|Home | About | Journals | Submit | Contact Us | Français|
NaVBP is the member of the bacterial voltage-gated Na+ channel superfamily found in alkaliphilic Bacillus pseudofirmus OF4. The alkaliphile requires NaVBP for normal chemotaxis responses and for optimal pH homeostasis during a shift to alkaline conditions at sub-optimally low [Na+]. We hypothesized that interaction of NaVBP with one or more other proteins in vivo, specifically methyl-accepting chemotaxis proteins (MCPs), is involved in activation of the channel under the pH conditions that exist in the extremophile and could underpin its role in chemotaxis; MCPs transduce chemotactic signals and generally localize to cell poles of rod-shaped cells. Here, immunofluorescence microscopy and fluorescent protein fusion studies showed that an alkaliphile protein (designated McpX) that cross-reacts with antibodies raised against Bacillus subtilis McpB co-localizes with NaVBP at the cell poles of B. pseudofirmus OF4. In a mutant in which NaVBP-encoding ncbA is deleted, the content of McpX was close to the wild-type level but McpX was significantly delocalized. A mutant of B. pseudofirmus OF4 was constructed in which cheAW expression was disrupted to assess whether this mutation impaired polar localization of McpX, as expected from studies in Escherichia coli and Salmonella, and, if so, whether NaVBP would be similarly affected. Polar localization of both McpX and NaVBP was decreased in the cheAW mutant. The results suggest interactions between McpX and NaVBP that affect their co-localization. The inverse chemotaxis phenotype of ncbA mutants may result in part from MCP delocalization.
Our understanding of the activation and physiological roles of bacterial channels lags far behind our knowledge of their structures (Booth et al., 2003; Booth et al., 2005; Kung & Blount, 2004). Recently, roles were demonstrated for the voltage-gated sodium channel, NaVBP, from alkaliphilic Bacillus pseudofirmus OF4 in pH homeostasis, motility and chemotaxis at alkaline pH (Ito et al., 2004b). The current studies probe cell biological properties of NaVBP that could underpin these physiological roles.
NaVBP is a member of the bacterial voltage-gated sodium channel superfamily, NaVBac whose founding member is the NaChBac channel from alkaliphilic Bacillus halodurans C-125 (Koishi et al., 2004; Ren et al., 2001). The strong resemblance of NaVBac channels to one of the 6-transmembrane segment repeats of biomedically important eukaryotic voltage-gated channels has led to intense investigative interest in these bacterial channels (Ren et al., 2001). Electrophysiological and mutational analyses have focused largely on NaChBac, either in lipid bilayers or expressed in eukaryotic cells (Blanchet et al., 2007; Chahine et al., 2004; Kuzmenkin et al., 2004; Pavlov et al., 2005; Richardson et al., 2006; Zhao et al., 2004). Since B. halodurans C-125, the native host of NaChBac, is not genetically accessible, it was not possible to evaluate the match between the properties that were determined in the non-native settings with conditions under which NaChBac could be shown to function physiologically. In our initial studies, we therefore turned to B. pseudofirmus OF4, in which the ncbA gene that encodes the NaChBac homologue NaVBP could be deleted in order to probe the physiological roles of the channel (Ito et al., 2004b).
B. pseudofirmus OF4 grows on non-fermentative carbon sources in a pH range from 7.5 to > 11. At pH values above 7.5, robust Na+/H+ antiporter (exchanger) activity, together with the activity of the proton-coupled ATP synthase, support net uptake of protons in respiring cells, thus achieving a cytoplasmic pH that is lower than the external pH. Antiport action reduces the proton motive force and establishes an inwardly directed sodium motive force. Na+-coupled solute uptake and flagellar rotation take advantage of this sodium motive force to support solute uptake and motility (Fujinami et al., 2007; Ito et al., 2004a; Krulwich et al., 2007; Padan et al., 2005). The Na+ that enters the cytoplasm during this Na+-coupled bioenergetic work completes the Na+ cycle and plays an important role in supporting the ongoing Na+/H+ antiport activity for pH homeostasis (Fig. 1).
