Effects of Cellular Memory and Adaptation Cost on Optimal Survival in Fluctuating Environments
Jain, P.; Jolly, M. K.; George, J. T.
Show abstract
Cells invariably encounter unpredictable changes in their microenvironment and adapt by orchestrating substantial alterations in their molecular states, often resulting in appreciable phenotypic changes. The timescale of molecular adaptation depends on how quickly a cell loses its molecular memory of past environmental encounters through the degradation rate of proteins unfavorable to the current environment. Concurrently, de novo synthesis of favorable biomolecules during adaptation imposes an energetic cost that impacts cellular fitness. Here, by developing a phenomenological model of intracellular processing of environmental signals and associated cell-state switching, we study the dynamical implications of cellular memory and adaptation cost on cellular responses to a changing environment. We find that while increasing cellular memory reduces cell fitness in periodic environments, counterintuitively, increasing adaptation cost can improve growth by minimizing mismatches between the environment and the cell state. Similarly, we observed a variable role of memory capacity and adaptation cost for stochastic correlated environments, with increasing memory and cost improving cellular fitness in negatively but not positively correlated environments. Lastly, we show that cellular memory and population heterogeneity in adaptation cost explained reported experimental observations in melanoma: increased population survival during drug treatment when the population was either enriched for rare cells expressing resistance marker genes or primed with low-dose drug treatment before exposure to high-dose treatment. Overall, this work establishes a foundational model for studying how cellular memory dynamics and adaptation cost drive cellular adaptation under different environmental conditions and explain complex cellular behaviors.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.