Genome sequence and phenotypic analysis of a first German Francisella sp. isolate (W12-1067) not belonging to the species Francisella tularensis

Background Francisella isolates from patients suffering from tularemia in Germany are generally strains of the species F. tularensis subsp. holarctica. To our knowledge, no other Francisella species are known for Germany. Recently, a new Francisella species could be isolated from a water reservoir of a cooling tower in Germany. Results We identified a Francisella sp. (isolate W12-1067) whose 16S rDNA is 99% identical to the respective nucleotide sequence of the recently published strain F. guangzhouensis. The overall sequence identity of the fopA, gyrA, rpoA, groEL, sdhA and dnaK genes is only 89%, indicating that strain W12-1067 is not identical to F. guangzhouensis. W12-1067 was isolated from a water reservoir of a cooling tower of a hospital in Germany. The growth optimum of the isolate is approximately 30°C, it can grow in the presence of 4–5% NaCl (halotolerant) and is able to grow without additional cysteine within the medium. The strain was able to replicate within a mouse-derived macrophage-like cell line. The whole genome of the strain was sequenced (~1.7 mbp, 32.2% G + C content) and the draft genome was annotated. Various virulence genes common to the genus Francisella are present, but the Francisella pathogenicity island (FPI) is missing. However, another putative type-VI secretion system is present within the genome of strain W12-1067. Conclusions Isolate W12-1067 is closely related to the recently described F. guangzhouensis species and it replicates within eukaryotic host cells. Since W12-1067 exhibits a putative new type-VI secretion system and F. tularensis subsp. holarctica was found not to be the sole species in Germany, the new isolate is an interesting species to be analyzed in more detail. Further research is needed to investigate the epidemiology, ecology and pathogenicity of Francisella species present in Germany.


Background
Francisella tularensis is an facultative intracellular pathogen that causes tularemia in humans and a wide range of animals [1]. Strains of F. tularensis subsp. (Ft.) tularensis can be lethal to humans and are mostly associated with cases of tularemia in the U.S.A. Doses as low as 10-20 bacteria can be infective [1]. Transmission mostly occurs via aerosol, alimentary ingestion or skin inoculation. In addition, F. tularensis is suspected as a potential bacterial biological weapon [2]. The species Ft. novicida is almost avirulent for humans in contrast to mice, and is thought to be an opportunistic pathogen [3,4]. Ft. novicida is assumed to constitute an environmental lineage along with F. philomiragia. In rare cases the latter has also been associated with human disease in immunocompromised individuals [4,5].
However, to our knowledge, no species other than Ft. holarctica has been identified in Germany until now. Therefore, our new isolate W12-1067 is the first aquatic isolate identified in Germany which does not belong to the species F. tularensis and is closest related to F. guangzhouensis.

Phenotypic assays
Growth without additional cysteine was done on BCYE agar plates without additional cysteine. Physiological characteristics of analyzed strains were determined by using API ZYM (bioMérieux Deutschland GmbH, Nürtingen, Germany).
Chitinase activity tests were done on 1.5% agarose plates containing 0.1% deacetylated glycol chitin. Chitinase activity experiments were done as described earlier [23], with the modification that the deacetylated glycol chitin was suspended in 0.01 M sodium phosphate (pH 5.5) by heating. In short, bacteria were grown in medium T overnight. The supernatant was precipitated by isopropanol and the protein pellet was resuspended in PBS (concentrated 40-fold). 50 μl were inoculated into agar plates (as described earlier [24]) and incubated for two days at 37°C. Degrading activity was visualized by incubation with 0.01% Calcofluor Brightener 28 (Sigma-Aldrich Chemie GmbH, Munich, Germany) for 10 min, washing two times with water and then incubation at room temperature (RT) overnight.
The NaCl sensitivity assays were done in 96-well plates in a total volume of 200 μl of medium T and~2 × 10 7 bacterial cells. Plates were incubated 2-3 days at 37°C and 5% CO 2 , and then the optical density (OD) at 600 nm was measured using an Infinite 200 reader (Tecan Deutschland GmbH, Crailsheim, Germany).

