG + and four UGCG -/clones were isolated, which stained positive and negative for CTX, respectively (Fig. 5f ). Loss of UGCG did not significantly alter expression of ACE2 (Fig. 5g) . A single UGCG + and UGCG -/clone were selected with similar ACE2 expression levels for further testing. The UGCG -/cells showed decreased RBD binding ( Fig. 5h) and were infected to a lesser extent by SARS-CoV-2 pseudotyped lentivirus (Fig. 5i) .

Overall, these results with genetic and pharmacological ablation of glycolipid biosynthesis suggest an important role for SA-containing glycolipids in SARS-CoV-2 infection of cells. representative results (a) and quantification of the MFI (b). c, RBD binding to ACE2 + HEK293 cells following GENZ-123346 treatment. d, SARS-CoV-2 pseudovirus infection in ACE2 + HEK293 cells following GENZ-123346 treatment. e, SARS-CoV-2 pseudovirus infection in CMAS + and CMAS -/-ACE2 + HEK293 cells following GENZ-123346 treatment. f, CTX staining of UGCG + and UGCG -/-ACE2 + HEK293 cell clones. g, ACE2 expression level on the UGCG + and UGCG -/-ACE2 + HEK293 cell clones. h, RBD binding to UGCG + and UGCG -/-ACE2 + HEK293 cells. i, SARS-CoV-2 pseudovirus infection of UGCG + and UGCG -/-ACE2 + HEK293 cells. Error bars represent ± standard deviation of three replicates. Statistical significance was calculated based on two-tailed unpaired Student's t-test.

In this study, we leveraged CaR-ESI-MS, a sensitive and label-and immobilization-free assay, to study defined and natural glycan libraries to identify human glycan structures recognized by SARS-CoV-2 that may facilitate viral infection. Screening a defined glycan library against SARS-CoV-2 RBD revealed that several different classes of structures are recognized. Notably, RBD binds a variety of acidic glycans, with the highest preference for the oligosaccharides on the gangliosides GM1 and GM2. Screening of nanodiscs containing a mixture of gangliosides confirmed RBD binding to GM1 and GM2, as well as GM3. This is the first report of RBD binding to monosialylated gangliosides embedded in a lipid bilayer. The affinities of the monosialylated ganglioside oligosaccharides for SARS-CoV 2 RBD (e.g., 160 ± 40 μM for GM1 pentasaccharide at 25 ºC) are similar and, in some instances, stronger than affinities reported for other gangliosidebinding viruses (e.g., respiratory syncytial virus and influenza virus strains) for which cell entry is SA-dependent 33, 34 . It is noteworthy that, while other members of the Coronavirus family also bind glycolipids, they do so through their NTD; therefore, our observation of binding glycolipids through the RBD is an entirely new finding. Interestingly Hao et al. did not detect any binding of S1 to the glycans of glycosphingolipids, including gangliosides, in glycan microarray screening 4 . This absence of binding (with the array) might be due to the relatively low affinities of these interactions, or deleterious effects of glycan labeling on binding.

SARS-CoV-2 primarily affects the respiratory system, but it can also invade multiple organ systems. Recently, it was revealed that SARS-CoV-2 can invade the central nervous system (CNS) 35 . Infection of neurons by SARS-CoV-2 is highly relevant to our findings as gangliosides are prominent in human brain, at 10-to 30-fold higher levels than other organs 36 . Therefore, gangliosides possibly play an important role in mediating SARS-CoV-2 infection of the CNS.

Acidic human milk oligosaccharides (HMOs) were also observed at high relative abundances in CaR-ESI-MS screening, raising the possibility (based on an assumption HMOs can act as viral inhibitors) that breast fed newborns may be protected against SARS-CoV-2 infection.

Indeed, in a study of 72 neonates breastfed by COVID-19 positive mothers, it was found that none tested positive for infection after 14 days 37 . Interestingly, some neutral glycans, including ABH blood group antigens (in particular A and H types 2, 3 and 4), were also found to exhibit moderate CaR-ESI-MS signal. Very recently, binding of RBD to blood group A was demonstrated 17 , which is consistent with our findings. Together, these may offer a clue as to why blood group A patients are at higher risk of hospitalization following SARS-CoV-2 infection 38 (1 μM) for heparin (average MW 15,700 Da) 5 . Notably, in the former SPR study, it was the RBD that was immobilized, while the glycan was immobilized in the latter. Therefore, simple avidity considerations likely explain this discrepancy in Kd values. Esko and co-workers reported that destruction of cellular glycosaminoglycans (GAGs) decreased SARS-CoV-2 infection, leading to the proposal that GAGs serve as a co-receptor. Notably, the affinities we report for ganglioside binding are as strong, or stronger, than for HS. Accordingly, our demonstration that reduction of SA or ganglioside levels on cells decreases RBD binding and infection of SARS-CoV-2 pseudotyped virus argues that sialoglycans may be equally as important as GAGs. Very recently, it was reported by others that, similar to what we observe, inhibition of UGCG diminished SARS-CoV-2 pseudotyped viral entry 39 . However, our findings go well beyond that by showing a direct binding interaction between the RBD and gangliosides, that inhibition of UGCG has no effect in cells lacking SA, and that genetic ablation of UGCG also causes a decrease in viral infection.

