The role of glycosylation in clinical allergy and immunology

Abstract

While glycans are among the most abundant macromolecules on the cell with widespread functions, their role in immunity has historically been challenging to study. This is in part due to difficulties assimilating glycan analysis into routine approaches used to interrogate immune cell function. Despite this, recent developments have illuminated fundamental roles for glycans in host immunity. The growing field of glycoimmunology continues to leverage new tools and approaches to uncover the function of glycans and glycan-binding proteins in immunity. Here we utilize clinical vignettes to examine key roles of glycosylation in allergy, inborn errors of immunity, and autoimmunity. We will discuss the diverse functions of glycans as epitopes, modulators of antibody function, and as regulators of immune cell function. Finally, we will highlight immune modulatory therapies which harness the critical role of glycans in the immune system.

Introduction

Glycosylation is among the most common post-translational modifications in the body; it is estimated that over 50% of proteins are glycosylated.¹ Glycosylation refers to the addition of carbohydrate structures (glycans) to asparagine (N-glycans) or serine/threonine residues (O-glycans) of newly synthesized proteins, thereby generating glycoproteins (Figure 1). Glycolipids are formed through the addition of glycans to lipid molecules.² Collectively these glycoconjugates (glycoproteins and glycolipids) are secreted or embedded in the cell membrane as part of the glycocalyx, a dense layer of glycoproteins, glycolipids, and proteoglycans that cover cells. Glycosylation impacts protein half-life, intrinsic aspects of their activity, and interaction with binding partners.³ In addition, glycans can represent a significant component of glycoprotein weight and, as they extend from the protein surface, can be the most accessible feature to its surroundings. Consequently, glycans influence cell and protein interaction with the environment.^4–7 For example, factor H, which facilitates complement inactivation, binds to sialic acids on the cell surface thereby protecting host cells from alternative complement activation,^{8, 9} emphasizing the role of sialic acids in distinguishing self from non-self.¹⁰ Engaging with glycoconjugates are glycan-binding proteins (GBP), also called lectins, that are key to decoding the glycans.^{9, 10} As such, glycans are critical structural features of glycoconjugates that shape their overall activity.¹¹

Figure 1: Glycans can shape immune cell activity.

(A) Glycosylation of proteins occurs through the attachment of glycans to asparagine (N glycosylation) or serine/threonine (O glycosylation) residues. N-glycosylation occurs in the endoplasmic reticulum (ER) and Golgi, whereas O-glycosylation occurs in the Golgi. During N-glycosylation, a precursor is attached to the glycoprotein, which is then trimmed and modified by a series of glycosidases and glycosyltransferases in a sequential manner until it reaches the trans-Golgi. From the trans-Golgi, the glycoprotein is transported through secretory vesicles to its destination. In contrast to N glycosylation, O glycosylation is initiated in the Golgi by the stepwise addition of monosaccharides. (B) Changes in cell surface glycosylation can impact the sensitivity of cell interactions with a glycan-binding protein (GBP). A cell that expresses glycans terminating in α2–6 linked sialic acid can interact with the siglec, CD22. In contrast, removal of sialic acid exposes galactose residues that can serve as ligands for galectins. (C) GBP and glycan regulation of host immunity. Dendritic cells and macrophages express C-type lectins that engage microbes, which can facilitate microbe uptake and/or change the behavior of the cell. Galectins can both directly kill microbes and regulate immune cell activity. Selectins facilitate leukocyte recruitment.

Many immune regulators that have been used as CD markers are either glycoconjugates or GBPs. Although these have been used for years to identify or modify unique immune cell populations, including the use of plant-derived GBPs as T cell mitogens long before the T cell receptor was defined,¹² the impact of glycans on cell surface glycoconjugates and the role of GBPs in shaping immune function have been difficult to define.^{7, 13–16} This does not reflect a lack of importance of glycosylation in orchestrating immune function, but is a direct outcome of the non-linear and non-templated nature of glycan synthesis and overall structure. While these very features of glycans likely contribute to their central and dynamic role in many immunological processes, from antibody effector function to chemokine receptor sensitivity ligand engagement,^{11, 17} these same properties have made them challenging to study.^{2, 3, 7, 9, 18, 19} However, recent advances in glycan sequencing and synthesis technologies, in addition to novel methods to study GBP-glycan interactions, have accelerated the field of glycosciences in general and the application of glycobiology to allergy and immunology.^20–26

In this review, we seek to illustrate the critical and diverse role of glycans in the immune system. Given the diverse role glycans and GBPs play in immune function, we will not attempt to provide a comprehensive review on the subject. Instead, we will provide an overview of glycosylation, followed by vignettes highlighting key features of glycoimmunology in the context of medicine. We will focus on the role of glycans as (1) epitopes, (2) modulators of antibody functions, and (3) regulators of immune cell development, activation, and differentiation. We conclude with a discussion of potential therapeutic applications within glycoimmunology.

