Keywords

1 Introduction

Since the first reports on glycans attached to immunoglobulin G (IgG) in the 1970s (Ciccimarra et al. 1976; Williams et al. 1973; Koide et al. 1977; Hymes et al. 1979) and the seminal papers by Parekh and al. on the association of a changed IgG glycome composition with a diseased status and aging, (Parekh et al. 1985, 1988) IgG glycans are today universally recognized as modulators of IgG activity (Yamaguchi and Barb 2020). The importance of IgG glycome composition is implied in various physiological and pathological states. IgG glycans are discussed as potential contributors to disease development and progression, as well as a diagnostic, prognostic, and follow-up biomarker. This chapter is a brief update and extension of our comprehensive review on IgG glycosylation in aging and diseases published 3 years ago (Gudelj et al. 2018a), with a focus on the potential functionality of the skewed IgG glycosylation pattern. The table presents the updated list of publications that examined IgG glycosylation in various diseased states.

2 IgG Glycans are an Integral Structural and Functional Part of the Molecule

IgG glycans represent about 15% of the molecule’s weight (Arnold et al. 2007). Each IgG molecule contains an N-glycan covalently attached to the conserved asparagine (Asn) at position 297 within the Fc region on each of the two heavy chains (Shade and Anthony 2013). In addition, 15–20% IgG molecules contain an N-glycan within the Fab region, attached to the asparagine within an N-glycosylation sequon formed by somatic hypermutation during affinity maturation (Dunn-Walters et al. 2000; van de Bovenkamp et al. 2016).

Fc N-glycans are placed in the cavity between the CH2 domains of the two opposing heavy chains (Pincetic et al. 2014; Deisenhofer et al. 1976) and are important for the molecule’s structural integrity, stability, and serum-half life (Boune et al. 2020; Cymer et al. 2018). They are also involved in the modulation of IgG effector functions, by affecting the molecule’s affinity toward its ligands and receptors: type I and type II Fc receptors, C1q complement component, mannan-binding lectin, etc. (Pincetic et al. 2014; Peschke et al. 2017; Malhotra et al. 1995; Dekkers et al. 2017). Although markedly less explored than Fc glycans, Fab glycans are also reported to affect IgG’s biological properties and effector functions, such as half-life, stability, solubility, and antigen-binding (van de Bovenkamp et al. 2016, 2018a; Wu et al. 2010; Wright et al. 1991; Higel et al. 2016; Liu 2015, 2018).

3 IgG Glycans Affect IgG Functions

The composition of both Fab and Fc glycans has been confirmed to influence IgG functionality and activity. Since this has been described in detail in Chap. 12, the main findings are only briefly summarized here as a reminder for the reader.

3.1 Fc Glycans

Due to the positioning of the Fc N-glycan at the Asn-297, structural differences of the N-glycans attached to the Fc region influence the affinity to the IgG ligands and receptors that interact with IgG at the CH2 domain and the CH2-CH3 domain interface (Dekkers et al. 2017; Reusch and Tejada 2015; Li et al. 2017; Wada et al. 2019; Vidarsson et al. 2014).

Core-Fucosylation

Contrary to most other plasma proteins, over 90% of all Fc glycans are core-fucosylated (fucosylated glycans, F) (van de Bovenkamp et al. 2016; Štambuk et al. 2020; Baković et al. 2013; Clerc et al. 2016). The lack of core fucose significantly increases the IgG’s affinity for the Fcγ receptor III (FcγRIII), both A and B, enhancing the FcγRIII-mediated effector functions, particularly the antibody-dependent cell-mediated cytotoxicity (ADCC) (Dekkers et al. 2017; Shields et al. 2002; Shinkawa et al. 2003). This prominent effect of alternative Fc glycosylation on the IgG function found its application in the industrial production of therapeutic monoclonal antibodies (Garber 2018).

