Glycan-mediated molecular interactions in bacterial pathogenesis

Abstract

Glycans are expressed on the surface of nearly all host and bacterial cells. Not surprisingly, glycan-mediated molecular interactions play a vital role in bacterial pathogenesis and host responses against pathogens. Glycan-mediated host–pathogen interactions can benefit the pathogen, host, or both. Here, we discuss (i) bacterial glycans that play a critical role in bacterial colonization and/or immune evasion, (ii) host glycans that are utilized by bacteria for pathogenesis, and (iii) bacterial and host glycans involved in immune responses against pathogens. We further discuss (iv) opportunities and challenges for transforming these research findings into more effective antibacterial strategies, and (v) technological advances in glycoscience that have helped to accelerate progress in research. These studies collectively offer valuable insights into new perspectives on antibacterial strategies that may effectively tackle the drug-resistant pathogens that are rapidly spreading globally.

Overview

Glycans are commonly found on the surface of host and bacterial cells. These glycans are highly diverse, yet commonalities between host and bacterial glycans exist [1]. Not surprisingly, glycointeractions are at the center of bacterial pathogenesis and host responses against invading pathogens [2] (Figure 1). Core concepts relevant to host–pathogen interactions are discussed in this review, with examples reported in the recent literature. We start the review by discussing concepts concerning bacterial glycans used for colonization and/or immune evasion. For instance, some bacterial pathogens express cell-surface glycans that are vital for successful colonization on/in host cells; some bacterial pathogens express molecular mimicry of host glycans on their cell surfaces, enabling them to disguise themselves from host immune surveillance; and some bacterial pathogens cover their cell surface with layers of glycans to hamper host recognition of common pathogen-associated molecular patterns (PAMPs) – such as lipoteichoic acid (LTA), lipopolysaccharides (LPS), and flagellin [1]. We continue by discussing concepts of bacterial utilization of host glycans for pathogenesis. Many bacterial pathogens and toxins are equipped with lectin-like features that recognize specific host glycans expressed on a set of host cells for their colonization and virulence [2]. Conversely, in the following section, we discuss concepts concerning host immune cells utilizing bacterial glycans as molecular patterns to trigger bactericidal immune responses [3] and utilizing host glycans for immune responses against pathogens whose outcomes can be altered by input from microbiota. As a result, glycointeractions between pathogens and the host can benefit bacteria, host, or both. The review continues by discussing recent examples of antibacterial strategies designed on the basis of the glycobiology of host–pathogen interactions, and finally, we deal with the technological advances in glycoscience that have helped to accelerate the progress in research.

Figure 1. Glycointeractions in bacterial pathogenesis.

(A) Bacterial surface glycans play a vital role in adherence and colonization on host cells, which in some cases results in bacterial invasion of host cells. Bacterial glycans are also crucial for biofilm formation. (B) Some bacterial pathogens express molecular mimicry of host glycans on their cell surface (bacteria with pink hexagons) that can disguise themselves from host immune surveillance. Some bacteria cover their cell surface with layers of glycans [capsular polysaccharide (CPS), light pink and/or pink hexagons] to hamper host recognition of common pathogen-associated molecular patterns (PAMPs). In comparison, the recognition of PAMPs (green hexagons) and downstream immune responses is depicted in the right half of the graphic. (C) Many bacterial pathogens and toxins are equipped with lectin-like features that recognize specific host glycans (depicted as a gray rod in the left panel or gray rod with codes for N-glycans in the right) expressed on a set of host cells for their colonization and/or virulence. (D) Conversely, host immune cells can utilize bacterial glycans as molecular patterns to trigger bactericidal immune responses (depicted in the left) and host glycans (blue), resulting in various immune responses and/or pathogen clearance. Multiple pathogens have developed evasion strategies. This review discusses the mechanisms involved.

