SummaryIntracellular bacterial pathogens generally express chaperones such as Hsp100s during multiplication in host cells, allowing them to survive potentially hostile conditions. Francisella tularensis is a highly infectious bacterium causing the zoonotic disease tularaemia. The ability of F. tularensis to multiply and survive in macrophages is considered essential for its virulence. Although previous mutant screens in Francisella have identified the Hsp100 chaperone ClpB as important for intracellular survival, no detailed study has been performed. We demonstrate here that ClpB of F. tularensis live vaccine strain (LVS) is important for resistance to cellular stress. Promoter analysis shows that the transcriptional start is preceded by a s 32-like promoter sequence and we demonstrate that expression of clpB is induced by heat shock. This indicates that expression of clpB is dependent on the heat-shock response mediated by s 32, the only alternative s-factor present in Francisella. Our studies demonstrate that ClpB contributes to intracellular multiplication in vitro, but is not essential. However, ClpB is absolutely required for Francisella to replicate in target organs and induce disease in mice. Proteomic analysis of membrane-enriched fractions shows that five proteins are recovered at lower levels in the mutant strain. The crucial role of ClpB for in vivo persistence of Francisella may be linked to its assumed function in reactivation of aggregated proteins under in vivo stress conditions.

Francisella tularensis live vaccine strain infection of mice has been established as an experimental model of tularemia that is suitable for studies of immune mechanisms against the intracellular pathogen. In this study, the model was used to explore immunogenic repertoire of F. tularensis with the aim of identifying new molecules able to activate the host immune system, potential bacterial markers with vaccine, and diagnostic applications. Immunoproteomic approach based on the combination of two-dimensional gel electrophoresis, immunoblotting, and mass spectrometry was applied. Globally, 36 different proteins were identified, which strongly reacted with sera from experimentally infected mice, including several putative virulence markers of intracellular pathogens as nucleoside diphosphate kinase, isocitrate dehydrogenase, RNA-binding protein Hfq, and molecular chaperone ClpB. Of them, 27 proteins are described for the first time as immunorelevant Francisella proteins. When comparing murine immunoproteome of F. tularensis with our previous data from human patients, 25 of the total of 50 identified murine sera immunoreactive spots were recognized by human sera collected from patients suffering from tularemia, as well. Immune sera from two Lps gene congenic strains of mice, C3H/HeN (Lpsn) and C3H/HeJ (Lpsd), represented murine immunoproteome in this study. The spectrum of immunoreactive spots detected by two-dimensional immunoblotting varied throughout the course of infection depending on murine strain. Nevertheless, the antibody patterns of the two strains showed significant homogeneity in being directed against almost identical subset of antigens.

Due to its sensitivity and productivity, bottom-up proteomics based on liquid chromatography−mass spectrometry (LC−MS) has become the core approach in the field. The de facto standard LC−MS platform for proteomics operates at sub-μL/min flow rates, and nanospray is required for efficiently introducing peptides into a mass spectrometer. Although this is almost a "dogma", this view is being reconsidered in light of developments in highly efficient chromatographic columns, and especially with the introduction of exceptionally sensitive MS instruments. Although conventional-flow LC−MS platforms have recently penetrated targeted proteomics successfully, their possibilities in discovery-oriented proteomics have not yet been thoroughly explored. Our objective was to determine what are the extra costs and what optimization and adjustments to a conventional-flow LC−MS system must be undertaken to identify a comparable number of proteins as can be identified on a nanoLC−MS system. We demonstrate that the amount of a complex tryptic digest needed for comparable proteome coverage can be roughly 5-fold greater, providing the column dimensions are properly chosen, extra-column peak dispersion is minimized, column temperature and flow rate are set to levels appropriate for peptide separation, and the composition of mobile phases is fine-tuned. Indeed, we identified 2 835 proteins from 2 μg of HeLa cells tryptic digest separated during a 60 min gradient at 68 μL/min on a 1.0 mm × 250 mm column held at 55 °C and using an aqua−acetonitrile mobile phases containing 0.1% formic acid, 0.4% acetic acid, and 3% dimethyl sulfoxide. Our results document that conventional-flow LC−MS is an attractive alternative for bottom-up exploratory proteomics.