RESUMO
The integrated stress response (ISR) is a conserved signaling network that detects aberrations and computes cellular responses. Dissecting these computations has been difficult because physical and chemical inducers of stress activate multiple parallel pathways. To overcome this challenge, we engineered a photo-switchable control over the ISR sensor kinase PKR (opto-PKR), enabling virtual, on-target activation. Using light to control opto-PKR dynamics, we traced information flow through the transcriptome and for key downstream ISR effectors. Our analyses revealed a biphasic, proportional transcriptional response with two dynamic modes, transient and gradual, that correspond to adaptive and terminal outcomes. We then constructed an ordinary differential equation (ODE) model of the ISR, which demonstrated the dependence of future stress responses on past stress. Finally, we tested our model using high-throughput light-delivery to map the stress memory landscape. Our results demonstrate that cells encode information in stress levels, durations, and the timing between encounters. A record of this paper's transparent peer review process is included in the supplemental information.