Glycan biosynthesis

Unlike the proteins they decorate, the composition of glycans is not directly encoded in the genome but is instead determined by a series of glycosyltransferases and glycosidases, in addition to available donor sugar nucleotides, that are present in a given cell.^{3, 7} This gives rise to a diverse set of glycans that can rapidly change in response to environmental cues by modulating the expression of glycan-modifying enzymes and the donor sugar nucleotides required for glycan elongation. These dynamic changes in turn affect protein function and alter a cell’s sensitivity to GBPs.^{3, 16, 27} N-glycan formation is initiated by the addition of a preformed oligosaccharide by the oligosaccharyltransferase (OST) complex that is then modified within the endoplasmic reticulum (ER) and Golgi apparatus (GA). O-glycans are formed stepwise in the GA, while glycolipids are likewise synthesized in a similar stepwise fashion. The final products are trafficked through secretory vesicles to the cell membrane or secreted into extracellular space (Figure 1). Monosaccharide donors are obtained through the metabolism of dietary sugars and from recycling glycans of endocytosed glycoproteins through the lysosome. Unlike secreted and membrane-bound proteins, glycosylation of intracellular proteins consists of a single monosaccharide attached to serine/threonine residues, O linked N-acetylglucosamine (GlcNAc).^28–31

Glycan-binding proteins in immune regulation

The first described GBP with a physiological role in the mammalian system was the asialoglycoprotein receptor. Expressed in the liver, it binds galactose and N-acetylgalactosamine (GalNAc) residues exposed due to loss of sialic acid on glycoproteins and clears proteins and platelets from circulation. The asialoglycoprotein receptor belongs to the calcium-dependent (C) – type lectins, one family of lectins with important functions in immunity.^{32, 33} Together with C-type lectins, sialic acid-binding immunoglobulin-like lectins (siglecs) and galectins are examples of GBPs that modulate both innate and adaptive immunity.⁹

C-type lectins can act as pattern-recognition receptors that recognize pathogen- and danger-associated molecular patterns, glycans expressed on the cell surface of the pathogen or exposed following cell damage, respectively (Figure 1). Examples include macrophage-inducible C-type lectin (Mincle), Dendritic cell-associated C-type lectin (Dectin)-1, Dectin-2, and DC-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN) which can engage extracellular bacteria and fungi. Upon recognition of their target glycan, C-type lectins facilitate phagocytosis and induction of an innate immune response. Additionally, some of these receptors, such as mannose-binding lectin, are secreted and bind extracellular pathogens and trigger complement. Pathogens can also exploit these receptors to enter the cell, for instance, DC-SIGN can be a critical vehicle for HIV infection.³⁴ C-type lectins can also regulate key immune cell functions, such as leukocyte trafficking, where selectins (E-, P-, and L-selectin) facilitate lymphocyte homing and neutrophil extravasation at sites of inflammation.^{34, 35}

Siglecs are membrane-embedded GBPs expressed by both innate and adaptive immune cells that interact with sialylated cellular targets. Most siglecs contain an immunoreceptor tyrosine-based inhibitory motif (ITIM) conferring suppressive activity (Figure 1). For example, CD22 (Siglec 2) recognizes sialic acid residues on host cells, inhibits B cells from responding to self-antigens by downregulating BCR signaling, and triggers apoptosis of the self-reactive B cells.^36–38 Siglecs can likewise regulate T cells, macrophages, eosinophils, neutrophils, and monocytes.^{39, 40}

Galectins are widely expressed in cells and tissues and have diverse functions. Extracellularly, galectins regulate immune cells by engaging polylactosamine (Poly-LacNAc) branches, a common modification that can be found on glycoconjugates.^41–44 Intracellularly they regulate immune cells through interaction with cytosolic proteins.⁴⁵ In addition to regulating host immune function, galectins can directly interact with microbes, which can protect a host through direct antimicrobial activity or can be exploited by pathogens to establish infection.^{19, 21, 26} Galectins negatively regulate T cell function by inhibiting activation, modulating cytokine expression, and inducing apoptosis in activated T cells.^{45, 46} Likewise, galectins can regulate neutrophils by inducing their turnover through apoptosis independent processes.⁴³ In immune cell development, galectins additionally can regulate T cell, B cell, dendritic cell, and macrophage maturation.¹⁴

Glycan epitopes

Anti-A and -B blood groups

The first anti-glycan antibodies discovered in the human population were directed against A and B blood group antigens on the surface of red blood cells. ABO blood groups were the first polymorphisms described in the human population and likely evolved because of infectious disease pressures that continue to be apparent as recently as COVID-19.^47–49 While other alloantibodies can certainly form against a variety of alloantigens following transfusion,^50–54 testing of ABO blood group antigens prior to transfusion and transplantation is the most common example of personalized medicine.^55–57 When these testing procedures fail, ABO incompatible blood transfusions can lead to rapid intravascular hemolysis, resulting in severe and often fatal transfusion reactions.^{53, 58} Although conflicting data exist,^{59, 60} the formation of anti-A and anti-B is thought to reflect exposure to glycans decorating gut microbiota.⁵⁵ However, anti-carbohydrate antibodies are certainly not limited to anti-A and anti-B. Anti-glycan antibodies can form against a wide variety of glycan-based blood group antigens and related structures. Unlike many RBC-induced alloantibodies, which tend to predominantly be immunoglobulin G (IgG),^{50–53, 61, 62} most anti-A and anti-B antibodies are IgM but can likewise be IgG with a small proportion of IgA antibodies.^50–52