Intracellular multiplication of Francisella in host cells
To determine which amoeba strain may be suitable to be used for replication assays, isolate W12-1067 (~10 10 cells) was suspended in 1 ml of dH 2 O and 100 μl were plated onto NN-agar plates (14 g/l of agar in dH 2 O). The amoeba strain (15 μl, A. castellanii ATCC 30010, A. castellanii ATCC 30234, A. castellanii 50739, A. lenticulata 45 ATCC 50703, A. lenticulata 118 ATCC 50706, Hartmannella vermiformis OS101, Hartmannella vermiformis ATCC 50256 and Naegleria gruberi ATCC 30244, respectively) was dropped onto the centre of the plates and incubated at RT or at 30°C for 7 days. The plates were inspected daily for movement and replication of the amoeba. All amoeba tested were motile and not killed by isolate W12-1067. Therefore, no further experiments were performed using amoebae as host cells.
For differentiation into macrophage-like cells, U937 cells were adjusted to 3 × 10 5 cells/ml and transferred into 100 ml of fresh RPMI medium containing 10% fetal calf serum (10% FCS), and PMA (phorbol-12-myristate-13-acetate, 1 mg/ml in dH 2 O [P-8139; Sigma-Aldrich Chemie]) was added at a concentration of 1:20,000. After incubation for 36 h at 37°C and 5% CO 2 , the supernatant was discarded and adherent cells were washed once with 10 ml of 0.2% EDTA in PBS. Cells were mechanically detached from the flask bottom with RPMI + 10% FCS, transferred into 50 ml tubes and centrifuged at 800 g for 10 min. All cells were counted after trypan blue staining in a Neubauer counting chamber and adjusted to 5 × 10 5 cells/ml with RPMI + 10% FCS. To each well of a 24-well plate 1 ml of the cell suspension was added and incubated for adhesion during 2 h at 37°C and 5% CO 2 . Macrophage-like mouse cell line J774A.1 was also grown in fresh RPMI medium containing 10% fetal calf serum and treated as described above, but without the differentiation step (no PMA treatment).
For both cell lines, stationary phase bacteria grown for 3 days on CHA or BCYE agar were diluted in plain RPMI medium and the infection was done with a multiplicity of infection (MOI) of 1, 10 or 100 (time point 0 h) for 2 h at 37°C and 5% CO 2 . Cells were washed three times with RPMI and incubated with 50 μg/ml of Gentamycin for 1 h to kill extracellular bacteria. Cells were washed again three times with RPMI and covered with 1 ml of RPMI + 10% FCS. For colony-forming unit (CFU) determination at various time points of infection, coincubations of cells and bacteria were lysed by addition of 10 μl of 10% Saponin (S4521, Sigma-Aldrich Chemie) for 5 min, and serial dilutions were plated on BCYE agar. In a control experiment we showed that Saponin treatment did not affect the number of remaining CFU of strain W12-1067 (data not shown).

Electron microscopy (EM)
J774A.1 cells were infected with Francisella strain W12-1067 (MOI of 10) at 37°C as described above. Cells were fixed 96 h post infection with 2.5% glutaraldehyde in 0.05 M HEPES buffer. Bacteria cultivated in medium T were fixed with 4% paraformaldehyde 5% glutaraldehyde in 0.05 M HEPES buffer for 2 h at RT. All samples were post-fixed with osmium tetroxide (1% in distilled water) and uranyl acetate (2% in distilled water), dehydrated stepwise in a graded ethanol series and embedded in LR White resin (Science Services GmbH, Munich, Germany) which was polymerized at 60°C overnight. Thin sections were prepared with an ultramicrotome (UC-T; Leica, Vienna, Austria) and counterstained with uranyl acetate and lead citrate.
Samples were examined using a transmission electron microscope (EM 902; Carl Zeiss Microscopy GmbH, Jena, Germany) at 80 kV, and the images were recorded using a slow-scan charge-coupled-device camera (Pro Scan elektronische Systeme GmbH, Lagerlechfeld, Germany).

