Francisella tularensis is a zoonotic bacterium that must exist in diverse environments ranging from arthropod vectors to mammalian hosts. To better understand how genes are regulated in these different environments, a transcriptional response- regulator gene (genome locus FTL0552) was deleted in F. tularensis live vaccine strain (LVS). The FTL0552 deletion mutant exhibited slightly reduced rates of extracellular growth but was unable to replicate or survive in mouse macrophages and was avirulent in the mouse model using either BALB/c or C57BL/6 mice. Mice infected with the FTL0552 mutant produced reduced levels of inflammatory cytokines, exhibited reduced histopathology and cleared the bacteria quicker than mice infected with LVS. Mice that survived infection with the FTL0552 mutant were afforded partial protection when challenged with a lethal dose of the virulent Schu S4 strain (4 of 10 survivors, day 21 post infection) when compared to naïve mice (0 of 10 survivors by day 7 post infection). Microarray experiments indicate that 148 genes are regulated in the FTL0552 mutant. Most of the genes are down regulated, indicating that FTL0552 controls transcription of genes in a positive manner. The list of down regulated genes includes genes located within the Francisella Pathogenicity Island (FPI) that are essential for intracellular survival and virulence of Francisella tularensis. Furthermore, a mutant in FTL0552 or the comparable locus in Schu S4 (FTT1557c) may be an alternative candidate vaccine for tularemia.

The phylum of Chlamydiota is composed of gram negative obligate intracellular bacteria that live as symbionts of diverse eukaryotes, from protists to animals and humans. Members of the phylum can be split into two groups: the environmental Chlamydia, which includes symbionts of amoeba, and the pathogenic Chlamydia, which includes species infecting animals, birds, and humans and includes Chlamydia trachomatis the leading cause of reportable, bacterial sexually transmitted infections and the ocular infection, trachoma. The characterized phylum members undergo a biphasic developmental cycle alternating between the infectious elementary body (EB) and the replicative reticulate body (RB), with each form having distinct morphological and physiological properties. Differentiation between these forms occurs within a host cell membrane-derived vacuole termed the inclusion. The molecular mechanisms governing and executing bacterial development and RB growth remain unclear. The essentiality and uniqueness of development makes it a prime target for the development of novel, chlamydial-specific therapeutics. Reductive evolution has resulted in the loss of or fragmentation of numerous metabolic pathways, particularly in the pathogenic Chlamydia (~1 Mbp genome) as compared to the environmental Chlamydia (~2.5 Mbp). We hypothesize that the bacterium senses environmental changes (host cytoplasm) to ensure that development and growth coincide with host cell energy and metabolite levels. We predict that an encoded partner switching mechanism (PSM) plays a key role in: 1) regulation of growth by acting as a molecular throttle through regulation of the housekeeping sigma factor, and 2) differentiation by impacting the composition of the sigma factor pool allowing for transcriptional changes needed for developmental transitions. We also predict that PSM regulation occurs through sensing of nucleotide triphosphates, TCA-cycle intermediates, metal concentrations, and redox. Canonical PSMs have a PP2C-type sensor phosphatase (SP), an anti-sigma factor (ASF, serine kinase), an anti-anti-sigma factor (AASF, substrate of the SP and ASF) and a stress-response related alternative sigma factor. The PSM in pathogenic Chlamydia is atypical, and despite its reduced genome, is comprised of two SPs (RsbU which responds to α-ketoglutarate and CTL0852), two AASFs (RsbV1 and RsbV2), one ASF (RsbW), and, unusually, the ASF regulates the availability of the “housekeeping” sigma factor, σ66. To test our hypotheses, we first constructed and purified a variety of amino acid point mutants of the two AASFs, ASF, and the SP for in vitro analyses. Kinase and phosphatase activity towards RsbV1/V2 was measured in the presence of different metals, phosphate donors, and pH and redox conditions. Phos-tag acrylamide gels were used to assess protein phosphorylation status. We discovered that metalation impacts enzyme activity and the substrate specificity of RsbU, and that RsbW can use multiple phosphate donors. Prior work, and our data, found that