Combining T-cell–specific activation and in vivo gene delivery through CD3-targeted lentiviral vectors

Purpose: Conventional lentiviral vectors (LVs) are not selective for T cells and cannot efficiently transduce resting or minimally stimulated T lymphocytes, which limits their use for direct in vivo T-cell gene delivery. There is a need for vectors that can both target T cells specifically and activate them sufficiently to permit gene transfer without prior ex vivo activation. ---

Hypothesis: Displaying agonistic CD3-specific single-chain variable fragments (scFvs) on Nipah virus (NiV)-based lentiviral vector particles will enable T-cell–specific activation and gene delivery. Such CD3-targeted LVs (CD3-LVs) should transduce nonactivated T cells, function in whole blood, and generate functional CAR T cells directly in vivo. ---

Aims: - Generate CD3-targeted LVs displaying agonistic anti-CD3 scFvs (TR66, TR66.opt, HuM291). - Assess selectivity and gene transfer into activated and nonactivated human T cells. - Evaluate transduction in whole human blood without additional stimuli. - Test in vivo gene delivery and CAR T-cell generation in humanized NSG mice. ---

Delivery system: Vector platform: NiV-based receptor-targeted lentiviral vectors (RT-LVs) pseudotyped with engineered Nipah virus glycoprotein (G_{mut}). Targeting ligand: Agonistic CD3-specific scFvs: - TR66 - TR66.opt (seven (V_L) residues converted to IgG1 consensus) - HuM291 The scFvs were C-terminally fused to (G_{mut}) and incorporated into LV particles. Payloads: - GFP reporter gene - Second-generation CD19-specific CAR with CD28 costimulatory domain and extracellular myc tag Production: LVs produced in HEK-293T or LentiX-293T cells via polyethyleneimine transfection. Entry mechanism: NiV glycoprotein-mediated fusion at neutral pH, not endocytosis. ---

Approach: In vitro models: - Jurkat (CD3(^+)) and Molt4.8 (CD3(^-)) cells - HUVEC, HepG2, Nalm-6 cells for off-target assessment - Primary human PBMCs from healthy donors - Whole human blood Activation/stimulation conditions: - αCD3/αCD28 antibody-activated PBMCs - IL-2–stimulated PBMCs - IL-7/IL-15–stimulated PBMCs - Unstimulated PBMCs - Whole blood without additional stimuli In vivo models: - NSG mice transplanted intraperitoneally with αCD3/αCD28-activated human PBMCs, then injected with TR66-LV or CD8-LV; analyzed 7 days later. - Humanized NSG mice transplanted with human CD34(^+) stem/progenitor cells; groups received IL-7 pre-treatment or PBS, then intravenous TR66.opt-LV or vehicle. CAR T cells and CD19(^+) B cells monitored in blood. ---

Key methods: - Flow cytometry for GFP, CAR, CD3, CD4, CD8, CD45, CD19, CD69, CD25, and exhaustion markers. - Vector copy number (VCN) by qPCR (WPRE/human albumin). - Cytokine quantification (IFN-γ, TNF-α, IL-2) from supernatants. - Nanoparticle tracking analysis or p24 ELISA for vector titers. - Immunoblot for glycoprotein incorporation. - Endocytosis inhibition using Bafilomycin A1 or NH(_4)Cl. - Statistical analysis with GraphPad Prism. ---

Key results: - Selectivity: CD3-LVs transduced CD3(^+) Jurkat cells but not CD3(^-) Molt4.8 cells. VSV-LV transduced HUVEC (86.74%), HepG2 (32.53%), and Nalm-6 (97.50%) cells, whereas CD3-LVs showed <0.3% transduction in these off-target cells. - Nonactivated T cells: CD3-LVs transduced freshly isolated PBMCs in the absence of cytokines or with IL-2 or IL-7/IL-15 at efficiencies similar to peak activation. VSV-LV was substantially less efficient under these conditions. - Whole blood: CD3-LVs mediated T-cell gene transfer in whole blood without additional stimuli. VCNs: TR66-LV 1.17 ± 0.57; TR66.opt-LV 1.23 ± 0.05; HuM291-LV 1.03 ± 0.87; VSV-LV 0.16 ± 0.28; VSV-LV + αCD3/αCD28 0.47 ± 0.59. - In vivo selectivity: In PBMC-NSG mice, ~14% of human CD3(^+) peritoneal cells were GFP(^+) with TR66-LV; <0.5% CD45(^-) mouse cells and 0.1% CD45(^+)CD19(^+) cells were GFP(^+). - In vivo CAR generation: CAR(^+) T cells were detected in blood of most vector-treated animals. With IL-7 + TR66.opt-LV, 5/6 mice had detectable CAR T cells; without IL-7, 3/5. CAR expression peaked at weeks 3–4 and declined by week 5. CD19(^+) B cells dropped concomitantly; by day 26, B cells were almost completely eliminated in 5/6 IL-7 + vector mice and 4/6 vector-only mice. Depletion was durable only in IL-7–pretreated mice. - Endocytosis inhibition: Bafilomycin A1 increased CD3-LV transduction up to sixfold for HuM291-LV and threefold for TR66-LVs in Jurkat cells, consistent with a neutral-pH fusion entry mechanism. ---

Interpretation: The authors conclude that CD3-LVs combine T-cell–specific activation with targeted gene delivery, enabling efficient transduction of nonactivated T cells, function in whole blood, and direct in vivo generation of functional CAR T cells. This approach may simplify CAR T-cell manufacturing, avoid off-target CAR gene transfer, and support direct in vivo T-cell engineering. ---

Limitations: - In vivo validation was performed in humanized NSG mice, not large immunocompetent animal models. - Safety, biodistribution, and efficacy studies in larger animals are still required. - GMP production for this novel vector type remains to be established. - Follow-up was relatively short (up to ~6 weeks). - Activation without costimulation can in principle lead to activation-induced cell death, although viability was not impaired in this setting. - IL-7 pre-treatment improved CAR T-cell persistence and B-cell depletion, but the requirement for IL-7 in clinical settings is uncertain. - CD3/TCR downmodulation differed by scFv and may affect repeated dosing or T-cell function.