Diversity-generating retroelements (DGRs) are family of retroelements that introduce nucleotide variability within defined protein-encoding DNA sequences. Sequence variation is site specific and occurs through a unique reverse transcriptase mediated process called mutagenic homing. DGRs were originally identified in the Bordetella phage BPP-1 and have since been identified in plasmids, bacteriophage and bacterial genomes. Moreover, DGRs were recently identified in Archaea and their viruses. Although DGRs are wide spread in nature and protein diversification has been demonstrated in both phage and bacterial systems, the precise mechanism of DGR mutagenic homing remains to be elucidated. We have demonstrated that mutagenic homing requires specific nucleotide sequence and structural elements, including target site recognition sequences, which include a DNA stem-loop/cruciform structure. We recently demonstrated that in addition to base pairing interactions in the stem, the specific sequence and length of the 4nt loop are critical for DGR function. In vitro and in vivo analyses of the stem-loop structure indicate that the loop nucleotide composition has a major effect on stem-loop/cruciform formation and stability, and is thus critical for DGR function. In addition to influencing structure stability, we demonstrate that the orientation of the loop nucleotide sequence determines target site recognition during mutagenic homing. Stem-loops have been identified in most DGRs and our analysis of similar elements from disparate species indicates that these conserved elements are functionally interchangeable and fundamental to target site recognition. We propose that the stem-loop/cruciform structure serves as a recognition element for DNA processing events that culminate in cDNA synthesis, diversification, then integration. Since bioinformatic and functional analysis of DGRs reveals the lack of sequences predicted to encode enzymes involved in DNA or RNA processing events, we postulated host-encoded trans-acting factors play a pivotal role in mutagenic homing. To identify host-encoded factors that directly or indirectly influence the Bordetella phage BPP-1 DGR homing, a random transposon-insertion library was created. Individual transposon mutants were screened for insertions that had a significant effect on DGR homing as measured by a quantitative Km resistance assay. As an alternative approach to identify host factors, we performed targeted mutagenesis of candidate genes involved in DNA- and RNA-processing activities, including but not limited to ssDNA-specific exonuclease, ATP-dependent DNA helicase, RNase H, RNase E, and other putative endoribonucleases. Mutants were screened for their ability to support mutagenic homing using phage tropism switching assays and we identified a subset of mutations in genes encoding DNA- and RNA-processing enzymes that decreased tropism switching. Taken together, our findings indicate that mutagenic homing involves both DGR-encoded and host-encoded factors that play a role in the diversification of target proteins, providing insight into a highly conserved mechanism for DNA editing.

Burkholderia pseudomallei (Bp), the etiologic agent of melioidosis, is increasingly recognized as a significant cause of morbidity and mortality in humans. Treatment requires up to nine months of antibiotic therapy and relapse is common. Clinical management is complicated by the fact that Bp is often resistant to multiple classes of antibiotics through both intrinsic and acquired mechanisms. Even with rigorous and appropriate antibiotic therapy, mortality rates are as high as 50%. Melioidosis is endemic in tropical and sub-tropical regions throughout the world. In addition to affecting local residents, individuals at risk include travelers and US military personnel deployed to tropical regions. The closely related pathogen, Burkholderia mallei (Bm), causes equine glanders and can also afflict humans. Although exceedingly rare, human glanders presents in forms similar to melioidosis with high rates of mortality. The lethal nature of these infections and the paucity of effective antibiotics or vaccines highlight the importance of understanding virulence mechanisms and developing means to protect populations at risk. This need is amplified by concerns regarding the potential use of these pathogens as biological weapons. Bm was, in fact, used for biowarfare in both World Wars I and II.

Intercellular spread is a hallmark of Burkholderia pathogenesis and represents an attractive therapeutic target as it is required for virulence. We have shown that efficient cell-cell spread depends on the ability of the bacterium to perform critical intracellular lifecycle steps, including invasion, endosome escape, intracellular motility, and membrane fusion. A striking feature of Burkholderia is their ability to fuse plasma membranes, forming multinucleated cells (MNCs). We have discovered that cell fusion represents the primary means of intercellular spread during infection of cell monolayers in vitro, and we have proposed a model that is similar to mechanisms used by enveloped viruses. Studies in our laboratory implicate a type VI secretion system (T6SS-5) as a key determinant in this process, and VgrG-5, a trimeric protein at the tip of the T6SS-5 apparatus. There remain significant unknowns about the mechanism of cell fusion and other critical intracellular lifecycle steps that promote cell-cell spread, including the full inventory of bacterial and host factors that participate in these processes. Furthermore, our data indicate that many important virulence determinants are highly conserved between Bp and Bm, suggesting the possibility of developing broadly acting therapeutic interventions, including small molecule inhibitors.