Bioenergetic studies on an ncbA mutant of B. pseudofirmus OF4, lacking NaVBP, showed that the channel plays a role in Na+ re-entry in support of pH homeostasis at pH 10.5 under low [Na+] conditions and low concentrations of solutes whose entry is coupled to Na+. The ncbA mutant also displayed a motility defect at high pH even though it possessed the motPS genes that are a required stator-force generating Na+ channel required for flagella rotation (Ito et al., 2004a; Ito et al., 2004b). The ncbA mutant exhibited motility only after prolonged growth in liquid culture or on soft agar plates. The third, striking phenotype that was observed at high pH in both poorly motile or highly motile variant ncbA mutants was a high frequency of tumbling (Ito et al., 2004b). The tumbly ncbA mutants exhibited an “inverse chemotaxis”, i.e. the mutants moved toward repellents and away from attractants.
The parallel electrophysiological studies of the channel in eukaryotic cells demonstrated that NaVBP is a Na+-specific voltage-gated channel whose activation is potentiated at high pH (Ito et al., 2004b). The electrophysiological characteristics indicated that in the natural alkaliphile host, the role of high external (and secondary elevation of cytoplasmic pH) in channel opening is supplemented by additional triggers (Ito et al., 2004b). Such an additional trigger could result from an interaction between the channel and either the chemotaxis signaling machinery or the flagellar motor (Fig. 1). A particularly attractive hypothesis was that NaVBP interacts with transmembrane chemoreceptors (also known as methyl-accepting chemotaxis proteins, MCPs), accounting both for triggering and channel effects in chemotaxis. The membrane-associated MCPs have periplasmic ligand-binding domains that monitor attractants and repellents on the outside of the cell and have signaling/adaptation domains in their cytoplasmic segments that communicate with the flagellar motor via a two-component signaling pathway (Szurmant & Ordal, 2004). Signaling is initiated by a signaling complex that is formed at conserved cytoplasmic domains of the MCPs (Wadhams & Armitage, 2004; Zhulin, 2001). In E. coli and Salmonella, MCPs form a complex with the histidine kinase CheA that mediates the signaling triggered at the MCP receptor and with adaptor protein CheW; this complex localizes to the cell poles (Gegner et al., 1992; Liu et al., 1997; Maddock & Shapiro, 1993; Sourjik & Berg, 2000); weaker lateral clusters also are observed (Lybarger & Maddock, 1999; Skidmore et al., 2000). In Bacillus subtilis, the core elements that mediate chemotaxis resemble those of the E. coli-Salmonella systems (Lamanna et al., 2005; Rao et al., 2004; Szurmant & Ordal, 2004; Weis, 2006). As in the enteric bacteria, MCPs are localized to polar end of other rod-shaped bacteria including Bacillus species (Gestwicki et al., 2000; Kirby et al., 2000; Lamanna et al., 2005). Polar localization and/or clustering of the complexes has been proposed to play a critical role in signal amplication in E. coli (Bray et al., 1998; Duke & Bray, 1999; Irieda et al., 2006; Lybarger & Maddock, 2001; Shapiro et al., 2002). If NaVBP interacts with polar MCPs and/or the kinase/adaptor proteins that are complexed with them in B. pseudofirmus OF4, the channels should also localize to the cell poles. On the other hand, if NaVBP interacts with flagellar motors, a delocalized pattern would be expected consistent with a peritrichous pattern of flagella in this alkaliphile (Fig. 1) (Fujinami et al., 2007; Sturr et al., 1994). We conducted an immunofluorescence microscopy (IFM) analysis of B. pseudofirmus OF4 cells to confirm the polar localization of MCPs and compare it with the localization pattern of NaVBP. MCP localization was assessed using monoclonal antibodies against NaVBP, and polyclonal antibodies that were raised against B. subtilis McpB, a chemoreceptor for Asn, Asp, Gln, Glu, and His (Hanlon & Ordal, 1994). The alkaliphile gene whose product cross-reacts with this antibody has not yet been identified; we refer to this putative MCP as McpX. Channel localization was also probed using a plasmid expressing a NaVBP-cyan fluorescent protein (CFP) fusion protein in B. pseudofirmus OF4. We further examined whether disruption of the alkaliphile cheAW locus decreases localization of McpX, NaVBP or both.
The results indicate that NaVBP co-localizes at cell poles with an alkaliphile protein that is recognized by antibodies raised against B. subtilis McpB. They further show that mutational loss of the channel leads to McpX delocalization, while cheAW disrupted mutant leads to significant delocalization of both the channel and McpX.