Bacterial glycans in bacterial colonization and/or immune evasion

Most bacterial glycoconjugates are located on cell surfaces and membranes. The membranes of Gram-positive bacteria consist of the inner membrane (IM) and cell wall comprised of peptidoglycan (PG), capsular polysaccharide (CPS), wall teichoic acid (WTA), and LTA (Figure 2). The roles of these bacterial glycans in host cell colonization or immune evasion are considerably well characterized. For instance, the functions of CPS in colonization, invasion of host cells, and immune evasion are perhaps best demonstrated in group A streptococci (GAS), such as S. pyogenes [4]. GAS glycans contain a polyrhamnose core where N-acetylglucosamine (GlcNAc) side chains are added by glycosyltransferase GacI, contributing to colonization and immune evasion [5], while ~25% of GlcNAc of GAS glycans can be further modified by glycerol phosphate via GacH, contributing to bacterial survival and pathogenesis [6]. Similarly, O-acetylation and deacetylation of glycans in the cell wall of Streptococcus pneumoniae contribute to enhancing resistance to lysozyme-mediated killing [7]. Glycosylated adhesins (e.g., cnm) on the surface of many Streptococcus spp. are vital for the binding to host cells and successful colonization [8,9]. Another example is Staphylococcus aureus WTA. The glycosylation of S. aureus WTA by TarS inhibits the binding of alternative penicillin-binding protein (PBP2a), contributing to resistance to β-lactam antibiotics, while methicillin-resistant S. aureus (MRSA) is known to utilize TarP-modified WTA to evade host defenses [10].

Figure 2. Common bacterial glycoconjugates located on bacterial cell surfaces and membranes.

As depicted on the left, the membranes of Gram-positive bacteria consist of the inner membrane (IM) and cell wall comprised of peptidoglycan (PG, a polymer of GlcNAc and MurNAc), capsular polysaccharide (CPS), wall teichoic acid (WTA), and lipoteichoic acid (LTA). As depicted on the right, the membranes of Gram-negative bacteria consist of the PG layer sandwiched between the IM and outer membrane (OM) with the space between the two membranes referred to as the periplasm. The outer surfaces of Gram-negative bacteria are coated by glycans that are part of the lipopolysaccharide (LPS) O-antigen, lipo-oligosaccharide (LOS) core-sugars, and/or the CPS.

The membrane structure of Gram-negative bacteria consists of the PG layer sandwiched between the IM and outer membrane (OM) with the space between the two membranes referred to as the periplasm. The outer surface of Gram-negative bacteria is coated by glycans that are part of the LPS O-antigen, core-sugars, and/or the CPS (Figure 2). The PG is a highly ordered glycan repeat synthesized by MurE and MurF. In the context of host–pathogen interactions, the PG can activate the immune system, and thus the modification of PG to noncanonical variants prevents this activation [11,12]. Likewise, noncanonical variants of LPS, such as extended LPS O-antigens and the lack of O-antigens, are expressed by multiple bacteria for immune evasion and/or successful colonization. For instance, Salmonella enterica serovar Paratyphi A produces very long O-antigen chains that resemble the length and function of the Vi CPS of S. Typhi, both of which hamper immune surveillance [13]. S. Typhi and S. Paratyphi cause the life-threatening systemic diseases typhoid fever and paratyphoid fever, respectively. A very long O-antigen is also observed in Pseudomonas aeruginosa that is controlled by the Wzz system [14], contributing to immune evasion [15]. Conversely, multiple bacteria produce lipo-oligosaccharides (LOS) in place of LPS, including species of Campylobacter and Neisseria. LOS is analogous to LPS for Lipid A and core oligosaccharides but lacks O-antigens (Figure 2). Some of these bacteria, such as Acinetobacter baumannii, have further evolved to carry either LOS modified with phosphatidylethanolamine and galactosamine (GalN) or can tolerate the total loss of LOS/LPS, contributing to resistance to antimicrobial peptides and/or antibiotics targeting Lipid A, such as a ‘last-resort’ antibiotic colistin [16,17].

In A. baumannii, several O-oligosaccharyltransferases (O-OSTs) glycosylate other surface molecules [18], which are also associated with antibiotic resistance and/or immune evasion. These include general O-OSTs, like PglL, that is responsible for glycosylating serine and threonine residues of multiple proteins, PglC, known for attaching GalNAc to the lipid carrier [18], and substrate-specific O-OSTs such as TfpO known for glycosylating C-terminal serine residues of type IV pilin [19]. Lastly, multiple Gram-negative bacteria have evolved to express CPS for pathogen survival and/or immune evasion. Notable examples are the CPSs of S. Typhi [20], A. baumannii, and Neisseria meningitidis [21,22], protecting the pathogens from complement-mediated bactericidal activities and preventing the recognition of PAMPs on bacterial cells. These bacteria can modify the CPSs by O-acetylation that is known to enhance the stability of these capsular glycans, contributing to the prolonged protection from bactericidal activities and the inhibition of triggering effective immune responses against the pathogens.