Bisection

Up to 10% of all IgG Fc glycans are bisected, i.e., contain a bisecting N-acetylglucosamine (GlcNAc) (bisected glycans, B) (van de Bovenkamp et al. 2016). Since the presence of GlcNAc and core fucose, to a degree, preclude each other during glycan synthesis (Benedetti et al. 2017; Schuster et al. 2005; Ferrara et al. 2006), the increase in binding affinity fo FcγRIII sometimes associated with bisected glycans (Umaña et al. 1999; Davies et al. 2001; Lifely et al. 1995) cannot be easily uncoupled from the same effect observed for core-fucosylated IgG glycans (Shinkawa et al. 2003).

Galactosylation

Galactosylation is the IgG glycosylation trait with the most pronounced inter-individual variation (Huhn et al. 2009; Gornik et al. 2012). On average, about 35% of IgG Fc glycans contain no terminal galactose residues (agalactosylated glycans, G0), about 35% contain one (monogalactosylated glycans, G1), and about 15% contain two terminal galactoses (digalactosylated glycans, G2) (Baković et al. 2013; Huffman et al. 2014). Terminal galactoses modulate IgG inflammatory potential by affecting binding affinities to complement components and FcγRs. Agalactosylated Fc glycans are considered to act pro-inflammatory by activating the complement through the alternative pathway (Banda et al. 2008), and the lectin pathway after binding to the mannose-binding lectin (Malhotra et al. 1995; Ji et al. 2002; Arnold et al. 2006). Galactosylation was also held responsible for the anti-inflammatory activity of immune complexes by binding to the inhibitory FcγRIIB (Karsten et al. 2012). However, Fc galactosylation has also been reported to enhance complement-dependent cytotoxicity (CDC) through the classical complement pathway by increasing the IgG’s affinity for the C1q complement component (Peschke et al. 2017; Boyd et al. 1995; Hodoniczky et al. 2005). Likewise, by increasing the affinity of IgG for FcyRs, it enhances the downstream processes mediated by FcyRs, in particular ADCC (Dekkers et al. 2017; Kumpel et al. 1994, 1995; Houde et al. 2010; Subedi and Barb 2016). We should therefore not rush to proclaim terminal IgG galactosylation simply “anti-inflammatory,” before considering the entire context and the nature and extent of IgG involvement in the process we are investigating.

Sialylation

On average, 10–15% of IgG Fc glycans carry a single terminal sialic acid (monosialylated glycans, S1) or two sialic acids (disialylated glycans, S2) (Baković et al. 2013; Huffman et al. 2014). Similar to terminal galactosylation, sialylation is most often discussed as a modulator of IgG functions regarding inflammation (Böhm et al. 2014).

The importance of sialylation became evident when the presence of the sialylated Fc fraction was discovered indispensable for the anti-inflammatory activity of the intravenous immunoglobulin (IVIg) preparation in a K/BxN serum-transfer mouse model of RA (Kaneko et al. 2006). Mouse studies on several antibody-dependent autoimmune disease models helped elucidate the mechanistic pathway for its activity, starting with the binding of the sialylated Fc fraction to specific ICAM-3 grabbing non-integrin-related 1 (SIGN-R1) on the surface of splenic macrophages and ending in enhanced FcγRIIB expression on the effector macrophages (Kaneko et al. 2006; Schwab and Nimmerjahn 2013; Anthony et al. 2008, 2011; Schwab et al. 2012, 2014; Washburn et al. 2015; Galeotti et al. 2017; Fiebiger et al. 2015). However, this finding did not hold in several other in vitro and in vivo models, nor human studies (Galeotti et al. 2017; Guhr et al. 2011; Leontyev et al. 2012; Campbell et al. 2014; Temming et al. 2019). This confirms the well-established notion that the IVIg mode of action is complex and tightly connected with the corresponding immune context.

Depending on the sialylation status of the Fc glycan, the Fc domain is suggested to adopt either an “open” or a “closed” conformation, for sialylated and asialylated glycans, respectively. The “open” conformation favors binding to the type I FcγRs, whereas the “closed” conformation favors the binding of type II FcRs (Pincetic et al. 2014; Sondermann et al. 2013). Terminal sialylation is thus proposed to act as a switch between two distinct immunological effector functions.

