Macrophage-Engaging IgG4 Antibody Triggers Cytotoxicity Against Integrin αvβ3+ Cancers

Abstract

Integrin αvβ3, absent in most normal cells, has emerged as both a marker and a driver of tumor stemness and drug resistance in epithelial cancers, making it an attractive therapeutic target. The humanized IgG1 anti-αvβ3 antibody, etaracizumab, was originally developed to exploit NK cell-mediated cytotoxicity against αvβ3-positive tumors. However, despite its favorable safety profile and clinical efficacy, its impact was insufficient for further development. We previously discovered that αvβ3-positive epithelial tumors exhibit a tumor-associated macrophage (TAM)-rich microenvironment with limited NK cell infiltration, potentially limiting the effectiveness of etaracizumab. Here, we hypothesized that re-engineering the anti-αvβ3 antibody to activate TAM-mediated cytotoxicity would enhance its anti-tumor activity. We developed a fully human IgG4 variant of etaracizumab (anti-αvβ3 G4) with identical affinity for integrin αvβ3, but optimized for activation of CD64, an immune effector cell-activating receptor, selectively expressed on macrophages. In organotypic cultures of lung cancer patients and mouse xenografts, anti-αvβ3 G4 demonstrated superior anti-tumor activity compared to its IgG1 counterpart. Mechanistically, this enhancement was driven by CD64 activation in TAMs, leading to robust upregulation of inducible nitric oxide synthase (iNOS), a pivotal enzyme for immune effector-mediated cytotoxicity. Our findings reveal a powerful strategy for targeting highly aggressive, drug-resistant integrin αvβ3-positive tumors by harnessing TAMs for antibody-mediated cancer therapy, and demonstrate that this Fc switch approach may be broadly applicable to other targets in TAM-enriched tumor microenvironments.

Figure 1.. Anti-αvβ3 G4 exhibited greater TAM-driven anti-tumor activity than anti-αvβ3 G1

A. Cancer tissue slices were incubated with the indicated antibodies (10 μg/mL) for 5 days (N = 3). Tissues were stained for integrin β3, CD68, or cleaved-caspase 3. % β3+, CD68+, or cleaved-caspase 3+ cells were measured. *P<0.05 using One-way ANOVA. Representative images are shown. Each dot represents a patient. Patient demographics are shown (upper left). B. Indicated target cells (luciferase+) were incubated with PBMC-derived macrophages or NK-92 cells with control IgG4 or anti-αvβ3 G4 (1 μg/mL) for 24 hours. At the end of the incubation, cells were subjected to Steady-Glo (luciferin). % antibody-dependent target cell death was calculated measuring luminescence and by normalizing to controls. *P<0.05 using student’s t-test comparing to the control IgG4. Error bars indicate standard deviation. Each dot corresponds to a biological replicate.

Figure 2.. Anti-αvβ3 G4 induced macrophage-mediated ADCC via CD64 activation

A. Cancer tissue slices from 3 patients were incubated with the indicated antibodies (10 μg/mL) for 5 days. Tissues were stained for cleaved-caspase 3, inducible nitric oxide synthase (iNOS), or a macrophage marker CD68. % cleaved-caspase 3+, iNOS+, or CD68+ cells were measured. Representative images are shown. *P<0.05 using One-way ANOVA. Each dot represents a patient. B. H1975 cells were incubated with PBMC derived macrophages (stained with CellTrace^™ Far Red) with or without anti-αvβ3 G4 (1 μg/mL) for 24 hours. Macrophages were isolated, and RNA was isolated. qPCR was performed to measure CD86, IL-1B, IL-6, NOS2, TNF, MRC1, and TGM2 levels. The error bar indicates standard deviation. Each dot represents a biological replicate. *P<0.05 using student’s t-test. C. H1975 cells were incubated with PBMC derived macrophages with human IgG4 or anti-αvβ3 G4 (1 μg/mL) with and without anti-CD16, -CD32, or -CD64 antibody (1 μg/mL) for 24 hours. % antibody-dependent target cell death was measured by reading luminescence and normalizing to controls. *P<0.05 using One-way ANOVA. Error bars indicate standard deviation. Each dot corresponds to a biological replicate. D. The FcγR activation by anti-αvβ3 G4, G1, and G1 with I332E mutation with the target cells (H1975) was measured. Error bars indicate standard deviations from four technical replicates. *P<0.05 using One-way ANOVA.

Figure 3.. Anti-αvβ3 G4 exhibited greater macrophage-mediated anti-tumor activity than anti-αvβ3 G1 in vivo

A. Nude mice bearing pancreatic (FG+β3, ectopic αvβ3+) or lung (H1975, endogenous αvβ3+) xenografts were treated with vehicle (FG+β3 n = 10, H1975 n = 9), anti-αvβ3 G1 (FG+β3 n = 8, H1975 n = 10, 10 mg/kg, twice/week), or anti-αvβ3 G4 (FG+β3 n = 6, H1975 n = 10, 10 mg/kg, twice/week). Tumor volumes are shown. Error bars indicate standard errors. *P<0.05 using Two-way ANOVA. B. Immunohistochemistry staining for cleaved-caspase 3 in pancreas and lung xenograft tumors was performed. % cleaved-caspase 3+ cells were measured for each animal. *P<0.05 using One-way ANOVA. Error bars indicate standard deviations. Each dot corresponds to each animal. Representative images are shown. C. Once the tumor volume reached 100 mm³, control (CTRL) or clodronate (CLOD) liposome administration was started (twice/week). 3 days after the initial liposome administration, vehicle or anti-αvβ3 G4 administration (10 mg/kg, twice/week) was started. Tumor volume is shown. n = 6-7/group. Error bars indicate standard errors. *P<0.05 using Two-way ANOVA. D. Once the tumor volume reached 100 mm³, PBS or anti-asialo GM1 (AAGM, twice/week) administration was started (twice/week). 3 days after the initial administration, vehicle or anti-αvβ3 G4 administration (10 mg/kg, twice/week) was started. Tumor volume is shown. n = 3-9/group. Error bars indicate standard errors. *P<0.05 using Two-way ANOVA.