FTL_0325 contributes to Francisella tularensis cell division, maintenance of cell shape, and structural integrity.

FTT0831c/FTL_0325 Contributes to Francisella tularensis Cell Division, Maintenance of Cell Shape, and Structural Integrity Gregory T. Robertson,a Elizabeth Di Russo Case,b* Nicole Dobbs,a Christine Ingle,a Murat Balaban,a* Jean Celli,b* Michael V. Norgarda Department of Microbiology, University of Texas Southwestern Medical Center, Dallas, Texas, USAa; Laboratory of Intracellular Parasites, NIAID, National Institutes of Health Rocky Mountain Laboratories, Hamilton, Montana, USAb

F

rancisella tularensis is a Gram-negative, facultative intracellular pathogen and the causal agent of the lethal zoonotic disease tularemia (1). Two subspecies of F. tularensis are the cause of the majority of human infections. Subspecies holartica is present throughout the Northern Hemisphere and is responsible for the majority of human infections, which are rarely fatal (1, 2). Subspecies tularensis is geographically restricted to North America, is capable of causing acute disease following pulmonary exposure to as few as 10 CFU, and exhibits a human mortality rate of 30% when untreated (1, 2). Natural exposure to tularemia is rare and usually is in the form of exposure to infected animals, especially rodents and lagomorphs, or through the bites of blood-feeding arthropods (3). However, because of its low infection dose and high morbidity and mortality, especially following aerosol exposure, F. tularensis has been designated by the Centers for Disease Control as a tier 1 biothreat agent with high potential for illegitimate use. It is generally believed that the highly infectious nature of the most virulent forms of F. tularensis results from a combination of successful phagosomal escape and intracellular replication but also the ability to avoid or limit early innate immune detection by the host (4). The former is dependent on genes contained within the Francisella pathogenicity island (FPI), a cluster of 16 to 19 genes which are postulated to encode a protein delivery system with distant sequence similarities to other known type VI secretion systems (5). In contrast, the factors that allow this pathogen to avoid early innate immune detection are as yet poorly defined (6–11). This has led many investigators to search for a bacterial factor or factors responsible for this early innate immunosuppression. In one such study, Weiss et al. identified mutants of strain F. tularensis subsp. novicida U112 that were attenuated in vivo and also hypercytotoxic in tissue culture (12). A hypercytotoxic phenotype also results from loss of oppB encoding a putative oligo-

July 2014 Volume 82 Number 7

peptide permease (13) or pepO, which may encode a metallopeptidase (13, 14). Similar observations were made for live vaccine strain (LVS) deficient in folate metabolism or pseudouridine synthase genes (15), mviN, a putative lipid flippase (16), ripA, a cytoplasmic protein of unknown function (17), and kdhAB, encoding a Kdo hydrolase (18), or for Schu S4 variants lacking genes involved in lipopolysaccharide (LPS) O-antigen and capsule biosynthesis (19). One interpretation of these results is that Francisella has the ability to actively limit host cell death and that modulation of these cell death pathways involves a broad number of Francisella gene products. Another possibility is that each of these individual mutations indirectly increases the overall cytotoxicity of the mutant strain for unrelated reasons. Indeed, recent work by Peng et al. (20) has demonstrated that multiple unrelated F. tularensis subsp. novicida U112 mutants, including those lacking

Received 18 January 2014 Returned for modification 20 February 2014 Accepted 20 April 2014 Published ahead of print 28 April 2014 Editor: S. M. Payne Address correspondence to Michael V. Norgard, [email protected]. * Present address: Elizabeth Di Russo Case, Department of Microbial Pathogenesis and Immunology, Texas A&M Health Science Center College of Medicine, Bryan, Texas, USA; Murat Balaban, Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, Nobels Vag, Stockholm, Sweden; Jean Celli, The Paul G. Allen School for Global Animal Health, College of Veterinary Medicine, Washington State University, Pullman, Washington, USA. Supplemental material for this article may be found at http://dx.doi.org/10.1128 /IAI.00102-14. Copyright © 2014, American Society for Microbiology. All Rights Reserved. doi:10.1128/IAI.00102-14

Infection and Immunity

p. 2935–2948

iai.asm.org

2935

Downloaded from http://iai.asm.org/ on March 14, 2015 by GEORGIAN COURT UNIV

The Francisella FTT0831c/FTL_0325 gene encodes amino acid motifs to suggest it is a lipoprotein and that it may interact with the bacterial cell wall as a member of the OmpA-like protein family. Previous studies have suggested that FTT0831c is surface exposed and required for virulence of Francisella tularensis by subverting the host innate immune response (M. Mahawar et al., J. Biol. Chem. 287:25216 –25229, 2012). We also found that FTT0831c is required for murine pathogenesis and intramacrophage growth of Schu S4, but we propose a different model to account for the proinflammatory nature of the resultant mutants. First, inactivation of FTL_0325 from live vaccine strain (LVS) or FTT0831c from Schu S4 resulted in temperature-dependent defects in cell viability and morphology. Loss of FTT0831c was also associated with an unusual defect in lipopolysaccharide O-antigen synthesis, but loss of FTL_0325 was not. Full restoration of these properties was observed in complemented strains expressing FTT0831c in trans, but not in strains lacking the OmpA motif, suggesting that cell wall contact is required. Finally, growth of the LVS FTL_0325 mutant in Mueller-Hinton broth at 37°C resulted in the appearance of membrane blebs at the poles and midpoint, prior to the formation of enlarged round cells that showed evidence of compromised cellular membranes. Taken together, these data are more consistent with the known structural role of OmpA-like proteins in linking the OM to the cell wall and, as such, maintenance of structural integrity preventing altered surface exposure or release of Toll-like receptor 2 agonists during rapid growth of Francisella in vitro and in vivo.

Robertson et al.

2936

iai.asm.org

phages (BMDMs) in vivo. The latter step precedes intracellular destruction in late forming LAMP-1-positive endosomes in vivo. We present evidence that these temperature-dependent growth defects result at least partly from altered cell division, culminating in the formation of highly irregular enlarged cells. These and other properties—including an unusual LPS O-antigen profile for the Schu S4-based, but not the LVS-based, FTT0831c/FTL_0325 mutants—are more consistent with this protein contributing to maintaining structural integrity rather than acting to subvert the innate immune response directly. MATERIALS AND METHODS Bacterial strains and culture conditions. F. tularensis subsp. tularensis strain Schus S4 (CDC1001) was obtained from the Centers for Disease Control and Prevention (Fort Collins, CO) in accordance with all federal and institutional select agent regulations and was manipulated under strict biosafety level 3 (BSL3) containment conditions. F. tularensis subsp. holarctica strain LVS was obtained from Timothy Sellati (Center for Immunology and Microbial Disease, Albany Medical College, Albany, NY) and manipulated under BSL2 containment conditions. For routine cultivation, F. tularensis was grown in sMHB or on supplemented MuellerHinton agar (sMHA) (32). A modified chocolate agar (CA⫹) agar was used for initial recovery of transconjugates (32). Brain heart infusion broth was prepared as previously described (33). Escherichia coli DH5␣ was used for routine plasmid manipulation. E. coli S17.1 was used as a host for bacterial conjugation. Where needed, Francisella growth media were supplemented with 200 mg of hygromycin/liter, 10 mg of kanamycin/ liter, 100 mg of polymyxin B/liter, 25 mg of ampicillin/liter, or 16 mg of vancomycin/liter, or 8% (wt/vol) sucrose. E. coli was grown using LuriaBertani broth or agar further supplemented with 200 mg of hygromycin/ liter, 30 mg of kanamycin/liter, or 100 mg of ampicillin/liter, as required. Owing to a marked growth restriction of Francisella strains lacking a functional FTT0831c gene, all Francisella strains prepared in the present study were routinely propagated at 30°C, except where indicated. Gene knockout and genetic complementation. Splicing-overlap extension (SOE) PCR (34) was used to generate an FTT0831c or FTL_0325 deletion-insertion cassette in which the majority of the coding region of the FTT0831c (FTL_0325 in LVS) open reading frame (ORF) (encompassing nucleotides 2 to 1168 of the 1,254-bp ORF) was replaced with the flippase recognition target (FRT)-flanked kanamycin resistance cassette (FRT-Pfn-kan-FRT) from pLG66a (35). Details on the primer pairs used for this construction are available in the supplemental material. Methods for SacB-based sucrose-assisted allelic replacement and subsequent flippase (FLP)-based excision of the FRT-flanked kanamycin-resistance cassette to generate markerless FTT0831c or FTL_0325 deletion mutations (henceforth referred to as ⌬0831 in Schu S4 and ⌬0325 in LVS) were as described previously (32). For genetic complementation, the FTT0831c ORF was placed under the control of the Francisella rpsL promoter (Pr) by SOE PCR (34). PstI and BamHI sites, engineered into the flanking primers, allowed directional cloning of the resultant Pr-FTT0831c into plasmid pUC18T-miniTn7T (36) cut with the same enzymes to yield pTP414. Addition of a BamHI-restricted kanamycin-resistance marker from pTP86 allowed selection for integration of the Pr-FTT0831c mini-Tn7 transposon at attTn7 (37) after electroporation into the LVS- and SchuS4-based ⌬0831::FRT mutants carrying the unstable helper plasmid pTP181 (32). The FTT0831c variant lacking the OmpA motif was constructed by inverse PCR of plasmid pTP414 with the primer pair GP276 (TCCCCCGGGTG TTTCGATTAGATCAGGTCCTGTTTGTT) and GP277 (AAAAGTAGA CTTATAGAGCAAATTGATAATATT). GP276 was designed to also carry an engineered SmaI site (underlined) to facilitate religation after PCR amplification and purification. This results in the addition of a single nontemplated proline residue at the site of the Asn67-Leu180 deletion. The addition of the kanamycin selection marker and integration into attTn7 was otherwise as described above.



genes whose products are involved in LPS biosynthesis or encode membrane proteins, were hypercytotoxic to macrophages in vitro, not because of loss of immune evasion factors but instead to increased intracytosolic bacteriolysis (resulting in the heightened release of pathogen-associated molecular pattern molecules [PAMPs]). Examination of these strains in vitro revealed aberrant cell morphology during growth in minimal medium (20), which was similarly reported for wild-type LVS cultivated under certain growth conditions (21). Ulland et al. reported similar morphological abnormalities and hyperinflammasome activation for an LVS strain deficient in a putative lipid flippase, mviN (16). In another example, LVS lacking kdhAB, a putative Kdo hydrolase, stimulated tumor necrosis factor alpha (TNF-␣) and interleukin-1␤ (IL-1␤) in a Toll-like receptor 2 (TLR2)-dependent fashion in vivo, owing to increased accessibility of surface proteins, possibly reflecting altered conformation of the outer membrane (18). Together, these data were interpreted to mean that certain conditions (specific mutations or growth conditions) alter surface characteristics or structural integrity or increase susceptibility to intracellular bacteriolysis in such a way as to promote heightened exposure of otherwise inaccessible PAMPs to germ line-encoded innate immune receptors. FTT0831c/FTL_0325 encodes a putative lipoprotein and shares significant homology (E values of 2.27E–3 to 4.10E–25 by BLAST) to orthologous proteins of the OmpA-like protein family (22–25). Proteins bearing OmpA-like structural motifs typically form tight, noncovalent interactions with the bacterial cell wall (26–28) and, in some instances, contribute directly to bacterial cell division by forming a dynamic molecular cross-bridge between the cell wall and the outer membrane (29). FTT0831c/FTL_0325 is not predicted to adopt a porin-like structure and lacks obvious motifs associated with other membrane-spanning proteins such as ␤-barrels. FTT0831c/FTL_0325 was previously reported to contribute to Francisella immune evasion by interfering with cytosolic AIM2 (absent in melanoma 2)- and NLRP3 (NLR-pyrin domain containing 3)-based inflammasome activation and nuclear NF-␬B signaling (30). However, this conclusion was based principally on observations of hyper-TLR2-dependent inflammatory stimulation following infection of host cells with FTT0831c/ FTL_0325-deficient bacteria and not on functional studies with FTT0831c protein per se (30). FTL_0325 was reported to be surface exposed, and the authors observed a modest, non-dose-dependent, effect on NF-␬B signaling following transfection of FTT0831c DNA into HEK293T cells (30). However, a precise immune evasion mechanism was not presented. An LVS FTL_0325 mutant was also impaired for virulence in mice and proved efficacious as a live attenuated vaccine to protect mice from Schu S4-mediated death following subsequent intranasal (i.n.) challenge (31), thus suggesting that the inflammatory nature of the LVS FTL_0325 mutant has practical implications as well. We sought here to reevaluate the biological roles of FTT0831c/ FTL_0325 by first constructing deletion mutants in both LVS and virulent Schu S4 backgrounds. Our data support a role for FTT0831c in intracellular survival and murine virulence of Schu S4, but we propose a different model to account for its hyperinflammatory nature. We demonstrate that the loss of FTT0831c/ FTL_0325 promotes profound temperature-dependent defects in cell viability and altered cell morphology during growth in supplemented Mueller-Hinton broth (sMHB) medium in vitro and during cytosolic replication in bone marrow-derived macro-

FTT0831c Is Required for Cell Integrity and Virulence


with AxioVision Rel.4.8 software (Carl Zeiss MicroImaging GmbH). In some cases, fixed cells were stained at room temperature with DAPI (4=,6=-diamidino-2-phenylindole) at 1 mg/liter. For Live/Dead BacLight staining (L7012; Molecular Probes), unfixed cells were recovered by centrifugation, washed in 0.85% saline, and stained according to the manufacturer’s instructions. Processed images were false colored and merged using ImageJ software (39). The relative viability of LVS ⌬0325 bacteria by live/dead staining was determined in a fluorescence microplate reader relative to that of suspensions of live and isopropyl alcohol-killed versions of the Tn7 transcomplemented LVS clone (Pr-0325⫹). For transmission electron microscopy (TEM), sMHB-grown stationary-phase cells, prepared and fixed in 10% BNF as described above, were washed twice by centrifugation in PBS, and the pellets were suspended in 0.01 ml of fixative solution (2.5% [wt/vol] glutaraldehyde, 0.1 M sodium cacodylate). This preparation was applied to positively charged Formvarcoated copper (200 mesh) grids (Electron Microscopy Sciences, Hatfield, PA) for 5 min. Excess liquid was blotted to cellulose filter paper (Whatman, Piscataway, NJ), and the samples were stained with 2% (wt/vol) uranyl acetate solution. Visualization was performed with a Tecnai G2 Spirit BioTWIN (FEI Company, Hillsboro, OR) transmission electron microscope. Flow cytometry. Aliquots (0.25 ml) of sMHB-grown stationary-phase cells, prepared and fixed in 10% BNF as described above, were further diluted in 10 ml of PBS and stained with 10 mg of ethidium bromide/liter for 5 min at room temperature or left unstained. Samples were washed twice by centrifugation, as described above, in 10 ml of PBS and suspended in 10 ml of PBS for flow cytometry. For each experiment, DNA content in a population of 200,000 cells was measured in a BD FACSCalibur flow cytometer at the UT Southwestern flow cytometry core. The data were collected and analyzed using FlowJo software (Tree Star, Inc., San Carlos, CA). Macrophage culture and infection. Bone marrow cells were isolated and differentiated essentially as described previously (40). Immediately prior to infection, a few colonies from a freshly streaked sMHA plate were suspended in sMHB, and the OD600 was measured to estimate bacterial numbers. Bacterial suspensions were then diluted in 1 g of glucose/liter in Dulbecco modified Eagle medium (DMEM; Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen), 10% L-conditioned medium, and 2 mM L-glutamine, and 0.5 ml was added to chilled BMDMs at a multiplicity of infection (MOI) of 50. Bacteria were centrifuged onto macrophages at 400 ⫻ g for 10 min at 4°C, and infected BMDMs incubated for 20 min at 37°C under 7% CO2 atmosphere, including an initial, rapid warm-up in a 37°C water bath to synchronize bacterial uptake. Infected BMDMs were then washed five times with DMEM to remove extracellular bacteria and incubated for 40 min in complete medium, followed by an additional 60 min in complete medium containing 100 mg of gentamicin/ liter to kill extracellular bacteria. Thereafter, infected BMDMs were incubated in gentamicin-free medium until processing. The number of viable intracellular bacteria per well was determined in triplicate for each time point. Infected BMDMs were washed three times with sterile PBS and then lysed with 1 ml of sterile 1% saponin for 3 min at room temperature, followed by repeated pipetting to complete lysis. Serial dilutions of the lysates were rapidly plated onto sMHA and incubated for 3 days at 37°C under 7% CO2 before enumeration of CFU. Infection of J774A.1 macrophages with LVS and derivatives was performed as described previously (32). Immunofluorescence microscopy. BMDMs grown on 12-mm glass coverslips in 24-well plates were infected, washed three times with PBS, fixed with 3% paraformaldehyde (pH 7.4) at 37°C for 20 min, washed three times with PBS, and then incubated for 10 min in 50 mM NH4Cl in PBS in order to quench free aldehyde groups. Samples were blocked and permeabilized in blocking buffer (10% horse serum and 0.1% saponin in PBS) for 30 min at room temperature. Cells were labeled by incubating inverted coverslips onto drops of primary antibodies diluted in blocking buffer for 45 min at room temperature. Primary antibodies used were mouse anti-F. tularensis LPS (US Biological, Swampscott, MA) and rat

iai.asm.org 2937


mCherry expressing Francisella were generated by placing a copy of the mCherry ORF under the control of the Francisella rpsL promoter (PrpsL) in a derivative of the stable hygromycin-resistant shuttle vector pMP831 (38) designated pTP266 (see the supplemental material). The resultant vector (pTP388) was introduced into LVS ⌬0325 by electroporation, followed by selection on sMHA supplemented with hygromycin. Animal care and use. All procedures involving animals were approved by the University of Texas Southwestern Medical Center (UT Southwestern) Medical Center Institutional Animal Care and Use Committee and the Biological and Chemical Safety Advisory Committee. Animals were housed in microisolator cages at the UT Southwestern Animal Resource Center and fed irradiated food and water ad libitum. Infection of C3H/HeN mice. Female 7- to 8-week-old C3H/HeN mice were used. All animals were housed in BSL3 facilities. Mice were anesthetized with ketamine plus xylazine and infected, dropwise, with 0.02 ml (0.01 ml per naris) for intranasal (i.n.) infections or by injection with 0.1 ml for intraperitoneal (i.p.) infections. Actual infection doses were determined by plating in triplicate onto sMHA and mice were monitored daily for signs of morbidity and mortality. Statistical comparisons were made using the log-rank (Mantel-Cox) test function of GraphPad Prism. For experiments requiring tissue harvest, lungs, spleens, and the left lateral lobe of the liver were aseptically harvested from mice and placed in Whirlpack bags (Nasco, Fort Atkinson, WI) and processed essentially as described previously (32), except all plates were incubated at 30°C in an atmosphere of 5% CO2. This lower recovery temperature was used for all strains to account for the temperature-sensitive phenotype of the Schu S4 ⌬0831 mutant. Cytokine quantitation. Sterile-filtered tissue homogenate and serum samples (day 5 postinfection [p.i.]) were analyzed for cytokine concentrations using the Bio-Plex Pro mouse cytokine 32-plex assay (Bio-Rad). Bio-Plex assay conditions were performed as indicated in the manufacturer’s instructions. Mouse cytokines granulocyte colony-stimulating factor (G-CSF), Eotaxin, GM-CSF, gamma interferon (IFN-␥), IL-1␣, IL-1␤, IL-2, IL-4, IL-3, IL-5, IL-6, IL-7, IL-9, IL-10, IL-12(p40), IL-12(p70), LIF, IL-13, LIX, IL-15, IL-17, IP-10, KC, macrophage chemoattractant protein 1 (MCP-1), macrophage inflammatory protein 1␣ (MIP-1␣), MIP-1␤, M-CSF, MIP-2, MIG, RANTES, VEGF, and TNF-␣ were analyzed for each sample. Growth curves, temperature sensitivity screens, and phase microscopy. Growth of the strains constructed herein was evaluated in sMHB with moderate aeration or on sMHA in an atmosphere of 5% CO2. For the former, cells were harvested from the surface of a sMHA plate grown at 30°C in an atmosphere of 5% CO2 for 72 h and back diluted in sMHB to give a routine starting optical density at 600 nm (OD600) of ⬃0.02 to 0.05. Cultures were either loosely capped (LVS-based) or grown with 0.22-␮mpore-size filter tops to allow free air exchange. For growth curve analysis, samples were grown as described above, but samples were also recovered at indicated times, serially diluted 10-fold in phosphate-buffered saline (PBS), and spread on sMHA for parallel CFU enumeration. For agarbased temperature sensitivity assays, the cells were grown and prepared as described above to give a starting OD600 of 0.01. This was serially diluted 10-fold in PBS, and 0.005 ml of the 10⫺2 through 10⫺4 sample was spotted and allowed to dry onto the surface of a sMHA plate. Duplicate plates were incubated at 30 or 37°C in an atmosphere of 5% CO2 for 72 to 96 h. For broth recovery assays, cells were grown in sMHB as described above at 30 or 37°C with aeration and then back-diluted 40-fold in fresh prewarmed sMHB medium to assess regrowth potential. Growth was monitored spectrophotometrically as the OD600. Fixation procedures and microscopy. Bacterial cells were collected by centrifugation at 16,000 ⫻ g for 5 min at room temperature, fixed in 10% buffered neutral formalin (BNF), and stored at 4°C until use. For phase and fluorescence microscopy, cells were immobilized by placing them on a microscope slide with a thin pad of 1% (wt/vol) agarose. Bacteria were observed with a Zeiss Axiovert 200 microscope with a ⫻63 objective. The images were acquired with the AxioCam MRm camera and processed

Robertson et al.

