Targeting virus–glycan recognition in influenza viruses and coronaviruses: from the molecular principles to glycomimetic antivirals strategies

Abstract

Glycans constitute the first molecular landscape encountered by viruses during infection and play fundamental roles in viral attachment, host adaptation, tissue tropism, and immune evasion. Far from being passive components of the cell surface, host glycans actively regulate virus–host interactions by functioning as attachment factors, entry receptors, or modulators of receptor engagement, while viral glycans contribute to immune escape and infectivity. Consequently, the molecular recognition of glycans has emerged as a major determinant of viral evolution and pathogenicity. This review presents our perspective on the structural and mechanistic principles that govern virus–glycan recognition, focusing on influenza viruses and coronaviruses as paradigmatic examples. We discuss how subtle variations in glycan structure, including glycosidic linkage, chain length, multivalent presentation, and chemical modifications such as O-acetylation, shape viral receptor specificity, host range, and cross-species transmission. Recent advances from structural biology, NMR spectroscopy, glycan arrays, computational modeling, and glycochemistry have substantially expanded our understanding of these dynamic recognition events and their evolutionary consequences. We further highlight how these insights are driving the development of glycomimetic antivirals designed to interfere with the earliest stages of infection. Approaches based on multivalent sialoglycan analogues, stabilized carbohydrate mimetics, and heparan sulfate-inspired scaffolds illustrate the growing potential of targeting glycan-mediated recognition as an alternative or complementary antiviral strategy. By integrating molecular recognition, viral adaptation, and medicinal chemistry, this review underscores the opportunities offered by glycobiology to develop innovative, potentially broad-spectrum antiviral therapeutics.

This review explores how virus–glycan recognition shapes influenza and coronavirus infection, host adaptation, and evolution, and how these molecular insights guide the design of glycomimetic antiviral strategies.

Introduction

Emerging and re-emerging viral diseases continue to pose major threats to human health, causing clinical outcomes ranging from mild infections to life-threatening complications.¹

The first molecular encounter between a virus and its host occurs at the glycocalyx, a dense and structurally diverse layer of glycoconjugates that covers virtually every mammalian cell. Far from representing a passive physical barrier, the glycocalyx provides a highly dynamic molecular landscape that regulates cell–cell communication, immune recognition, and pathogen interactions. Consequently, viruses must navigate this complex glycan-rich environment and have therefore evolved to exploit host glycans in multiple ways, using them as attachment factors, entry receptors, co-receptors, or modulators of receptor accessibility.² Conversely, glycans decorating viral envelope proteins influence receptor engagement, membrane fusion, immune evasion, and viral fitness. Together, these reciprocal interactions place glycans at the center of virus–host recognition.

Virus–glycan interactions are fundamentally different from most protein–protein recognition events. Individual carbohydrate–protein interactions are intrinsically weak and often transient, yet they achieve remarkable specificity through multivalency, spatial organization, and the extraordinary structural diversity of glycans. Subtle variations in glycosidic linkage, branching, chain length, chemical modification, or glycan density can profoundly alter viral tropism, host specificity, and transmission. Therefore, during evolution, adaptation to the host glycome has emerged as a major evolutionary force, shaping viral emergence and zoonotic spillover. On one hand, viral surface proteins are extensively decorated with host-derived glycans that mask immunogenic epitopes and facilitate immune evasion, a strategy known as glycan shielding. In this way, viral glycans can actively promote infection through interactions with host lectins and other glycan-binding receptors.^3,4

Alternatively, host glycans themselves frequently participate in the early stages of infection. For many viruses, host cell-surface glycans function either as attachment factors or as entry receptors. Attachment factors facilitate infection by concentrating viral particles at the cell surface, thereby increasing the likelihood of productive interactions with entry receptors.⁵ This strategy is employed by several coronaviruses, including middle east respiratory syndrome CoronaVirus (MERS-CoV) and severe acute respiratory syndrome CoronaVirus 2 (SARS-CoV-2), which utilize sialylated glycans (sialoglycans) as attachment factors, while relying on specific protein receptors for cell entry.⁶ In contrast, other viruses, such as influenza A virus, directly recognize host glycans as functional receptors that mediate viral attachment and initiate infection.

Recent advances in structural biology, glycan synthesis, glycan arrays, nuclear magnetic resonance spectroscopy, cryo-electron microscopy, computational chemical biology, and chemical glycobiology have transformed our understanding of these molecular recognition processes. These complementary approaches now permit virus–glycan interactions to be characterized at unprecedented atomic resolution, revealing the structural basis by which viruses recognize complex glycans and adapt to changing host glycomes.

Beyond their fundamental biological importance, these discoveries have important therapeutic implications. The earliest stages of viral infection are increasingly recognized as attractive targets for intervention, stimulating the development of glycomimetic compounds capable of disrupting virus attachment before irreversible entry occurs. Progress in medicinal chemistry has led to increasingly sophisticated glycan-inspired ligands, multivalent inhibitors, and heparan sulfate mimetics that exploit the molecular principles governing virus–glycan recognition.

In this review, we discuss our perspective on the molecular basis of glycan recognition during viral attachment and entry, focusing on influenza viruses and coronaviruses as paradigmatic examples of distinct glycan-dependent infection strategies. We examine how glycan recognition drives viral evolution, host adaptation, and interspecies transmission, and highlight recent advances in the design of glycomimetic antivirals targeting these early recognition events. Together, these examples illustrate how integrating glycobiology, structural biology, and medicinal chemistry is opening new opportunities for the development of broad-spectrum antiviral therapies.

Host glycan recognition by influenza viruses

Influenza viruses are enveloped, negative-sense RNA viruses of the family Orthomyxoviridae. Among the four influenza virus types (A–D), influenza A and B viruses are the principal causative agents of seasonal epidemics in humans, whereas influenza C generally causes mild disease and influenza D primarily infects cattle.

