Isogenic bacteria can exhibit a range of phenotypes, even in homogeneous environmental conditions. Such non-genetic individuality has been observed in a wide range of biological processes, including differentiation and stress response(1). A striking example is the heterogeneous response of bacteria to antibiotics, whereby a small fraction of drug-sensitive bacteria can persist under extensive antibiotic treatments. Recently, a renewed interest in the persistence phenomenon has revealed that non-genetic heterogeneity might be one of the main reasons for the failure of antibiotic treatment in infections such as tuberculosis, where a single persistent bacterium can re-start an infection(2). Persistence is typically observed through the monitoring of the survival fraction of a bacterial population exposed to antibiotics. The initially rapid killing of the bacteria is followed by a significantly reduced killing rate which indicates the presence of a persistent sub-population. When cells grown from this persistent sub-population are subjected again to antibiotics, the same bi-phasic killing curve is obtained, suggesting that the persistent sub-population is not genetically different from the original population. We have previously shown that persistent bacteria enter a phenotypic state, identified by slow growth or dormancy, which protects them from the lethal action of antibiotics(3). Here we studied the effect of persistence on the interaction between Escherichia coli and phage lambda. We focused on two different variations of this well-studied predator-prey system: (a) a phage that is present in the genome of each bacterium and can cause bacterial death by a process called "prophage induction" and (b) a lytic phage that attacks bacteria from the outside, infects them and them kills them. The effect of the persistent phenotype was studied in those systems and the experimental results obtained were then implemented in a mathematical description of these interactions(4). We used long-term time-lapse microscopy to follow the expression of GFP under the phage lytic promoter, as well as cellular fate, in single infected bacteria. We found that dormancy that protects bacteria under antibiotic treatments also protects them against prophage induction. A competition experiment run between a low persistence population and a high persistence one demonstrated a clear advantage to the latter. This suggests that persistence might have evolved under the evolutionary pressure of prophage stress. Intriguingly, we found that, while persistent bacteria are protected from prophage induction, they are not protected from lytic infection. Quantitative analysis of gene expression revealed that the expression of lytic genes is suppressed in persistent bacteria. However, when persistent bacteria switch to normal growth, the infecting phage resumes the process of gene expression, ultimately causing cell lysis. Despite its mild effect on the short-term survival of the population, the delayed cell lysis of persistent bacteria needs to be taken in account. Using a mathematical model for this predator-prey interaction, we found that the bacteria's non-genetic individuality can significantly affect the population dynamics, and might be relevant for understanding the co-evolution of bacterial hosts and phages. References 1. Rando, O. J. & Verstrepen, K. J. (2007) Timescales of genetic and epigenetic inheritance Cell 128, 655-668. 2. Stewart, G. R., Robertson, B. D., & Young, B. D. (2003) Tuberculosis: A problem with persistence Nature Reviews: Microbiology 1, 97-105. 3. Balaban, N. Q., Merrin, J., Chait, R., Kowalik, L., & Leibler, S. (2004) Bacterial persistence as a phenotypic switch Science 305, 1622-1625. 4. Pearl, S., Gabay, C., Kishony, R., Oppenheim, A., & Balaban, N. Q. (2008) Nongenetic individuality in the host-phage interaction PLoS Biol 6, e120.