RESULTS

Genetic inactivation and complementation of FTT0831c/ FTL_0325. To gain more insight into the role of FTT0831c/ FTL_0325 in the biology and pathogenesis of F. tularensis, we inactivated FTL_0325 from the LVS and FTT0831c from the Schu S4 backgrounds using homologous recombination with a FRTflanked kanamycin resistance cassette and SacB-assisted allelic replacement, followed by FLP-based excision of the antibiotic resistance marker as described previously (32). This results in a markerless deletion of FTL_0325 or FTT0831c, respectively, leaving only a short FRT scar behind, which is diagnostic for gene inactivation. To verify that any phenotype resulting from loss of FTT0831c/FTL_0325 was due to absence of the protein and not other unrelated mutations, a modified Tn7 delivery system (32) was next used to insert a wild-type copy of FTT0831c or a variant lacking the canonical OmpA motif [FTT0831c ⌬(Asn67-Leu180)] (Fig. 1A) under the control of the Francisella rpsL promoter (Pr) in attTn7 near the glmS gene (32, 37). Loss of FTT0831c expression was confirmed by immunoblot of the ⌬0831 and ⌬0325 mutants and restoration of FTT0831c/FTL_0325 expression to slightly elevated levels was observed for the Tn7 transcomplemented clone (referred to here as Pr-0831⫹ and Pr-0325⫹, respectively) and to levels equivalent to that of wild-type for the variant lacking the ompA motif [Pr-⌬(OmpA)] (see Fig. 5A). FTL_0325 encodes a bacterial lipoprotein. Although FTT0831c/ FTL_0325 is highly conserved among sequenced Francisella species (⬎75% identity, nonvirulent subspecies; ⬎99% identity virulent subspecies) and bears a highly conserved OmpA structural motif (22–25), the remainder of this protein shares much lower overall sequence identity to other proteins in the nonredundant database (data not shown). FTT0831c/FTL_0325 was localized to the outer membrane by sucrose density gradient fractionation and immunoblot (30; G. T. Robertson, unpublished observations) and

2938

iai.asm.org

FIG 1 FTL_0325 encodes a lipoprotein. (A) Schematic depiction of FTT0831c/ FTL_0325 and flanking genes (based on Schu S4 nomenclature). The OmpA motif (encompassing asparagine 67 to leucine 180) is marked with a black bar. Putative pseudogenes are indicated with parentheses. (B) The N terminus of FTT0831c/FTL_0325 contains a lipobox motif (boldface type), including a canonical cysteine (underlined) at position 20. (C) Autoradiograph demonstrating in vivo incorporation of [3H]palmitic acid into polypeptides in the LVS or the FTL_0325 null mutant (⌬0325). FTL_0325 (0325) is indicated with an arrow. The asterisk signifies an unrelated radiolabeled protein present in whole-cell lysates and serves as an internal label incorporation control. Abbreviations: WCL, whole-cell lysate; IP-␣8, immunoprecipitate using rat antiFTT0831c antiserum-coupled Dynabeads; IP-NS, immunoprecipitate using naive rat sera coupled Dynabeads.

is predicted to encode a lipoprotein based on the presence of a putative signal peptidase II (SPII) cleavage site and conserved cysteine residue at position 20 in the FTT0831c/FTL_0325 coding sequence (Fig. 1B). Consistent with this proposal, FTL_0325 was found to partition into the detergent phase of the nonionic detergent Triton X-114 (data not shown), which is thought to solubilize bacterial lipoproteins owing to the amphipathic properties imparted by the covalently attached long-chain fatty acids (41). To confirm these phase partitioning results, LVS or LVS ⌬0325 were grown in CDM medium and then pulsed for ⬃18 h with the radiolabeled long-chain fatty acid precursor, [3H]palmitic acid. Growth in the presence of [3H]palmitate resulted in the appearance of labeled proteins, but none that obviously correlated with the predicted size (⬃42 kDa) of mature FTL_0325 (Fig. 1C). We therefore performed an immunoprecipitation of these labeled whole-cell lysates using anti-FTT0831c sera (␣8) or with control preimmune sera as was previously described (32). As shown in Fig. 1C, we observed the specific enrichment of a protein corresponding to the predicted size of FTL_0325 in LVS after immunoprecipitation with ␣8 but not after immunoprecipitation with control preimmune sera (Fig. 1C). As expected, no such band was seen for the LVS ⌬0325 null mutant (Fig. 1C), which does not produce FTL_0325 protein (see Fig. 5A). Parallel immunoblots using these same sera confirmed the identity of the precipitated [3H]palmitate labeled protein as FTL_0325 (data not shown). Taken together, these data are consistent with the in silico analyses identifying FTT0831c/FTL_0325 as a lipoprotein. FTT0831c is a required for full lethality and dissemination of Schu S4 in mice. Because FTL_0325 was previously shown to be required for in vivo virulence of LVS in mice (31) and FTT0831c and FTL_0325 were required for intracellular survival of Schu S4 and LVS in vitro (30), we next sought to confirm the role of FTT0831c in F. tularensis pathogenesis by performing i.n. infections of C3H/HeN mice with our Schu S4 ⌬0831 null mutant. Mice were infected dropwise via the i.n. route with the indicated infection dose and monitored for signs of illness or death for up to 3 weeks p.i. (Fig. 2A). These time-to-death assays revealed significant attenuation of the Schu S4 ⌬0831 null mutant, which was



anti-mouse LAMP-1 (clone 1D4B, developed by J. T. August and obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by the University of Iowa, Department of Biological Sciences, Iowa City, IA). Bound antibodies were detected by incubation with 1:500 dilutions in blocking buffer of Alexa Fluor 488-conjugated donkey anti-mouse and Alexa Fluor 568-conjugated donkey anti-rat antibodies for 45 min at room temperature. Cells were washed twice with 0.1% saponin in PBS, once in PBS, and once in H2O and then mounted in Mowiol 4-88 mounting medium (Calbiochem, Gibbstown, NJ). Samples were observed on a Carl Zeiss LSM 710 confocal laser scanning microscope for image acquisition. Confocal images of 1,024 ⫻ 1,024 pixels were acquired and assembled using Adobe Photoshop CS3. To quantify the escape of Francisella from its initial phagosome, phagosomal integrity assays were performed as described previously (40). Briefly, infected BMDMs on 12-mm glass coverslips were selectively permeabilized by incubation with 50 ␮g of digitonin (Sigma)/ml for 1 min at room temperature. Rabbit polyclonal anti-calnexin (Stressgen Biotechnologies) and Alexa Fluor 488-conjugated mouse monoclonal anti-F. tularensis LPS antibodies (US Biological) were delivered to the macrophage cytosol for 12 min at 37°C to label the endoplasmic reticula of permeabilized cells and accessible intracellular bacteria, respectively. The coverslips were then washed, fixed, and processed for microscopy as described above. Bound anti-calnexin antibodies were detected using cyanin 5-conjugated donkey anti-rabbit antibodies (Jackson ImmunoResearch Laboratories), and all intracellular bacteria were labeled using Alexa Fluor 568conjugated anti-Francisella antibodies. Samples were observed using a Nikon Eclipse E800 epifluorescence microscope equipped with a Plan Apo ⫻60/1.4 objective for quantitative analysis.


dissemination assays following intranasal administration to mice. (A) Groups of 6 to 8 C3H/HeN mice were infected intranasally with the indicated dose of Schu S4 or mutant bacteria and monitored for signs of morbidity for up to 3 weeks p.i. The data are representative of two independent experiments. (B) Groups of mice were infected intranasally with ⬃102 CFU Schu S4 or mutant bacteria. Bacterial burdens were determined by serial dilution and plating of organ homogenates for CFU on days 3 and 5 p.i. Absence of any remaining bacteria for ⌬0831was determined on day 35 p.i. The horizontal line indicates the mean result. The limit of detection (LOD) was 150 CFU/organ for spleens and lungs and 446 CFU/organ for livers. ns, not significant; **, P ⬍ 0.01; ***, P ⬍ 0.001 (two-way analysis of variance [ANOVA]).

attenuated at 11,200 CFU (4 of 7 mice survived; P ⬍ 0.001 versus Schu S4) and avirulent at 163 CFU (8 of 8 mice survived; P ⬍ 0.001 versus Schu S4). The wild-type Schu S4 parent was lethal at 120 CFU (0 of 8 mice survived, median survival time, 5 days), which is ⬎5 times the minimum i.n. lethal dose for Schu S4 in our hands (see Fig. 2A). Restoration of full virulence was observed for the complemented strain (Pr-0831⫹) at 350 CFU (0 of 7 mice survived, median survival time, 5 days; no difference versus Schu S4 and P ⬍ 0.001 versus ⌬0831) but not in an otherwise identical strain expressing a derivative of FTT0831c [Pr-⌬(OmpA)] lacking the putative OmpA motif when administered at 1,017 CFU (5 of 6 mice survived; P ⬍ 0.001 versus Schu S4 and no difference versus ⌬0831) (see Fig. 2A). These data are interpreted to mean that FTT0831c is required for lethal pulmonary infection of type A Schu S4 in C3H/HeN mice and that the OmpA motif is required for FTT0831c to function in this capacity. Organ CFU burdens in the lungs, livers, and spleens of parallel groups of mice infected with similar infection doses of ⬃102 CFU revealed marked differences in the colonization patterns of the tested strains. Whereas increasing concentrations of bacteria were recovered from the lungs, spleens, and livers of Schu S4-infected mice at days 3 and 5 p.i., the Schu S4 ⌬0831 mutant was found at significantly lower levels in the lungs of mice and was cleared by


FIG 3 Attenuation of the Schu S4 FTT0831c null strain in time-to-death and dissemination assays after i.p. administration to mice. (A) Groups of eight C3H/HeN mice were infected i.p. with the indicated dose of Schu S4 or mutant bacteria and monitored for signs of morbidity for up to 2 weeks p.i. (B) Bacterial burdens were determined by serial dilution and plating of organ homogenates for CFU from two mice each on day 3 p.i. (all groups) and day 5 p.i. for the Schu S4 ⌬0831 mutant. The horizontal line indicates the mean result. *, P ⬍ 0.05 (two-way ANOVA).

day 35 p.i. (Fig. 2B). In contrast, little to no detectable Schu S4 ⌬0831 bacteria were detected in the livers and spleens of these animals at days 3 or 5 (Fig. 2B), indicating an appreciable defect in either dissemination from the lung, or replication in these more distal tissues. To distinguish between these two possibilities, we performed a second experiment in which eight mice each were infected systemically by the i.p. route with 209 CFU Schu S4, 205 CFU ⌬0831, or 112 CFU Pr-0831⫹. Infection by this route bypasses the lung barrier and is highly lethal for mice. Indeed, all of the Schu S4- or Pr-0831⫹-infected mice rapidly succumbed to disease (median survival time, 4 days; no statistical difference between Schu S4 and Pr-0831⫹), whereas none of the mice infected with the Schu S4 ⌬0831 null mutant died (P ⬍ 0.001 versus Schu S4 or Pr-0831⫹) or showed any outward signs of illness (Fig. 3A). To this end, whereas Schu S4 and Pr-0831⫹ showed extensive replication and were recovered at equivalent levels to one another in the lungs, livers, and spleens of infected animals on day 3 p.i., the Schu S4 ⌬0831 mutant was detected at low levels in the spleen and was present at, or below, the limit of detection in the lung or liver (Fig. 3B). Importantly, no net replication was observed in any of these tissues between days 3 and 5 for the Schu S4 ⌬0831 mutant (Fig. 3B). These data indicate that the defect in systemic colonization of mice by the Schu S4 ⌬0831 null mutant was not the due to an inability to escape the lung per se but instead reflected an inability of this mutant to thrive in these more distal tissues.

iai.asm.org 2939


FIG 2 Attenuation of the Schu S4 FTT0831c null strain in time-to-death and

Robertson et al.

infected mice was measured by Bio-Plex Pro mouse cytokine 32-plex assay. The data are presented as a ratio of cytokine levels stimulated by the Schu S4 ⌬0831 mutant over that stimulated by the Schu S4 parent. Dashed lines indicate an arbitrary 1.7-fold difference cutoff.

The Schu S4 FTT0831c mutant hyperstimulates proinflammatory cytokines in the lungs of mice. Infection of macrophages in vitro with LVS lacking FTL_0325 or Schu S4 lacking FTT0831c was shown previously to result in hyperproinflammatory cytokine release in a manner that was dependent on TLR2 (30). To investigate whether similar responses were observed with our mutant in vivo, we examined cytokine levels in filtered lung homogenates from C3H/HeN mice by using a Bio-Plex Pro mouse cytokine 32-plex assay 5 days after i.n. infection with Schu S4 ⌬0831 or the wild-type parent Schu S4 (Fig. 4). We elected to examine the cytokine response in lungs at this time because (i) the lung is the site of initial colonization following i.n. infection and (ii), unlike the liver or spleen, CFU differences between these strains were less dramatic by day 5 p.i. in these tissues (see Fig. 2B). Cytokine quantitation revealed a bipolar response in the lungs of mice 5 days p.i. with Schu S4 or Schu S4 ⌬0831; heightened production of IL-17, IFN-␥, IL-1␣, IL-1␤, TNF-␣, LIF, RANTES, and IP-10 was disproportionately observed for the Schu S4 ⌬0831 mutant compared to the Schu S4 parent, whereas higher levels of G-CSF, MCP-1, IL-10, and IL-5 were detected for the virulent parent versus the attenuated Schu S4 ⌬0831 mutant (Fig. 4). The complete cytokine profile, including what appears to be a Schu S4-induced “cytokine storm” in the livers, spleens, and sera of these same animals (which were severely ill at this time), is available in the supplemental material. These data are interpreted to mean that i.n. infection of mice with Schu S4 0831 promotes hyperinduction of a unique set of proinflammatory cytokines in lung tissues in vivo and is consistent with limited in vitro and in vivo data published elsewhere (30, 31, 42). Altered surface properties of the Schu S4, but not LVS-based, FTT0831c/FTL_0325 null mutants. Hyperproinflammatory cytokine production following infection with FTT0831c/FTL_0325deficient LVS or Schu S4 variants were previously proposed to result from loss of surface-exposed protein and not heightened release of TLR2 agonists owing to surface changes or loss of structural integrity (30). Curiously, however, our immunoblots revealed an unexpected loss of high-molecular-weight O-antigen production for our Schu S4-derived ⌬0831 strain, but not the otherwise identical LVS-derived ⌬0325 mutant (Fig. 5A). This is noteworthy, inasmuch as mutants of F. novicida and F. tularensis lacking key O-antigen biosynthetic genes were previously shown to be hypercytotoxic to BMDMs (19, 20). In our study, the lack of

2940

iai.asm.org

O-antigen biosynthesis in the Schu S4 ⌬0831 background was fully reversed in the complemented clone (Pr-0831⫹) or in a second complemented clone isolated through a separate, wholly independent, transformation experiment (Fig. 5A, left panel). Further, the expression of a mutant form of FTT0831c lacking the OmpA motif, or the empty vector alone, failed to reverse this defect in O-antigen production. Therefore, although it is currently unclear why the loss of FTT0831c/FTL_0325 affects LPS synthesis differently in these two closely related strain backgrounds, the genetic complementation data indicate that it is the loss of FTT0831c, and not a secondary spontaneous mutation, that results in this LPS O-antigen defect in the Schu S4 background. Importantly, this unusual LPS defect did not impart increased sensitivity to complement-mediated killing to the Schu S4 ⌬0831 mutant; control strains (i.e., E. coli DH5␣ or a spontaneous deep rough Schu S4 isolate) were readily killed by complement-preserved, but not heat-inactivated, human serum (Fig. 5B). Deletion of FTT0831c/FTL_0325 alters cell growth and morphology of F. tularensis at physiologic temperatures. Previous studies failed to detect any growth defects of LVS FTL_0325 mutants grown in a variety of liquid media at physiologic temperatures (30, 31). We elected to investigate this in more detail, since in our hands, passage of either the LVS ⌬0325 or Schu S4 ⌬0831 null mutants on sMHA in an atmosphere 5% CO2 at 37°C consistently resulted in the appearance of smaller colonies relative to that of the parent or the complemented mutant (data not shown). This phenotype was even more striking for LVS ⌬0325 cells were first diluted in PBS (or sMHB) and spotted onto the surface of sMHA at 37°C in an atmosphere of 5% CO2, which resulted in a 2- to 3-log reduction in viability (Fig. 6A). The Schu S4 ⌬0831 mutant also showed a strong growth restriction and a significant (P ⬍ 0.05, Student t test) reduction in plating efficiency (i.e., 30%) at 37°C (data not shown). The basis for the difference in apparent magnitude of this effect between the LVS and Schu S4 strain backgrounds is at present unknown. However, the effect in both cases was reversed by growth at 30°C or by genetic complementation (Fig. 6A and data not shown), thus indicating that the growth restriction phenotype is temperature dependent and results from the loss of FTT0831c/FTL_0325 and not polar effects on downstream genes. The presence of the OmpA motif was also required to rescue the temperature-sensitive growth of both the Schu S4 ⌬0831 and the LVS ⌬0325 mutants in vitro, which is interpreted to



FIG 4 The Schu S4 ⌬0831 strain hyperstimulates proinflammatory cytokine production in the lungs of mice. Cytokine production at day 5 p.i. in lungs of


variant. Anti-F. tularensis LPS O-antigen monoclonal antibody (FB11) or monospecific, polyclonal antibodies to FTT0831c (␣8) or FTT0825c (␣0825) were used to assay LPS O-antigen production, or FTT0831c/FTL_0325 levels, or FTT0825 levels (loading control), respectively, in Schu S4, Schu S4 lacking FTT0831c (⌬0831), the complemented clone (Pr-0831⫹), a variant lacking only the putative OmpA motif [Pr-⌬(OmpA)], the ⌬0831 variant expressing the kanamycin resistance gene (Pg-aphA), or an independently isolated complemented clone (Pr-0831⫹#8) (left panels) or LVS, LVS lacking FTL_0325 (⌬0325), the complemented clone (Pr-0325⫹), or a variant lacking only the putative OmpA motif [Pr-⌬(OmpA)] (right panels). (B) Resistance of Schu S4 or derivative bacteria to killing by human serum. Bacteria were incubated for 60 min at 37°C in RPMI containing 10% (vol/vol) fresh human serum (hatched bars) or human serum heat inactivated (solid bars) at 56°C for 30 min prior to use. E. coli DH5␣ and a spontaneous deep rough variant of Schu S4 (rough S4) were used as positive controls for complement-mediated killing. (C) Mutation of FTL_0325 in LVS leads to enhanced accessibility of the bacterial cell surface. F. tularensis LVS, the ⌬0325 mutant, and the complemented clone (Pr-0325⫹) were tested for surface binding of anti-F. tularensis LPS monoclonal antibody (LPS) or polyclonal antiFTT0831c (0831), anti-FopA (FopA), anti-EfTu (EfTu) antibodies (ab) in surface accessibility assays (SAA). Bacteria with surface-bound antibodies were lysed, and the proteins were resolved by SDS–12.5% PAGE and transferred to nitrocellulose. Immunoblots were performed using peroxidase-conjugated secondary antibodies to mouse IgG or rat IgG. Reactions with the heavy chain of IgG (IgG HC) and total Tul4A protein levels (loading control) are shown.

mean that contact with the peptidoglycan cell wall is required for FTT0831c/FTL_0325 activity in vitro as well (Fig. 6A and data not shown). Surprisingly, no such growth defect was initially evident when LVS ⌬0325 cultures were examined for increases in the optical density (OD) during a single round of aerobic cultivation in sMHB liquid medium at 37°C (Fig. 6B). This result is in agreement with that reported previously (30, 31) and may partially explain the discrepancy between these two studies. However, in striking contrast to the parent strain or the complemented strain, the LVS ⌬0325 mutant showed no corresponding increase in CFU during peak log phase growth in sMHB (Fig. 6C). Similarly, both the Schu S4 ⌬0831 mutant and a strain lacking the putative OmpA motif showed a substantial, but somewhat lower, ⬃90% (1-log) reduction in viability during growth to saturation in broth culture relative to the complemented clone or the Schu S4 parent (data not shown). These findings are at odds with that reported previously


(30), where disproportionately elevated log10 CFU values were reported per OD equivalent when growth was assayed in microtiter plate format and not under typical aerobic broth conditions as used herein. Microscopic examination of these cells after growth to saturation in sMHB under permissive (aerobic growth, 30°C) and nonpermissive (aerobic growth, 37°C) conditions also revealed strikingly different cell morphologies. Whereas wild-type LVS or the complemented mutant were found to exhibit characteristic pleomorphic rod-shaped morphology by TEM (Fig. 6E and I), the LVS ⌬0325 cells grown at 37°C were spherical in nature and approximately three to five times the size of the parent or the complemented clone (Fig. 6F). The spherical forms of the LVS ⌬0325 mutant were often phase bright by standard phase-contrast microscopy (data not shown). No such irregularities were seen for the LVS ⌬0325 cells when cultivated at 30°C (permissive growth conditions) (Fig. 6G) and, importantly, the transition to the larger spherical shape occurred concomitantly with active growth in

iai.asm.org 2941


FIG 5 Altered surface properties of Schu S4 ⌬0831 or LVS ⌬0325. (A) Immunoblot demonstrating altered LPS O-antigen production in the Schu S4 ⌬0831