To summarize—agalactosylated, asialylated, and bisected IgG molecules are often simply described as “pro-inflammatory,” and terminally galactosylated and sialylated IgG molecules as “anti-inflammatory,” while afucosylated IgG has an augmented capacity to trigger ADCC through enhanced FcγRIIIA binding. We should, however, always bear in mind that this generalization is a simplification, and exercise caution when considering its implications.

3.2 Fab Glycans

As expected, Fab glycans are mostly reported to affect antigen-binding (Wright et al. 1991; Coloma et al. 1999; Schneider et al. 2015; Wallick et al. 1988; Tachibana et al. 1997; Leibiger et al. 1999; Khurana et al. 1997; Man Sung Co et al. 1993; Fujimura et al. 2000; Van De Bovenkamp et al. 2018b). Besides the obvious, they are also suggested to influence IgG aggregation and precipitation (Courtois et al. 2016), immune complex formation (Gutierrez et al. 2006), and have a role in the IVIg mode of action (Käsermann et al. 2012; Wiedeman et al. 2013; Massoud et al. 2014; Séïté et al. 2010, 2014).

4 Regulation of IgG Glycosylation

IgG glycosylation is a complex trait, influenced by both, genetics (Menni et al. 2013; Pučić et al. 2011; Klarić et al. 2020) and the environment (Štambuk et al. 2020; Yu et al. 2016; Krištić et al. 2014; De Jong et al. 2016). More precisely, the compound IgG glycosylation pattern seems to be, to different degrees, modulated by IgG aminoacid sequence (Lund et al. 1996; Zaitseva et al. 2018; Yu et al. 2013), the intra- and extracellular milieu affecting the glycosylation machinery (Ohtsubo and Marth 2006; Oefner et al. 2012; Bartsch et al. 2020; Hess et al. 2013; Canellada et al. 2002; Gutiérrez et al. 2001; Miranda et al. 2005; Wang et al. 2011; Pfeifle et al. 2017; Liu et al. 2014; Fan et al. 2015), and environmental factors (Novokmet et al. 2014; Greto et al. 2020; Ercan et al. 2017; Engdahl et al. 2018; Klasić et al. 2018; Tijardović et al. 2019; Sarin et al. 2019; Peng et al. 2019). Solving the outstanding question of IgG glycosylation regulation would likely bring us one step closer to understanding the possible functionality of changes in IgG glycan composition in different physiological and pathological states.

5 Common IgG Glycosylation Pattern in Inflammatory Diseases and Aging

Advances in the development of high-throughput glycomic and glycoproteomic analyses (Huhn et al. 2009; Mariño et al. 2010; Trbojević-Akmačić et al. 2016, 2017) have enabled a significant number of large-scale epidemiological studies examining total IgG glycosylation pattern in diseased vs. healthy control subjects (Štambuk et al. 2020; Singh et al. 2020; Lemmers et al. 2017; Menni et al. 2018; Šimurina et al. 2018; Theodoratou et al. 2016; Wahl et al. 2018). In many of the diseases that were studied, a similar pattern emerged: diseased subjects often exhibited a decreased abundance of galactosylated, sialylated, and—occasionally—an increased abundance of bisected bulk IgG glycans when compared to healthy controls (Fig. 13.1). In addition, the trend was often associated with disease severity and reverted to baseline values upon successful application of therapy. Interestingly, the same pattern that was observed in diseases with an inflammatory component was also evident in aging subjects (Fig. 13.1) (Gudelj et al. 2018a; Lauc 2016). This “pro-inflammatory IgG glycome composition” is likely associated with the common background inflammatory component of the studied diseases. In some cases, it might reflect a predisposition toward disease development, or/and even be involved as an effector of inflammation. Additionally, it might represent a consequence of environmental exposure to antigens through a lifetime or unhealthy lifestyle choices.