Influenza A viruses are classified into subtypes according to the antigenic properties of their two surface glycoproteins, hemagglutinin (HA) and neuraminidase (NA), which are the principal mediators of virus–host interactions and the major targets of the host immune response. HA and NA play complementary roles during infection: HA recognizes sialylated glycans on the host cell surface to initiate viral attachment, whereas NA cleaves terminal sialic acid residues to facilitate viral spread.⁷ The functional interplay between these two glycoproteins is central to influenza virus infectivity, host adaptation, and transmission (Fig. 1).^8–11

Fig. 1. Functional interplay between hemagglutinin (HA) and neuraminidase (NA) during the influenza A virus replication cycle. HA initiates infection by recognizing sialylated glycans on the host cell surface (1), promoting viral attachment and endocytosis (2) and subsequent membrane fusion (3). Following viral genome replication and transcription (4), protein synthesis (5), and virion assembly (6), progeny viruses bud from the plasma membrane (7), where NA cleaves terminal sialic acid residues from cellular and viral glycoconjugates to prevent virion aggregation and facilitate viral release (8). Successful infection relies on a finely balanced interplay between HA-mediated receptor binding and NA-mediated receptor destruction, which optimizes viral attachment, dissemination, and host adaptation. In addition, both glycoproteins are extensively glycosylated by the host biosynthetic machinery and are subject to continuous immune pressure, driving their antigenic evolution. Created with https://BioRender.com.

The interaction between HA and host glycans exemplifies a fundamental principle of glycan recognition. At the monovalent level, protein–glycan interactions are inherently weak, typically displaying dissociation constants in the high µM to mM range.^12–14 Influenza viruses overcome this limitation through multivalency.^5–17 The viral architecture acts in concert with the high density and spatial organization of sialylated glycans within the host glycocalyx, allowing the simultaneous engagement of multiple HA proteins, thereby transforming numerous weak interactions into stable virus attachment and, thus ensuring efficient host-cell recognition and productive infection.^18,19

The host glycome is a highly dynamic and structurally diverse interface that profoundly influences viral evolution. Consequently, influenza viruses must continuously adapt to the structural diversity, density, and spatial organization of host glycans. Evolutionary changes in HA receptor specificity and NA enzymatic activity enable influenza viruses to navigate this glycan-rich environment, contributing to efficient transmission, host adaptation, and zoonotic emergence. Understanding how influenza viruses recognize and adapt to host glycans is therefore central for elucidating the molecular basis of host specificity and interspecies transmission, and for developing strategies to predict and counteract future pandemic threats.

Sialic acid glycosidic linkage encodes IAV species specificity

Over the past century, four influenza A virus (IAV) pandemics have emerged through the cross-species transmission and subsequent adaptation of avian or avian-derived swine influenza viruses in immunologically naïve human populations. Although pandemic emergence is a multifactorial process involving the adaptation of multiple viral genes, the receptor-binding properties of hemagglutinin (HA) are major determinants of host specificity, tissue tropism, and zoonotic potential.^20–22

One of the principal determinants of influenza host range is the glycosidic linkage through which sialic acid is presented on host glycans. Avian-adapted HAs preferentially recognize α2,3-linked sialoglycans (Neu5Acα2-3Gal), which are predominantly expressed in the intestinal epithelium of birds, the principal site of viral replication in their natural reservoir. In contrast, human-adapted HAs preferentially bind α2,6-linked sialoglycans (Neu5Acα2-6Gal), which are abundantly expressed on epithelial cells of the human upper respiratory tract.^10,23 These distinct glycan landscapes constitute an important species barrier that limits cross-species transmission.²⁴

Studies conducted in ferrets have demonstrated that acquisition of human-type receptor specificity is a key prerequisite for efficient airborne transmission.²⁵ Consistent with these observations, the emergence of α2,6-linked receptor specificity is now considered an important criterion in the assessment of the pandemic potential of avian influenza viruses.^26–28

The segmented genome of IAVs facilitates genetic reassortment when distinct viral strains co-infect the same host cell. This process has played a central role in the emergence of pandemic strains, including H2N2 in 1957 and H3N2 in 1968, both of which arose through reassortment between avian and human influenza viruses.^29,30 However, successful adaptation to humans requires also optimization of HA receptor specificity. Comparative analyses of pandemic HAs and their avian counterparts indicate that influenza virus evolution in humans reflects a delicate balance between immune escape and enhanced recognition of human-type receptors.³¹ Remarkably, numerous studies have shown that only a small number of amino acid substitutions—and in some cases a single mutation—are sufficient to shift HA specificity from avian-type α2,3-linked to human-type α2,6-linked sialoglycans.^21,32,33 Indeed, a recent study by Wilson and Paulson further demonstrated that the Q226L substitution in an H5 HA is sufficient to switch receptor preference from avian- to human-type glycans.³⁴ These findings highlight the structural plasticity of the HA receptor-binding site and its pivotal role in host adaptation.

A classic example of receptor adaptation is provided by the 1968 H3N2 pandemic virus. This strain emerged through reassortment between avian and human influenza viruses, acquiring an avian-derived H3 hemagglutinin and PB1 polymerase gene while retaining the N2 neuraminidase and the remaining internal genes from the previously circulating human H2N2 lineage. Notably, the HA of the pandemic H3N2 virus differed from its putative avian precursor by only seven amino acid substitutions. Among these, Q226L and G228S were particularly important, as they conferred a strong preference for human-type α2,6-linked sialoglycans and facilitated adaptation to mammalian hosts.^31,35

Together, these findings establish HA receptor specificity as a central determinant of influenza A virus evolution, host adaptation, and pandemic emergence.

Continuous adaptation to the human glycome: A/H3N2 lineage evolution

Since its introduction into the human population in 1968, the H3N2 lineage has undergone continuous evolution under the pressure of immune selection imposed by vaccination and prior infection, while progressively adapting to the human glycan landscape (Fig. 2). Antigenic drift, driven by the accumulation of amino acid substitutions within the globular head of hemagglutinin (HA), constitutes the primary mechanism of immune escape. Importantly, these mutations not only remodel antigenic epitopes, but can also alter glycan-binding properties, thereby linking immune evasion to receptor adaptation.^9,33

