Dynamics of mechanosensing in the bacterial flagellar motor

Abstract

Mechanosensing by flagella is thought to trigger bacterial swarmer-cell differentiation, an important step in pathogenesis. How flagellar motors sense mechanical stimuli is not known. To study this problem, we suddenly increased the viscous drag on motors by a large factor, from very low loads experienced by motors driving hooks or hooks with short filament stubs, to high loads, experienced by motors driving tethered cells or 1-μm latex beads. From the initial speed (after the load change), we inferred that motors running at very low loads are driven by one or at most two force-generating units. Following the load change, motors gradually adapted by increasing their speeds in a stepwise manner (over a period of a few minutes). Motors initially spun exclusively counterclockwise, but then increased the fraction of time that they spun clockwise over a time span similar to that observed for adaptation in speed. Single-motor total internal reflection fluorescence imaging of YFP-MotB (part of a stator force-generating unit) confirmed that the response to sudden increments in load occurred by the addition of new force-generating units. We estimate that 6-11 force-generating units drive motors at high loads. Wild-type motors and motors locked in the clockwise or counterclockwise state behaved in a similar manner, as did motors in cells deleted for the motor protein gene fliL or for genes in the chemotaxis signaling pathway. Thus, it appears that stators themselves act as dynamic mechanosensors. They change their structure in response to changes in external load. How such changes might impact cellular functions other than motility remains an interesting question.

Keywords: Escherichia coli; mechanical load; stator remodeling.

Fig. 1.

Adaptation to mechanical stimuli: (A) A cell is brought up to a 1-μm latex bead held in an optical trap (trap not shown). The bead binds to a short sticky-filament stub and its rotation is monitored. (B) A cell with a short sticky-filament stub is allowed to settle onto the surface of a glass cover-slip and self-tether. The rotation of the cell body is monitored. (C) Bead rotational speeds measured from the time of attachment, plotted as a function of time. Positive values are CCW; negative values are CW. The black line is the mean value of the absolute speed for the interval between successive steps. The arrow represents the time of mechanical stimulus. