Hematogenous dissemination is important for infection by many bacterial pathogens, but is poorly understood because of the inability to directly observe this process in living hosts at the single cell level. All disseminating pathogens must tether to the host endothelium despite significant shear forces caused by blood flow. However, the molecules that mediate tethering interactions have not been identified for any bacterial pathogen except E. coli, which tethers to host cells via a specialized pillus structure that is not found in many pathogens. Furthermore, the mechanisms underlying tethering have never been examined in living hosts. We recently engineered a fluorescent strain of Borrelia burgdorferi, the Lyme disease pathogen, and visualized its dissemination from the microvasculature of living mice using intravital microscopy. We found that dissemination was a multistage process that included tethering, dragging, stationary adhesion and extravasation. In the study described here, we used quantitative real-time intravital microscopy to investigate the mechanistic features of the vascular interaction stage of B. burgdorferi dissemination. We found that tethering and dragging interactions were mechanistically distinct from stationary adhesion, and constituted the rate-limiting initiation step of microvascular interactions. Surprisingly, initiation was mediated by host Fn and GAGs, and the Fn- and GAG-interacting B. burgdorferi protein BBK32. Initiation was also strongly inhibited by the low molecular weight clinical heparin dalteparin. These findings indicate that the initiation of spirochete microvascular interactions is dependent on host ligands known to interact in vitro with numerous other bacterial pathogens. This conclusion raises the intriguing possibility that fibronectin and GAG interactions might be a general feature of hematogenous dissemination by other pathogens.

Persistent infection by pathogenic organisms requires effective strategies for the defense of these organisms against the host immune response. A common strategy employed by many pathogens to escape immune recognition and clearance is to continually vary surface epitopes through recombinational shuffling of genetic information. Borrelia burgdorferi, a causative agent of Lyme borreliosis, encodes a surface-bound lipoprotein, VlsE. This protein is encoded by the vlsE locus carried at the right end of the linear plasmid lp28-1. Adjacent to the expression locus are 15 silent cassettes carrying information that is moved into the vlsE locus through segmental gene conversion events. The protein players and molecular mechanism of recombinational switching at vlsE have not been characterized. In this study, we analyzed the effect of the independent disruption of 17 genes that encode factors involved in DNA recombination, repair or replication on recombinational switching at the vlsE locus during murine infection. In Neisseria gonorrhoeae, 10 such genes have been implicated in recombinational switching at the pilE locus. Eight of these genes, including recA, are either absent from B. burgdorferi, or do not show an obvious requirement for switching at vlsE. The only genes that are required in both organisms are ruvA and ruvB, which encode subunits of a Holliday junction branch migrase. Disruption of these genes results in a dramatic decrease in vlsE recombination with a phenotype similar to that observed for lp28-1 or vls-minus spirochetes: productive infection at week 1 with clearance by day 21. In SCID mice, the persistence defect observed with ruvA and ruvB mutants was fully rescued as previously observed for vlsE-deficient B. burgdorferi. We report the requirement of the RuvAB branch migrase in recombinational switching at vlsE, the first essential factor to be identified in this process. These findings are supported by the independent work of Lin et al. in the accompanying article, who also found a requirement for the RuvAB branch migrase. Our results also indicate that the mechanism of switching at vlsE in B. burgdorferi is distinct from switching at pilE in N. gonorrhoeae, which is the only other organism analyzed genetically in detail. Finally, our findings suggest a unique mechanism for switching at vlsE and a role for currently unidentified B. burgdorferi proteins in this process.

Spirochetes causing Lyme borreliosis are obligate parasites that can only be found in a tick vector or a vertebrate host. The ability to survive in these two disparate environments requires up and downregulation of specific genes by regulatory circuits that remain largely obscure. In this work on the Lyme spirochete, B. burgdorferi, we show that a disruption of the hrpA gene, which encodes a putative RNA helicase, results in a complete loss in the ability of the spirochetes to infect mice by needle inoculation. Studies of protein expression in culture by 2D gels revealed a change in the expression of 33 proteins in hrpA clones relative to the wild-type parent. Quantitative characterization of protein expression by iTRAQ analysis revealed a total of 187 differentially regulated proteins in an hrpA background: 90 downregulated and 97 upregulated. Forty-two of the 90 downregulated and 65 of the 97 upregulated proteins are not regulated under any conditions by the previously reported regulators in B. burgdorferi (bosR, rrp2, rpoN, rpoS or rrp1). Downregulated and upregulated proteins also fell into distinct functional categories. We conclude that HrpA is part of a new and distinct global regulatory pathway in B. burgdorferi gene expression. Because an HrpA orthologue is present in many bacteria, its participation in global regulation in B. burgdorferi may have relevance in other bacterial species where its function remains obscure. We believe this to be the first report of a role for an RNA helicase in a global regulatory pathway in bacteria. This finding is particularly timely with the recent growth of the field of RNA regulation of gene expression and the ability of RNA helicases to modulate RNA structure and function.