Clinical allergy testing and glycan epitopes

While most anti-glycan antibodies are thought to be predominately IgM and form in the absence of CD4⁺ T cell help, glycan-specific antibodies have been discovered for all classes of antibodies, including IgE.⁶³ The discovery of glycan-specific IgE antibodies is intrinsically linked to the development and widespread use of radioallergosorbent allergy testing. In 1967, Wide et al. reported a new in-vitro method, the radioallergosorbent test, to detect allergen-specific antibodies of a new immunoglobulin class,IgE.⁶⁴ As this testing became widely available, multiple groups noted cross-reactivity between different types of food, grass pollen, birch, papain, and bromelain. In 1981, Aalberse et al. studied several individuals with cross-reactive serum, concluding that the IgE antibodies were likely binding to carbohydrate epitopes found in numerous foods, in particular plants, and insect venoms.⁶⁵ These cross-reactive epitopes (later termed cross-reactive carbohydrate determinants or CCDs) were subsequently confirmed to be distinct carbohydrate chains (glycans) found in insect venoms, parasites, and food.^66–68 Unlike the clinical relevance of anti-ABO (H) antibodies, numerous studies suggested that, despite their high prevalence, these anti-glycan IgE antibodies had little, if any, clinical relevance. Most patients with anti-plant glycan IgE antibodies do not report corresponding symptoms of food allergy.^{65, 66, 69} Patients with anti-plant glycan antibodies were evaluated by skin prick testing, in-vitro basophil histamine release assay, and double-blinded placebo-controlled oral challenges. In the subset of patients who underwent skin testing and/or oral challenge, there was no evidence of allergy with negative skin testing and no allergic symptoms with oral challenge. In an in vitro basophil histamine release assay, the majority of anti-glycan IgE antibodies from these patients’ serum induced histamine release but at concentrations 5–6 times the positive control used, major grass pollen allergen rPhl p 5.^{66, 70} Therefore, the initial focus on anti-glycan IgE antibodies was mitigating their potential for causing false positive allergy testing.^71–73

Alpha-gal and Cetuximab

In contrast to these earlier studies, Chung et al. reported that IgE anti-galactose α1,3 galactose (α-Gal) antibodies could trigger anaphylaxis.⁷⁴ This study stemmed from the observation of increased rate of severe hypersensitivity reactions to Cetuximab, a chimeric mouse-human IgG1 monoclonal antibody used in colorectal and squamous cell cancers, in North Carolina, Arkansas, Missouri, Virginia and Tennessee.^{75, 76} One study found a 22% rate of severe hypersensitivity reactions in Tennessee and North Carolina, in contrast to a rate of <1% in the Northeast.⁷⁷

Chung et al. hypothesized that the increased rate of sever hypersensitivity reactions was due to preexisting IgE anti-Cetuximab antibodies. Their study confirmed this hypothesis and identified the epitope as a glycan, α-Gal. First described by Karl Landsteiner as the “B-like antigen”, α-Gal is a terminal glycan modification expressed in all mammals, except old world monkeys and humans, and has structural similarities to the blood group B antigen.^{78, 79} Cetuximab has an α-Gal moiety on each of its fragment antigen-binding regions (Fab region). The structure and number of α-Gal molecules on cetuximab results in increased avidity, allowing for cross-linking of IgE and subsequent severe allergic symptoms.^{80, 81}

Given that the glycosylation of proteins is determined by the cell line expressing it, Chung et al. expressed Cetuximab in a cell line lacking α-1,3-galactosyltransferase. The patients’ anti-Cetuximab antibodies no longer bound to it, confirming that the epitope for Cetuximab is α-Gal. They also found that anti-cetuximab IgE antibodies bound to cat, dog, and beef proteins, all of which are non-primate mammalian proteins.⁷⁴ Based on these findings, Commins et al. hypothesized that IgE antibodies binding to α-Gal could cause food allergy to red meat. Indeed, they found specific IgE antibodies against α-Gal in patients who developed urticaria, angioedema, and/or anaphylaxis most commonly 3–6 hours after eating red meat. This IgE-mediated reaction was not typical of anaphylaxis due to the delayed timing of the reaction and the epitope, which was a glycan rather than a protein.^82–84 Commins et al. subsequently demonstrated that these patients were sensitized to α-Gal after multiple bites from the Lone Star tick, which has a high concentration of α-Gal in its saliva. Lone star ticks are most prevalent in the southeastern United States, likely explaining the increased prevalence of hypersensitivity to Cetuximab in that region.^{81, 85} Alpha-Gal syndrome thus highlights how the environment interplays with the body’s response to glycoconjugates (Figure 2). Anti-α-Gal antibodies not only can contribute to drug and red meat allergies, but recent studies have shown that they can also facilitate IgG anti-drug antibody development and may be responsible for the premature failure of porcine-derived heart valve replacement.^86,87

Figure 2: Glycans as epitopes, glycan regulation of antibody activity, and the impact of glycosylation on immune cell function.

(A) Upper panel: In α-galactose (Gal) allergy, an IgE anti-α-Gal antibody response is triggered by a tick bite. Subsequently, IgE-mediated activation of mast cells triggers an allergic response to α-Gal epitopes found in dietary meat. Middle panel: In Guillain-Barre Syndrome, anti-carbohydrate antibodies form following exposure to Campylobacter jejuni, which expresses surface glycans similar to the ganglioside GM1 found on neurons. This results in cross-reactivity with neuronal GM1, causing neuronal injury. Lower panel: In IgA nephropathy, increases in total IgA result in a corresponding increase in Tn-bearing IgA. The latter are bound by naturally occurring anti-Tn IgM antibodies which form immune complexes that can deposit in glomeruli, causing kidney injury. (B) IgG antibodies possess an asparagine (N) 297 glycan in the Fc domain that can regulate Fc gamma receptor engagement and complement activation. IgE can also be glycosylated on N265, N371, N383, and N394. (C) Changes in glycosylation on the proteins and lipids expressed on the cell surface can impact a variety of fundamental cellular activities. These include receptor residence time, receptor sensitivity to ligand-induced signaling, cellular activation states, and leukocyte trafficking.