Fig. 2. Evolution of glycan recognition during adaptation of the A/H3N2 influenza virus lineage to the human glycome. Following its introduction into the human population in 1968, H3N2 viruses underwent a progressive shift in receptor specificity driven by antigenic drift. Early strains, represented by A/Hong Kong/1/1968 (HK68), displayed broad receptor recognition and interacted primarily with the terminal sialic acid residue of both α2,3- and α2,6-linked sialosides. Subsequent variants such as A/Netherlands/816/1991 (NL91) evolved increased specificity for α2,6-linked human-type receptors. Around the early 2000s, H3N2 viruses underwent a major shift in receptor-binding specificity, exemplified by strains such as A/Netherlands/109/2003 (NL03), A/Netherlands/761/2009 (NL09), and A/Singapore/INFIMH-16-0019/2016 (SG16), which preferentially recognized extended α2,6-linked glycans containing multiple N-acetyllactosamine (LacNAc) repeats. More recent variants, including A/Darwin/6/2021 (Darwin21), have been proposed to partially revert receptor specificity toward shorter N-glycans containing di-LacNAc motifs. The recently emerged H3N2 subclade K highlights the continued evolution of this lineage, although its precise glycan-binding preferences remain unknown yet. Collectively, these evolutionary transitions demonstrate how antigenic drift reshapes HA glycan-binding preferences, enabling long-term adaptation of H3N2 viruses to the complex glycan landscape of the human airway. Created with https://BioRender.com.

Over the past five decades, human H3N2 viruses have exhibited a remarkable evolution in receptor recognition. Early pandemic strains circulating between 1968 and 1979 displayed dual specificity toward both α2,3- and α2,6-linked sialic acids. Consistent with this phenotype, structural and biophysical studies revealed that the HA of the prototype A/Hong Kong/1/1968 (HK68) strain interacted almost exclusively with the terminal sialic acid residue.³⁶ This relatively simple binding mode enabled recognition of both avian-type (Neu5Acα2-3Gal) and human-type (Neu5Acα2-6Gal) receptors and likely facilitated the interspecies transition of the virus.

Subsequent human-adapted variants progressively evolved toward exclusive recognition of α2,6-linked sialoglycans.³⁷ The structural basis for this specificity was elucidated using the A/Netherlands/816/1991 (NL91) strain. These studies showed that efficient receptor engagement depends not only on interactions with the terminal sialic acid but also on contacts with the adjacent galactose residue, which can be optimally established in α2,6-linked, but not α2,3-linked, glycans.³⁸ Consistent with their relatively broad receptor-binding properties, these early H3N2 viruses efficiently agglutinated erythrocytes and replicated in laboratory propagation systems such as embryonated chicken eggs and MDCK cells. Glycomic analyses have shown that these systems predominantly present short α2,6-sialylated glycans containing a single LacNAc repeat, which are sufficient to support infection by early H3N2 strains.³⁹

A major shift in glycan recognition emerged during the early 2000s. Contemporary H3N2 isolates progressively lost the ability to agglutinate erythrocytes and became increasingly difficult to propagate in conventional laboratory hosts. Initially, these observations led to the assumption that the viruses had reduced affinity for human receptors. However, subsequent studies demonstrated that they retained strong specificity for α2,6-linked sialosides while evolving a pronounced preference for extended glycans containing multiple N-acetyllactosamine (LacNAc) repeats.^40,41

This transition represents a refinement rather than a loss of receptor recognition, whereby viruses increasingly target a narrower subset of physiologically relevant human glycans. Structural and biophysical analyses of post-2003 strains, including A/Netherlands/109/2003 (NL03), A/Netherlands/761/2009 (NL09), and A/Singapore/INFIMH-16-0019/2016 (SG16), revealed that these viruses engage substantially larger glycan epitopes and frequently require tri-LacNAc or longer sialylated chains for optimal binding.²² This extended binding mode is accompanied by remodeling and expansion of the HA receptor-binding site, enabling direct interactions with internal regions of the glycan chain rather than exclusively with the terminal sialic acid, reflecting adaptation to the complex and heterogeneous glycan environment of the human airway.³⁸

The increasing dependence on extended glycans has important practical consequences. Because such receptors are poorly represented in eggs and standard cell culture systems, viral propagation in these hosts imposes strong selective pressures that frequently favor adaptive mutations restoring recognition of shorter glycans. As a result, laboratory-passaged viruses may exhibit receptor-binding properties that do not accurately reflect those of circulating strains. This mismatch contributes to reduced viral growth and has been implicated in the diminished effectiveness of some egg-based influenza vaccines.^42–44

Recent evolutionary trends suggest that receptor recognition continues to diversify. All H3N2 viruses emerging after SG16 harbor additional substitutions within HA that have been proposed to influence the flexibility and preorganization of the receptor-binding site, promoting a partial shift in specificity toward shorter N-glycans containing di-LacNAc motifs. Glycan-array analyses, tissue-binding studies, and molecular dynamics simulations indicate that epistatic interactions among multiple residues within the receptor-binding site collectively reshape glycan.⁴⁵ Nevertheless, the lack of high-resolution structural information currently limits a detailed mechanistic understanding of these newly emerged binding phenotypes.

Overall, the evolution of glycan recognition in the H3N2 lineage illustrates a progressive transition from broad, largely structure-independent interactions to highly selective recognition of complex human glycans. This evolutionary trajectory exemplifies the co-evolution of influenza viruses with the human glycome, whereby immune-driven mutations are constrained and shaped by the structural requirements of receptor engagement (Fig. 2). Importantly, these adaptations arise not from individual substitutions alone but from epistatic networks of mutations that collectively remodel the HA receptor-binding site while preserving viral fitness and transmissibility.⁴⁶

The continuous emergence of antigenically distinct H3N2 variants highlights that this evolutionary process remains highly dynamic. During the 2025–2026 Northern Hemisphere influenza season, the H3N2 subclade K rapidly became a dominant circulating lineage. This variant harbors multiple HA substitutions relative to previous vaccine strains that affect key antigenic regions and have been associated with reduced vaccine effectiveness.^47–49 Although disease severity remains comparable to that of previous seasonal strains, its emergence underscores the continuous adaptation of influenza viruses under the combined pressures of host immunity and receptor optimization, reinforcing the need for global surveillance and regular vaccine updating.

More broadly, influenza viruses provide one of the best-characterized models of glycan-dependent host adaptation. The remarkable evolution of HA receptor specificity demonstrates how subtle changes in glycan recognition can reshape viral tropism, transmissibility, and pandemic potential, providing a conceptual framework for understanding glycan-mediated infection strategies across diverse viral families.