Host glycans utilized by bacteria for pathogenesis

Physical barriers to the entry of pathogens, including the gastrointestinal (GI) tract, urogenital tract, and respiratory tract, are heavily glycosylated, which is vital for the host barrier function. Paradoxically, those host glycans benefit multiple pathogens for colonization, tropism, and/or a source of nutrients during infection (Figure 3). Relating to colonization, in the GI tract, S. Typhimurium Std fimbriae [23] and Bacillus subtilis YesU [24] use fucosylated glycans abundant in the gut to adhere and colonize. Vibrio cholerae adhesins, such as V. cholerae cytolysin (VCC) and RbmC, bind complex-type N-linked glycans expressed on intestinal epithelial cells to support colonization [25], while enterotoxigenic Escherichia coli (ETEC) adhesin EtpA binds blood group A glycans to adhere and deliver AB₅ toxins, heat-labile, and heat-stable toxins [26]. Helicobacter pylori infection alters glycan expression in the gastric tissues, which elevates an autophagy response in infected cells, supporting H. pylori replication [27].

Figure 3. Utilization of host glycans by bacterial pathogens.

Physical barriers to the entry of pathogens (the light blue shaded area in this graphic), including the gastrointestinal (GI) tract, urogenital tract, and respiratory tract, are heavily glycosylated, which is vital for the host barrier function. Paradoxically, those host glycans benefit multiple pathogens for colonization, tropism, and/or as a source of nutrients during infection. Bacterial names and virulence factors involved are indicated in the top portion of this graphic. Some notable examples are highlighted in this figure and are discussed under the heading ‘Host glycans utilized by bacteria for pathogenesis’. Abbreviations: S. Typhimurium, Salmonella enterica serovar Typhimurium; V. cholerae, Vibrio cholerae; UPEC, uropathogenic Escherichia coli; N. meningitidis, Neisseria meningitidis; VCC, Vibrio cholerae cytolysin; RbmC, the biofilm matrix protein.

Glycan-mediated tropisms of pathogens are exemplified in the urogenital tract [28]. Predominant glycans differently expressed in cells and tissues are associated with tropism of pathogens affecting these tissues. For example, uropathogenic E. coli (UPEC) adhesins type 1 and type P fimbriae bind mannose-containing glycans and galactose (Gal)α1–4Gal moieties of glycolipids predominantly expressed on the lower and upper urinary tract epithelium, respectively [29], corresponding to their primary target sites. Similarly, N. gonorrhoeae binds to mannosyl glycans expressed on cervical and urethral epithelial cells, while Chlamydia trachomatis recognizes N-glycosylated galectin-1 (Gal1) expressed on the genital tract, promoting bacterial attachment and entry in their primary infection sites [30,31]. Furthermore, the recognition of triantennary sialylated N-glycans containing poly-N-acetyllactosamine by N. meningitidis type IV pili Tfp is an example of glycan-mediated pathogen tropism to the central nervous system (CNS) [32].

In the respiratory tract, among several invasive groups of streptococci named groups A, B, C, and G, the GAS bind glycans terminating in sialic acid (N-acetylneuraminic acid, Neu5Ac) and GalNAc to colonize in the oral cavity and respiratory tract [33,34]. Coinfection of group B streptococci (GBS) with Influenza A virus (IAV), that utilizes α2,3-linked sialic acids/Neu5Acs on epithelial cells in the respiratory tract, results in more severe infectious outcomes [35]. Intriguingly, as the infection progresses, some streptococci produce bacterial glycoside hydrolases (GHs) to degrade host glycans to utilize as nutrients to benefit their growth during infection [36,37]. Similarly, P. aeruginosa, also infecting the respiratory tract, can exploit monosaccharides digested from host mucins as sources of nutrients for successful colonization [38].

Bacterial toxins

Given that almost all AB toxins use host glycans as cellular receptors [39], bacterial AB toxins are prime examples that help us to understand multiple important concepts of glycan-mediated host–pathogen interactions, including the roles in tropism to a specific set of target host cells, host adaptation, and toxin trafficking pathways (Figure 4). For instance, Salmonella typhoid toxin, an A₂B₅ toxin consisting of two enzymatic subunits CdtB and PltA linked to a homopentamer of glycan receptor binding subunits PltB₅ [40], exhibits tropism to specific target host cells. The in vivo tropism of typhoid toxin is achieved through high-affinity multivalent binding to glycan receptor moieties [41]. Consistently, multiantennary N-linked glycans, terminated in Neu5Acα2–6 or α2–3-linked to the underlying Gal and GlcNAc disaccharides, serve as cellular receptors for typhoid toxin, contributing to in vivo tropism to a range of host cells in the GI tract, systemic sites, and biliary tract that match with the dynamic infectious cycle of the bacteria [40–42]. 9-O-acetylation of the terminal sialic acid Neu5Ac further increases the binding affinity of typhoid toxin to glycan receptor moieties, also contributing to in vivo tropism of typhoid toxin [41,43]. Intriguingly, in certain conditions, typhoid toxin CdtB and PltA subunits are known to be able to assemble an alternate A₂B₅ toxin with alternate glycan receptor-binding subunits PltC₅ [44]. Glycans recognized by PltC subunits remain to be identified, but we predict that the glycan target matches with the infectious cycle of this pathogen.

