UC Berkeley

Toxin-Antitoxin Systems are evolutionarily successful genetic modules that have proliferated across almost all free-living bacteria. The ubiquitous Type II modules encode a toxic protein and a corresponding antitoxin protein that forms a tight complex with the toxin to inhibit its activity. Originally thought to be merely selfish genetic elements, TA systems are now implicated in a wide variety of processes from biofilm formation to bacterial persistence. Persistence is low frequency phenomenon that permits a small subpopulation within a larger growing population to survive various stresses, including antibiotic treatment. These tolerant persister cells enter a transient dormant state and are genetically identical to the larger population. Once the antibiotic is removed, persister cells resume normal growth and become sensitive to the antibiotic once again. A popular mainstream model describing how TA systems interface with bacterial persistence has come under considerable scrutiny and major tenets of the model have been invalidated. Thus, there is a critical need to carefully reassess the biological roles of TA systems.

This dissertation focuses on the hipBA TA system; the link between the hipBA module and persister cell formation was first reported over thirty years ago and endures these recent challenges. The hipBA operon encodes a HipA toxin and a HipB antitoxin. The HipA toxin is a serine/threonine kinase that functions as a protein synthesis inhibitor. In Escherichia coli, HipA phosphorylates the glutamyl tRNA synthetase, GltX, to inhibit tRNA charging. The buildup of uncharged tRNAs activates the Stringent Response, a highly conserved stress response that is implicated in persister cell formation. The precise mechanism by which HipA mediates persister cell formation is under debate and the hipBA system has only been rigorously studied in E. coli; thus, there is a need to validate the role of hipBA in bacterial persistence in other organisms and dissect the mechanism of persister cell formation. Caulobacter crescentus is an aquatic alphaproteobacteria that is adapted to survive in environments with very low nutrient availability. Curiously, the C. crescentus NA1000 chromosome contains three hipBA modules. This dissertation characterizes each of these modules and assesses their contribution to bacterial persistence in C. crescentus. Furthermore, this work dissects the biological role of each module in helping C. crescentus survive its stressful environment.

The first chapter, to be submitted as a primary research article, focuses on bacterial persistence in C. crescentus. Each hipBA module is experimentally validated as a functional TA system with a kinase HipA toxin capable of inhibiting growth and a corresponding HipB antitoxin. All three HipAs inhibit protein synthesis and several tRNA synthetases are identified as HipA kinase targets. This work is the first published report of persistence in C. crescentus. The persister fraction is quantified in rich PYE media; The persister frequency was consistently ~10-5 - 10-6 during exponential growth and ~10-4 - 10-5 in stationary phase against various antibiotics: carbenicillin, streptomycin and vancomycin. Two of the modules, hipBA1 and hipBA2, are responsible for the increase in persister frequency between exponential growth and stationary phase. Without these modules, C. crescentus does not accumulate persister cells in stationary phase cultures grown in PYE media. Curiously, all three HipAs are capable of increasing persister frequencies ~10-100 fold when ectopically expressed. HipA mediated persister cell formation is dependent on the Stringent Response. While we validate the biological role of hipBA modules in bacterial persistence through a similar pathway as that in E. coli, persistence is never abolished even in a strain that entirely lacks the Stringent Response. Therefore, there are multiple pathways through which persister cells form in C. crescentus.

The second chapter contains unpublished material examining the hipBA3 module. While the role of hipBA1 and hipBA2 in increasing persister cell frequencies in stationary phase cultures grown in PYE media was identified, no biological role was assigned to hipBA3. Here, we report a novel integration of a hipBA module into stress tolerance. Instead of being stochastically active in a small subpopulation, hipBA3 responds to environmental cues. The HipB3 antitoxin is degraded by the HslUV protease in response to phosphate limitation. Strains lacking the hipBA3 module are ultrasensitive to peroxide stress during the exponential-to-stationary phase transition in cultures grown in PYE media. Supplementing phosphate or functional HipA3 kinase complements this ultrasensitivity. Dps is a ferritin-like protein implicated in C. crescentus survival against peroxide stress. Ectopic Dps expression complements the peroxide ultrasensitivity in the ΔhipBA3 strain. dps transcript abundance is reduced in a ΔhipBA3 strain and ectopic HipA3 expression upregulates dps. Thus, HipA3 phosphorylates a downstream target that results in the transcriptional upregulation of dps. A model detailing how phosphate potentiates Dps function and the importance of translating Dps as phosphate becomes limiting is proposed.