Mixed IgG Fc immune complexes exhibit blended binding profiles and refine FcR affinity estimates

Introduction:: Antibodies are both a core component of adaptive immunity and a versatile platform for developing therapies. IgG antibodies coordinate immune effector responses by selectively binding to target antigens and then interacting with various effector cells via the Fcγ receptors. FcγR activation is initiated through IgG-mediated clustering, which in turn is caused by the engagement of several antibodies on an antigen target forming an immune complex (IC) [1]. The Fc domain of IgG can promote or inhibit distinct effector responses across several different immune cell types through variation in subclasses and Fc domain glycosylations. Extensive characterization of these interactions has revealed how the inclusion of certain Fc subclasses or glycans results in distinct immune responses, but the effect of IgG Fc mixtures has not been well quantified [2]. During an immune response, IgGs are almost always produced with mixtures of Fc domains. In this work, we examined the binding properties of ICs with mixed IgG Fc composition by applying a multivalent binding model to capture such hetero-valent IC mixtures binding [3,4]. Additionally, we examined whether and how this mixed composition influences immune effector responses by estimating their effector cell interactions. We linked their binding behavior to individual-level target cell depletion by analyzing humanized mice in vivo platelet depletion experiments with this binding model.


Materials and Methods:: We systematically measured the in vitro binding of mixed Fc IgG immune complexes formed by each pair of IgG subclasses to FcγR-expressing CHO cells. To capture the binding behavior of mixed IgG ICs, we extended a multivalent binding model so it could accurately account for the binding of an IgG cocktail to cells expressing one or more FcγRs (Fig. a) [4]. We fit the measurements to this model with a Monte Carlo Markov Chain (MCMC) (Fig. b). This fitting process inferred posterior (updated) distributions from documented ones (priors) for not only the experimental parameters such as immune complex concentration and receptor expression, but also the IgG-FcγR binding affinities. To examine IgG-elicited immune effector leukocyte responses, we measured the binding of each subclass to classical and nonclassical monocytes and neutrophils in vitro, then compared them with the model prediction. We then combined the model with generalized linear regression to quantify how the response of each immune effector cell contributes to antibody-mediated platelet depletion in humanized mice (Fig. f) [5].


Results, Conclusions, and Discussions:: We first validated our mixed-composition IgG binding measurements and demonstrated that the binding properties of mixed Fc immune complexes fell along a continuum between those of the corresponding pure cases. Fitting these measurements to the mechanistic binding model using documented affinities for each interaction, our model captured overall binding trends, with some outliers on several low-affinity interactions mostly involving IgG2 (Fig. c). We uncovered that the model discrepancies could be explained by inaccurate estimates of those low-affinity Fc receptor interactions such as IgG2 (Fig. d). Using Bayesian inference with MCMC, we refined those affinities, mostly on IgG2 and IgG4, and demonstrated that IgG2-FcγRI, a previously reported nonbinding pair, is actually the highest-affinity interaction for IgG2 (Fig. e). We validated revised affinities within an independent dataset and found it greatly improved concordance with the data there as well [3]. This shows that due to the heightened avidity, multivalent ICs are better at detecting low-affinity IgG-Fc receptor interactions. Examining binding through ICs also better simulates the relevant structure of Fc-FcR interactions in vivo.

To explore antibody-mediated effector cell response, we measured the in vitro IgG-effector cell binding and compared them with the model predictions made by documented and updated affinities [2]. The updated affinities matched the experiments better, with most improvements occurring on IgG2. Finally, we demonstrated that the binding model could predict antibody-driven effector-cell elicited platelet depletion in humanized mice, with the model inferring the relevant cell populations (Fig. g). While the updated affinities did not change the agreement of the model with the data, it did change the interpretation of IgG2’s small effect on depletion—rather than not binding to classical monocytes, IgG2 binds strongly when in a larger IC, but platelets might provide insufficient avidity to observe sufficient engagement.

Overall, our work demonstrates that computational methods greatly facilitate quantitative reasoning about the Fcγ receptor pathway's complex signaling at both cellular and organismal levels. We anticipate that this study will continue to grow in its utility for developing more sophisticated antibody-based therapeutics in the age where mixtures of subclasses, Fc glycosylation, and IgG allotypes can all be harassed for optimal effect.


Acknowledgements (Optional): : This work was supported by NIH U01-AI-148119 to A.S.M. and F.N.


References (Optional): : [1] Nimmerjahn, F. & Ravetch, J. V. Divergent Immunoglobulin G Subclass Activity Through Selective Fc Receptor Binding. Science 310, 1510–1512 (2005).

[2] Bruhns, P. et al. Specificity and affinity of human Fcγ receptors and their polymorphic variants for human IgG subclasses. Blood 113, 3716–3725 (2009).

[3] Robinett, R. A. et al. Dissecting FcγR Regulation through a Multivalent Binding Model. Cell Syst. 7, 41-48.e5 (2018).

[4] Tan, Z. C. & Meyer, A. S. A general model of multivalent binding with ligands of heterotypic subunits and multiple surface receptors. Math. Biosci. 342, 108714 (2021).

[5] Schwab, I., Lux, A. & Nimmerjahn, F. Pathways Responsible for Human Autoantibody and Therapeutic Intravenous IgG Activity in Humanized Mice. Cell Rep. 13, 610–620 (2015).