Robertson et al.

broth, since these cells were essentially indistinguishable from that of the parent upon initial inoculation or during early log phase growth (i.e., ⬍3 doublings) (Fig. 6H and see Fig. 8A). Finally, this defect was not unique to LVS, since similar alterations in cell morphology were observed for the Schu S4 ⌬0831 strain (Fig. 6K) or this same strain expressing an allele of FTT0831c lacking the OmpA motif [Pr-⌬(OmpA)] when cultivated at 37°C in sMHB (Fig. 6M and data not shown). Hence, in two different strain backgrounds, the loss of functional FTT0831c/FTL_0325 protein results in highly irregular cell morphology following growth at physiologic temperature in vitro. Given that cell mass (OD) increased linearly during log phase growth (see Fig. 6B), but viable CFU did not (Fig. 6C), we hypothesized that loss of FTT0831c/FTL_0325 was promoting a defect in normal cell division under these nonpermissive growth conditions. Similar results have been reported for mutants lacking other OmpA motif containing bacterial lipoproteins (i.e., Pal) (29). To assess this directly, we next examined the relative number of genome equivalents (i.e., DNA content) of the LVS parent or the LVS ⌬0325 mutant cultivated under permissive (aerobic growth, 30°C) or nonpermissive (aerobic growth, 37°C) temperatures. If

2942

iai.asm.org

our hypothesis was correct, we would predict a corresponding increase in ethidium bromide fluorescence (i.e., nucleic acid content) equivalent to the overall cell mass increase for an individual LVS ⌬0325 bacterial cell. Indeed, using flow cytometry we observed an ⬃3.7-fold shift in mean fluorescence values for LVS ⌬0325 grown under nonpermissive temperatures, relative to that of the wild-type parent or the LVS ⌬0325 mutant grown under permissive conditions (Fig. 6D). This observed ⬃3.7-fold-increase in genome equivalents is, in general, in good agreement with the approximately 3- to 5-fold increase in cell surface area observed in TEM, thus supporting our hypothesis that the disconnect in cell mass and cell viability can be at least partly attributed to cell division defects at 37°C resulting from the loss of FTL_0325. Similar increases in ethidium bromide staining (i.e., genomic DNA content) were observed for the Schu S4 ⌬0831 mutant or a strain lacking the OmpA structural motif [⌬(OmpA)] but were restored to wild-type levels in the complemented clone (0831⫹) (Fig. 6D). Intracytosolic growth of the Schu S4 ⌬0831c mutant promotes altered bacterial morphology and eventual killing in LAMP-1-positive vacuoles. To determine the intracellular fate of



FIG 6 Loss of FTT0831c/FTL_0325 imparts growth defects at physiologic temperatures, aberrant morphology, and altered cell division. (A) Tenfold serial dilutions corresponding to the 10⫺2 (lane ⫺2) the 10⫺3 (lane ⫺3), and 10⫺4 (lane ⫺4) of LVS, LVS lacking FTL_0325 (⌬0325), the complemented clone (Pr-0325⫹), or a variant lacking only the putative OmpA motif [Pr-⌬(OmpA)] grown previously on sMHA at 30°C were spotted for growth at 30 or 37°C on duplicate sMHA plates. (B and C) Disconnect between increases in cell mass (OD600) and cell viability (CFU) in the LVS ⌬0325 variant. Aerobic growth of LVS or mutant bacteria in sMHB was monitored spectrophotometrically by direct observation of the OD600 in 1.5-cm tubes (B). Cell viability was determined in parallel by serial dilution in PBS and plating of samples onto sMHA at 30°C for CFU determinations (C). (D) Flow cytometric analysis showing the increase in ethidium bromide staining (DNA content) of the LVS ⌬0325 and Schu S4 ⌬0831 mutants or Schu S4 expressing a variant of FTT0831c lacking the putative OmpA motif [⌬(OmpA)] relative to wild-type or a complemented mutant (0831⫹) when grown at 37°C. Relative numbers of genome equivalents were calculated as the ratio of the mean fluorescence peak of each sample over that observed for the parent grown at 37°C. The LVS ⌬0325 mutant grown at 30°C (permissive temperature) was included as a control. (E to I) Bacterial morphology of LVS and derivatives visualized via transmission electron microscopy. (E) LVS parent at stationary phase at 37°C; (F) LVS ⌬0325 at stationary phase at 37°C; (G) LVS ⌬0325 at stationary phase at 30°C; (H) LVS ⌬0325 at early log phase at 37°C; (I) the complemented LVS ⌬0325 strain (Pr-0325⫹) at stationary phase at 37°C. All TEM images were scaled to the same extent. A scale bar is included for reference. (J to M) Bacterial morphology of Schu S4 and derivatives visualized by fluorescence microscopy of DAPI-stained cells fixed previously in 10% buffered neutral formalin. (J) Schu S4 parent at stationary phase at 37°C; (K) Schu S4 ⌬0831 at stationary phase at 37°C; (L) the complemented Schu S4 ⌬0831 strain (Pr-0831⫹) at stationary phase at 37°C; (M) Schu S4 [⌬(OmpA)] at stationary phase at 37°C. All fluorescence microscopy images were scaled to the same extent.


Schu S4 lacking FTT0831c, we infected BMDM and measured intracellular growth by a gentamicin protection assay and endosomal trafficking using a previously described phagosomal integrity assay coupled to confocal immunofluorescence microscopy of bacterial colocalization with LAMP-1-positive membranes (as a measure of vacuolar versus cytosolic location) (43). Consistent with previous reports (30), we observed appreciable defects in intracellular replication of the Schu S4 ⌬0831 strain in BMDMs (Fig. 7A). These defects manifested as a reduced apparent intracellular growth rate, and an appreciable loss of viability between 16 and 24 h p.i. Similar intracellular growth defects were observed for the LVS ⌬0325 strain at 22 h in J774A.1 macrophages (5.9% intracellular survival relative to the LVS parent; data not shown). Hence, our results from both LVS and Schu S4 are wholly consistent with that reported elsewhere (30). Importantly, these defects were fully reversed in the complemented mutants (Fig. 7A and data not shown). We also show that this defect did not correlate with an inability to escape the phagosome, since the Schu S4 ⌬0831 strain was found free in the cytosol with kinetics that were indistinguishable from that of the virulent parent or the complemented mutant (Fig. 7B). Consistent with our previous in vitro findings, however, intracytosolic growth of the Schu S4 ⌬0831 strain also resulted in the formation of enlarged irregular cells that


stained poorly with anti-O-antigen LPS antibody (Fig. 7C, middle panels, see insets). This suggests that, like the situation in vitro, rapid growth of Schu S4 during intracytosolic residence at 37°C was also associated with alterations in cell morphology in vivo and the altered LPS profile affected intracellular staining with anti-LPS antibody. Strikingly, by 16 h p.i. the Schu S4 ⌬0831 cells were observed in association with LAMP-1-positive vacuoles where active bacterial destruction was observed (Fig. 7C, far right panel). The association of Schu S4 ⌬0831 bacteria with LAMP-1-positive vacuoles corresponded well with the appreciable loss of cell viability observed in gentamicin protection assays (compare Fig. 7A and C, far right panel) and since these same bacteria were previously cytosolic, must mean that the cells had reentered the endocytic pathway, possibly by autophagy (44), to be destroyed in association with LAMP-1-positive vacuoles. Taken as a whole, these results further substantiate the role of FTT0831c in the pathogenesis of virulent F. tularensis by suggesting an unusual intracellular fate resulting not from failure to escape the phagosome, but rather, altered growth and morphology and subsequent destruction of otherwise replicating intracytosolic bacteria by late-forming LAMP-1-positive vacuoles. Progressive changes in cell morphology and loss of membrane integrity. Alterations in cell morphology were observed for

iai.asm.org 2943


FIG 7 Aberrant cell morphology accompanies intracytosolic replication of the Schu S4 ⌬0831 strain and eventual lysis in late forming LAMP-1-positive endosomes. (A) Schu S4, the ⌬0831 null mutant, or the ⌬0831 null mutant complemented in trans from attTn7 were used to infect BMDM seeded in 24-well plates at an MOI of 50. Intracellular CFU were enumerated at various times p.i. The data are presented as means ⫾ the standard deviations (SD) from a representative experiment performed at least twice. (B) Intracellular trafficking of Schu S4 and derivatives in BMDM. At various times p.i., infected macrophages were subjected to a phagosomal integrity assay to enumerate the percentage of cytosolic bacteria. The data are means ⫾ the SD of three independent experiments. (C) Representative confocal micrographs of BMDM infected for 1 or 10 h with Schu S4 and derivatives or for 16 h with the ⌬0831 null strain. Samples were processed for immunofluorescence labeling of bacteria (green) and LAMP-1-positive vacuoles (red). Single-channel images of the boxed areas are shown in the magnified insets. White arrow indicates bacteria of interest.

Robertson et al.

physiologic temperatures. (A) LVS ⌬0325 cells constitutively expressing mCherry were visualized by immunofluorescence microscopy during growth in sMHB at 37°C. Growth (i.e., doublings) was monitored spectrophotometrically by direct observation of the OD600 in 1.5-cm tubes, and samples were removed and fixed in 10% buffered neutral formalin prior to microscopic examination. Averages of bacteria exhibiting normal, irregular, or spherical morphology were determined by counting ⬎100 bacteria in at least two different fields. (B) Behavior of LVS (black circles), the LVS ⌬0325 mutant (blue circles), or the complemented clone (Pr-0325⫹) (red circles) during serial passage in sMHB at 30°C (upper panel) or 37°C (lower panel). (C) Differences in OD600 (closed circles, dashed lines) and viable counts (open circles, solid lines) for the LVS ⌬0325 mutant with serial passage at 30°C (upper panel) or 37°C (lower panel). (D) Live/Dead staining of LVS, the LVS ⌬0325 (⌬0325) mutant or the complemented clone (Pr-0325⫹) grown to saturation at 37°C in sMHB. The percentage of viable LVS ⌬0325 bacteria relative to the Tn7 complemented clone (Pr-0325⫹) is shown.

