Therapeutic Targeting of Viral N-Glycosylation Modification: From Molecular Mechanisms to Clinical Application Prospects

Abstract

Viral diseases represent a significant global health challenge. N-glycosylation, a critical post-translational modification, plays diverse and essential roles throughout the viral life cycle. This review outlines the initiation of N-glycosylation, elucidating its role in facilitating viral protein synthesis within the endoplasmic reticulum (ER). This process ensures proper protein folding and functionality through stringent quality control mechanisms. Consequently, targeting N-glycosylation offers substantial potential for developing antiviral therapies. Specific antiviral strategies include inhibiting glycosyltransferase activity to block the initial glycosylation step and employing glucosidase inhibitors to disrupt viral protein glycosylation, leading to envelope protein misfolding. Additionally, this review explores other mechanisms of glucosidase inhibitors, including their dual role in modulating the ER-associated degradation (ERAD) pathway. The clinical evaluation of investigational antiviral agents is also addressed, providing a comprehensive analysis of their efficacy in suppressing viral replication and associated adverse effects. To advance precise antiviral interventions, future research must deepen the understanding of these mechanisms while optimizing the balance between efficacy and long-term safety in drug design. Such efforts are crucial for translating laboratory findings into effective clinical applications.

Key Summary Points

N-Glycosylation, a vital post-translational modification in viral pathogenesis, is a promising target for broad-spectrum antiviral agents.

Inhibitors exhibiting varying inhibitory activities, such as those targeting glycosyltransferases, glucosidases, and mannosidases, are able to inhibit the viral N-glycosylation process. This impairs the normal viral life cycle, thereby exerting antiviral activity.

A major challenge in developing N-glycosylation inhibitors is optimizing the selectivity index (SI), which measures the balance between antiviral efficacy and host cytotoxicity.

Research on N-glycosylation inhibitors shows promise but is still early-stage, with limited clinical trial data and no drugs approved yet. Advances in structural optimization, targeted inhibitors, and combination therapies could make these inhibitors key tools for broad-spectrum antiviral treatment.

Introduction

Viral diseases have exerted a profound impact on global health. In recent years, a series of viral infectious diseases, including influenza, acquired immunodeficiency syndrome (AIDS), Ebola virus disease, Zika virus disease, and coronavirus disease 2019 (COVID-19), have emerged worldwide, resulting in substantial casualties and economic losses [1–4]. These viruses are highly contagious and pathogenic. Additionally, their propensity for mutation often reduces the effectiveness of existing vaccines and therapeutics over time [5, 6]. Moreover, viral diseases can trigger societal panic and strain public health systems, amplifying their global health impact [7–9].

Direct-acting antivirals (DAAs) target specific molecules or structures critical to the viral life cycle, inhibiting replication [10]. For instance, nucleoside analogs such as zidovudine and lamivudine mimic natural nucleotides, integrating into viral DNA or RNA chains to terminate chain extension and thereby inhibit viral nucleic acid synthesis [11–13]. Similarly, protease inhibitors, such as lopinavir, block viral protease activity, preventing the cleavage and maturation of viral polyproteins. This inhibition, as seen in the impaired processing of the human immunodeficiency virus type 1 (HIV-1) Gag polyprotein, resulted in immature, noninfectious viral particles [14, 15]. However, the specificity of DAAs makes them vulnerable to resistance due to viral mutations [16]. These limitations have spurred exploration of alternative strategies. Host-targeting antivirals (HTAs) offer a promising approach owing to their broad-spectrum activity, reduced risk of resistance, and enhanced safety profile, particularly against emerging or drug-resistant viruses [9, 17, 18]. Furthermore, research into these host-directed therapies has highlighted the critical role of viral N-glycosylation modification.

N-Glycosylation, recognized as a vital post-translational modification, has been demonstrated to be indispensable for the functionality of numerous viral proteins [19, 20]. Through the attachment of oligosaccharide chains to specific asparagine residues, critical cellular processes such as protein folding, intracellular transport, structural stability, and functional activity are systematically regulated. This post-translational modification has been widely recognized as critical for fundamental viral processes, encompassing virion assembly, maturation cycles, and infectivity mechanisms [21–23]. Notable examples include the hemagglutinin (HA) protein of influenza virus and the envelope glycoprotein (gp120) of HIV-1, both reliant on N-glycosylation for proper structure and function [24, 25]. Consequently, inhibiting N-glycosylation or disrupting its pathway is a promising strategy for novel antiviral therapies [26, 27]. As N-glycosylation is critical for viral replication, its inhibition can significantly impair infectivity and transmission. Furthermore, targeting this host-mediated process may lower the risk of viral resistance compared with DAAs, as viruses struggle to mutate around essential host-dependent modifications [26]. Strategies targeting N-glycosylation may also offer broader activity against multiple viruses and reduce resistance development. The molecular mechanisms underlying N-glycosylation modification are systematically examined in this review, along with its critical involvement in viral replication being elucidated. Furthermore, the therapeutic potential of this post-translational modification process is highlighted as a promising target for developing innovative antiviral strategies.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

N-Glycosylation in the Viral Life Cycle

Initiation of N-Glycosylation

N-Glycosylation represents a pivotal post-translational modification characterized by the covalent attachment of oligosaccharide moieties at conserved asparagine residues located within the canonical Asn-X-Ser/Thr sequon (X ≠ proline) of nascent polypeptides [28–30]. The process of modification exerts multifaceted regulatory effects on tertiary structure formation, conformational stability, and biological activity of glycoproteins. From a virological perspective, this biochemical process has been demonstrated to be indispensable for viral pathogenesis through mediating critical virological processes including but not limited to virion morphogenesis, structural maturation, and host cell entry mechanisms.