The bacterial strains and plasmids used in this study are shown in Table 1. E. coli and B. subtilis strains were grown routinely in LB or LBK medium (Goldberg et al., 1987). Alkaliphilic B. pseudofirmus OF4 wild-type (strain 811M) (Clejan et al., 1989) and its derivative strains were grown aerobically in semi-defined malate-yeast extract (MYE) medium (pH 10) (Sturr et al., 1994) at 37°C. pH 10 medium replaced the pH 10.5 medium used in most of our studies because the SC34 strain that lacked a functional NaVBP exhibited a significant deficit in mid-logarithmic growth at the higher pH value; earlier studies of SC34 growth had examined yield at stationary phase and a defect at high pH was not observed (Ito et al., 2004b). Nifedipine-treated wild-type (811M) cells were prepared by adding 50 μM nifedipine to mid-logarithmic phase cells (OD600=0.6). Growth was continued for one hour, and the cells were then used for Immunofluorescence microscopy. Antibiotics were added at 5 μg ml−1 for chloramphenicol or 0.3 μg ml−1 for erythromycin.
The chromosomal DNA fragment that contains the cheAW genes (accession number DQ150110) was replaced entirely by a chloramphenicol acetyltransferase gene (cat) using the gene SOEing method described by others (Horton, 1996). Primers used are shown in Table 2. Two independent PCRs were performed on B. pseudofirmus OF4 wild-type chromosomal DNA template with the primer sets del-AW-F1-BH1 and del-AW-R1, del-AW-F2 and del-AW-R2-ER1. Another PCR was performed on pC194 (Horinouchi & Weisblum, 1982) DNA template with the primer set del-AW-F-CAT and del-AW-R-CAT to amplify the cat gene. The three purified PCR products were used as templates for a second PCR with primer sets del-AW-F1-BH1 and del-AW-R2-ER1. The purified product of this reaction was digested with BamHI and EcoRI and cloned into BamHI- and EcoRI-digested pG+host4, yielding pG4ΔcheAWCAT. The plasmid was transformed into B. pseudofirmus OF4-811M protoplasts. The protocol for isolation of single crossover candidates was previously described (Ito et al., 1997). From the two kinds of single crossover candidates, analyses of the cheAW disruption mutant indicated that the recombination occurred in the region upstream of the cheA gene (region between del-AW-F1-BH1 and del-AW-R-CAT) as was confirmed by PCR using primer sets OF4cheA24 and OF4cheA25, OF4cheA3 and OF4cheA25, OF4cheA24 and OF4cheA15. The cheAW disruption mutant was designated strain 811M-cheAW.
The large fragment (4.8 kb) of XbaI- and HindIII-digested shuttle vector pYH30 (Bechhofer & Wang, 1998) was ligated with XbaI- and HindIII-digested pMW118, yielding pYM1. Plasmid pYM1 is a shuttle vector in E. coli and Bacillus species that is low copy in both hosts (~6 copies/cell) (Ito et al., 2004a). The ncbA-ecfp fragment that enhanced the cyan fluorescent protein gene (ecfp) was fused just before the stop codon of ncbA gene using the gene SOEing method (Horton, 1996). A schematic diagram of ncbA-ecfp fragment is shown in Fig. 2A. The primers used for constructions are shown in Table 2. Two independent PCRs were performed on B. pseudofirmus OF4 wild-type chromosomal DNA template with the primer sets NC-F1-SC2 and NC-R1-1, NC-F2-1 and NC-R2-BH1. Another PCR was performed on pECFP plasmid DNA template with the primer set NC-F-CFP1 and NC-R-CFP1 to amplify the ecfp gene. The three purified PCR products were used as templates for a second PCR with primer sets NC-F1-SC2 and NC-R2-BH1. The purified product of this reaction was digested with EcoRI and cloned into EcoRI-digested pMW118, yielding pMWSC-CFP. pMWSC-CFP was digested with BamHI and HindIII and ligated with the large fragment (5.0 kb) of BamHI- and HindIII-digested pYM1, yielding pSC-CFP. In this plasmid, NaVBP-CFP is expressed under the control of the native ncbA promoter.