Host glycan recognition by CoronaViruses

Whereas influenza viruses directly exploit host sialoglycans as functional receptors, coronaviruses have evolved more diverse glycan-dependent strategies in which glycans primarily function as attachment factors that facilitate engagement of protein receptors. This distinction illustrates two fundamentally different solutions to the common challenge of navigating the host glycocalyx.

Coronaviruses (CoVs) are enveloped, positive-sense single-stranded RNA viruses classified into four genera (α, β, γ, and δ). They infect a broad range of vertebrate hosts and cause respiratory, enteric, hepatic, and neurological diseases of varying severity. Human coronaviruses include endemic strains, such as HCoV-229E, HCoV-NL63, HCoV-OC43, and HCoV-HKU1, which are generally associated with mild respiratory disease, as well as highly pathogenic zoonotic viruses, including severe acute respiratory syndrome coronavirus (SARS-CoV), middle east respiratory syndrome coronavirus (MERS-CoV), and SARS-CoV-2. Despite their biological diversity, many coronaviruses exploit host glycans during the earliest stages of infection, using them to promote viral attachment, modulate spike-protein conformation, and enhance engagement of their primary protein receptors.

Coronavirus attachment and entry are primarily mediated by the spike glycoprotein, which recognizes host receptors and drives membrane fusion. A common feature of many coronaviruses is their ability to exploit host glycans during the early stages of infection. In several species, including HCoV-HKU1 and MERS-CoV, the spike protein contains glycan-binding sites that facilitate viral attachment by engaging cell-surface sialoglycans prior to interaction with proteinaceous entry receptors (Fig. 3). In addition, several β-coronaviruses encode a hemagglutinin-esterase (HE) protein that contributes to sialoglycan recognition and receptor destruction, functionally paralleling the coordinated HA–NA system of influenza viruses.⁵⁰

Fig. 3. hCoV attachment receptors on host cells and glycan recognition motifs. hCoV variants differ in their infection strategies: the S spike protein of OC43 binds to 9AcO-diSia on the host cell, whereas other hCoVs utilize glycans as attachment factors while relying on specific protein receptors for cell entry, such as TMPRSS2 or DPP4 for HKU1 and MERS-CoV, respectively; or the combination of ACE2 as entry receptor and TMPRSS2 as a spike-activating protease in SARS-CoV-2. On the bottom panel, proposed glycan binders for each of the viruses are shown. Both OC43 and HKU1 are specific for O-acetylated sialic acids, particularly 9-O-acetylated α2–8-linked disialosides. MERS-CoV displays preference for α2,3-Sia motifs, like the sulfated sialyl Lewis X (sLe^x) and sulfated sialyl LacNAc antigen, and sialylated bi- and tri-antennary complex N-glycans. SARS-CoV-2 binds in a similar fashion to α2,3- and α2,6-linked sialyl LacNAc/Lac and heparan sulfate, whereas the beta subvariant strain 501Y.V2-1 recognizes 9AcO-diSia epitopes. Created with https://BioRender.com.

The diversity of coronavirus–glycan interaction reflects the remarkable structural complexity of the mammalian sialome. Beyond differences in glycosidic linkage, sialic acids undergo extensive chemical modifications that greatly expand the repertoire of potential viral receptors. Among these, O-acetylation has emerged as a major determinant of coronavirus recognition. The tissue- and species-specific distribution of O-acetylated sialic acids strongly influences viral attachment and is increasingly recognized as an important factor governing host susceptibility, tissue tropism, and viral adaptation.^51,52

O-acetylated sialic acids as determinants of coronavirus attachment and host adaptation

The importance of O-acetylated sialic acids is particularly evident among embecoviruses, a subgenus of β-coronaviruses that includes bovine coronavirus (BCoV), rabbit coronavirus (RbCoV), equine coronavirus (ECoV), canine respiratory coronavirus (CRCoV), murine hepatitis virus (MHV), and the human coronaviruses HCoV-OC43 and HCoV-HKU1. These viruses exhibit a conserved dependence on O-acetylated sialoglycans for host-cell attachment despite infecting diverse host species.^53,54

Coronavirus host adaptation is closely associated with the receptor specificity of their viral lectins. Structural and glycan-binding studies have revealed a remarkable complementarity between the O-acetyl substituents of host sialoglycans and hydrophobic pockets within the viral glycan-binding site. Although subtle differences in binding specificity exist among viruses infecting different species, 9-O-acetylated sialic acid emerges as a common recognition motif across most embecoviruses examined to date. The conservation of this glycan-binding strategy across animal and human coronaviruses suggests that recognition of 9-O-acetylated sialic acids represents an evolutionarily conserved mechanism of host attachment.⁵⁵

The evolutionary implications of glycan recognition are particularly evident in HCoV-OC43 and HCoV-HKU1. Despite arising from independent zoonotic introductions into humans, both viruses convergently evolved toward selective recognition of 9-O-acetylated α2–8-linked disialosides.^56–59 Because receptor specificity is a major determinant of host range and tissue tropism, understanding how coronaviruses recognize structurally distinct glycans is essential for predicting viral adaptation and zoonotic emergence.

Consistent with this model, glycan microarray studies have demonstrated that HCoV-HKU1 preferentially recognizes 9-O-acetylated α2–8-linked disialosides, whereas HCoV-OC43 displays broader specificity and can additionally engage α2–3- and α2–6-linked 9-O-acetylated monosialylated glycans.⁶⁰ These studies further revealed that modifications in the acetylation pattern profoundly influence receptor recognition, whereas variations in glycosidic linkage exert comparatively modest effects on binding. Together, these findings identify O-acetylation as a dominant molecular determinant of coronavirus–sialoglycan interactions.

Physiologically, α2–8-linked disialic acid motifs are abundant in gangliosides such as GD3.⁶¹ Experimental overexpression of the ganglioside biosynthetic enzyme ST8SIA1, which catalyzes the conversion of GM3 into GD3 and related disialylated gangliosides, enhances HCoV-OC43 and HCoV-HKU1 entry, supporting the notion that gangliosides function as physiologically relevant attachment factors.⁶⁰ The addition of O-acetyl groups is catalyzed by the sialate O-acetyltransferase CASD1, currently the only mammalian enzyme known to mediate 9-O-acetylation.⁶² These observations further suggest that glycan recognition contributes to tissue tropism by directing viral attachment toward cells enriched in specific ganglioside species.

