Interference with bacterial sociobiology: Inhibition of common good production as an antimicrobial strategy with a low risk of resistance development.

The rising threat of antibiotic resistance leaves most antibiotics obsolete and once easy-to-treat infectious diseases may become life-threatening. There is thus an urgent need for new antimicrobials and more importantly, for antimicrobials for which resistance is counterselected. A possible solution to the problem has been proposed, where the production of public goods of bacteria are inhibited. Public goods are group-beneficial traits that are produced by one cell but also benefit surrounding cells, e.g. secretion of nutrient-releasing enzymes or iron-scavenging siderophores. Social evolution theory predicts that inhibiting public goods inhibits bacterial growth and survival, but that strains resistant to such inhibitor will be counterselected. The rationale is that a resistant strain will produce the public good and pay a cost to do so, while susceptible strains will be able to use the public good without paying the cost. Theoretically then, we should be able to design drugs where resistance is counterselected. However, this has never been put in practice.

Next to the rapid resistance development, bacteria are able to form biofilms, which are structured communities of cells embedded in a self-produced matrix. Important features of biofilms are their increased tolerance against antimicrobials and environmental stresses and their ability to attach to all sorts of surfaces. Biofilms are the predominant life style of bacteria and it is suggested that social interactions, such as public good production, are enhanced within a biofilm. In this PhD thesis, we focused on the exopolymeric substances (EPS) of Salmonella Typhimurium as a public good.

In the first part of this thesis, the public good character of the EPS of S. Typhimurium was demonstrated using an in vitro biofilm model in petridishes. This set-up is representative for a broad range of environmental and industrial environments and provides a first approximation for the in vivo situation in the gut. Our study revealed that (i) EPS increases biofilm formation and antimicrobial tolerance, (ii) EPS is costly to producing cells and that (iii) EPS made by one cell benefits other cells. Moreover, it was demonstrated that resistance does not evolve during a long-term evolution experiment in the presence of an in house developed EPS inhibitor. In addition, it was shown that a strain, resistant to an EPS inhibitor, is outcompeted by an inhibitor-susceptible strain in the presence of that inhibitor. In a next step, this antimicrobial strategy was extended towards another industrial application and towards an application in the clinical practice. Using a flow cell, we leaned towards an industrial

application where biofilms are formed inside pipelines, whereas with the in vivo Salmonella Typhimurium mouse model we shifted towards a clinical application for the treatment of Salmonella infections. It was demonstrated that EPS production is crucial both for colonization of the surface inside the flow cell as during infection of mice. However, using these models, we were not able to demonstrate that EPS is exploitable under the applied conditions.

In the second part, a single gene-deletion mutant library in Salmonella Typhimurium was screened for novel interesting targets for public good inhibition. Mutants were screened for their capability to form biofilms and their ability to exploit the wild type producing strain. Using this assay, 7 mutants were found that have a reduced biofilm fitness and that are able to overgrow and outcompete the producing wild type strain. In a next step, we excluded these mutants that exert their biofilm defect through EPS deficiency. Overall, two possible targets for public good inhibitors were identified, that may have a lowered chance of resistance evolution.

In conclusion, in this thesis it was successfully demonstrated that interference with EPS production is an effective antimicrobial strategy in both industrial and clinical settings. Moreover, using an in vitro biofilm model in petridishes, it was demonstrated that EPS inhibitors have the capability to select against resistance.