Optimal metabolic strategies for microbial growth in stationary random environments

Abstract

In order to grow in any given environment, bacteria need to collect information about the medium composition and implement suitable growth strategies by adjusting their regulatory and metabolic degrees of freedom. In the standard sense, optimal strategy selection is achieved when bacteria grow at the fastest rate possible in that medium. While this view of optimality is well suited for cells that have perfect knowledge about their surroundings (e.g. nutrient levels), things are more involved in uncertain or fluctuating conditions, especially when changes occur over timescales comparable to (or faster than) those required to organize a response. Information theory however provides recipes for how cells can choose the optimal growth strategy under uncertainty about the stress levels they will face. Here we analyse the theoretically optimal scenarios for a coarse-grained, experiment-inspired model of bacterial metabolism for growth in a medium described by the (static) probability density of a single variable (the `stress level'). We show that heterogeneity in growth rates consistently emerges as the optimal response when the environment is sufficiently complex and/or when perfect adjustment of metabolic degrees of freedom is not possible (e.g. due to limited resources). In addition, outcomes close to those achievable with unlimited resources are often attained effectively with a modest amount of fine-tuning. In other terms, heterogeneous population structures in complex media may be rather robust with respect to the resources available to probe the environment and adjust reaction rates.

How total mRNA influences cell growth

“…It should be noted that while growth rate optimality is generally a very good guiding principle for growth laws (10; 41; 42), it is also clear to us that on biological grounds, things are more complex (43). Indeed, growth rate is not the sole quantity that allocation of resources such as ribosomes and catabolic enzymes allocation should optimize, and that other quantities should be relevant such as ability to respond (or to hedge their bets) in changing environments, to filter out “false-alarms” due to noise, to resist famine conditions (44). Additionally, mechanistic regulation might be trying to achieve growth-rate optimization heuristically, and sub-optimal behavior may be determined by architectural limits of cellular regulation.…”