Dual receptor mechanisms

Recent studies have revealed that coronavirus attachment is often a multistep process involving both glycan and protein receptors. HCoV-HKU1 provides a compelling example of this mechanism. In addition to recognizing 9-O-acetylated sialoglycans as attachment factors, HCoV-HKU1 utilizes the host transmembrane protease serine 2 (TMPRSS2) for both spike protein cleavage and as a direct receptor.⁶³ High-resolution cryo-electron microscopy studies demonstrated that binding of O-acetylated gangliosides induces conformational rearrangements within the spike trimer, facilitating exposure of receptor-binding domains required for protein receptor engagement.^64,65 Likewise, HCoV-OC43 undergoes conformational changes upon carbohydrate binding.⁶⁶ These findings support a dual-receptor model in which glycans act not only as attachment factors, but also as allosteric regulators that prime the spike protein for subsequent receptor interactions.

This paradigm extends beyond endemic human coronaviruses. MERS-CoV employs a two-step attachment mechanism in which sialylated glycans function as attachment factors together with the protein receptor dipeptidyl peptidase 4 (DPP4).^57,67 Structural studies identified a glycan-binding pocket within the N-terminal domain of the MERS-CoV spike that recognizes sialylated glycans and enhances viral attachment before engagement of DPP4.^68,69 These observations highlight the widespread use of glycan-assisted entry strategies among coronaviruses.

The contribution of glycans to SARS-CoV-2 infection remains an active area of investigation. Although angiotensin-converting enzyme 2 (ACE2) is the primary receptor mediating viral entry and TMPRSS2 acts as a spike-activating protease,^70,71 accumulating evidence suggests that glycan-mediated interactions contribute to viral attachment and tropism. Nuclear magnetic resonance, glycan array, and atomic force microscopy studies indicate that the SARS-CoV-2 spike protein can interact with multiple sialylated glycans, including α2–3- and α2–6-linked sialylated N-acetyllactosamine structures^6,72,73 and, potentially, 9-O-acetylated sialic acids, which has been proposed as an initial binding site for the SARS-CoV-2 variant 501Y.V2-1.⁵⁹ Additional studies have implicated gangliosides such as GM1 in facilitating viral uptake in specific cell types.⁷⁴ Although experimental evidence supports interactions between SARS-CoV-2 and sialoglycans, the biological significance of these interactions in vivo remains uncertain. Therefore, the proposed role of glycans as auxiliary attachment factors should be interpreted with caution until supported by definitive in vivo evidence, unlike ACE2 engagement or the well-characterized recognition of O-acetylated sialic acids by embecoviruses.

Collectively, these studies demonstrate that coronavirus attachment depends not only on the presence of sialic acids but also on their precise chemical modifications and structural presentation within host glycoconjugates. In particular, the selective recognition of O-acetylated sialoglycans illustrates how subtle changes in glycan chemistry can profoundly influence viral attachment, host adaptation, tissue tropism, and interspecies transmission. Consequently, these findings identify glycan-mediated recognition as a key determinant of coronavirus biology and provide a compelling rationale for the development of glycomimetic compounds capable of disrupting the earliest stages of viral infection. Such strategies offer promising opportunities for the design of novel antiviral agents targeting conserved mechanisms of virus–host recognition.

Glycomimetics

The growing understanding of virus–glycan recognition has stimulated the development of glycomimetic compounds, and more generally, inhibitors designed to interfere with the earliest stages of viral infection. By disrupting glycan-mediated attachment, these molecules aim to prevent productive virus–host interactions before irreversible cell entry occurs. Although this strategy has been explored most extensively for influenza viruses, it is increasingly being extended to coronaviruses and other glycan-dependent pathogens, highlighting the broad therapeutic potential of targeting conserved mechanisms of viral recognition.

For influenza viruses, glycomimetic approaches have primarily focused on the two viral surface glycoproteins responsible for receptor engagement and release described above: hemagglutinin (HA) and neuraminidase (NA). HA-directed inhibitors seek to block viral attachment to sialylated receptors, whereas NA inhibitors prevent the enzymatic cleavage of terminal sialic acids required for efficient viral dissemination. However, the rapid antigenic evolution of HA, together with the intrinsically weak affinity of monovalent HA–glycan interactions, has complicated the development of broadly effective HA-targeted therapeutics.

Influenza virus inhibitors

Early efforts to develop HA-targeting glycomimetics focused on identifying the structural features of sialic acid required for receptor recognition. Systematic analyses of the effects of anomeric configuration, acetylation patterns, and neighboring monosaccharide residues demonstrated that these modifications influence HA binding but do not substantially increase affinity, which generally remains in the low-millimolar range.¹³ These observations highlighted the intrinsically weak nature of monovalent HA–glycan interactions and underscored the need for alternative design strategies. Subsequent studies explored non-hydrolyzable sialic acid analogues capable of resisting neuraminidase-mediated cleavage while retaining HA recognition, as well as libraries of structurally modified and fluorinated sialic acid derivatives generated through structure-guided and computational approaches (Fig. 4A).^75–77 Although these compounds provided valuable insights into the molecular determinants of receptor recognition, their inhibitory potency remained limited.

Fig. 4. Chemical structure of glycomimetics targeting IAV. (A) Non-hydrolysable monovalent glycomimetic developed by Weinhold and Knowles.⁷⁵ (B) Trivalent glycomimetic by Lu et al.⁸⁰ (C) Structures of the three FDA approved influenza A virus neuraminidase inhibitors.^89–91.

The trimeric architecture of HA and the high density of HA molecules on the viral surface motivated the development of multivalent glycomimetics designed to exploit avidity effects. By simultaneously engaging multiple receptor-binding sites, these compounds can achieve substantially enhanced binding compared with monovalent ligands. A variety of multivalent scaffolds displaying α2,6-sialylated glycans have been developed, including nucleic acid-based constructs, glycoclusters, protein conjugates, and self-assembling supramolecular systems (Fig. 4B).^78–83 These approaches have yielded affinity enhancements ranging from two to three orders of magnitude and, in some cases, potent inhibition of viral infection at nanomolar concentrations. A representative example is provided by Bandlow et al., who used structural analyses to define the spatial organization of HA receptor-binding sites. Adjacent receptor-binding sites within the HA homotrimer are separated by approximately 42 Å, whereas the centers of neighboring HA trimers are spaced by approximately 102 Å (Fig. 5). Based on these structural constraints, nucleic acid-based scaffolds displaying 6′SLN ligands at defined spacings ranging from 23 to 101 Å were designed. Binding studies revealed that optimal affinity was achieved with ligand spacings of 52–59 Å, consistent with the simultaneous engagement of two adjacent receptor-binding sites within a single HA trimer. The design was subsequently extended to enable simultaneous binding to two HA trimers; however, this modification did not produce a significant additional increase in affinity.

