A hallmark of B-cell immunity is the generation of a diverse repertoire of antibodies from a limited set of germline V(D)J genes. This repertoire is usually defined in terms of amino acid composition. However, variable domains may also acquire -linked glycans, a process conditional on the introduction of consensus amino acid motifs (-glycosylation sites) during somatic hypermutation. High levels of variable domain glycans have been associated with autoantibodies in rheumatoid arthritis, as well as certain follicular lymphomas. However, the role of these glycans in the humoral immune response remains poorly understood. Interestingly, studies have reported both positive and negative effects on antibody affinity. Our aim was to elucidate the role of variable domain glycans during antigen-specific antibody responses. By analyzing B-cell repertoires by next-generation sequencing, we demonstrate that -glycosylation sites are introduced at positions in which glycans can affect antigen binding as a result of a specific clustering of progenitor glycosylation sites in the germline sequences of variable domain genes. By analyzing multiple human monoclonal and polyclonal (auto)antibody responses, we subsequently show that this process is subject to selection during antigen-specific antibody responses, skewed toward IgG4, and positively contributes to antigen binding. Together, these results highlight a physiological role for variable domain glycosylation as an additional layer of antibody diversification that modulates antigen binding.

Immunoglobulin G (IgG) antibodies are symmetrical molecules that may be regarded as covalent dimers of 2 half-molecules, each consisting of a light chain and a heavy chain. Human IgG4 is an unusually dynamic antibody, with half-molecule exchange ("Fab-arm exchange") resulting in asymmetrical, bispecific antibodies with two different antigen binding sites, which contributes to its anti-inflammatory activity. The mechanism of this process is unknown. To elucidate the elementary steps of this intermolecular antibody rearrangement, we developed a quantitative real-time FRET assay to monitor the kinetics of this process. We found that an intrinsic barrier is the relatively slow dissociation of the CH3 domains that noncovalently connect the heavy chains, which becomes rate determining in case disulfide bonds between the heavy chains are reduced or absent. Under redox conditions that mimic the previously estimated in vivo reaction rate, i.e., 1 mM of reduced glutathione, the overall rate is ca. 20 times lower because only a fraction of noncovalent isomers is present (with intra- rather than interheavy chain disulfide bonds), formed in a relatively fast pre-equilibrium from covalent isomers. Interestingly, Fab arms stabilize the covalent isomer: the amount of noncovalent isomers is ca. 3 times higher for Fc fragments of IgG4 (lacking Fab domains) compared to intact IgG4, and the observed rate of exchange is 3 times higher accordingly. Thus, kinetic data obtained from a sensitive and quantitative real-time FRET assay as described here yield accurate data about interdomain interactions such as those between Fab and/or Fc domains. The results imply that in vivo, the reaction is under control of local redox conditions.

Human IgG4, normally the least abundant of the four subclasses of IgG in serum, displays a number of unique biological properties. It can undergo heavy-chain exchange, also known as Fab-arm exchange, leading to the formation of monovalent but bispecific antibodies, and it interacts poorly with FcγRII and FcγRIII, and complement. These properties render IgG4 relatively “non-inflammatory” and have made it a suitable format for therapeutic monoclonal antibody production. However, IgG4 is also known to undergo Fc-mediated aggregation and has been implicated in auto-immune disease pathology. We report here the high-resolution crystal structures, at 1.9 and 2.35 Å, respectively, of human recombinant and serum-derived IgG4-Fc. These structures reveal conformational variability at the CH3–CH3 interface that may promote Fab-arm exchange, and a unique conformation for the FG loop in the CH2 domain that would explain the poor FcγRII, FcγRIII and C1q binding properties of IgG4 compared with IgG1 and -3. In contrast to other IgG subclasses, this unique conformation folds the FG loop away from the CH2 domain, precluding any interaction with the lower hinge region, which may further facilitate Fab-arm exchange by destabilisation of the hinge. The crystals of IgG4-Fc also display Fc–Fc packing contacts with very extensive interaction surfaces, involving both a consensus binding site in IgG-Fc at the CH2–CH3 interface and known hydrophobic aggregation motifs. These Fc–Fc interactions are compatible with intact IgG4 molecules and may provide a model for the formation of aggregates of IgG4 that can cause disease pathology in the absence of antigen.