In summary, our findings suggest that sialylated glycans, which are abundant on all human cells, are bound through the RBD of SARS-CoV-2, thereby facilitating viral entry. These finding may have important implications in the tissue tropism of SARS-CoV-2, and could lead to new therapeutic approaches. Proteins. Detailed information on the proteins used in this study and the preparation of stock solutions are given as Supplementary Information. Defined glycan library. The structures and MWs of the oligosaccharides contained within the defined library are described in Figure 1 and Table S1 (Supporting Information), along with their sources. With the exception of the HS compounds (39) (40) (41) (42) and N-glycan standard (141), the glycans were divided into 16 MW-unique sub-libraries (Library A -P, Figure S1 ). To prepare the sub-libraries, a 1 mM stock solution of each glycan was first prepared by dissolving known mass of the glycan in ultrafiltered Milli-Q water (Millipore, Billerica, MA, USA). Aliquots of these solutions were then mixed, with ultrafiltered Milli-Q water, to give 50 μM stock solutions of each sub-library. All stock solutions were stored at -20 °C until needed.

Additional details for preparation of these compounds is described in the supporting information. (5-acetamido-4,7,8,9-tetra-O-acetyl-3-dehydro-3,5-dideoxy-3-fluoro-5-b-D- 

This compound was prepared similarly to as described elsewhere 40 . Cells were centrifuged at (300 rcf, 5 min) and the cell pellets were resuspended in the flow buffer.

Differences in the median fluorescence intensity of AF647 were measured on the flow cytometer. Research Chemicals) in cell-culture grade dimethyl sulfoxide (DMSO) were prepared at 5 mM and stored at -20 ºC. 100,000 cells in 250 µL of media were seeded in a 24-well plate. The Genz stocks, or DMSO as a negative control, were added to wells at a 1:1000 dilution to achieve a final concentration of 5 µM. Cells were incubated with GENZ-123346 at 37 ºC, 5% CO2 for 48 h to deplete glycolipids.

A Student's t-test was used to assess statistical significance. All assays were conducted with replicates of n = 3. Data Availability. The authors declare that all data supporting the findings of this study are available within the paper and its supplementary information files. Please contact the corresponding authors (J.S.K) for access of raw data. This is stored electronically, and will be made available upon reasonable requests. 

The tissue slices were incubated with 3% BSA in PBS-T overnight at 4 °C. The next day, purified viral spike proteins (50 μg/ml) or antibodies were added to the tissues for 1 h at RT. With rigorous washing steps in between the secondary antibodies were applied for 45 min at RT. Where indicated tissue slides were treated with 2 mU of Vibrio cholerae neuraminidase in 10 mM potassium acetate and 2.5 mg mL -1 Triton X-100, pH 4.2 at 37 °C O/N. Pharmacological depletion of sialic acid. Stocks of peracetylated Neu5Ac or 3F-Neu5Ac were prepared in cell-culture grade dimethyl sulfoxide (DMSO) at 300 mM and stored at -20 o C. 100,000 cells in 250 µL of media were seeded in a 24-well plate. The compound stocks were added to wells at a 1:1000 dilution to achieve a final concentration of 300 µM

crRNA was designed to target human CMAS (TGAGACGCCATCAGTTTCGA, Integrated DNA Technologies; IDT) or human UGCG (CCGATTACACCTCAACAAGA, IDT). HEK293T cells (500,000) expressing ACE2 were plated in a 6-well tissue culture plate in 1.5 mL of media

16 µL CRISPRMAX reagent (ThermoFisher) in 600 µL of Opti-MEM medium (Gibco) was added to cells and incubated at 37 ºC, 5% CO2 for 24 h. The cells were removed from the 6-well plate via trypsin digestion, centrifuged (300 rcf, 5 min), then resuspended in 400 µL FACS buffer (HBSS containing 1% FBS and 500 µM EDTA). The top 5% ATTO-550 positive cells were sorted on a BD FACSMelody™ Cell Sorter into 96-well flat-bottom plates containing 200 µL of media at one cell per well and References 1. Maginnis, M.S. Virus-receptor interactions: the key to cellular invasion

