SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The Salmonella flagellar motor consists of a rotor and about a dozen stator elements. Each stator element, consisting of MotA and MotB, acts as a proton channel to couple proton flow with torque generation. A highly conserved Asp33 residue of MotB is directly involved in the energy coupling mechanism, but it remains unknown how it carries out this function. Here, we show that the MotB(D33E) mutation dramatically alters motor performance in response to changes in external load. Rotation speeds of the MotA/B(D33E) and MotA(V35F)/B(D33E) motors were markedly slower than the wild-type motor and fluctuated considerably at low load but not at high load, whereas the rotation rate of the wild-type motor was stable at any load. At low load, pausing events were frequently observed in both mutant motors. The proton conductivities of these mutant stator channels in their ‘unplugged’ forms were only half of the conductivity of the wild-type channel. These results suggest that the D33E mutation induces a load-dependent inactivation of the MotA/B complex. We propose that the stator element is a load-sensitive proton channel that efficiently couples proton translocation with torque generation and that Asp33 of MotB is critical for this co-ordinated proton translocation.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The bacterial flagellum, which enables bacteria to swim towards regions where conditions are more favourable, consists of a rotary motor, a universal joint and a helical propeller. The flagellar motor of Salmonella enterica serovar Typhimurium (hereafter referred to as Salmonella) consists of a reversible rotor and about a dozen stator elements. It is powered by downhill proton translocation along the electrochemical potential of protons across the cytoplasmic membrane (Berg, 2003; Minamino et al., 2008; Sowa and Berry, 2008). The stator element is composed of four copies of MotA and two copies of MotB (Kojima and Blair, 2004), and it acts as a proton channel to couple the proton flow through a proton channel with torque generation (Blair and Berg, 1990; Stolz and Berg, 1991). The MotA/B complex is thought to be anchored to the peptidoglycan layer through an OmpA-like motif within the C-terminal periplasmic domain of MotB (MotBC) upon assembly into the motor (De Mot and Vanderleyden, 1994; Roujeinikova, 2008; Kojima et al., 2009). It has been shown that the flagellar motor regulates the number of stator elements around the rotor in response to changes in external load (Lele et al., 2013; Tipping et al., 2013). FliG, FliM and FliN form the C ring on the cytoplasmic face of the MS ring of the basal body, which is made of FliF (Francis et al., 1994). The MS-C ring complex acts as a reversible rotor. Electrostatic interactions between MotA and FliG are responsible not only for torque generation (Zhou et al., 1998a) but also for stator assembly around the rotor (Morimoto et al., 2010b; 2013).

Asp33 of Salmonella MotB, which is located near the cytoplasmic end of its transmembrane helix, is absolutely conserved among MotB orthologues and is functionally critical for proton translocation through the MotA/B proton channel complex (Zhou et al., 1998b). The MotB(D33N) mutation totally abolishes motor function (Zhou et al., 1998b). Residues 52 to 71 of MotB act as a plug that prevents the MotA/B complex from leaking protons when it is not assembled into the motor (Hosking et al., 2006; Morimoto et al., 2010a). Both protonation and deprotonation of this aspartic acid residue cause conformational changes in the cytoplasmic loop of MotA that may drive flagellar motor rotation (Kojima and Blair, 2001). The conserved Pro173 residue of MotA is in relatively close proximity to MotB-Asp33 and facilitates the conformational changes of the stator that support the rapid mechanochemical cycle of the motor (Kim et al., 2008; Nakamura et al., 2009b).

