Het genetisch landschap verantwoordelijk voor persistentie in Escherichia coli

Phenotypic switching has facilitated bacterial survival and colonization of diverse environments. In antibiotic-rich conditions, bacteria can survive treatment by adopting a temporarily antibiotic-tolerant state. When only a subfraction of an isogenic bacterial population is transiently antibiotic tolerant, this phenomenon is referred to as persistence. Persistence is often called a dormant state, as it is associated with reduced metabolic activity, decreased ATP levels and minimized or no bacterial growth. During favourable conditions, persister cells can recover and resume growth, thereby establishing a new bacterial population. Persistence has been associated with chronic infections and facilitates the emergence of antibiotic resistance. Given its clinical importance, fundamental insights in the genes and mechanisms underlying persistence are required for the development of anti-persister therapies. In this research, we studied a promising regulator of persistence that had previously been found in our lab, and we searched for new genetic determinants involved in persistence by employing a CRISPRi screen. In a first part of this work, we assessed a potential role for YbiB in persistence. YbiB had previously been found to interact with ObgE, which is a well-known regulator of persistence. We discovered that overexpression of ybiB significantly increases persister levels, independently of ObgE. The mechanisms by which YbiB influences persistence remain unknown, but acceleration of aggregate formation may play a role. In the second part of the work, we aimed to identify novel genetic determinants of persistence. While persister levels result from a dynamic interplay between persister formation and recovery, limitations in established genetic screens and experimental setups have up until now prevented a detailed examination of both processes individually. The latter heavily relies on the availability of a genome-wide gene-silencing technique that can be activated at a time of interest. In the specific case of persistence, activation prior to antibiotic treatment allows to study the effect of gene silencing on persister formation, while activation after treatment allows to investigate persister recovery and regrowth. Ideally, this technique should allow to study both essential and non-essential genes, since both have previously been implicated in persistence. To meet these requirements, we made use of CRISPRi, which is a gene-silencing approach that can be adjusted to function as an inducible technique. CRISPRi employs a gRNA, which serves as a complementary sequence to the gene of interest, and a dCas9 enzyme, which is a catalytically inactive Cas9 enzyme. The gRNA directs dCas9 to the gene of interest, where dCas9 blocks transcription initiation or elongation through steric hindrance. We developed an inducible CRISPRi approach by controlling the expression of dcas9 by an optimized Ptet* promoter, while the gRNAs are continuously expressed. Additionally, we developed a CRISPRi protocol suitable for identifying both essential and non-essential genes involved in either persister formation or recovery. Using the inducible CRISPRi approach, we screened for genes involved in persister formation and recovery upon amikacin or ciprofloxacin exposure. We discovered that reducing metabolic activity is not sufficient to survive antibiotic treatment. Moreover, specific metabolic pathways need to be downregulated, and the particular pathways involved depend on the antibiotic class used. Furthermore, we distinguished additional antibiotic-specific pathways, including import inhibition for amikacin treatment and DNA repair for ciprofloxacin exposure. Genes involved in DNA repair could be identified during both persister formation and recovery screens. This indicates that persister cells require active mechanisms to repair damage inflicted by the DNA damaging antibiotic, and that persister recovery is not solely reversing pathways that were involved in the transition to the persister state. Additionally, we evaluated if persister cells are multidrug tolerant. We discovered that the persister population is heterogenous with a fraction of the cells being multidrug tolerant while others are killed by dual treatment. With this research, we provide a comprehensive overview of the genetic landscape underlying persistence to aminoglycoside and quinolone treatment. These fundamental insights contribute to understanding the mechanisms involved in the persister phenotype and might serve as a basis for future anti-persister therapies.