The initial stage of N-glycosylation is localized within the ER (Fig. 1), where biosynthesis of the oligosaccharide precursor is initiated through its assembly on the lipid carrier dolichol phosphate (Dol-P). This precursor structure, composed of two N-acetylglucosamine (GlcNAc) units, nine mannose (Man) residues, and three glucose (Glc) molecules arranged in the configuration (GlcNAc)₂Man₉Glc₃, is systematically constructed through stepwise enzymatic reactions [30, 32]. The sequential addition of monosaccharides is mediated by membrane-associated glycosyltransferases, with specific sugar residues being transferred from nucleotide-activated donors including uridine diphosphate (UDP)-GlcNAc and guanosine diphosphate (GDP)-Man to the dolichol-bound intermediate. Following complete assembly, the oligosaccharide moiety is translocated en bloc from its dolichol phosphate anchor to acceptor polypeptides through a transamidation reaction. This critical transfer event is catalyzed by the oligosaccharyltransferase (OST) complex, which is strategically positioned within the ER membrane [33]. Recognition of the canonical Asn-X-Ser/Thr motif on nascent polypeptides is accomplished by the OST machinery, enabling precise covalent conjugation of the oligosaccharide’s reducing terminus to the amide nitrogen of target asparagine residues [34, 35]. The entire catalytic cycle is thereby completed through these coordinated membrane-dependent processes.

Fig. 1.

Simplified pathway for biosynthesis of N-linked glycans. Reproduced from Ref. [31]: Sobala et al. Structure of human endo-α-1,2-mannosidase (MANEA), an antiviral host-glycosylation target. Proc. Natl. Acad. Sci. U.S.A. 2020;117(46):29,595–29,601, 10.1073/pnas.2013620117. Licensed under CC BY-NC-ND 4.0, https://creativecommons.org/licenses/by-nc-nd/4.0/. No changes were made to the original figure

Following the primary glycosylation event, nascent glycoproteins are translocated to the Golgi apparatus, where comprehensive glycan remodeling is executed through sequential enzymatic modifications. During this post-ER processing phase, surplus glucose and mannose residues are enzymatically excised by specific glycosidases, while supplementary carbohydrate units such as galactose and sialic acid are stereoselectively incorporated through glycosyltransferase-mediated reactions, culminating in the formation of mature complex/hybrid-type glycans [35–37]. These post-translational modifications are essential for the functional diversification of glycoproteins, influencing their cellular localization, secretion efficiency, and interactions within the extracellular environment.

Protein Folding and Quality Control in the ER

Protein folding and quality control is a complex, multifaceted process primarily occurring in the ER, involving molecular chaperones, glycan-trimming enzymes, and degradation pathways to ensure the proper maturation of proteins. This system is critical for maintaining cellular homeostasis by promoting correct protein folding and targeting irreparably misfolded proteins for degradation [38, 39].

Glucosidase I and II are α-glucosidases located in the ER that are integral to the N-glycosylation-dependent protein folding quality control system [40]. These enzymes sequentially cleave glucose residues from nascent glycoproteins, modulating their interactions with the molecular chaperones calnexin (CNX) and calreticulin (CRT). This trimming process is critical for maintaining the equilibrium between protein folding and degradation pathways. Glucosidase I (Glu I) initiates the trimming by hydrolyzing the outermost α-1,2-linked glucose from the initial N-glycan precursor, (GlcNAc)₂Man₉Glc₃, yielding (GlcNAc)₂Man₉Glc₂ (Fig. 1). This step is essential for subsequent processing and activates the CNX/CRT cycle, as elucidated in structural studies of glucosidase II [41]. Subsequently, glucosidase II (Glu II) removes the α-1,3-linked glucose from (GlcNAc)₂Man₉Glc₂, producing the monoglucosylated (GlcNAc)₂Man₉Glc, which serves as the ligand recognized by CNX and calreticulin. Glu II can further excise the remaining glucose, generating (GlcNAc)₂Man₉, which results in the dissociation of the glycoprotein from these chaperones [42]. In the protein folding surveillance mechanism, Glu II collaborates with UDP-glucose:glycoprotein glucosyltransferase (UGGT). If a glycoprotein is released from CNX or CRT in a misfolded state following glucose trimming, UGGT reglucosylates the N-glycan to (GlcNAc)₂Man₉Glc, facilitating re-association with the chaperones for additional folding attempts [43–45]. Thus, Glu II functions as a pivotal regulator, governing the timing of chaperone release and enabling multiple folding cycles.

The CNX/CRT cycle constitutes a critical quality control mechanism within the ER to ensure the proper folding of glycoproteins destined for secretion or membrane integration [46]. This process hinges on the dynamic regulation of N-glycan processing and the coordinated activity of lectin chaperones. Specifically, calnexin, a membrane-bound chaperone, and calreticulin, its soluble counterpart, bind to monoglucosylated N-glycans on glycoproteins via their lectin domains, thereby shielding hydrophobic regions of unfolded or partially folded proteins to prevent aggregation. These chaperones further recruit auxiliary proteins, such as the thiol-oxidoreductase ERp57, to facilitate the formation of correct disulfide bonds, a pivotal step in achieving proper protein conformation [47]. Upon attaining its native structure, the glycoprotein is acted upon by glucosidase II, which removes the terminal glucose residue from the N-glycan, triggering dissociation from CNX/CRT and enabling subsequent transport to the Golgi apparatus. Conversely, if folding remains incomplete, UGGT reglucosylates the N-glycan, permitting the protein to re-associate with CNX/CRT for additional folding attempts. This iterative cycle of folding, quality assessment, and refolding persists until the glycoprotein either achieves its correct conformation or is deemed irreparable and targeted for degradation [48–50].

Collectively, N-glycosylation is a highly regulated post-translational modification that increases the functional diversity of proteins through the attachment of specific glycan structures. It is essential for protein folding quality control and numerous cellular processes, and is extensively exploited by viruses to ensure proper folding and stabilization of viral proteins, as well as to facilitate viral particle assembly, secretion, and evasion of host immune responses.

