These data suggest that glucose exhaustion itself is not a trigger of the colR mutant lysis; rather, this mutant cannot respond adequately to a certain glucose concentration range which finally causes Metformin cell death. This scenario also allows to explain the absence of the lysis phenotype in liquid glucose medium. Obviously, the period of nutrient limitation is transient in liquid batch culture and could have

been easily missed in our experiments. Literature data suggest that bacteria growing under suboptimal levels of nutrient, i.e. under conditions between the feast and the famine, express cellular responses that are significantly different from those of rapid growth and starvation [3, 48]. Under conditions of hunger when a nutrient becomes limiting but is not yet depleted, bacteria increase permeability of the membrane to facilitate nutrient entry. For instance, a significantly increased MDV3100 mw expression of the OprF porin and the LamB-Mgl high-affinity glucose uptake system is considered to be the hunger response of E. coli under glucose limitation [5]. Analogously, we detected essential nutrient concentration-dependent changes in the OM protein composition of the glucose-grown P. putida. We found that the abundance of the sugar channel OprB1 was significantly increased and that of OprE was drastically decreased under low glucose concentrations (Figure 6A). Interestingly, in addition

D-malate dehydrogenase to being modulated by glucose, the abundance of OprE also responds to anaerobiosis [54, 55] suggesting that this outer membrane channel contributes to the adaptation to various environmental conditions. OprB1 is known to mediate high-affinity glucose transport both in P. putida and P. aeruginosa [25, 56, 57]. While OprB1 is not essential for the glucose transport at higher substrate concentrations, it becomes rate-limiting in nutrient uptake at micromolar (1-10 μM) glucose concentrations [25, 57]. Therefore, the up-regulation of OprB1 at low glucose concentrations can be considered an adaptive response of hungry bacteria to stimulate glucose acquisition. However, our results show that the spatiotemporal expression of OprB1 generates

spatiotemporal lethal toxicity for colR-deficient bacteria, which implies that the ColRS two-component system is an essential regulator of the hunger response of the glucose-growing P. putida. Our data demonstrate that the up-regulation of OprB1 in response to hunger is controlled post-transcriptionally and that catabolite repression control (CRC) protein Crc is one of the factors involved in this regulation (Figure 7). CRC is an important global control system in bacteria allowing hierarchical assimilation of substrates under simultaneous presence of several possible carbon sources. Interestingly, while in many bacteria glucose is a preferred carbon source, Pseudomonas prefers organic acids and amino acids to glucose [51, 58].