key: cord-0851351-c10xeyr7
authors: Lam, Su Datt; Waman, Vaishali P; Orengo, Christine; Lees, Jonathan
title: Insertions in the SARS-CoV-2 Spike N-Terminal Domain May Aid COVID-19 Transmission
date: 2021-12-07
journal: bioRxiv
DOI: 10.1101/2021.12.06.471394
sha: 18af7bb35193c2378bf9699630d428a68780c5a9
doc_id: 851351
cord_uid: c10xeyr7

Coronavirus disease 2019 (COVID-19) caused by SARS-CoV-2 is an ongoing pandemic that causes significant health/socioeconomic burden. Variants of concern (VOCs) have emerged which may affect transmissibility, disease severity and re-infection risk. Most studies focus on the receptor-binding domain (RBD) of the Spike protein. However, some studies suggest that the Spike N-terminal domain (NTD) may have a role in facilitating virus entry via sialic-acid receptor binding. Furthermore, most VOCs include novel NTD variants. Recent analyses demonstrated that NTD insertions in VOCs tend to lie close to loop regions likely to be involved in binding sialic acids. We extended the structural characterisation of these putative sugar binding pockets and explored whether variants could enhance the binding to sialic acids and therefore to the host membrane, thereby contributing to increased transmissibility. We found that recent NTD insertions in VOCs (i.e., Gamma, Delta and Omicron variants) and emerging variants of interest (VOIs) (i.e., Iota, Lambda, Theta variants) frequently lie close to known and putative sugar-binding pockets. For some variants, including the recent Omicron VOC, we find increases in predicted sialic acid binding energy, compared to the original SARS-CoV-2, which may contribute to increased transmission. We examined the similarity of NTD across a range of related Betacoronaviruses to determine whether the putative sugar-binding pockets are sufficiently similar to be exploited in drug design. Despite global sequence and structure similarity, most sialic-acid binding pockets of NTD vary across related coronaviruses. Typically, SARS-CoV-2 possesses additional loops in these pockets that increase contact with polysaccharides. Our work suggests ongoing evolutionary tuning of the sugar-binding pockets in the virus. Whilst three of the pockets are too structurally variable to be amenable to pan Betacoronavirus drug design, we detected a fourth pocket that is highly structurally conserved and could therefore be investigated in pursuit of a generic drug. Our structure-based analyses help rationalise the effects of VOCs and provide hypotheses for experiments. For example, the Omicron variant, which has increased binding to sialic acids in pocket 3, has a rather unique insertion near pocket 3. Our work suggests a strong need for experimental monitoring of VOC changes in NTD.

Coronavirus disease 2019 , first detected in December 2019, is an ongoing pandemic situation and a cause of significant health and socioeconomic burden globally. COVID-19 is caused by a novel coronavirus namely SARS-CoV-2, which has led to the loss of the lives of around 4.2 million people worldwide (World Health Organization, 2021) .

Vaccines are being successfully administered, however the emergence of new variants of concern (VOCs) raises key questions about their phenotypic impact on transmissibility, disease severity, risk of re-infection and impact on diagnostics (World Health Organization, 2021) . So far, five VOCs have been detected namely: Alpha variant (reported in 197 countries), Beta (147 countries), Gamma (105 countries), Delta (in 201 countries) and Omicron variant (in 19 countries) (World Health Organization, 2021) . All VOCs are reported to carry novel mutations particularly in the domains of their spike protein, along with a few additional mutations in other parts of the virus genome (http://sars2.cvr.gla.ac.uk/cog-uk/;https://www.who.int/en/activities/tracking-SARS-CoV-2variants/).

The Spike protein (S) is the key antigen and plays a major role in virus host cell attachment and entry and fusion with the host cell membrane. It is the primary target for neutralising antibodies during infection and the majority of vaccines. It is a homo-trimeric protein and in SARS-CoV-2, it is made up of two subunits, S1 and S2 . The S1 subunit comprises two domains: the NTD (N-terminal domain) and the RBD (receptor binding domain). Structural studies have revealed mechanisms by which the RBD of the SARS-CoV-2 spike protein interacts with human ACE2 protein to gain entry into the host cell (Lan et al., 2020; Shang et al., 2020) .

A recent study by Dicken and co-workers (2021) showed that the N-terminal domain (NTD) of the spike protein might also have a role in facilitating efficient virus entry. Their experimental study of the Alpha variant investigated how amino acid deletions in NTD resulted in more efficient entry into human cells and infection, as compared to the reference strain (Wuhan-Hu-1). Furthermore, the NTD domain is observed to be a sequence diversity hotspot. The comprehensive sequence-based analyses of the Sarbecovirus subgenus of the Betacoronaviruses (BCoVs) by Dicken et al. (2021) showed five distinct regions insertions and deletions (indels) in the NTD domains of different BCoVs. These were found to be associated with flexible loops in the NTD structure. Intriguingly, indel regions 1, 3 and 5 correspond to an NTD antigenic supersite (McCallum et al., 2021 ) (see Figure 4B ), that is a target of all known NTD-specific neutralising antibodies, thus supporting the role of NTD in eliciting protective immunity. McCallum et al. (2021) recently revealed that most of the VOCs harbour novel mutations in this NTD antigenic supersite.

The NTD domain of the Spike protein has been proposed to bind to C-type lectin receptors such as L-SIGN/CD-SIGN (Amraei et al., 2020) . Furthermore, the role of NTD in binding to sialic acidcontaining glycoproteins and gangliosides has been reported in several coronaviruses and other viruses (Fantini et al., 2020; Seyran et al., 2021) . There is also some evidence from experimental studies of other BCoVs that changes in NTD may improve affinity for sugars on the host cell surface and therefore increase the affinity of the virus for the specific host cell (Hulswit et al., 2019; Qing et al., 2020) . Sialic acids comprise a nine-carbon backbone and are located at the terminal end of glycans. Several in silico structural studies have searched for and analysed pockets in NTD. Three pockets have been found, to date, two of which have been shown to bind sialic acid (Awasthi et al., 2020; Fantini et al., 2020; Robson, 2020; Baker et al., 2021; Bò et al., 2021; Di Gaetano et al., 2021; Milanetti et al., 2021) .