bacteria lacking FTT0831c/FTL_0325 in vitro in rich media and in vivo in BMDMs at physiologic temperatures. Overall, these changes appeared to result from altered cell division at 37°C that was not similarly apparent at lower growth temperatures in vitro. The most logical explanation would be that these morphological abnormalities were progressive in nature, and not the result of a singular cell wall defect. To determine whether this was true, we examined LVS ⌬0325 cells expressing mCherry for morphological changes over time during active aerobic growth at 37°C in sMHB in vitro. Consistent with our previous studies (see Fig. 6H), the LVS ⌬0325 mutant initially presented as short, pleomorphic rods (through ⬃3 doublings) (Fig. 8A), typical of wild-type LVS bacteria (see Fig. 6E). After 5 doublings, however, cellular abnormalities became apparent, and ⬃33% of the surveyed cells were irregular in nature, which included the appearance of cells with enlarged ends and perpendicular blebs near the cellular midpoint (Fig. 8A). These changes were progressive, since more than 45% of surveyed cells were irregular after 7 doublings with the initial appearance of large spherical cells at this point (i.e., ⬃9% of cells counted) (Fig. 8A). After overnight growth (⬃13 doublings), the majority of the cells were irregular, or more often (i.e., 91% of cells), large and spherical in nature (Fig. 8A). Taken together, these data indicate that the changes in LVS ⌬0325 morphology are progressive in nature and result in at least two distinct morphotypes during different stages of rapid growth in sMHB medium in vitro. We next sought to determine whether the LVS ⌬0325 cells grown to saturation were capable of renewed replication upon

2944

iai.asm.org

dilution in fresh medium or whether the spherical cells represented a terminal form of aborted division. Whereas all strains grew normally and were not impacted in further growth upon serial passage at 30°C (Fig. 8B, upper panel), the LVS ⌬0325 mutant showed negligible increases in OD600 (i.e., cell mass) upon serial passage at 37°C in two independent experiments (Fig. 8B and C, lower panels). No such defect was observed for the LVS parent or the complemented mutant at 37°C (Fig. 8B, lower panel). To ensure that this defect was not simply due to loss of viability upon growth to saturation in sMHB under nonpermissive growth temperatures, we repeated the serial passage studies and recovered cells for CFU enumeration after serial dilution and plating at 30°C. Consistent with our earlier observation (Fig. 6B and C), we observed an inverse correlation between OD600 increases and viable CFU counts following growth of the LVS ⌬0325 mutant to saturation at 37°C but not at 30°C (Fig. 8C). Although similar morphological abnormalities are seen during growth of the Schu S4 ⌬0831 strain in vitro and within macrophages (see Fig. 6K and 7C), it is not yet known whether the Schu S4 ⌬0831deficient bacteria exhibit a similar fate during serial passage in vitro. However, given the similarities in altered cell division observed between these two closely related species, this seems likely to be the case. Because these cells appeared morphologically intact, we next used a Live/Dead BacLight bacterial viability staining kit (Molecular Probes) to assess cell viability, and hence, the membrane integrity of individual cells. As shown in Fig. 8D, we observed ap-



FIG 8 Progressive changes in morphology and reduced membrane integrity and cell death are associated with rapid growth of the LVS ⌬0325 null mutant at


bacterial lipoprotein that likely forms noncovalent interactions between the outer membrane (OM) and the peptidoglycan cell wall (PG). The inner membrane is shown (IM). Deletion of FTT0831c/FTL_0325 results in loss of critical OM and cell wall contacts promoting the formation of OM perturbations (especially at the midpoint or cell poles, where active or recent cell constriction has occurred) that may allow release of PAMPs directly or increase host access to these molecules. Diagrammatic representations of the various cell morphotypes associated with rapid growth of the F. tularensis FTT0831c/FTL_0325-deficient bacteria are shown.

preciable decreases in viability (i.e., green cells) for individual LVS ⌬0325 cells (50.3% viable) relative to the complemented mutant (Fig. 8D). The LVS ⌬0325 bacteria were not, however, more susceptible to hydrophobic compounds and detergents (i.e., sodium dodecyl sulfate, ethidium bromide, vancomycin, deoxycholate, gentamicin, and bacitracin) when tested at 5 ⫻ 105 CFU/ml in microtiter plate-based MIC assays in vitro (data not shown). To investigate this further, we next evaluated the ability of Tul4A antibodies to bind to normally occluded Tul4A at the cell surface of live cells in vitro using a previously described surface antigen binding assay (18). While LPS was readily detected at the cell surface by mouse anti-LPS (FB11) antibody and Western blotting, FTL_0325, FopA, and Ef-Tu were not detected at the cell surface of any strain tested by Western blotting (Fig. 5C). In contrast, and consistent with the proposal that loss of FTL_0325 alters surfaceexposed constituents, antibodies to Tul4A reproducibly detected sufficient quantities of this outer-membrane-anchored lipoprotein in the LVS ⌬0325 mutant but not in the LVS parent or the complemented clone (Fig. 5C). This indicates that loss of FTL_0325 promotes altered cell morphology and loss of viability without obviously compromising cell membrane integrity but instead alters outer membrane envelope structure sufficiently to allow surface exposure (or release) of some normally occluded protein constituents, including those likely to stimulate innate immune receptors (i.e., PAMPs) (Fig. 9). DISCUSSION

F. tularensis is considered a “stealth” pathogen, owing to its ability to establish infection without significant early host detection. F. tularensis modifies its Lipid A to avoid immunodetection by pattern recognition receptors (PRRs) such as TLRs (45). However, the mechanisms by which F. tularensis avoids and suppresses other aspects of the host innate immune response are poorly defined (6–11). Surveys for F. tularensis mutants that show heightened cytotoxicity or TLR2-dependent inflammatory responses has led to the identification of multiple unrelated gene products that contribute to this response (15–19). Although it is possible that some of the gene products identified are directly involved in subverting


host innate immune response, the prevailing model is that these mutations alter the structural integrity of the cell resulting in either increased bacteriolysis during intracytosolic residence (20) or altered cell properties culminating in increased access of PRRs to otherwise inaccessible TLR ligands (18). One common feature of such mutants is an increase in early host recognition and proinflammatory cytokine response (e.g., TNF-␣ and IL-1␤) but also decreased pathogenesis and reduced bacterial dissemination (16, 18). FTT0831c/FTL_0325 was previously reported as a virulence factor for F. tularensis, contributing to intracellular survival and murine pathogenesis (30, 31, 42). Our studies support and extend these findings by demonstrating that loss of FTT0831c severely impairs murine pathogenesis of Schu S4 and promotes a strong proinflammatory response (i.e., IL-17, IFN-␥, IL-1␣, IL-1␤, TNF-␣, LIF, RANTES, and IP-10) in the lungs of mice following primary pulmonary infection (see Fig. 2A and 4). We further show that Schu S4 ⌬0831 bacteria show reduced, but significant, replication in the lungs of mice following i.n. infection, but with little to no dissemination or replication in more distal tissues (Fig. 2B). This may suggest that the early innate immune response to ⌬0831 bacteria is critical in controlling further systemic spread. Indeed, similar results were observed for LVS bacteria lacking the Kdo hydrolase genes kdhAB which elicits a similar strong early inflammatory response in the lung following i.n. infection (18). Other studies have shown that prior administration of known TLR agonists reduces overall organ burdens and increases survival following subsequent F. tularensis infection (46–48). However, other properties likely also contribute to the limited systemic replication of the ⌬0831 bacteria, since the Schu S4 ⌬0831 strain also failed to replicate in the spleens and livers of mice when administered systemically via i.p. injection, which bypasses the lung barrier. This defect is not due to increased susceptibility to complement-mediated killing, inasmuch as the Schu S4 ⌬0831 bacteria were fully resistant to the action of preserved human serum in vitro, despite possessing an altered LPS profile (Fig. 5A). The basis for the latter difference between the LVS ⌬0325 and Schu S4 ⌬0831 mutants is at present unknown, but our genetic data strongly suggest that the

iai.asm.org 2945


FIG 9 Model for FTT0831c/FTL_0325 contribution to cell division and inflammatory immune responses in vivo. FTT0831c/FTL_0325 (0831) encodes a

Robertson et al.

2946

iai.asm.org

self, is difficult to reconcile given that FTT0831c/FTL_0325 would be required to broadly inhibit both nuclear NF-␬B signaling (30) and cytosolic AIM2 and NLRP3-inflammasome signaling (42) for FTT0831c/FTL_0325 to exert its proposed effects. Further, although FTT0831c/FTL_0325 clearly is outer membrane associated, it likely is tethered via its long-chain fatty acids to the inner leaflet of the outer membrane (Fig. 9); lipoproteins are not commonly surface exposed in Gram-negative bacteria, inasmuch as the translocation of lipoproteins to the outer leaflet is not thermodynamically favorable, and thus likely requires a highly specialized pathway that is yet to be characterized. To this end, we failed to detect surface-exposed FTL_0325 in our surface accessibility assays (see Fig. 5C). Our data thus suggest a different model to account for the hyperinflammatory nature of FTT0831c/ FTL_0325 mutants. First, rescue of in vitro growth at 37°C or in vivo growth in mice requires the OmpA structural motif, which is highly conserved in FTT0831c/FTL_0325 and typically required for noncovalent interactions with the peptidoglycan cell wall. Based on these data alone, it is difficult to understand how these features might contribute to the proposed biological role (30) of FTT0831c/FTL_0325 as a surface-bound (or secreted) innate immune evasion factor, unless some as-yet-unknown alternate processing event results in the release of membrane-bound protein for transport or translocation to the bacterial cell surface. Second, inactivation of FTT0831c in the Schu S4 background results in prominent defect in LPS O-antigen synthesis, which is fully restored when wild-type FTT0831c is expressed in trans. Lastly, FTT0831c/FTL_0325 was required for normal growth at physiologic temperature in rich medium or during intracytosolic residence in BMDMs. These growth defects were accompanied by prominent morphological irregularities that were found to correlate with heightened defects in structural (i.e., membrane) integrity, based on failure to exclude propidium iodide in Live/Dead staining (see Fig. 8D) or surface accessibility assays (Fig. 5C) following growth of the LVS-based ⌬0325 mutant to saturation at physiologic temperatures in vitro. Similar structural changes resulting in increased inflammatory properties can arise during growth of Francisella under certain in vitro cultivation conditions (21) or in the presence of a number of genetic mutations (20). Further, we propose that these growth defects arise due to loss of molecular interactions between the OmpA structural motif and peptidoglycan since mutants lacking this motif phenocopy that of a null mutant. As such, we suggest that FTT0831c/FTL_0325, like other well-characterized OmpA motif-containing proteins (i.e., Pal), acts principally as a structural protein, serving to tether the OM to the peptidoglycan during rapid cell growth conditions in vitro and in vivo (Fig. 9). Physical contact between OM-anchored FTT0831c/FTL_0325 and the peptidoglycan cell wall would serve to link the two layers together and facilitate normal OM invagination during active cell cytokinesis (Fig. 9). Without such interactions, FTT0831c/FTL_0325-deficient bacteria undergo cell division defects and exhibit altered structural properties that are prominent at cell midpoles and termini (the sites of active or recent cell constriction events) and the eventual formation of large round cells during rapid log growth, resulting in release or enhanced presentation of PAMPs (Fig. 9) such as Tul4A (see Fig. 5C). This in turn gives rise to the heightened induction of an inflammatory cytokine response observed in vivo during active bacterial growth conditions (see Fig. 4). In contrast, this proposed outer-membrane–peptidoglycan interaction may prove less criti-



