Background/Question/Methods Correctly responding to environmental stress is crucial for an organism’s survival and reproduction. Adapting to an unpredictable stress can be problematic, however. The response that maximizes fitness can depend critically on whether the stress turns out to be short or long-term. As a result, the best strategy can be for a single genotype to bet-hedge by responding to stress in two or more ways. The symbiotic bacterium Sinorhizobium meliloti can spend years in the soil, where carbon supply may not meet demand, between legume hosts. When food is available, S. meliloti stores energy in the lipid polyhydroxybutyrate (PHB). Using flow cytometry, we found that genetically-uniform S. meliloti form two distinct phenotypes upon starvation: 35-90% of the population forms ‘growers’, using most of their stored PHB for reproduction (increasing short-term fitness, but putting long-term survival at risk). The remainder form ‘persisters’, using stored PHB to maintain long-term viability instead of to reproduce. In a series of experiments, we used flow cytometry and fluorescence microscopy to a) characterize PHB use by each phenotype during starvation, b) determine the fitness of each phenotype during long-term starvation, and c) understand how a genetically uniform population can form a stable mix of persisters and growers.

Results/Conclusions Persisters and growers use PHB for different purposes. At the beginning of starvation, all cells contained PHB in excess of 50% cell dry weight. After 18 hours of starvation, all cells in the population had consumed ~50% of this PHB to fuel a population doubling. Yet for the rest of this experiment (114h), persisters actually accumulated PHB (r²=.87, n=18) gaining 55% more, while growers used virtually all remaining PHB for reproduction. After 16 months of starvation, 56% of persisters remained viable, while only 10% of growers were. To elucidate the behaviors of individual cells that result in a stable mix of persisters and growers, we followed the fate of single cells during starvation. When dividing, PHB is not reallocated evenly between parent and daughter-cells. After 24h of starvation, 93.3% (n=15) of daughter cells contained little PHB while parent cells possessed large PHB granules. Collectively, our results describe a behavior by S. meliloti that allows individual cells to bet-hedge, maximizing fitness in an environment where starvation is episodic and of unpredictable duration.