The N-terminal domain of SARS-CoV-2 Spike protein is a beta sandwich protein belonging to the "Spike glycoprotein, N-terminal domain" superfamily (ID: 2.60.120.960) in the CATH structural classification (Sillitoe et al., 2021) . The sugar-binding protein human Galectin is also found within the same fold group (2.60.120) and the two domains are structurally similar. Studies have shown that the NTD of HKU-23 (a BCoV) is structurally related to the Galectin fold (e.g., Galectin-3), possibly co-opted by the virus from the host (Li, 2016) . More recently, Cheng et al. showed that BCoVs (BovineCoV-NTD, PHEV-NTD, HCoV-OC43-NTD, HCoV-HKU23-NTD, HKU1-NTD, MHV-NTD, SARS-CoV-NTD and MERS-NTD) bind to sugars such as sialic acid-containing glycoproteins and gangliosides in an equivalent binding pocket to Galectin-3 (referred to as 'Pocket 1') (Cheng et al., 2019) .

Interestingly, Behloul and co-workers found a GTNGTKR motif, a known sugar binding motif (located at residues 72-78 in SARS-CoV-2 Spike protein) within a different pocket on the protein (referred to as 'Pocket 2') suggesting that this region may also bind sugars (Behloul et al., 2020) . Awasthi et al (2020) performed a comparative structural analysis of pocket 2 in the NTD of SARS-CoV-2 and SARS-CoV, using molecular dynamics and docking approaches. Their analyses highlighted the flexible nature of the loop (L244-G261, part of pocket 2) allowing pocket 2 of SARS-CoV-2:NTD to bind sialosides more strongly than that of SARS-CoV (Awasthi et al., 2020) . However, human Galectin-3 does not bind sugar in this pocket 2 region, it solely utilises pocket 1 (Johannes et al., 2018) . Pocket 2 is detected in other coronaviruses including BovineCoV-NTD, PHEV-NTD, HCoV-OC43-NTD, HCoV-HKU23-NTD, HKU1-NTD, MHV-NTD but not well defined in SARS-CoV-NTD nor MERS-NTD (Cheng et al., 2019) . Furthermore, the structure of HCoV-OC43 has been determined with a sialic acid (i.e., 9-O-acetyl-sialic acid) bound in Pocket 2 (Tortorici et al., 2019) . More recently, a third druggable pocket was detected in NTD domain of SARS-CoV-2, but there is no indication to date on whether it can bind sugars (Di Gaetano et al., 2021) .

Our work builds on these previous studies and further explores the role of NTD in sialic acid binding. We explore the characteristics of all putative sialic binding pockets in NTD and examine the effects of variants (residue insertions/deletions) in these pockets on sialic acid binding. In particular we use docking studies to characterise the pockets and observe that pockets 2 and 3 bind sialic acid more strongly than pocket 1. We also detect a fourth pocket (pocket 4) that can bind sialic acid but not as strongly as pockets 2 and 3.

Since Awasthi et al. (2020) showed that insertions in SARS-CoV-2 relative to SARS-CoV enhanced sialic binding, we structurally compared all 4 of the pockets across related coronaviruses to determine whether (and if so how) these pockets had evolved to enhance binding to polysaccharides. Our results support the work of Awasthi et al, showing that SARS-CoV-2 possesses additional loops in pocket 2 that extend the pocket to increase contact with polysaccharides. We also found extensions in pocket 3. We then determined whether variants, found in VOCs and VOIs of SARS-CoV-2, including insertions/deletions in the loops defining the NTD sugar binding pockets could result in enhanced binding of the spike protein to sialic acid (and therefore to the host membrane).

Finally, we assessed which of the pockets were druggable and sufficiently structurally conserved across related coronaviruses for their structural features to be exploited in the design of generic drugs against these BCoV viruses. Whilst there are some therapeutic strategies targeted at pocket 1 in SARS-CoV (Fantini et al., 2020) , this pocket is too variable across BCoVs. Pocket 2 and 3 are even more structurally variable and again not amenable to generic drug design. However, the fourth druggable pocket, pocket 4, appears to be highly structurally conserved and could therefore be further investigated in pursuit of a generic (pan BCoVs) drug. Sialic acid targeting, or mimicking drugs could serve as good candidates for antiviral strategies (Heida et al., 2021) .

To explore the sequence and structure conservation of BCoVs, we selected a set of representative sequences, expected to be closely related to SARS-CoV-2 and also strains that have caused human disease . We obtained 33 nucleotide sequences of the Sarbecovirus subgenus of the BCoV and SARS-related coronaviruses from NCBI (Sayers et al., 2020) and GISAID (Elbe and Buckland-Merrett, 2017; Shu and McCauley, 2017) databases (See Supplementary Table 1) . We extracted the Spike protein sequences by scanning the sequence of the SARS-CoV-2 Spike protein (YP_009724390.1) against the nucleotide database of BCoV sequences using NCBI BLAST v2.6 tblastn (Altschul et al., 1990) . For each of the Spike protein sequences extracted, we then extracted the N-terminal domain (residues 1-303).

We also obtained BCoV NTD domain sequences for these proteins from the CATH "Spike glycoprotein, N-terminal domain" domain superfamily (2.60.120.960) (Sillitoe et al., 2021) .

