Document resource
Background The limited durability of responses to immunotherapy is often attributed to the immunosuppressive tumor microenvironment (TME) and tumor heterogeneity. A major challenge in overcoming this resistance is the rational identification of TME-associated targets whose simultaneous modulation can reverse immune suppression and enhance therapeutic efficacy. Critically, the optimal combination of immune targets varies across tumor types, reflecting distinct TME landscapes. To address this challenge, we developed enLIGHTEN™, a discovery platform that integrates machine learning, viral vector engineering, and rapid payload validation to design programmable viral immunotherapies tailored to modulate multiple immune pathways within the TME with the goal of overcoming resistance and improving responses.Methods Using harmonized TCGA RNA sequencing data, enLIGHTEN™ employed linear modeling and machine learning to identify genes linked to clinical response in breast cancer. The replication-defective herpes simplex viral vector Alpha-201 was chosen for its immunostimulatory and dual oncolytic/gene therapy properties. Candidate payloads were predicted in silico and validated in vitro using a proprietary library of Alpha-201 vectors encoding single genes. Best performing combinations were then tested in vivo in indication-specific tumor models.Results Tumor/PBMC co-culture assays demonstrated that enLIGHTEN™-predicted therapeutic payloads induced PBMC-mediated tumor cell killing in an infection- and payload-dependent manner. Correlation analysis revealed that the expansion and activation of CD8+ T cells, natural killer (NK) cells, and dendritic cells were associated with effective tumor cell killing (Lasso regression, Alpha=0.1, R 2=0.79). The combination of Alpha-201 vectors encoding IL-12 and IL-15 elicited the strongest immunostimulatory activity and was selected for in vivo evaluation. In EMT6 tumor-bearing mice, IL-12/IL-15 combination therapy significantly suppressed tumor growth (60.0% ± 12.6 vs. vehicle, p=0.002) and showed evidence of positive synergistic interaction (coefficients: 0.34; p=0.20). This treatment increased circulating Ki67+ CD8+ T cells (2.7-fold, p=0.001), Ki67+ NK cells (1.6-fold, p=0.0002), and cDC2s (2.4-fold, p=0.0004). RNA sequencing of tumors revealed upregulation of inflammatory response pathways, including IFN- gamma/alpha responses and allograft rejection signatures. Less pronounced anti-tumor and immunostimulatory effects were observed in LLC1 tumor-bearing mice, consistent with in silico predictions of reduced efficacy of the selected payload in lung cancer.Conclusions enLIGHTEN™ enables the rational design of multimodal viral immunotherapies by integrating computational predictions with experimental validation. The designed asset ‘Alpha-201-IL-12/IL-15’ showed strong immune activation and tumor suppression in breast cancer, demonstrating potential of enLIGHTEN™ to guide indication-specific immunotherapy development.