CAR architecture has evolved through four generations. The first relied solely on CD3ζ and lacked potency. Second-generation receptors added either CD28 or 4-1BB costimulation, striking a balance between cytolytic vigor and persistence that remains the clinical standard. Third-generation designs combine two costimulatory domains for stronger activation, whereas fourth-generation "armored" CARs secrete cytokines such as IL-12, IL-15, or IL-18 to remodel the tumor microenvironment. Extracellularly, single-chain variable fragments (scFvs) have been supplemented by nanobodies, shark VNARs, and lamprey VLRs, each influencing spatial conformation and signaling. Intracellularly, domains from ICOS, MyD88, OX40, or CD3ε are being tested to fine-tune activation thresholds.
Solid tumors, however, present physical barriers, antigen heterogeneity, and immunosuppressive milieus that blunt CAR-T efficacy. Dense stroma and abnormal vasculature restrict infiltration; down-regulation of chemokines impedes recruitment; heterogeneous antigen expression enables escape; regulatory T cells, tumor-associated macrophages, and TGFβ drive exhaustion. Persistent antigen exposure combined with constitutive low-level signaling—termed tonic signaling—accelerates T cell dysfunction and limits in vivo persistence. Understanding and modulating this baseline signaling has therefore become a central challenge for next-generation designs.
Tonic signaling is not unique to CARs. Resting T and B lymphocytes sustain low-grade antigen-receptor signals that maintain survival and homeostatic proliferation. In the thymus, pre-TCR and pre-BCR complexes oligomerize spontaneously through electrostatic interactions between positively charged residues; mutations that disrupt these patches abrogate development. Analogously, CARs lacking antigen can cluster via positively charged patches (PCPs) on their scFv surfaces, driving CD3ζ phosphorylation and downstream activation. GD2-specific CARs exhibit strong clustering and rapid exhaustion, whereas CD19-specific CARs display diffuse membrane distribution and sustained function. Quantifying PCP area enables prediction of tonic strength; engineering scFv charge, linker length, hinge composition, or costimulatory domains allows precise tuning.
A bell-shaped "Peak Theory" describes the relationship between tonic strength and therapeutic efficacy. Signals that are too weak curtail proliferation and memory formation; signals that are too strong provoke exhaustion and toxicity. Empirical data support this model: mutating surface charges of GD2 CARs to lower tonic signaling reduces exhaustion and improves tumor control, whereas enhancing tonic signaling of CD19 CARs increases persistence without triggering dysfunction. Similarly, shortening the VH-VL linker or lengthening the hinge to an intermediate size yields optimal signaling amplitude. Even ubiquitination-resistant CARs that recycle back to the membrane prolong signaling duration yet remain within the beneficial range, reinforcing the concept that both intensity and duration must be balanced.
Tonic signaling also governs antigen sensitivity, metabolic reprogramming, and differentiation fate. Higher basal phosphorylation correlates with elevated CD5 expression, heightened responsiveness to low-density antigens, and up-regulation of glycolytic and anabolic pathways. Conversely, strong signals drive cells toward effector memory phenotypes with increased PD-1 and TIM-3 and decreased TCF7 and CCR7. Moderate elevation of tonic signaling therefore offers a lever to counteract tumor heterogeneity while maintaining stem-like self-renewal.
Extension of these principles to alternative cellular platforms—CAR-Treg, CAR-NK, and CAR-macrophage—remains exploratory. Regulatory T cells require tonic signals that preserve suppressive function without inducing conversion to pathogenic phenotypes. Natural killer cells, which naturally lack long-term persistence, might gain proliferative fitness through calibrated tonic activation, reducing the need for repeated dosing. Macrophages, which polarize between tumoricidal M1 and tumor-supportive M2 states, could be biased toward sustained M1 function by manipulating CAR-induced baseline signaling. Each lineage, however, possesses distinct signaling architectures and functional endpoints, demanding tailored optimization strategies grounded in quantitative models.
In summary, tonic signaling functions as a molecular rheostat that determines CAR-T fitness, tumor recognition, and longevity. By decoding the electrostatic and structural determinants of spontaneous receptor clustering, and by mapping the resulting signaling intensities to phenotypic outcomes, researchers can move beyond binary on/off designs toward predictive fine-tuning. The same lever that currently moves blood cancers toward cure is being lengthened, calibrated, and redirected to dislodge solid tumors, quell autoimmunity, reverse fibrosis, and perhaps delay aging itself.
