key: cord-0870079-qg8ir99t
authors: Guo, Lei; Liang, Yan; Li, Heng; Zheng, Huiwen; Yang, Zening; Chen, Yanli; Zhao, Xin; Li, Jing; Li, Binxiang; Shi, Haijing; Sun, Ming; Liu, Longding
title: Epigenetic glycosylation of SARS-CoV-2 impact viral infection through DC&L-SIGN receptors
date: 2021-11-11
journal: iScience
DOI: 10.1016/j.isci.2021.103426
sha: 3360e3f6860061ca6e18a02bf8f2ea3087b5fd5b
doc_id: 870079
cord_uid: qg8ir99t

Glycosylation of SARS-CoV-2 Spike glycoprotein mediate viral entry and immune escape. While glycan site is determined by viral genetic code, glycosylation is completely dependent on host cell post-translational modification. Here, by producing SARS-CoV-2 virions from various host cell lines, viruses of different origins with diverse spike protein glycan patterns were revealed. Binding affinities to C-type lectin receptors (CLRs) DC&L-SIGN differed in the different glycan pattern virions. Although none of the CLRs supported viral productive infection, viral trans&cis-infection mediated by the CLRs were substantially changed among the different virions. Specifically, trans&cis-infections of virions with a high-mannose structure (Man5GlcNAc2) at the N1098 glycan site of the spike postfusion trimer were markedly enhanced. Considering L-SIGN co-expression with ACE2 on respiratory tract cells, our work underlines viral epigenetic glycosylation in authentic viral infection and highlights the attachment co-receptor role of DC&L-SIGN in SARS-CoV-2 infection and prevention.

Over one and a half years after its outbreak, the coronavirus disease-2019 pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) remains ongoing. As an emerging highly infectious virus, what we know about viral pathogenesis is still little despite vaccine accessibility. The viral surface trimeric spike (S) glycoprotein binds to its receptor ACE2, and after binding the S trimer undergoes proteolytic cleavage and conformational change from the prefusion to postfusion form which triggers membrane fusion and delivers the viral genome into the cytosol to initiate replication (Lan et al., 2020) . As a glycoprotein, S undergoes extensive glycosylation per protomer and glycans present on both prefusion and postfusion trimers of viral surface Watanabe et al., 2020; Yao et al., 2020; Zhao et al., 2020) . Heavy glycosylation of viral entry protein is considered a way of virus immune escape by forming a glycan shield (Watanabe et al., 2019) . Moreover, glycan presented on S protein has been suggested to support S-ACE2 binding in conformation and mediate infection as ligands for lectin receptor binding (Casalino et al., 2020; Evans and Liu, 2021) .

Glycosylation is a highly diverse process that produces abundant and highly complex glycans that are covalently attached to proteins, lipids, and even RNAs present on host cells and viruses (Flynn et al., 2021; Monteiro and Lepenies, 2017) .

Instead of template-determination, glycosylation relies on a post-translational modification (PTM) process from endoplasmic reticulum to Golgi apparatus by glycosyltransferases and glycosidases with randomness. The host cell type, cell metabolic level, cell surrounding, and cell stimuli all have a strong influence on glycosylation of a specific glycoprotein (Butler and Spearman, 2014; Goh and Ng, 2018) . Viruses hijack the host cell glycosylation machine for their glycoprotein (Monteiro and Lepenies, 2017) ; thus, the glycosylation profile of viruses may differ J o u r n a l P r e -p r o o f upon infection with all types of host cells under various physiological conditions. 22 N-linked glycosylations were assessed in S glycoprotein of SARS-CoV-2 (Watanabe et al., 2020) . The 22 N-glycan sites remain highly conserved among the prototype virus and the emerging highly contagious viruses, including the alpha to delta variants.

Only the T20N substitution in the gamma variant seems to acquire a new N-glycan site based on the N-glycosylation principle. The overall glycosylation states of the S protein produced from recombinant expression and viral infection were similar, however, composition of high-mannose structure differed (Watanabe et al., 2020; Yao et al., 2020) .