Precise measurement of the motor rotation gives us physiological insights into the mechanism of motor rotation. At low speed near stall, the motor operates close to thermodynamic equilibrium, and hence the rotation rate of the flagellar motor is simply limited by external load rather than by the rates of internal processes in the mechanochemical cycle of the motor, including the proton conductivity of the MotA/B channel complex. At high speed under low load, however, the motor runs far from equilibrium, and hence the rate of high-speed motor rotation under low-load conditions is limited by the rate of the mechanochemical cycle of the motor (Chen and Berg, 2000a; 2000b). We have previously shown that the MotB(D33E) mutation causes c. 50% reduction in near-stall torque and a sharp decline in the torque-speed curve, with an apparent maximal rotation rate of c. 20 Hz. The suppressor mutations including the MotA(V35F) substitution restore the near-stall torque to the wild-type level in conjunction with the MotB(D33E) replacement, but it does not reverse the sharp decline of the torque-speed curve and the maximum rotation rate associated with the MotB(D33E) substitution. These results suggest that the MotB(D33E) mutation decreases proton conductivity and the rate of conformational changes of the stator complex involved in torque generation steps and that the suppressor mutations restore nearly wild-type performance in the torque generation step but not in the proton translocation step (Che et al., 2008). However it remained unknown whether the proton conductivity was actually the limiting factor of the high-speed rotation at low load.

To understand the energy coupling mechanism of the flagellar motor, we analysed motor performance of fully assembled MotA/B(D33E) and its MotA(V35F)/B(D33E) suppressor mutant motors in more detail. We show that the proton conductivities of unplugged proton channels of the MotA/B(D33E) and MotA(V35F)/B(D33E) stator complexes are c. 50% and 60% of the unplugged wild-type MotA/B channel activity. High-resolution bead assays show that rotation speeds of the MotB(D33E) and MotA(V35F)/MotB(D33E) motors fluctuate considerably at low load but not at high load near stall, whereas the rotation speeds of the wild-type motor are stable under any load conditions. These results suggest that the MotA/B complex is a load-sensitive proton channel that efficiently couples proton translocation with the conformational changes that generate torque. The mechanical reaction in response to load facilitates progress of the coupled steps, and the MotB(D33E) mutation somehow changes the conformational dynamics to reduce the load sensitivity of the channel complex.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Effect of the MotB(D33E) mutation on stator assembly around the rotor

To obtain precise physical parameters for the MotB(D33E) and MotA(V35F)/MotB(D33E) motors, we performed rotation measurements of individual motors of these mutants by bead assays with polystyrene beads with a diameter of 1.0 μm, using various concentrations of Ficoll to change the external load on the motor (Fig. 1). This method allowed for more reliable and quantitative measurements than our previous experiments, in which the rotation rates of many motors labelled with beads of different sizes were measured to collect the data for torque-speed curves (Che et al., 2008). Like the wild-type MotA/B complex (Ryu et al., 2000; Yuan and Berg, 2008), upon induction of these mutant MotA/B complexes in the motAB null strain with 0.1 mM arabinose, a stepwise increment in speed was observed, and then the rotational speeds of these motors reached their maximum levels (data not shown), indicating that they are fully assembled motors with the maximal number of stators. Figure 1A is a record of an experiment showing the rotation speed of a single 1.0 μm bead attached to a partially sheared sticky filament of the fully assembled wild-type motor in motility buffers containing Ficoll with a range of concentrations from 0% to 12% (w/v). As the concentration of Ficoll was increased step-by-step by 3% at each step, the motor speed decreased from ∼ 80 to ∼ 15 Hz. The speed recovered when the medium was exchanged back to the motility buffer without Ficoll, indicating that motor function was not irreversibly perturbed by the buffer exchanges. The torque of the wild-type motor was approximately constant over a wide range of speeds under our experimental conditions (Fig. 1B and Fig. S1). In contrast, the stall torque produced by the fully assembled MotB(D33E) mutant motor was approximately half of the wild-type level, and the motor torque declined almost linearly when the motor speed increased by lowering the concentration of Ficoll (Fig. 1B). The MotA(V35F) suppressor mutation recovered the stall torque to the wild-type level but not to the maximum speed at zero load (Fig. 1B). These results are in good agreement with previous data (Che et al., 2008).

figure

Figure 1. Motor speed under different fluid viscosity, and torque–speed relationships of the wild-type and two mutant motors.

A. Speed of a single wild-type motor measured using a 1.0 μm (diameter) bead in media containing 0%, 3%, 6%, 9% and 12% Ficoll (w/v).