Figure 4. Utilization of host glycans by bacterial AB toxins.

Almost all bacterial AB toxins use host glycans as cellular receptors. Toxin names and their glycan receptor moieties identified are indicated in this graphic. The light blue shaded area depicts target host cells. Some notable examples are highlighted in this figure and are discussed under the heading ‘Bacterial toxins’. Abbreviations: LDLR, low-density lipoprotein receptor; sGAGs, sulfated glycosaminoglycans.

Salmonella A₂B₅ toxins also help us to understand the role of glycan-mediated host–pathogen interaction in cell, tissue, and host adaptations. Typhoid toxin orthologs with some amino acid sequence variations on subunits, including their glycan-receptor binding subunits, are encoded in the genomes of some nontyphoidal Salmonella (NTS) serovars. One such toxin characterized at the molecular level is Javiana toxin expressed by S. Javiana [42]. The glycan receptor-binding subunit of Javiana toxin JaPltB contains three amino acid sequence variations that are sufficient for changing the tropism of this toxin to intestinal epithelial cells. This change is considered a niche-specific cell/tissue adaptation because the GI tract is the primary infection site of NTS serovars that, in general, do not cause systemic infection in immunocompetent humans [42].

About host adaptation, typhoid toxin recognizes sialosides terminated in Neu5Ac but not to ‘nonhuman-type’ N-glycolylneuraminic acid (Neu5Gc) that is identical to Neu5Ac in structure except for the hydroxyl group added at the C₁₁ position of Neu5Ac [45]. Neu5Gc is often called ‘nonhuman-type’ because, unlike various animal species, humans cannot synthesize Neu5Gc due to the lack of the functional cytidine monophospho-Neu5Ac hydroxylase (CMAH) that converts Neu5Ac to Neu5Gc. Thus, humans can make Neu5Ac only. The preference of typhoid toxin to ‘human-type’ Neu5Ac correlates to the human-specific feature of S. Typhi [45]. This aspect of typhoid toxin is often contrasted to subtilase cytotoxin (SubAB) from the enteric pathogen Shiga-toxigenic Escherichia coli (STEC) to highlight host adaptation of bacterial toxins. SubAB is an AB₅ toxin that binds glycans terminating in Neu5Gc and exhibits a strong binding preference toward Neu5Gc over similar glycans but terminating in Neu5Ac [46]. As noted previously, human intestinal epithelial cells cannot synthesize Neu5Gc, but a small amount of Neu5Gc can be incorporated in human cells from diets such as red meat [46], thus serving as target host cells for this toxin.

Typhoid toxin, and possibly SubAb, can control their trafficking mechanisms through toxin–glycan interactions. A characteristic two-stage trafficking mechanism is involved in typhoid toxin, consistent with the observation that typhoid toxin is produced only by S. Typhi which has invaded host cells. S. Typhi situated in the extracellular environment does not produce typhoid toxin. Environmental cues found in the Salmonella-containing vacuole (SCV) are vital for inducing the expression of typhoid toxin subunits [47,48]. The first stage of typhoid toxin trafficking is toxin export from infected host cells to the extracellular environment. The second stage of toxin trafficking is toxin-receptor binding and endocytosis [47,49]. PltB binding to its specific glycan receptor moieties expressed on the SCV membrane in infected host cells regulates the first stage toxin of export [49]. The interaction of the exported toxin with multiantennary N-linked glycans on target host cells is essential for the second stage of toxin endocytosis [40,41]. SubAB toxin is known to use complex-type N-glycans and core 1 O-glycans for entry and subsequent cellular toxicity [50]; it is intriguing to hypothesize that the glycan-binding preferences of SubB₅ to these glycans change during the retrograde toxin trafficking to facilitate the process.