High-mannose structure (Man5-9GlcNAc2) are main ligands for two C-type lectin receptors (CLRs), DC-SIGN and L-SIGN (Guo et al., 2004; Mitchell et al., 2001) . As pattern recognition receptors (PRRs), these CLRs sense glycans present on the surface of pathogens to activate antiviral immune responses. Moreover, some viruses, including HIV-1 (Geijtenbeek et al., 2000) , Ebola virus (Alvarez et al., 2002) , influenza virus (Wang et al., 2008) , human cytomegalovirus (Halary et al., 2002) , Dengue virus (Tassaneetrithep et al., 2003) , and SARS-CoV (Jeffers et al., 2004) , have evolved to exploit CLRs as additional receptors for viral trans/cis-infection. DC-SIGN expressed by immature or mature DCs and specialized monocytes/macrophages (Khoo et al., 2008) . L-SIGN, which shares 77% amino acid sequence identity with DC-SIGN, is present on endothelial cells in the liver, lymph nodes, lungs, and placenta (Khoo et al., 2008) . L-SIGN has been found to bind SARS S glycoprotein and support viral infection as a functional viral receptor (Chan et al., 2006; Jeffers et al., 2004) . In view of this, works have been performed to explore the role of DC/L-SIGN as a SARS-CoV-2 receptor. Studies by Amraie et al. and Soh et al. indicate that DC-SIGN and L-SIGN act as alternative entry receptors for SARS-CoV-2 infection using pseudo-type virus system and authentic virions. (Soh et al., 2020; Amraei et al., 2021; ) . However, other works revealed that DC/L-SIGN alone does not allow direct cell infection and proliferation; in contrast, they mediate SARS-CoV-2 infection in the presence of ACE2 as co-receptor or auxiliary receptor (Lempp et al., 2021; Thepaut et al., 2021; Kondo et al., 2021) . Monocyte-derived DCs J o u r n a l P r e -p r o o f (MDDCs) expressing DC-SIGN capture SARS-CoV-2 virions and promote virus transfer to infect ACE2+ Calu-3 cells (trans-infection) (Thepaut et al., 2021) . A recent work found that overexpression of DC/L-SIGN in 293T cells enhances viral infection, suggesting the possibility of cis-infection by DC/L- SIGN (Lempp et al., 2021) . Thus, DC-and L-SIGN plays alternative receptors and/or co-receptors in SARS-CoV-2 infection. The discrepancy of DC-SIGN and L-SIGN receptors in mediating SARS-CoV-2 infection is possible due to the heterogeneity of high-mannose binding ligands present on S glycoprotein of different produced virions. Thus, to explore epigenetic glycosylation of SARS-CoV-2 and the implication in viral infection, the virions were prepared from various host cell lines expressing ACE2 receptor and viral S glycan profiles and viral infections through these two CLRs were analyzed.

Affinities of S glycoprotein and DC/L-SIGN were measured by microscale thermophoresis (MST) assay and the results showed that both DC-SIGN and L-SIGN interact with recombinant S protein with a relatively lower affinity (Kd, 28.8 nM, 43.7 nM, respectively) than that of the ACE2 receptor (Kd, 3.04 nM) (Figures 1A, 1B, and 1C) . Furthermore, both the S1 and S2 subunits of S protein interacted with DC-SIGN and L-SIGN with affinities varying from 3.27 to 277 nM (Figures 1D, 1E, 1F, and 1G) .

After treating viral proteins with PNGasF which specifically cleaves N-linked oligosaccharides, ELISA showed that the binding of recombinant S, S1, and S2 proteins to DC-SIGN and L-SIGN was eliminated, while the binding of ACE2 was retained ( Figure 1H , 1I, and 1J). Moreover, the binding of SARS-CoV-2 virions produced from Vero cells infected with DC-SIGN and L-SIGN was also abolished when virions were preincubated with PNGasF ( Figure 1K ). Thus, the results suggest that SARS-CoV-2 virus binds to DC-SIGN and L-SIGN with S glycoprotein which is dependent on its N-glycans.

A549 cells and MLE-12 cells that did not express human ACE2 receptor were used J o u r n a l P r e -p r o o f to stably express DC-SIGN or L-SIGN for SARS-CoV-2 infection (Figure 2A and S1).

