https://orcid.org/0000-0002-1586-4464
Figures
Abstract
Highly pathogenic H5Ny influenza A viruses are causing unprecedented, season-independent outbreaks across avian and mammalian species, including dairy cattle, a novel reservoir. The sialoside-binding properties of influenza A hemagglutinin (HA) are strongly related to its ability to infect and transmit between hosts. Mucin-like O-glycans, omnipresent in respiratory tracts, have been understudied as viral receptors due to their complexity. To address this, we synthesized 25 O-linked glycans with diverse sialosides, including modifications by fucosides and sulfates. Our findings reveal that H5Ny 2.3.4.4b viruses bind core 3 sialyl-Lewisx and Sia-Gal-β3GalNAc, O-linked glycans not recognized by classical H5 or other avian viruses. By determining crystal structures, we resolved the structural features of four glycans in an H5 hemagglutinin (HA) from a 2016 2.3.4.4b virus. While these viruses do not bind human-type receptors, their broad receptor specificity enhances binding to human tracheal tissues, suggesting that O-glycan recognition could contribute to the continues spillover of this clade.
Author summary
Avian influenza A viruses of the H5Ny family have caused repeated outbreaks in birds, diverse mammals and, in rare cases, humans. A central question is how these viruses evolve the ability to bind to host cells and spread across species. In our study, we show that H5Ny viruses belonging to clade 2.3.4.4 have acquired a previously unrecognized binding specificity; they can attach to mucin-like O-glycans, core 3 sialyl-Lewisx and Sia-Gal-β3GalNAc, which are not bound by classical H5 or other avian influenza strains. This broadened glycan recognition may help these viruses infect novel tissues or host species. Our results highlight the molecular mechanism by which avian influenza viruses increase their adaptability by diversifying the array of cell-surface glycans that they can possibly exploit.
Citation: Weber J, Ponse NLD, Zhu X, Ríos Carrasco M, Han AX, Funk M, et al. (2026) The receptor binding properties of H5Ny influenza A viruses have evolved to bind to avian-type mucin-like O-glycans. PLoS Pathog 22(1): e1013812. https://doi.org/10.1371/journal.ppat.1013812
Editor: Miklos Guttman, University of Washington, Seattle, UNITED STATES OF AMERICA
Received: June 3, 2025; Accepted: December 15, 2025; Published: January 20, 2026
Copyright: © 2026 Weber et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The atomic coordinates and structure factors are being deposited in the Protein Data Bank (PDB) under accession codes 9NRR, 9NRS, 9NRT, 9NRU, and 9NRV for H5FR H5 HA apo-form, and in complex with compounds 26, 25, 7, and LSTa, respectively. All other relevant data are in the paper and its supporting information files.
Funding: JW is supported by the Austrian Science Fund (FWF, J-4260-B21). This research was made possible by funding from ICRAD, an ERA-NET co-funded under the European Union’s Horizon 2020 research and innovation program. (https://ec.europa.eu/programmes/horizon2020/en), under Grant Agreement n°862605 (Flu-Switch) to RPdV. The glycan array setup was supported by the Netherlands Organization for Scientific Research (NWO, TOP-PUNT 718.015.003 to G.-J. B.). MRC and RPdV are supported by an NWO-M2 (OCENW.M20.106). This research was partially funded by the National Institutes of Health NIAID Centers of Excellence for Influenza Research and Response contract 75N93021C00015 / PENN CEIRR (I.A.W.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Influenza A viruses (IAV) are maintained in wild waterfowl and regularly spillover to poultry and, occasionally to mammalian species, including humans [1]. Only a few lineages of influenza A viruses exist in mammalian hosts, such as H3N8 viruses in horses, H1 swine viruses, and pandemic H1N1, H2N2, and H3N2 viruses in humans [2].
The H5 HA clade 2.3.4.4 viruses paired with several NA subtypes including H5N2, H5N6 and H5N8 emerged in 2014 [3,4]. The 2.3.4.4b H5N1 viruses now circulate year-round from the northern to the southern hemisphere and cause high mortality in wild birds [5,6], and regularly infect poultry. They have been transmitted to several species of mammals [7–9], with sustained mammal-to-mammal transmission in mink farms in Europe and South American marine mammals, a first for H5N1 viruses [10]. A recent spillover into dairy cattle represents an exceptional novel viral reservoir in important livestock with occasional transmission to dairy farm workers and domestic animals [11]. With humans being immunologically naïve to the H5 HA in H5N1 and other H5 viruses, it is vital to establish molecular determinants important for the increased spillover of these viruses.
A widely accepted paradigm is that avian viruses require terminal 2,3-linked sialosides (SIAs) as receptors for cell binding and entry, while human IAVs employ α2,6-linked SIAs as receptors [12,13]. Recent studies have indicated that this notion is an oversimplification and other features, such as fucosylation, sulfation, and the length of N-acetyl lactosamine (LacNAc, galactose-β1–4-N-acetylglucosamine) chains, can determine or modulate receptor specificity and may represent barriers to cross-species infection [14,15]. Only one sialyl transferase (ST6Gal1) can transfer terminal α2,6-linked SIAs onto oligosaccharide chains, and this enzyme preferentially modifies LacNAc moieties of N-glycans [16,17]. On the other hand, several sialyl transferases can add α2,3-linked SIAs onto N- and O-linked glycans and glycolipids, resulting in a far greater structural diversity [18]. A range of N-linked glycans has been synthesized and employed for glycan array development, which allowed the investigation of how N-glycan structural complexity can influence receptor specificity [13,19–22]. It has, for example, been found that an N-glycan having an α2,6-sialoside at a di-LacNAc moiety is a commonly employed receptor by human influenza A viruses [23].
Several sialyl transferases can catalyze the addition of α2,3-linked SIAs onto O-glycans, and this type of sialoside is particularly abundant on mucins [24,25]. IAVs have evolved an ability to attach and penetrate the mucus layer of respiratory and enteric tissues [26,27] and several reports indicate that IAVs can bind core 2, 3, and 4 O-glycan structures [28, 29]. There are, however, no systematic investigations of binding specificities of IAVs to biologically relevant O-glycans.
To address this knowledge gap, we developed a chemoenzymatic approach providing a library of 25 O-linked glycopeptides with different cores and terminal epitopes. These compounds were printed as a glycan microarray, which was employed to probe receptor specificities of evolutionary distinct H5 influenza viruses. We focused on H5 influenza viruses of the 2.3.4.4b clade, including a virus representing the HA of a human isolate of the dairy cattle outbreak [30], and examined differences in receptor specificities from their predecessors. The receptor specificity of 2.3.4.4b viruses coincided with binding to human upper respiratory tract tissues, while binding to chicken trachea was maintained. We determined crystal structures of a representative 2.3.4.4b HA with specific mucin-like O-glycans and demonstrated how a fucoside can be accommodated in 2.3.4.4b viruses.
Results
Chemoenzymatic synthesis of a library of O-linked glycopeptides
We embarked on the chemoenzymatic synthesis of a panel of O-linked glycopeptides displaying different core structures and terminal modifications (Fig 1). O-glycan biosynthesis is initiated by adding a single GalNAc residue to Ser/Thr to form a so-called Tn-antigen [31]. The latter moiety can be extended by a β1,3Gal moiety to form core 1, which subsequently can be extended to core 2 by adding β1,6GlcNAc to the GalNAc residue. Core 3 has a β1,3GlcNAc linked to the Tn-antigen. The core structures can be modified by sialosides, blood groups, and Lewis antigens resulting in various epitopes. Core 1 and 2 containing structures are commonly found on glycoproteins and mucins of many different cell types, whereas core 3 display is more restricted, and these structures are mainly found on gastrointestinal tissues [31].
(A) Enzymatic synthesis of core 1 (2), core 2 (3) and extended core 2 O-glycan structures (4-22) (B) Enzymatic synthesis of extended core 3 O-glycans (23- 26).
Compounds 2, 3, and 23 represent core 1, 2, and 3 structures, respectively (Fig 1). We targeted compounds with terminal epitopes that may function as receptors of IAV. They include 2,3-linked sialosides and derivatives with additional fucosylation and/or sulfation to give sialyl Lewisx (SLex) and 6-sulfo-SLex and 6-sulfo-sialyl-LacNAc derivatives. There are data to support that IAV can employ such structures as receptors [15,32–34] and, therefore, were included as target compounds. The attraction of this series of glycopeptides is that these epitopes are presented on different core structures, making it possible to examine whether these differences influence HA binding.
The synthesis of the core 1 and 2 structures started with the assembly of glycopeptide 1 by solid phase peptide synthesis (SPPS) using a Rink amide AM resin, Fmoc protected amino acids, and Fmoc-Thr-(3,4,6-tri-O-acetyl-α-D-GalNAc) (Scheme S1). After assembly, the glycopeptide was cleaved and purified by HPLC employing a reverse-phase C18 column. The N-terminal α-amine of the glycopeptides was left free to facilitate immobilization onto NHS ester-activated microarray slides for glycan array development.
The enzyme glycoprotein-N-acetyl galactosamine 3-β-galactosyltransferase 1 (C1GalT1) can convert a Tn antigen into a core 1 structure (2). It requires an underlying peptide chain, and earlier studies have shown that the peptide sequence FVTIG is appropriate for this enzyme [35]. As anticipated, treatment of 1 with uridine-5´-diphospho-galactose (UDP-Gal) in the presence of recombinant C1GalT1 resulted in the facile formation of glycopeptide 2. The latter compound was converted into core 2 by β-1,6-N-acetylglucosaminyltransferase (GCNT1) and uridine-5´-diphospho-glucosamine (UDP-GlcNAc) (Fig 1A). The resulting glycopeptide 3 was the starting point for the installation of various terminal epitopes. For example, the GlcNAc moiety of 3 could be extended by a β(1,4)-galactoside to give LacNAc-containing derivative 4 by treatment with β-1,4-galactosyltransferase 4 (B4GalT4) and UDP-Gal. The resulting terminal Gal moiety of glycopeptide 4 could selectively be modified as an α2,3-sialoside by α2,3-sialyltransferase 4 (ST3Gal4) and cytidine-5′-monophospho-N-acetylneuraminic acid (CMP-Neu5Ac) to form glycopeptide 5. SLex containing glycopeptide 6 was formed by treatment of 5 with α1,3- fucosyltransferase 5 (FUT5) and guanosine-5´-diphospho-β-l-fucose (GDP-Fuc).
In an alternative sequence of reactions, the β1,3-galactoside of core 2 glycopeptide 3 was selectively α2,3-sialylated to give 7 by treatment with α2,3-sialyltransferase 1 (ST3Gal1) in the presence of CMP-Neu5Ac. This compound could be converted into a variety of structures; for example, the enzyme B4GalT1 could convert the terminal GlcNAc moiety into LacNAc 8, which could be sialylated further by ST3Gal4 to give di-sialylated glycopeptide 9. Treatment of 8 or 9 with FUT5 or resulted in the formation of Lewisx (Lex) and SLex containing compounds 10 and 11, respectively.