Fig. 13.1
figure 1

General patterns of IgG glycosylation changes are similar in several diseases and aging. The effect (shown on the y-axis) is shown as the difference between means of case and control populations (for aging, population over vs. population under 50 years of age), expressed in standard deviations. The whiskers represent the 95% confidence interval. The reference populations for disease cohorts are age- and sex-matched healthy controls. SLE systemic lupus erythematosus, RA rheumatoid arthritis, CD Crohn’s disease, UC ulcerous colitis, T2D type 2 diabetes mellitus, CRC colorectal carcinoma. G0 agalactosylated glycans, G2 digalactosylated glycans, B bisected glycans, S sialylated glycans, F core-fucosylated glycans. Reused with permission from Lauc (2016)

Full size image

Indeed, the composition of IgG glycome was reported to associate with many physiological and biochemical traits, as well as with traits correlated to inflammation and poor metabolic health (Gudelj et al. 2018a) and the expected lifespan (Štambuk et al. 2020). IgG glycome was thus suggested to be a biomarker of general immune activation (De Jong et al. 2016), while we propose total IgG glycoprofile can be positioned as a read-out of a general state of health, i.e. biological age (Vilaj et al. 2019).

6 Role of Skewed IgG Glycosylation in Diseases

When we take into account the complexity of the IgG’s multiple roles in our immune system, it is no wonder there is no single common interpretation of the altered IgG glycopattern across the wide spectrum of diseases (Table 13.1). The multiple pleiotropic loci, i.e. shared associations of IgG glycome composition and autoimmune, inflammatory, and other diseases (Klarić et al. 2020; Lauc et al. 2013), as well as the changes in IgG glycopattern preceding disease development—such as in the case of RA (Gudelj et al. 2018b) and cardiovascular diseases (Menni et al. 2018)—suggest that a skewed bulk serum IgG glycoprofile might reflect a disease risk or predisposition. This predisposition can manifest through an inherited (Klarić et al. 2020; Lauc et al. 2013) or acquired propensity for inflammation modulation (Franceschi et al. 2018).

Table 13.1 Diseases exhibiting an altered serum IgG glycosylation profile
Full size table

In most other cases, when it comes to glycosylation of the bulk serum IgG, the role of a shifted glycosylation pattern is not clear. As already mentioned, decreased galactosylation and sialylation often accompany diseases that involve an inflammatory immune response. The evidence that would enable us to unambiguously determine whether the “pro-inflammatory” IgG glycoforms represent one of many drivers of disease pathology or merely byproducts of the inflamed immune system is still lacking. The current consensus is that total IgG glycopattern is likely relevant in the general modulation of the immune activation threshold.

In some cases, however, a clear link/evidence for the functionality of IgG glycans is provided. A mouse study investigating the link between obesity and the development of hypertension resulted in a very intriguing observation. Hyposialylated IgG from mice in which obesity was induced by a high-fat diet (HFD) induced an elevated blood pressure when transferred to IgG-deficient mice. Moreover, supplementing HFD-feed mice with a sialic acid precursor, N-acetyl-d-mannosamine (ManNAc), resulted in the restoration of the baseline level of IgG sialylation and protected them from obesity-induced hypertension development (Peng et al. 2019). This finding thus demonstrated the functional role of IgG glycans in the development of hypertension. Interestingly, the same treatment restored IgG sialylation and reduced tumor load and bone loss in a mouse model of myeloma (Westhrin et al. 2020).

On the level of total serum IgG, increased level of glycosylation of the Fab region observed in some malignant diseases (Zhu et al. 2002, 2003; Radcliffe et al. 2007; Coelho et al. 2010; McCann et al. 2008) is proposed to contribute to disease development and progression by enhancing tumor cell persistence and expansion (Coelho et al. 2010; Amin et al. 2015).

Glycosylation changes on antigen-specific IgG are more likely to be directly involved in disease pathology in case of antibody-mediated auto- or alloimmune diseases or defense from pathogens in case of infectious diseases. The role of differential IgG glycosylation in these cases corresponds to the specifics of a particular disease and the molecular mechanisms underlying its pathology.