Although the virus bound to CLR expressing cells after incubation (2 h post-infection, 2 h.p.i.), DC-SIGN or L-SIGN does not support viral replication and proliferation based on the detection of viral loads, viral nucleocapsid protein expression, and viral titers at 24 h.p.i. (Figure 2B and S2). Considering the glycosylation heterogenicity of S protein from different production systems, virions were propagated from various SARS-CoV-2 permissive host cell lines, including 293T, HepG-2, Caco-2, Calu-3, Huh-7, A549-ACE2, MLE-12-ACE2, and 16HBE-ACE2 cells (Methods). The binding of the virions with various origins to ACE2, DC-SIGN, and L-SIGN receptors were evaluated with equal viral PFUs, and the results showed that all of the virions were able to bind the three receptors ( Figure 2C , 2D, and 2E). While binding capacity of the different virions to ACE2 receptor tend to be consistent ( Figure 2C ), the binding affinities varied from 1.3 to 2.1 at peak values for DC-SIGN and L-SIGN receptors ( Figure 2D amd 2E). None of the SARS-CoV-2 viruses could productively infect DC/L-SIGN expressing MLE-12 cells compared to ACE2 expressing cells based on a viral proliferation assay (24 h.p.i. viral titer detection) ( Figure 2F ). The above results indicate that although the binding capacities of DC-and L-SIGN receptors differed from those of SARS-CoV-2 viruses of various origins, the two CLRs were not able to support productive viral infection like ACE2 receptor.

MDDCs that expressed DC-SIGN ( Figure S3 ) and MLE-12-L-SIGN cells were challenged with the virions (MOI=1) for 1.5 h, and after intensive washing, they were co-cultivated with permissive Vero cells. The results showed that both CLRs can N-glycosylation sites in S protein were revealed ( Figure 4A and supplemental Excel).

The overall S protein N-glycan modifications of the tested viruses were mainly highly processed complex glycan types ( Figure 4A ). Distribution of underprocessed oligomannose and hybrid types were scattered in S1, S2 and also stalk subunits of S protein with relatively lower levels of oligomannose glycosylation at N61, N122, N234, N603, N709, N717, and N801 compared with the recombinant S trimer protein (Watanabe et al., 2020; Zhao et al., 2020) . Among the S proteins of virions of different origins, glycosylation processing levels and types at each N-glycan site differed from each other, even though the S protein from virions of Vero cell origin presented an inconsistent glycosylation state with another native Vero origin S protein (Yao et al., 2020) , highlighting that SARS-CoV-2 S glycosylation was essentially affected by host glycan epigenetic modification.

High-mannose glycans are binding ligands of DC-and L-SIGN receptors, and structures of Man5-12GlcNAc2 were observed across the S protein of the 4 virions ( Figure 4B ). High-mannose glycans (Man5, 8, 9, 11, 12GlcNAc2) were concentrated at N165, N234, N282, N331, and N616 of the S1 subunit, with abundances varying from 10% to 60%. Three N-glycan sites (N165, N234, and N331) are close to the ACE2 receptor binding domain (RBD) in steric ( Figure 4C ), which supports the notion that high-mannose structures at these sites help to stabilize receptor binding via RBD (Casalino et al., 2020) . High-mannose structures (Man5-9GlcNAc2) were detected at N717, N801, N1074, N1098, and N1134 of the S2 subunit ( Figure 4B ). The overall frequency of high-mannose glycans in S2 subunits was relatively lower than that of S1 subunits in virions from Vero, 293T, and A549(ACE2) cells, while a high proportion of Man5GlcNAc2 structures was detected at N1098 (100% occupation) and J o u r n a l P r e -p r o o f N717 (> 40%) in virions from 16HBE(ACE2) cells ( Figure 4B ). Notably, high-mannose structures were mainly distributed at the S2 subunit of virions from 16HBE(ACE2) cells, in contrast to Vero, 293T, and A549(ACE2) cell origin virions, in which high-mannose glycans dominated the S1 subunit.

Apparently, high-mannose glycans present on the top of the S1 head, such as N165, N234, and N331, instead of N717 and N1098, which are located close to the stalk domain of the S protein, are more easily to access for DC-and L-SIGN binding based on the S protein prefusion structure ( Figure 

CLRs DC-and L-SIGN were found to bind to SARS-CoV-2 S glycoprotein (Gao et al., 2020a; Amraei et al., 2021; Kondo et al., 2021; Lu et al., 2021; Thepaut et al., 2021) , while their function in inducing the proinflammatory response was revealed (Lu et al., 2021) , their role as viral infection receptors was contradictory. Unlike ACE2, which recognizes the S protein RBD, DC-and L-SIGN bind specific viral protein PTM molecules, high-mannose glycans. Since glycosylation is not a template driven process but rather depends on host PTM machine, the glycan profile of viral glycoproteins can be diverse and complex. Our results underline host glycosylation epigenetic modification on SARS-CoV-2 S protein glycan composition. Consequently, different glycan compositions of viruses from various host cells will contribute to distinctly exploring DC-and L-SIGN receptors for infection, which is well documented in our work. Our work revealed that a high-mannose structure oligomannose-type glycans than the native S trimer (Watanabe et al., 2020; Zhao et al., 2020) which is advantageous for inducing more protective high-mannose antibodies.