Several concepts described previously are also shared among other bacterial AB toxins. Glycan receptor binding profiles of cholera toxin, an AB₅ toxin secreted by the enteric pathogen V. cholerae, have been revisited recently. In addition to GM1 gangliosides, fucosylated glycans linked to glycoproteins are proposed as a preferred glycan receptor for subsequent endocytosis processes in intestinal epithelial cells [51]. The binding pockets for fucosylated glycoprotein glycans are located on the lateral side of the B subunit pentamer, apart from the binding pockets for GM1 situated on the bottom side of the pentamer [52], suggesting multiple combinations that can result in multivalent high-affinity bindings between the toxin and glycans. The similar multivalent interaction mechanisms to intestinal epithelial cells through multiple binding pockets available in the B subunit pentamer have been shown using two Salmonella A₂B₅ toxins, typhoid toxin and Javiana toxin [41,42]. Furthermore, the roles of toxin–glycan interactions in toxin tropism are highlighted in the following recent discoveries. Bacterial pore-forming toxins, such as Tc toxin complexes and cholesterol-dependent-cytolysins (CDCs), have been characterized for their uses of glycans as cellular receptors and toxin tropism [53,54]. Clostridium difficile toxin A (TcdA) is a single-chain AB-type toxin contributing to the disruption of the colonic epithelium during C. difficile infection, causing fluid secretion, inflammation, and cell death [55]. TcdA contains two glycan receptor-binding domains which use Galα1–3Galβ1–4GlcNAc glycans, sulfated glycosaminoglycans (sGAGs), and low-density lipoprotein receptors (LDLRs) as its cellular receptor expressed on colonic epithelial cells [56,57].

Bacterial and host glycans for immune responses against pathogens

Bacterial glycans

Host immune cells recognize bacterial glycans and induce bactericidal immune reactions. C-type, Siglecs, and galectins are three major classes of lectins involved in this process, preferentially binding to mannose-, sialic acid-, and N-acetyllactosamine-rich glycoconjugates, respectively [58]. C-type lectins recognize many pathogens and induce proinflammatory responses, although not surprisingly, many pathogens have evasion mechanisms [58]. The mannose-binding lectin (MBL), a kind of C-type lectin, recognizes a variety of pathogen oligosaccharides, including those expressed on Borrelia burgdorferi, and induces bactericidal responses [59]. Another type of C-type lectin, the macrophage galactose-binding lectin (MGL), recognizes GalNAc on S. aureus and induces cytokine production, which may affect downstream adaptive immune responses and pathogen clearance [60]. Siglecs-mediated interactions to GBS, N. meningitidis, and N. gonorrhoeae, can induce host inflammatory responses [58]. Galectins bind to glycans of several important pathogens, including the opportunistic P. aeruginosa, and mediate the up-take by phagocytic cells [61].

Host glycans

Host glycans are crucial for innate and adaptive immune responses against bacteria. Regarding the roles in innate immune responses, lectins can control pathogens after the invasion of host cells by modulating cellular autophagy processes by recognizing specific types of host glycan. Intracellular galectins recognize host glycans expressed on vacuoles harboring intracellular pathogens, resulting in antibacterial autophagy [62]. For instance, the cytosolic galectin-3 recognizes damage-exposed host glycans, which induces autophagic responses against damaged lysosomes, offering protection against Mycobacterium tuberculosis [63].

An additional example highlighting the role of host glycan in innate immunity is the mucus that is produced by many pathogen entry sites in the body; it guards against pathogens while serving as a niche for microbiota and a protective layer to keep critical organs sterile [64]. For instance, mucin glycans can directly inhibit the expression of virulence genes of pathogens, including P. aeruginosa [65]. Furthermore, the microbiota can modulate host immune responses to infection by regulating cytokine expression [66,67], alleviating neurotoxicity [68], or influencing mucosal immunity [69]. One mechanism is by using host- and/or diet-derived glycans; gut microbiota, such as Bacteroides, uses enzymes encoded in the polysaccharide utilization loci (PULs) to utilize available glycans [70,71].

Host glycans also play a critical role in the development of immune cells and adaptive immune responses. For instance, the role of N-glycan branching in thymocyte development has been established [3]. N-glycan branching of the T cell receptor and CD4/CD8 coreceptors modulates the affinity for the positive selection of T cells. Their role in B cell development was also reportedly characterized in studies in which a lack of branched N-glycans resulted in the impaired maturation of B cells [72]. N-glycan branching played a critical role in B cell receptor and coreceptor expression modulation, essential in promoting B cell selection. Glycosaminoglycans (GAGs), such as heparin, chondroitin, and dermatan sulfate, bind to chemokine receptors, such as CXCL1 and CXCL5, to play a diverse role in the immune system [73]. Furthermore, critical functions regarding the N-glycosylation of immunoglobulins in antibody stability and placental IgG transfer efficiency have been demonstrated [74,75].

