Enhanced efficacy of TRAIL or TRAIL-CD28 chimera armored anti-BCMA CAR T cells in multiple myeloma

Abstract

Abstract Background: While BCMA-targeted CAR-T cell therapies have shown promising response rates in multiple myeloma (MM), most patients ultimately relapse. Tumoral intrinsic and extrinsic factors contribute to CAR T cell resistance. Amongst CAR T cell-intrinsic limitations, poor expansion and persistence (typically 3–6 months in MM), exhaustion, and loss of early memory phenotypes, are major drivers of treatment failure. Our single-cell transcriptomic profiling of T cells from MM patients with durable remissions post anti-BCMA CAR T therapy identified high expression of the death receptor ligand TNFSF10 (encoding for TRAIL) as well as the T cell costimulatory receptor CD28 in their CAR T cells. Previous report also identified that FADD (FAS associated death domain) or TNFRSF10B (encoding DR5 also know TRAIL-R2) knock-out abrogated anti-CD19 CAR T cell efficacy. Given these observation, we first sought to explorewhether armoring anti-BCMA CAR-T cells with TRAIL or a chimeric TRAIL-CD28 fusion will enhance their cytolytic function, and second to examine whether the TRAIL-CD28 chimera will attenuate any potential fratricide effect induced by TRAIL overexpression. Methods: We engineered third-generation anti-BCMA CAR constructs (scFv-4-1BB-CD3ζ) into the pLX307 lentiviral backbone, with or without co-expression of TRAIL or a chimeric TRAIL-CD28 molecule (TRAIL extracellular/transmembrane domain fused to CD28 intracellular domain). Functional assays were performed in MM cell lines, including teclistamab-resistant and NFκB-activated (TRAF3 CRISPR knockout) MM cell lines that were in vitro derived to acquire BCMA-antigen independent relative resistance to TCE and CAR T cells. Co-cultures were performed under standard and exhaustion-inducing conditions (e.g., high tumor burden; E:T = 1:5 and repeated stimulation). Cytotoxicity, CAR-T cells phenotype, and viability were assessed using flow cytometry. Co-cultures with healthy donor PBMCs and normal bone marrow cells were conducted to evaluate off-tumor toxicity. In vivo functional evaluations of these TRAIL engineered CAR were performed in NSG mice that were systemically injected OPM2 TRAF3KO cells (stably transduced with firefly luciferase). Results: We first evaluated the of role of the death receptors axis and granzymes in mediating anti-BCMA CAR T cells cytolytic activity. To this extent KMS12BM myeloma cell lines were engineered to KO FADD or TRAIL-R2 (DR5) or overexpress Cowpox virus protein CrmA to inhibit caspase 8 downstream of death receptors or overexpress the granzyme B inhibitor Serpin B9. Co-culture studies of these cell lines with anti-BCMA CAR T unarmored or TRAIL-armored cells confirmed the significant contribution of TRAIL to CAR T cell activity since death receptor blockade (in FADD- or TRAIL-R2 KO cells) and similar to granzyme B inhibition, significantly attenuated MM cell death. Consistent with these observations, TRAIL-armored anti-BCMA CAR-T cells exhibited enhanced cytotoxicity against a library of MM cell lines including cells with lower TRAIL-R1/TRAIL-R2 expression. This enhanced activity of the TRAIL armored CAR was also particularly evident under stress conditions such as chronic stimulation and low E:T ratios. This effect extended to teclistamab-resistant and NFκB-activated MM cell lines models where unarmored CAR-T cells showed limited efficacy. Importantly, no cytotoxicity of the TRAIL-armored CAR was observed against healthy PBMCs or bone marrow–resident cells, consistent with their low or absent TRAIL-R1/TRAIL-R2 expression. As anticipated, TRAIL overexpression induced mild CAR T cells fratricide (< 10%) however only at high CAR-T cell culture densities. This fratricide effect was substantially abrogated in anti-BCMA CARs transduced with the TRAIL-CD28 chimera, which also provided CD28-mediated pro-survival signals. TRAIL-CD28–armored CAR-T cells maintained robust BCLxL expression and a favorable phenotypes with enriched naïve and central memory cells. Lastly, the in vivo NSG mice studies, demonstrated that both TRAIL- and TRAIL-CD28–armored CAR-T cells induced superior tumor clearance compared to unarmored CAR-T cells. Conclusion: TRAIL and TRAIL-CD28 armoring significantly enhances anti-BCMA CAR-T cell efficacy, especially in resistant MM models, through death receptor-mediated tumor killing and CD28-driven survival signaling. These constructs demonstrate activity, supporting their development as next-generation CAR-T therapies.