Based on the role of the glycosylation postfusion S trimer in mediating viral infeciton, using the full-length S protein as an immunogen may be better than using only the S1 or RBD subunit, the former of which could induce antibodies against the postfusion S trimer.

Overall, our work revealed that DC-and L-SIGN plays an attachment co-receptor 

To confirm that the high-mannose N1098 virus binds to DC&L-SIGN receptors and 

The authors declare no competing interests. 

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Longding Liu (liuld@imbcams.com.cn).

This study did not generate new unique reagents.

All data reported in this paper will be shared by the lead contact upon request. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. 

Vero (African green monkey, kidney,), 293T (human, kidney), Huh-7 (human, liver), HepG-2 (human, liver), Caco-2 (human, colon), Calu-3 (human, lung), A549 (human, were purified using anti-human CD14 antibody-labeled magnetic beads and EasySep Magent (Stemcell, Canada) from PBMCs. Differentiation to immature MDDCs was achieved by incubation of CD14+ monocytes at 37˚C with 5% CO2 for 7 days and activation with GM-CSF (1000 U mL -1 ) and IL-4 (500 U mL -1 ) (Novoprotein, China) every second day.

The viral strain SARS-CoV-2-KMS1/2020 (GenBank accession number: their titers were assessed. Among them, Vero, Caco-2, and Calu-3 cells support efficient SARS-CoV-2 proliferation (titer, > 10 7 ml -1 ), while 293T (titer, 10 5 ml -1 ), Huh-7 (titer, 10 5.5 ml -1 ), HepG-2 (titer, 10 4 ml -1 ), A549 (titer, 10 2 ml -1 ), and 16HBE (titer, < 10 2 ml -1 ) cells support moderate to limited viral proliferation. Viral titers were under detection of MLE-12 and RPMI-2650 cells. A549, 16HBE, and MLE-12 cells were conducted to stable express ACE2 by lentiviral transduction, and the corresponding A549-ACE2, 16HBE-ACE2, and MLE12-ACE2 cells efficiently supported SARS-CoV-2 virus with titers > 10 7 ml -1 .

The binding affinities of the ligand to the receptor were measured using the MST method on a Monolith NT.115 instrument (NanoTemper Technologies, Germany). The ligand, purified recombinant S, S1, and S2 proteins were fluorescently labeled with 

Recombinant SARS-CoV-2 S trimer (DRA49), S1 protein (DRA47), S2 protein (DRA48), and human ACE2 protein (C419) were produced by a mammalian expression system (HEK293 cell) purchased from Novoprotein (Shanghai, China). 

The SARS-CoV-2 spike proteins were digested by trypsin&Asp-N, trypsin&Glu-C, and chymotrypsin. Gel slices were cut into 1 mm3 cubes, and the gel cubes were transferred to a 1.5 mL microcentrifuge tube. The tube was centrifuged for 1-2 sec to spin the gel slices to the bottom of the tube. Then, 50 μL of 30 mmol/L K3Fe(CN)6: 100 mmol/L Na2S2O3 = 1:1 (vol/vol) was added, followed by washing until the brown disappeared and removing the supernatant immediately. Next, 200 μL of water was adeed to stop the reaction for 10 min, the supernatant was removed, and 100 μL of 100 mm NH4HCO3 was added. The solution was allowed to stand for 20 min, the supernatant was removed. 500 μL of acetonitrile was added and the solution was incubated for 10 min. The gel pieces should become opaque and stick together.

The acetonitrile was removed using a pipettor with a clean pipette tip, and the gel slices were rehydrated in 10 mM DTT/50 mM ammonium bicarbonate, followed by the addition of enough solution to completely cover the gel slices. After incubation at 56°C for 1 hour, the supernatant was removed, and 500 μL of acetonitrile was added, followed by incubation for 10 min. The gel pieces should become opaque and stick together. The acetonitrile was removed using a pipettor with a clean pipette tip. Then, 50 mM IAA and 50 mM ammonium bicarbonate were added to completely cover the gel slices. After incubation for 1 h at room temperature in the dark, the IAA/ammonium bicarbonate was removed using a clean pipette tip, 500 μL of acetonitrile was added, and the solution was incubated for 10 min. The gel pieces should become opaque and stick together. The acetonitrile solution was removed, and just enough enzyme digestion solution was added to cover the gel slices. The gel pieces were incubated on ice for 45 min, and more digestion solution was added if all of the initial solution was absorbed by the gel pieces. Then, 5-20 μL of enzyme digestion solution was added to keep the gel pieces wet during enzymatic digestion.

