AbstractsEngineering

Summary Antibodies act as adapter molecules that link recognition of foreign substances, such as bacteria and viruses, to elimination mechanisms mediated by cells and molecules of the innate immune system. Cells that express the classical Fcγ receptors (FcγRs) bind immunoglobulin G (IgG) and IgG containing immune complexes (IC). This, and the fact that IgG has long serum half-life, has made IgG the first choice for antibody-based therapeutics. The serum half-life of IgG is regulated by a recycling pathway that involves the neonatal Fc receptor (FcRn), which binds the Fc region of IgG. Thus, there is a great interest in the design of novel IgGs with altered interaction with FcRn and increased or shortened serum half-life. However, such Fc-engineering should necessarily not affect binding to the classical FcγRs, whenever the therapeutic effects of IgGs depend on optimal binding to these receptors. Thus, the aim of this study was to perform a thorough investigation of how Fc-engineered IgG1 variants with altered binding to FcRn bind to the mouse and human FcγRs. A range of such recombinant human Fc-engineered IgG1s variants were made, chosen from previously published reports, and screened for binding to soluble recombinant mouse and human FcγRs using in vitro interaction assays. Although FcRn and the FcγRs bind to separate binding sites on Fc, we demonstrate that Fc-engineering of the interaction with FcRn most often negatively modulates binding to the different mouse and human FcγRs. Specifically, mutations in the FcRn binding site (IHH) reduced binding less than mutations in amino acids in close proximity to the FcRn binding site (MN, HN, MST and MST/HN). The MST/HN mutations reduced binding to all Fcγ receptors, and the greatest effect was on binding to the inhibitory FcγRIIb, as well as the activating receptors, FcγRIIIa and FcγRIIIb. As the IgG variant is not designed for target cell killing, this may not prevent clinical use. MN reduced binding to all receptors, except FcγRIIb, and in particular, the binding to FcγRIIIa of the F158 allotype was affected. This may well reduce its ability to direct efficient target cell killing. Likewise, HN showed decreased binding to all receptors, and again, binding to the FcγRIIIa of the F158 allotype was affected. Importantly, we demonstrate that such Fc-engineering has less significant effect on binding to the mouse receptors, a feature that must be considered in regard to pre-clinical evaluations of human IgG1 Fc-engineered variants in rodents. 10 Furthermore, we show that IgG1 Fc-mutations of positions that bind zinc ions affect binding to FcRn and the FcγRs with different degree, which may imply a regulatory role of Zn ion binding to the Fc residues. Here, mutation of H268 to alanine was shown to improve binding by 2-fold to the inhibitory FcγRIIb, while the binding to the activating receptor FcγRIIIa of the F158 allotype and also to FcγRIIIb was reduced by 50%. Taken together, our findings show that engineering of the FcRn-IgG interaction may greatly influence binding to…