Evaluation of an optimized CD3ζ truncCAR transgene and its impact on CAR-T cell function in vitro. (A) dsDNA templates for targeted delivery of a CAR or truncCAR respectively into TRAC (left) or CD3ζ (middle), as in Figure 1A, and for targeted delivery of a GSG-P2A-linker-modified truncCAR into CD3ζ (right). (B) Top: Mean fluorescence intensity (MFI) determined by flow cytometry at steady state (n = 4 biological replicates in 4-6 technical replicates in 2 independent experiments, data normalized to mean of TRAC for each donor; mixed-effects analysis with Geisser-Greenhouse correction followed by Holm-Šídák multiple comparison test, with individual variances computed for each comparison). Bottom: dynamics of CAR MFI after CAR-stimulation using CD19+ Nalm-6 tumor cells. (n = 3-4 biological replicates in 1-2 technical replicates). (C) Relative cytotoxicity assessed in a 6-hour VITAL assay (similar to Figure 1C, n = 4 biological replicates in 3 technical replicates; two-way ANOVA followed by Holm-Šídák multiple comparison test with a single pooled variance.). (D) Cytokine expression in CAR+ cells in response to control (CD19−) cell or target (CD19+) cell encounter (n = 3 biological replicates). (E-H) CAR-T-cell rechallenge in serial cocultures with Nalm-6 target cells. (E) Top: CAR MFI normalized to TRAC condition at steady state (n = 2 biological replicates in 4 technical replicates; statistics as in B). Bottom: dynamics of CAR MFI after target cell engagement (n = 2-4 biological replicates in 1-2 technical replicates). (F) 6-hour VITAL assay. (n = 3 biological replicates in 3-4 technical replicates; two-way ANOVA followed by Holm-Šídák multiple comparison test with a single pooled variance.). (G) Top: relative expansion of CAR+ T cells. Bottom: CAR+ frequency within T-cell products. (n = 4 biological replicates). (H) Cell surface expression of inhibitory receptors (LAG-3, PD-1, TIM-3; means of n = 4 biological replicates). (I) In vivo CAR-T-cell efficacy tested in Nalm-6 acute lymphoblastic leukemia xenograft mouse model (n = 5-6 mice/group; multiple log-rank tests).