defect is due to loss of FTT0831c and dependent on the OmpAsequence motif and therefore is not due to a secondary spontaneous mutation(s). As such, we favor another model that the elevated temperature of the mouse (ca. 37°C to 39°C) promotes cell division and growth defects similar to that observed during in vitro passage of this mutant at physiologic temperatures in vitro (see Fig. 6K). Indeed, it is anticipated that the lung growth environment, wherein significant replication of the mutant is observed in vivo (see Fig. 2B), would be closer to the permissive temperature for the mutant in vitro owing to the cooling effects of ambient room temperature air exchange and respiration. This is further supported by the observation that the mutant is not pathogenic and fails to replicate when administered to mice via the i.p. route (see Fig. 3). Further studies would be necessary to fully test this proposal. One curious difference between our work and that reported elsewhere (30, 31) for an LVS FTL_325 transposon mutant is the apparent lack of acellular growth defects in the latter. In those studies, cell growth was measured during incubation for 36 h in rich or defined medium in microtiter plates at 37°C (30, 31). In our hands, FTT0831c/FTL_0325 is essential for normal cell growth, division, and viability of Schu S4 and LVS during growth at physiologic temperatures in vitro (Fig. 6 and 8) and during intracytosolic residence in vivo (Fig. 7). This loss of viability correlated with the appearance of obvious morphological changes, which were both progressive in nature and identical between FTT0831c/FTL_0325 null mutants and variants lacking only the OmpA structural motif. Thus, the OmpA motif is required for normal cellular activity and virulence, leading us to propose that FTT0831c/FTL_0325 activity requires contact with the peptidoglycan cell wall. In contrast, these defects were not observed when these cells were grown at 30°C, possibly reflecting the lower growth rate of cells under these conditions. Although our CFU data clearly indicate some heightened loss of viability with growth of the LVS ⌬0325 mutant to saturation in sMHB at 37°C in vitro, these data are not sufficient to account for the negligible increase in OD observed upon secondary cultivation of these same cells after initial passage at 37°C (see Fig. 8B and C). Indeed, saturated sMHB LVS ⌬0325 cells passaged once at 37°C could not be recovered by secondary passage at 30°C (i.e., permissive temperature) in sMHB (data not shown). This is instead interpreted to mean that the growth defect resulting from loss of FTL_0325 in vitro leads to a form of terminal cell division block from which further cell growth is inhibited even under otherwise permissive growth conditions. Similar growth defects are observed for E. coli and other Gram-negative bacteria entering into SulA-mediated SOStype DNA repair (reviewed in reference 49). The role of FTT0831c in the virulence of Schu S4 was confirmed in these studies, as was the hyperinduction of a strong proinflammatory response in the lungs of mice following i.n. infection. This response was not due to differences in bacterial burden, since the numbers of Schu S4 and the ⌬0831 bacteria were similar at that time. In a model presented by Mahawar et al. (30), it was proposed that the basis for the TLR2-dependent proinflammatory cytokine response from BMDMs following infection with attenuated LVS FTL_0325 or Schu S4 FTT0831c mutants was not due to altered host response to these attenuated pathogens but instead to the physical loss of surface-exposed FTT0831c/ FTL_0325 protein, which in some manner, is proposed to act as a specific innate immune evasion factor. This notion, in and of it-


ACKNOWLEDGMENTS We thank Felix Yarovinsky (UT Southwestern, Department of Immunology) for helpful discussions and for technical assistance with flow cytometry and Neal Alto (UT Southwestern, Department of Microbiology) for the gift of mCherry. This study was supported by grant U54 AI057156 from National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH). The contents are solely the responsibility of the authors and do not necessarily represent the official views of the RCE Programs Office, NIAID, or NIH.

REFERENCES 1. Foley JE, Nieto NC. 2009. Tularemia. Vet. Microbiol. 140:332–338. 2. McLendon MK, Apicella MA, Allen LA. 2006. Francisella tularensis: taxonomy, genetics, and Immunopathogenesis of a potential agent of biowarfare. Annu. Rev. Microbiol. 60:167–185. http://dx.doi.org/10.1146 /annurev.micro.60.080805.142126. 3. Ellis J, Oyston PC, Green M, Titball RW. 2002. Tularemia. Clin. Microbiol. Rev. 15:631– 646. http://dx.doi.org/10.1128/CMR.15.4.631-646.2002. 4. Celli J, Zahrt TC. 2013. Mechanisms of Francisella tularensis intracellular pathogenesis. Cold Spring Harbor Perspect. Med. 3:a010314. http://dx .doi.org/10.1101/cshperspect.a010314. 5. Broms JE, Sjostedt A, Lavander M. 2010. The role of the Francisella tularensis pathogenicity island in type VI secretion, intracellular survival, and modulation of host cell signaling. Front. Microbiol. 1:136. http://dx .doi.org/10.3389/fmicb.2010.00136. 6. Jones CL, Napier BA, Sampson TR, Llewellyn AC, Schroeder MR, Weiss DS. 2012. Subversion of host recognition and defense systems by Francisella spp. Microbiol. Mol. Biol. Rev. 76:383– 404. http://dx.doi.org /10.1128/MMBR.05027-11. 7. Bosio CM, Bielefeldt-Ohmann H, Belisle JT. 2007. Active suppression of the pulmonary immune response by Francisella tularensis Schu4. J. Immunol. 178:4538 – 4547. http://dx.doi.org/10.4049/jimmunol.178.7.4538. 8. Crane DD, Ireland R, Alinger JB, Small P, Bosio CM. 2013. Lipids derived from virulent Francisella tularensis broadly inhibit pulmonary inflammation via Toll-like receptor 2 and peroxisome proliferator-activated receptor alpha. Clin. Vaccine Immunol. 20:1531–1540. http://dx.doi.org /10.1128/CVI.00319-13. 9. Hall JD, Woolard MD, Gunn BM, Craven RR, Taft-Benz S, Frelinger JA, Kawula TH. 2008. Infected-host-cell repertoire and cellular response in the lung following inhalation of Francisella tularensis Schu S4, LVS, or U112. Infect. Immun. 76:5843–5852. http://dx.doi.org/10.1128/IAI .01176-08.


10. Malik M, Bakshi CS, Sahay B, Shah A, Lotz SA, Sellati TJ. 2006. Toll-like receptor 2 is required for control of pulmonary infection with Francisella tularensis. Infect. Immun. 74:3657–3662. http://dx.doi.org/10.1128/IAI .02030-05. 11. Mares CA, Ojeda SS, Morris EG, Li Q, Teale JM. 2008. Initial delay in the immune response to Francisella tularensis is followed by hypercytokinemia characteristic of severe sepsis and correlating with upregulation and release of damage-associated molecular patterns. Infect. Immun. 76: 3001–3010. http://dx.doi.org/10.1128/IAI.00215-08. 12. Weiss DS, Brotcke A, Henry T, Margolis JJ, Chan K, Monack DM. 2007. In vivo negative selection screen identifies genes required for Francisella virulence. Proc. Natl. Acad. Sci. U. S. A. 104:6037– 6042. http://dx.doi.org /10.1073/pnas.0609675104. 13. Brotcke A, Weiss DS, Kim CC, Chain P, Malfatti S, Garcia E, Monack DM. 2006. Identification of MglA-regulated genes reveals novel virulence factors in Francisella tularensis. Infect. Immun. 74:6642– 6655. http://dx .doi.org/10.1128/IAI.01250-06. 14. Hager AJ, Bolton DL, Pelletier MR, Brittnacher MJ, Gallagher LA, Kaul R, Skerrett SJ, Miller SI, Guina T. 2006. Type IV pili-mediated secretion modulates Francisella virulence. Mol. Microbiol. 62:227–237. http://dx .doi.org/10.1111/j.1365-2958.2006.05365.x. 15. Ulland TK, Janowski AM, Buchan BW, Faron M, Cassel SL, Jones BD, Sutterwala FS. 2013. Francisella tularensis live vaccine strain folate metabolism and pseudouridine synthase gene mutants modulate macrophage caspase-1 activation. Infect. Immun. 81:201–208. http://dx.doi.org/10 .1128/IAI.00991-12. 16. Ulland TK, Buchan BW, Ketterer MR, Fernandes-Alnemri T, Meyerholz DK, Apicella MA, Alnemri ES, Jones BD, Nauseef WM, Sutterwala FS. 2010. Cutting edge: mutation of Francisella tularensis mviN leads to increased macrophage absent in melanoma 2 inflammasome activation and a loss of virulence. J. Immunol. 185:2670 –2674. http://dx.doi.org/10 .4049/jimmunol.1001610. 17. Huang MT, Mortensen BL, Taxman DJ, Craven RR, Taft-Benz S, Kijek TM, Fuller JR, Davis BK, Allen IC, Brickey WJ, Gris D, Wen H, Kawula TH, Ting JP. 2010. Deletion of ripA alleviates suppression of the inflammasome and MAPK by Francisella tularensis. J. Immunol. 185:5476 –5485. http://dx.doi.org/10.4049/jimmunol.1002154. 18. Okan NA, Chalabaev S, Kim TH, Fink A, Ross RA, Kasper DL. 2013. Kdo hydrolase is required for Francisella tularensis virulence and evasion of TLR2-mediated innate immunity. mBio 4:e00638 – 00612. 19. Lindemann SR, Peng K, Long ME, Hunt JR, Apicella MA, Monack DM, Allen LA, Jones BD. 2011. Francisella tularensis Schu S4 O-antigen and capsule biosynthesis gene mutants induce early cell death in human macrophages. Infect. Immun. 79:581–594. http://dx.doi.org/10.1128/IAI .00863-10. 20. Peng K, Broz P, Jones J, Joubert LM, Monack D. 2011. Elevated AIM2mediated pyroptosis triggered by hypercytotoxic Francisella mutant strains is attributed to increased intracellular bacteriolysis. Cell Microbiol. 13: 1586 –1600. http://dx.doi.org/10.1111/j.1462-5822.2011.01643.x. 21. Singh A, Rahman T, Malik M, Hickey AJ, Leifer CA, Hazlett KR, Sellati TJ. 2013. Discordant results obtained with Francisella tularensis during in vitro and in vivo immunological studies are attributable to compromised bacterial structural integrity. PLoS One 8:e58513. http://dx.doi.org/10 .1371/journal.pone.0058513. 22. Dabo SM, Confer AW, Quijano-Blas RA. 2003. Molecular and immunological characterization of Pasteurella multocida serotype A:3 OmpA: evidence of its role in P. multocida interaction with extracellular matrix molecules. Microb. Pathog. 35:147–157. 23. Ristow P, Bourhy P, da Cruz McBride FW, Figueira CP, Huerre M, Ave P, Girons IS, Ko AI, Picardeau M. 2007. The OmpA-like protein Loa22 is essential for leptospiral virulence. PLoS Pathog. 3:e97. http://dx.doi.org /10.1371/journal.ppat.0030097. 24. Serino L, Nesta B, Leuzzi R, Fontana MR, Monaci E, Mocca BT, Cartocci E, Masignani V, Jerse AE, Rappuoli R, Pizza M. 2007. Identification of a new OmpA-like protein in Neisseria gonorrhoeae involved in the binding to human epithelial cells and in vivo colonization. Mol. Microbiol. 64:1391–1403. http://dx.doi.org/10.1111/j.1365-2958.2007 .05745.x. 25. Wang Y. 2002. The function of OmpA in Escherichia coli. Biochem. Biophys. Res. Commun. 292:396 – 401. http://dx.doi.org/10.1006/bbrc.2002 .6657. 26. Bouveret E, Benedetti H, Rigal A, Loret E, Lazdunski C. 1999. In vitro

