CELL INTELLIGENCE · 肿瘤细胞治疗研究

Chimeric antigen receptor (CAR)-T cell therapies targeting CD19 have demonstrated remarkable clinical efficacy in B-cell malignancies. However, substantial differences exist in therapeutic outcomes, partly due to differences in CAR design and surface expression. Current methods for CAR detection, including flow cytometry, do not allow direct quantification of receptor density. Here, we establish a linker-targeted direct stochastic optical reconstruction microscopy ( d STORM) approach for quantitative assessment of CAR surface expression across structurally diverse CD19 CAR-T cell products. By targeting conserved scFv linker regions, including (G4S)3 and Whitlow linkers, we enable antigen-independent detection using commercially available antibodies. We generated primary human T cells expressing constructs resembling approved CD19 CAR-T cell products and compared CAR detection by flow cytometry and d STORM. While flow cytometry enables detection, d STORM demonstrated superior sensitivity, allowing reliable visualization of CAR expression levels insufficient for flow cytometry. Direct CAR detection via linker-targeting antibodies revealed construct-dependent differences in CAR surface density, with a (G4S)3-containing construct exhibiting higher receptor densities compared to two Whitlow-based designs. d STORM resolved these differences more reliably at low expression levels, where flow cytometry yielded more CAR-negative events, suggesting that transfection marker staining and flow-based analyses may not fully capture CAR surface expression. Validation in clinically relevant CAR products underscores the robustness and versatility of this toolbox for CAR-specific staining across constructs. Overall, linker-targeted d STORM represents a highly sensitive and broadly applicable platform for quantification of CAR surface expression, offering new opportunities to understand how CAR design influences receptor density, spatial organization, and therapeutic function.