Unanswered questions regarding glycan epitopes in allergy

The sharp contrast in immunogenicity of α-Gal compared to most IgE glycan epitopes raises several questions. First, are there other glycan epitopes that lead to clinically significant allergic symptoms? There is evidence for glycans acting as clinically significant epitopes, e.g. peanut allergy⁸⁸ and insect venom allergy.⁶⁶ Glycan modifications can also act as adjuvants, increasing the immunogenicity of diverse antigens including dust, ⁸⁹ peanut, ⁹⁰ microbes,⁹¹ and therapeutic recombinant factor VIII in hemophilia patients.⁸⁷ This is important in the context of using recombinant proteins for allergy testing. Depending on the cell line used to express a recombinant allergen, its glycosylation may differ from the natural allergen.⁹² Therefore, glycan epitopes can both be the cause of false positive and, more rarely, false negative allergy testing.^{66, 72, 93} Antibodies can also form against N-glycolylneuraminic acid (Neu5Gc), another glycan present in red meat. In contrast to α-Gal, although Neu5Gc is not synthesized by human cells, if introduced through oral intake it can become incorporated into human cell glycoconjugates. These anti-Neu5Gc antibodies may contribute to chronic inflammation and cancer.⁹⁴ Emerging data suggests a role for Neu5GC in allergy, but future studies are certainly needed.⁹⁵

Secondly, how are anti-glycan IgE antibodies formed? For most anti-glycan IgE’s specific for CCD’s the route of sensitization is thought to be inhalant allergens, or insect stings.^{66, 96, 97} The exact mechanism for anti-α-Gal IgE antibody formation remains a debated question. While additional studies are needed, early studies suggest that the microbiota influence anti-α-Gal IgG antibody formation, similar to their potential role in anti-ABO antibody development. ^{26, 63, 98, 99} It may be that the Lone Star tick bite triggers class switching of IgG to IgE anti-α-Gal antibodies although the mechanism for this remains poorly understood.^{80, 100} Alternatively, repeated Lone Star tick bites may lead to anti-alpha gal IgE antibodies, IgE sensitization, which subsequently can trigger allergic reactions to red meat or cetuximab.^{81, 85, 101–104} Third, why do some glycans induce severe allergic responses while the majority do not? There are several potential explanations for this, none of which have been fully tested. One compelling argument is that the ability to cross-link IgE determines the potential pathogenicity of a glycan epitope. This means that the number of glycan binding sites on a protein as well as their steric location determines if a glycan epitope can cause IgE cross-linking and subsequent histamine release.^{74, 105} In this model, if glycans bind with lower affinity to IgE or the density of the glycan epitope is suboptimal, cross-linking may not occur. However, even with lower affinity IgE interactions, optimal glycan epitope presentation may allow for sufficient avidity to drive histamine release.⁶⁶ Of note, although most CCDs are thought to have low affinity for IgE, multiple studies have suggested this is not always the case.^{106, 107} Clinically, anti-glycan epitopes remain a diagnostic challenge.^{66, 68, 71–73, 108}

Glycan epitopes and autoimmunity

IgA nephropathy

In addition to serving as targets in allergies, glycan modifications can also be the primary epitopes in a variety of autoimmune diseases. IgA nephropathy, first described in 1968, is a clinical diagnosis based on IgA and IgG deposits in the glomeruli.^{109, 110} Anti-glycan antibodies against the hypoglycosylated Tn antigen, a simple monosaccharide O-glycan modification within the hinge region of IgA, can result in immune complex formation that leads to antibody deposition in glomeruli, resulting in renal pathology.¹¹¹ The level of anti-Tn antibodies correlates with disease severity and predicts relapse post kidney transplant. Pathogenesis is thought to follow a two-step process. First, elevations in total IgA increase the relative abundance of hypoglycosylated IgA1, resulting in higher exposure to cryptic Tn antigen GalNAc residues. Second, anti-Tn autoantibodies, which are mainly IgM, cause immune complex formation that can deposit in glomeruli leading to kidney damage (Figure 2). These immune complexes can trigger complement activation, which may contribute to the underlying pathophysiology.

It is unclear how hypoglycosylated IgA antibodies are formed. IgA is galactosylated by core 1β1,3-galactosyltransferases (C1GalT1) with the help of a chaperone, Cosmc.¹¹² Early hypotheses suggested Tn formation on IgA was a product of reduced Cosmc expression. More recent data suggests that the IgA1 hinge region is less accessible to C1GalT1. Consistent with this, while total IgA is increased in patients with IgA nephropathy, the percentage of Tn+ IgA does not change.¹¹³ Environmental triggers which lead to acute increases in IgA1 production coupled with the presence of naturally occurring anti-Tn antibodies may ultimately converge to drive anti-Tn-IgA1 antibody immune complex formation and associated pathology.