Fig. 5. Adapted from Bandlow et al.⁷⁸ Distance affinity relationship of bivalent conjugated sialosides to influenza A (H3N2) X31 virus. The expected ligand–receptor interaction as function of linker length in the proposed inhibitor molecules is shown. On the right bottom, a scheme showing the distances between adjacent receptor-binding sites within the HA homotrimer (42 Å) and between two neighboring HA trimers (102–154 Å) of influenza A (H3N2) X31 virus.

An alternative approach has focused on the development of S-linked sialosides, in which the glycosidic oxygen is replaced by sulfur. These analogues preserve recognition by HA while exhibiting increased resistance to neuraminidase-mediated hydrolysis. Both monovalent and multivalent S-sialoglycomimetics have been reported, including fullerene-based glycoconjugates, protein-linked sialosides, and amphiphilic self-assembling systems.^84–86 Although their antiviral activities vary considerably, these compounds illustrate the potential of simultaneously targeting viral attachment and receptor-cleavage processes.

Collectively, these findings highlight the importance of matching ligand spacing and linker flexibility to the spatial organization of viral receptor-binding sites rather than simply increasing ligand valency. Although many of these multivalent constructs are relatively large and therefore may present challenges for therapeutic development, they provide important proof-of-concept for exploiting the nanoscale organization of HA on the viral surface. Future development will require balancing multivalent avidity with physicochemical properties, stability, tissue distribution, and pharmacokinetic profiles suitable for therapeutic application.

Despite the avidity gains achieved by multivalent glycomimetics, most of these molecules function as reversible inhibitors that competitively occupy the HA receptor-binding sites. Consequently, their antiviral activity depends on sustained target engagement, and viral particles may regain infectivity upon inhibitor dissociation, potentially limiting therapeutic efficacy in vivo. To address this limitation, recent multivalent designs have incorporated virucidal mechanisms to irreversibly inactivate viral particles. For example, cyclodextrin-based macromolecules displaying either human- or avian-type sialylated receptors exhibited broad-spectrum antiviral activity against influenza viruses in vitro, ex vivo, and in vivo, achieving nanomolar potency and therapeutic efficacy in mouse models of influenza infection.⁸⁷

A conceptually distinct strategy has focused on targeting the host receptor rather than the virus itself. DAS181, a recombinant sialidase fusion protein, removes terminal sialic acids from the respiratory epithelium, thereby preventing viral attachment independently of HA subtype or antigenic variation. Clinical studies of DAS181 have demonstrated antiviral activity and reduced viral loads in influenza-infected patients, providing proof-of-concept that host-directed glycan-targeting therapies can effectively interfere with influenza infection.⁸⁸

In parallel with efforts directed toward HA, neuraminidase has emerged as a highly successful target for antiviral drug development. Unlike HA, the catalytic site of NA is relatively conserved among influenza strains, facilitating the design of broad-spectrum inhibitors. Currently, three neuraminidase inhibitors have received regulatory approval for clinical use: zanamivir (Relenza), oseltamivir (Tamiflu), and peramivir (Rapivab).^89–91 These compounds act by occupying the catalytic site of NA and preventing the cleavage of terminal sialic acid residues from host glycoconjugates, thereby impairing viral release and limiting viral spread (Fig. 4C). Despite their structural differences, all three inhibitors preserve key molecular features involved in sialic acid recognition, including a negatively charged carboxylate that mimics the natural substrate and strategically positioned polar functionalities that reproduce critical interactions within the NA active site.

Additional antiviral agents approved for the treatment of influenza include the M2 ion-channel inhibitors amantadine and rimantadine, as well as the cap-dependent endonuclease inhibitor baloxavir marboxil (Xofluza).^92,93 Although these compounds do not directly target glycan-mediated processes, they represent important components of the current anti-influenza therapeutic arsenal and illustrate the diversity of strategies available to interfere with the viral replication cycle.

Coronavirus inhibitors

The growing understanding of glycan-mediated attachment mechanisms in coronaviruses has opened new opportunities for antiviral intervention. Unlike influenza viruses, where therapeutic strategies have largely focused on sialic acid recognition, coronaviruses exploit multiple glycan-dependent pathways that can be targeted using glycomimetic compounds. These approaches primarily aim to disrupt either the interaction of viral proteins with sialylated glycans (at the N-terminal domain) or the engagement of heparan sulfate proteoglycans which involve the receptor-binding domain and adjacent regions extending toward the N-terminal domain and the S1/S2 cleavage site, and that function as important attachment factors during the earliest stages of infection (Fig. 6). By interfering with these initial recognition events, glycomimetics have the potential to reduce viral attachment, impair cellular entry, and complement antiviral therapies directed against protein receptors or viral replication.

Fig. 6. Cryo-electron microscopy structure of SARS-CoV-2 spike glycoprotein (PDB ID: 6VXX).⁹⁴ Left: Structure highlighting the N-terminal domain (NTD, blue) and receptor binding domain (RBD, green) regions. Right: Electrostatic surface representation. Structure generated from PDB entry 6VXX.⁹⁴.

One strategy has focused on modulating cellular sialylation. Because several coronaviruses exploit sialylated glycans during attachment, reducing the abundance of cell-surface sialic acids can impair viral infection. Inhibition of sialyltransferase activity using fluorinated sialic acid analogues such as 3Fax-peracetyl-Neu5Ac decreases the incorporation of sialic acid into host glycoconjugates and has been shown to reduce replication of both human coronavirus OC43 and influenza A virus.⁹⁵ These findings further support the importance of sialylated receptors during the early stages of viral infection.

