Hassan, J.; Reda, O.; Irie, N.; Pedersen, M.; Foo, S.; Appleton, L.; Okazaki, I.-m.; Okazaki, T.; Satou, Y.; Harrington, K.; Melcher, A.; Ono, M.

In cancer immunotherapy, T-cell functional states change dynamically, and no single marker or gene fully defines them. Static molecular profiles may therefore be insufficient to resolve these states. Capturing signalling history together with ongoing activity could provide a temporal framework for their dissection. Here, we developed CanonicalTockySeq, which integrates a molecular clock of T-cell receptor (TCR) signalling based on the Fluorescent Timer reporter Nr4a3-Tocky (Timer-of-cell-kinetics-and-activity) with scRNA-seq to establish an experimentally anchored temporal reference. Using landmark Tocky fractions as biological ground truths, CanonicalTockySeq constructs a transcriptomic manifold in canonical space with conical geometry. In a murine melanoma model treated with combination anti-PD-L1 and anti-CTLA-4, this framework separates temporal progression (geodesic angle) from signalling strength (radial intensity), enabling time-resolved analysis of gene-expression programmes at single-cell resolution. Application to scRNA-seq data from patients with melanoma further identifies distinct temporal programmes associated with clinical response. These analyses indicate that effective combination immunotherapy is associated with decreased persistence of antigen engagement, suppression of exhaustion-associated TCR signalling programmes, and maintenance of progenitor-like features linked to durable antitumour responses. Collectively, our findings identify temporal-state control as a key component of immunotherapy outcome and establish CanonicalTockySeq as a framework for resolving T-cell response states in vivo.