Molecular mimicry and Guillain-Barre syndrome

One of the classic examples of the breakdown of identifying self vs. non-self toward carbohydrate antigens is Guillain-Barre syndrome (GBS), the most common cause of pediatric acute flaccid paralysis. Like the formation of anti-A and anti-B blood group antibodies, microbial molecular mimicry is thought to be the root of the pathogenesis of GBS. The majority of patients with GBS have an antecedent microbial infection, most commonly Campylobacter jejuni. It is thought that lipo-oligosaccharides and lipopolysaccharides found on the cell membrane of C. jejuni mimic the structure of GM1 gangliosides, a type of glycolipid found on the cell surface of neural axons. Anti-GM1 glycolipid antibodies that form in response to C. jejuni can engage neurons, leading to neuronal damage and severe life-threatening peripheral neuropathy (Figure 2)¹¹⁴. Intravenous immunoglobulin (IVIG), as well as plasmapheresis, are used to treat GBS.¹¹⁵

Glycans modify protein (immunoglobulin) function

IVIG

Glycans not only function as epitopes, but they also play a critical role in modulating the function of immunoglobulins. Although all immunoglobulins are glycosylated, this has been best studied for IgG. IgG antibodies are N-glycosylated on their Fc portion with a biantennary glycan modifying each of the two heavy chains at asparagine 297 (ASN297) (Figure 2), although Fab fragments can be variably glycosylated as noted above for Cetuximab. The immune modulatory role of immunoglobulin glycosylation was first identified for IVIG.¹¹⁶ First given to patients with deficiencies in humoral immunity, including X-linked agammaglobulinemia and common variable immunodeficiency, IVIG was originally purified as a therapeutic strategy to restore antibody levels.¹¹⁷ Two children with congenital agammaglobulinemia and severe thrombocytopenia treated with IVIG had resolution of thrombocytopenia. This led to the hypothesis that IVIG alone may enhance platelet counts in patients with immune thrombocytopenia (ITP). Treatment of 13 patients with ITP with IVIG lead to significant increases in platelet counts,¹¹⁸ providing some of the earliest data suggesting that IVIG could be used to treat autoimmunity. However, the features of IVIG responsible for favorably manipulating autoimmunity were unknown.

Using a murine model of ITP to investigate the anti-inflammatory mechanism of IVIG, the Fc portion of IgG was found to be sufficient for its anti-inflammatory activity. Additional studies demonstrated that Fc region sialyation, which is a terminal glycan modification, is a key mediator of its anti-inflammatory properties.^{116, 119} Although the anti-inflammatory mechanisms of IVIG remain an active debate, in animal models of arthritis and ITP, hyper-sialylated IVIG had enhanced anti-inflammatory activity.¹²⁰ In addition, using multiple autoimmune models, Pagan et al. converted endogenous IgG to anti-inflammatory IgG through sialylation.¹²¹ It is now well-established that N-glycan modification of the Fc portion of immunoglobulins impacts the conformation, stability, binding affinity, and effector functions of IgG antibodies.^{122, 123} However, the impact of distinct glycoforms on individual IgG isotypes remains incompletely defined. Recent findings also demonstrate that Fc receptor glycosylation itself not only can impact IgG engagement, but that the glycosylation of recombinant forms of Fc receptors commonly used to study IgG-Fc receptor interactions may not always faithfully recapitulate native glycosylation states.¹²⁴ As a result, additional studies are needed to fully define the complex regulation of IgG and Fc receptor glycosylation on the outcomes of IgG interactions.

IgE glycosylation

Although IgE has the most glycosylation sites of the immunoglobulins (Figure 2), less is known about its glycosylation. To investigate the clinical relevance of IgE glycosylation, Shade et al. studied the glycosylation content of IgE molecules in a cohort of individuals with and without atopy. They found that those with peanut allergy had increased sialylation of IgE, while those without atopic disease had increased bisecting GlcNAc and terminal galactose, exposed when sialic acid is absent. In addition, galactose and sialic acid content of total IgE were predictive for allergic disease. Using passive immunization in mice, removal of sialic acid from IgE prior to sensitization weakened the allergic response. Removal of the sialic acid did not impact binding to the allergen, but instead negatively modulated signaling through the FcεRI receptor. Ultimately, using an enzyme that cleaves sialic acid (sialidase), removal of sialic acid from IgE-bearing cells in-vivo attenuated anaphylaxis, suggesting a potential therapeutic strategy of glycoengineering immunoglobulins as a treatment for atopic disease.¹²⁵

Congenital disorders of glycosylation (CDG)

Glycans can also regulate fundamental aspects of cellular function. However, as changes in cellular glycosylation can impact many different cellular processes, the consequences of glycan changes are highly pleiotropic and therefore much more complex than those observed for an individual class of proteins, such as immunoglobulins (Figure 2).^{92, 126} CDGs are a rare group of monogenic disorders caused by pathogenic variants in pathways important for glycan biosynthesis. In 2020, Pascoal et al reviewed immune involvement in CDGs. Of the 133 CDGs described at that time, 23 had an immune phenotype including immunodeficiency, autoinflammation, or autoimmunity. The immunodeficient phenotype ranges from mild, susceptibility to specific pathogens, to severe, severe combined immunodeficiency (SCID). There is a wide spectrum of autoimmune and autoinflammatory phenotypes including inflammatory bowel disease (IBD), celiac disease, recurrent fever, and aphthous ulcer.¹²⁷

Leukocyte Adhesion Deficiency (LAD) type II

LAD type II is a well-described CDG with a severe and distinct immune phenotype. Two unrelated boys from consanguineous families presented with short stature, intellectual disability, recurrent bacterial infections without pus, distinct facial appearance, elevated neutrophil count, and the Bombay (hh) blood type. Their recurrent bacterial infections without pus formation and elevated neutrophil count were reminiscent of a recently described disorder termed leukocyte adhesion deficiency (LAD). In LAD, the integrin CD18 is not expressed leading to defective neutrophil adhesion. Neutrophil studies of the two boys demonstrated intact CD18. The researchers hypothesized it might be due to a defect in glycosylation because of the presence of the Bombay blood type. The Bombay blood type is characterized by a lack of the H antigen due to a defect in fucosylation critical for ABO blood group formation. Ultimately, a lack of sialyl lewis-X, a fucosylated glycan and ligand for selectins, was found on the boys’ neutrophils leading to a neutrophil trafficking defect. ^{128, 129}

As discussed before, leukocyte extravasation is mediated by selectins. Selectins are membrane-bound proteins; L-selectin resides on leukocytes, whereas E-selectin and P-selectin are expressed on activated vascular endothelial cells. Selectins mediate rolling of leukocytes along vascular endothelial, which is followed by integrin-mediated firm adhesion and subsequent extravasation.^{35, 130} LAD type II illustrates that sialyl lewis-X is required for neutrophil adhesion to endothelial cells through binding to selectins. Deficiency of sialyl lewis-X therefore leads to a severe defect in neutrophil trafficking.