We (Berman et al., 2000) .

Structural models of other BCoV NTDs were built using an in-house FunMOD modelling pipeline (Lam et al., 2017; Sillitoe et al., 2020) , provided they had more than 40% sequence identity to the template (see Supplementary Table 2 for details). Otherwise, models were built using AlphaFold2 (Jumper et al., 2021 , 'AlphaFold2_mmseqs2' notebook available from https://github.com/sokrypton/ColabFold, Mirdita et al., 2021 ) (see Supplementary Table 3 for details). FunMOD generated query-template alignments using HH-suite version 3 (Steinegger et al., 2019) , which were then used as input to the MODELLER v.9.23 program (Webb and Sali, 2016) . We used the 'very_slow' schedule for model refinement. Ten models were generated for each query and we then selected the model with the lowest normalised DOPE score (nDOPE) (Shen and Sali, 2006) , which reflects the quality of the model. Positive scores are likely to be poor models, while scores lower than −1 are likely to be native-like.

To build a Maximum likelihood tree, we aligned all NTD amino acid sequences using CLUSTAL-OMEGA (Madeira et al., 2019) . A phylogenetic tree was inferred according to the Maximum Likelihood method. Genetic distance was computed using the Whelan and Goldman model (Whelan and Goldman, 2001 ) and gamma-distributed rate variation among sites (WAG + G). The evolutionary history was inferred using the Maximum Likelihood method and the analyses were conducted using MEGA X (Kumar et al., 2018) .

We used ESPript3 webserver (http://espript.ibcp.fr/ESPript/ESPript/) (Robert and Gouet, 2014) to produce structure-based multiple sequence alignment. The secondary structure elements were defined based on the SARS-CoV-2 NTD structure (PDB ID 6ZGE).

The ScoreCons method (Valdar, 2002) was used to calculate the sequence conservation of residues at a particular position. ScoreCons reports a score between 0-1 which is robust for multiple sequence alignments with high information content (DOPS score >70). A ScoreCons value above 0.7 suggests a highly conserved position.

Protein structures were compared using our in-house SSAP algorithm (Taylor and Orengo, 1989) . The SSAP score ranges from 0 to 100. Structures with a SSAP score above 80 are considered to be highly similar. Protein structures were rendered using PyMOL (Schrödinger and DeLano, 2020) and UCSF Chimera (Pettersen et al., 2004) .

We used CavityPlus (Xu et al., 2018) to predict druggable pockets using chain A of SARS-CoV-2 structure (PDB ID: 7C2L). The CavityPlus DrugScore denotes the druggability of a particular pocket with more positive scores indicating more druggable sites.

docking was done using the HADDOCK 2.4 webserver (https://wenmr.science.uu.nl/haddock2.4/) (Honorato et al., 2021) . HADDOCK was one of the topperforming protein-ligand binding methods in the D3R Grand Challenge 2 and 3 (Kurkcuoglu et al., 2018; Gaieb et al., 2019) . The sialic acid used was 9-O-acetylated sialic acid (PubChem ID 71312953). This ligand is the sialic acid bound to the experimental structure of HCoV-OC43 (PDB ID 6NZK, Tortorici et al., 2019) . The HADDOCK score is a linear combination of van der Waals, electrostatics, and desolvation energy terms. A lower HADDOCK score signifies stronger binding. The binding affinity of sialic acid-NTD was predicted using HADDOCK's PRODIGY-LIGAND server (https://wenmr.science.uu.nl/prodigy/) (Vangone et al., 2019) .

We used the LIGPLOT program of PDBsum (Laskowski and Swindells, 2011) to examine residue interactions between sialic acid and the NTDs.

We performed phylogenetic studies on the NTD domain to explore the relationships between the different BCoVs. We obtained sequences of the NTD domain for 26 BCoVs from the NCBI GenBank database and GISAID database.

BCoVs have been classified into 5 lineages: Embecovirus (lineage A), Sarbecovirus (lineage B), Merbecovirus (lineage C), Nobecovirus (lineage D) and Hibecovirus (Woo et al., 2010; Zhao et al., 2020) . SARS-CoV-2 and SARS-related coronaviruses belong to the Sarbecovirus lineage (lineage B). HCoV-OC43 belongs to Embecovirus (lineage A), while MERS-CoV was classified into Merbecovirus (lineage C). So far, there have been no reports of human-infecting BCoVs from other lineages. Figure 1 demonstrates that the phylogenetic tree of SARS-CoV-2 based on the NTD domain, is similar to those constructed from other parts of the Spike protein i.e. SARS-CoV-2, BatCOV-RaTG13, PangolinCoV and BatCoV-ZXC21 and BatCoV-ZC45 fall into the same clade in the tree. In contrast, SARS-CoV, Civet-SARSr-CoV and BatCoV-RmYN02 were found to belong to another clade. HCoV-OC43, which is thought to have emerged in the 1950s (Vijgen et al., 2005; Lau et al., 2011) and MERS-CoV are more distant from the others. See Supplementary Figure 1 for a bootstrap tree.

Phylogenetic tree of BCoV NTD amino acid sequences. The phylogenetic tree was inferred according to the Maximum Likelihood method. Genetic distance was computed using the Whelan and Goldman model and gamma-distributed rate variation among sites (WAG + G).

The structures of the Spike proteins from the Wuhan strain of SARS-CoV-2 and several other strains (i.e. SARS-CoV, PangolinCoV-GX, PangolinCoV-GD, BatCov-RaTG13) have been determined by experimental methods (Song et al., 2018; Wrapp et al., 2020; Wrobel et al., 2021; Zhang et al., 2021) . The NTD domain interacts with a number of other domains within the Spike protein (subdomains SD1 and SD2, which serve as a hinge for RBD up-movement (Berger and Schaffitzel, 2020; Peters et al., 2020) ), however a significant proportion of the domain is surface accessible and able to intera with other compounds and proteins on the host cell surface (see Figure 2 ).

Structure of the SARS-CoV2 Spike protein highlighting the different domains. The RB domain is coloured in blue, the NTD domain in purple, the SD1 domain in grey and the SD1 doma in black. The trimer complex is shown in (A). The S1 region of a Spike protein monomer is shown (B). The interactions of the NTDs with SD1 and SD2 domains of the Spike protein trimer are show in (C). PDB structure 6VSB.

Structure analyses of the 4 BCoVs that are in the same lineage (B) (see above) as SARS-CoV showed that the NTD domains are all assigned to the same evolutionary superfamily (2.60.120.96 in the in-house CATH domain structure classification (Sillitoe et al., 2021) . This superfamily falls the fold group (2.60.120) containing the human sugar-binding protein Galectin-3. To determine ho conserved this domain is across these 4 BCoVs and SARS-CoV-2 are, and identify any variab regions, we performed a structure superposition using our in-house protein structure alignment to SSAP (see methods). Overall, the structures are very similar (see Figure 3 ) with an average SSAP 86 (range is 0 to 100) suggesting a high global conservation of this domain across these BCoV However, whilst a large core of the domain is highly conserved, there are extensions in some of t loops. The variable nature of these loop regions across multiple BCoV NTD domains was al identified in a multiple sequence alignment (Garry et al., 2021) , but the authors did not perfor teract RBD main wn in hown V-2, .960) lls in how iable t tool P of oVs. f the also form structural analyses to assess their proximity to putative sugar-binding regions. However, there some evidence in other BCoVs that these changes may improve affinity for sugars on the host c surface and therefore increase the affinity of the virus for the specific host cell (Hulswit et al., 201 Qing et al., 2020) . 

As mentioned above, there is evidence of at least three pockets in SARS-CoV-2 (Awasthi et a 2020; Fantini et al., 2020; Baker et al., 2021; Di Gaetano et al., 2021) , two of which are known bind sialic acid/sugars. To confirm and further characterise these and search for additional pock we used cavity detection and docking tools. Analysing the structure of the SARS-CoV-2 NT domain using CavityPlus (Xu et al., 2018) , we identified the known sugar binding pocket shared w Galectin-3 (pocket 1, Figure 4 ). Two other cavities were confirmed (pocket 2 (red) and 3 (blue) Figure 4 ) that share, through a common loop, a reported conserved sugar-binding motif (comprisi 7 residues -G72, T73, N74, G75, T76, K77 and R78). Furthermore, in agreement with McCallum al. (2021) we observe that the antigenic NTD supersite coincides with sugar-binding pockets 1 and Finally, a fourth, previously undefined pocket was detected (pocket 4, colour cyan) shown in figure below. Supplementary Table 4 shows the SARS-CoV-2 NTD residues that form the pockets detected by o study. Some of these (pocket1 and 2) are thought to interact with sugars (e.g. sialic acid) (Fantini al., 2020; Behloul et al., 2020 , Baker et al., 2021 . Docking studies performed by us (see Section 3 below) also suggest pocket 3 and 4 may bind sialic acid. Studies that previously identified commented on these pockets are cited in the table.

Garry et al. performed a detailed sequence analysis of multiple BCoVs and strains of SARS-CoV and demonstrated the suitability of this approach to highlight regions varying across the BCoVs. W have expanded this alignment to 31 BCoVs and used a structure guided approach to ensure accura alignment of more distant BCoVs ( Figure 5 ). Garry et al (2021) observed that lineage indels are common feature of the Spike protein in BCoVs, with seven major loops most frequently affecte Five out of seven loops involve the NTD. We aligned the sequences of the 31 BCoVs included in t phylogenetic analysis using CLUSTAL-OMEGA, excluding the MERS-CoV and HCoV-OC BCoVs as they are from a different lineage and disrupt the alignment. Since we had structural da for SARS-CoV-2, we exploited this data to produce a structure-based multiple sequence alignme Whilst residues in most regions of the protein are highly conserved (ScoreCons values > 80), t residues contributing to sugar binding pocket 1 are somewhat less conserved (pocket 1 -avera motif (P1riant, ), the erage ScoreCons of 70) and even more so for pocket 2 (average ScoreCons of 61). Pocket 2 has changed in some BCoVs (SARS-CoV, civet SARS-CoV and BatCoV-RmYN02). Cheng et al. also compared BCoVs NTD structures and found pocket 2 of SARS-CoV to be less well-defined with the loops not giving such a well-formed pocket. Our docking studies (see Section 3.6 below) suggest that pocket 3 is also able to bind sialic acid/sugars. This pocket is also quite variable with average ScoreCons values of 73. However, pocket 4 is very conserved relative to the other pockets with an average ScoreCons value of 87.

The proximity of the indel regions to residues in these known and putative sugar-binding pockets is shown in Table 1 , below. Most are close to pockets 2 and 3. Indel region 2 (close to pockets 2 and 3), indel region 3 and indel region 5 (close to pocket 3) are the most variable regions across these BCoVs. Table 1 Proximity of indel regions to residues in the sugar binding pockets. If the indel region lies close to multiple pockets (<5Å), minimal residue distances to all pockets are provided.

Minimal residue distance to known and putative sugar binding pockets (Å) We see that much of the indel innovation between lineages occurs in loop regions associated with the NTD binding pockets 2 and 3 suggesting that these binding pockets are hotspots for viral evolution ( Figure 6 ). Indeed, most of the indels occur on this side of the NTD domain (See Figure 6 ) and seem to target the sugar-binding pockets, possibly tuning the interactions with sugars such as sialic acids. 

We analysed the structural features observed in different BCoV domains in the CATH superfam containing NTD domains (see Figure 7) . We also compared against the human Galectin-3 prote which is in the same fold group as this superfamily. It can be seen that loop insertions in the BCo near the sugar-binding pocket 1 appear to enhance the contact of the protein with the ligand and t contacts increase going from human Galectin-3 to HCoV-OC43 (a BCoV that emerged in the 19 century) up to SARS-CoV-2.

As mentioned in the introduction, Behloul and co-workers found a GTNGTKR motif (located with a different pocket in the BCoVs (Pocket 2) suggesting that this region may also bind sugars (Behlo et al., 2020). Human Galectin-3 does not bind sugar at this pocket (Johannes et al., 2018) and we al see loop extensions when comparing SARS-CoV-2 to human Galectin-3 in the region around pock 2 (see Figure 7 ).

were 1 in mily otein CoVs d the 19th ithin hloul also ocket Figure 7 . Evolution of the sugar-binding pocket from human Galectin-3 to SARS-CoV-2 a emergence of other sugar-binding regions. We only show the sugar-binding regions. We coloured t human Galectin-3 NTD structure in green, hCoV-OC32 NTD structure in blue, SARS-CoV NT structure in yellow and SARS-CoV-2 NTD structure in red. PDB structures 1A3K,7C2L,6AC 6NZK respectively.

Loops in the region of pocket 2 seem to be highly structurally variable and this pocket is less w defined in SARS-CoV. Comparing the structures and sequences of HCoV-OC43 and SARS-CoV we see strong innovation in pocket 2 in SARS-CoV-2 suggesting alterations in this pocket we important for the evolution of this virus. Similarly, structural analyses comparing pocket 3 across t coronaviruses suggests that this pocket can be highly variable too, and it is also found to be w defined in SARS-CoV-2 offering another potential sialic acid binding site. We used docking stud to explore and contrast the ability of the different pockets in SARS-CoV-2 to bind sialic acid.

Previous studies had indicated that pocket 1, which is similar to the sugar binding pocket in Galact 3, could bind sugar (Fantini et al., 2020) and that pocket 2 could possibly bind sugars. For example structure of a BCoV (HCoV-HKU23 -which has 25% sequence identity to SARS-CoV-2) with sia acid bound in pocket 2 has been solved (Tortorici et al., 2019) . Therefore, whilst there is som indication, including from MD studies (Awasthi et al., 2020; Fantini et al., 2020; Behloul et al., 202 Baker et al., 2021 ) that SARS-CoV-2 can bind sialic acid in pockets 1 and 2, further computation analyses were used to explore and contrast the binding of sialic acid to all pockets in the NT domain. and d the NTD CC, well oV-2 were s the well udies ctinple, a sialic some 020; ional NTD We first determined the binding energy of sialic acid bound to pocket 2 of HCoV-HKU23 using the PRODIGY score of HADDOCK and obtained a value of -7.1 kcal/mol. The LigPlot for the structureligand complex is shown in Supplementary Figure 3 . We then used HADDOCK to dock sialic acid into all four of the pockets in the experimental structure of the NTD domain of SARS-CoV-2, and calculated the PRODIGY predicted binding energy of sialic acid. The results suggest that sialic acid binds most strongly to pockets 2 and 3 (see Table 2 ). Awasthi and co-workers also demonstrated the ability of other sialic acids (such as 5-N-acetyl neuraminic acid (Neu5Ac), α 2,3-sialyl-N-acetyllactosamine (2,3-SLN), α 2,6-sialyl-N-acetyl-lactosamine (2,6-SLN), 5-N-glycolyl neuraminic acid (Neu5Gc), and sialyl LewisX (sLeX)) to bind into pocket 2 of SARS-CoV-2 using in silico molecular docking (2020). Since pockets 2 and 3 show the most variability across BCoVs and bind sialic acid more strongly than the other pockets we subsequently used LigPlot+ to further analyse the interactions of sialic acid with residues in these pockets. For pocket 2 (see Supplementary Figure 2 ), 3 residues of the sugar binding motif (T76, K77 and R78) are involved in binding with sialic acid. K77 forms a hydrogen bond and hydrophobic interaction with sialic acid, while T76 and R78 form hydrophobic interactions with the compound. For pocket 3, LigPlot showed no polar interactions of the ligand with residues in the sugar-binding motif. Yet, we also observe hydrogen bonds and Van der Waals interactions between sialic acid and the NTD. cavities were identified by CavityPlus and are shown using surface representation and coloured bl (pocket 2) and red (pocket 3) respectively. C) the structure is oriented to show both cavities and w give the PRODIGY predicted binding energy.

) the s. B) blue d we

Since pocket 2 and 3 are the most variable and less frequently studied pockets, we compared these putative sialic binding pockets in SARS-CoV-2 to SARS-CoV and Pangolin-CoV-GX. We used HADDOCK to dock sialic acid into the experimental structures of the NTD domains of these BCoVs. We selected the same binding residues as pocket 2 and 3 in SARS-CoV-2. None of these proteins binds sialic acid as strongly as SARS-CoV-2 (see Table 3 ). This result is not surprising as previous structural analysis by Cheng et al. found pocket 2 of SARS-CoV to be less-defined than other BCoVs (2019). We also compared the LigPlots for sialic acid docked to SARS-CoV and Pangolin-CoV-GX with those for SARS-CoV-2 (See Supplementary Figure 3 and Supplementary Figure 4) . For both pockets, we observe more interactions with sialic acid and more hydrogen bonds formed in SARS-CoV-2 NTD than SARS-CoV NTD. The number of sialic acid interacting residues for Pangolin-CoV-GX and SARS-CoV-2 are similar. 

In order to assess the possible impact of recent variants of concern/interest in relation to these sugar binding pockets, we performed molecular docking of sialic acid into all 3 pockets using the original Wuhan-1 strain and assessed the impacts on the binding for the various mutations in the VOCs and VOIs. We did not consider pocket 4 as this is highly conserved across BCoVs and no VOC variants occurred in this pocket. We list all the NTD mutations/insertions found in the different VOCs/VOIs in Supplementary Table 5 . Many of the mutations/insertions lie close to the 3 known and putative sugar binding pockets discussed in this and related studies.

We report the PRODIGY binding energy for the individual NTD mutations in Supplementary Table  6 . Overall, most of the mutations in the VOCs and VOIs were not predicted to drastically alter or enhance sialic acid binding. However, two mutations, one observed in pocket 1: E154K (found in Kappa variant) and the other observed in pocket 3: T95I (found in Kappa, Delta, Delta Plus (AY.1, AY.2) Iota, Mu and B.1.1.263 variants) were predicted to increase binding affinity (≥0.5 kcal/mol changes), as compared to the original Wuhan-1 strain of SARS-CoV-2. See Supplementary Figure 6 and Supplementary Figure 7 for LigPlots.

We also examined the impact on binding energies for NTD domains in different strains, when all the mutations and indels in the domain were considered (see Table 4 ). For pocket 1, the binding of the B.1.617.2 Delta strain, AY.2 Delta plus strain and two of the Gamma P.1 strains to sialic acid are stronger than the original Wuhan-1 strain (an increase of 0.5 kcal/mol) (See Supplementary Figure 8 for LigPlot). For pocket 2, there is no significant difference among the SARS-CoV-2 strains. However, for pocket 3 we see a significant increase in binding energy, compared to the original SARS-CoV-2, in the following strains of SARS-CoV2: B. Figure 9 for LigPlots). However, it is important to note that these predictions need to be verified by careful experimental work to confirm these effects. 

Structural superpositions of the known (PDB) structures of BCoV NTD domains are shown in Figure  9B . We also modelled the structures of all BCoVs classified in the same superfamily in CATH, using AlphaFold2. Only good quality models (predicted IDDT score >75, See Supplementary Table 3 for details) were used and we superposed these to assess the conservation of the structures (see Figure  9B ). Overall, the conservation of the structures is high with an average SSAP of 87 (out of a maximum score of 100). We also calculated the structure conservation for the pockets. Pocket 1 is conserved with an average SSAP of 83.50. There have been attempts to design drugs for this pocket and chloroquine and hydroxychloroquine have been tested and shown to bind in this pocket [Fantini et al (2020) ]. These drugs have been used for the treatment of malaria but are not being used for Covid-19. On the other hand, the structural conservation for pockets 2 and 3 is low, with average SSAP scores of 65-73.

However, analysis by CavityPlus identified a fourth pocket druggable pocket (CavityPlus Drug Score of 1838, a highly positive DrugScore suggests high druggability), which binds sialic acids less strongly than the other pockets. The structural conservation of this pocket is very high (average SSAP score of 93.80). The high structural conservation across BCoVs and the druggability of this pocket suggest that it could be a good pan coronavirus drug target. Figure 9 . Structure superposition of CoV NTDs. A) comprises the 16 BCoV NTDs in the multip sequence alignment (see Figure 5) . B) constitutes 24 BCoV NTDs from the CATH family. For and B), we used known structures and structural models (built using AlphaFold2). We found a high conserved pocket (in the box) (score of 1838 -highly positive DrugScore constitutes hi druggability) that could be a good pan coronavirus drug target. C) demonstrates this pocket predict by CavityPlus in blue. D) We computed the structural conservation of pockets by calculating t average SSAP score.

Several experimental studies have reported that BCoVs use sialic acid-containing receptors for en into the host cell (Tortorici et al., 2019; Peng et al., 2021) . In the case of SARS-CoV-2, sequence a structure based studies reported critical residues involved in putative sugar binding pockets (pocke and 2) (Behloul et al., 2020; Fantini et al., 2020; Baker et al., 2021) .

We have built on these studies and examined all known and putative binding pockets in the NT domain of SARS-CoV-2 and related coronaviruses. Whilst pockets 1 and 2 previously had detail structural analyses including MD and docking (Awasthi et al., 2020; Fantini et al., 2020) , suggesti sialic acid binding, pocket 3 whilst detected (Di Gaetano et al., 2021) had not been characterised f ltiple or A) ighly high icted g the entry e and ket 1 NTD tailed sting d for sialic acid binding. Furthermore, to our knowledge pocket 4 had not been reported by other groups. We found pocket 2 and 3 to bind sialic acid more strongly than pocket 1. Pocket 3 binds sialic acid as strongly as pocket 2 and like pocket 2 is highly variable across BCoVs. Unlike other studies on these pockets, our study also includes comprehensive analyses of Variants of Concern (VOCs) and Variants of Interest (VOIs) in these pockets. When considering all changes in the VOC, some significant changes in binding energy were linked to pocket 3 (B. 

The NTD of SARS-CoV-2 shares a similar fold to the human Galectin-3 domain (CATH topology (2.60.120), which is known to bind sugars. In this study, we predicted four binding pockets (pockets 1-4) in NTD of SARS-CoV-2, three of which were previously known. The predicted sugar binding 'pocket 1' in SARS-CoV-2, corresponds to the galectin-binding pocket in human Galectin-3. Fantini et al. (2020) identified critical sialic acid-binding residues in this pocket in SARS-CoV-2. Our analyses, comparing SARS-CoV-2 to related coronaviruses, indicates that insertions in NTD of SARS-CoV-2 gave additional loops that extended pocket 1, enhancing contact and binding with sialic acid. Other indel regions that emerged in SARS-CoV-2 NTD co-localise at or near the other putative sugar binding pockets, possibly also tuning the interactions of these pockets with sialic acids.

Pockets 2 and 3 were found to be even more structurally variable, across the BCoVs than pocket 1. These pockets have been shown by our studies and others (Awasthi et al., 2020) to be evolutionary hotspots due to significant indel innovations. Our work highlights how the IR2 indel region can affect both pockets 2 and 3 simultaneously since the loop sits between these 2 pockets. IR2 contains a wellknown sugar binding motif for sialic acids (72-GTNGTKR-78 motif) that is conserved in different BCoVs supporting a functional role for this region in infection (Behloul et al., 2020) . Awasthi et al. (2020) , also showed by docking and simulations using sialiosides and NTD structures of SARS-CoV-2 and MERS-CoV that the flexibility of the loop (formed by 244L-261G in pocket 2) could enable binding to a wide range of sialosides.

Recent experimental studies have suggested that the deletions (at positions 69 and 70) in pocket 3 in the SARS-CoV-2 Alpha variant affect the infectivity of the virus (Kemp et al., 2021; Liu et al., 2021) . Important to note is that these positions (69-70) bridge both pockets 2 and 3. Notably, a recent study in Brazil (Resende et al., 2021) reported communitary transmission of different lineages of the variant of concern (VOC) gamma (P.1), which harbours NTD indels 69-70 in pocket 3 and an insertion at 214, also in pocket 3. These VOCs were thought to be responsible for the widespread transmission of SARS-CoV-2 in Brazil. The same study also reported a new variant of interest (VOI) N.10, emerging from the B.1.1.33, which harbours NTD deletions that map to both pocket 2 (residues 256-258) and pocket 3 (residue 211) and suggests that these newly emerging variants are likely to be more resistant to antibody neutralisation than the parental variants of concern. An even more recent VOC, Omicron, also has mutations and deletions in pocket 2 and 3 suggesting similar issues and reasons for concern.

All of the structural innovations that change the binding pockets may have an important role in altering sialic acid binding thereby changing the infectivity profile. Recent SARS-CoV-2 variants show wide changes in infectivity (Dicken et al., 2021) . The potential role of the NTD in altered infectivity is supported by our binding energy studies which predicted that some VOC variants, particularly in pocket 3, could increase binding to sugars, discussed again in detail below.

Some of these changes could represent an ongoing adaptation of the virus to human cells with their specific sugar modifications. These observations combined with the Dicken et al. study (2021) on increased efficiency of cell entry for the Alpha VOC supports the need for further studies to characterise these binding pockets and monitor any ongoing (pocket 2 and 3) changes as new VOCs emerge.

We used established computational tools (HADDOCK and its scoring tool PRODIGY) to analyse changes in binding energy for different BCoVs binding to sialic acid. We found that the NTD domain in the Spike protein of SARS-CoV-2 binds sialic acid more strongly than SARS-CoV, in sugarbinding pockets 1 to 3. Furthermore, some VOCs and VOIs also bind more strongly in pocket 1 (Kappa variant; E154K) and pocket 3 (T95I found in Delta, Iota, Mu variants), compared to the original Wuhan-1 strain of SARS-CoV-2 and SARS-CoV (see Table 5 ). T95I mutation has been reported to be one of the most prevalent mutations in the new (Delta Plus) strains of the Delta lineage (namely AY.2 or B.1.617.2.1) (Kannan et al., 2021) . It is worth noting that the proximity of mutations to binding pockets may also alter the accessibility to the binding pocket or may alter binding to other regions proximal to the pocket (sialic acids are found as part of bigger macromolecules) even if not directly altering sialic acid binding energy.

We propose continuous monitoring of NTD mutations and indels in the context of newly emerging variants and their impacts on sugar-binding. For example, as regards pocket 1 and 3, our studies found that newly emerging strains such as from Brazil (P.1), South Africa (B.1.1.263 and Omicron B.1.1.529) and Delta plus (AY.2) have enhanced binding energy (as compared to original Wuhan strain) that may be associated with higher levels of transmission. Our structural study helps to rationalise potential effects of new VOCs and provide hypotheses for experiments. For example, the Omicron variant has a rather unique 3 amino acid insertion at position 214 which lies near to pocket-3.

Concern regarding the putative increase in transmission of the Omicron VOC highlights the need for further experimental characterisation of the roles of these NTD pockets in infection. In particular, the dramatic changes in virus entry efficiency achieved by changes to IR2 should be studied further, especially by experimental study and in relation to pockets 2 & 3, to help rationalise likely impacts of emerging VOCs.

We observe high conservation in pocket 1 (galectin-like binding pocket, structural similarity score of 83.50 out of 100) and pocket 4 (score 93.80 out of 100). The druggable nature of pocket 1 is also supported by the study by Fantini et al. (2020) which showed that the drugs chloroquine and hydroxychloroquine fully mimic the way in which the sialic acid binds in pocket 1 of NTD, and in the presence of these drugs SARS-CoV-2 no longer binds to sialic acid. Dually targeting galectins and the sialic acid-binding domain of SARS-CoV-2 has been suggested as a promising strategy for COVID-19 for preventing viral entry and modulating the host immune response (Caniglia et al., 2020) .

In addition to pocket 1, we suggest targeting pocket 4, which is much more highly conserved among coronaviruses. Designing inhibitors that target structurally conserved regions or pockets in the BCoVs, such as the conserved sialic acid binding pockets (1 and 4) found in SARS-CoV-2, seems to be a promising strategy to inhibit protein function and block viral entry (Caniglia et al., 2020) .

Basic local alignment search tool

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

The Sialoside-Binding Pocket of SARS-CoV-2 Spike Glycoprotein Structurally Resembles MERS-CoV

Correction to "The SARS-COV-2 Spike Protein Binds Sialic Acids, and Enables Rapid Detection in a Lateral Flow Point of Care Diagnostic Device

Role of the GTNGTKR motif in the N-terminal receptor-binding domain of the SARS-CoV-2 spike protein

The SARS-CoV-2 spike protein: balancing stability and infectivity

The Protein Data Bank

Exploring the Association Between Sialic Acid and SARS-CoV-2 Spike Protein Through a Molecular Dynamics-Based Approach

Immunopathology of galectin-3: an increasingly promising target in COVID-19

Crystal structure of the S1 subunit N-terminal domain from DcCoV UAE-HKU23 spike protein

More Is Always Better Than One: The N-Terminal Domain of the Spike Protein as Another Emerging Target for Hampering the SARS-CoV-2 Attachment to Host Cells

Characterisation of B.1.1.7 and Pangolin coronavirus spike provides insights on the evolutionary trajectory of SARS-CoV-2. Microbiology

Data, disease and diplomacy: GISAID's innovative contribution to global health: Data, Disease and Diplomacy

Structural and molecular modelling studies reveal a new mechanism of action of chloroquine and hydroxychloroquine against SARS-CoV-2 infection

Predicting the zoonotic capacity of mammals to transmit SARS-CoV-2

D3R Grand Challenge 3: blind prediction of protein-ligand poses and affinity rankings

Spike protein mutations in novel SARS-CoV-2 'variants of concern'commonly occur in or near indels

Advances in the development of entry inhibitors for sialic-acid-targeting viruses

Structural Biology in the Clouds: The WeNMR-EOSC Ecosystem

Human coronaviruses OC43 and HKU1 bind to 9-O-acetylated sialic acids via a conserved receptorbinding site in spike protein domain A

Galectins at a glance

Highly accurate protein structure prediction with AlphaFold

Evolutionary analysis of the Delta and Delta Plus variants of the SARS-CoV-2 viruses

SARS-CoV-2 evolution during treatment of chronic infection

MEGA X: Molecular Evolutionary Genetics Analysis across Computing Platforms

Performance of HADDOCK and a simple contact-based protein-ligand binding affinity predictor in the D3R Grand Challenge 2

An overview of comparative modelling and resources dedicated to large-scale modelling of genome sequences

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor

LigPlot+: multiple ligand-protein interaction diagrams for drug discovery

Molecular epidemiology of human coronavirus OC43 reveals evolution of different genotypes over time and recent emergence of a novel genotype due to natural recombination

Structure, Function, and Evolution of Coronavirus Spike Proteins

An infectivity-enhancing site on the SARS-CoV-2 spike protein targeted by antibodies

The EMBL-EBI search and sequence analysis tools APIs in 2019

Nterminal domain antigenic mapping reveals a site of vulnerability for SARS-CoV-2

In-Silico Evidence for a Two Receptor Based Strategy of SARS-CoV-2

ColabFold -Making protein folding accessible to all. bioRxiv

Cell entry by SARS-CoV-2

Static all-atom energetic mappings of the SARS-Cov-2 spike protein and dynamic stability analysis of "Up" versus "Down" protomer states

UCSF Chimera--a visualisation system for exploratory research and analysis

Distinct Roles for Sialoside and Protein Receptors in Coronavirus Infection

The ongoing evolution of variants of concern and interest of SARS-CoV-2 in Brazil revealed by convergent indels in the amino (N)-terminal domain of the spike protein

Deciphering key features in protein structures with the new ENDscript server

Bioinformatics studies on a function of the SARS-CoV-2 spike glycoprotein as the binding of host sialic acid glycans

Insights on cross-species transmission of SARS-CoV-2 from structural modeling

The structural basis of accelerated host cell entry by SARS-CoV-2 †

Structural basis of receptor recognition by SARS-CoV-2

Statistical potential for assessment and prediction of protein structures

GISAID: Global initiative on sharing all influenza data -from vision to reality

Genome3D: integrating a collaborative data pipeline to expand the depth and breadth of consensus protein structure annotation

CATH: increased structural coverage of functional space

Cryo-EM structure of the SARS coronavirus spike glycoprotein in complex with its host cell receptor ACE2

HH-suite3 for fast remote homology detection and deep protein annotation

Protein structure alignment

Structural basis for human coronavirus attachment to sialic acid receptors

Scoring residue conservation

Largescale prediction of binding affinity in protein-small ligand complexes: the PRODIGY-LIG web server

Complete genomic sequence of human coronavirus OC43: molecular clock analysis suggests a relatively recent zoonotic coronavirus transmission event

Comparative Protein Structure Modeling Using MODELLER

A general empirical model of protein evolution derived from multiple protein families using a maximum-likelihood approach

Coronavirus genomics and bioinformatics analysis. Viruses

Weekly Epidemiological Update on COVID-19

Classification of Omicron (B.1.1.529): SARS-CoV-2 Variant of Concern

COVID-19 weekly epidemiological update on COVID-19 -24

Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation

Structure and binding properties of Pangolin-CoV spike glycoprotein inform the evolution of SARS-CoV-2

Conformational dynamics of SARS-CoV-2 trimeric spike glycoprotein in complex with receptor ACE2 revealed by cryo-EM

CavityPlus: a web server for protein cavity detection with pharmacophore modelling, allosteric site identification and covalent ligand binding ability prediction

Bat and pangolin coronavirus spike glycoprotein structures provide insights into SARS-CoV-2 evolution

2020 update on human coronaviruses: One health, one world

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. 

The Supplementary Material for this article can be found online at:

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.