Temporal and Clonal Resolution of Cellular Evolution Under Stress

Dissecting cellular evolution under transient versus persistent stress is essential for understanding both healthy and pathological processes. Yet cellular stress responses involve nonlinear dynamics and survival bottlenecks, making them challenging to study. We develop a framework that combines single-cell multiomic lineage tracing with statistical modeling to quantify how individual cells contribute to the future population, by explicitly modeling exponential expansion and intra-clonal heterogeneity. Applied to a cancer model under short- and long-term treatment, we identify clones primed to endure treatment versus ones that survive by producing diverse progeny, enabling characterization of molecular features enriched in each clonal survival strategy. This framework can quantify treatment-driven selection and adaptation, revealing that their relative contributions to population dynamics vary between treatments. Furthermore, short- and long-term treatment results in expansion of distinct clones characterized by different gene programs and transcription factor activity. Notably, certain pathways such as AP-1 signaling can limit initial cancer cell growth but promote later resistance, and such opposing associations are consistently observed in independent experiments and clinical data. Together, this study introduces a generalizable approach for investigating cellular evolution in a time- and clone-resolved manner, and highlights time-dependent molecular programs that underlie stress responses.