X-linked immunodeficiency with magnesium defect, EBV infection, and Neoplasia (XMEN)

In 2011, XMEN, a new inborn error of immunity, was described. XMEN is caused by a hemizygous loss-of-function mutation in magnesium transporter 1 (MAGT1). Patients can present with combined immunodeficiency marked by persistent EBV infection, EBV-associated B cell malignancies, and lymphoproliferation. Immune phenotyping demonstrated an inverted CD4/CD8 ratio, elevated B cell counts, and decreased surface expression of NKG2D (natural killer group 2) on NK cells and CD8 T cells.¹³¹ Initially, these findings were attributed to a defect in magnesium transport. Subsequently, a larger cohort of patients was described with a broader range of phenotypes including lymphadenopathy, cytopenias, liver disease, cavum septum pellucidum, and increased CD3⁺CD4⁻CD8⁻TCRαβ⁺ (double negative) T cells.¹³² Investigators hypothesized that a magnesium defect was not sufficient to fully explain this complex phenotype. As MAGT1 is also known to be a subunit of the OST complex that initiates N-glycosylation, the complex phenotype might be due to abnormal N-glycosylation. Indeed, loss of MAGT1 leads to hypoglycosylation of multiple immune proteins. Most notably, decreased glycan expression leads to reduced cell surface retention of NKG2D, which results in decreased cytotoxic effector function of NK and CD8⁺ T cells. ^133–135

Phosphoglucomutase 3 (PGM3) deficiency

Another example of a CDG with a dominant immune phenotype is PGM3 deficiency. In 2014, Sassi et al. reported 9 individuals from 4 consanguineous families with Hyper-IgE syndrome (HIES) and biallelic hypomorphic variants in PGM3. All individuals in the study had markedly increased serum IgE levels, recurrent skin abscesses, and pneumonia with no variants identified in genes associated with HIES. Functional testing demonstrated defective catalytic activity of PGM3 with reduced production of complex branched N-glycans.¹³⁶ Stray-Pedersen et al. subsequently reported three unrelated patients with biallelic variants in PGM3 deficiency with T⁻B⁻NK⁺ SCID, congenital neutropenia with progression to bone marrow failure, two of the three patients having skeletal dysplasia. Only one patient had elevated serum IgE levels. Two of the patients underwent curative hematopoietic stem cell transplantation (HSCT). The third most severely affected patient died as an infant from severe infection. Functional testing again demonstrated reduced enzymatic activity of PGM3, with one variant completely inactivating the enzyme.¹³⁷ To date, four patients with PGM3 deficiency have received HSCT, three of which with good outcomes. The fourth had severe neurological impairment and multiple post-HSCT complications, dying from infection 8 months post-HSCT. There is some evidence that residual PGM3 activity correlates with severity of presentation but why some patients develop SCID while others present with HIES remains incompletely understood. ^136–138

Diminished levels of PGM3 disrupt the hexosamine pathway of glycolysis, which is important for generating a substrate, uridine diphosphate (UDP)–GlcNAc, needed for protein glycosylation, which leads reduced N- and O-glycans on immune proteins.¹³⁹ Patients with PGM3 deficiency have reduced numbers of specific N- and O-glycans that are dependent on UDP-GlcNAc. Specifically, neutrophils and B cells in these patients have reduced branched N-glycans, glycan moieties that are important in immune regulation.¹³⁶ Such changes likely influence the surface resident time of key glycoprotein receptors and sensitivity to GBPs, culminating in immune dysregulation.

Polygenic/complex trait diseases and N-glycan branching

Ulcerative Colitis (UC)

Defective N-glycan branching has also been implicated in polygenic diseases such as ulcerative colitis (UC) and multiple sclerosis (MS). T cells from patients with active UC carry less branched N-glycans compared to those with well-controlled disease. Supplementing patient T cells in vitro with GlcNAc restored N-glycan branching and inhibited activation and proliferation of T cells.^{140, 141} In patients with severe UC, the level of branched N-glycans was predictive of response to standard therapy, defined as 5-aminosalicylate, corticosteroids, and immunosuppressants. When looking at the disease course on a 5-year timespan 80% of patients with high levels of branched N-glycans had a good response to standard therapy over the course of 5 years. In contrast, of patients with low levels of branched N-glycans, 88% of patients needed to divert from standard therapy and receive anti-TNF therapy. In a separate study, T cells isolated from colon biopsies from patients with UC had lower expression of glycosyltransferase Mgat5 and fewer N-glycans compared to healthy controls.¹⁴² Finally, single nucleotide polymorphisms (SNPs) in Mgat5 were associated with reduced expression of Mgat5 in T cells of UC patients. These SNPs correlated with disease severity and the need for biologics.¹⁴³ Although these research studies are promising, the use of branched N-glycans as biomarkers has never been tested in a clinical trial.¹⁴⁴

Multiple Sclerosis (MS)