The ecfp fragment (Fig. 2B), pMWCFP, and pCFP were constructed by using a parallel method. In pCFP, CFP is expressed under the control of the native ncbA promoter. The inserts of pSC-CFP and pCFP were mutation free. The plasmids were transformed into B. pseudofirmus OF4-SC34 protoplasts (Ito et al., 1997).
Rabbit polyclonal antibodies against the abundant McpB of B. subtilis (Hanlon & Ordal, 1994; Szurmant & Ordal, 2004) were obtained from G. W. Ordal (University of Illinois). An anti-NaVBP mouse monoclonal antibody (L30/34; IgG2b isotype) was generated against a recombinant, bacterially expressed full-length NaVBP protein. Mice were immunized and hybridomas generated and screened by standard methodology (Bekele-Arcuri et al., 1996; Trimmer et al., 1985). L30/34 hybridomas were grown in roller bottles containing Dulbecco's Modified Eagle Medium (Gibco) supplemented with 10% Fetal Bovine Serum (Gibco). Supernatants were collected and monoclonal antibodies purified with Protein G Agarose (Amersham Pharmacia) following the manufacturer's protocol.
B. pseudofirmus OF4 wild-type and its derivative mutant strains were grown overnight on MYE medium as described above. 1 ml of overnight culture was inoculated into 100 ml of fresh MYE medium pH 10, and grown to OD600=0.6. Cells were harvested and washed in TSE buffer (50 mM Tris-HCl pH 8.0, 10% sucrose, 1mM EDTA). Cells were suspended in the same buffer and a protease inhibitor cocktail (Sigma-Aldrich) and DNase (50 μg ml−1) was added. Cells were disrupted by sonication and unbroken cells were removed by centrifugation at 9,100 g for 15 min at 4°C, yielding whole cell suspension. The membrane fraction was harvested by centrifuging at 40,000 rpm for 90 min in a Beckman 70Ti rotor at 4°C and suspended in TSE buffer. A protease inhibitor cocktail (Sigma-Aldrich) was added. The protein concentration of the whole cell suspension and membrane fraction was measured by the Lowry method (Lowry et al., 1951) with BSA as a standard. The same volume of SDS loading buffer was added to each sample which were boiled for 3 min at 100°C, after which the proteins were separated by 12% polyacrylamide SDS gels (Schagger & von Jagow, 1987). The gels were then transferred to nitrocellulose filters (Bio-Rad) electrophoretically overnight in Tris–glycine–methanol buffer (25 mM Tris, 192 mM glycine, 20% (v/v) methanol, pH 8.3).
The following antibodies and reagents were used for detection in immunoblot analyses: for McpB (B. subtilis) and McpX (B. pseudofirmus OF4), 1/2000 rabbit anti-B. subtilis McpB antibody and 1/3000 Goat anti-Rabbit HRP (Bio-Rad); and for NaVBP-CFP and CFP, 1/250 (whole cell analyses) or 1/1000 (membrane fraction analyses) rabbit anti-GFP polyclonal antibody (Clontech), 1/3000 Goat anti Rabbit HRP (Bio-Rad), and Can Get SignalTM Immunoreaction Enhancer Solution (TOYOBO). Detection and analyses of chemiluminescence images were conducted using the quantitative imaging system Fluor-S MAX (Bio-Rad) according to the protocol provided in the manufacturer's instructions (Amersham Biosciences). NaVBP could not be detected in immunoblot analyses of either whole cells or membranes using the anti-NaVBP mouse monoclonal L30/34 anti-body (data not shown).
The method described by Hiraga et al. (Hiraga et al., 1998) was adapted for B. pseudofirmus OF4 strains. Wild-type and mutant strains were grown as described above for immunoblot analyses. One ml of the final culture was briefly added to 10 ml of 70% ethanol, mixed gently and left for 1 hr at room temperature. The fixed cells were harvested by centrifugation (700 g for 15 min at 4°C) and suspended in 1 ml of 70% ethanol. A glass slide, S-2215 (Matsunami Glass), was covered by 20 μl of poly-L-lysine hydrobromide (1 mg ml-1) and left for 5 min, washed with distilled water, and air dried. 10 μl of the ethanol-fixed cell suspension was then dropped on the poly-L-lysine coated slide and air dried for 20 min. The slide was covered with 100 μl of lysozyme solution (2 mg/ml in 25 mM Tris-HCl, pH 8.0, 50 mM glucose, 10 mM EDTA) and incubated for 5 min at room temperature. The slide was washed with 5ml PBSTE (140 mM NaCl, 2 mM KCl, 8 mM Na2HPO4, 1.5 mM KH2PO4, 0.05% Tween 20, 10 mM EDTA, pH 8) 3-times and then incubated with PBSTE-BSA (PBSTE containing 2% bovine serum albumin) for 15 min at room temperature.