RsbW and RsbU have higher enzymatic activity towards RsbV1 than RsbV2, leading us to explore the importance of RsbV2 in chlamydial biology. Genome gazing revealed that environmental Chlamydia possess a single AASF, and bioinformatic analyses support that it is more similar to RsbV2 than RsbV1 suggesting that the pathogenic Chlamydia gained RsbV1. Comparing the biochemical features of the two AASFs provides potential reasons for the different enzyme affinities. To flesh out the in vivo importance of each AASF, we characterized bacterial growth, infectious progeny production, and the levels of RsbV1/V2 in a cell culture infection model using a collection of C. trachomatis L2 rsbV1 null or rsbV2 knockdown strains. We also overexpressed the AASFs in strains grown with different glucose levels. Note that C. trachomatis is an auxotroph for glucose-6-phosphate. In normal chlamydial culture glucose medium levels, the rsbV1 null strain showed an ~1 log reduction in infectious progeny numbers while the rsbV2 knockdown or AASF overexpression strains had no defects. We also observed that the rsbV1 null strain has a developmental delay and exhibits growth differences in response to glucose levels, i.e. a functional PSM seems to set a “growth cap” in response to different glucose availability. Immunoblotting analysis of RsbV1/V2 demonstrated the presence of both proteins throughout development, and protein levels remained the same in low or high glucose levels and in the wild type or rsbV1 null strains (measuring RsbV2 only for the RsbV1 null strain). These results tell us that the AASF levels have minimal impact on chlamydial biology, suggesting that phosphorylation status is key to regulation. To assess phosphorylation, we used protein pulldown assays and Phos-tag gels to assess RsbV1 and RsbV2 phosphorylation during development. Both RsbV1 and RsbV2 were phosphorylated during the EB stage, which is similar to our prior results using Chlamydia caviae. In conjunction with the in vivo phosphorylation data, we hypothesize that stage-dependent inhibition of AASF/RsbW interactions frees RsbW to sequester σ66. Reduced pools of σ66 would promote RB-EB conversion through increased RNAP binding to the late gene sigma factors σ54 and σ28. Supporting this model, overexpression of a non-phosphorylatable RsbV2 S55A mutant (an RsbW “trap”), but not overexpression of RsbV1 S56A, resulted in a 3 log reduction in infectious progeny production without gross changes in inclusion morphology or bacterial numbers, while causing a reduction in σ54 and σ28 regulated EB-specific proteins and inhibition of RB-EB transition shown via transmission electron microscopy. As an alternative approach to assess the consequence of reduced “free” RsbW, we used a CRISPRi knockdown system targeting rsbW and observed a reduction in infectious progeny production under some conditions, which is consistent with the RsbV2 S55A expression strain results. The rsbW CRISPRi-associated phenotype was weaker than the RsbV2 S55A phenotype. As bacterial redox status changes throughout development (RBs are reduced and EBs are oxidized), we also assessed whether the cysteine-rich proteins RsbV2 and RsbW were redox responsive. In parallel to the unique AASF expansion in the pathogenic Chlamydia, RsbV2 in the pathogenic Chlamydia has a CXCC motif that is not found in the RsbV homolog in the environmental Chlamydia. Our in vitro studies found that, under oxidizing conditions, RsbV2 is dimerized, and the dimer form inhibits phosphorylation of RsbV2 by RsbW. We predict that retention of RsbV2 after RsbV1 acquisition has been selected for, in part, owing to a unique redox-sensing role compared to RsbV1 and that the presence of two AASFs enables more sensitive tuning of growth and development in response to metabolite levels. The different phenotypes when overexpressing non-phosphorylatable RsbV1 and RsbV2 also hints at a potential non-PSM or expanded PSM role for RsbV1. The in vitro redox findings need to be further explored in an in vivo model. Collectively, we think the expansion of the PSM, in addition to other gene gain events, facilitated infection of multi-cellular organisms. Additionally, our data support that the PSM regulates growth/cell differentiation in response to energy/nutrients, and that redox levels and biochemical features of RsbV1 and RsbV2 govern PSM-component interactions. As disruption of normal PSM function significantly reduces production of infectious progeny, compounds targeting the PSM components could serve as novel, narrow spectrum inhibitors.

