Fig. 5: The L32P substitution enhances antitumor efficacy of 28z-based CAR T cells in vivo by increasing memory formation.

a, CD69 expression levels in mock 1928z as compared to L32P-modified 1928z CAR T cells determined by flow cytometry without further antigen stimulation (n = 11 biologically independent donors, mean ± s.d.; two-sided paired Student’s t-test). b, n-fold CAR T cell expansion for L32P-edited relative to unmodified 1928z CAR T cells 7 days after one (day 7), two (day 14) and three (day 21) stimulation(s) with Nalm6 cells (two-sided one-sample t-test; n = 5 biologically independent donors, mean ± s.d.). c, Fold enrichment of the central memory (CD62L⁺CD45RA⁻) population in L32P-edited 1928z relative to mock 1928z CAR T cells within the CD8⁺ (left) and CD4⁺ (right) CAR⁺ compartment. Statistical analysis was performed using a two-sided one-sample Wilcoxon test (n = 9 biologically independent donors). d–f, Luciferase-expressing Sh-Sy5y cells were i.v. injected into NSG mice followed by i.v. administration of UT T cells or B7H3-28z CAR T cells with or without L32P substitution. d, Experimental timeline. e, Bioluminescence imaging of Sh-Sy5y-bearing mice (representative of f) treated with UT, mock B7H3-28z and L32P-edited B7H3-28z CAR T cells at the indicated time points (ventral view). f, Survival of the cohort (L32P B7H3-28z: n = 9, mock B7H3-28z: n = 9 mice, UT: n = 4 mice per group; T cells from two biologically unrelated healthy donors; two-sided Mantel–Cox test). g, Data from the liver of mice (n = 5) treated as shown in (d) and collected 10 days after injection of either mock B7H3-28z or L32P-modified B7H3-28z CAR T cells. Absolute count of SCM CD4⁺ and CD8⁺ CAR T cells in spleen and blood (n = 5 mice per group, mean ± s.d.; two-sided unpaired Student’s t-test).