The preferential recognition of 9-O-acetylated sialic acids by several coronaviruses has also inspired the development of sialic acid-based glycomimetics. Multivalent glycoclusters displaying Neu5Ac or 9-O-acetylated Neu5Ac on a variety of molecular scaffolds, including pillararenes, calixarenes, fullerenes, and porphyrins, have been shown to compete with host receptors and inhibit viral attachment.^96,97 In general, incorporation of the 9-O-acetyl substituent significantly enhances inhibitory potency, consistent with the conserved preference of many coronaviruses for O-acetylated sialoglycans. Among the different architectures explored, porphyrin-based glycoclusters have exhibited particularly promising activity, inhibiting attachment of multiple SARS-CoV-2 variants at low-micromolar concentrations. Further optimization of linker length and glycan presentation has yielded compounds with improved antiviral potency across distinct viral lineages.⁹⁸

Despite their promising activity, O-acetylated sialic acid derivatives suffer from a major limitation: the intrinsic lability of ester groups toward hydrolysis. To address this issue, chemically stabilized analogues have been developed in which the 9-O-acetyl substituent is replaced by bioisosteric functionalities. In particular, 9-N-acetyl sialic acid derivatives retain recognition by coronavirus lectins while displaying increased resistance to hydrolysis by viral hemagglutinin-esterases.⁹⁹ These compounds provide attractive building blocks for the in-silico design and future development of stable multivalent glycomimetics targeting O-acetyl-sialic acid-binding coronaviruses.^100,101

Beyond sialoglycans, increasing attention has been directed toward heparan sulfate-mediated attachment.¹⁰² Several coronaviruses, including SARS-CoV-2, utilize heparan sulfate proteoglycans to increase their local concentration at the cell surface and facilitate subsequent engagement of entry receptors. Although unfractionated heparin can inhibit viral attachment, its clinical utility is limited by anticoagulant activity and bleeding risk.^103,104 Consequently, considerable efforts have focused on the development of non-anticoagulant heparan sulfate mimetics capable of preserving antiviral activity while minimizing adverse effects.

Among these, highly sulfated polyglycerols have emerged as promising candidates. Both linear and hyperbranched polysulfated polyglycerols efficiently bind the SARS-CoV-2 spike protein and prevent its interaction with host cells.¹⁰⁵ Notably, highly sulfated linear polyglycerols display inhibitory activities that substantially exceed those of heparin, highlighting the importance of charge density and multivalent electrostatic interactions in spike recognition. Similarly, the heparan sulfate mimetic pixatimod (PG545) has demonstrated potent activity against multiple SARS-CoV-2 variants (Fig. 7A).¹⁰⁶ In addition to directly competing with host glycans for spike binding, pixatimod appears to induce conformational changes within the receptor-binding domain, thereby reducing engagement of the ACE2 receptor. Its broad antiviral activity, favorable safety profile, and limited anticoagulant properties have positioned pixatimod as one of the most advanced glycomimetic candidates currently under investigation for coronavirus infections. Naturally occurring polysaccharides and their synthetic derivatives have also attracted significant interest. In sulfated polysaccharide-based glycomimetics, the sulfate groups provide the principal negatively charged functionalities that engage positively charged residues on the spike surface through electrostatic interactions, while the extended glycan scaffold enables multiple contacts along the protein surface. Thus, both the density and spatial presentation of sulfate groups and the length and architecture of the glycan scaffold contribute to binding and antiviral activity. Sulfated chitosan derivatives engineered to mimic the charge distribution of heparan sulfate exhibit antiviral activity against multiple SARS-CoV-2 strains through direct interaction with the spike receptor-binding domain (Fig. 7B).^107,108 Similar effects have been reported for modified alginates as well as sulfated derivatives of heparin and enoxaparin.^109,110 Sulfated polysaccharides have also emerged as promising broad-spectrum inhibitors of coronaviruses by targeting viral attachment through electrostatic interactions with the spike glycoprotein. Among these, iota-carrageenan has shown the greatest antiviral activity, inhibiting SARS-CoV-2 infection in vitro with low-microgram per milliliter potency. In addition, intranasal iota-carrageenan has demonstrated clinical efficacy against common cold viruses, including endemic human coronaviruses, supporting its potential as a broad-spectrum antiviral. However, the antiviral activity of carrageenans is strongly influenced by their heterogeneous composition, complicating structure–activity relationship studies and necessitating rigorous characterization of polymer preparations.¹¹¹

Fig. 7. Chemical structure of glycomimetics targeting CoV. (A) Structure of Pixatimod (PG545).¹⁰⁶ (B) General chemical structure of the chitosan derivatives synthesized by Revuelta et al.¹⁰⁸.

Collectively, these studies further support a length- and sulfation-dependent interaction between heparan sulfate mimetics and the SARS-CoV-2 spike protein and demonstrate that highly sulfated polysaccharide scaffolds represent a versatile platform for the development of broad-spectrum inhibitors of coronavirus attachment.

Overall, advances in coronavirus glycomimetics underscore the therapeutic potential of targeting glycan-mediated recognition during viral attachment. The successful exploitation of both sialoglycan- and heparan sulfate-dependent pathways illustrates how structural glycobiology and medicinal chemistry can converge to generate innovative antiviral strategies. As our understanding of virus–glycan recognition continues to expand, glycomimetic design is likely to play an increasingly important role in the development of broad-spectrum antivirals directed against conserved mechanisms of virus–host interaction.

Closing perspectives and broader implications

Glycans as viral recognition elements: extending beyond influenza and coronaviruses

The examples discussed throughout this review illustrate that glycans are far more than passive components of the cell surface. They constitute a dynamic molecular language that governs viral attachment, host adaptation, tissue tropism, and transmission. Collectively, by functioning as attachment factors, entry receptors, co-receptors, or modulators of receptor accessibility, glycans complement protein–protein interactions and frequently determine the efficiency, specificity, and outcome of virus–host recognition. Consequently, understanding the molecular principles of glycan recognition has become essential for deciphering viral evolution and pathogenesis.

Importantly, glycans influence viral infection from both sides of the virus–host interface. Beyond serving as host-cell receptors, they are integral structural components of viral envelope glycoproteins, where they regulate receptor engagement, membrane fusion, immune evasion, and antigenicity. Together, these complementary roles establish glycans as active regulators of viral infectivity rather than passive structural decorations.