iai.asm.org 2947


cal during slow-growth conditions (e.g., 30°C) when alternate cell wall contacts possibly via other OmpA motif-containing proteins (i.e., Pal and FopA) might suffice. Further, given that growth in macrophages resulted in similar altered bacterial morphology followed by subsequent intracellular destruction in late-forming LAMP-1-positive vacuoles, it is possible that autophagy maybe the principal clearance and or detection mechanism for structurally compromised F. tularensis Schu S4. This proposal is consistent with recent observations that nonviable mutants are captured within LAMP-1-positive autophagosomes as a clearance mechanism for damaged cytosolic Francisella (44). However, based on our data alone, we cannot exclude the possibility that this instead results from hypercytotoxicity of the Schu S4 ⌬0831 mutant toward macrophages in vitro and the release and subsequent secondary engulfment of gentamicin-killed bacteria. Further studies will seek to define the role of autophagy and TLR2 in this response and also the mechanistic basis for the cell division defects associated with loss of FTT0831c/FTL_0325 at physiologic, but not lower, growth temperatures. Such studies will be particularly informative in understanding the role of this protein in biology and cell cytokinesis of this intracellular pathogen.

Robertson et al.

27.

28.

29.

31.

32.

33.

34. 35. 36.

37.

38.

2948

iai.asm.org

39. 40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

for Francisella tularensis genetics. FEMS Microbiol. Lett. 291:95–102. http: //dx.doi.org/10.1111/j.1574-6968.2008.01440.x. Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9:671– 675. http://dx.doi.org/10 .1038/nmeth.2089. Chong A, Wehrly TD, Nair V, Fischer ER, Barker JR, Klose KE, Celli J. 2008. The early phagosomal stage of Francisella tularensis determines optimal phagosomal escape and Francisella pathogenicity island protein expression. Infect. Immun. 76:5488 –5499. http://dx.doi.org/10.1128/IAI .00682-08. Radolf JD, Chamberlain NR, Clausell A, Norgard MV. 1988. Identification and localization of integral membrane proteins of virulent Treponema pallidum subsp. pallidum by phase partitioning with the nonionic detergent triton X-114. Infect. Immun. 56:490 – 498. Dotson RJ, Rabadi SM, Westcott EL, Bradley S, Catlett SV, Banik S, Harton JA, Bakshi CS, Malik M. 2013. Repression of inflammasome by Francisella tularensis during early stages of infection. J. Biol. Chem. 288: 23844 –23857. http://dx.doi.org/10.1074/jbc.M113.490086. Checroun C, Wehrly TD, Fischer ER, Hayes SF, Celli J. 2006. Autophagy-mediated reentry of Francisella tularensis into the endocytic compartment after cytoplasmic replication. Proc. Natl. Acad. Sci. U. S. A. 103:14578 –14583. http://dx.doi.org/10.1073/pnas.0601838103. Chong A, Wehrly TD, Child R, Hansen B, Hwang S, Virgin HW, Celli J. 2012. Cytosolic clearance of replication-deficient mutants reveals Francisella tularensis interactions with the autophagic pathway. Autophagy 8:1342–1356. http://dx.doi.org/10.4161/auto.20808. Kanistanon D, Hajjar AM, Pelletier MR, Gallagher LA, Kalhorn T, Shaffer SA, Goodlett DR, Rohmer L, Brittnacher MJ, Skerrett SJ, Ernst RK. 2008. A Francisella mutant in lipid A carbohydrate modification elicits protective immunity. PLoS Pathog. 4:e24. http://dx.doi.org/10.1371 /journal.ppat.0040024. Pyles RB, Jezek GE, Eaves-Pyles TD. 2010. Toll-like receptor 3 agonist protection against experimental Francisella tularensis respiratory tract infection. Infect. Immun. 78:1700 –1710. http://dx.doi.org/10.1128/IAI .00736-09. Rozak DA, Gelhaus HC, Smith M, Zadeh M, Huzella L, Waag D, Adamovicz JJ. 2010. CpG oligodeoxyribonucleotides protect mice from Burkholderia pseudomallei but not Francisella tularensis Schu S4 aerosols. J. Immune Based Ther. Vaccines 8:2. http://dx.doi.org/10.1186/1476 -8518-8-2. West TE, Pelletier MR, Majure MC, Lembo A, Hajjar AM, Skerrett SJ. 2008. Inhalation of Francisella novicida ⌬mglA causes replicative infection that elicits innate and adaptive responses but is not protective against invasive pneumonic tularemia. Microbes Infect. 10:773–780. http://dx.doi .org/10.1016/j.micinf.2008.04.008. Lewis K. 2000. Programmed death in bacteria. Microbiol. Mol. Biol. Rev. 64:503–514. http://dx.doi.org/10.1128/MMBR.64.3.503-514.2000.



30.

characterization of peptidoglycan-associated lipoprotein (PAL)peptidoglycan and PAL-TolB interactions. J. Bacteriol. 181:6306 – 6311. Cascales E, Lloubes R. 2004. Deletion analyses of the peptidoglycanassociated lipoprotein Pal reveals three independent binding sequences, including a TolA box. Mol. Microbiol. 51:873– 885. Parsons LM, Lin F, Orban J. 2006. Peptidoglycan recognition by Pal, an outer membrane lipoprotein. Biochemistry 45:2122–2128. http://dx.doi .org/10.1021/bi052227i. Gerding MA, Ogata Y, Pecora ND, Niki H, de Boer PA. 2007. The trans-envelope Tol-Pal complex is part of the cell division machinery and required for proper outer-membrane invagination during cell constriction in Escherichia coli. Mol. Microbiol. 63:1008 –1025. http://dx.doi.org /10.1111/j.1365-2958.2006.05571.x. Mahawar M, Atianand MK, Dotson RJ, Mora V, Rabadi SM, Metzger DW, Huntley JF, Harton JA, Malik M, Bakshi CS. 2012. Identification of a novel Francisella tularensis factor required for intramacrophage survival and subversion of innate immune response. J. Biol. Chem. 287: 25216 –25229. http://dx.doi.org/10.1074/jbc.M112.367672. Mahawar M, Rabadi SM, Banik S, Catlett SV, Metzger DW, Malik M, Bakshi CS. 2013. Identification of a live attenuated vaccine candidate for tularemia prophylaxis. PLoS One 8:e61539. http://dx.doi.org/10.1371 /journal.pone.0061539. Robertson GT, Child R, Ingle C, Celli J, Norgard MV. 2013. IglE is an outer membrane-associated lipoprotein essential for intracellular survival and murine virulence of type A Francisella tularensis. Infect. Immun. 81: 4026 – 4040. http://dx.doi.org/10.1128/IAI.00595-13. Hazlett KR, Caldon SD, McArthur DG, Cirillo KA, Kirimanjeswara GS, Magguilli ML, Malik M, Shah A, Broderick S, Golovliov I, Metzger DW, Rajan K, Sellati TJ, Loegering DJ. 2008. Adaptation of Francisella tularensis to the mammalian environment is governed by cues which can be mimicked in vitro. Infect. Immun. 76:4479 – 4488. http://dx.doi.org/10 .1128/IAI.00610-08. Horton RM. 1995. PCR-mediated recombination and mutagenesis. SOEing together tailor-made genes. Mol. Biotechnol. 3:93–99. Gallagher LA, McKevitt M, Ramage ER, Manoil C. 2008. Genetic dissection of the Francisella novicida restriction barrier. J. Bacteriol. 190: 7830 –7837. http://dx.doi.org/10.1128/JB.01188-08. Choi KH, Gaynor JB, White KG, Lopez C, Bosio CM, KarkhoffSchweizer RR, Schweizer HP. 2005. A Tn7-based broad-range bacterial cloning and expression system. Nat. Methods 2:443– 448. http://dx.doi .org/10.1038/nmeth765. LoVullo ED, Molins-Schneekloth CR, Schweizer HP, Pavelka MS, Jr. 2009. Single-copy chromosomal integration systems for Francisella tularensis. Microbiology 155:1152–1163. http://dx.doi.org/10.1099/mic.0 .022491-0. LoVullo ED, Sherrill LA, Pavelka MS, Jr. 2009. Improved shuttle vectors

B-cell epitopes in GroEL of Francisella tularensis.

Lipidation of the FPI protein IglE contributes to Francisella tularensis ssp. novicida intramacrophage replication and virulence.

Comparative review of Francisella tularensis and Francisella novicida.

Comparative Transcriptional Analyses of Francisella tularensis and Francisella novicida.

Virulence factors of Francisella tularensis.

Acid resistance in Francisella tularensis.

Control of intracellular Francisella tularensis by different cell types and the role of nitric oxide.

B cell subsets are activated and produce cytokines during early phases of Francisella tularensis LVS infection.

Structure and function of REP34 implicates carboxypeptidase activity in Francisella tularensis host cell invasion.

Single-Cell Lineage Tracing Reveals that Oriented Cell Division Contributes to Trabecular Morphogenesis and Regional Specification.

cByJ Mice.

Iron and Virulence in Francisella tularensis.

Francisella tularensis Modulates a Distinct Subset of Regulatory Factors and Sustains Mitochondrial Integrity to Impair Human Neutrophil Apoptosis.

NF-κB signaling contributes to glioblastoma stem cell maintenance.

Cell division and the maintenance of epithelial order.

Francisella tularensis intracellular survival: To eat or to die.

Francisella tularensis subsp. tularensis group A.I, United States.

TolC-dependent modulation of host cell death by the Francisella tularensis live vaccine strain.

Chromosome replication, cell growth, division and shape: a personal perspective.

Cellular fatty acid composition of Francisella tularensis.

Mechanisms of heme utilization by Francisella tularensis.

Biochemical and structural characterization of polyphosphate kinase 2 from the intracellular pathogen Francisella tularensis.

Hare-to-human transmission of Francisella tularensis subsp. holarctica, Germany.

Non-muscle tropomyosin (Tpm3) is crucial for asymmetric cell division and maintenance of cortical integrity in mouse oocytes.