The C-6 hydroxyl of the GlcNAc moiety of core 2 structures can be sulfated by GlcNAc-6-O-sulfotransferases 2 and 6 (CHST2 and 6) [36]. This enzyme only modifies terminal GlcNAc moieties, and as anticipated, treatment of 7 with 3′-phosphoadenosine-5′-phosphosulfate (PAPS) in the presence of CHST2 resulted in the facile formation of sulfated glycopeptide 12. The galactosyl transferase B4GalT4 can attach a 1,4-linked galactoside to 6-sulfo-GlcNAc residues and, as expected, could readily convert 12 into 13, which is an appropriate substrate for ST3Gal4 to give glycopeptide 14 having 6-sulfo-sialyl-LacNAc moiety. The latter compound could be fucosylated with FUT6 to provide 6-sulfo-SLex containing glycopeptide 15. A sulfo-sialyl-LacNAc moiety is an appropriate substrate for keratan sulfate galactose 6-sulfotransferase (KSGal6ST or CHST1), which in the presence of PAPS can sulfate the C-6 hydroxyl of the galactoside and this enzyme could readily convert 14 into the di-sulfate 16. Compound 13 was also fucosylated with FUT6 to provide 6-sulfo-Lex containing glycopeptide 17. A similar sequence of reactions starting from 18, which was synthesized by sulfation of 3 with CHST2 and PAPS, provided compounds 19, 20, 22, and 21.
Acetylgalactosaminyl-O-glycosyl-glycoprotein β-1,3-N-acetylglucosaminyltransferase (B3GNT6) is the enzyme that can convert core 1 (2) into a core 3 structure 23 (Fig 1). This enzyme has limited in vitro activity [37] and could not successfully be expressed; therefore, we chemically synthesized appropriately protected GlcNAcGalNAc-Thr (S1 & S3), which was employed in glycopeptide assembly to give the core 3 containing structure 23 (Scheme SA). This compound could be extended by galactosylation by B4GalT1 in the presence of UDP-Gal to form glycopeptide 24. α2,3-Sialylation of 24 was achieved by using ST3Gal4 and CMP-Neu5Ac to obtain 25. Treatment of 25 with FUT5 and GDP-Fuc resulted in the formation of 26 (Fig 1B). An attempt to sulfate core 3 (7) with PAPS and CHST2 or CHST6 failed, and probably these compounds are not appropriate substrates for these enzymes.
Glycan microarray development and screening
The library of 25 O-glycans having various terminal epitopes was printed on amine-reactive NHS-activated glass slides (Fig 2A), resulting in an O-glycopeptide microarray. The resulting array slides were probed by several plant lectins [38], which confirmed spot integrity and proper substrate deposition and revealed fine specificities of glycan-binding proteins examined for O-glycans (Fig S1A). The lectins MAL-I and MAL-II have well-known affinity for α2,3-sialylated glycans and are commonly used for detecting avian IAV receptors on tissues [39]. On the array, MAL-I showed robust binding to α2,3-sialylated LacNAc structures on O-glycan cores 2 and 3 (5, 9, 25) and tolerated a 6- sulfate on GlcNAc (14, 20) (Fig S1B). MAL-II exhibited a more expanded response to sulfated core 2 O-glycans with an α2,3-linked SIA on the β3-branch (12, 13) and to sulfated core 2 di-sialosides (14–16), and the results highlight that the specificities of MAL-I and -II are substantially different. A mouse IgM antibody (CD15s, clone CSLEX1)) was used to detect SLex epitopes, which are present on structures 6 and 11 (Fig S1B). Notably, glycan 26 was not recognized by this IgM. The lectin ECA detected non-sialylated epitopes [40] only binding to terminal type 2 LacNAc structures. It tolerated a 6-sulfate but not fucose on GlcNAc.
Monosaccharides nomenclature is shown. (B) Phylogenetic tree of H5 viruses, including those used in this work, reconstructed temporally using TreeTime. Different colors represent different antigenic clades (C) The specificity of several influenza A H5 viruses, with relative fluorescence units on the y-axis and glycan numbering on the X-axis as in (A). Shown is a representative of two biological replicates, which were tested as two experimental replicates. Standard deviation is from 6 technical replicates with the highest and lowest data points removed.
Additionally, lectins Jacalin and PNA were used to analyze non-sialylated epitopes, specifically terminal Tn-antigen [41,42] (Fig S1C). Interestingly, both lectins recognized different O-glycan structures since Jacalin did not tolerate 6-sulfation on the GlcNAc and recognized 2, 23, and 24. In contrast, PNA more selectively recognized all terminal T-antigen epitopes. Ricin was used to analyze type 2 LacNAc [43], which bound the unmodified terminal type 2 LacNAc. AAL was used to detect fucosylated O-glycans [44] and indeed recognized all fucosylated structures.
H5Ny 2.3.4.4 HAs have a broad, yet specific receptor-binding repertoire
Next, we focused on receptor binding selectivity of different H5Ny viruses (Fig 2B). Clade 2.3.4.4b viruses have accumulated substantial changes in the receptor binding site (RBS) in the 130, 150, 220-loops, and 190-helix (3) (Fig S2). Specifically, positions 132, 133a, 137, 158–160, 193, 196, 222, and 227 in the receptor binding site have changed between classical (viruses isolate before 2010) and current H5 viruses and can affect receptor binding properties [12]; however, amino acid changes conferring human-type receptor binding are absent.
Recombinant viruses that contained the HA (without a multibasic cleavage site) of interest in the background of the attenuated vaccine strain A/Puerto Rico/8/1934 (PR/8) were handled under biosafety level (BSL) 2 conditions in agreement with national regulations on genetically modified viruses (GMO-99–090 and IG 15–160_IIk). All viruses in this study are thus recombinant, but we do refer to them with the full virus name throughout the manuscript. We employed the viruses in this study on the array with an HA titer of 128. We used two pre-2.3.4.4 viruses, A/Indonesia/5/2005 (2.1.3.2) and 2.3.4 A/Anhui/1/2005 (Fig 2B) that are known to bind terminal α2,3-linked SIAs on N-glycans [45,46]; however, detailed binding specificities of O-glycans have not been determined.
The A/Indonesia/5/2005 virus bound to core 2 O-glycans with a sialylated LacNAc containing a 6-sulfated GlcNAc (15–17 and 20–22), with slight preference if these structures displayed 2 sialic acids, compound 9 over 5, 11 over 6, 8 over 9 and 11 over 10 (Fig 2C). The direct precursor 2.3.4 virus (A/Anhui/1/2005) also displays a conventional avian-type (α2,3-linked SIA) receptor binding profile, very similar to A/Indonesia/5/2005, and is not able to bind the unmodified sialylated LacNAc (5, 9, and 25). The A/Anhui/1/2005 virus preferred sialylated LacNAc in which the GlcNAc moiety is sulfated regardless of whether the LacNAc is fucosylated (15, 16, 20–22). Notably, the non-sulfated SLex displays minimal responsiveness (6, 11, and 26). Interestingly, the only Sia-Gal-β3GalNAc structures bound contain a sulfated Lex (17).
We selected several 2.3.4.4 viruses including an early strain from 2014 and a representative of a European strain. Both 2.3.4.4 viruses, A/Sichuan/26221/14 H5N6 and A/European polecat/NL/1/22 H5N1, bind every available α2,3-linked SIA (Fig 2C). A/Sichuan/26221/14 H5N6 represents an older 2.3.4.4a subclade (Fig S3), while the A/European polecat/Netherlands/1/2022 is a 2.3.4.4b virus. The differentiating features between non-2.3.4.4 and 2.3.4.4 viruses include simple mucin-like core 2 O-linked structures displaying Sia-Gal-β3GalNAc epitopes (7, 8–10, 12, and 13) and core 3 O-glycans, with the non-modified core 3 structure 25 efficiently bound as well as 26. Sulfated structures are less critical for 2.3.4.4b H5 viruses, as compounds 5–11 are bound similarly to sulfated structures. Finally, we tested a representative virus from the current bovine outbreak in the US, A/Texas/37/2024. Compared to the classical A/Indonesia/5/2005 and A/Anhui/1/2005 viruses, the binding profile is extended, with high responsiveness to SLex epitopes (6, 11, and 26). However, compared to the early and the European 2.3.4.4b viruses, the A/Texas/37/2024 did not bind modified epitopes equally well when the Gal-β3GalNAc was not sialylated (20 and 21) and hardly bound 25. In summary, 2.3.4.4 viruses have a broadened receptor repertoire compared to their predecessors, exemplifying receptor binding evolution and possibly explaining the spillover events of these viruses.
H5Ny 2.3.4.4 recombinant HAs display near-identical receptor binding specificities compared to viruses
Recombinant HAs are valuable tools for quickly and safely analyzing receptor binding properties of viruses that spill over in different hosts, as they are surrogates for whole viruses. Previously, we have shown that avian-like receptor binding is very similar between recombinant HAs and viruses, with the multivalency effect leading to higher relative fluorescent intensities detected when using viruses [45].
To compare the data obtained using recombinant virus data with those obtained with recombinant protein and confirm the broadened receptor binding specificities, we tested recombinant classical and 2.3.4.4.b HA H5 hemagglutinins. We employed the HAs from A/duck/Mongolia/54/2001 H5N2 [33], A/Vietnam/1203/2004 H5N1 (clade 1) [45], and A/Indonesia/5/2005 H5N1. The latter is identical at the amino-acid level to the recombinant virus HA, which is commonly used in receptor-binding studies [47]. Furthermore, we investigated two HA proteins from the 2.3.4.4 subclade, A/chicken/Kumamoto/1–7/2014 is a representative strain for subclade 2.3.4.4c, and we used a 2.3.4.4b subclade (A/duck/France/161108h/2016, H5FR), in which the multibasic cleavage site was removed [48], which represents currently circulating panzootic viruses in Europe. The relation between the viruses used in this work is shown in the phylogenetic tree in Fig 2B.
We observed that the HAs of the classical H5s preferentially bound core 2 O-glycans with sialylated LacNAc containing a 6-sulfated GlcNAc (14–16, 20 and 22) (Fig 3A and 3B). Indeed, the literature indicates that many avian viruses, including H5, can bind 6-sulfate GlcNAc-containing sialosides [15,49]. Interestingly, structures only displaying Sia-Gal-GalNAc (black bars) are hardly bound by these three HAs; similarly, compounds 25 and 26 are not bound. The extra SIA on the β3-branch is required for binding since glycan 9 is bound, even without a 6-sulfated GlcNAc, in contrast to 5.
(A) The shown O-glycans were printed on a glycan microarray and the terminal epitopes are indicated in different colors. (B) The specificity of classical H5 HAs, (C) A/Indonesia/5/2005 and two 2.3.4.4 H5 proteins and (D) H5FR mutant proteins. Relative fluorescence units are on the y-axis and glycan numbering on the X-axis as in (A). Shown is a representative of two biological replicates that were tested as two experimental replicates, standard deviation from 6 technical replicates with the highest and lowest data points removed.
The A/Indonesia/5/2005 protein (Fig 3C) is identical on the amino acid level compared to the recombinant virus in Fig 2C. The main difference is the lack of binding to 21 and 22, while their counterparts (15 & 16) with an additional SIA are bound by both the virus and protein. The presence of an extra SIA on the β3-branch also allows binding to 9 and 11. 11 has a sLex epitope on the β6-branch, while the presence of solely the SIA on the β3-branch (7, 8, and 10) or the sLex on the β6-branch (5) does not allow for binding for classical H5s. Interestingly, structures 13 and 17 contain a 6-sulfate on the GlcNAc and showed increased binding compared to 7 and 8, indicating that this charged structure is vital for binding. At the same time, this arm lies outside of the RBS as the position of SIA in the RBS is conserved.