In MS patients, SNPs in Mgat5 have also been associated with MS disease severity; SNPs that correlated with reduced branched N-glycans were enriched in patients with the most severe disease.¹⁴⁵ PL/J mice develop spontaneous autoimmunity that is similar to chronic MS. Grigorian et al. showed that in these mice development of disease was suppressed following adoptive transfer of T cells fed in vitro with GlcNAc, which raised levels of UDP-GlcNAc and increased N-glycan branching by Mgat5.¹⁴⁶ Oral treatment with GlcNAc increased branched N-glycans and was able to suppress EAE progression in mice.¹⁴⁷ In the same mouse model of MS, N-glycan branching was eliminated through KO of glycosyltransferases Mgat1 or Mgat2 in B cells, eliminating or truncating branched N-glycans respectively. These mice had increased demyelination driven by Th1 and Th17 cells that were activated by autoreactive B cells.¹⁴⁸ Deficient N-glycan branching is likely an important risk factor for these polygenic diseases.

Glycosylation in cancer

Abnormal glycosylation is a hallmark of cancer.¹⁴⁴ Changes in glycosylation can result in the expression of non-self-glycans, termed tumor-associated carbohydrate antigens (TACAs). TACAs can intrinsically enhance neoplastic transformation.¹⁴⁹ These modifications can also engage immune modulatory receptors on immune cells, possibly impeding anticancer immune responses and create immune suppressive environments.¹⁵⁰ Localized changes in the expression of GBPs, particularly galectins, can likewise serve to modulate the tumor microenvironment.¹⁵¹ The expression of these neoantigens is often restricted to the tumor cells, and may therefore also be specifically targeted without harming healthy cells.^{150, 152} In addition to triggering direct cytotoxicity, anti-TACA therapies may improve anti-cancer immune responses by alleviating immune suppression.^{150, 152}

Therapeutic applications of glycobiology

Dinutuximab

The first FDA approved immunotherapy targeting a glycan epitope, Dinutuximab, was also the first immunotherapy to be approved for pediatric cancer. It is a chimeric monoclonal antibody against glycolipid ganglioside G2 (GD2). GD2 expression is mainly restricted to cancer cells with the exception of the central nervous system (CNS). GD2 is expressed on the surface of neuroectodermal tumors, such as neuroblastomas, gliomas, melanomas, sarcomas, as well as triple-negative breast cancer cells. Upon binding GD2 on neuroblastoma cells, Dinutuximab induces potent antibody-dependent cell-mediated cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC).^{153, 154} The relatively restricted expression of GD2 in healthy cells allows anti-GD2 monoclonals to more effectively target cancer cells, although there are significant side effects due to binding of the antibody to normal CNS cells.¹⁵⁵ The underlying features of anti-GD2 therapy that facilitate effective treatment outcomes remains a matter of active study.^155–157

Sweet CAR T cells

In 2016, Posey et al. reported a glycan-binding chimeric antigen receptor (CAR) T cell targeting a widely expressed tumor-specific epitope, GalNAxa1-O-Ser/Thr (Tn) - Mucin1 (Tn-MUC1). In vitro these CAR T cells had no reactivity against healthy human cells, but effectively induced tumor cell lysis in vitro and survival in murine models of T cell leukemia and disseminated pancreatic cancer. These CAR T cells can target multiple types of cancer which express Tn-MUC1. This method is being harnessed in phase 1 clinical trials for several solid tumors.^{158, 159} The unique expression of the glycan Tn antigen in cancer cells is particularly promising, given that it is highly expressed in the majority of carcinomas, and is rarely expressed in normal tissue.^{160, 161} It is a good potential biomarker given that it correlates with poor prognosis in a number of cancer types.¹⁶¹ While therapeutics targeting Tn have been hindered by a historical lack of high affinity specific antibodies against Tn, newly developed anti-Tn antibodies are promising.^{162, 163}

Glycoengineering of antibodies: Benralizumab

Glycans, in addition to serving as targets for therapeutics, can also be modified to optimize the function of the therapeutics themselves. As previously discussed, N-glycan modifications of the Fc portion of IgGs are critical for their confirmation, binding affinity, and effector function.¹²² This has led to glycoengineering of monoclonal antibodies by changing the Fc N-glycan to optimize their function. The most successful example of this to date has been the removal of core fucose on a monoclonal antibody (afucosylation) that increased ADCC activity.¹⁶⁴ In most cases, the N-glycan modification impacts conformation stability leading to differential affinity and binding to Fc receptors. In these cases, the N-glycan does not directly interact with the Fc receptor. In contrast, for both naturally occurring and glycoengineered antibodies, afucosylated antibodies bind with higher affinity to FcγRIIIa and FcγRIIIb through direct carbohydrate-carbohydrate interaction.^{122, 165}

Reduced or afucosylated anti-CD20 monoclonal antibodies are one of the best examples of deliberate monoclonal glycoengineering improving therapeutic potential.¹⁶⁶ These glycoengineered therapeutics show increased efficacy with improved B cell depletion and tumor growth inhibition in B cell malignancies.¹⁶⁶ Benralizumab is an afucosylated humanized anti-IL-5a antibody that triggers ADCC-mediated lysis of eosinophils and blocks IL-5R signaling. Clinically, it reduces asthma exacerbations in patients with severe eosinophilic asthma and is currently being evaluated in clinical trials against eosinophilic chronic rhinosinusitis, chronic obstructive pulmonary disease, and chronic idiopathic urticaria. ¹⁶⁷