In addition to the change in total IgG, multiple infectious diseases are characterized by a distinct glycosylation pattern of relevant antigen-specific IgG in comparison to total IgG (Table 13.1). This implies a distinct regulation of IgG glycosylation, depending on both the disease and the antigen (Ackerman et al. 2013), even within a single individual (Mahan et al. 2016). This supports the notion that IgG glycome relevance should be interpreted in the disease-specific functional context.

One of the rare instances where the role of IgG glycosylation is mechanistically explained is once more linked to the enhanced affinity of afucosylated IgG molecules for FcγRIIIA. In the case of dengue fever, occasionally a secondary, heterologous dengue infection results in severe dengue hemorrhagic fever and dengue shock syndrome. This is attributed to antibody-dependent enhancement (ADE) of the disease by cross-reactivity of afucosylated anti-dengue IgG with platelet antigens, resulting in platelet depletion (Wang et al. 2017). Additionally, the enhanced binding of afucosylated IgG to FcγRIIA and FcγRIIIA promotes the FcγR-mediated viral entry and signaling in cells bearing these receptors on their surface, primarily monocytes and macrophages, resulting in infection progression (Thulin et al. 2020).

A similar relevance for afucosylated antigen-specific IgG is observed in COVID-19 patients. Anti-SARS-CoV-2 IgG with a higher core-fucosylation level is associated with unaided clearance of the infection (Larsen et al. 2020). By contrast, critically ill patients display lower levels of fucosylated anti-SARS-CoV-2 IgG (Larsen et al. 2020; Chakraborty et al. 2021). Furthermore, in in vitro studies afucosylated anti-S/-RBD antibodies were shown to induce enhanced natural killer (NK) cell degranulation (Chakraborty et al. 2021) and elevated production of pro-inflammatory cytokines by primary monocytes and alveolar macrophages, which is likely the background of the severe disease phenotype associated with this glycoprofile in vivo (Larsen et al. 2020; Chakraborty et al. 2021; Hoepel et al. 2020).

Similarly, afucosylated antigen-specific IgG in fetal and neonatal alloimmune thrombocytopenia (FNAIT) and hemolytic disease of the fetus and newborn (HDFN) are thought to contribute, again through enhanced FcγRIIIA-mediated mechanisms, namely phagocytosis and ADCC, to the more severe disease phenotype (Kapur et al. 2014a, b; Sonneveld et al. 2016, 2017a).

In lupus nephritis, a serious complication of SLE, the presence of core fucose was shown to induce upregulated calcium/calmodulin kinase IV expression in podocytes, leading to podocyte injury and limited nephrin synthesis. In the same experimental setting, the presence of terminal galactoses acted protectively (Bhargava et al. 2021).

An interesting recent finding on the importance of Fab glycans emerged in the most explored disease in the context of IgG glycosylation. In RA, a high percentage of anti-citrullinated protein antibody (ACPA) is additionally glycosylated at the Fab region (Rombouts et al. 2016; Hafkenscheid et al. 2017), a feature distinguishing RA patients from ACPA+ but healthy subjects (Kissel et al. 2019; Hafkenscheid et al. 2019). This suggests Fab glycosylation of ACPA might be mechanistically involved in RA development (Rombouts et al. 2016).

7 Perspectives for IgG Glycosylation in Precision Medicine

A skewed IgG glycoprofile in comparison to the personal baseline value (requiring longitudinal monitoring) or in comparison to ethnicity-, age-, and sex-matched subjects (requiring a population baseline cohort) in a cross-sectional experimental design, might indicate an increased risk for disease development (Gudelj et al. 2018b), or disease progression (Gudelj et al. 2018a). However, since the alterations in bulk serum IgG glycome composition are not disease-specific, they cannot be used as a stand-alone diagnostic marker. A total IgG glycoprofile of the composition significantly removed from the baseline can instead be used as an indication of a necessity for an examination by an expert clinician.

In case of an established diagnosis, bulk IgG glycome might serve as a predictor of disease progression—e.g., decreased IgG2/3 galactosylation in patients progressing from undifferentiated to rheumatoid arthritis (Sénard et al. 2021). Similarly, IgG glycome is proposed to bear potential for a useful add-on tool for monitoring functional disease progression and response to therapy (Parekh et al. 1988; Kanoh et al. 2004a, 2008; Váradi et al. 2015; Collins et al. 2013; Van Zeben et al. 1994; Rook et al. 1994; Pasek et al. 2006; Gindzienska-Sieskiewicz et al. 2007; Croce et al. 2007; Ercan et al. 2010).

The relevance and biomarker potential of IgG glycome analysis is more evident in some cases of antigen-specific IgG. For instance, due to the increased level of ACPA Fab glycosylation in individuals at risk for RA development, IgG glycome analysis might in the future provide the currently missing understanding (and biomarker) for the first determining pathogenic event leading to disease development (Rombouts et al. 2015; Scherer et al. 2010). Furthermore, as already mentioned, in several diseases a particular antigen-specific IgG glycopattern is associated with a risk for the severe phenotype (Kapur et al. 2014a, b; Sonneveld et al. 2016; Sonneveld et al. 2017a). Similarly, following the mechanistical explanation for the role of afucosylated anti-dengue IgG described in the previous section, afucosylated maternal anti-dengue IgG is proposed to denote a susceptibility to symptomatic dengue infection in infants (Thulin et al. 2020). The knowledge that a particular glycan profile of antigen-specific IgG, including post-vaccination status for some infectious diseases, is related to the risk of developing (the severe form of) a disease might in the future enable or aid the stratification of patients at risk and timely preventive action.

Another sought-after biomarker type is the one enabling patient stratification aiming at improved differential (sub-)diagnosis and subsequent selection of appropriate therapeutic measures. Differential IgG glycosylation was also suggested as a possibility for such applications. Indeed, the IgG sialylation level predicted response to therapy in Kawasaki disease (Ogata et al. 2013), and the galactosylation level response to anti-tumor necrosis factor (TNF) therapy in RA and Crohn’s disease (Váradi et al. 2015), and response to methotrexate therapy in RA (Lundström et al. 2017). Having the means to distinguish non-responders before the very initiation of long and expensive therapeutic treatments is truly an exciting prospect.

In summary, there are multiple possibilities for IgG glycosylation to enter the arena of clinical disease management. Currently, all of the possible applications mentioned here are still at the level of basic research and further studies are necessary to validate the initial findings and propel the IgG glycome analysis to the status of a full-fledged clinical biomarker.

8 Conclusions

IgG glycans can modulate virtually all of its numerous effector roles, the specifics depending on the disease and immune context. The associations of multiple IgG glycosylation traits with an immense array of heterogeneous diseases and their different stages imply that there is no single pathway connecting IgG glycome composition and disease development and progression.

Many inflammatory, autoimmune, infectious, cardiometabolic, and neoplastic diseases share a common IgG glycosylation profile of bulk (total) serum IgG, also characteristic for aging and often described as “pro-inflammatory”: a decreased level of galactosylated and sialylated glycans, and (sometimes) an increased level of bisected IgG glycans. This pattern is presumably associated with an inflammatory disease component as a part or consequence of disease pathology, or environmental events, such as antigen exposure. It might be mechanistically involved in disease advancement through modulation of inflammation, and, in some cases, manifest before the occurrence of symptoms, thus representing disease predisposition or mark the risk for disease development or progression.

When it comes to a distinct glycosylation profile of antigen-specific versus total serum IgG, IgG glycans are more likely to be directly involved in disease pathogenesis and progression through disease-specific effector mechanisms. This is often the case with afucosylated IgG glycans enhancing the affinity of IgG toward FcγRIIIA.

The read-out of IgG glycosylation has a potential for an (add-on) biomarker helping improve current algorithms for disease prediction and diagnosis, patient stratification, monitoring of disease progression, and response to therapy.