The binding preference of HAs of 2.3.4.4 viruses, A/chicken/Kumamoto/1–7/2014 and H5FR is remarkably different. Substantial changes in amino acids are present in the HA1 receptor binding domains of these viruses compared to earlier classical H5s, including A132T, S133aL, S137A, NST158–160NDA, K/R193N, Q196K, K222Q, and S227R (Fig S1) [33,47]. However, none of the canonical changes for α2,3-linked vs. α2,6-linked SIA specificities are present at positions 226 and 228 [12], and the H5FR indeed does not bind human-type receptors [50]. The representative H5 HAs from subclade 2.3.4.4 displayed a broader receptor binding repertoire compared to other H5 HAs, with a nearly identical trend compared to the viruses (Fig 3C). The separating features between classical and 2.3.4.4b H5s include binding to 7, 8, 25 and 26, representing Sia-Gal- β3GalNAc and SLex epitopes. These epitopes have been identified in different cells and organs from various species [29,51,52]. We titrated an H5FR HA protein and a virus, A/European polecat/NL/1/22, on the array to check if we could identify preferred receptors, but none were found (Fig S4). Thus, the host range of these viruses might be explained by the broad binding specificity to the Sia-Gal-β3GalNAc structure or O-glycan Core 3 sLex.
Bovine-specific mutations T160A and T199I do not affect receptor binding specificities when introduced in H5FR HA
Clearly, classical and currently circulating 2.3.4.4 H5N1 viruses have different receptor binding profiles. Indeed, several recent reports demonstrated that contemporary 2.3.4.4 H5N1 viruses have a broadened receptor binding profile [53–55]. Moreover, one report suggested that this broader receptor binding profile compared to their European counterparts is due to T199I [56]. We, however, do not observe differences between the A/Texas/37/2024 virus and its European, A/European polecat/NL/1/22 H5N1, nor its 10-year-old Asian, A/Sichuan/26221/14 H5N6, predecessors. We introduced T199I into H5FR to analyze this discrepancy and determine if this mutant would further increase the binding repertoire (Fig 3D). The mutants displayed a receptor-binding repertoire very similar to the wild type. While circulating in cattle, small groups of viruses have been identified containing T160A, an antigenic change that removes an N-glycosylation site but also affects human-type receptor binding in human-adapted H5N1 viruses [57]. When we introduced the T160A mutations in both the wildtype and T199I mutant, we observed some decrease in binding but no substantial change in binding specificity on the O-glycan array (Fig 3D). Notable diminished binding was observed for compounds 5, 7, 8, 10, 20, 21, and 25, but binding to 6 and 26 is maintained. We conclude that the receptor binding specificity of 2.3.4.4b HAs is conserved.
The broad O-glycan specificity results in binding to human respiratory tract tissues
To test the possibility that the broad receptor binding profile of 2.3.4.4 H5Ny HAs can result in binding to sialosides in the human respiratory tract, we tested their ability to bind chicken and human tracheal tissues as representatives of upper respiratory tract organs. As controls, we analyzed a human H3N2 H3 protein (A/Netherlands/109/2003) that indeed only bound the human trachea tissue (Fig 4). We selected the H5FR HA as it is representive of the RBS of contemporary 2.3.4.4b HA proteins [58]. The H5FR HA protein bound to chicken with high responsiveness, and human tracheal tissue epithelial cells with moderate responsiveness in an SIA-dependent manner. The H5FR fails to bind α2,6-linked SIA containing-receptors [55]; thus, the human respiratory tract might display O-glycans that H5FR can bind. None of the classical H5 proteins could bind human tracheal epithelial cells. They did, however, efficiently bind to the chicken trachea in an SIA-dependent manner as neuraminidase treatment abrogated binding (Fig 4). We also analyzed the A/Texas/37/2024 H5 protein for binding to chicken and human tracheal tissues with near identical results compared to the H5FR (Fig S5). Finally, we determined if the A/Indonesia/5/2005, H5FR and A/Texas/37/2024 H5 proteins and H3N2 H3 protein (A/Netherlands/109/2003) could bind duck and goose tracheal sections. The human H3 protein failed to bind duck and goose trachea, whereas all H5 proteins bound the trachea of both wild bird species (Fig S6 i). Notably, a drop off in signal was detected in the goose trachea from A/Indonesia/5/2005 and the 2.3.4.4b HAs.
The binding to human and chicken tracheal tissue was investigated for different influenza A H5 HAs. The HA from A/NL/109/2003 was used as a positive control for human trachea binding. Sialic acid dependency was shown using tissue slides that were pre-treated with Vibrio cholerae neuraminidase (VCNA). AEC staining was used to visualize tissue binding.
To analyze the presence of SLex in these tissues, we employed the anti-SLex CD15s antibody (Fig S7). Importantly, this antibody does not accommodate sulfated SLex structures (Fig S1) and thus should be an appropriate control for the 2.3.4.4 H5Ny proteins to differentiate between SLex and their sulfated counterpart bound by non-2.3.4.4 HA proteins. The CD15s antibody bound efficiently to the chicken trachea [33] but failed to attach to the human trachea. Thus, the human trachea does not display canonical SLex epitopes, although structure 26 (core 3 SLex) is not recognized by the CD15s antibody (Fig S1) and can still be present. To further confirm our findings, we analyzed a variety of other H5 HAs (Fig S7) and used human trachea from another donor (Fig S8). Again, only the human H3 control and the H5FR bound human tracheal epithelial tissues.
Crystal structures of 2.3.4.4b H5FR H5 HA in complex with core 3 compounds 25 and 26, core 2 compound 7, as well as avian receptor analog LSTa
To understand the expanded receptor binding repertoire of H5 2.3.4.4b HA, we determined crystal structures of H5FR H5 HA in apo-form (1.94 Å resolution) and in complex with O-linked glycan compounds 26 (2.50 Å), 25 (1.98 Å) and 7 (2.40 Å), as well as avian receptor analog LSTa (NeuAcα2–3Galβ1–3GlcNAcβ1–3Galβ1–4Glc) (2.90 Å) (Figs 5 and S9 and Table S1).
(A) Core 3 compound 26 bound to H5FR HA. (B) Core 3 compound 25 bound to H5FR HA. (C) Core 2 compound 7 bound to H5FR HA. (D) Avian receptor analog LSTa bound to H5FR HA. (E) Avian receptor analog 3´-SLN bound to VN1194 H5 HA from A/Vietnam/1194/2004 (H5N1) (PDB ID 4BGY). (F) Superimposition of compound 25 (cyan) and 3´-SLN (grey, from PDB ID 4BGY) in the HA RBS.
To resolve the molecular determinants of SLex specificity for 2.3.4.4b viruses, we started with compound 26. Compound 26 shares the same structure as 25 but with an additional fucose 5 (Fuc-5) attached to GlcNAc-3 (Figs 1 and S9). Indeed, Fuc-5 mediates additional hydrogen bonds with Q222 and the main-chain carbonyl of G225 (Fig 5A) but not with R227, although it was proposed that Q222 and R227 confer sLex specificity [33,59]. While G225 is conserved between H5FR and H5VN H5 HAs, H5VN HA has K222 (Fig S2). These additional interactions between Fuc-5 and H5FR HA enhance the binding of 26; we even observed weak density for the threonine (T3) in peptide linker moiety of the glycopeptide (Fig S9A).
H5FR H5 HA undergoes little conformational change upon binding to other glycans; for example, when 25 binds, only very slight movements or rotamer differences of side chains, such as Gln226 and Leu194, are observed in the RBS (Figs S10 and S9B). The O-linked glycans 25 and 26 share a common terminal sequence SIA 1 (Sia-1), galactose 2 (Gal-2), and N-acetylglucosamine 3 (GlcNAc-3) with the avian receptor analog 3´-SLN (NeuAcα2–3Galβ1–4GlcNAc), while compound 7 shares the first two monosaccharides Sia-1 and Gal-2 with 3´-SLN (Fig 1). For 25 in complex with H5FR HA, the conformations of Sia-1, Gal-2 and GlcNAc-3 are very similar to 3´-SLN in its complex with VN1194 H5 HA from A/Vietnam/1194/2004 (H5N1) (Fig 5). VN1194 H5 HA and H5VN HA would share the same receptor specificity as the only amino acid difference between these two HAs is located in the HA stem, and is HA1 T46 (H3 numbering) for VN1194 H5 HA and K46 for H5VN HA [60]. In the H5FR HA–25 complex, 25 adopts a trans configuration in the Siaα2–3Gal linkage (Fig 5B), similar to 3´-SLN in its complex with VN1194 H5 HA [61] (Fig 5E and 5F). Compound 25 Sia-1 and Gal-2 form hydrogen bonds with RBS A136, A137 and E190, while the N-acetylgalactosamine 4 (GalNAc-4 that connects to Thr or Ser in an O-linked glycan) projects out of the RBS and the peptide linker residues FVTIG (Phe-Val-Thr-Ile-Gly) are not visible in the electron density map (Figs 5B and S9B). Previously, we determined a crystal structure of human Sh2 H7 HA from A/Shanghai/2/2013 (H7N9) with a similar avian O-linked glycan, glycan #21 (NeuAcα2–3Galβ1–4GlcNAcβ1–3GalNAcα-Thr), which shares the same glycan structure with 25 but with Thr only for the peptide linker (Fig S11) [50]. In contrast, glycan #21 adopts a cis configuration in the Siaα2–3Gal linkage with Sh2 H7 HA (Fig S11A), and the receptor exits over the 220 loop, facilitating contacts between GlcNAc-3 and HA Q222 of Sh2 H7 HA (Fig S11A). The different binding modes of the O-linked glycan (Fig S11B) highlight the diversity in HA-glycan recognition, which is not predictable without determining crystal structures.
Compound 7 is a branched glycan (Fig 1) with one arm containing Sia-1, Gal-2, and GalNAc-3, and the other arm has only GlcNAc, which is not modeled due to poor electron density (Fig S9C). Similar to 26 and 25, the overall binding of 7 to H5FR HA is through Sia-1 and Gal-2, with no substantial glycan-HA interactions visualized for GalNAc-3 (Fig 5C).
As a positive control, LSTa without GalNAc, a commonly used avian receptor analog for structural and functional studies, was investigated (Fig 5D). Sia-1, Gal-2 and GlcNAc-3 were ordered in LSTa-H5FR HA complex structure (Figs 5D and S9D), while in the LSTa complex with VN1194 H5 HA (PDB 3ZP0, 2.51 Å resolution), only Sia-1 and Gal-2 were visible [62]. Structural comparison of H5FR HA with LSTa versus H5FR HA in complex with 26, 25, and 7 reveals the similar key interactions between glycan and H5FR HA through Sia-1 and Gal-2 (Fig 5).
In summary, Fuc-5 of compound 26 directly interacts with the H5FR HA, facilitating stronger binding than compound 25 (Fig 5A and 5B). Compounds 25 and 7 bind H5FR HA mainly through interactions of Sia-1 and Gal-2 (Fig 5B and 5C). RBS residue differences between different H5 HA strains, such as residue 222, as well as 137, 193, and 227 that differ between 2.3.4.4 H5 strains and other classical H5 strains (Fig S2), need to be further interrogated to determine their roles in binding specificity of these O-glycans.
The peptide linker FVTIG of glycopeptides binds to the H5FR H5 HA stem domain and head domain in the crystal structures
Unexpectedly, when we determined crystal structures of H5FR HA with glycan compounds 25, 26, and 7 by soaking H5FR HA crystals in these glycans at 5 mM for 10 minutes, strong additional density was observed in the H5FR HA stem domain for all three glycan complexes and in the head domain with the 25 complex (Fig S12). The substantial densities were from the binding of these glycopeptides to the H5 HA (Figs S12D, G and J) via their peptide linker FVTIG moiety (Fig 1). The peptide FVTIG in 25, 26, and 7 was found to bind H5FR HA stem region with the same constellation of HA residues in each structure (Figs S12D, E, G, H, J and K). Among key peptide-HA interactions, the peptide V2 main chain forms two hydrogen bonds with the side chain of HA2 N53, and peptide G5 hydrogen bonds to the side chain of HA1 H38. Peptide residue V2 also interacts with HA1 M292 and HA2 V52 and I56, and the peptide I4 makes hydrophobic interactions with HA2 W21, I45 and V48. The H5FR HA contacting residues to peptide FVTIG are fully conserved in H5VN H5 HA and overlap with the CR9114 epitope in H5VN H5 HA except for HA2 N53, which does not contact CR9114 [63].
The peptide linker FVTIG in 25 was also found to bind H5FR HA1 head residues K165 to T171 primarily through main-chain hydrogen bond interactions of peptide V2 with HA1 K165 and S167, peptide I4 with HA1 N169, as well as peptide G5 with the HA1 side chain T171 (Fig S12B and C). The lack of apparent electron density for peptide FVTIG from 26 and 7 binding to H5FR HA head may be attributed to the weaker binding affinity of the peptide to the HA head than to the HA stem. The potential impact on HA function by peptide binding to HA1 fragment 165–171 needs to be further investigated (Figs 3 and S12).
In glycan array analysis, the peptide moiety of FVTIG might not directly bind HA due to the method of glycan immobilization. The peptide with a primary amine on one end was immobilized onto NHS ester-activated microarray slides in the glycan array and was not freely available for HA binding. In HA crystal soaking, the FVTIG moieties of glycan compounds are accessible, as we observe in the complex structures (Fig S12). To estimate the binding affinity of the peptide FVTIG to HA, we conducted surface plasmon resonance (SPR) binding analysis of H5FR HA to core 1 glycan compound 2. Glycopeptide 2 contains the FVTIG moiety but lacks SIA (Fig 1), which renders no binding to H5FR HA in our glycan array studies (Fig 2). In our SPR analysis in which the FVTIG moiety of 2 is accessible, the binding of 2 to H5FR HA was weak as the estimated apparent KD value was greater than 650 μM (Fig S13).
A/black swan/Akita/1/2016 2.3.4.4e needs two amino acid mutations to confer binding to human tracheal epithelial cells
In the previous glycan array and tissue binding experiments, we employed an H5 HA of the 2.3.4.4e virus A/black swan/Akita/1/2016. This virus contains a representative RBS for 2.3.4.4e viruses, with a clade-specific deletion at position L133aΔ and an R227Q mutation (Figs S2 and S3). This HA did not bind broadly as it did not recognize Sia-Gal-β3GalNAc and core 3 SLex structures efficiently (Fig 6A) and failed to bind human tracheal epithelial cells. At the same time, it was biologically active, as evidenced by its classical avian-type receptor binding profile on the glycan array and binding to the chicken trachea. The clade-specific L133aΔ and R227Q mutations are directly in the RBS, and we considered whether a double mutant towards the H5FR sequence (Δ133aL/ Q227R) would convert to broad glycan binding properties and result in binding to human tracheal tissues. Indeed, the double mutant, Δ133aL/ Q227R, was able to bind Sia-Gal-β3GalNAc and core 3 SLex, similar to 2.3.4.4b HA proteins and viruses, and bound human tracheal tissues (Fig 6B). Thus, distinct amino acids on position 133a and 227 are essential for the receptor binding phenotypes of 2.3.4.4b H5 viruses.
(A) Binding of the A/black swan/Akita/1/2016 to the O-glycan array as shown in Fig 2A was investigated. Relative fluorescence units are on the y-axis and glycan numbering on the X-axis as in Fig 3A. Shown is a representative of two biological replicates which were tested as two experimental replicate. The standard deviation is from 6 technical replicates with the highest and lowest data points removed. Binding of this HA to chicken and human tracheal tissue was also investigated. AEC staining was used to visualize tissue binding. (B) The same assays were performed with the Δ133aL and Q227R mutant of A/black swan/Akita/1/2016.
Sia-Gal-6-S-GlcNAc as a universal receptor for avian HAs
Next, we focused on the O-glycan specificities of a wide variety of IAV HAs, as only a handful of studies have previously investigated the involvement of O-glycans as receptors [26]. H3N8 is a subtype that has been able to infect many avian and mammalian species, including humans [64,65]. We therefore compared the glycan binding specificities of H3N8 isolates from different species: A/duck/Ukraine/1/1963, A/harbor seal/Massachusetts/1/2011, A/equine/Miami/1/1963, and A/canine/Texas/1/2004 [22,45] (Fig 7A and 7B). Again the 6-sulfate on GlcNAc is important as the di-sialosides 14–16 are bound by all H3 proteins (Fig 7B). Interestingly, both the duck and seal HAs had the broadest receptor binding profile bound to some Sia-Gal-β3GalNAc epitopes, if a sulfate was present (12, 13, and 17), but with low responsiveness. Both HAs also displayed binding to a non-modified Sia-LacNAc (9) and SLex (11). The equine and canine strains displayed more restricted binding profiles by not binding any Sia-Gal-β3GalNAc epitope. The canine isolate is bound more broadly compared to the equine as it can bind structures 20-22, which are the same as 14–16, except for the additional Sia in the Sia-Gal-β3GalNAc epitope. Analyzing two other IAV HAs of the H1 (A/duck/Hokkaido/111/2009 H1N5) and H4 (A/duck/Hokkaido/138/2007 H4N6) subtypes further consolidate the preference for 6S-GlcNAc containing structures, such as 14–16, as these displayed the highest responsiveness (Fig 7C). These receptor binding profiles are more similar to classical H5s and A/duck/Ukraine/1/1963, as some Sia-Gal- β3GalNAc epitopes are bound with a sulfated non-sialylated β6-arm (12, 13, and 17). Furthermore, non-modified Sia-LacNAc (9) was bound as well. Next, we focused on two H6 isolates described with differential specificity to 6-sulfated GlcNAc structures [34]. The A/chicken/Tainan/V156/1999 H6N1 HA entirely depended on the presentation of sulfated sLex in di-sialosides (Fig 7D), while the A/duck/Hong Kong/960/1980 H6N2 HA was able to bind non-modified Sia-LacNAc (9) and SLex (11). In conclusion, the receptor binding properties of 2.3.4.4 H5 viruses are unique in their receptor binding properties.
(A) The shown O-glycans were printed on a glycan microarray and the terminal epitopes are indicated in different colors. (B) The binding of several H3 HAs, (C) H1, H4, and (D) H6 HAs were determined. Relative fluorescence units are on the y-axis and glycan numbering on the X-axis as in (A). Shown is a representative of two biological replicates which where tested as two experimental replicates. The standard deviation is from 6 technical replicates with the highest and lowest data points removed.
Discussion
In this study, we created an O-glycan array by combining chemical and enzymatic synthesis. We synthesized core 1, 2, and 3 O-glycans encompassing various sialylated LacNAc structures with modifications such as sulfates and fucosides. We discovered that the Sia-Gal- β3GalNAc epitope in core 2 O-glycans and core 3 SLex are bound by contemporary 2.3.4.4b H5Ny viruses, whereas all other HAs tested failed to do so. The distinct O-glycan specificity of the widespread 2.3.4.4 H5Ny viruses may be relevant for the host range and the observed gain of binding to human tracheal tissues. Such a specificity likely arose during continuous circulation in wild bird species in which the SLex epitope is particularly common [66].
Our findings on the binding of distinct O-glycans emphasize the need to develop new methods to create O-glycans. Reported enzymes for extending O-glycans are often of microbial origin as they are generally easier to express than mammalian enzymes. However, microbial glycosyltransferases that operate early in the O antigen pathway require sugar-diphosphate-lipids as acceptors [37], which were unsuitable in our synthesis since we aimed to synthesize glycopeptides that were printable on the glycan microarray slides. For instance, the microbial transferase WbwC can add a galactose on a GalNAc-α-diphosphate-lipid acceptor but not on GalNAc-α-peptides, and we could not utilize this enzyme. Instead, we used the corresponding mammalian enzyme C1GalT1, which acts specifically on GalNAc-α-peptides.
O-glycans are present in several glycan arrays that evaluate IAV receptor binding [46,59,67,68]. However, these are often not discussed in detail, and it is difficult to get a proper overview of IAV receptor binding specificities to these glycan cores [29]. Here, we show the importance of investigating O-glycans as potential receptors for IAV. For contemporary circulating H5N1 viruses, we not only demonstrated increased binding to O-glycans on the array, but we also observed binding of H5 HAs to human tracheal tissue, not only of H5FR but also for A/Texas/37/2024, and H5AK with mutations Δ133aL and Q227R. Similar human respiratory tissue binding properties have been described previously [55,69]. Indicating that binding to human tracheal epithelial cells could also be due to binding α2,3 sialosides on O-glycans.
The Sia-Gal-β3GalNAc structures, which we show to be bound by 2.3.4.4b viruses, are uniquely found in O-linked glycans [29,51,52], whereas SLex epitopes are also found in N-linked structures. The ability of some H5 viruses to bind sLex has been shown previously, but it was thought not to be essential [59]. Here, we demonstrate that core 3 O-glycans (25 and 26) are unique as they are only bound by 2.3.4.4b H5Ny viruses, indicating that the β3-linkage to the core GalNAc is essential. H5FR H5 HA from clade 2.3.4.4b H5N8 virus binds to glycans 25 and LSTa similarly to VN1194 HA or H5VN HA from clade 1 H5N1 viruses, and subtle amino acid differences around RBS and beyond may determine the binding specificity to O-glycans, such as Q222 in H5FR HA that might be critical for its enhanced binding to compound 26 (Fig 3). This specific O-glycan is present in the human respiratory tract [70], but how it precisely relates to the spillover properties of 2.3.4.4b H5Ny viruses remains unknown. Furthermore, it is unknown if neuraminidases can cleave sialic acids on these O-glycan cores. Thus, studies into neuraminidase specificity and how SLex binding relates to an increased host range are warranted.
There have been few changes in the RBS of 2.3.4.4b H5Ny viruses over the past years. Notably, several hallmark mutations that are described that confer avian- to human-type receptor specificity with E190D and G225D for H1 and Q226L and G228S for H3 [71], are not present in contemporary H5 viruses. However, the introduction of the H3N2 pandemic came with additional amino acid mutations [72], and literature suggests that H5 proteins also need three or more amino acid mutations to bind human-type receptors [46,57,73,74]. We recently showed that 2.3.4.4e only needs the single Q226L mutation to bind human-type receptors on linear structures [58]. It has also been recently demonstrated that the A/Texas/37/2024 virus HA protein, identical to the virus used in this study (except for the multibasic cleavage site), only requires the single Q226L mutant to bind human-type receptors presented on complex N-glycans [75]. Human-type receptor binding was specific to branched poly-LacNAc containing symmetrical N-glycans, rare in the human upper respiratory tract [29,76]. Significantly, however, all previous studies have been limited in their use of biologically relevant O-glycans.
In conclusion, synthesizing biologically relevant O-glycans with diverse LacNAc modifications led us to the discovery that the current 2.3.4.4 H5Ny influenza A viruses tolerate a variety of sialylated glycans. Thus, while these viruses maintain an overall avian-type receptor specificity, their apparent spillover properties in multiple avian and mammalian hosts can potentially be explained by their novel O-glycan specificities. Importantly, this study is limited in looking at binding patterns to fixed tissue specimens only, and other functional assays would provide additional important information.
Materials and methods
Expression and purification of trimeric influenza A hemagglutinins
Recombinant trimeric IAV hemagglutinin (HA) ectodomain proteins were cloned into the pCD5 expression vector (an example is Addgene plasmid #182546 [45]) in frame with a GCN4 trimerization motif (KQIEDKIEEIESKQKKIENEIARIKK), a super folder GFP [77] or mOrange2 [78] and the Twin-Strep-tag (WSHPQFEKGGGSGGGSWSHPQFEK); IBA, Germany). Mutations in HAs were generated by site-directed mutagenesis. The trimeric HAs were expressed in HEK293S GnTI(-) cells with polyethyleneimine I (PEI) in a 1:8 ratio (µg DNA:µg PEI) for the HAs as previously described [79]. The transfection mix was replaced after 6 hours by 293 SFM II suspension medium (Invitrogen, 11686029), supplemented with sodium bicarbonate (3.7 g/L), Primatone RL-UF (3.0 g/L, Kerry, NY, USA), glucose (2.0 g/L), glutaMAX (1%, Gibco), valproic acid (0.4 g/L) and DMSO (1.5%). According to the manufacturer’s instructions, culture supernatants were harvested five days post-transfection and purified with sepharose strep-tactin beads (IBA Life Sciences, Germany).
Glycan microarray binding studies
All compounds were printed on NHS-ester activated glass slides (NEXTERION Slide H, Schott Inc.) using a Scienion sciFLEXARRAYER S3 non-contact microarray equipped with a Scienion PDC80 nozzle (Scienion Inc.). Individual compounds were dissolved in sodium phosphate buffer (250 mM, pH 8.5) at a concentration of 100μM and were printed in replicates of 6 with spot volume ∼400pL, at 20°C and 50% humidity. Each slide has 24 subarrays in a 3x8 layout. After printing, slides were incubated in a humidity chamber for 8 h and then blocked for 1 h with a 50mM ethanolamine in Tris buffer (pH 9.0, 100 mM) at 50°C. Blocked slides were rinsed with DI water, spun dry, and kept in a desiccator at room temperature for future use.
Viral isolates diluted in PBS-T (PBS + 0.1% Tween) to an HA titer of 128 were applied to subarrays in the presence of oseltamivir (200 nM) in a humidified chamber for 1 h. Next, the microarray slide was rinsed with PBS-T (PBS + 0.1% Tween), PBS, and deionized water (2x) and dried by centrifugation. The slide was incubated for 1 h in the presence of the CR6261 stem specific antibody (100 µL, 5 µg mL − 1 in PBS-T) and washed as described above. A secondary goat anti-human AlexaFluor-647 antibody (100 µL, 2 µg mL − 1 in PBS-T) was applied and the resulting slide was incubated for 1 h in a humidified chamber and then washed by the standard procedure. HAs (50 µg/ml) were pre-complexed with human anti-streptag and goat anti-human-Alexa555 (#A21433, Thermo Fisher Scientific) antibodies in a 4:2:1 molar ratio, respectively, in 50 µL PBS with 0.1% Tween-20. Biotinylated lectins (5 µg/ml) were pre-complexed with streptavidin-Alexa555 (#S32355, Thermo Fisher Scientific) in a 5:1 weight ratio. The following biotinylated lectins from Vector Laboratories were used: MAL-I (B-1315–2), MAL-II (B-1265–1), ECA (B-1145–5), Jacalin (B-1155–5), PNA (B-1075–5), Ricin (B-1085–5), and AAL (B-1395–1). Siglec-9 (50µg/mL) was precomplexed with pA-LS at a 1:1 mass ratio. The mixtures were incubated on ice for 15 minutes and then on the array surface for 90 minutes in a humidified chamber. Then, slides were rinsed successively with PBS-T (0.1% Tween-20), PBS, and deionized water. For the anti-sialyl-Lewisx antibody CD15S (clone CSLEX1, #551344, BD Biosciences), a solution of 50 µg/ml in 40 µL PBS with 0.1% Tween-20 was first incubated on the slide for 90 min in a humidified chamber. After washing successively with PBS-T (0.1% Tween-20), PBS, and deionized water, a mixture of 10 µg/ml goat anti-mouse IgM-HRP (#1021-05, Southern Biotech) and 5 µg/ml donkey anti-goat IgG-Alexa555 (#A21432, Thermo Fisher Scientific) in 40 µL PBS with 0.1% Tween-20 was incubated on the slide for 90 min in a humidified chamber. Afterward, the slides were rinsed successively with PBS-T (0.1% Tween-20), PBS, and deionized water. Slides were dried by centrifugation after the washing steps and scanned immediately using an Innopsys Innoscan 710 microarray scanner. Various gains and PMT values were used to ensure the signals were in the linear range and to avoid saturation of the signals. Images were analyzed with Mapix software (version 8.1.0 Innopsys) and processed with an Excel macro (https://github.com/enthalpyliu/carbohydrate-microarray-processing). The average fluorescence intensity and SD were determined for each compound after removal of the highest and lowest intensities from the spot replicates to give n = 4.
Expression, crystallization, and crystal structure determination of H5FR H5 HA and its glycan complexes
The ectodomain of H5FR H5 HA was expressed in a baculovirus system essentially as previously described [80,81]. Briefly, the cDNA corresponding to residues 11–327 of HA1 and 1–174 of HA2 (H3 numbering) of wildtype H5FR H5 HA from A/duck/France/161108h/2016 (H5N8) (GISAID accession number EPI869809) was incorporated into a baculovirus vector, pFastbacHT-A (Invitrogen) with an N-terminal gp67 signal peptide, a C-terminal foldon trimerization domain, and a His6-tag, as well as a thrombin cleavage site between the HA ectodomain and the foldon trimerization domain. The constructed plasmid was used to transform DH10bac competent bacterial cells, and the purified recombinant bacmids were used to transfect Sf9 insect cells for overexpression. HA protein was produced in suspension cultures of High Five insect cells. Soluble HA protein was harvested in supernatants and purified by metal affinity chromatography using Ni-nitrilotriacetic acid (Ni-NAT) resin (Qiagen). For crystallization, the foldon trimerization domain and His6-tag were cleaved from the HA ectodomain with thrombin, and the HA ectodomain was purified further by size-exclusion chromatography on a Hiload 16/90 Superdex 200 column (GE Healthcare) in 20 mM Tris, pH 8.0, 150 mM NaCl and 0.02% NaN3.
Crystallization experiments were set up using the sitting drop vapor diffusion method on our robotic CrystalMation system (Rigaku). For structure determination of the apo-form H5FR H5 HA and its complex with glycan compounds 7 and 25, purified protein at 11.1 mg/ml was crystallized in 0.1 M Tris, pH 7.0, 0.11 M lithium sulfate, and 2.0 M ammonium sulfate at 20°C. For structure determination of the H5FR HA with compound 26, the purified HA at 11.1 mg/ml was crystallized in 0.1 M sodium citrate, pH 5.6, 0.2 M potassium sodium tartrate, and 2.0 M ammonium sulfate at 20°C. For structure determination of the H5FR HA in complex with compound LSTa (NeuAcα2–3Galβ1–3GlcNAcβ1–3Galβ1–4Glc), purified H5FR HA at 11.1 mg/ml was crystallized in 1.6 M sodium citrate, pH 6.5 at 20°C. H5FR HA-ligand complexes were obtained by soaking HA crystals in the well solution containing ligands and cryo-protectants. The final concentrations of compounds 7, 25, 26, and LSTa were all 5 mM, and the soaking times were all about 10 min. The crystals were flash-cooled at 100K in mother liquid, and 15% ethylene glycol (v/v) was added as a cryoprotectant for apo-form H5FR HA, and its complex with compounds 7 and 25, and 20% glycerol (v/v) was added as a cryoprotectant for H5FR HA in complex with compounds 26 and LSTa.
Diffraction data were collected at Stanford Synchrotron Radiation Lightsource (SSRL) beamline 12–1 (Table S1). Data were integrated and scaled with HKL3000 [82]. Data and refinement statistics are summarized in Table S1 .
The crystal structure of H5FR H5 HA was determined by molecular replacement (MR) using the program Phaser [83] with the crystal structure of H5 HA from A/duck/Eastern China/L0230/2010 (H5N2) (PDB code 7DEB) as input MR model. Initial rigid body refinement was performed in REFMAC5 [84], and further restrained refinement, including TLS refinement, was carried out in Phenix [85]. The model building between rounds of refinement was carried out with Coot [86]. The final statistics are summarized in Table S1. The structure quality was analyzed using the JCSG validation suite (qc-check.usc.edu/QC/qc_check.pl) and MolProbity [87]. All structural figures were generated with PyMol (www.pymol.org).
Surface Plasmon Resonance (SPR) binding analysis of H5FR H5 HA and glycan 2
SPR was conducted on BiaCore S200 instrument (Cytiva Inc.). High-density poly-NTA HC hydrogel chips (NiHC1500M, XanTec) were used to capture polyhistidine-tagged HA protein. The protein immobilization level reached more than 20,000 RU in the sample flow cell. All experiments were performed at 25°C, with PBS, pH 7.4 as the running buffer. 1 mM of glycan 2 solution was prepared by more than 100 times dilution from stock solution with running buffer. Further sample concentrations were prepared by serial dilution four times from a 1 mM solution. Samples were injected over the reference paired 2 flow cells for 120 s followed by 240 s dissociation time. Five sample concentrations were injected, with 3 cycles of zero concentration in between to remove residual samples left on the chip surface or protein pocket. All signals were reference corrected, and the negative sample was subtracted before analysis. Signals from 5 s before injection stop were used for steady-state analysis.
Protein histochemical tissue staining
Sections of formalin-fixed, paraffin-embedded chicken (Gallus gallus domesticus) were obtained from the Division of Pathology, Department of Biomolecular Health Sciences, Faculty of Veterinary Medicine of Utrecht University, the Netherlands. Sections of formalin-fixed, paraffin-embedded human tissues were obtained from the UMC Utrecht, Department of Pathology, Utrecht, the Netherlands (TCBio-number 22–599). In the figures, representative images of at least two individual experiments are shown. Protein histochemistry was performed as previously described [88,89]. In short, tissue sections of 4 µm were deparaffinized and rehydrated, after which antigens were retrieved by heating the slides in 10 mM sodium citrate (pH 6.0) for 10 min. Endogenous peroxidase was inactivated using 1% hydrogen peroxide in MeOH for 30 min at RT. When neuraminidase treatment was performed, slides were incubated overnight at 37°C with neuraminidase from Vibrio cholerae (#11080725001, Roche) diluted 1:50 in a solution of 10mM potassium acetate and 0.1% Triton X-100 at pH 4.2. Non-treated slides in experiments with neuraminidase were incubated with only buffer. Tissues were blocked at 4°C using 3% BSA (w/v) in PBS for at least 90 minutes. For plant lectin stains, blocking was performed using carbo-free blocking solution (SP-5040–125; Vector Laboratories, Burlingame, CA, USA) instead. Subsequently, slides were stained for 90 minutes using anti- sLex antibody or pre-complexed HAs as previously described for the glycan microarray. For A/duck/France/161108h/2016, 2 µg/ml of H5 HA was used, while for the other H5 HAs, 2.5 µg/ml was used. For A/NL/109/2003, 5.0 µg/ml H3 HA were used. For antibody control images, only antibodies and no HAs have been used. The detection of sLex was performed with the anti- sLex antibody CD15S (clone CSLEX1, #551344, BD Biosciences), which was diluted 1:1000 in PBS and precomplexed with goat anti-mouse IgM-HRP (#1021-05, Southern Biotech) in a 1:100 dilution.
H5 HA phylogenetic trees
We downloaded all available influenza H5Ny HA sequences and metadata that were available up until 1 February 2025 from the Global Initiative on Sharing All Influenza Data (GISAID) (https://gisaid.org/) EpiFlu database. Low-quality (>1% ambiguous nucleotide) and incomplete sequences (<95% of HA segment length) were removed. Besides the viruses studied in this work, we randomly subsampled 5–20 representative sequences per H5 clade as assigned by GISAID. We aligned sequences using MAFFT v7.520 [90], and reconstructed the maximum-likelihood phylogenetic tree with IQ-TREE v2.2.2.6 [91], using the general time reversible (GTR) model. We then reconstructed the temporally resolved phylogenies using TreeTime v0.11.4 [92]. We used Baltic (https://github.com/evogytis/baltic/tree/master) python library to produce all tree visualizations.
Supporting information
S1 Scheme. Overview of chemical synthesis of compounds 1 and 23 and intermediates.
https://doi.org/10.1371/journal.ppat.1013812.s001
(TIF)
S1 Table. Data collection and refinement statistics.
https://doi.org/10.1371/journal.ppat.1013812.s002
(DOCX)
S1 Fig. O-glycan characterization using plant lectins
A) The O-glycans were printed on a glycan microarray, and the terminal epitopes are indicated in different colors. (B) O-glycan binding of plant lectins MAL-I, MAL-II (both binding α2,3 linked SIA), ECA (binding terminal LacNAc), and a mouse IgM antibody, CD15s, clone CSLEX1 (binding SLex) (C) The binding of plant lectins Jacalin [binds to terminal T-antigen (Gal-β1,3-GalNAc)]2, PNA (peanut agglutinin, binds to binds terminal T-antigen [Gal-β1,3-GalNAc)3, Ricin (binds to terminal type 2 LacNAc)4, and AAL (Aleuria aurantia lectin, binds to fucose)5.
https://doi.org/10.1371/journal.ppat.1013812.s003
(JPG)
S2 Fig. HA RBS amino acid alignment of the H5 hemagglutinins used in this study
Alignment of the RBS residues with amino acid positions (H3 numbering) indicated above the alignment, non-conserved residues highlighted in a black background, dots indicating identical amino acids. 2024, dairy cattle outbreak specific amino acid changes, are highlighted with a gray background.
https://doi.org/10.1371/journal.ppat.1013812.s004
(JPG)
S3 Fig. Phylogenetic tree of 2.3.4.4.b H5 HAs.
The viruses that were studied in this work are indicated.
https://doi.org/10.1371/journal.ppat.1013812.s005
(JPG)
S4 Fig. Titration of HA and virus on the glycan microarray.
Top panel is the O-linked glycan ligands on the array and bottom panels are the binding of these ligands to both HA and virus titrated from 40 to 2.5 µg/mL and 16–1 hemagglutination units. For reference, in Fig 2, 50 µg/mL and 32 HAU units were used respectively.
https://doi.org/10.1371/journal.ppat.1013812.s006
(JPG)
S5 Fig. The HA of A/Texas/37/2024 binds human tracheal tissue.
The binding to human and chicken tracheal tissue was investigated for A/Texas/37/24 H5 HA. AEC staining was used to visualize tissue binding.
https://doi.org/10.1371/journal.ppat.1013812.s007
(JPG)
S6 Fig. HA binding to duck and goose trachea.
The binding to duck and goose tracheal tissue was investigated for A/NL/109/03, A/Indonesia/5/05, A/duck/France161108h/16, and A/Texas/37/24 H5 HA. AEC staining was used to visualize tissue binding.
https://doi.org/10.1371/journal.ppat.1013812.s008
(JPG)
S7 Fig. The HA of A/duck/France/161108h/2016 H5N8 is the only H5 HA that binds human tracheal tissue.
The binding to human and chicken tracheal tissue was investigated for different influenza A H5 HAs. The HA from A/NL/109/2003 was used as a positive control for human trachea binding. The antibody CD15S was used to visualize sialyl-Lewisx epitopes. AEC staining was used to visualize tissue binding
https://doi.org/10.1371/journal.ppat.1013812.s009
(JPG)
S8 Fig. Tissue binding analyses with additional human donor trachea.
Binding experiments with several influenza H5 HAs, anti-sialyl-Lewisx antibody, and human tracheal tissues (from a different patient than in Fig 4) were performed. The H3 HA from A/NL/109/03 was used as a positive control. AEC staining was used to visualize tissue binding.
https://doi.org/10.1371/journal.ppat.1013812.s010
(JPG)
S9 Fig. Simulated annealing omit Fo-Fc electron density maps of glycan compounds bound in H5FR H5 HA RBSs.
(A) Compound 26 (2.50 Å resolution). (B) Compound 25 (1.98 Å resolution). (C) Compound 7 (2.40 Å resolution). (D) Glycan LSTa (2.90 Å resolution). Some residues from the peptide linker moeity (FVTIG) of the glycan compounds (Fig 1) could also be modeled, albeit into weak electron density. The compounds are colored with cyan carbon atoms. Simulated annealing omit Fo-Fc maps are represented in grey mesh and contoured at 2.5 σ.
https://doi.org/10.1371/journal.ppat.1013812.s011
(JPG)
S10 Fig. Structural comparison of apo-form H5FR H5 HA and its complex with glycan compound 25.
The apo HA is in pink carbon atoms, the complexed HA in yellow carbon atoms, and the ligand in cyan carbon atoms. The side chains of Q226 and L194 with very minor conformational changes are shown for comparison to the unaltered backbones.
https://doi.org/10.1371/journal.ppat.1013812.s012
(JPG)
S11 Fig. Structural comparison of human Sh2 H7N9 HA in complex with avian O-linked glycan #21 with H5FR H5 HA in complex with glycan compound 25.
Avian O-linked glycan #21 (NeuAca2-3Galb1-4GlcNAcb1-3GalNAca-Thr) shares the same glycan structure with compound 25, but with Thr only in the peptide linker. (A) Glycan #21 bound to human H7 HA (PDB ID 4N63). (B) Superposition of human H7 HA with #21 compared to H5FR HA with compound 25. The H7 HA is in pink carbon atoms with ligand #21 in grey carbon atoms, and the H5FR HA in yellow carbon atoms with ligand 25 in cyan carbon atoms.
https://doi.org/10.1371/journal.ppat.1013812.s015
(JPG)
S12 Fig. Crystal structures of H5FR H5 HA binding sites outside of RBS for glycan compounds 25, 26 and 7 via the peptide linker moiety (FVTIG).
(A) to (E) H5FR H5 HA in complex with compound 25. In all of these figures, compound binding to the RBS has been removed for clarity. One HA protomer is colored with HA1 in green and HA2 in cyan and the other two protomers in grey. Compound 25 is colored in yellow carbon atoms, and the electron density (simulated annealing omit Fo-Fc map) for the glycopeptide is represented in a grey mesh and contoured at 2.5 σ. The same coloring scheme is used throughout this figure. (A) Overall structure of H5FR HA with 25. (B) Compound 25 also binds to a non-RBS site in the apex of the HA head domain. (C) Electron density map (as indicated above) for 25 in the HA head. (D) Compound 25 also binds to the HA stem domain. (E) Electron density map for 25 in the HA stem. (F) to (H) H5FR HA with compound 26 bound to the HA stem. (F) Overall structure of H5FR HA with 26. (G) Compound 26 interaction in the HA stem. (H) Electron density map for 26. (I) to (K) H5FR HA with compound 7 bound to the HA stem. (I) Overall structure of H5FR with 7. Compound 7 interacts with the HA stem. Electron density map for 7.
https://doi.org/10.1371/journal.ppat.1013812.s013
(JPG)
S13 Fig. SPR analysis of glycan compound 2 binding to H5FR H5 HA protein.
Signals (represented as response units, RU) were detected when serially diluted compound 2 solution was flowed through HA-immobilized sensor chips. Signals were recorded over time. Different colored lines represent different concentrations of 2, as indicated in the legend. KD value was estimated from steady-state affinity analysis. Since the upper end of the fitting curve did not saturate, we can only estimate the value of KD > 650 μM.
https://doi.org/10.1371/journal.ppat.1013812.s014
(JPG)
Acknowledgments
We thank Takahiro Hiono, Masatoshi Okamatsu, Yoshihiro Sakoda (Hokkaido University), and Romain Volmer (Université de Toulouse) for generously sharing reagents.
We gratefully acknowledge all data contributors, i.e., the authors and their originating laboratories responsible for obtaining the specimens and their submitting laboratories for generating the genetic sequence and metadata and sharing via the GISAID Initiative, on which part of this research is based.
We thank Henry Tien for automated robotic crystal screening. The SSRL is a Directorate of SLAC National Accelerator Laboratory and an Office of Science User Facility operated for the U.S. Department of Energy Office of Science by Stanford University. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research, and by the National Institutes of Health, National Institute of General Medical Sciences (including P41GM103393) and the National Center for Research Resources (P41RR001209).
References
- 1. Joseph U, Su YCF, Vijaykrishna D, Smith GJD. The ecology and adaptive evolution of influenza A interspecies transmission. Influenza Other Respir Viruses. 2017;11(1):74–84. pmid:27426214
- 2. Yoon S-W, Webby RJ, Webster RG. Evolution and ecology of influenza A viruses. Curr Top Microbiol Immunol. 2014;385:359–75. pmid:24990620
- 3. Lewis NS, Banyard AC, Whittard E, Karibayev T, Al Kafagi T, Chvala I, et al. Emergence and spread of novel H5N8, H5N5 and H5N1 clade 2.3.4.4 highly pathogenic avian influenza in 2020. Emerg Microbes Infect. 2021;10(1):148–51. pmid:33400615
- 4. Xie R, Edwards KM, Wille M, Wei X, Wong S-S, Zanin M, et al. The episodic resurgence of highly pathogenic avian influenza H5 virus. Nature. 2023;622(7984):810–7. pmid:37853121
- 5. Ariyama N, Pardo-Roa C, Muñoz G, Aguayo C, Ávila C, Mathieu C, et al. Highly pathogenic avian influenza A(H5N1) clade 2.3.4.4b virus in wild birds, Chile. Emerg Infect Dis. 2023;29(9):1842–5. pmid:37487166
- 6. Yang J, Zhang C, Yuan Y, Sun J, Lu L, Sun H, et al. Novel avian influenza virus (H5N1) clade 2.3.4.4b reassortants in migratory birds, China. Emerg Infect Dis. 2023;29(6):1244–9. pmid:37209677
- 7. Floyd T, Banyard AC, Lean FZX, Byrne AMP, Fullick E, Whittard E, et al. Encephalitis and death in wild mammals at a rehabilitation center after infection with highly pathogenic avian influenza A(H5N8) virus, United Kingdom. Emerg Infect Dis. 2021;27(11):2856–63. pmid:34670647
- 8. Hiono T, Kobayashi D, Kobayashi A, Suzuki T, Satake Y, Harada R, et al. Virological, pathological, and glycovirological investigations of an Ezo red fox and a tanuki naturally infected with H5N1 high pathogenicity avian influenza viruses in Hokkaido, Japan. Virology. 2023;578:35–44. pmid:36462496
- 9. Agüero M, Monne I, Sánchez A, Zecchin B, Fusaro A, Ruano MJ, et al. Highly pathogenic avian influenza A(H5N1) virus infection in farmed minks, Spain, October 2022. Euro Surveill. 2023;28(3):2300001. pmid:36695488
- 10. Peacock TP, Moncla L, Dudas G, VanInsberghe D, Sukhova K, Lloyd-Smith JO, et al. The global H5N1 influenza panzootic in mammals. Nature. 2025;637(8045):304–13. pmid:39317240
- 11. Burrough ER, Magstadt DR, Petersen B, Timmermans SJ, Gauger PC, Zhang J, et al. Highly pathogenic avian influenza A(H5N1) clade 2.3.4.4b virus infection in domestic dairy cattle and cats, United States, 2024. Emerg Infect Dis. 2024;30(7):1335–43. pmid:38683888
- 12. Paulson JC, de Vries RP. H5N1 receptor specificity as a factor in pandemic risk. Virus Res. 2013;178(1):99–113. pmid:23619279
- 13. Byrd-Leotis L, Gao C, Jia N, Mehta AY, Trost J, Cummings SF, et al. Antigenic pressure on H3N2 influenza virus drift strains imposes constraints on binding to sialylated receptors but not phosphorylated glycans. J Virol. 2019;93(22):e01178-19. pmid:31484755
- 14. Broszeit F, van Beek RJ, Unione L, Bestebroer TM, Chapla D, Yang J-Y, et al. Glycan remodeled erythrocytes facilitate antigenic characterization of recent A/H3N2 influenza viruses. Nat Commun. 2021;12(1):5449. pmid:34521834
- 15. Gambaryan A, Yamnikova S, Lvov D, Tuzikov A, Chinarev A, Pazynina G, et al. Receptor specificity of influenza viruses from birds and mammals: new data on involvement of the inner fragments of the carbohydrate chain. Virology. 2005;334(2):276–83. pmid:15780877
- 16. Joziasse DH, Schiphorst WE, van den Eijnden DH, van Kuik JA, van Halbeek H, Vliegenthart JF. Branch specificity of bovine colostrum CMP-sialic acid: N-acetyllactosaminide alpha 2----6-sialyltransferase. Interaction with biantennary oligosaccharides and glycopeptides of N-glycosylproteins. J Biol Chem. 1985;260(2):714–9. pmid:3881425
- 17. Liu L, Prudden AR, Capicciotti CJ, Bosman GP, Yang J-Y, Chapla DG, et al. Streamlining the chemoenzymatic synthesis of complex N-glycans by a stop and go strategy. Nat Chem. 2019;11(2):161–9. pmid:30532014
- 18. Mohamed KA, Kruf S, Büll C. Putting a cap on the glycome: dissecting human sialyltransferase functions. Carbohydr Res. 2024;544:109242. pmid:39167930
- 19. Jia N, Byrd-Leotis L, Matsumoto Y, Gao C, Wein AN, Lobby JL, et al. The human lung glycome reveals novel glycan ligands for influenza A virus. Sci Rep. 2020;10(1):5320. pmid:32210305
- 20. Canales A, Sastre J, Orduña JM, Spruit CM, Pérez-Castells J, Domínguez G, et al. Revealing the specificity of human H1 influenza A viruses to complex N-glycans. JACS Au. 2023;3(3):868–78. pmid:37006776
- 21. Peng W, de Vries RP, Grant OC, Thompson AJ, McBride R, Tsogtbaatar B, et al. Recent H3N2 viruses have evolved specificity for extended, branched human-type receptors, conferring potential for increased avidity. Cell Host Microbe. 2017;21(1):23–34. pmid:28017661
- 22. Spruit CM, Zhu X, Tomris I, Ríos-Carrasco M, Han AX, Broszeit F, et al. N-glycolylneuraminic acid binding of avian and equine H7 influenza A viruses. J Virol. 2022;96(5):e0212021. pmid:35044215
- 23. Ma S, Liu L, Eggink D, Herfst S, Fouchier RAM, de Vries RP, et al. Asymmetrical biantennary glycans prepared by a stop-and-go strategy reveal receptor binding evolution of human influenza A viruses. JACS Au. 2024;4(2):607–18. pmid:38425896
- 24. Delaveris CS, Webster ER, Banik SM, Boxer SG, Bertozzi CR. Membrane-tethered mucin-like polypeptides sterically inhibit binding and slow fusion kinetics of influenza A virus. Proc Natl Acad Sci U S A. 2020;117(23):12643–50. pmid:32457151
- 25. Linden SK, Sutton P, Karlsson NG, Korolik V, McGuckin MA. Mucins in the mucosal barrier to infection. Mucosal Immunol. 2008;1(3):183–97. pmid:19079178
- 26. Mayr J, Lau K, Lai JCC, Gagarinov IA, Shi Y, McAtamney S, et al. Unravelling the role of O-glycans in influenza A virus infection. Sci Rep. 2018;8(1):16382. pmid:30401951
- 27. Nakamura S, Horie M, Daidoji T, Honda T, Yasugi M, Kuno A, et al. Influenza A virus-induced expression of a GalNAc transferase, GALNT3, via microRNAs is required for enhanced viral replication. J Virol. 2015;90(4):1788–801. pmid:26637460
- 28. Nycholat CM, McBride R, Ekiert DC, Xu R, Rangarajan J, Peng W, et al. Recognition of sialylated poly-N-acetyllactosamine chains on N- and O-linked glycans by human and avian influenza A virus hemagglutinins. Angew Chem Int Ed Engl. 2012;51(20):4860–3. pmid:22505324
- 29. Walther T, Karamanska R, Chan RWY, Chan MCW, Jia N, Air G, et al. Glycomic analysis of human respiratory tract tissues and correlation with influenza virus infection. PLoS Pathog. 2013;9(3):e1003223. pmid:23516363
- 30. Caserta LC, Frye EA, Butt SL, Laverack M, Nooruzzaman M, Covaleda LM, et al. Spillover of highly pathogenic avian influenza H5N1 virus to dairy cattle. Nature. 2024;634(8034):669–76. pmid:39053575
- 31. Brockhausen I, Wandall HH, Hagen KGT, Stanley P. O-GalNAc glycans. In: Varki A, Cummings RD, Esko JD, Stanley P, Hart GW, Aebi M, et al., editors. Essentials of glycobiology. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press. p. 117–28. (Cold Spring Harbor Laboratory Press Copyright 2022 The Consortium of Glycobiology Editors, La Jolla, California; published by Cold Spring Harbor Laboratory Press; doi:10.1101/glycobiology.4e.10. All rights reserved.)
- 32. Gambaryan AS, Tuzikov AB, Pazynina GV, Desheva JA, Bovin NV, Matrosovich MN, et al. 6-sulfo sialyl Lewis X is the common receptor determinant recognized by H5, H6, H7 and H9 influenza viruses of terrestrial poultry. Virol J. 2008;5:85. pmid:18652681
- 33. Hiono T, Okamatsu M, Igarashi M, McBride R, de Vries RP, Peng W, et al. Amino acid residues at positions 222 and 227 of the hemagglutinin together with the neuraminidase determine binding of H5 avian influenza viruses to sialyl Lewis X. Arch Virol. 2016;161(2):307–16. pmid:26542967
- 34. Kikutani Y, Okamatsu M, Nishihara S, Takase-Yoden S, Hiono T, de Vries RP, et al. E190V substitution of H6 hemagglutinin is one of key factors for binding to sulfated sialylated glycan receptor and infection to chickens. Microbiol Immunol. 2020;64(4):304–12. pmid:31943329
- 35. Marcos NT, Pinho S, Grandela C, Cruz A, Samyn-Petit B, Harduin-Lepers A, et al. Role of the human ST6GalNAc-I and ST6GalNAc-II in the synthesis of the cancer-associated sialyl-Tn antigen. Cancer Res. 2004;64(19):7050–7. pmid:15466199
- 36. Wu Y, Bosman GP, Chapla D, Huang C, Moremen KW, de Vries RP, et al. A biomimetic synthetic strategy can provide keratan sulfate I and II oligosaccharides with diverse fucosylation and sulfation patterns. J Am Chem Soc. 2024;146(13):9230–40. pmid:38494637
- 37. Brockhausen I. Crossroads between bacterial and mammalian glycosyltransferases. Front Immunol. 2014;5:492. pmid:25368613
- 38. Bojar D, Meche L, Meng G, Eng W, Smith DF, Cummings RD, et al. A useful guide to lectin binding: machine-learning directed annotation of 57 unique lectin specificities. ACS Chem Biol. 2022;17(11):2993–3012. pmid:35084820
- 39. Nicholls JM, Chan RWY, Russell RJ, Air GM, Peiris JSM. Evolving complexities of influenza virus and its receptors. Trends Microbiol. 2008;16(4):149–57. pmid:18375125
- 40. Iglesias JL, Lis H, Sharon N. Purification and properties of a D-galactose/N-acetyl-D-galactosamine-specific lectin from Erythrina cristagalli. Eur J Biochem. 1982;123(2):247–52. pmid:6978812
- 41. Kabir S, Daar AS. The composition and properties of jacalin, a lectin of diverse applications obtained from the jackfruit (Artocarpus heterophyllus) seeds. Immunol Invest. 1994;23(3):167–88. pmid:8070846
- 42. Lotan R, Skutelsky E, Danon D, Sharon N. The purification, composition, and specificity of the anti-T lectin from peanut (Arachis hypogaea). J Biol Chem. 1975;250(21):8518–23. pmid:811657
- 43. Klamer Z, Haab B. Combined analysis of multiple glycan-array datasets: new explorations of protein-glycan interactions. Anal Chem. 2021;93(31):10925–33. pmid:34319080
- 44. Kochibe N, Furukawa K. Purification and properties of a novel fucose-specific hemagglutinin of Aleuria aurantia. Biochemistry. 1980;19(13):2841–6. pmid:7397108
- 45. Broszeit F, Tzarum N, Zhu X, Nemanichvili N, Eggink D, Leenders T, et al. N-Glycolylneuraminic acid as a receptor for influenza A viruses. Cell Rep. 2019;27(11):3284-3294.e6. pmid:31189111
- 46. Peng W, Bouwman KM, McBride R, Grant OC, Woods RJ, Verheije MH, et al. Enhanced human-type receptor binding by ferret-transmissible H5N1 with a K193T mutation. J Virol. 2018;92(10):e02016–7.
- 47. Eggink D, Spronken M, van der Woude R, Buzink J, Broszeit F, McBride R, et al. Phenotypic effects of substitutions within the receptor binding site of highly pathogenic avian influenza H5N1 virus observed during human infection. J Virol. 2020;94(13):e00195-20. pmid:32321815
- 48. Bessière P, Figueroa T, Coggon A, Foret-Lucas C, Houffschmitt A, Fusade-Boyer M, et al. Opposite outcomes of the within-host competition between high- and low-pathogenic H5N8 avian influenza viruses in chickens compared to ducks. J Virol. 2022;96(1):e0136621. pmid:34613804
- 49. Gambaryan AS, Tuzikov AB, Pazynina GV, Webster RG, Matrosovich MN, Bovin NV. H5N1 chicken influenza viruses display a high binding affinity for Neu5Acalpha2-3Galbeta1-4(6-HSO3)GlcNAc-containing receptors. Virology. 2004;326(2):310–6. pmid:15302215
- 50. Xu R, de Vries RP, Zhu X, Nycholat CM, McBride R, Yu W, et al. Preferential recognition of avian-like receptors in human influenza A H7N9 viruses. Science. 2013;342(6163):1230–5. pmid:24311689
- 51. Chan RWY, Karamanska R, Van Poucke S, Van Reeth K, Chan IWW, Chan MCW, et al. Infection of swine ex vivo tissues with avian viruses including H7N9 and correlation with glycomic analysis. Influenza Other Respir Viruses. 2013;7(6):1269–82. pmid:24001121
- 52. Jia N, Barclay WS, Roberts K, Yen H-L, Chan RWY, Lam AKY, et al. Glycomic characterization of respiratory tract tissues of ferrets: implications for its use in influenza virus infection studies. J Biol Chem. 2014;289(41):28489–504. pmid:25135641
- 53. Chopra P, Ray SD, Page CK, Shepard JD, Kandeil A, Jeevan T, et al. Receptor-binding specificity of a bovine influenza A virus. Nature. 2025;640(8059):E21–7. pmid:40240861
- 54. Pulit-Penaloza JA, Belser JA, Brock N, Kieran TJ, Sun X, Pappas C, et al. Transmission of a human isolate of clade 2.3.4.4b A(H5N1) virus in ferrets. Nature. 2024;636(8043):705–10. pmid:39467572
- 55. Song H, Hao T, Han P, Wang H, Zhang X, Li X, et al. Receptor binding, structure, and tissue tropism of cattle-infecting H5N1 avian influenza virus hemagglutinin. Cell. 2025;188(4):919-929.e9. pmid:39848246
- 56. Good MR, Fernández-Quintero ML, Ji W, Rodriguez AJ, Han J, Ward AB, et al. A single mutation in dairy cow-associated H5N1 viruses increases receptor binding breadth. Nat Commun. 2024;15(1):10768. pmid:39737954
- 57. de Vries RP, Zhu X, McBride R, Rigter A, Hanson A, Zhong G, et al. Hemagglutinin receptor specificity and structural analyses of respiratory droplet-transmissible H5N1 viruses. J Virol. 2014;88(1):768–73. pmid:24173215
- 58. Ríos Carrasco M, Lin T-H, Zhu X, Gabarroca García A, Uslu E, Liang R, et al. The Q226L mutation can convert a highly pathogenic H5 2.3.4.4e virus to bind human-type receptors. Proc Natl Acad Sci U S A. 2025;122(16):e2419800122. pmid:40232794
- 59. Guo H, de Vries E, McBride R, Dekkers J, Peng W, Bouwman KM, et al. Highly pathogenic influenza A(H5Nx) viruses with altered H5 receptor-binding specificity. Emerg Infect Dis. 2017;23(2):220–31. pmid:27869615
- 60. Zhu X, Guo Y-H, Jiang T, Wang Y-D, Chan K-H, Li X-F, et al. A unique and conserved neutralization epitope in H5N1 influenza viruses identified by an antibody against the A/Goose/Guangdong/1/96 hemagglutinin. J Virol. 2013;87(23):12619–35. pmid:24049169
- 61. Xiong X, Coombs PJ, Martin SR, Liu J, Xiao H, McCauley JW, et al. Receptor binding by a ferret-transmissible H5 avian influenza virus. Nature. 2013;497(7449):392–6. pmid:23615615
- 62. Crusat M, Liu J, Palma AS, Childs RA, Liu Y, Wharton SA, et al. Changes in the hemagglutinin of H5N1 viruses during human infection--influence on receptor binding. Virology. 2013;447(1–2):326–37. pmid:24050651
- 63. Dreyfus C, Laursen NS, Kwaks T, Zuijdgeest D, Khayat R, Ekiert DC, et al. Highly conserved protective epitopes on influenza B viruses. Science. 2012;337(6100):1343–8. pmid:22878502
- 64. Yang R, Sun H, Gao F, Luo K, Huang Z, Tong Q, et al. Human infection of avian influenza A H3N8 virus and the viral origins: a descriptive study. Lancet Microbe. 2022;3(11):e824–34. pmid:36115379
- 65. Collins PJ, Vachieri SG, Haire LF, Ogrodowicz RW, Martin SR, Walker PA, et al. Recent evolution of equine influenza and the origin of canine influenza. Proc Natl Acad Sci U S A. 2014;111(30):11175–80. pmid:25024224
- 66. Guan M, Olivier AK, Lu X, Epperson W, Zhang X, Zhong L, et al. The sialyl lewis X glycan receptor facilitates infection of subtype H7 avian influenza A viruses. J Virol. 2022;96(19):e0134422. pmid:36125302
- 67. Watanabe T, Kiso M, Fukuyama S, Nakajima N, Imai M, Yamada S, et al. Characterization of H7N9 influenza A viruses isolated from humans. Nature. 2013;501(7468):551–5. pmid:23842494
- 68. Maines TR, Chen L-M, Van Hoeven N, Tumpey TM, Blixt O, Belser JA, et al. Effect of receptor binding domain mutations on receptor binding and transmissibility of avian influenza H5N1 viruses. Virology. 2011;413(1):139–47. pmid:21397290
- 69. Bauer L, Leijten L, Iervolino M, Chopra V, van Dijk L, Power M, et al. A 2022 avian H5N1 influenza A virus from clade 2.3.4.4b attaches to and replicates better in human respiratory epithelium than a 2005 H5N1 virus from clade 2.3.2.1. bioRxiv. 2024:2024.2011.2027.625596.
- 70. Yu H, Gonzalez-Gil A, Wei Y, Fernandes SM, Porell RN, Vajn K, et al. Siglec-8 and Siglec-9 binding specificities and endogenous airway ligand distributions and properties. Glycobiology. 2017;27(7):657–68. pmid:28369504
- 71. Matrosovich M, Tuzikov A, Bovin N, Gambaryan A, Klimov A, Castrucci MR, et al. Early alterations of the receptor-binding properties of H1, H2, and H3 avian influenza virus hemagglutinins after their introduction into mammals. J Virol. 2000;74(18):8502–12. pmid:10954551
- 72. West J, Röder J, Matrosovich T, Beicht J, Baumann J, Mounogou Kouassi N, et al. Characterization of changes in the hemagglutinin that accompanied the emergence of H3N2/1968 pandemic influenza viruses. PLoS Pathog. 2021;17(9):e1009566. pmid:34555124
- 73. Herfst S, Schrauwen EJA, Linster M, Chutinimitkul S, de Wit E, Munster VJ, et al. Airborne transmission of influenza A/H5N1 virus between ferrets. Science. 2012;336(6088):1534–41. pmid:22723413
- 74. Imai M, Watanabe T, Hatta M, Das SC, Ozawa M, Shinya K, et al. Experimental adaptation of an influenza H5 HA confers respiratory droplet transmission to a reassortant H5 HA/H1N1 virus in ferrets. Nature. 2012;486(7403):420–8. pmid:22722205
- 75. Lin T-H, Zhu X, Wang S, Zhang D, McBride R, Yu W, et al. A single mutation in bovine influenza H5N1 hemagglutinin switches specificity to human receptors. Science. 2024;386(6726):1128–34. pmid:39636969
- 76. Liang R, Peccati F, Ponse NLD, Uslu E, de Rooij AJH, Han AX, et al. Epistasis in the receptor-binding domain of contemporary H3N2 viruses that reverted to bind sialylated di-LacNAc repeats. Cell Rep. 2025;44(8):116007. pmid:40679911
- 77. Nemanichvili N, Tomris I, Turner HL, McBride R, Grant OC, van der Woude R, et al. Fluorescent trimeric hemagglutinins reveal multivalent receptor binding properties. J Mol Biol. 2019;431(4):842–56. pmid:30597163
- 78. Shaner NC, Lin MZ, McKeown MR, Steinbach PA, Hazelwood KL, Davidson MW, et al. Improving the photostability of bright monomeric orange and red fluorescent proteins. Nat Methods. 2008;5(6):545–51. pmid:18454154
- 79. de Vries RP, de Vries E, Bosch BJ, de Groot RJ, Rottier PJM, de Haan CAM. The influenza A virus hemagglutinin glycosylation state affects receptor-binding specificity. Virology. 2010;403(1):17–25. pmid:20441997
- 80. Zhu X, Viswanathan K, Raman R, Yu W, Sasisekharan R, Wilson IA. Structural basis for a switch in receptor binding specificity of two H5N1 hemagglutinin mutants. Cell Rep. 2015;13(8):1683–91. pmid:26586437
- 81. Zhu X, Han J, Sun W, Puente-Massaguer E, Yu W, Palese P, et al. Influenza chimeric hemagglutinin structures in complex with broadly protective antibodies to the stem and trimer interface. Proc Natl Acad Sci U S A. 2022;119(21):e2200821119. pmid:35594401
- 82. Minor W, Cymborowski M, Otwinowski Z, Chruszcz M. HKL-3000: the integration of data reduction and structure solution--from diffraction images to an initial model in minutes. Acta Crystallogr D Biol Crystallogr. 2006;62(Pt 8):859–66. pmid:16855301
- 83. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007;40(Pt 4):658–74. pmid:19461840
- 84. Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr. 1997;53(Pt 3):240–55. pmid:15299926
- 85. Liebschner D, Afonine PV, Baker ML, Bunkóczi G, Chen VB, Croll TI, et al. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct Biol. 2019;75(Pt 10):861–77. pmid:31588918
- 86. Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of Coot. Acta Crystallogr D Biol Crystallogr. 2010;66(Pt 4):486–501. pmid:20383002
- 87. Williams CJ, Headd JJ, Moriarty NW, Prisant MG, Videau LL, Deis LN, et al. MolProbity: more and better reference data for improved all-atom structure validation. Protein Sci. 2018;27(1):293–315. pmid:29067766
- 88. Bouwman KM, Parsons LM, Berends AJ, de Vries RP, Cipollo JF, Verheije MH. Three amino acid changes in avian coronavirus spike protein allow binding to kidney tissue. J Virol. 2020;94(2):e01363-19. pmid:31694947
- 89. Wickramasinghe INA, de Vries RP, Gröne A, de Haan CAM, Verheije MH. Binding of avian coronavirus spike proteins to host factors reflects virus tropism and pathogenicity. J Virol. 2011;85(17):8903–12. pmid:21697468
- 90. Katoh K, Misawa K, Kuma K, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002;30(14):3059–66. pmid:12136088
- 91. Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, von Haeseler A, et al. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era. Mol Biol Evol. 2020;37(5):1530–4. pmid:32011700
- 92. Sagulenko P, Puller V, Neher RA. TreeTime: maximum-likelihood phylodynamic analysis. Virus Evol. 2018;4(1):vex042. pmid:29340210