Genome sequencing, ORF finding and annotation
Genome sequencing of chromosomal DNA of isolate W12-1067 was performed by Eurofins MWG Operon (Eurofins Medigenomix GmbH, Ebersberg, Germany): (i) Short insert shotgun library (FLX + library): 1 μg of DNA was fragmented using a Covaris E210 instrument (Covaris Inc., Woburn, MA) according to manufacturer's instructions. End-repair, dA-tailing and ligation of barcoded adapter were performed following New England Biolabs' instructions (New England Biolabs GmbH, Frankfurt/ Main, Germany). Emulsion-based clonal amplification (emPCR amplification) was performed following Roche's instructions (Roche Diagnostics GmbH, Mannheim, Germany). FLX + Sequencing: Sequencing was performed on an FLX + platform according the manufacturer's instructions using 1/8 plate. The sequencing process was controlled by the Roche 454 software gsRunProcessor v2.8 (shotgun signal processing pipeline). (ii) 8 kb mate-pairlike library (Long-Jumping-Distance library): Creation of the 8 kb mate-pair-like library was done at Eurofins MWG Operon (Ebersberg, Germany) using their proprietary protocol. (iii) Illumina Sequencing: For sequencing, the library was loaded on an Illumina MiSeq machine. Cluster generation and paired-end sequencing was performed using the manufacturer's instructions. MiSeq Control Software 2.2.0 was used for sequencing. For processing of raw data RTA version 1.17.28 and CASAVA 1.8.2 were used to generate FASTQ-files. (iv) Data analysis: A two-step hybrid de novo assembly was conducted using the sequencing data of the two libraries. First, the FLX + data (long reds, single-end) has been assembled separately using the Roche 454 software Newbler (v2.6). The resulting contigs as well as the Illumina longjumping-distance pairs (Illumina mate-pair-like) were then assembled together with a hardware-accelerated assembly pipeline based on the Convey hardware and software tools (http://www.conveycomputer.com) that mimic a standard de novo assembly using the Velvet assembler (Eurofins proprietary assembly pipeline) [25,26]. The draft genome (all contigs) was annotated by using the RAST server, freely available at www.patricbrc.org [27].

Phylogenetic analysis
16S rDNA gene and the multi-gene locus (in frame gene sequence) of genes fopA, gyrA, rpoA, groEL, sdhA and dnaK of strain W12-1067 and available homologous sequences from Francisella species and L. pneumophila Paris (obtained from GenBank) were used for nucleotide comparison. The multi-gene locus of L. pneumophila Paris exhibited no fopA gene because no homolog is present within the genome. Phylogenetic analysis (phylogenetic tree) was generated by using the ClustalW program (ClustalW 2.1; www.patricbrc.org).

Accession number
This whole genome shotgun project has been deposited at DDB/EMBL/GenBank under accession AWHF00 000000. The version described in this paper is version [AWHF01000000].

Results and discussion
During 2012 health authorities of the city of Heilbronn (Germany) observed some coincident spots of Legionnaires' disease (LD). Therefore, different putative sources were screened for the presence of Legionella species using GVPC (glycine-vancomycin-polymyxin-cycloheximide) agar plates. The investigation of a water reservoir of a cooling tower led to the isolation of strain W12-1067. The isolate W12-1067 was first thought to belong to the genus Legionella because of its habitus and growth on GVPC agar plates, but PCR analysis did not support this finding. Preliminary 16S rDNA PCR analysis performed by the German Consultant Laboratory for Legionella (Dresden, Germany) revealed that this strain may belong to the genus Francisella. The identified isolate was not involved in the LD outbreak. However, the strain was send for further analysis to the Centre for Biological Threats and Special Pathogens at the Robert Koch Institute (Berlin, Germany). Here, W12-1067 was identified to be the first German isolate of the genus Francisella which did not belong to the species F. tularensis.