The slide was treated for 1 hr at room temperature with the primary antibody solution (1/1000 mouse anti-NaVBP monoclonal L30/34 antibody or 1/2000 rabbit anti-B. subtilis McpB polyclonal antibody in PBSTE-BSA), and covered with a cover glass in a shading moisture chamber. The slide was washed with 5 ml of PBSTE 3 times (the cover glass was removed by washing). After the slide was washed with PBSTE, it was again incubated with PBSTE-BSA for 15 min at room temperature. It was then treated for 1 hr at room temperature with a secondary antibody solution (1/1000 Alexa Fluor 488 rabbit anti-mouse IgG (Molecular Probes) for the NaVBP, and Alexa Fluor 546 goat anti-rabbit IgG (Molecular Probes) for the McpX in PBSTE-BSA). The slide was then covered with a cover glass in a shading moisture chamber and was again washed with 5 ml of PBSTE 3 times. After washing with PBSTE, it was covered with a cover glass and sealed with enamel. The microscopic images were obtained by Imaging workstation FW4000 (Leica Geosystems), and processed with Photoshop CS software (Adobe Systems). In each experiment, over 30 cells were assessed with respect to the subcellular localization of NaVBP and McpX, and the length of the cells. All the results shown are a total of three independent experiments. The localization of NaVBP and McpX was classified by two researchers (S.F. and M.I.) as: A:One pole; B:Two poles; C:One pole and side(s); D:Two poles and side(s); E:Side(s); or diffuse (Table 3 and and44).
B. pseudofirmus OF4-SC34 cells (ΔncbA) were transformed with plasmid pSC-CFP (yielding the SC34/pSC-CFP transformant) or pCFP (yielding the SC34/pCFP transformant). The transformants were grown as described above for immunoblot analyses. Small aliquots of the culture were spotted onto glass slides coated with 0.5% agarose. Microscopic images were obtained on an imaging workstation FW4000 (Leica Geosystems). Photoshop CS software (Adobe Systems) was used to manipulate images, e.g. to create overlays, without altering the images or patterns of the cells themselves. In each experiment, over 30 cells were classified with respect to the subcellular localization of NaVBP-CFP and CFP. All the results shown are a total of 3 independent experiments.
Migration in soft agar was assayed by the method described previously (Fujinami et al., 2007). B. pseudofirmus OF4 wild-type and its derivative mutant strains were grown in MYE medium overnight and then grown to logarithmic phase on MYE, pH 10, as described above. Migration assay of B. pseudofirmus OF4 wild-type and its derivative strains in soft agar were conducted on plates containing MYE medium (pH 10) (Sturr et al., 1994) and solidified by the addition of 0.3% Noble agar. One μl of a liquid culture was spotted on to the center of the plate. After incubation of the plates at 37°C for 10 hr, the diameter of the colony was measured. All results shown are the averages of three independent experiments.
B. pseudofirmus OF4 wild-type and its derivative strains were grown as described above. Qualitative assessment of tumbling bias in liquid MYE medium (pH 10) was carried out by the hanging drop method using a Leica DMLB100 dark field microscope (400×) and Leica DC300F camera, Leica IM50 version 1.20 software (Leica Geosystems).