Beyond their role in receptor recognition, glycans can also display essential structural and functional roles within viral entry machineries. In hantaviruses, for example, N-linked glycans present on the envelope glycoprotein Gc are required for the conformational rearrangements that drive membrane fusion. Unlike HIV, where glycans primarily mediate interactions with host lectins,^112,113 hantaviruses rely on glycosylation to maintain the structural integrity and functionality of the viral fusion apparatus.¹¹⁴ Comparable glycan-dependent recognition mechanisms are also found in numerous viral families beyond those discussed in detail throughout this review. Noroviruses recognize histo-blood group antigens (HBGAs), whereas rotaviruses interact with either HBGAs or sialylated glycans in a genotype-dependent manner during viral attachment. Likewise, polyomaviruses utilize specific gangliosides as functional entry receptors, while herpes simplex virus exploits cell-surface heparan sulfate proteoglycans as initial attachment factors to restrict viral mobility and facilitate subsequent engagement with protein receptors.^115–117 Taken together, this mechanistic diversity underscores how glycans contribute to viral infectivity, acting either as mediators of host recognition or as critical regulators of viral protein function. The dual involvement of glycans at both the host and viral levels offers unique opportunities for antiviral intervention. Carbohydrate-binding agents, glycomimetics, and inhibitors of glycan biosynthesis or processing have collectively demonstrated that glycan-dependent stages of infection are amenable to therapeutic targeting. Importantly, many of these recognition mechanisms are conserved across diverse viral families, raising the prospect of developing broad-spectrum antiviral strategies that interfere either with host glycan recognition pathways or with glycosylation-dependent functions of viral envelope proteins. As structural and mechanistic insights into virus–glycan interactions continue to emerge, rational glycomimetic design is expected to become an increasingly powerful approach for the development of next-generation antiviral therapeutics. Indeed, the successful development of multivalent glycomimetics for the inhibition of cholera toxin and Shiga toxin suggests that analogous design principles could also be exploited to improve the efficacy of antiviral glycan-based therapeutics.¹¹⁸

Overall, advances in glycobiology have transformed our understanding of glycans from passive structural components into active regulators of virus–host interactions. By integrating glycochemistry, structural biology, virology, and medicinal chemistry, the field is now uncovering the molecular principles that govern viral recognition, adaptation, and evolution. These discoveries not only deepen our understanding of viral pathogenesis but also establish glycan-mediated recognition as a rich and largely untapped source of therapeutic opportunities. Harnessing this knowledge will accelerate the development of innovative antivirals capable of targeting conserved mechanisms of infection and may ultimately provide new strategies to combat both emerging and re-emerging viral diseases.

Conflicts of interest

There are no conflicts to declare.

Biographies

Biography

Cristina Fernández Pérez obtained her Bachelor's Degree in Chemistry from the University of Granada (UGR) (2019–2023). She subsequently pursued a Master's Degree in Biotechnology at UGR (2023–2024). In September 2024, she joined CIC bioGUNE as a predoctoral researcher and began her PhD under the supervision of Dr Luca Unione in the Chemical Glycobiology Laboratory. Her doctoral research focuses on elucidating how antigenic drift reshapes influenza glycan-binding preferences and contributes to viral host adaptation.

Biography

María Emilia Cuervo Bustamante obtained her Bachelor's Degree in Biochemistry from Universidad Juan Agustín Maza (Argentina) in 2017. She subsequently worked as a clinical biochemist until 2022, when she joined the Laboratory of Biochemistry and Immunity at IMBECU-CONICET as a cell culture technician. In 2024, she obtained a Master's Degree in Clinical Research from the National University of Cuyo (Argentina). In 2025, she joined the Chemical Glycobiology Lab (CIC bioGUNE) as a PhD student under the supervision of Dr Luca Unione. Her doctoral research aims to elucidate how glycosylation modulates the structure and function of glycoproteins of biomedical relevance.

Biography

Iris A. Bermejo studied Chemistry at the University of La Rioja (Spain), where she received her PhD in 2018, funded by a PhD fellowship from the Spanish Association Against Cancer (AECC), working on cancer immunotherapy and the synthesis of glycomimetic-based cancer vaccines. In 2020, she was awarded a Marie Skłodowska-Curie Action (MSCA) for a postdoctoral stay at the University of Vienna (Austria). Since 2023, she has been a researcher at CIC bioGUNE, supported by a second MSCA. Her research focuses on glycans as regulators of biomedical processes, particularly in human coronaviruses, and on glycan–protein interactions using a multidisciplinary approach.

Biography

Jon I. Quintana studied Chemistry in the University of the Basque Country. In 2017, he joined the Chemical Glycobiology Lab at CIC bioGUNE and began his PhD in 2018 under Prof. Jesús Jiménez-Barbero and Dr Ana Arda, studying lectin–glycan interactions through NMR, ITC, and molecular dynamics simulations. During a 2021 secondment at the Max Planck Institute of Colloids and Interfaces, he worked on solid-phase glycan synthesis. He subsequently joined CIC biomaGUNE as a postdoctoral researcher, working on the synthesis and remodelling of N-glycans. He later returned to CIC bioGUNE, where he continues investigating the structural and functional roles of glycans.

Biography

Jesús Jiménez-Barbero has been an Ikerbasque Research Professor and Scientific Director of CIC bioGUNE since 2014. His research focuses on Chemical Biology, especially in Glycosciences. Major contributions include the unravelling of the molecular basis of glycan recognition by lectins at atomic resolution, in solution and on-cell, using a multidisciplinary approach combining chemical synthesis, protein biochemistry and molecular biology, biophysics, molecular modeling, and NMR, using a wide network of collaborations worldwide.

Biography

Luca Unione is an Ikerbasque Associate Professor and Associate Principal Investigator at CICbioGUNE, leading research on glycan–receptor interactions in infectious diseases. Trained as a chemist in Italy, he earned his PhD in Pharmacy in Madrid. His career includes translational pharmaceutical research, and an HFSP Fellowship at Utrecht University to develop chemo-enzymatic tools to study glycans and viral infections. He joined CICbioGUNE in 2022 and received an ERC Starting Grant in 2023 to investigate glycans in infectious diseases at atomic resolution. In 2025, he was awarded a Ramón y Cajal Fellowship. He is Visiting Professor at the University of Deusto.

Data availability

No primary research results, software or code have been included, and no new data were generated or analyzed as part of this review.