We developed a high-throughput cell-based phenotypic assay that comprehensively assesses the Burkholderia intercellular lifecycle and used it to screen 1] a saturating transposon mutant library in Burkholderia thailandensis, a BSL2 surrogate, to identify bacterial factors required for Burkholderia intercellular spread, 2] an siRNA host kinase library for host factors involved in facilitating Burkholderia intercellular spread, and 3] a small molecule library comprising ~220,000 small molecules to identify compounds capable of disrupting intercellular spread by Burkholderia. We identified several novel bacterial factors, and one host factor, which are necessary for efficient intercellular spread of Burkholderia. In addition, our small molecule screen uncovered 91 small molecules that inhibit cell-cell spread by Bt, 27 of which also disrupt the intercellular lifecycles of Bp and/or Bm. Among these were burkfloxacin, a fluoroquinolone analog with potent intracellular bactericidal activity, and an FDA-approved drug. We found that this FDA-approved drug inhibits Burkholderia-mediated membrane fusion by targeting the secretion activity, but not expression, of the type VI secretion system-5 (T6SS-5), a critical virulence determinant. Bacterial metabolism of this drug is required for activity and ensures selective toxicity. These results suggest that burkfloxacin and this FDA-approved drug should be advanced as potential therapeutic countermeasures for melioidosis, and demonstrate the utility of cell-based phenotypic screening for select agent drug discovery.

Diversity-generating retroelements (DGRs) are distinguished by their ability to iteratively diversify defined DNA sequences which encode the ligand binding domains of target proteins (TP). Diversification occurs through a template-dependent, error prone reverse-transcriptase mediated process, termed mutagenic homing, which introduces nucleotide substitutions into a variable repeat (VR) while preserving cis- and trans- acting elements needed for future rounds of diversification. The process of DNA diversification requires a DGR-encoded reverse transcriptase (RT), an accessory variability determinant (Avd), and a template repeat (TR)-derived RNA intermediate. The archetype DGR is found within the Bordetella bacteriophage BPP however, over 300 putative DGRs have been identified within the bacterial domain as well as within a species of archaea. We have identified DGRs within the opportunistic human pathogen Legionella pneumophila (Lp) as well as within Legionella tunisiensis which encode nearly identical diversification machinery, yet each has a VR with a unique pattern of adenine mutagenesis suggesting individualized diversification in response to selection. We analyzed the DGR within Lp strain Corby and identified that the genetic requirements for mutagenic homing were similar to those of BPP, suggesting all DGRs might function through a conserved mechanism. Using in vitro growth conditions, we observed elevated levels of diversification during Lp transition from exponential to stationary growth phase suggesting that DGR mutagenic homing might be modulated by host regulatory networks. To investigate this hypothesis we generated deletions of key factors, relA and spoT, which are critical for coordinating phenotypic differentiation of Lp. PCR analysis to detect levels of mutagenic homing in wt and mutant Lp cells revealed an increase in relA spoT double deletion mutants. qRT-PCR analysis of wt and mutant cells showed that double deletion mutants had an increase in TR-RNA transcripts, while expression of avd and RT was unchanged. Over-expression of TR-RNA increased levels of mutagenic homing above those observed in wt cells, but not to the extent observed with the overexpression of avd, TR, and RT. Cumulatively these data suggest that in Lp, the abundance of TR-RNA transcripts partially controls rates of mutagenic homing, indicating regulation of these elements on multiple levels. Lp DGRs have the potential to generate ~1019 distinct polypeptide sequences within C-type lectin (CLec) domains of their TPs. The Lp Corby TP, ldtA, expresses a surface displayed outer membrane (OM) lipoprotein whose CLec domain is exposed to the extra-cellular milieu. Translocation of LdtA across the inner membrane requires the twin-arginine translocation (TAT) machinery where we hypothesize it is recognized, modified, and transported to the outer membrane by the localization of lipoproteins (Lol) system. To fully understand the pathways required for surface display of a protein, we identified and analyzed the contribution of a non-canonical lipobox with conserved targeting residues at +2/+3 positions in LdtA. Mutagenesis of the lipobox conserved cysteine as well as replacement of targeting residues with amino acids shown to result in sorting by Lol to the OM all resulted in retention of LdtA in the inner membrane. Furthermore, cleavage of LdtA from pro- to mature-peptide was found to depend on the characteristic of the +2 residue suggesting that maturation and subsequent translocation of LdtA does not follow the established convention for lipoproteins. Furthermore, we demonstrated surface display of LdtA in several species of gram negative bacteria suggesting its localization requires conserved pathways common to bacteria. These results suggest that DGR TPs in Lp are diversified in response to the physiological state of the host and trafficked to the surface by an unusual Lol-related mechanism.

Burkholderia pseudomallei and Burkholderia thailandensis are related facultative intracellular pathogens that infect a variety of cell types, escape into the cytoplasm of host cells, and spread from cell-to-cell via type 6 secretion system (T6SS)-mediated cell fusion. Recent work has demonstrated a new mechanism of rapid intracellular motility dependent on a cryptic flagellar locus, fla2. In the first section of this dissertation, I show that the Fla2 system is present and expressed in Australian strains of B. pseudomallei and B. thailandensis. The results demonstrated that the fla2 locus encodes lateral flagella expressed intracellularly. This Fla2 system promotes swarming on semisolid agar. Lateral flagella encoded by the fla2 locus may provide a growth advantage for Burkholderia in environmental conditions, as wild-type B. thailandensis outcompeted a nonmotile ∆fliC2 derivative in a semisolid agar swarming competition assay.

In the next section of this dissertation I tested the effect of intracellular expression of lateral flagella on innate immune detection of Burkholderia. Innate immune mechanisms, particularly NLRC4-driven pyroptosis and production of IL-1β, are critical for the host response to B. pseudomallei infection. Such innate immune responses were previously attributed to type 3 secretion system (T3SS) stimuli deployed intracellularly. Here, I show that this response is at least partially dependent on fliC2, the flagellin gene encoded in the fla2 cluster in B. thailandensis and Australian strains of B. pseudomallei. While deletion of fliC2 in B. thailandensis had no measurable effect on pyroptosis or intracellular bacterial growth, it led to a nearly 33% decrease in IL-1β production in wild-type (wt) murine macrophages, without affecting IL-1β release in nlrc4-deficient macrophages. FliC2 did not appear to contribute similarly to IL-1β secretion during B. pseudomallei NAU14B-6 infection despite the immunogenicity of B. pseudomallei FliC2 ectopically expressed in a B. thailandensis ∆fliC2 strain. By quantitative RT-PCR, I demonstrate that intracellular expression of fliC2 is lower in the B. pseudomallei strains tested than in B. thailandensis. This difference may provide an explanation for the discrepancy between the innate immune response observed against FliC2 in B. thailandensis versus B. pseudomallei, leaving open the possibility that differences in fliC2 expression levels between B. pseudomallei strains may lead to strain-specific innate immune responses.

In the final section of this dissertation, I describe the role of a novel fla2 regulator on the global regulatory cascade involved in the B. pseudomallei intracellular life cycle. Overexpression of the flagellar response regulator frr activates the expression of lateral flagella, but decreases plaquing efficiency in both Fla2+ and Fla2- strains of B. pseudomallei and B. thailandensis. I show through quantitative RT-PCR that frr affects not only expression of lateral flagella components, but also polar flagella components from the Fla1 system utilized for extracellular swimming by Burkholderia and type 3 secretion system components required for escape from an initial endosome and from autophagosomes. These results suggest that frr plays a role as a master regulator controlling multiple aspects of Burkholderia virulence. In this chapter, I also investigate the role of other B. pseudomallei master regulators on subsets of virulence genes and demonstrate that each of these regulators exhibits crosstalk with other regulatory systems involved in the intracellular life cycle.

Together, the results described in this dissertation define and characterize a novel intracellular motility system in B. pseudomallei. The expression of lateral flagella by Burkholderia in the host cytoplasm is a unique and surprising feature for an intracellular pathogen, given the cytosolic surveillance mechanisms evolved in mammalian hosts to respond to flagella. In the concluding remarks of this dissertation, I discuss the status of B. pseudomallei as an “accidental” pathogen and how mechanisms evolved to survive and persist in specific ecological niches can contribute to or stand in the way of mammalian pathogenesis.