Promoting the Correct Folding and Stability of Viral Proteins

N-Glycosylation facilitates the recruitment of molecular chaperones in the ER, such as CNX and CRT, which are critical for assisting viral proteins in achieving their correct three-dimensional conformation. This proper folding is essential for the biological functionality of viral proteins, as many require specific spatial structures to fulfill their roles [51, 52]. Additionally, N-glycosylation prevents misfolding and subsequent degradation. Incorrectly folded proteins are often targeted for degradation via the host’s ERAD pathway. By enhancing folding efficiency, N-glycosylation reduces the likelihood of degradation, ensuring that viral proteins successfully progress to subsequent secretion and assembly stages [43].

Furthermore, N-glycosylation enhances the thermal and chemical stability of viral proteins. It also protects these proteins from protease-mediated degradation by shielding specific surface regions with glycan chains, thereby decreasing recognition and cleavage by host proteases [53, 54]. Consequently, glycosylated viral proteins exhibit an extended half-life, providing the virus with additional time for replication, assembly, and release. Owing to the hydrophilicity of glycans, oligosaccharides shield cleavage sites to enhance the solubility and/or proteolytic stability of proteins, thereby extending their half-life. Additionally, glycans stabilize proteins by engaging in key interactions with adjacent amino acids, thus improving their thermal stability and slowing down aggregation kinetics [55, 56]. Moreover, a deletion at amino acid position 131 in the head region of the H5N6 virus HA protein resulted in the formation of a novel N-linked glycosylation site at position 129. This modification significantly enhanced the adaptability of the H5N6 virus to mammalian hosts by lowering the pH threshold required for membrane fusion and concurrently improving the acid and thermal stability of the HA protein. These findings indicated that N-linked glycosylation of the HA protein plays a critical role in regulating viral adaptability through the modulation of protein stability [57].

Disguising Viral Proteins to Evade Immune System Recognition

N-Glycosylation is a critical mechanism that enables viral proteins to evade immune recognition, facilitating viral survival and dissemination within the host. This process involves the attachment of glycans to viral proteins, which serves multiple immunomodulatory functions [20, 58, 59]. Specifically, glycans can shield antigenic epitopes on the surfaces of viral proteins, impairing recognition by the host immune system and reducing antibody binding efficiency.

Bedsides, glycosylation can disrupt the activation of the complement system, allowing viruses to avoid immune-mediated clearance [60]. Notable instances include the HA of the influenza virus and the gp120 protein of HIV-1 [24, 61]. These proteins are extensively glycosylated, with glycans forming a protective shield that effectively conceals antigenic epitopes. Sagar et al. reported that N-linked glycosylation of the HIV-1 V1–V2 envelope loop alters the virus’s sensitivity to neutralizing antibodies during infection [62]. Similarly, Wolk et al. demonstrated that N-linked glycans within the V1–V2 loop of 15 variants of HIV-1 NL4–3 gp120 are essential for viral infectivity and confer reduced sensitivity to serum antibodies [63].

Mediating Viral Receptor Recognition and Invasion

Glycoproteins on enveloped viruses bind to specific host cell receptors, initiating membrane fusion to facilitate viral entry. N-Glycans on these glycoproteins are crucial for receptor recognition [58]. A prominent example is HIV-1 gp120, one of nature’s most highly glycosylated proteins. gp120 not only modulates viral envelope conformation but is also pivotal for viral entry and infectivity. The well-established mechanism involves sequential binding of gp120 to the cluster of differentiation 4 (CD4) receptor and chemokine co-receptors (CCR5 or CXCR4) on target cells [64]. Research demonstrated that removal of N-glycans from gp120 via endoglycosidase treatment or mutagenesis of N-glycosylation sites significantly impaired CD4 binding [65, 66]. Similarly, the HA protein of the H5N1 avian influenza virus illustrates the critical role of glycosylation in receptor recognition and pathogenicity. The glycosylation state of HA, particularly at the 158N site, directly influences H5N1’s interaction with host cell receptors. Glycosylation at this site alters the HA protein’s spatial conformation, modulating the virus’s affinity for α-2,3 sialic acid receptors (common in avian hosts) versus α-2,6 sialic acid receptors (prevalent in humans). The presence of the 158N glycan typically reduces affinity for human receptors, maintaining preference for avian receptors, which may limit human-to-human transmission [67, 68]. Conversely, deglycosylation at this site can expose antigenic epitopes, potentially increasing immunogenicity but also enhancing pathogenicity in animal models such as ferrets, owing to altered interactions with the host immune system.

However, the role of N-glycans varies considerably among viruses. Wen et al. demonstrated that elimination of N-glycosylation sites on the Zika virus (ZIKV) envelope protein did not affect ZIKV infection or replication in mammalian cell culture systems but prevented the virus from traversing the mosquito midgut barrier. This barrier is critical for flavivirus transmission, as mosquitoes acquire the virus during blood feeding, with the midgut serving as the primary infection site [69]. Conversely, removal of N-glycans proximal to the highly conserved receptor-binding domain of the Ebola virus (EBOV) glycoprotein in Vero cells enhances viral entry efficiency [70]. Zhang et al. demonstrated that the absence of the 158N glycosylation site in H5N1 increased the pathogenicity in mice [67]. Similarly, within the same evolutionary clade, Gao et al. found that the loss of the 158N glycosylation site on the HA protein of the H5N6 subtype of avian influenza virus also enhanced pathogenicity in mice and elicited stronger host immune responses in mammals [71, 72]. Collectively, these findings indicated that N-glycans exhibit diverse, context-dependent regulatory functions in viral recognition and invasion.