Analysis of strain W12-1067
First we performed 16S rDNA PCR and sequenced the PCR product. The phylogenetic analysis of the 16S rDNA revealed that isolate W12-1067 is a close representative of the recently identified new environmental Francisella species F. guangzhouensis [20]. The phylogenetic tree of 16S rDNA of different Francisella strains is given in Figure 1A, corroborating the close relationship of isolate W12-1067 with F. guangzhouensis and other Chinese cooling tower isolates of this species (99% identity). The other Francisella strains analyzed revealed DNA identities of 16S rDNA sequences of 94-95% with 16S rDNA of isolate W12-1067 and 83% identity with the 16S rRNA gene of L. pneumophila Paris.
We then sequenced the whole genome of W12-1067, resulting in an annotated draft genome of this strain. We used the sequences of the genes fopA, gyrA, rpoA, groEL, sdhA and dnaK (see Table 1) to build a multigene locus (9918 bp for strain W12-1067) for each strain analyzed. With these 6-loci concatenated sequences we performed a further phylogenetic analysis ( Figure 1B). The overall DNA identity was only 89%, indicating that W12-1067 is not identical to strain F. guangzhouensis, but that F. guangzhouensis is the closest relative identified to date. The overall DNA identity of the gene cluster of isolate W12-1067 to other Francisella strains analyzed was 80-81% and that to L. pneumophila Paris 67% ( Figure 1B).
Similar to strain W12-1067, F. guangzhouensis strains had been isolated from water of air conditioning systems of cooling towers in China, during a routine investigation to detect Legionella [20]. The growth optimum of this species ranged between 25 and 28°C and it showed growth on BCYE-alpha (minus cysteine) Legionella-agar plates. Furthermore, no virulence to mice was found for this strain [20,28]. No further information about virulence properties of this species was available yet.
sp. strain W12-1067 grew well on BCYE, GVPC and CHA plates. The isolate grew faster in medium T than in AYE medium, and it did not grow in the cell culture medium RPMI (data not shown). Whereas L-cysteine within BCYE agar plates is necessary for the growth of L. pneumophila Paris, it stimulates the growth of F. philomiragia, and the growth of strain W12-1067 was nearly similar with or without additional cysteine (see Additional file 1: Figure S1A). Growth in medium T revealed that the growth optimum of strain W12-1067 is about 30°C (Figure 2A), but growth of strains F. philomiragia and Ft. novicida in general was faster and reached a higher cell density compared to that of the growth of strain W12-1067 ( Figure 2B-D). Growth of strain W12-1067 in media with NaCl was reduced from a concentration of 4-5% NaCl, which was comparable to growth of strains F. philomiragia and Ft. novicida (up to 6-7% NaCl), but more resistant than the non-halotolerant strain Ft. holarctica LVS (see Additional file 1: Figure S1B). Using the API ZYM assay kit (bioMérieux), W12-1067 showed a profile similar to F. guangzhouensis (data not shown). In contrast to F. guangzhouensis, isolate W12-1067 was negative for alkaline phosphatase activity and showed only very low activity of the acid phosphatase, which is in good agreement with only one putative phosphatase encoding gene (peg_768) present within the genome sequence (see below).
These experiments were followed by co-culture studies using macrophage-like cell lines of human (U937) or mice (J774A.1) origin as host cells. We found that strain W12-1067 was able to persist within U937 cells (data not shown), and it replicated intracellularly in J774A.1 cells (Figure 3). Replication within J774A.1 was slower than that of L. pneumophila Paris. However, W12-1067 multiplied about 16-fold within 4 days of co-incubation, with a 24 h lag-phase at the beginning of the infection (Figure 3), indicating that the new isolate is able to infect and multiply within eukaryotic cells. Cells of strain W12-1067 grown in medium T at 37°C possessed a rodshaped, slightly pleomorphic morphology ( Figure 4A).  EM analysis of the infection assay revealed that strain W12-1067 was localized intracellularly within a vacuole 96 h after infection ( Figure 4B-D). The data indicate that W12-1067 is able to replicate intracellularly, but from EM analysis it was not clear yet whether the bacteria were able to escape from the vacuole. Experiments are under way to investigate this question further.