The MCPs of B. pseudofirmus OF4, and the genes encoding them, have not yet been characterized, because the genome-sequence of this microorganism has not been determined. We anticipated that antibodies raised against a major MCP from B. subtilis could be used to study the localization of an alkaliphile MCP since it has been reported that antibodies raised against an MCP can be used to detect MCPs of different organisms by immunoblotting and immunofluorescence microscopy (IFM) (Gestwicki et al., 2000). Before undertaking IFM, confirmation was sought that anti-B. subtilis McpB antibody could detect an MCP of B. pseudofirmus OF4 by assessing the heterogeneity of any cross-reacting proteins in the alkaliphile on immunoblot analyses (Fig. 3). The bands in expected positions of McpA and McpB were detected in whole cells of B. subtilis 168 (Kirby et al., 2000). The location of the band corresponding to McpB is indicated with an arrow in Fig. 3. In whole cells of wild type B. pseudofirmus OF4 (811M), a single band was detected at almost the same location as McpB of B. subtilis 168. We designated this MCP of B. pseudofirmus OF4 as McpX. Bands corresponding to McpX were also detected in whole cells of the channel mutant, SC34, and in a channel mutant strain (SC34-R) to which ncbA had been restored to the chromosome (Fig. 3). The expression level of McpX was similar in the 3 alkaliphile strains.
IFM of NaVBP and McpX was then conducted in B. pseudofirmus OF4 wild-type and its mutant derivatives to assess the localization of these proteins. Both McpX and NaVBP localize extensively to cell poles of the wild-type strain (Fig. 4). Quantification of the co-localization of NaVBP and McpX is shown in Table 3 (top rows of each section). “Total cells” designates the total number of the cells that were analyzed. Subcellular localization of NaVBP was observed in 92% of wild-type (811M) cells. In 82% of 811M cells, NaVBP was localized at a cell pole, and in 90% of such cells, polar co-localization of NaVBP and McpX was observed. In the channel mutant, SC34, no localized signal was observed with the antibody against the channel, as expected, except for a small signal that reflects the difficulty of setting the parameters “counted” as fluorescence with complete precision. Unexpectedly, the polar localization of McpX was also significantly decreased in the channel mutant, from 78% in wild-type to 15% in SC34 (Fig. 4 and Table 3, second rows of each section). A channel mutant with a restored ncbA gene at the native location (SC34-R) exhibited fluorescence with 76% polar localization, 89% of which was co-localized with McpX. McpX fluorescence in the SC34-R strain that was localized at the poles rose from 15% in SC34 to 71% in the SC34-R strain and 96% of that was co-localized with NaVBP (Fig. 4 and Table 3, third rows of each section). Measurements of 50 cells of each strain showed that there were no significant differences in the cell length or width of the different strains. The average cell length (μm) of 811M, SC34, SC34-R, 811M-AW, SC34/pSC-CFP and SC34/pCFP was 1.95±0.44, 1.65±0.36, 2.43±0.55, 2.39±0.63, 2.25±0.45 and 2.13±0.55, respectively. The average cell width (μm) of 811M, SC34, SC34-R, 811M-AW, SC34/pSC-CFP and SC34/pCFP was 0.82±0.05, 0.82±0.05, 0.84±0.05, 0.84±0.06, 0.84±0.05 and 0.85±0.04, respectively.
The finding that mutational loss of NaVBP led to delocalization of McpX raised the possibility that the inverse chemotaxis phenotype observed in the ΔncbA strain could result entirely from secondary delocalization of MCPs rather than from a channel function of NaVBP itself. Evidence against this line of reasoning was the earlier observation that the NaVBP channel inhibitor nifedipine caused inverse chemotaxis behavior of wild-type B. pseudofirmus OF4 toward the chemoattractant aspartate when the inhibitor added to the chemotaxis assay buffer (at 50 μM) (Ito et al., 2004b). IFM experiments were conducted to test whether comparable exposure of the wild-type strain to nifedipine caused delocalization of NaVBP or McpX. These experiments revealed no such delocalization (Fig. 4B).
Transformants of B. pseudofirmus OF4-SC34 (ΔncbA) cells with plasmid pSC-CFP (SC34/pSC-CFP), in which CFP is fused to the channel, or control pCFP (SC34/pCFP) were grown in MYE, pH 10, and subjected to immunoblot analyses using antibodies against GFP. As shown in Fig. 5, bands corresponding to the expected location for NaVBP-CFP fusion were detected in whole cell and membrane fractions of the SC34/pSC-CFP transformant (white arrows). The band for CFP was detected in whole cells of SC34/pSC-CFP but not in membrane fractions (filled arrows). Anti-NaVBP mouse monoclonal antibody was also used for this detection. However, the antibody was not able to detect denatured NaVBP for immunoblot analyses. This monoclonal antibody can probably recognize an antigen only when NaVBP maintains three-dimensional structure.
