Dual Graphene Oxide-Based Multigene Delivery for Cancer Elimination via Stromal and Immune Reprogramming

Cancer therapies are often limited by poor cellular specificity and tumor microenvironment-mediated immunosuppression. We developed a dual-function graphene oxide (GO)-based gene delivery platform combining direct cancer cell elimination with stromal and immune reprogramming. CD47-targeting siRNA and ANT1-encoding plasmid DNA selectively induced apoptosis in tumor cells, while TGFβ-targeting siRNA altered cancer-associated fibroblasts (CAFs) phenotype and was associated with a shift in macrophage polarization toward a pro-inflammatory, tumor-restraining phenotype. We evaluated this approach in a 3D multicellular lung cancer model consisting of A549 cells, human lung fibroblasts, and THP1-derived macrophages. TGFβ silencing alone modulated stromal and immune features without direct cytotoxicity, whereas CD47/ANT1 modulation reduced overall cell viability to ∼51%. When both therapeutic functions were applied together, overall cell viability was further reduced to ∼29%. EGFR-targeted GO-CD47-ANT1 (GE11-peptide modified formulation) enhanced tumor-cell uptake and selective apoptosis in CK7⁺ cancer cells to 91%, while sparing CK7^- nonmalignant populations (16%). Mechanistically, combined therapy suppressed immunosuppressive macrophage markers (CD163, IL10, TGFβ), restored epithelial integrity (E-cadherin 3.5-fold), and attenuated CAF activation (αSMA 0.3-fold; vimentin 0.6-fold). GE11 peptide functionalization shifted nanocarrier internalization toward EGFR-mediated cancer cell uptake, minimizing off-target delivery to fibroblasts and macrophages. These findings demonstrate that synergistic modulation of cancer cells and the tumor microenvironment is essential for robust and selective tumor elimination. Our study establishes a modular, mechanism-informed GO platform integrating dual therapeutic functions-targeted cancer cell killing and microenvironmental reprogramming-highlighting its translational potential in physiologically relevant tumor models.