Modulating cell behavior with glycan-binding proteins (Siglecs)

Changes in the glycans expressed on the cell surface can alter the sensitivity to glycan-binding proteins such as galectins and siglecs. In turn, these GBPs can modulate immune cell signaling. For example, therapeutics have been developed to recruit siglecs to immune-activating receptors and inhibit downstream cell signaling. Islam et al. used anti-IgD antibodies linked to a high-affinity siglec ligand (anti-IgD-CD22L) to inhibit B cell signaling in-vitro. In their in-vitro model, calcium flux was used as a measure of B cell activation. They showed inhibition of calcium flux when stimulating murine B cells with anti-IgD-CD22L. By repeating the experiment using B cells from a CD22 knockout mouse they demonstrated the specificity of this inhibition, with anti-IgD-CD22L now inducing robust calcium flux.¹⁶⁸ To further leverage siglec biology to treat allergic disease, the ability of anti-human anti-IgE antibodies bound to a high-affinity siglec ligand (anti-IgE-CD33L) was tested in an in vivo anaphylaxis model. In both the passive cutaneous anaphylaxis model and a passive systemic anaphylaxis model in “humanized” mice, anti-IgE-CD33L inhibited mast cell degranulation and prevented degranulation when mice were re-challenged with anti-IgE antibodies. Numerous studies have also demonstrated therapeutic potential for galectins in a broad range of diseases, including infection, autoimmunity, and even cancer. We anticipate that these strategies can be used to modulate cell signaling in many other disorders of allergy and immunology.^{168, 169.}

Targeting protein-glycan interactions using monoclonal antibodies has been proven effective for other diseases as well. Crizanlizumab, a P-selectin inhibitor, was one of the first FDA-approved drugs for the treatment of vaso-occlusive crisis in patients with sickle cell disease. Vaso-occlusive crises occur when sickled red cells adhere to the endothelium and block blood flow in the vessels of patients with sickle cell disease. This in turn triggers inflammation, mainly driven by neutrophils. Blocking P-selectin interaction with neutrophils using Crizanlizumab successfully reduced the rate of pain episodes in patients with sickle cell disease.¹⁷⁰

Conclusion

The examples outlined above are just a few examples of glycans’ diverse roles in the immune system. As noted earlier, many questions remain. The overall mechanisms responsible for the formation of anti-glycan antibodies, including exposures and immune regulators that govern class switching to IgE, remain relatively unexplored. Similarly, while the impact of glycosylation on IgG function has been extensively examined, the factors that regulate alterations in immunoglobulin glycosylation, in addition to the role of glycans in regulating IgE function is only beginning to emerge. Given the ability of glycosylation to impact numerous cellular functions, including cellular engagement by entire families of GBPs, immune regulation by cell surface glycosylation remains a formidable, yet exciting frontier. With new technologies becoming readily available, the application of glycosciences to understanding fundamental features of immune function holds significant promise in providing new therapies and understanding fundamental features of allergy and immunology.

Abbreviations:

ADCC

Antibody-dependent cell-mediated cytotoxicity

α-Gal

alpha-galactose

C1GalT1

core 1β1,3-galactosyltransferases

C3

Complement factor 3

C-type lectin

Calcium-dependent type lectin

CAR T

Chimeric Antigen Receptor T cell

CCD

cross-reactive carbohydrate determinants

CD

Cluster of Differentiation

CD3⁺CD4⁻CD8⁻TCRαβ⁺

Double Negative

CDC

Complement Dependent Cytotoxicity

CDG

Congenital Disorders of Glycosylation

CNS

Central Nervous System

DC-SIGN

DC-specific intercellular adhesion molecule-3-grabbing non integrin

Dectin

Dendritic cell associated C-type lectin

EAE

Experimental Autoimmune Encephalomyelitis

EBV

Epstein-Barr Virus

ER

Endoplasmic Reticulum

Fab region

Fragment antigen-binding region

GA

Golgi Apparatus

GalNAc

N-Acetylgalactosamine

GBS

Guillain-Barre Syndrome

GBP

Glycan-Binding Protein

GD2

Glycolipid ganglioside 2

GlcNAc

N-Acetylglucosamine

GM1

Monosialotetrahexosylganglioside

HIES

Hyper-IgE Syndrome

HSCT

Hematopoietic Stem Cell Transplantation

HSP

Henoch-Schönlein Purpura

IBD

Inflammatory Bowel Disease

Ig

Immunoglobulin

ITIM

Immunoreceptor Tyrosine-based Inhibitory Motif

ITP

Immune Thrombocytopenia

IVIG

Intravenous Immunoglobulin

LAD

Leukocyte Adhesion Deficiency

LPS

Lipopolysaccharide

MAC

Membrane Attack Complex

MAGT1

Magnesium Transporter 1

MBL

Mannose-Binding Lectin

Mgat

Mannosyl-glycoprotein 2-beta-N-acetylglucosaminyltransferase

Mincle

Macrophage-inducible C-type lectin

MS

Multiple Sclerosis

Neu5Ac

N-glycolylneuraminic acid

OST Complex

Oligosaccharyltransferase complex

Poly-LacNAc

Polylactosamine

PGM3

Phosphoglucomutase 3

SCID

Severe Combined Immunodeficiency

Siglec

Sialic acid-binding immunoglobulin-like lectins

SNP

Single nucleotide polymorphisms

TACA

Tumor-associated carbohydrate antigens

UC

Ulcerative Colitis

UDP

Uridine diphosphate

XMEN

X-linked immunodeficiency with magnesium defect, EBV infection, and Neoplasia