Antibacterial strategies

Host glycan mimicry (Table 1)

Table 1.

Examples of antibacterial strategies targeting glycan-mediated host–pathogen interactions, which are discussed in this review

Names	Targets	Challenges
Glycan mimicry
Aryl galactoside and GalNAc	FmlH adhesin of the F9 pilus used by UPEC [76,77]	Requires structure-based optimizations of high-affinity ligands to antagonize bacterial lectins [76].
Zanamivir sialic acid mimic	Periodontal bacterial sialidases [78]
STARFISH Gb3 mimic	Homopentameric receptor-binding B subunits of Shiga-like toxin I	Must overcome the low intrinsic affinity of carbohydrate-protein interactions through the use of an oligovalent ligand [79].
L-fucose analogs	Homopentameric receptor-binding B subunits of cholera toxin	Requires optimization of the analog candidates through the structure-activity relationship analysis of the CTB-fucose interaction [80].
Antibodies
Anti-CPS IgG MAbs	CPS of carbapenem-resistant K. pneumoniae	Only elicit passive immunity against the bacterial target [81].
Anti-KDO MAb	KDO added to a terminal N-acetyllactosamine structure of LOS on NTHi	NTHi expresses KDO only under certain conditions in vivo, and bactericidal effects were observed only in a limited number of strains [82].
Anti-PltB MAbs	Glycan receptor-binding pockets on the receptor-binding subunits [83,84]	A-subunit-mediated asymmetry-associated interference of antibody binding to the lateral side-located epitopes on the receptor-binding subunits [84].
Vaccines
Pneumococcal vaccines	CPSs of specific S. pneumoniae strains	Serotype-specific protection has prompted changes in serotype prevalence, requiring multivalent vaccines incorporating synthetic glycoconjugates to cover more serotypes [85].
Vi CPS vaccine	Vi polysaccharide of S. Typhi	Emergence of S. Typhi strains lacking Vi polysaccharide [87].
Hypervirulent K. pneumoniae bioconjugate bivalent vaccine candidate	CPSs of the K1 and K2 K. pneumoniae serotypes	Uses a relatively new, in vivo method of glycan-carrier protein conjugation [88,89].
C. difficile synthetic glycoconjugate vaccine candidates	C. difficile surface glycans PS-I, PS-II, and PS-III	Requires synthetic glycan conjugation to a carrier protein [90].
Gonococcal peptide mimotope vaccine candidate	Glycan epitope on the LOS of N. gonorrhoeae	Requires the development of a stable and homogeneous peptide mimotope derivative of the glycan epitope [91].

High-affinity synthetic glycans resembling host glycans recognized by bacteria or toxins have been demonstrated as effective antibacterial strategies. Various mannoside compounds effectively compete over endogenous host glycans recognized by the type 1 fimbriae of UPEC and N. gonorrhoeae [30,76,77]. A similar strategy has been developed against the adhesin FmlH in F9 fimbriae of UPEC; aryl galactoside and GalNAc, high-affinity binders to FmlH designed through structure-guided analyses, inhibit the binding of UPEC to human kidney tissue [76,77]. Porphyromonas gingivalis, Tannerella forsythia, and Treponema denticola, oral pathogens capable of causing severe periodontitis, secrete sialidases as a major virulence factor(s) that remove sialic acids from sialylated glycoproteins expressed on host cells. Consistently, Zanamivir, an FDA-approved influenza drug that inhibits the enzymatic activity of sialidases, effectively inhibited P. gingivalis sialidase SiaPG, resulting in the inhibition of biofilm formation, initial attachment, and invasion into oral epithelial cells [78]. Glycan mimicry has also been effective against bacterial toxins that use host glycans as cellular receptors. Shiga-like toxin 1 (STL-1) of E. coli, an AB₅ toxin, contains the characteristic B₅ homopentamer that recognizes its glycan receptor globotriaosylceramide (Gb3), whose interactions were effectively interrupted by a pentavalent, water-soluble glycan mimicry dubbed STARFISH [79]. The homopentameric receptor-binding B subunits (CTB₅) of cholera toxin utilize sialylated, fucosylated N-linked glycans expressed on intestinal epithelial cells; L-fucose analogs were able to inhibit the binding of CTB to host cells, indicating the potential of a non-natural fucose-containing polymer as an antitoxin strategy [80].