Following incubation overnight at 37°C, a pipettor and a clean pipette tip were used to recover the supernatant and transfer it into a fresh 1.5 mL microcentrifuge tube.

Then, 100 μL of 50 mM ammonium bicarbonate/acetonitrile solution (1:2, vol/vol) was added to cover the gel slices, followed by incubation for 1 hour at 37°C. The solution was extracted and transferred to a 1.5 mL microcentrifuge tube, and then, the extracted peptides were lyophilized to near dryness.

Nano LC-MS/MS analysis was conducted by BiotechPack Scientific (Beijing, China).

Before analysis, the peptides were reconstituted in 10 μL of 0.1% formic acid. 

The raw MS files were analyzed and searched against the target protein database based on the species of the samples using Byonic software (V3.6, Protein Metrics, USA). The mass tolerance was set to 20 ppm and 0.02 Da for the precursor and the fragment ions, respectively, with up to two missed cleavages allowed.

Carbamidomethyl (+57.021 Da) (C) was used as a fixed modification, while oxidation (+15.995 Da) (M) and N-glycan 309 mammalian no sodium.txt@NGlycan were used as variable modifications. The results of protein identification were filtered with the criteria of a mass tolerance less than 10 ppm for peptides and a false positive rate less than 1% at the protein level. Only peptides with high confidence were chosen for downstream protein identification analysis. N-linked glycans were categorized into 5 major classes according to the composition detected. HexNAc(2)Hex(1-12) were classified as oligomannose type with HexNAc(2)Hex(5-12) classified as high-mannose type; HexNAc(3)Hex(5-9)Fuc(0-1)NeuAc(0-1) were classified as hybrid type; and HexNAc(>3) and HexNAc(3)Hex(3-4) were classified as complex type. The rest of the types including HexNAc(1) and HexNAc(2) were classified as others.

C-type lectins DC-SIGN and L-SIGN mediate cellular entry by Ebola virus in cis and in trans

CD209L/L-SIGN and CD209/DC-SIGN act as receptors for SARS-CoV-2

The choice of mammalian cell host and possibilities for glycosylation engineering

Distinct conformational states of SARS-CoV-2 spike protein

Beyond Shielding: The Roles of Glycans in the SARS-CoV-2 Spike Protein

Homozygous L-SIGN (CLEC4M) plays a protective role in SARS coronavirus infection

Small RNAs are modified with N-glycans and displayed on the surface of living cells

SARS-CoV-2 Spike Protein Interacts with Multiple Innate Immune Receptors. bioRxiv

Development of an inactivated vaccine candidate for SARS-CoV-2

DC-SIGN, a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells

Impact of host cell line choice on glycan profile

Structural basis for distinct ligand-binding and targeting properties of the receptors DC-SIGN and DC-SIGNR

Human cytomegalovirus binding to DC-SIGN is required for dendritic cell infection and target cell trans-infection

SARS-CoV-2 reverse genetics reveals a variable infection gradient in the respiratory tract

DC-SIGN (CD209) mediates dengue virus infection of human dendritic cells

The N-terminal domain of spike glycoprotein mediates SARS-CoV-2 infection by associating with L-SIGN and DC-SIGN

DC/L-SIGN recognition of spike glycoprotein promotes SARS-CoV-2 trans-infection and can be inhibited by a glycomimetic antagonist

DC-SIGN mediates avian H5N1 influenza virus infection in cis and in trans

Site-specific glycan analysis of the SARS-CoV-2 spike

Exploitation of glycosylation in enveloped virus pathobiology

Inhibition of SARS-CoV-2 (previously 2019-nCoV) infection by a highly potent pan-coronavirus fusion inhibitor targeting its spike protein that harbors a high capacity to mediate membrane fusion

Molecular Architecture of the SARS-CoV-2 Virus

Virus-Receptor Interactions of Glycosylated SARS-CoV-2 Spike and Human ACE2 Receptor

Monoclonal antibodies for the S2 subunit of spike of SARS-CoV-1 cross-react with the newly-emerged SARS-CoV-2

Single-cell RNA-seq data analysis on the receptor ACE2 expression reveals the potential risk of different human organs vulnerable to 2019-nCoV infection