B. Torque-speed relationship of the wild-type (closed circles), MotA/B(D33E) (closed triangles) and MotA(V35F)/B(D33E) (closed squares) motors. Each symbol represents the average value for five motors.

Download figure to PowerPoint

It has been shown that there are as many as 11 stator elements around the rotor (Reid et al., 2006). The stall torque produced by the flagellar motor is proportional to the number of the stator elements around the rotor (Ryu et al., 2000). It has been shown that GFP-MotB forms a punctate localization pattern at the base of a filament labelled with a fluorescent dye (Morimoto et al., 2010b). To investigate whether the D33E substitution affects the number of stators around the rotor, we constructed Salmonella strains that encode GFP-MotB(D33E) on the chromosome under the control of the promoter of the mocha operon, which includes motAB as its first two genes. We analysed their subcellular localization of the GFP-tagged MotB by epi-illumination fluorescence microscopy and measured the number and intensity of fluorescent spots of GFP-MotB(D33E) (Fig. 2). We used a Salmonella strain encoding GFP-tagged wild-type MotB as a positive control. Similar punctate patterns were observed for GFP-MotB and GFP-MotB(D33E) (Fig. 2A). More than 85% of the cells expressing GFP-MotB and GFP-MotB(D33E) had more than one fluorescent spot, and the number of spots in cell expressing GFP-MotB(D33E) was essentially the same as with GFP-MotB (Fig. 2B). The average intensity of the GFP-MotB(D33E) spots was also essentially the same as that of the GFP-MotB spots (Fig. 2C). These results indicate that the MotB(D33E) mutation does not affect the subcellular localization of GFP-MotB. In agreement with this, MotA/B(D33E) exerted a dominant negative effect on motility of wild-type cells (Fig. S2). Therefore, we conclude that the D33E replacement does not significantly decrease the number of MotA4MotB2 stator elements around the rotor.

figure

Figure 2. Effect of the MotB(D33E) mutation on GFP-MotB localization.

A. Epifluorescence images of YVM003 (GFP-MotB) and YVM008 [GFP-MotB(D33E)]. The cells were incubated in LB at 30°C for 12 h and observed by epifluorescence microscopy. Measurements were performed at 23°C.

B. Fraction of cells with different numbers of GFP-MotB and GFP-MotB(D33E) spots in each strain. More than 280 cells were counted.

C. Relative fluorescence intensities of GFP-MotB and GFP-MotB(D33E) spots in each strain. Fifty spots were counted.

Download figure to PowerPoint

Effect of the MotB(D33E) substitution on conductivity of the proton channel

Because the maximal motor rotation speed is thought to be determined by the conductivity of the proton channel, we tested whether the MotB(D33E) substitution significantly decreased the proton-channel activity of the MotA/B complex and whether the MotA(V35F) suppressor mutation partially restored it. Because it has been shown that the plug segment of MotB suppresses massive proton flow through the MotA/B proton channel complex when it is not assembled into a motor (Hosking et al., 2006; Morimoto et al., 2010a), we deleted residues 52 to 71, including the plug segment, from the MotB protein and analysed the growth rate of Salmonella cells harbouring pBAD24-based plasmids encoding the unplugged MotA/B(Δ52–71/D33E) or MotA(V35F)/B(Δ52–71/D33E) proton channel (Fig. 3A). We used the unplugged MotA/B(Δ52–71) and MotA/B(Δ52–71/D33N) proton channels as the positive and negative controls respectively. Induction of MotA/B(Δ52–71) by adding 1 mM arabinose totally inhibited cell growth whereas that of MotA/B(Δ52–71/D33N) did not, in agreement with previous reports (Hosking et al., 2006; Morimoto et al., 2010a). Induction of MotA/B(Δ52–71/D33E) or MotA(V35F)/B(Δ52–71/D33E) still resulted in a significant growth defect, although not to the level of MotA/B(Δ52–71). These results suggest that the D33E mutation does not suppress the massive proton flow through the unplugged channel that causes growth impairment.

figure

Figure 3. Examination of the proton conductivity of unplugged wild-type and mutant MotA/B proton channels.

