(530d) A Tumor-Targeted Cytokine/Antibody Fusion to Stimulate Anti-Cancer Immunity

Significance. Interleukin-2 (IL-2) is a pleiotropic cytokine that plays a critical role in T cell function and has been clinically useful in promoting anti-cancer immunity. However, the clinical efficacy of IL-2 therapy is limited by both the short serum half-life and systemic toxicity of IL-2, along with the cytokine’s concurrent activation of both immunostimulatory effector cells (i.e., natural killer and T effector cells) and immunosuppressive regulatory T (T_Reg) cells. IL-2 signals through an intermediate-affinity (K_D~100 pM) ternary complex that consists of the IL-2 receptor-Β (IL-2Rβ) subunit and common gamma chain (γ_c) or a high-affinity (K_D~1 nM) quaternary complex that also includes the IL-2Rα subunit. T_Reg cells express high levels of the IL-2Rα subunit, making them much more responsive to IL-2 compared to naïve effector cells. An anti-human IL-2 (hIL-2) antibody known as 602 was found to bias IL-2 activity to stimulate effector over T_Reg cells by simultaneously blocking the IL-2/IL-2Rα interaction and allosterically enhancing the IL-2/IL-2Rβ interaction. To extend the serum half-life and selectively bias the immunostimulatory functions of IL-2 in a stable unimolecular format, we created a single-chain fusion protein, denoted 602 immunocytokine (602 IC), that intramolecularly fuses human IL-2 to the 602 antibody. A variant of the 602 IC, denoted F10 IC, was evolved to have superior Eff biasing capability using yeast surface display. To enhance translational relevance, we generated a chimeric F10 IC with human constant domains. We hypothesized that targeting the effector cell-biased activity of chimeric F10 IC to the tumor microenvironment would result in development of a potent anti-cancer therapeutic. To this end, we fused chimeric F10 IC to collagen-targeting proteins, so that they would specifically home to and be retained in the tumor, where collagen is abundant and readily exposed due to the leaky vasculature.

Methods. Fusion proteins between chimeric F10 IC and a collagen binding domain (CBD) were produced in a mammalian cell expression system and purified via protein G affinity chromatography followed by size exclusion chromatography. The tumor-targeted recombinant chimeric ICs were then biophysically characterized via bio-layer interferometry using an Octet® instrument. To quantify IL-2 signaling induced by our engineered chimeric F10 IC-CBD fusion proteins, phosphorylation of signal transducer and activator of transcription 5 (STAT5) was measured on IL-2Rα^- and IL-2Rα⁺ YT-1 human natural killer cells as surrogates for stimulation of immune effector and T_Reg cells, respectively. Affinity for collagen was determined by enzyme-linked immunoassay (ELISA).

Results. Bio-layer interferometry studies confirmed that chimeric F10 IC and tumor-targeted F10 IC-CBD fusion protein chimeras do not bind the IL-2Rα subunit and enhance interaction with the IL-2Rβ subunit. YT-1 cell signaling studies demonstrated these constructs maintain biased activation towards effector cells, eliminating the T_Reg cell advantage typically conferred by IL-2-Rα. Collagen-binding ELISA assays demonstrated that the tumor-targeted ICs specifically bind collagen.

Conclusions. Our work illustrates feasibility of our approach to engineer, for the first time, tumor-targeted ­­ICs with biased IL-2 activity. Moreover, we demonstrate that collagen binding is intact and that the ICs maintain effector cell-biased immune activity. Building on these promising results, we are now assessing the in vivo activity and therapeutic efficacy of these chimeric fusion proteins in various mouse tumor models to illustrate the potential of tumor-targeted chimeric F10 IC as a powerful new anti-cancer drug.