Logo: to the web site of Uppsala University

Contact-Dependent growth inhibition (CDI) was discovered in 2005 in the E. coli isolate EC93. Since then our knowledge of CDI systems and their impact on bacterial communities have increased exponentially. Yet many biological aspects of CDI systems are still unknown and their impact on complex microbial communities have only just begun to be studied. CDI systems require the function of three proteins; CdiBAI. The outer-membrane transport protein, CdiB, allows for the transportation of the toxin delivery protein CdiA to the cell surface of an inhibitor cell. Through a contact- and receptor-dependent interaction with a target cell the toxic C-terminal domain of CdiA is cleaved off and delivered into the target cell were it mediates a growth arrest. Different CdiA-CT domains encodes for diverse toxic activities, such as nucleases and membrane ionophore toxins. Each unique CdiA-CT toxin has a cognate immunity protein (CdiI) that binds and neutralize against its toxic activity, thus preventing a possible self-inhibition.

In this thesis I have studied the effect of CDI system(s) on both single cell and population level, within both intra- and interspecies bacterial communities. The findings presented here shows that multiple class I cdiBAI loci within a cell can function in a synergetic manner and act as versatile interbacterial warfare systems able to inhibit the growth of rival bacteria, even when CdiA expression is low. We also show that class II CdiA receptor-binding domains can mediate broad-range cross-species toxin delivery and growth inhibition, even when a non-optimal target cell receptor is expressed at a low level. Additionally, we show that the cdiA gene supports the expression of two separate proteins. The full-length CdiA protein, able to mediate an extracellular toxin delivery, but also the discrete CdiA-CT toxin domain. This stand-alone CdiA-CT expression was stress-dependent and together with its cognate CdiI immunity protein functioned as a selfish-genetic element. Moreover, we show that CDI systems can increase bacterial stress tolerance via an extracellular toxin delivery between kin-cells. This stress tolerance phenotype only occurred under conditions when we also observed a selective degradation of the CdiI immunity protein. Therefore, we suggest that a selective CdiI degradation allows for a sub-population of cells to self-intoxicate, thereby becoming transiently dormant, which confers an increase in stress tolerance. The findings presented in this thesis collectively suggest that CDI systems could function as a pseudo-quorum sensing system able to mediate behavioral changes and stress tolerance within a sub-population of cells in a bacterial community.