A. Growth curve of SJW2241 (ΔmotAB) cells transformed with pBAD24 (vector, labelled V) (open triangles), pYC109 [MotA/B(Δ52–71), indicated as WT] (closed circles), pYC110 [MotA/B(Δ52–71/D33E), indicated as D33E] (closed triangles), pYC111 [MotA(V35F)/B(Δ52–71/D33E), indicated as V35F/D33E] (closed squares) or pYC112 [MotA/B(Δ52–71/D33N), indicated as D33N] (open circles). The arrow indicates the time when 1 mM arabinose was added.

B. Measurement of intracellular pH change of YSC2302 (ΔmotA-B, ΔfliC::pHluorin) carrying the above plasmids after addition of 1 mM arabinose at an external pH of 7.0. The intracellular pH was measured using the pHluorin probe in 15 min after induction of MotA/B(Δ52–71), MotA/B(Δ52–71/D33E) or MotA(V35F)/B(Δ52–71/D33E).

C. Immunoblotting, using polyclonal anti-MotB antibody, of whole cell proteins of the same transformants. The arrowhead indicates the position of MotB(Δ52–71).

Download figure to PowerPoint

To test whether this growth defect is due to proton leakage, we measured the intracellular pH of Salmonella cells at an external pH value of 7.0 after 15 min of induction of MotA/B(Δ52–71), MotA/B(Δ52–71/D33E) or MotA(V35F)/B(Δ52–71/D33E) with arabinose (Fig. 3B). Although the intracellular pH of cells carrying pBAD24 (vector) or pYC112 encoding MotA/B(Δ52–71/D33N) did not change significantly, the intracellular pH decreased by approximately 0.51 units after induction of MotB(Δ52–71). The intracellular pH of cells expressing MotA/B(Δ52–71/D33E) and MotA(V35F)/B(Δ52–71/D33E) dropped by c. 0.27 and 0.32 units respectively (Fig. 3B). The expression level of MotB was slightly higher in these two mutants than in the motB(Δ52–71) mutant and almost the same as that seen in the motB(Δ52–71/D33N) mutant (Fig. 3C). From the expression level and the decrease in the intracellular pH, the proton conductivities of unplugged proton channels of the MotA/B(D33E) and MotA(V35F)/B(D33E) complexes were estimated to be c. 50% and 60% of the unplugged wild-type channel respectively.

Effect of the D33E mutation on the stability of rotation speed

It has been reported that accelerations, decelerations and pauses of motor rotation are occasionally observed during rotation measurements (Eisenbach et al., 1990; Kudo et al., 1990; Muramoto et al., 1995; Bai et al., 2010). Therefore, we measured the rotation speed (ω) of fully assembled wild-type and mutant motors at several different loads (Figs S3 and S4) and calculated the average speed (ωav) and the standard deviation of ω (σω) (Fig. 4). The ratio of σω/ωav was plotted as a function of external load (Fig. 4, third row). The value of σω/ωav for the wild-type motor was about 0.075 over a wide range of load (Fig. 4A). This indicates that the wild-type motor can spin stably regardless of the load level, in agreement with a previous report (Yuan and Berg, 2008). Rotation speeds of the MotA/B(D33E) and MotA(V35F)/B(D33E) motors were stable at high load (Fig. 4B and C, first row) but not at low load (Fig. 4B and C, second row). The values of σω/ωav for the MotA/B(D33E) and MotA(V35F)/B(D33E) motors gradually increased with decreasing load and ranged from 0.17 to 0.44 and from 0.08 to 0.36, respectively, when the filaments were labelled with 0.6 μm beads (Fig. 4B and C, third rows). These results indicate that their rotation rates fluctuate considerably at low load but not at high load near stall. When we measured rotation speeds of these two mutant motors for a much longer period of time, pauses were observed frequently at low load but not at high load, whereas the wild-type motor did not show any pauses (Fig. 5 and Fig. S5). These results suggest that these mutant motors cannot produce torque constantly at low load.