Diversity in cell surface sialic acid presentations: implications for biology and disease

Sialic acid receptors of viruses

Binding of the SARS-CoV-2 spike protein to glycans

Heparan sulfate proteoglycans as attachment factor for SARS-CoV-2

Characterization of heparin and severe acute respiratory syndromerelated coronavirus 2 (SARS-CoV-2) spike glycoprotein binding interactions

Virus recognition of glycan receptors

Identification of sialic acid-binding function for the Middle East respiratory syndrome coronavirus spike glycoprotein

Structural basis for human coronavirus attachment to sialic acid receptors

Targeting heparin and heparan sulfate protein interactions

Human coronavirus NL63 utilizes heparan sulfate proteoglycans for attachment to target cells

The utility of native MS for understanding the mechanism of action of repurposed therapeutics in COVID-19: heparin as a disruptor of the SARS-CoV-2 interaction with its host cell receptor

Sulfated polysaccharides effectively inhibit SARS-CoV-2 in vitro

SARS-CoV-2 infection depends on cellular heparan sulfate and ACE2

Human Sialome and Coronavirus Disease-2019 (COVID-19) Pandemic: An Understated Correlation?

The SARS-COV-2 spike protein binds sialic acids and enables rapid detection in a lateral flow point of care diagnostic device

The SARS-CoV-2 receptor-binding domain preferentially recognizes blood group A

Inhibition of SARS-CoV-2 viral entry upon blocking N-and O-glycan elaboration

Subtle influence of ACE2 glycan processing on SARS-CoV-2 recognition

Applications of a catch and release electrospray ionization mass spectrometry assay for carbohydrate library screening

Mass spectrometry-based shotgun glycomics for discovery of natural ligands of glycan-binding proteins

Screening natural libraries of human milk oligosaccharides against lectins using CaR-ESI-MS

GlyGen: Computational and Informatics Resources for Glycoscience

Virus-receptor interactions of glycosylated SARS-CoV-2 spike and human ACE2 receptor

Mucin-type O-glycosylation Landscapes of SARS-CoV-2 Spike Proteins

Deducing the N-and Oglycosylation profile of the spike protein of novel coronavirus SARS-CoV-2

N-and O-glycosylation of the SARS-CoV-2 spike protein

Method for distinguishing specific from nonspecific protein-ligand complexes in nanoelectrospray ionization mass spectrometry

A rapid high-resolution high-performance liquid chromatographic method for separating glycan mixtures and analyzing oligosaccharide profiles

Protein-glycolipid interactions studied in vitro using ESI-MS and nanodiscs: insights into the mechanisms and energetics of binding

Multimerization-and glycosylation-dependent receptor binding of SARS-CoV-2 spike proteins

Global metabolic inhibitors of sialyl-and fucosyltransferases remodel the glycome

The greater affinity of JC polyomavirus capsid for alpha2,6-linked lactoseries tetrasaccharide c than for other sialylated glycans is a major determinant of infectivity

Binding of influenza viruses to sialic acids: reassortant viruses with A/NWS/33 hemagglutinin bind to alpha2,8-linked sialic acid

Neuroinvasion of SARS-CoV-2 in human and mouse brain

Gangliosides in the brain: physiology, pathophysiology and therapeutic applications

Neonatal management and outcomes during the COVID-19 pandemic: an observation cohort study

Association between ABO blood groups and risk of SARS-CoV-2 pneumonia

Glucosylceramide synthase inhibitors prevent replication of SARS-CoV-2 and Influenza virus

Synthesis of a New Series of Sialylated Homoand Heterovalent Glycoclusters by using

Synthesis of oligomers derived from amide-linked neuraminic acid analogues

Nonselective and efficient fluorescent labeling of glycans using 2-amino benzamide and anthranilic acid

A quantitative, high-throughput method identifies protein-glycan interactions via mass spectrometry

Measuring SARS-CoV-2 neutralizing antibody activity using pseudotyped and chimeric viruses

Development of a high cell density transient CHO platform yielding mAb titers greater than 2 g/L in only 7 days

A high-coverage shRNA screen identifies TMEM129 as an E3 ligase involved in ER-associated protein degradation

The authors acknowledge the Canadian Glycomics Network, the Natural Sciences and Engineering