Whole genome sequencing and sequence analysis of strain W12-1067 General features The genomic DNA was sequenced using 454 shotgun reads and paired-end Illumina reads (for details, see Methods), followed by de novo assembly of the sequence, which was done by Eurofins (Ebersberg, Germany), resulting in 73 contigs and a chromosomal size of approximately 1,704,745 bp. The draft genome was annotated by using the RAST server (freely available at http://rast.nmpdr.otg). All genes encoded by the draft genome sequence of isolate W12-1067 (GenBank AWHF01000000) are given in the Additional file 2: Table S1. The draft genome exhibits 1,541 protein-coding genes (peg), three copies of the 16S-23S-5S rRNA gene locus (plus one additional 5S rRNA gene) and 37 tRNA genes ( Table 2). The G + C content is 32.2% and therefore not significantly different from various other Francisella strains already sequenced [12,[29][30][31][32][33][34][35][36][37]. However, the number of protein-coding genes and of tRNA genes seemed to be lower compared with other Francisella strains (  Figure S2. The genome sequence of strain F. guangzhouensis 08HL01032 T was not yet available for a comparative analysis, but a G + C content of about 32.5% of its draft genome was reported [20].

Mobile elements
In F. tularensis strains six different insertion sequences (IS) elements (ISFtu1-ISFtu6) have been described. IS elements are important elements of F. tularensis genomes and are thought to be generally stable among different isolates [36,40]. Within the genome of Francisella isolate W12-1067, we identified several mobile elements and transposases (Table 3). For the element ISFw3 (1,177 bp in length) we identified 18 copies within the genome sequence. The element encodes a putative integrase of 359 amino acids with 70% amino acid identity to the protein encoded by the IS481 element of Wolbachia endosymbiont of Drosophila simulans. All other elements are only present as a single copy. There is only one element (ISFw7) with significant similarity to one of the ISFtu elements of F. tularensis. ISFw7 encodes a putative transposase exhibiting 78% amino acid identity to ISFtu1 of Ft. holarctica strains. Of the 14 mobile elements identified, six (ISFw4, 5, 7, 9, 10 and peg_1255) exhibit homologs within Francisella strains (Table 3).

Virulence factors and secretion systems
A number of genes involved in Francisella pathogenesis have been identified in various different studies, including in vivo negative selection screens of transposon mutant libraries of Francisella strains [41][42][43][44], reviewed in [45] and [46]. We therefore looked for homologs of these virulence genes within the genome sequence of Francisella strain W12-1067. Some of them are shown in Table 4, and others, like LPS, capsule, type IV pili or FPI-associated genes, are described below. Various known virulence factors of Francisella exhibit homologs in strain W12-1067. For the most important virulence factors of F. tularensis, encoded by the genes of the Francisella Pathogenicity Island (FPI), no close homologs could be identified within the genome. However, we identified an FPI-like island at the genome sequence which will be discussed later. Interestingly, the genome of isolate W12-1067 exhibit six different chitinases of different sizes and putative location (cytoplasmic, periplasmic or extracellular) (Table 4), which was in good agreement with the phenotypically identified chitinase activity within the supernatant of Francisella W12-1067 cells (see Additional file 1: Figure S1C). Four chitinases (peg_818, 1009, 1044 and 1477) exhibit a signal peptide and therefore could be secreted by the general secretion system (Sec). Two of the chitinases (peg_0816 and peg_0818) are separated by a gene encoding a putative DNA-invertase (peg_0817). We could  not identify a homolog of this protein in the available genome sequences of Francisella strains. We were unable to determine whether the invertase is involved in DNA inversion in Francisella and whether this may influence the expression of the nearby chitinase genes. Chitinase peg_0816 is 76% identical to chitinase peg_0818. Both chitinases exhibit a homolog in F. philomiragia 25017 (Fphi_0512) and 25015, but not in the other Francisella strains, whereas homologs of the chitinases 1, 4 and 5 could be found in the genomes of Ft. tularensis, F. holarctica and F. philomiragia. Chitinases (FTN_0627 and FTN_1744) have been detected as essential virulence factors of Ft. novicida and for biofilm formation [42,44,47]. In L. pneumophila a chitinase was also found to be involved in the infectivity of Legionella for mice [48]. It would be interesting to further analyze the role of the different chitinases in strain Francisella W12-1067.