figure

Figure 4. Dependence of flagellar rotation rate and speed fluctuation on external load. Rotation measurements of YSC2123 [ΔmotAB, fliC(Δ205–293), ΔcheY, ΔfimA] cells transformed with (A) pYC20 (MotA/B, indicated as WT), (B) pYC21 [MotA/B(D33E), indicated as D33E] or (C) pYC22 [MotA(V35F)/B(D33E), indicated as V35F/D33E] were carried out at c. 25°C by tracking the position of 2.0 μm (upper panel) and 0.6 μm (middle panel) beads attached to partially sheared flagellar filaments. Speed histograms are shown on the right. Speed fluctuations (σω/ωav) are shown as a function of external load (lower panel). The values of the average speeds (ωav) and their standard deviations (σω) were obtained by tracking the position of 2.0 μm (filled circles), 1.5 μm (open circles), 1.1 μm (filled triangles), 0.8 μm (open triangles) and 0.6 μm (filled squares) beads for 7 s.

Download figure to PowerPoint

figure

Figure 5. Pausing in motor rotation under different load conditions. Fraction of time spent pausing during 1 min of rotation is shown as a function of external load for YSC2123 cells expressing (A) MotA/B (WT), (B) MotA/B(D33E) (D33E) or (C) MotA(V35F)/B(D33E) (V35F/D33E).

Download figure to PowerPoint

To examine whether the function of these mutant motors is irreversibly abolished at low load, we monitored the performances of individual flagellar motors labelled with 0.8 μm beads in the presence and absence of 10% Ficoll (Fig. 6). The wild-type motor rotated at c. 20 Hz in the presence of 10% Ficoll. When the motors were exposed to motility buffer without Ficoll, the rotation rate of the motors increased to c. 130 Hz without showing much fluctuations, and when the medium was exchanged back to motility buffer containing 10% Ficoll, the motor speed was reduced to its original level (upper panel). In the presence of 10% Ficoll, the MotA/B(D33E) and MotA(V35F)/B(D33E) mutant motors rotated stably at c. 6 Hz and 14 Hz respectively. When the medium was exchanged to motility buffer without Ficoll, their rotation rates increased and fluctuated considerably, but when the medium was exchanged back to motility buffer with Ficoll, they rotated slower and stably again. These results demonstrate that the speed fluctuations of these mutant motors are load dependent.

figure

Figure 6. Reversibility of the effect of viscosity on the rotation rate of the motor. Rotation rates were measured by tracking the position of a 0.8 μm bead attached to the sticky filament of YSC2123 cells carrying (A) pYC20, (B) pYC21 or (C) pYC22 in motility buffer containing 10% Ficoll. Then, the medium was changed to motility buffer without Ficoll and then replaced with buffer containing 10% Ficoll.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The ability of molecular motors to sense force is critical for various cellular processes. Myosin, which is an ATP-driven linear motor, alters its ATPase activity and mechanical properties in response to changes in load (Veigel et al., 2003; Laakso et al., 2008). The bacterial flagellum acts as a mechanosensor to detect changes in the environment (Anderson et al., 2010). Restricting the rotation of the flagella by increasing the viscosity of the medium induces cell differentiation and hyper flagellation, allowing cells to move on the surface of agar plates (Harshey, 2003). Recently, it has been shown that the number of stator elements driving the flagellar motor is much less at low load than at high load (Lele et al., 2013; Tipping et al., 2013), suggesting that the stator elements also act as dynamic mechanosensors to regulate the stator number around the motor in response to changes in external load. It has been reported that a decrease in the sodium ion influx through the sodium ion channel of the Vibrio alginolyticus polar flagellar motor triggers cell differentiation and induces lateral flagellar gene expression (Kawagishi et al., 1996), raising the possibility that the ion-conducting activity of the stator complex is directly linked to biological processes that induce cell differentiation, but it remains unknown how. In this study, we analysed the fully assembled MotA/B(D33E) and MotA(V35F)/MotB(D33E) mutant motors and found evidence suggesting that the MotB-Asp33 residue plays important roles not only in proton translocation through the MotA/B ion channel but also in a load-dependent coupling mechanism for torque generation.

