Glycosylation of Fcγ receptors influences their interaction with various IgG1 glycoforms

Highlights

• Glycosylation of both Immunoglobulins G1 and Fcγ receptors affects their kinetics of interactions.

• We performed site-directed mutagenesis along with glycoengineering on Fcγ receptors (FcγRI, FcγRIIa/b and FcγRIIIa_V158/F158).

• We here report, within the same study, the kinetics of interactions of Fcγ receptorsglycovariants with 3 IgG glycovariants.

Abstract

Most of therapeutic monoclonal antibodies belong to the immunoglobulin G1 (IgG1) family; they interact with the Fcγ receptors (FcγRs) at the surface of immune cells to trigger effector functions. The IgG1-Fc N-glycans impact the interaction with FcγRs and are considered a critical quality attribute. Pioneer studies on FcγR N-glycans have unveiled an additional complexity in that the N-glycan linked on the Asn-162 of FcγRIIIa was shown to be directly involved in the strong affinity for afucosylated IgG1. The last few years have thus seen the emergence of many studies investigating the complex influence of FcγRIIIa N-glycans on the interaction with IgG1 through their glycosylation sites or their glycoprofiles. In this context, we performed site-directed mutagenesis along with glycoengineering on FcγRs (FcγRI, FcγRIIa_H131/b and FcγRIIIa_V158/F158) in an effort to elucidate the impact of FcγRs N-glycans on the interaction with IgG1. Furthermore, we assessed their binding to various trastuzumab glycoforms with an enhanced surface plasmon resonance assay. The FcγRIIIa N-glycans had the highest impact on the interaction with IgG1. More specifically, the N162 glycan positively influenced the affinity (15-fold) for afucosylated IgG1 while the N45 glycan presented a negative impact (2-fold) regardless of the IgG1 glycoforms. Interestingly, only the FcγRIIIa glycoprofile had an impact on the interaction with IgG1 with a 1.5-fold affinity increase when FcγRIIIa displays high-mannose glycans. These results provide invaluable insights into the complex and strong influence of N-glycosylation upon FcγRs/IgG1 binding and are instrumental to further understand the impact of FcγRs N-glycosylation in their natural forms.

Introduction

Since the first monoclonal antibody (mAb) approval in 1986, this class of biotherapeutics has grown continuously and now dominates the biopharmaceutical market (Walsh, 2018). Recombinant mAbs belong mostly to the immunoglobulin G1 (IgG1) family. They recognize antigens via their Fab regions while their Fc region binds to proteins of the immune system, including the Fcγ receptors (FcγRs) (Jefferis, 2010). Depending on the immune cell type and the FcγRs present, the interactions between the immune complexes and the FcγRs can trigger different effector functions (Lejeune et al., 2010). The IgG1-Fc region is composed of two identical heavy chains that both present a highly conserved N-glycosylation site at the asparagine residue 297 (Asn-297) (Jefferis, 2009). N-glycosylation is considered a critical quality attribute (CQA) for therapeutic mAbs (Zhang et al., 2016) as it can impact mAb serum half-life (Liu, 2015). More importantly, IgG1-Fc N-glycans also influence the interactions between the IgG1 and FcγRs. For example, deglycosylation of IgG1 leads to the abrogation of the interaction with FcγRIIIa (Subedi and Barb, 2015) and drastically decreases the IgG1 affinity for FcγRI (Dorion-Thibaudeau et al., 2016; Oganesyan et al., 2015). Besides, the N-glycan profile can significantly modulate FcγR interactions and consequently the triggering of effector functions, including the antibody-dependent cell-mediated cytotoxicity (ADCC) and the complement-dependent cytotoxicity (CDC) (Wada et al., 2019; Dekkers et al., 2017). Therapeutic agents with enhanced clinical efficacy have thus been developed via the selective enrichment of desired glycoforms of IgG1s while FcγRs are commonly used to investigate mAbs safety and efficacy (Hayes et al., 2014a). Furthermore, an increasing number of studies have also highlighted the complex and significant impact of the FcγR N-glycosylation on the FcγR / IgG1 interaction and its potential role in modulating the immune system activation or inhibition (Hayes et al., 2017; Hayes et al., 2016; Subedi and Barb, 2018; Ferrara et al., 2011).

FcγRs are glycoproteins classified into three families: FcγRI (or CD64), FcγRIIa-c (or CD32a-c) and FcγRIIIa/b (or CD16a/b) (Nimmerjahn et al., 2015). FcγRI has a high affinity for IgG1s (apparent K_D of 10⁻⁸ – 10^-9 M) and interacts with monomeric IgG1 (Lu and Sun, 2015). On the other hand, the type II and type III receptors display lower affinities (apparent K_Ds of ≈ 10^-6 – 10^-7 M) and must be in the presence of an immune complex to be engaged (Nimmerjahn and Ravetch, 2007). At last, FcγRs have between 2–7 N-glycosylation sites (i.e., 7 for FcγRI, 2 for FcγRIIa, 3 for FcγRIIb, 5 for FcγRIIIa and 6 for FcγRIIIb) (Hayes et al., 2016). Glycosylation variations occur in the final structure of the attached glycan as well as in the site-occupancy, those variations being referred to as microheterogeneity and macroheterogeneity, respectively (Butler, 2006).

Several studies have highlighted the influence of FcγR glycosylation on IgG1 binding by sequentially knocking out FcγR glycosylation sites (Ferrara et al., 2006; Shibata-Koyama et al., 2009). In particular, it has been demonstrated that the affinity for afucosylated IgG1s is decreased by over one order of magnitude for FcγRIIIa lacking the Asn-162 (N162) glycan. Initially, two independent studies suggested that the absence of core fucose on the IgG1-Fc diminished steric hindrance, hence promoting carbohydrate-carbohydrate interactions combined to carbohydrate-protein and protein-protein interactions (Ferrara et al., 2011; Mizushima et al., 2011). However, another mechanism, which does not imply intermolecular interactions with the IgG1-Fc has recently been proposed (Falconer et al., 2018). More precisely, it has been suggested that the core fucose on the IgG1-Fc N-glycan would restrict the conformations sampled by the FcγRIIIa-N162 glycan and would partially inhibit the binding with FcγRIIIa. In contrast, the glycan on the Asn-45 (N45) of FcγRIIIa led to a decrease of IgG1 binding affinity (Shibata-Koyama et al., 2009). Finally, a decrease in the binding levels and an increase in the dissociation rate of antibody have been observed when FcγR glycans are removed (Hayes et al., 2014b).

The N-glycan patterns have also been investigated by global (Hayes et al., 2017; Hayes et al., 2014b; Takahashi et al., 2002; Cosgrave et al., 2013) or site-specific (Zeck et al., 2011; Kawasaki et al., 2014) N-glycosylation profile analysis of the FcγRs expressed in different cell types. Of interest, these studies have shown that the FcγR N-glycosylation is cell-type specific and that these variations influence the IgG1 binding kinetics (Hayes et al., 2017; Zeck et al., 2011). In addition, site-specific N-glycan structures have been observed (Zeck et al., 2011; Kawasaki et al., 2014). Although valuable, these results are related to recombinant FcγRs expressed in mammalian cell lines (i.e., CHO, HEK, NSO and BHK) (Hayes et al., 2016). These recombinant FcγRs display N-glycan structures that do not necessarily reflect the natural N-glycans and their potential impact (Patel et al., 2018). Thus, to better understand how the immune system could use the FcγR N-glycosylation as a mechanism to fine-tune the antibody response, the investigation of the natural N-glycosylation forms of the FcγRs present on each immune cell, both in disease and healthy states, is necessary (Hayes et al., 2014a). However, FcγRs N-glycosylation characterization in native human tissues represents a real challenge given the difficulty in obtaining enough material for analysis (Patel et al., 2018). Nonetheless, a study has recently performed a site-specific N-glycosylation analysis of the soluble FcγRIIIb in human serum (Yagi et al., 2018) while another achieved the N-glycan analysis of the FcγRIIIa from primary human NK cells (Patel et al., 2018). Of salient interest, Patel et al. have demonstrated that the FcγRIIIa N-glycans at the surface of NK cells display higher level of high-mannose and hybrid structures (Patel et al., 2018), starkly contrasting with previous studies on the recombinant FcγRIIIa N-glycosylation, which identified predominantly complex structures (Hayes et al., 2017; Hayes et al., 2014b; Cosgrave et al., 2013; Zeck et al., 2011). Further analysis mixing N-glycosylation site substitution and N-glycan composition modification have then indicated that the FcγRIIIa with an oligomannose structure have 10 to 50-fold higher affinity for various IgG1 glycoforms (Subedi and Barb, 2018). These results thus validate the conclusion proposed by Edberg et al. that FcγR-mediated antibody response could be tuned by cell-specific glycome (Edberg and Kimberly, 1997).

With the aim of understanding the influence of FcγR N-glycosylation and thus potentially helping at improving immunotherapy, we herein performed site-directed mutagenesis, along with glycoengineering, on five FcγRs (FcγRI, FcγRIIa_H131/b and FcγRIIIa_V158/F158). We then assessed in detail the impact of the FcγR N-glycosylation on binding to three distinct IgG1 glycoform lots (reference, afucosylated and galactosylated). The IgG1 / FcγR interaction affinities were measured by surface plasmon resonance (SPR) biosensing, a biophysical technique commonly used to study the IgG1 / FcγR interactions and the impact of N-glycosylation on binding (Hayes et al., 2017; Subedi and Barb, 2018; Ferrara et al., 2006; Shibata-Koyama et al., 2009; Hayes et al., 2014b; Zeck et al., 2011; Patel et al., 2018; Bruhns et al., 2009; Cambay et al., 2019). To our knowledge, only few studies have combined glycoengineering on both IgG1 and FcγRs to investigate the impact of glycosylation upon interaction (Subedi and Barb, 2018; Ferrara et al., 2011; Shibata-Koyama et al., 2009), leading to ambiguous conclusions, which may be attributed to either the inherent complexity of the interaction or biases from the SPR assay. In our case, the binding affinities of each FcγR mutant for each IgG1 glycoform were measured with confidence using our validated SPR assay relying on coiled-coiled interactions to capture the various FcγR ectodomains at the biosensor surfaces (Cambay et al., 2019). Overall, this work helps elucidate the complex influence of FcγR and IgG1 N-glycans, along with the global and site-specific N-glycan composition, upon antibody-receptor interactions.