CD64 binding potential does not translate into enhanced therapeutic efficacy for anti-IL-23 antibodies under physiologically relevant conditions.

backgroundInterleukin (IL)-23 neutralizing antibodies are clinically efficacious but differ in the functionality of their Fc portion. Guselkumab (GUS) and ustekinumab (UST) have a wild-type (WT) Fc portion, permitting native binding to Fc gamma receptors (FcγRs), while risankizumab (RZB) lacks FcγR binding due to the L234A to L235A (LALA) mutation. Recently, GUS was reported to neutralize IL-23 more effectively than RZB in vitro, owing to GUS-mediated binding of FcγRI (CD64) on macrophages. However, these findings do not account for the fact that FcγRI (CD64) is competitively occupied by endogenous immunoglobulin (Ig)Gs in vivo, as it is a high-affinity receptor for monomeric IgGs.

methodsTo investigate the contribution of the LALA mutation in the Fc portion to IL-23 neutralization in vivo, we administered anti-IL-23 antibodies bearing either WT or LALA-modified Fc portions to Ig-competent (Il10−/−) and Ig-deficient (Rag2−/−) preclinical colitis mouse models. Building on these in vivo experiments, physiological competition between therapeutic antibodies and endogenous IgGs on FcγRI (CD64) binding was next assessed using in vitro binding assays. Specifically, GUS, RZB, and UST antibodies were incubated with CHOK1 cells expressing FcγRI (CD64) or with activated primary human monocytes (CD64high) in RPMI medium, in the presence or absence of plasma. To understand if IL-23 is exclusively produced by cells that may bind monomeric, therapeutic IgG, single-cell RNA sequencing (scRNAseq) and spatial transcriptomic analysis were used to evaluate the overlap of IL23A and FCGR1A (CD64) transcripts in cells isolated from the skin of patients with psoriasis (PsO) and intestine of patients with inflammatory bowel disease (IBD), respectively.

resultsIn the Il10−/− and Rag2−/− murine models, treatment with anti-mouse IL-23 WT or LALA IgG2a mouse monoclonal antibodies similarly suppressed proinflammatory gene expression and inflammatory cell infiltration. In vitro assays using activated primary human monocytes (FcγRIhigh [CD64high]) showed that GUS, UST, and control human IgG1 bound to FcγRI (CD64), whereas RZB did not. However, the presence of plasma inhibited the binding of all therapeutic antibodies to FcγRI (CD64). We confirmed that endogenous IgGs in plasma saturated inflammatory macrophages FcγRI (CD64). Finally, in patient samples, the IL23A transcript was not exclusive to FCGR1A+ (CD64+) cells.

conclusionsThese data support the notion that IL-23 is efficiently neutralized by anti-IL-23 monoclonal antibodies, independent of Fc-mediated binding to FcγRI (CD64), under physiologically relevant conditions. Additionally, coexpression of IL-23A and FCGR1A was rarely observed in tissues from patients with PsO or IBD.