Evolutionary Origins and Physiological Architecture of Cellular Adaptability

Cells in nature face many simultaneous environmental challenges, and their responses to combined stresses rarely add up the way responses to single stresses would predict. This dissertation argues that cellular adaptability is structured, multiscale, and shaped by evolutionary history — an architecture cells inherit rather than a property they possess. I develop this argument across three scales of biological organization in Saccharomyces cerevisiae, presented here in the order the dissertation takes them up. At the cellular level, I show that responses scale coherently across the whole transcriptome. Single-cell RNA sequencing across 20 environments revealed that environmental information is hierarchically ordered across the principal components of transcriptional variation: carbon source first, then media, osmolarity, temperature, and reactive oxygen species. Under combined stress, cells mount the higher-ranked program and suppress the lower one. The molecular router for this prioritization is differential collapse of polysomes, which gates the heat shock response's condensate cascade and determines which response can deploy. At the evolutionary level, more than 3,000 generations of selection under constant osmotic stress collapsed this hierarchy. Carbon, media, and ROS information compressed into the leading components, while osmolarity — the selected axis — was demoted to deeper ones. Sustained selection does not amplify the representation of an environmental variable but converts it from a regulated input into a constitutive condition, trading broad adaptability for specialized fitness. At the molecular level, I then ask what lies beneath the translation-initiation node that does this prioritizing, recasting the heat shock response (HSR) as a convergent stress cascade. A single biophysical pathway — orphan ribosomal protein condensation, chaperone redistribution, and Hsf1 transcriptional condensate formation — integrates inputs from heat, ethanol, and respiratory stress, with stressor identity encoded both at the input, where metabolic state tunes upstream signals, and at the output, where stressor-specific feedback modules tailor the response. Together, these three scales show that adaptability is not a fixed capacity of cells but rather, it is an architecture built by selective history.