The rate of protonation and deprotonation of MotB-Asp33 is limited by the conformational dynamics of the MotA/B complex

Electrostatic interactions between a cytoplasmic loop of MotA and FliG are required for normal torque generation (Zhou et al., 1998a). Both protonation and deprotonation of the highly conserved MotB-Asp33 residue, which is the major proton binding site in the MotA/B channel, cause conformational changes of the cytoplasmic loop in MotA that interacts with FliG. Thus, MotB-Asp33 is critical for energy coupling between proton translocation through the channel and torque generation (Kojima and Blair, 2001). The MotB(D33E) mutation causes a 50% reduction in stall torque and a 90% reduction in the maximal rotation rate under low load (Che et al., 2008). The compensating suppressor mutations including the MotA(V35F) mutation restore stall torque to the wild-type level but do not support the maximal rotation rate at low load (Che et al., 2008). Here, we showed that the D33E substitution does not decrease the number of stator elements around the rotor (Fig. 2) and confirmed that the decrease in stall torque is not due to the reduction in the number of stator elements around the motor. Because the rotation rate of the motor is not limited by the rate of the mechanochemical cycle of the motor at low speed near stall (Chen and Berg, 2000b), we suggest that in the MotB(D33E) motor the cytoplasmic loop of MotA is misaligned relative to FliG at the stator–rotor interface. This mismatch causes the 50% reduction in the energy coupling efficiency of the motor. In this view, the MotA(V35F) suppressor mutation readjusts the alignment, thereby restoring the coupling efficiency to the wild-type level as reported by Garza et al. (1996a,b).

The near zero-load speed of the wild-type motor was c. 250 Hz (Fig. S4), whereas those of the MotA/B(D33E) and MotA(V35F)/B(D33E) mutant motors were estimated to be 20 Hz and 30 Hz respectively (Fig. 1B). The recovery in the zero-load speed by the MotA(V35F) suppressor mutation is rather small, although the torque generation step that requires precise stator–rotor interactions is restored by the MotA(V35F) mutation as described above. The rate of proton flow through the unplugged ion channels of the MotA/B(D33E) and MotA(V35F)/B(D33E) complexes were estimated to be c. 50% and c. 60% of that through unplugged wild-type channel (Fig. 3). Thus, the decreased proton conductivity through the channel itself is not the main cause of the marked reduction in the maximal rotation speed at low load. The flagellar motor driven by a single stator element shows 26 steps per revolution (Sowa et al., 2005; Nakamura et al., 2010). Each step consists of at least two distinct processes: a torque generation step involving stator–rotor interactions and a step for recovery of the original stator–rotor geometry (Sowa et al., 2005; Nakamura et al., 2010). Because proton translocation through the MotA/B proton channel complex is coupled with the mechanochemical cycle of the motor, the rate of proton flow must be limited by the rate of conformational change of the MotA/B complex. Therefore, we suggest that the 90% decrease in the maximal rotation speed of the MotB(D33E) motor is due to a decreased rate of the conformational change of the MotA/B complex coupled with protonation and deprotonation of MotB-Glu33 in the mechanochemical cycle of the motor.