Furthermore, we identified three hypothetical proteins (peg_523, 567 and 1109) containing ankyrin-repeat domains. These proteins did not exhibit significant homology to any known protein (Table 4). For L. pneumophila it was shown that ankyrin-repeat-containing proteins are involved in the pathogen-host interplay during intracellular replication [49,50]. Furthermore, we looked for genes encoding putative antibiotic resistance proteins. We identified ten putative multidrug resistance proteins (peg_126, 152, 681, 682,  683, 764, 1134, 1430, 1444 and 1445). In addition, we identified a Chloramphenicol acetyltransferase (peg_183), a Chloramphenicol phosphotransferase (peg_1413) and a Streptomycin-6 kinase (peg_1416) without a homolog in any of the available Francisella genomes. We performed growth inhibition experiments with the isolate W12-1067 and found that levels of resistance to erythromycin, chloramphenicol and streptomycin were comparable to those of F. philomiragia (data not shown). We also identified a putative Acriflavin resistance protein (peg_810) exhibiting 85% amino acid identity to the AcrB protein (Fphi_1007) of F. philomiragia 25017.
Since we identified putative signal peptides at the N-terminus of some of the chitinases and chitinase activity within the supernatant, we searched the genome sequence for genes of the general Sec system. The identified proteins are given in the Additional file 4: Table S2. We identified all proteins necessary for a putative functional Sec system plus two signal sequence recognition proteins (SRP) and three different signal peptidases. Therefore, strain W12-1067 seems to encode a functional Sec system for the transport of proteins across the inner membrane. We could not detect genes encoding homologs of a type-II secretion system (T2SS), but proteins (HlyB, HlyD, TolC2) of a putative T1SS (Table S2). We were also able to identify a putative functional Tol-Pal system generally involved in vitamin B12 or colicin translocation, but also in capsule synthesis and outer membrane vesicle formation [51,52]. The presence of a putative T6-like SS will be discussed in the next section.
A new putative homolog of the Francisella pathogenicity island (FPI) and regulatory proteins T6SS are widely distributed amongst diverse Gramnegative species. It is a complex molecular machine which injects effector proteins to target cells or bacteria [53]. T6SS are involved in virulence and in eukaryotic cell targeting. They are also reported to have antibacterial activity [54,55]. Most systems are able to function "antieukaryotically" or "anti-bacterially", and one system has been reported to be able to do both [53,56,57]. It was also proposed that T6SS may target and defend predatory eukaryotes in the environment [58]. For an overview see [53]. Francisella strains exhibit an FPI encoding a functional T6SS, needed for preventing phagolysosomal fusion and escape into the cytosol, which is therefore essential for intracellular replication and pathogenicity of Francisella [59][60][61][62][63]. The island is~33 kb in length, encodes for 15-19 open reading frames (ORFs) and is present in Ft. tularensis strains, Ft. novicida and F. philomiragia ( Figure 5A). In strains of F. tularensis this island is duplicated [36,[64][65][66]. However, it is still unclear yet whether both copies are needed for full virulence of these strains. Genes of the locus were named igl (intracellular growth locus, [66]) and pdp (pathogenicity determinant proteins, [65]). The complete locus was identified in 2004, and it seems to be acquired via horizontal gene transfer because of a lower G + C content compared with the core genome [60,64,65].