Affecting Viral Release and Infection Efficiency

Enveloped viruses typically release progeny virions through budding [73, 74]. During this process, the viral envelope is derived from the host cell membrane, and N-glycans on viral proteins can influence the efficiency of virion release. Glycoprotein C (gC) of HSV-1 mediates viral attachment to susceptible host cells by interacting with cell surface glycosaminoglycans (GAGs). Notably, gC contains a mucin-like domain (MLD) adjacent to the GAG-binding site. Compared with wild-type HSV-1, mutant strains lacking the MLD in gC display reduced binding affinity to host cells and diminished release of viral particles from infected cells [75]. West Nile virus (WNV) encodes two envelope proteins: the premembrane protein (prM) and the envelope protein (E). Deletion of the glycosylation site in prM in WNV genotype I or II strains reduces the release of subviral particles (SVPs). Similarly, removal of the glycosylation site in the E protein of genotype I strains yields comparable effects, whereas introduction of this site into genotype II strains that naturally lack it enhances SVP production [76]. These findings indicate that glycosylation of WNV prM and E proteins modulates viral release and infection efficiency.

Mechanisms of Antiviral Activity by Targeting N-Glycosylation

Inhibition of Glycosyltransferase Activity

Glycosyltransferases (GTs) are pivotal enzymes that catalyze the transfer of sugar moieties from donor to acceptor molecules, thereby establishing glycosidic linkages [77]. In the context of viral infections, these enzymes assume a critical role as viruses frequently co-opt the host’s glycosylation machinery for their own N-glycosylation processes [78, 79]. Despite extensive research, the development of effective competitive inhibitors targeting GTs remains limited, and overall progress in inhibition strategies has been modest. Nonetheless, several strategies employing small molecules, including alternative primers, chain terminators, and modified substrates, have been investigated [11]. Oligosaccharyltransferase (OST), a pivotal glycosyltransferase complex, catalyzes the initial step of N-glycosylation within the ER. In mammalian systems, OST comprises two catalytic subunits: STT3A and STT3B [80, 81]. Using pull-down assays, Huang et al. demonstrated that severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) structural proteins interact with both STT3A and STT3B [82]. Notably, the N-glycosylated envelope (E), membrane (M), and spike (S) proteins of SARS-CoV-2 exhibited robust interactions with both isoforms. Moreover, inhibition of either STT3A or STT3B via small interfering (siRNA) or clustered regularly interspaced short palindromic repeats (CRISPR)-CRISPR-associated protein 9 (Cas9) single guide RNA (sgRNA) resulted in a significant reduction in SARS-CoV-2 infection. Collectively, these findings suggest that both OST isoforms modulate the functions of the E, M, and S proteins, thereby underscoring the essential role of OST in the pathogenesis of SARS-CoV-2.

NGI-1, an aminobenzamide sulfonamide (Fig. 2A), is a small-molecule inhibitor of OST. This compound demonstrates inhibitory effects against SARS-CoV-2 spike protein N-glycosylation, exhibiting a half-maximal inhibitory concentration (IC₅₀) of 0.860 μM and a selectivity index of 509. Quantitative mass spectrometry was utilized to assess N-glycosylation at spike protein sites N1074 and N1194. The results indicated a significant reduction in N-glycosylated peptide proportions: from 12.3% to 2.1% at N1074 and from 87.4% to 65.9% at N1194. These data confirm that NGI-1 effectively suppresses SARS-CoV-2 spike protein N-glycosylation [82]. Notably, NGI-1 also displays broad-spectrum antiviral activity against flaviviruses including dengue and Zika viruses, though its mechanism of action in these contexts may involve host factors distinct from N-glycosylation inhibition [83].

Fig. 2.

Structures of NGI (A) and tunicamycin (B)

Tunicamycin, a nucleoside antibiotic mixture (Fig. 2B), exhibits significant biological activity by specifically inhibiting N-acetylglucosamine-1-phosphotransferase (GPT). This inhibition occurs through structural mimicry of UDP-N-acetylglucosamine (UDP-GlcNAc), effectively preventing the initial step of N-glycosylation, the transfer of GlcNAc-1-phosphate to the dolichol phosphate carrier [84, 85]. Owing to its broad inhibition of host cell glycosylation pathways, tunicamycin induces substantial cytotoxicity, thereby precluding direct clinical applications [86]. Consequently, it serves primarily as an experimental probe for investigating protein glycosylation mechanisms.

Gabriela et al. synthesized thioglycosyl substrate analogs designed to competitively inhibit glycosyltransferases [87]. Specifically, novel compounds 13 and 14 exhibited potent inhibition of the envelope glycoproteins E2 from classical swine fever virus (CSFV) and E1 from hepatitis C virus (HCV). This activity likely stems from competitive blockade of viral glycosylation pathways. Critically, both glycoproteins feature extensive N-glycosylation, with E2 and E1 containing six and five N-glycosylation sites, respectively. Notably, the precise glycosyltransferase(s) targeted by these compounds remains unidentified.

Glucosidase Inhibitors Inhibit Viral N-Glycosylation Trimming

Glucosidase inhibitors primarily exert their antiviral effects by disrupting the early processing of N-linked glycans. Following the attachment of the initial glycan, inhibition of glucosidase I and II prevents the removal of glucose residues [88, 89]. This retention of glucose on the viral protein impairs its interaction with chaperone proteins such as calnexin and calreticulin, leading to misfolding and subsequent degradation by the proteasome [90].

The HIV-1 envelope glycoprotein, comprising gp120 (surface subunit) and gp41 (transmembrane subunit), undergoes extensive N-glycosylation with approximately 26–30 N-linked glycosylation sites occupying 50–70% of its surface [91, 92]. These N-glycans include both high-mannose and complex-type structures, the latter containing N-acetylglucosamine, galactose, and sialic acid (Fig. 3A) [93]. Studies demonstrated that N-butyl-deoxynojirimycin (NB-DNJ) treatment inhibits syncytia formation and infectious HIV-1 production [94]. In contrast, deoxymannojirimycin (DMJ), an inhibitor of Golgi-resident mannosidase I, did not impair viral secretion. This differential effect indicated that α-glucosidase-mediated glycan trimming is essential for productive CNX/CRT-assisted folding. Conformation-specific antibody analysis further revealed that NB-DNJ induced structural alterations in the gp120 V1/V2 loop, demonstrating that α-glucosidase inhibition resulted in localized misfolding [95]. Although gp120 trafficking to the plasma membrane and viral budding remained unaffected, the conformational changed prevent cell-surface exposure of gp41, thereby blocking viral fusion. Collectively, these findings established that NB-DNJ reduced infectious HIV-1 secretion by disrupting post-CD4-binding conformational changes, specifically impairing gp41 exposure required for membrane fusion [96].