The subcellular localization of NaVBP-CFP of SC34/pSC-CFP and CFP of SC34/pCFP was observed by fluorescence microscopy (Fig. 6). NaVBP-CFP was localized at cell poles in SC34/pSC-CFP (Fig. 6A), and CFP was not localized in SC34/pCFP (Fig. 6B). Polar localization of NaVBP-CFP was observed in 88% of SC34/pSC-CFP cells (Table 4).
The alkaliphile gene encoding the MCP monitored in the IFM experiments has not yet been identified. It was therefore impossible to disrupt this gene and examine whether the disruption caused NaVBP delocalization. However, it was possible to test whether the delocalization of MCPs that results from mutational loss of CheAW would lead to NaVBP localization. We therefore introduced a cheAW disruption into the wild type strain, producing strain 811M-CheAW. Immunoblot assays showed that the mutation in 811M-CheAW did not affect the expression level of McpX (Fig. 3). IFM studies of the subcelluar localization of McpX and NaVBP showed a decrease in the polar localization of both McpX and NaVBP in 811M-CheAW in comparison with the wild-type parent (Fig. 4A and Table 3, fourth rows of both sections). The polar localization of McpX decreased from 78% in wild-type to 33% in 811M-CheAW, with only 51% co-localized with NaVBP in the mutant. The polar localization of NaVBP decreased from 82% in wild-type to 33% 811M-CheAW, with only 50% co-localized with McpX in the mutant.
Migration assays in soft agar and observations of the tumbling bias in liquid medium were conducted on B. pseudofirmus OF4 wild-type and its derivative mutant strains to ascertain whether the effects of ncbA status on polar localization of McpX correlated with swarming capacity in soft agar and tumbliness (Fig. 7, Table 5). As shown in earlier studies of an up-motile alkaliphile variant and its mutant derivative (Ito et al., 2004b), deletion of ncbA (in SC34) from the wild-type strain led to a smaller colony diameter and a more tumbly phenotype relative to wild-type. Restoration of ncbA (in SC34-R), restored wild-type motility on soft agar and reduced the tumbliness. Transformation of the channel mutant, SC34, with a multicopy plasmid encoding ncbA-ecfp (SC34/pSC-CFP) led to almost wild-type motility and tumbliness. Control plasmid pCFP, encoding ECFP, did not restore motility or reverse the tumbly phenotype of SC34 strain.
NaVBP had earlier been shown to have a role in pH homeostasis, motility and chemotaxis in B. pseudofirmus OF4 (Ito et al., 2004b). The role in pH homeostasis was presumed to reflect a contribution of NaVBP to entry of Na+ in support of the Na+/H+ antiport that supports pH homeostasis during alkalinization of the medium, e.g. during a pH 8.5 → 10.5 shift in the outside pH. The upward pH change would potentiate opening of the channel (Ito et al., 2004b). The role of NaVBP observed in the pH shift experiments was especially notable in the presence of low Na+ and when the medium contained low concentrations of solutes that enter the cell together with Na+. No deficit was observed in the growth yield of the channel mutant at pH 10.5 in MYE medium with optimal Na+ and solutes that are co-transported with Na+, so no role for NaVBP in pH homeostasis under optimal conditions was indicated in the earlier studies. Here we found that the channel mutant, SC34, exhibited a defect in growth to the mid-logarithmic phase in the Na+-replete, malate-containing medium at pH 10.5. These results indicated that NaVBP does have a role in alkaline pH homeostasis even when adequate Na+ and solutes are present. This raises the possibility that the lag in motility in the channel mutant could be a result of sub-optimal pH homeostasis since the pH range for optimal swimming of B. pseudofirmus OF4 is below the highest values for optimal non-fermentative growth (Fujinami et al., 2007). However, the role of NaVBP in pH homeostasis cannot explain the dramatic effect of channel loss on chemotaxis. Most of the original assays of chemotaxis were conducted at pH 8.5, where the pronounced inverse chemotaxis phenotype was consistently observed in the mutant although no pH homeostasis problem would exist (Ito et al., 2004b). Rather, we hypothesize that the effect of NaVBP status on the chemotaxis behavior of B. pseudofirmus OF4 is related to the interaction between NaVBP and chemotaxis proteins.