We were not able to identify a close homolog of the FPI within the genome of strain W12-1067. However, we identified two FPI-like clusters which seem to encode a further putative T6SS ( Figure 5). Clusters I ( Figure 5B) and II ( Figure 5C) each code for 16 ORFs. Homologs of iglABC, dotU and pdbB are present, and the whole island I is similar to an FPI-like element (FTN_0038-0054) recently identified within the genome of Ft. novicida, but not yet discussed further [60,64]. Within the genome sequence, island I is also localized between the pyrD and tyrA genes and island II between rpsU and glmS ( Figure 5B and C). All genes, features of the ORFs and closest homologs of both systems identified within the genome of W12-1067 are given in Table 5. In a recently published model of an FPI-encoded T6SS-like apparatus IglAB (TssBC), IglC (TssD), DotU (TssL), PdpB (TssM) and VgrG (TssI) were identified as core proteins of the T6SS [67]. We identified similar proteins within the identified loci of strain W12-1067. In strain F-TX077308, the FPI-like gene cluster is localized at a different site within its genome sequence ( Figure 5B). Altogether these findings led us to hypothesize that these loci may encode a further T6-like SS in Francisella strain W12-1067. As mentioned above, infection assays and EM analysis indicateed that strain W12-1067 is able to replicate intracellularly (Figures 3 and 4), which also suggest that there should be a functional T6SS in this strain. However, since the EM analysis did not yet clearly demonstrate whether the strain is released into the cytosol, further experimental data are needed to verify this hypothesis. Various regulatory proteins are known for Francisella, and the regulators MglAB, SspA, PmrA, FevR, MigR and Hfq were identified to be involved in the regulation of genes of the FPI [59,61,[68][69][70][71][72][73][74][75][76]. In F. tularensis the genes of the FPI are upregulated during intracellular growth within macrophages [61,68,[77][78][79][80]. In the genome of strain W12-1067 we identified only homologs of MglAB and Hfq, but also two further response regulator proteins (OmpR1 and R2) as well as two sigma factors (Sig-70 and Sig-32) and homologs of IscR, ArsR, Crp and Fur (see Additional file 5: Table S3). In addition, proteins involved in the stringent response could be identified (SspB, SpoT and RelA).
Surface structures: The wbt locus, capsule and type IV pili (i) LPS. The lipid A core portion of the LPS anchors the lipopolysaccharide structure to the outer membrane, whereas the O-polysaccharide chain is the predominant epitope recognized by the immune system and specifies antigenicity. Ft. tularensis subspecies-specific antisera have been generated and applied to antigen detection in F. tularensis [81]. Furthermore, LPS is used as an antigen in seroprevalence studies and for diagnostic of human tularemia [6,82,83]. However, the LPS produced by F. tularensis is less endotoxic compared to other Gram-negative bacteria, such as E. coli, a phenotype also known for L. pneumophila LPS [84][85][86]. Genes probably involved in the biosynthesis of the O-polysaccharide chain are given in supplementary Table 4 (see Additional file 6: Table S4). Obviously, there is one cluster of genes (peg_0636-0646) for which homologs were found in Ft. novicida and F. philomiragia, whereas for genes peg_0609-0615 and peg_0628-0631 the homologs were only found Figure 5 Genetic organization of T6 secretion systems in Francisella. Organization of the FPI island of Ft. novicida U112 (A) and of genomic islands I (B) and II (C) encoding putative T6-like secretion systems of Francisella isolate W12-1067. Genomic island I is integrated between genes pyrD and tyrA in W12-1067 and Ft. novicida U112, whereas it is integrated between F7308_1884 und F7308_1920 in Francisella strain TX077308 (F-TX077308). Genomic island II is integrated between rpsU and glmS in W12-1067 and is not present in Ft. novicida U112 and Francisella strain TX077308. Genes (ORFs) are indicated by arrows. Gene names are given below the genes, and the protein-encoding gene (peg) numbers are given in Table 5. Genes encoding homologous proteins are boxed in the same color, with the exception of pink (mobile elements) and black (conserved core genome genes).