Fig. 3.

(A) Visual representation of the HIV-1BaL BaL/SUPT1-R5 cell line gp120 secondary structure. The general type of glycosylation (high mannose, complex, hybrid) observed at each site is indicated by color coding, with the key given within the figure. Reproduced from Ref. [113]: Maria Panico et al. Mapping the complete glycoproteome of virion-derived HIV-1 gp120 provides insights into broadly neutralizing antibody binding. Sci. Rep. 6, 32,956 (2016), 10.1038/srep32956. Licensed under CC BY 4.0, http://creativecommons.org/licenses/by/4.0/. No changes were made to the original figure. (B) HBV envelope proteins. All three proteins share a common N-linked glycosylation site at amino acid position 146 (labeled Asn-146) within the S domain. The MHBs protein contains an additional glycosylation site at amino acid position 4 of the pre-S2 domain (labeled Asn-4)

The hepatitis B virus (HBV) genome encodes three envelope glycoproteins: large (LHBs), middle (MHBs), and small (SHBs). These proteins are translated from a single open reading frame but employ distinct start codons (Fig. 3B) [97, 98]. In sharp contrast to the HIV-1 envelope glycoprotein, HBV envelope glycoproteins possess only two conserved N-glycosylation sites: Asn-146 in the S domain and Asn-4 in the preS2 domain [99, 100]. Treatment with NB-DNJ, a competitive α-glucosidase inhibitor, significantly reduces secretion of HBV virions. This suppression is most pronounced for MHBs, which displays heightened sensitivity to glucosidase inhibition. Critically, viral DNA secretion exhibits sustained suppression for multiple days post-inhibitor withdrawal [101].

Dengue virus (DENV), a member of the Flaviviridae family, is a single-stranded, positive-sense RNA virus primarily transmitted by mosquitoes and responsible for diseases such as dengue fever and severe dengue [102]. DENV comprises four serotypes (DENV-1 to DENV-4), and its genome encodes three structural proteins (capsid protein C, premembrane/membrane proteins prM/M, and envelope protein E) and seven nonstructural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5). In DENV, glycosylation primarily occurs on the E, prM/M, and NS1 proteins, whereas other proteins such as NS4B may exhibit minor glycosylation [103]. The E protein contains two potential N-glycosylation sites: Asn-67 and Asn-153. The Asn-67 site is unique to DENV, whereas the Asn-153 site is conserved across most flaviviruses. The prM protein contains an N-glycosylation site (with positions varying by serotype, such as Asn-69 in DENV-2); glycosylation facilitates proper folding of prM and protects viral particles. The NS1 protein contains two highly conserved N-glycosylation sites: Asn-130 and Asn-207 [103, 104]. Early in vitro studies demonstrated that the α-glucosidase inhibitor castanospermine inhibits all four DENV serotypes in a dose-dependent manner, with efficacy superior to that of DNJ [105, 106]. Castanospermine disrupts glucose trimming, thereby impairing the folding efficiency of prM and E proteins; this results in delayed and unstable formation of prME heterodimers and a significant reduction in DENV particle secretion. Subsequently developed α-glucosidase inhibitor derivatives (such as Bu-CAST, CM-10-18, and UV-4) exhibit antiviral efficacy that varies by DENV serotype. Bu-CAST shows in vitro efficacy against DENV-2 up to 100-fold higher than that of CAST and approximately 2-fold higher in vivo efficacy, while CM-10-18 and UV-4 demonstrate potent in vitro inhibition with low cytotoxicity. Early perspectives considered inhibition of α-glucosidase II as the key mechanism for anti-DENV activity, as Bu-CAST exhibits superior inhibition of this enzyme and antiviral effects compared with NB-DNJ. However, a single high-dose administration of UV-4B, through selective inhibition of α-glucosidase I, prevented mouse mortality when given 48 h after lethal DENV infection, thereby challenging the previous viewpoint [107, 108].

Severe fever with thrombocytopenia syndrome virus (SFTSV), a member of the Bunyaviridae family, is a single-stranded, negative-sense RNA virus [109]. Its genome consists of three segments (L, M, and S), with the M segment encoding an envelope glycoprotein precursor that is processed by host cell proteases to form the N-terminal glycoprotein (Gn) and C-terminal glycoprotein (Gc). Gn has two N-linked glycosylation sites at Asn-33 and Asn-63. Gc exhibits three N-linked glycosylation sites in its extracellular domain: Asn-853, Asn-914, and Asn-936 [110, 111]. Studies have found that NN-DNJ, an α-glucosidase inhibitor, reduces the production of infectious SFTSV, suggesting that the glycosylation process of SFTSV Gn may be inhibited [112]. The relevant mechanisms require further in-depth research to elucidate.