<p>Prokaryotes are continually exposed to variations in their environment. Survival in unstable milieu requires a wide range of transcriptional regulators (TRs) that respond to specific environmental and cellular signals by modulating gene expression and provide an appropriate physiological response to external stimuli. These adaptive responses to environmental signals are mostly mediated by TRs from one of two families: the single or the two component signal transduction systems (1CSTS; 2CSTS). In this thesis the structural analysis of two 1CSTS – DntR and NikR − are presented. One study was carried out to try to develop a bacterial biosensor for synthetic dinitrotulenes compounds, the other to characterise the Ni-sensing mechanism that contributes to the acid adaptation of the human pathogen<i> Helicobacter pylori.</i> DntR belongs to the LysR family and the crystal structures obtained have allowed the proposal a model of the interaction of DntR with salicylate inducer as well as giving insights into the signal propagation mechanism in LysR-type transcription factors (<b>paper I</b>). DntR mutant crystal structures combined with the modelling of DntR-2,4-dnt interactions led to the design of a DntR mutant that has a limited response to 2,4-dnt in a whole cell biosensor system (<b>paper 2</b>). Crystal structures of apo-NikR from <i>H. pylori </i>(HpNikR) and of Ni-bound intermediary states of the protein were obtained. The latter have helped in unravelling the Ni incorporation and selectivity mechanisms of NikRs and have shown a strong cooperativity between conformational changes in the Ni binding domain with movements of the DNA binding domain (<b>paper 3</b>). Biochemical studies and comparisons of the HpNikR crystal structures with those of NikR homologues strongly suggest that HpNikR has evolved different surface properties (<b>paper 4</b>) and a new mode of DNA binding. </p>

Prokaryotes are continually exposed to variations in their environment. Survival in unstable milieu requires a wide range of transcriptional regulators (TRs) that respond to specific environmental and cellular signals by modulating gene expression and provide an appropriate physiological response to external stimuli. These adaptive responses to environmental signals are mostly mediated by TRs from one of two families: the single or the two component signal transduction systems (1CSTS; 2CSTS). In this thesis the structural analysis of two 1CSTS – DntR and NikR − are presented. One study was carried out to try to develop a bacterial biosensor for synthetic dinitrotulenes compounds, the other to characterise the Ni-sensing mechanism that contributes to the acid adaptation of the human pathogen Helicobacter pylori. DntR belongs to the LysR family and the crystal structures obtained have allowed the proposal a model of the interaction of DntR with salicylate inducer as well as giving insights into the signal propagation mechanism in LysR-type transcription factors (<b>paper I</b>). DntR mutant crystal structures combined with the modelling of DntR-2,4-dnt interactions led to the design of a DntR mutant that has a limited response to 2,4-dnt in a whole cell biosensor system (<b>paper 2</b>). Crystal structures of apo-NikR from H. pylori (HpNikR) and of Ni-bound intermediary states of the protein were obtained. The latter have helped in unravelling the Ni incorporation and selectivity mechanisms of NikRs and have shown a strong cooperativity between conformational changes in the Ni binding domain with movements of the DNA binding domain (<b>paper 3</b>). Biochemical studies and comparisons of the HpNikR crystal structures with those of NikR homologues strongly suggest that HpNikR has evolved different surface properties (<b>paper 4</b>) and a new mode of DNA binding.