Vaccines and antibodies targeting bacterial glycans (Table 1)

Antibodies designed to target specific bacterial cell-surface glycans are shown to be effective against bacterial pathogens, including drug-resistant bacteria. For example, two anti-CPS IgG monoclonal antibodies (MAbs) against carbapenem-resistant Klebsiella pneumoniae ST258, capable of binding to the CPS of the clinically relevant clade 2, promoted opsonophagocytic killing and protected mice infected intratracheally with the bacteria, demonstrating passive immunotherapy potential [81]. An MAb recognizing the ketodeoxyoctanoate (KDO) within the LOS of the respiratory pathogen nontypeable Haemophilus influenzae (NTHi) was able to kill the bacteria, providing insights into the development of a multivalent vaccine(s) against NTHi using a glycoconjugate made of multiple KDOs or a KDO-N-acetyllactosamine conjugated to an immunogenic protein [82]. Also, MAbs recognizing the glycan receptor-binding pockets of Salmonella typhoid toxin neutralize the toxin with various efficacy depending on their epitope locations [83,84].

Glycan-based vaccines against the CPSs of S. pneumoniae and S. Typhi are perhaps best known. S. pneumoniae vaccines that target the CPS afford serotype-specific protection [85]. Unfortunately, yet not unexpectedly, there are 98 recognized polysaccharide serotypes among S. pneumoniae; the wide use of vaccines has prompted bacterial evolution to overcome vaccine-induced immune responses, suggesting necessary refinement of pneumococcal vacci-nation strategies [85]. Coformulation of licensed pneumococcal vaccines with synthetic glycoconjugates representing serotypes not covered by existing vaccines has shown potential in multivalent vaccine production [86]. Subunit vaccines against S. Typhi are based on the CPS Vi antigen, which affords serotype-specific protection. Similar to pneumococcal CPS vaccines, S. Typhi strains that lack Vi have emerged in endemic areas [87].

Similar to the previously-mentioned examples, glycan-mediated vaccine development against many other bacteria is in progress, including bioconjugate vaccines against hypervirulent K. pneumoniae serotypes [88] and polyvalent pneumococcal bioconjugate vaccines [89]. Additional promising examples include glycoconjugate vaccine candidates against C. difficile GI infection designed to resemble surface glycans PS-I, PS-II, and PS-III using four synthetic antigens ranging in size from disaccharides to hexasaccharide [90] and a peptide mimotope vaccine candidate of the glycan epitope on the LOS of N. gonorrhoeae against gonococcal genital tract infection [91].

The development of glycan-mediated vaccines would likely be accelerated as relevant technologies advance. Some of the aforementioned vaccines have been produced using glycoengineered E. coli cells (rather than conventional chemical methods), highlighting alternative vaccine production platforms. Also, newly characterized capsular polymerases that synthesize GBS capsule polymers could serve as another attractive tool to produce capsule polymers useful in glycoconjugate vaccine formulations [92]. Likewise, the recent development of a robust exponential glycan synthesis model, already shown to be capable of synthesizing 128-mer and creating a product relevant to the O-antigen of Bacteroides vulgatus, could serve as a tool for obtaining long, specific glycans [93], suggesting its broader application in multivalent glycan-based vaccine synthesis.

Technological advances in glycoscience

High-throughput approaches

Glycan microarrays are a high-throughput means of analyzing glycan-binding profiles of microbes, protein virulence factors, and others. A range of mammalian glycan microarrays has been instrumental in assigning carbohydrate specificities to several glycan-binding proteins ranging from microbial agents, such as bacterial toxins and viruses, to mammalian surface receptors for cell-to-cell interactions [94–96]. Advancements in bacterial glycan microarrays have been hampered by the bacterial glycome diversity that is even more diverse than mammalian counterparts [1]. Recent efforts have been directed towards constructing microbe-focused glycan arrays. For instance, combined complementary glycan synthesis strategies were used to create a pathogen-focused valuable platform for future investigations of serum antiglycan antibodies, innate immune receptors, and bacterial virulence factors [97]. Besides conventional glycan microarrays, electrospray ionization (ESI)-mass spectrometry (MS) has been demonstrated to be suitable for providing the interactions of glycan-binding proteins with their ligands in a high-throughput, quantitative manner. The use of an oligosaccharide library with common affinity tags against a series of glycan-binding proteins allowed for rapid identification and quantification of ligands [98]. Glycomics uses mass spectrometry to gain insight into the repertoire of oligosaccharides present, while computational analytic tools are essential for analyzing and quantifying glycans [99,100]. E-infrastructure has been established to overcome deficits in data and experimental transparency by enabling a workflow implementing the standardized submission of MS-based glycomics information into the public repository UniCarb-DR, MIRAGE (Minimum Requirement for A Glycomics Experiment) reporting guidelines, storage of unprocessed MS data in the GlycoPOST repository, and glycan structure registration using the GlyTouCan registry [99].