Other Potential Mechanisms of Glucosidase Inhibitors

Woodchuck hepatitis virus (WHV) exhibits high genomic, replicative, and pathological similarity to HBV, establishing WHV-infected woodchucks as a validated animal model for studying HBV infection, chronic hepatitis, and hepatocellular carcinoma (HCC) [114, 115]. Research demonstrated that the glucosidase inhibitor N-nonyl-DNJ reduced viremia in chronically infected woodchucks in a dose-dependent fashion. Notably, N-nonyl-DNJ (Fig. 4) displays 100- to 200-fold greater anti-HBV potency than NB-DNJ [116]. Serological analysis further revealed that N-nonyl-DNJ specifically impeded HBV glycoprotein maturation without altering secretion of other glycoproteins. Although targeting ER α-glucosidases, N-nonyl-DNJ maintained antiviral efficacy at sub-enzyme-inhibitory concentrations. Crucially, whereas glucosidase inhibition (e.g., by NB-DNJ) induced intracellular accumulation of defective virions and elevated HBV replication intermediates, N-nonyl-DNJ treatment markedly reduced HBV replication forms. Current hypotheses propose that N-nonyl-DNJ may interfere with pregenomic RNA encapsidation, accelerate the degradation of viral core particles, or affect the primer reaction required for reverse transcription. Alternatively, it may inhibit interactions between viral core proteins and specific cellular factors, leading to improper nucleocapsid assembly and stability [117]. Collectively, N-nonyl-DNJ represents a novel anti-HBV agent class that operates through distinct, glucosidase-independent mechanisms, potentially offering complementary activity to nucleoside analogs.

Fig. 4.

Structures of iminosugars as glucosidase inhibitors

Targeting the Mannosidase Modification Process

Endo-α-1,2-mannosidase (MANEA), the sole endoglycosidase involved in N-linked glycan trimming within the Golgi apparatus, trims glycoproteins that bypass the modification processes mediated by glucosidases I and II in the ER [31, 118]. Modulating MANEA activity establishes an alternative glycoprotein maturation route, termed the endogenous mannosidase pathway, which operates independently of classical glucosidase processing [119]. Research indicated that MANEA inhibitors, exemplified by α-d-glucopyranosyl-1,3-deoxymannojirimycin (GlcDMJ, Fig. 4), did not impact viral RNA synthesis. Nevertheless, plaque reduction assays demonstrated that GlcDMJ exhibited potential as a broad-spectrum antiviral agent, as alterations in envelope protein glycosylation diminish the infectivity of progeny viruses, including bovine viral diarrhea virus (BVDV) and dengue virus [31]. Furthermore, the superior antiviral efficacy of GlcDMJ compared with analogous inhibitors may be attributed to factors such as enhanced enzyme binding affinity and improved cellular permeability. GlcDMJ mimics the substrate structure and enhances cellular permeability through the addition of a glucose moiety at the 3-position of DMJ, while also leveraging interactions with the −2 subsite of enzyme. These modifications enhanced the specificity and functionality of GlcDMJ, notwithstanding its overall reduced binding affinity [120]. Consequently, mannosidase-targeting inhibitors may disrupt the N-glycosylation of viral glycoproteins by host enzymes, thereby attenuating viral infectivity.

Recent advances have elucidated the role of MANEA in HBV morphogenesis. Under pharmacological inhibition of α-glucosidases (e.g., miglustat), HBV particles were devoid of M protein and exhibit impaired secretion of DNA-containing virions [121]. This phenotype arises from differential sensitivity among the three HBV envelope glycoproteins (LHBs, MHBs, and SHBs) to glucosidase activity. Notably, dual blockade of α-glucosidase I/II induced retention of triglucosylated glycan moieties on MHBs, resulting in aberrantly processed glycoproteins that undergo lysosomal targeting. In contrast, LHBs and SHBs maintain mature glycan structures through MANEA-mediated trimming in the Golgi compartment [97].

Targeting the ERAD Pathway

ERAD is a fundamental protein quality control mechanism that facilitates the clearance of misfolded or improperly assembled proteins from the endoplasmic reticulum via the ubiquitin–proteasome system (Fig. 5). This pathway comprises several essential components, including valosin-containing protein (VCP), E3 ubiquitin ligases (e.g., HRD1), and α-mannosidases (e.g., EDEM1 and EDEM2). Research exploring antiviral strategies targeting ERAD has highlighted the potential of this approach, albeit with variable efficacy depending on the specific viral species. Notably, ERAD inhibition may serve as an effective antiviral strategy against certain viruses, whereas augmenting ERAD activity could enhance host defense mechanisms against other viral infections.

Fig. 5.

a Schematic of the ERAD pathway. Major steps include: (1) recognition and recruitment of misfolded proteins to the ERAD machinery; (2) retrotranslocation of substrates across the ER membrane; (3) ubiquitination of substrates by E3 ligase; (4) extraction of polyubiquitinated substrates by the ATPase valosin-containing protein (VCP) complex; and (5) subsequent degradation by the 26S proteasome. b Schematic of the mammalian SEL1L–HRD1 core complex and associated cofactors, including OS9 and ERLEC1 (substrate adaptors), DERL (a retrotranslocation facilitator), HERP and FAM8A1 (regulatory components), the E2 ubiquitin-conjugating enzyme UBE2J1, and the AAA⁺-ATPase VCP complex. c A model of the yeast Hrd1–Hrd3 complex, which can exist in either monomeric or homodimeric configuration. d Cryo-electron microscopy structure of the monomeric yeast Hrd1–Hrd3–Der1–Usa1 complex. e Representative list of mammalian SEL1L–HRD1 ERAD substrates, including those linked to genetic mutations (indicated by an asterisk). Reproduced from Ref. [122]: Wang et al. SEL1L–HRD1-mediated ERAD in mammals. Nat. Cell. Biol., 27, 1063–1073 (2025). 10.1038/s41556-025-01690-1. Copyright © 2025, Springer Nature Limited. No changes were made to the original figure

