|Home | About | Journals | Submit | Contact Us | Français|
It was recently shown that B. melitensis is flagellated. However, the flagellar structure remains poorly described.
We analyzed the structure of the polar sheathed flagellum of B. melitensis by TEM analysis and demonstrated that the Ryu staining is a good method to quickly visualize the flagellum by optical microscopy. The TEM analysis demonstrated that an extension of the outer membrane surrounds a filament ending by a club-like structure. The ΔftcR, ΔfliF, ΔflgE and ΔfliC flagellar mutants still produce an empty sheath.
Our results demonstrate that the flagellum of B. melitensis has the characteristics of the sheathed flagella. Our results also suggest that the flagellar sheath production is not directly linked to the flagellar structure assembly and is not regulated by the FtcR master regulator.
Brucellae are gram-negative, intracellular pathogenic bacteria that cause brucellosis in a variety of mammals, including humans. During a long time, they were considered as unflagellated. However, the presence of a sheathed flagellum has recently been discovered in Brucella melitensis [1,2].
The fact that the flagellum was never previously observed in Brucella is probably due to the very short period (the early exponential growth phase in rich liquid medium) during which the flagellum is produced. The availability of the genome sequences of Brucella melitensis and Brucella suis revealed the presence of all the flagellar genes needed for the construction of a functional flagellum, except the genes encoding the chemotactic system [3,4]. Today, the sequence of the genome of B. abortus, B. ovis, B. canis and B. microti are available, showing that flagellar genes are also present in these species [5-7]. Interestingly, the presence of mutations in several flagellar genes in the different Brucella species suggest that the flagellar regulon contains pseudogenes and that the filamentous appendage observed is a flagellum-derived structure .
The flagellum is a complex machine divided in three main structures . The basal body structure is embedded in the membranes and functions like a tiny engine containing a motor, a stator and a secretion system. The hook structure is a universal joint transmitting the energy generated by the basal body to the filament. Finally, the filament is the most visible part of the flagellum and is made of about 20,000 monomers of flagellin with a length of several μm. The control of the flagellar assembly is a fine-tuned mechanism well described in Salmonella enterica serovar Typhimurium, Escherichia coli and Caulobacter crescentus (for reviews, see the references [9-12]).
It was recently described that flagellar regulation of B. melitensis was controlled by the FtcR flagellar master regulator which activates flagellar expression by binding directly on fliF promoter . The mutation of ftcR decreases fliF expression and induces the extinction of the flagellin and hook proteins. The two LuxR-type regulators VjbR and BlxR are also required for flagellar activation [14,15]. VjbR is involved in the quorum sensing of B. melitensis and mediates the inhibitory effect of the N-dodecanoyl-DL-homoserine lactone (C12-HSL) . In the early steps of the hierarchical flagellar regulation of B. melitensis, it was proposed that VjbR controls the expression of ftcR . More recently, it was proposed that flagellar hierarchy is divided in three classes in B. melitensis (J. Ferooz et al., unpublished). In this later work, FlbT was described as the checkpoint regulator between the class II and III genes, and activates the flagellin production.
Here, we demonstrate that the extracellular appendage produced by B. melitensis is a flagellum with all characteristics of the sheathed flagella. Moreover, ΔftcR, ΔfliF, ΔflgE, and ΔfliC mutants still produce a filament, probably an empty sheath regulated independently of the ftcR and vjbR flagellar pathway.
All strains used in this study are listed in Table Table1.1. All Brucella strains used in this study derive from B. melitensis 16 M Nalr (spontaneous nalidixic acid resistant mutant selected from B. melitensis 16 M, received from A. Macmillan, Central Veterinary Laboratory, Weybridge, UK). The growth was measured by reading the optical density of the cultures at 600 nm. B. melitensis 16 M growth curves in rich medium (1% yeast extract, 1.6% peptone, 0.5% NaCl) were performed from a late-exponential overnight culture obtained in liquid 2YT medium. B. melitensis 16 M strains grew with shaking at 37°C in rich medium containing appropriate antibiotics from an initial OD600 of 0.05. C. crescentus CB15N grew in peptone-yeast extract (PYE complex media) at 30°C. Antibiotics were used at the following final concentrations: kanamycin, 50 μg ml-1; nalidixic acid, 25 μg ml-1.
Bacterial flagella were stained using the "Ryu staining" method as described previously [17-19]. The Ryu stain has two components. Solution I (the mordant) contains 10 ml of 5% aqueous solution of phenol, 2 g of tannic acid, and 10 ml of saturated aqueous solution of aluminum potassium sulfate-12 hydrate. Solution II (the stain) is a saturated ethanolic solution of crystal violet (3 g in 25 ml of 95% ethanol). The final stain was prepared just before use by mixing 1 part of solution II with 10 parts of solution I and then by filtering the mixture through filter paper to remove coarse precipitate. A drop of cell culture was transferred onto the clean slide and covered with a cover slip. After 5 to 10 min, two drops of Ryu stain were applied to the edge of the cover slip and flowed under the cover slip by capillarity and mixed with the cell suspension. The cells were examined for flagella after 5 to 15 min at room temperature under a phase-contrast microscope. The edges of the cover slip on the slide were sealed with nail varnish.
The bacteria were grown in rich medium at 37°C to an OD600 of 0.25. Bacteria were centrifuged at 1000 r.p.m. for 20 min (Jouan), washed in PBS and fixed for 20 min in 50 μl of 4% paraformaldehyde pH 7.3. Samples were stored at 4°C. A drop (15 to 35 μl) of a solution of 1% Alcian blue was placed on a sheet of Parafilm. A carbon Formvar-coated grid was placed on a drop of solution for 5 min, carbon side down, washed five times in distilled water and then placed on a drop of bacterial suspension for 10 min on the same parafilm sheet. Grids with adherent cells were either (i) negatively stained for direct visualization on transmission electron microscope (TEM) or (ii) labeled with the anti-Brucella LPS O-chain, M epitope, mAb A156b3b2 [20,21] before staining as described previsously by Fretin et al.. For immunolabeling, after 10 min on the drop of bacterial suspension, the carbon Formvar-coated grid was placed onto drops (15 to 35 μl) of the following reagents on the same parafilm sheet: 5 washes in phosphate-buffered saline (PBS)-glycine 5% (3 sec each), PBS-bovine serum albumin (BSA)(1 min), monoclonal antibody A156b3b2 diluted 1/20 in PBS-BSA 5% (1 h), five washes in PBS (10 sec each), rabbit anti-mouse immunoglobulin conjugated to ± 15 nm colloidal gold diluted 1/20 in PBS-BSA (1 h), three washes in PBS (10 sec each), 2 washes in distilled water (10 sec each). After a completed treatment protocol, the grid was negatively stained with 2% aqueous solution of uranyl acetate for 10 sec, the excess fluid was removed with a filter paper, and the grid was air dried. Specimens were examined using a Philips Technai 10 TEM. Note that protein A-colloidal gold were also used instead of rabbit anti-mouse immunoglobulin with the same results.
Due to the short period of flagellar production (only at early exponential growth) and the low percentage of flagellated bacteria in B. melitensis, the visualization of the flagellum by TEM is difficult . To easily detect the flagellated Brucella, we used the Ryu staining. An advantage of that technique is that the flagellum of B. melitensis can be visualized directly without any prior centrifugation (required for the TEM technique), minimizing the manipulation of the culture containing the pathogen. As positive control, we used Caulobacter crescentus because its flagellar expression is cell-cycle dependent and the flagella were previously visualized by Ryu staining . We were able to visualize the flagellum of B. melitensis using the Ryu staining by phase-contrast microscopy (Figure (Figure1A).1A). Although the number of flagellated B. melitensis in the population is fewer than the flagellated C. crescentus (Figure (Figure1B),1B), the stained flagellum produced by B. melitensis is clearly detected by optical microscope. This result demonstrates that the Ryu staining is a simple technique allowing a rapid visualization of the flagellum of B. melitensis and with minimal sample manipulation.
We used the transmission electron miscroscopy (TEM) to analyze precisely the ultrastructure of the flagellum in B. melitensis. A sample of a B. melitensis culture at the early exponential growth phase in rich medium was stained with uranyl acetate 2% and visualized by TEM as described in the Materials and Methods section. We also labeled the bacteria with an anti-Brucella lipopolysaccharide (LPS) antibody and confirmed the presence of LPS on both the cell surface and the sheathed flagellum of B. melitensis (Figure (Figure1C).1C). B. melitensis produces a polar sheathed flagellum of about 50 nm diameter, showing an inner filament of about 11 nm diameter clearly surrounded by a sheath, which seems to be most likely a continuous extension of the bacterial outer membrane (Figure (Figure1C1C and and2).2). As usually seen in other bacterial species producing a sheathed flagellum, the flagellum of B. melitensis is ending by a club-like structure (Figure (Figure3A).3A). A polar bended structure is also usually observed, like a nascent flagellum during the first steps of flagellar assembly (Figure (Figure3B).3B). This bended structure, and sometimes a flagellum, is visualized at the septum of division between two daughter cells, suggesting that flagellar assembly occurs at this site (Figure (Figure3C).3C). Taken together, these data clearly demonstrate that the appendage produces by B. melitensis has the typical features of the sheathed flagella and not a pilus-like structure or a remaining part of the flagellum.
To determine whether the mutation of flagellar genes affect flagellar assembly, ΔfliF, ΔflgE and ΔfliC mutants were analyzed by TEM. These mutants were created by gene replacement with the aphA4 cassette (J. Ferooz et al., unpublished). Surprisingly, these three mutants still produce a filamentous structure (Figure 4A, B and and4C).4C). We previously demonstrated that FlgE and FliC protein cannot be detected by Western blot analysis in ΔflgE and ΔfliC respectively. Therefore, we propose that the appendage observed in ΔfliF, ΔflgE and ΔfliC is an empty sheath. The genome of B. melitensis contains the clusters of genes coding for only two extracellular structures: a type IV secretion system (T4SS) expressed at the stationary growth phase and a flagellum. In order to know whether the structure seen is a T4SS or not, a ΔvirB mutant was visualized by TEM. This ΔvirB mutant was made by removing the 12 open reading frames (ORFs) encoding the T4SS of B. melitensis 16 M . As seen in the wild-type strain (Figure (Figure4D),4D), ΔvirB also produces a similar flagellar structure at the early exponential growth phase (Figure (Figure4E).4E). This result demonstrates that the extracellular appendage observed in flagellar mutants is not a T4SS.
The flagellar expression is not affected by mutation of the structural genes, indeed ΔfliF and ΔflgE still produce the flagellin FliC (J. Ferooz et al., unpublished). However, the flagellar master regulator FtcR is needed for fliF expression and the FlgE and FliC synthesis . By intensively observing ΔftcR by TEM, we also found some bacteria producing a filament-like structure, suggesting that sheath synthesis and assembly is likely independent of the FtcR pathway (Figure (Figure4F4F).
Although it was demonstrated that B. melitensis produces a polar sheathed flagellum under specific conditions , there were still some doubts concerning the production of a classical flagellum by Brucella [5,23]. The conditions in which the flagellum is observed are not optimal because Brucella produces a flagellum only at the early exponential growth phase in rich medium . Moreover, Brucella must be handled according to level 3 biosafety precautions , making difficult its manipulation for the preparation of sample for TEM analysis. This is the reason why we tested the Ryu staining technique, that allows an easier detection of flagellum than TEM as described in several flagellated bacteria such as C. crescentus, L. monocytogenes, Bacillus subtilis and Salmonella Typhi [19,25-27]. Interestingly, even with a low number of flagellated bacteria, this technique allows a quick and direct manipulation of the culture before the extinction of flagellar expression. In this work, we showed that the flagellum of B. melitensis can be easily visualized using the Ryu staining.
The structure of the flagellum produces by B. melitensis was analyzed by the TEM technique and several features of the sheathed flagella of other species were highlighted. At the present time, little is known about the formation, composition or function of flagellar sheaths in bacteria, and interestingly, Brucella is the only one rhizobiale to produce a sheathed flagellum [28,29]. Due to the sheath around the filament, the visible flagellum of B. melitensis has a diameter of 50 nm, which is larger than an unsheathed flagellum. However, the diameter of the bacterial filament is usually about 20 nm but we showed that the diameter of the inner filament in the sheath of B. melitensis is only 11 nm. This smaller diameter could be linked to the shorter amino acid sequence of the unique flagellin composed of only 282 amino acids in B. melitensis. In comparison, the flagellin sequence of E. coli, S. enterica serovar Typhimurium and Pseudomonas aeruginosa are between 488 and 498 amino acids.
Among the features of sheathed flagellum, the shape at the end of the flagella of B. melitensis is similar to a club-like structure, also described in other bacteria like B. bacteriovorus, H. pylori and V. fischeri [30-32]. This extension can be viewed as the continuity of the sheath at the end of the filament. The genome of Brucella lacks the FliD (or HAP2) homologue, a protein involved in the flagellin assembly at the top of the filament. One hypothetical function of this club-like structure could be the formation of a confined space facilitating the self-assembly of flagellin monomers into a filament, despite the lack of a FliD homologue. Similarly in Vibrio alginolyticus, it was assumed that the sheath could trap excreted flagellin to allow polymerization independently of FliD . Secondly, the sheath of the flagellum of B. melitensis is likely an extension of the outer membrane and contains LPS, which is also observed in H. pylori, B. bacteriovorus and some Vibrio spp., [1,32,34-38]. Even if the sheath contains LPS, the sheath composition of Vibrio and Helicobacter is different from the outer membrane [30,39,40]. To the best of our knowledge, all these features have never been observed in a pilus-like structure, confirming that this appendage is a flagellum and not a pilus or flagellum-like structure.
Similarly to other α-proteobacteria, Brucella and C. crescentus are morphologically asymmetric . C. crescentus couples flagellar biogenesis with cell cycle allowing the production of a flagellum at the swarmer pole of predivisional cell and a stalk at the other pole . Surprisingly, we occasionally observed a bended structure or a flagellum at the septum division of B. melitensis rather than at the poles of predivisional cells. The localization of the flagellum at the septum of division is rarely observed in bacteria but is not unusual and is described in B. bacteriovorus and Treponema phagedenis [42,43].
In the last part of this work, we demonstrated that production of the sheath is probably unlinked to flagellar assembly in B. melitensis. Indeed, ΔfliF, ΔflgE and ΔfliC structural flagellar mutants still produce a filamentous appendage despite the absence of FlgE or FiC proteins. However, persistence of an empty sheath in flagellar mutants was often described in bacteria producing a sheathed flagellum as Vibrio species and H. pylori [44-48]. In B. bacteriovorus, deletion of flagellin fliC3 caused the synthesis of copious, disordered, tubular material resembling outer membrane [49,50]. The authors suggest that this structure is a disordered sheath without a normal flagellum inside. It is important to note that counting the number of flagellated bacteria in the population of the mutants in order to estimate a percentage of flagellation compared to the wild-type strain is not relevant due to the variability and the low number of flagellated bacteria detected between samples. Thus, it is possible that the population of flagellar mutants is less flagellated than the wild type population, but better conditions to enhance flagellation of B. melitensis will be needed.
In spite the fact that FtcR is the master regulator of flagellar assembly in Brucella, persistence of sheath production in ΔftcR mutant also demonstrate that sheath expression is not dependent of FtcR and suggests that another unknown regulator is involved. Since the sheath surrounds the filament and is produced at one pole of the cell, even in flagellar mutants, it is possible that a flagellar pole marker coordinates both flagellum and sheath biogenesis at the cell pole before the FtcR activation. In Vibrio alginolyticus and C. crescentus, the localization of the flagellum at the pole is dictated by FlhF and PflI regulators respectively [51,52]. Polar localization of the pole marker PflI is independent of FliF, whose oligomerization into the MS-ring probably allows the definition of the site of flagellar synthesis, suggesting that PflI acts before or independently of this event . Similarly, a regulator could couple flagellar assembly and sheath production independently of FtcR in B. melitensis.
Altogether, the data presented in this study proved that B. melitensis produces a flagellum with the characteristics of sheathed flagella described in other organisms. Flagellar assembly and sheath production are a complex regulatory mechanism that remains to be further investigated to gain a better understanding of the flagellum's function during Brucella's infection.
The authors declare that they have no competing interests.
JF and JJL designed the experiments and interpreted the data. JF performed the experiments. JF wrote the manuscript. All authors read and approved the final manuscript.
We thank Dr Valérie Haine for her critical and careful reading of this manuscript and helpful discussion. We are grateful to Bettina Battisti for the helpful manuscript revision. We thank the anonymous reviewers for their insightful comments and suggestions. We acknowledge the 'Service Interfacultaire de Microscopie Electronique' of the University of Namur and for its expertise in TEM. This work was supported by grants to Jonathan Ferooz from the Fonds Adrien Bauchau and the FNRS (Fonds National de la Recherche Scientifique). This work was supported by the Commission of the European Communities (contract no. QLK2-CT-1999-00014), the FRFC (Fonds de la Recherche Fondamentale Collective, conventions 2.4521.04 and 2.4521.08) and the ARC fellowship (Actions de Recherches Concertées, conventions 04/09-325 and 08/13-015). Jonathan Ferooz held a Belgian specialization grant from the FRIA (Fonds pour la Formation à la Recherche dans l'Industrie et dans l'Agriculture).