Mechanistic and translational studies

Lectins are glycan-binding proteins specifically recognizing one or more glycan structures. Many bacterial, viral, plant, and animal lectins are characterized, collectively serving as a valuable research tool. Lectins are particularly useful for mapping glycan structures expressed on cells and tissues and the tropism of particular pathogens, critical for understanding glycan-mediated host–pathogen interactions [1]. Glycosidases are one group of carbohydrate-active enzymes (CAEs) which hydrolyze the glycosidic bond between two or more glycans in a substrate-specific manner. For instance, sialic acids, often the terminal glycan of various glycoconjugates, are very diverse in structure, evident by more than 30 variants of sialic acids known thus far, which plays a critical role in the cell, tissue, or host tropism of many pathogens [101,102]. Various sialidases serve as an important research tool for investigating the tropism of pathogens and virulence factors [1]. PNGase F, also commonly used, is specific to N-linked glycoproteins and allows for research tailored for N-linked glycoproteins [103]. Many in vitro and in vivo research tools, such as glycoengineered knockout and knockin cells and mouse lines, have been helpful for understanding structure–function correlations. Moreover, computational methods, including a deep-learning tool, help us to understand glycan function in the context of host–pathogen interactions by using a large, curated glycan dataset in predicting glycan immunogenicity and the pathogenicity of bacterial strains [104]. Web resources such as GlycoSuiteDB and ProGlycProt V2.0 (for prokaryotic origin) have been designed to consolidate findings from the scientific literature on glycoprotein-derived glycan structures, their biological sources, the references in which the glycan was described, and the methods used to determine the glycan structure [105,106]. Furthermore, we predict that the expansion and streamlining of glycan synthesis methods would offer more readily accessible glycan products useful for in-depth structure–function studies, glycan arrays, vaccine development, and therapeutic development. For instance, streamlined one-pot procedures with automatic and programmable potential that integrate chemical, enzymatic, or chemoenzymatic synthesis methods have been established for producing complex glycans in a more effective manner [107,108].

Concluding remarks

Understanding the glycobiology of host–pathogen interactions is important and offers insights into antibacterial strategies. Successful interventions using glycan-based vaccines and glycan mimicry against several bacteria or bacterial toxins promise the fruitful outcomes of similar interventions against many other bacterial infections. Therefore, further collective efforts in this research area are expected to continually tackle clinical challenges posed by the global spread of multidrug-resistant pathogens (see Outstanding questions).

Outstanding questions.

Is there any way to improve the overall efficacies of glycan-mediated vaccines and/or glycan mimics?

Is there any way to slow down the emergence of bacterial pathogen variants that can evade vaccines targeting bacterial glycans and/or glycan mimics?

Is there any cost-effective way to implement a regular surveillance program to monitor the emergence of bacterial pathogen variants that can evade vaccines targeting bacterial glycans and/or glycan mimics?

Is there any cost-effective way to implement a basic research pipeline to rapidly determine newly emerged glycointeractions between bacterial pathogen variants and host?

Is there any cost-effective way to implement a translational pipeline to rapidly develop intervening strategies targeting newly emerged glycointeractions between bacterial pathogen variants and host?

Would the use of strategies targeting glycointeractions between pathogen and host in combination with other antibacterial strategies result in a better outcome in treating infection?

Highlights.

Glycans are common on the cell surface of host and bacterial cells, and not surprisingly, glycan-mediated molecular interactions play a critical role in bacterial pathogenesis and host responses against pathogens.

Bacterial glycans play a vital role in bacterial colonization, invasion, and/or immune evasion.

Many bacterial pathogens and their virulence factors exploit host glycans for pathogenesis and virulence.

Host glycans and lectins play a crucial role in inducing bactericidal immune responses, although many pathogens have developed evasion mechanisms.

Targeting glycan-mediated interactions between host and pathogen has emerged as an effective antibacterial strategy.

Technological advances in glycoscience have helped accelerate research progress in the field.