in F. philomiragia or Ft. novicida U112, respectively. In addition, there is another cluster of genes (peg_0616-0627) which seems to have no homologs in F. tularensis or F. philomiragia, but homologs were identified in different bacterial species, as in Vibrio, Pseudomonas, Sulfurovum or Acinetobacter (see Additional file 6: Table S4). The LPS structure of strain W12-1067 has not yet been analyzed further. (ii) Capsule. Electron microscopy of strain W12-1067 grown on agar plates or in medium revealed the absence of a capsule ( Figure 4A and data not shown). However, we identified three ORFs encoding homologs of capBCA genes ( Table 4). The capBCA locus of Francisella is similar to determinants encoding the poly-gamma-glutamic capsule in Bacillus anthracis [87]. These genes have been shown to be essential for the virulence of F. tularensis in a murine model of tularemia [44,87]. Further experiments will be needed to analyze the role and structure of the capsule of Francisella strain W12-1067 and to identify conditions necessary for the putative induction of capsule gene expression. (iii)Type IV pili. Electron microscopy of strain W12-1067 did not show any pili on the surface of the bacteria grown in medium at 37°C ( Figure 4A) or on agar plates (data not shown). However, we identified homologs of the type IV pilus (Tfp) encoding loci of F. philomiragia ATCC 25017 in the genome sequence of W12-1067 (see Additional file 7: Table S5) [88]. We could not identify a PilA homolog, a homolog of the second PilW protein Fphi_0522 and of the two additional PilA/PilE pilus assembly proteins (Fphi_0424/0449). Tfp systems are known to be involved in bacterial pathogenesis, bacterial adhesion and twitching motility [89]. In L. pneumophila Tfp pili are required for twitching motility, natural competence, biofilm formation and are involved in the attachment to host cells [90][91][92][93].
In type II systems, the antitoxin (small unstable protein) sequesters the toxin (stable protein) through protein complex formation (reviewed in [99]). Peg_0599 encodes a protein (84 amino acids, putative toxin), exhibiting a Pfam_Plasmid_Txe domain and 67% amino acid identity to the YoeB toxin of Pleurocapsa sp. PCC7319. peg_0600 encodes the respective putative antitoxin (83 amino acids), exhibiting a Pfam_PhdYeFM (type II toxin-antitoxin) and shows 60% amino acid identity to the prevent-host-death protein of Methylocystis rosea. The second system is composed of protein Peg_0704 (96 amino acids), exhibiting a HTH-XRE motif and a HigA-antidote (VapI) domain and shows 65% amino acid identity to protein LLO_065 of Legionella longbeachae NSW150. The respective putative toxin (HigB, 97 amino acids) is encoded by peg_0705, exhibiting a Pfam_plasmid killer domain, and shows 62% amino acid identity to the plasmid maintenance system killer protein of Deferribacter desulfurricans SSM1. Both described systems seemed to have no homolog in the sequences of Francisella available so far and are localized on contig_34 of the draft genome of strain W12-1067.
The third system is localized on contig_46. peg_1296 encodes for a protein (85 amino acids, putative antitoxin), exhibiting a Pfam_PhdYeFM domain (type II toxin-antitoxin system), and shows 66% amino acid identity to the plasmid-encoded (pF243) protein F243_0001 of F. philomiragia ATCC 25017 and 65% identity to the Phd protein (pFNL10_p3) of Ft. novicida. The respective putative toxin (84 amino acids) is encoded by peg_1297, exhibits a Pfam_Plasmid_Txe (YoeB) domain and shows 79% amino acid identity to F243_0002 of F. philomiragia. Plasmid pF243 is 5,072 bp long and was predicted to encode six putative ORFs [100]. ORFs F243_0001 and F243_0002 are organized in an operon that is similar to the phd-doc post-segregation killing system operon of pFNL10 [100,101]. This post-segregation killing mechanism relies on the difference in stability of the antitoxin and toxin. In the daughter cells the plasmid-free bacteria will be killed by the activity of the toxin [102,103]. Chromosomally encoded toxin-antitoxin systems may stabilize chromosomal regions during evolution and seem to be involved in host regulatory networks or fitness advantages [102]. Less is known so far about toxin-antitoxin systems in Francisella, but they have been used to construct plasmids which are stable without a selective marker gene [101,104].