Regarding inhibition of the ERAD pathway to exert antiviral activity, studies have demonstrated that the ERAD pathway plays a pivotal role in ZIKV and DENV replication, with disruption of ERAD flux significantly impairing viral propagation. For instance, eeyarestatin I (EEY), an inhibitor of VCP within the ERAD pathway, exhibited potent antiviral activity against Zika and Usutu viruses [123]. Similarly, the proteasome inhibitor bortezomib demonstrated antiviral efficacy against Zika and dengue viruses by impairing ERAD-mediated protein degradation. Mechanistic studies have revealed that, rather than directly inhibiting NS2B3 protease activity, bortezomib could trigger extensive ubiquitination and subsequent aggregation of NS3 viral protein through coordinated action of two E3 ubiquitin ligases: HRD1 and RNF126. This post-translational modification cascade results in significant attenuation of NS3 protease functionality within host cells, consequently disrupting proper processing of viral polyprotein precursors and effectively suppressing viral replication [124]. Moreover, EDEM proteins (EDEM1, EDEM2, and EDEM3), critical regulators of the ERAD pathway, mediate the proteasomal degradation of misfolded glycoproteins through demannosylation activity (Fig. 5). Experimental evidence has demonstrated that EDEM1, EDEM2, and ER α-1,2-mannosidase I (ERManI) specifically target the hemagglutinin (HA) glycoprotein of influenza A virus (IAV) via ERAD-dependent mechanisms. Notably, shRNA-mediated knockdown of these ERAD components led to sustained HA accumulation, accompanied by a significant increase in viral titers, observed 36 to 48 h post-infection [124].

Regarding activation of the ERAD pathway to exert antiviral activity, viral infections commonly activate ER stress responses in host cells, thereby initiating the ERAD pathway to ubiquitinate and degrade misfolded or unassembled glycoproteins through proteasomal mechanisms [125, 126]. In hepatitis C virus (HCV) infection, studies have demonstrated ER stress activation via the IRE1–XBP1 arm of the unfolded protein response (UPR), with ERAD components EDEM1 and EDEM3 critically regulating post-translational processing of viral glycoproteins [120]. Experimental evidence revealed that siRNA-mediated knockdown of EDEM1 and EDEM3 enhanced HCV production by 3.1-fold and 2.3-fold, respectively, whereas ectopic EDEM1 expression reduced viral yield by 2.4-fold. Inhibition of the ERAD pathway with kifunensine, a potent inhibitor of ER mannosidases, increased HCV secretion in a dose-dependent manner [124]. Similar to HCV, HBV envelope proteins are subject to degradation via the ERAD pathway. HBV infection upregulates the levels of EDEM proteins, particularly EDEM1. Studies have demonstrated that EDEM1 interacts with HBV envelope proteins regardless of viral replication or nucleocapsid protein expression. Overexpression of EDEM1 accelerated the degradation of HBV envelope proteins, while suppression of endogenous EDEM1 enhanced their stability, thereby modulating the pool of native peptides available for subviral particle (SVP) assembly and viral envelopment [120]. Notably, mitochondrial translocator protein (TSPO) has emerged as an ERAD-associated regulator of HIV-1 replication. Mechanistic analyses indicated that TSPO impaired HIV-1 Env glycoprotein folding, triggering ERAD-mediated degradation and subsequent suppression of viral replication. Elevated TSPO expression correlated with accelerated Env proteostasis failure, characterized by misfolded protein accumulation and ERAD-dependent clearance. Crucially, kifunensine-mediated ERAD inhibition rescued Env expression and restored viral replication kinetics, confirming the pathway’s central role. These findings position TSPO as a promising target for antiretroviral therapies exploiting host protein quality control networks [127, 128].

Clinical Efficacy and Adverse Effects of Investigational Antiviral Agents

Miglustat (NB-DNJ), an approved therapeutic for Gaucher disease and Niemann–Pick type C, reduces glycolipid synthesis via inhibition of glucosylceramide synthase [129–132]. Recent investigations have highlighted its antiviral potential against pathogens such as HIV-1 and SARS-CoV-2 [133, 134]. Common adverse effects include nausea, vomiting, abdominal pain, diarrhea, weight loss, headache, muscle cramps, dizziness, and asthenia [130, 135, 136]. These symptoms typically manifest during initial treatment phases and may be mitigated through dietary modifications (e.g., reduced carbohydrate intake) or adjunct medications (e.g., loperamide) [137]. However, phase I/II clinical trials in HIV-1 demonstrated limited efficacy, attributable to suboptimal potency and challenges in achieving steady-state therapeutic concentrations [137].

Celgosivir, a prodrug of castanospermine (Fig. 4), inhibits α-glucosidase I to disrupt viral N-glycosylation processes, primarily targeting HIV-1, HCV, and dengue virus [138–140]. In dengue phase I trials, celgosivir exhibited favorable tolerability, with mild-to-moderate diarrhea as the predominant adverse event and no severe cases reported. Notably, dengue-associated symptoms (e.g., fever and myalgia) or known celgosivir-related side effects showed no significant exacerbation [124]. While celgosivir monotherapy demonstrated limited efficacy in HCV, synergistic effects were observed in combination regimens with pegylated interferon alpha (IFNα)-2b and ribavirin. Its potential utility in acute dengue infections remains investigational, given the requirement for short-term therapy compared with chronic HIV-1 or hepatitis C protocols. A phase 1b, randomized, double-blind, placebo-controlled proof-of-concept study conducted in Singapore evaluated the efficacy and safety of celgosivir in patients with acute dengue [124]. The celgosivir group received an initial dose of 400 mg, followed by 200 mg every 12 h for a total of nine doses over 5 days, while the placebo group received a matching placebo. The primary endpoints included the mean virological log reduction (VLR) from baseline on days 2, 3, and 4, and the area under the fever curve (AUC, > 37 °C) from 0 to 96 h, with results showing no significant differences. This outcome may reflect accelerated viral clearance in secondary dengue infections, potentially masking celgosivir’s therapeutic effect owing to robust immune responses that enhance natural viral suppression. As an α-glucosidase inhibitor, celgosivir targets glycosylation of dengue viral prM, E, and NS1 proteins but may concurrently modulate ER stress-related pathways [106, 108, 141]. The indirect effect of celgosivir on ER stress may modulate virus replication, which may be one of the reasons why the reduction in viremia was less than expected [124].