MotB-Asp33 contributes to co-ordinated proton translocation in response to external load

The stator element has a high duty ratio and hence remains attached to the rotor for most of the torque-generation cycle (Ryu et al., 2000; Yuan and Berg, 2008). As shown previously (Yuan and Berg, 2008), the rotation speed of the wild-type motor is very stable over a wide range of rotation rates (Fig. 4). In contrast, the rotation speeds of the MotA/B(D33E) and MotA(V35F)/B(D33E) motors showed large fluctuations at low load although they were stable at high load (Figs 4 and 6). Frequent pauses were also observed for these mutant motors at low load (Fig. 5 and Fig. S5), indicating that these mutant motors have difficulty maintaining the torque generation cycle when the load against rotation is small. We analysed the speed fluctuation of the wild-type motor when PMF was reduced to half of the physiological level by CCCP and found that the wild-type motor showed no speed fluctuation (data not shown). These results suggest that the wild-type MotA/B complex can sense even a small load to continue the mechanochemical reaction cycle for torque generation by efficiently coupling proton translocation with its conformational change. The MotB(D33E) mutation somehow changes the conformational dynamics of the MotA/B complex to reduce its sensitivity to external load, and therefore the conformational changes of the MotA/B(D33E) and MotA(V35F)/B(D33E) mutant complexes cannot be efficiently driven at low load.

The rotation rate of the proton-driven flagellar motor is proportional to the proton motive force (PMF) across the cytoplasmic membrane (Manson et al., 1980; Gabel and Berg, 2003). It has been shown that acceleration and deceleration of flagellar motor rotation can be induced by an external electric field (Kami-ike et al., 1991; Fung and Berg, 1995). But, as the PMF values of the two mutants we used were the same as that of wild-type cells (data not shown) and because the PMF should not change with changing external load on the motor, we can exclude the possibility that the fluctuations of motor speed at low load result from changes in PMF. It has been shown that the stator elements act as dynamic mechanosensors and change their structure in response to changes in external load (Lele et al., 2013; Tipping et al., 2013). We therefore propose that the stator element acts as a load-sensitive proton channel that efficiently couples proton translocation with torque generation and that MotB-Asp33 is important for this co-ordinated proton translocation.

Mechanosensitive ion channels have been discovered in both eukaryotes and prokaryotes. Mechanosensitive ion channels respond to a wide range of external mechanical stimuli to control their conductivity. In eukaryotes, mechanosensitive ion channels require are tethered to the cytoskeleton and/or extracellular matrix in order to sense the force that induces gate opening (Sackin, 1995). The MotA/B stator complex is anchored to the peptidoglycan layer through the MotBC domain (De Mot and Vanderleyden, 1994). When bacteria swim through viscous environments, torque generated by the stator–rotor interactions should produce an equal and opposite force on the MotA/B complex because it is anchored to the peptidoglycan layer. Therefore, we propose that stator–rotor interactions directly transmit force to the proton channel of the MotA/B complex to co-ordinate proton translocation across the cytoplasmic membrane in response to external load.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Bacterial strains, plasmids, P22-mediated transduction, DNA manipulations and media

Salmonella strains and plasmids used in this study are listed in Table 1. P22-mediated transductional crosses were carried out using p22HTint as described (Yamaguchi et al., 1984). DNA manipulations were carried out as described before (Saijo-Hamano et al., 2004). Site-directed mutagenesis was carried out using the QuikChange Site-Directed Mutagenesis method as described in the manufacturer's instruction (Stratagene). Salmonella YVM008 [gfp-motB(D33E)] strain was constructed using the λ Red homologous recombination system (Datsenko and Wanner, 2000) as described previously (Hara et al., 2011). DNA sequencing was performed with an ABI 3130 Genetic Analyser (Applied Biosystems). L-broth, T-broth and motility medium were prepared as describe before (Minamino and Macnab, 1999; Minamino et al., 2003). A 100 μg ml−1 ampicillin solution was added to the medium as needed.