In summary, the major findings on the localization of alkaliphile NaVBP and an MCP that emerge from this work are as follows. (i) The putative alkaliphile MCP designated McpX localized at cell poles as anticipated from similar findings in other systems (Kirby et al., 2000). Some McpX appeared to localize at the sides of the cell (Fig. 4, Table 3). MCPs are translocated by the helically distributed Sec machinery (Shiomi et al., 2006). It is possible that the experiments captured McpX that was in-transit or that was localized to a central region destined to be a new pole. (ii) NaVBP also exhibited localization to the cell poles, as shown both by IFM (Fig. 4, Table 3) and fluorescence microscopy on cells expressing CFP fused to the channel (Fig. 6, Table 4). The expression of the fused channel restored wild-type motility properties as assessed in soft agar plates and reversed tumbly properties as assessed qualitatively by microscopic examination (Table 5). (iii) A significant percentage of the polarly localized McpX and NaVBP exhibited co-localization (Fig. 4, Table 3). On the other hand, the pattern of NaVBP localization did not correlate well with the pattern of peritrichous flagellum location in wild-type B. pseudofirmus OF4. The wild-type strain typically has 1-2 flagella/cell and these are located at various locations around the cell body (Fujinami et al., 2007). (iv) The expression level of McpX was not markedly changed in the ncbA mutant that lacks NaVBP, but the McpX was significantly delocalized. This effect was reversed appreciably in a strain in which ncbA was restored to the chromosome (Fig. 4, Table 3). (v) Significant delocalization of both NaVBP and McpX was observed when cheAW genes of the wild-type were disrupted (Table 3). This delocalization was not as complete as in E. coli in which CheAW had been shown to form a complex with MCPs that is required for polar localization as well as clustering (Irieda et al., 2006; Kentner et al., 2006). Perhaps association of NaVBP and McpX is sufficient to allow a low level of polar co-localization in B. pseudofirmus OF4.
The finding that mutational loss of NaVBP leads to delocalization of McpX raises the possibility that such delocalization is in part responsible for the altered chemotactic responses rather than loss of channel function itself being completely responsible. However, the functional capacity of NaVBP is also essential for normal chemotaxis. Nifedipine addition to chemotaxis assay buffers results in inverse chemotaxis (Ito et al., 2004b) but does not result in delocalization of NaVBP and McpX. This supports the possibility that NaVBP-mediated fluctuations in membrane potential could play a role in the signaling pathway for chemotaxis of B. pseudofirmus OF4, as was raised for channels not yet identified in earlier studies of chemotaxis in Spirochaeta aurantia (Goulbourne & Greenberg, 1983) and E. coli (Tisa et al., 1993; Tisa et al., 2000). The co-localization of NaVBP and MCPs further raises the possibility that modulation of NavBP function could occur via dynamic changes in NavBP phosphorylation as mediated by chemoreceptor activation of the CheA kinase and CheZ phosphatase system that presumably are also present at the cell poles (Baker et al., 2006). This could lead to changes in NavBP gating as occurs in many eukaryotic voltage-gated ion channels (Levitan, 1999). The cytoplasmic C-terminus of NavBP contains four aspartate residues that could serve as potential phosphoacceptors for the CheA kinase; changes in C-terminal phosphorylation of eukaryotic channels can profoundly affect channel gating (Park et al., 2006). Further work will be needed to define the nature of MCP and channel interactions in the alkaliphile and how such interactions affect channel function and chemotaxis. Finally, these studies raise the question of whether, as inferred by the earlier investigators cited above, voltage-gated ion channels are involved in signaling events or other features of bacterial chemotaxis in bacteria beyond a sub-set of extremophiles.
We are grateful to Prof. G. W. Ordal at University of Illinois for providing polyclonal antibodies that react with McpB of B. subtilis. This research was supported by a National Institutes of Health grants GM28454 (to T.A.K.), GM68658 (to D.E.C.), and NS34383 (to J.S.T.), and a Grant-in-Aid for Scientific Research (C) (17613004) and the 21st Century Center of Excellence program of the Ministry of Education, Culture, Sports, Science and Technology of Japan and a special Grant-in-Aid from the Toyo University (to M.I.).