Future Prospects

N-Glycosylation, a pivotal post-translational modification in viral pathogenesis, presents an attractive therapeutic target for developing broad-spectrum antiviral agents. Emerging evidence demonstrated that N-glycosylation inhibitors exhibit remarkable antiviral potential against diverse enveloped viruses including HIV-1, SARS-CoV-2, influenza virus, and dengue virus, offering promising solutions for combating emerging/reemerging viral threats. However, optimizing the selectivity index (SI), a crucial pharmacological parameter reflecting the therapeutic window between antiviral efficacy and host cytotoxicity, remains a fundamental challenge in drug development [107]. Structural optimization of α-glucosidase inhibitors (e.g., iminosugars) reveals critical structure–activity relationships: The head group primarily mediates enzyme inhibition, while the N-alkyl chain length modulates lipophilicity and cellular pharmacokinetics [116, 142]. Notably, elongation of the N-alkyl chain generally enhances the antiviral potency but concomitantly increases the cytotoxicity. This paradox is exemplified by DNJ, which demonstrates 20–40-fold greater potency against BVDV and 100-fold enhanced activity against HBV compared with its N-butyl counterpart (NB-DNJ) [143, 144]. Intriguingly, N-nonyl-DNJ manifests glucosidase-independent antiviral mechanisms, underscoring the complexity of structure–activity relationships in this drug class [145].

Another critical aspect in enhancing the therapeutic potential of antiviral agents lies in the development of specific inhibitors. Recent studies, particularly post the COVID-19 pandemic, have highlighted the significance of targeting conserved viral lifecycle pathways. N-linked glycosylation, a process frequently exploited by viruses, positions STT3A/B as a promising therapeutic target. However, given the pivotal role of glycosylation in fundamental human physiological processes, the toxicological implications of STT3A/B inhibition must be carefully balanced against its antiviral efficacy. The study revealed that compound 17, the most potent N-linked glycosylation inhibitor among NGI-1 derivatives, nearly completely suppressed neuronal firing in both neurons and astrocytes following a 48-h incubation period [146]. The findings demonstrated that chronic interference with the N-linked glycosylation pathway under conditions mimicking STT3A/B inhibitor exposure may induce neurotoxic effects. Therefore, the development of specific inhibitors represents a key approach in antiviral therapeutic strategies. Previous studies have shown that NGI-1 inhibits glycosylation in both STT3A- and STT3B-knockout cell lines, indicating that this small molecule simultaneously targets both OST complexes [147]. Structure–activity relationship analysis of NGI-1 revealed that its heterocyclic amine is crucial for inhibitory activity against STT3A. An NGI-1 analog, compound 19, featuring a reversed amide and phenyl substitution, exhibits STT3B-specific inhibition while maintaining comparable potency [147]. Unlike NGI-1, which reduces EGFR glycosylation and cell surface expression, compound 19 does not affect EGFR glycosylation or activation. For COX-2, which possesses an STT3B-dependent C-terminal N-linked glycosylation site essential for protein stability, compound 19 reduces its N-linked glycosylation, thereby stabilizing the protein and increasing overall COX-2 activity. Furthermore, unlike tunicamycin, which inhibits glycan precursor synthesis and induces a significant UPR and cytotoxicity, small molecules like compound 19 can reduce N-linked glycosylation without substantially activating the UPR or causing cytotoxicity. It is also important to note that the loss of N-linked glycosylation in certain proteins, such as STT3B target proteins, may have detrimental physiological effects.

Furthermore, coordinated potentiation represents an effective strategy for optimizing antiviral therapeutic approaches. Studies have demonstrated that simultaneous knockout of both ER α-glucosidases I and II failed to generate viable cells, suggesting significant cytotoxicity may occur when these enzymes were fully inhibited in vitro [148]. This underscores the necessity of combination therapy to mitigate potential cytotoxicity associated with enhanced ER α-glucosidase suppression. Combination regimens employing α-glucosidase inhibitors alongside established antiviral agents, such as ribavirin, PEGylated IFNα-2b, or IFN-α, may achieve enhanced or synergistic antiviral activity through dual or triple drug combinations [148–151]. Importantly, such combinatorial strategies reduce peak dosages of individual drugs, thereby balancing antiviral efficacy with pharmacological toxicity.

It is critical to emphasize that O-glycosylation of proteins is as significant as N-glycosylation in viral biology. For example, the HIV envelope glycoprotein gp120 exhibits extensive N- and O-glycosylation. Notably, O-glycosylation of the Ebola virus glycoprotein plays a more pivotal role in its in vivo infectivity than N-glycosylation. Thus, O-glycosylation, alongside N-glycosylation, serves as a crucial regulatory mechanism in virus–host interactions. Future antiviral strategies targeting glycosylation should prioritize a deeper understanding of specific glycan chain structures and functions on viral proteins, the development of selective regulatory tools, and the exploration of combination therapies that integrate these approaches with existing treatments to enhance therapeutic efficacy.

Conclusions

Research into antiviral drugs targeting N-glycosylation modification holds significant promise, although it remains in its early stages. In vitro and in vivo experiments have demonstrated the potential of candidate drugs such as STT3A/B inhibitors and α-glucosidase inhibitors. However, clinical trial data are currently limited, and no such drugs have yet been approved. Looking ahead, advancements through structural modifications, targeted inhibitors, and combination therapies could position N-glycosylation inhibitors as crucial tools for broad-spectrum antiviral treatment. Nonetheless, challenges related to specificity, safety, and resistance must be further addressed to facilitate a successful transition from laboratory research to clinical application.

Author Contributions

Dan Wang: conceptualization, formal analysis, investigation, writing—original draft. Zihao He: conceptualization, formal analysis, investigation. Zehong Chen: conceptualization, investigation, visualization, supervision, validation.

Funding

No funding or sponsorship was received for this study or publication of this article. The Rapid Service Fee was funded by the authors.

Data Availability

All data generated or analyzed during this study are included in this published article/as supplementary information files.

Declarations

Conflict of Interest

Dan Wang, Zihao He and Zehong Chen declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.