Table 1. Strains and plasmids used in this study
Strains and plasmidsRelevant characteristicsSource or reference
Escherichia coli  
NovaBlueRecipient for cloning experimentsNovagen
Salmonella  
YSC2123ΔmotAB, fliC(Δ204–292), ΔcheY, ΔfimAMorimoto et al. (2013)
YSC2302ΔmotAB, ΔfliC::pHluorinMorimoto et al. (2013)
YVM003gfp-motBMorimoto et al. (2010b)
YVM008gfp-motB(D33E)This study
Plasmids  
pBAD24Cloning vectorGuzman et al. (1995)
pYC20pBAD24/MotA + MotBMorimoto et al. (2010a)
pYC21pBAD24/MotA + MotB(D33E)This study
pYC22pBAD24/MotA(V35F) + MotB(D33E)This study
pYC98pBAD24/MotA + MotB(D33N)This study
pYC109pBAD24/MotA + MotB(Δ52–71)Morimoto et al. (2010a)
pYC110pBAD24/MotA + MotB(Δ52–71/D33E)This study
pYC111pBAD24/MotA(V35F) + MotB(Δ52–71/D33E)This study
pYC112pBAD24/MotA + MotB(Δ52–71/D33N)This study

Fluorescence microscopy

Salmonella cell bodies and epifluorescence of GFP-MotB and GFP-MotB(D33E) were observed as described previously (Morimoto et al., 2010b). Fluorescence image processing was carried out with the ImageJ version 1.47 software (National Institutes of Health) as described before (Morimoto et al., 2013).

Measurement of intracellular pH

Intracellular pH measurements with a ratiometric fluorescent pH indicator protein, pHluorin (Miesenböck et al., 1998), were carried out as described before (Morimoto et al., 2010a). Intracellular pH measurements for YSC2302 (ΔmotAB, ΔfliC::pHluorin) expressing MotA/B and its mutant variants were carried out at an external pH value of 7.0 as described before (Nakamura et al., 2009a; Morimoto et al., 2010a).

Immunoblotting

Whole cellular proteins were prepared as described before (Minamino and Macnab, 1999). After SDS-PAGE, immunoblotting with polyclonal anti-MotB antibody was done as described (Minamino and Macnab, 1999). Detection was performed with an ECL plus immunoblotting detection kit (GE Healthcare).

Bead assay for motor rotation

Salmonella YC2123 [ΔmotAB ΔcheY ΔfimA fliC(Δ204–292)] strain harbouring a pBAD24-based plasmid encoding the MotA/B complex were grown in T-broth containing 50 μg ml−1 ampicillin and 0.1 mM l-arabinose at 30°C for 5 h with shaking. Because YC2123 lacks the cheY gene, the motors rotate exclusively CCW. Bead assays with polystyrene beads with diameters of 2.0 μm, 1.5 μm, 1.0 μm, 0.8 μm or 0.6 μm (Invitrogen) were carried out as described (Che et al., 2008; Nakamura et al., 2009a). All data were recorded at 1.0 ms intervals by a 16-bit A-D board (Microscience), using the LaBDAQ software package (Matsuyama Advance). Torque calculation was done as described before (Che et al., 2008).

The ratio of σω/ωav was calculated as described before (Muramoto et al., 1995). Speed (ω) were computed from a peak in the FFT spectrum, using data windows of length 1 s beginning at intervals of 0.1 s as described (Reid et al., 2006). The statistical values used in this work were calculated as follows:

  • display math
  • display math

ωav is average of ω for 7 s data, σω is standard deviation of ωav, N is total number of FFT and n is the event number.

To determine duration of pause during a sampling period of time (60 s), we obtained motor angles by fitting an ellipse to bead trajectories under the assumption that trajectories represent the projection of a circular orbit onto a quadrant-type photodiode and then calculated angler velocity. We fitted the angular velocity using linear regression with 100 ms width and 50 ms interval. The pausing states were defined as periods, of which rotational fluctuations were smaller than the standard deviation of the X and Y signals of beads attached to the flagellar filament of a motAB null mutant. Pausing frequency means a sum of duration of pause during the sampling period.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank May Macnab, Fan Bai and Katsumi Imada for critical reading of the manuscript and helpful comments. This work was supported in part by JSPS KAKENHI Grant No. 24770141 (to S.N.), and 21227006 and 25000013 (to K.N.), and by a Grant-in-Aid for Scientific Research on Innovative Areas ‘Spying minority in biological phenomena’ (23115008 to T.M.) of MEXT, Japan.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
mmi12453-sup-0001-si.pdf1122KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.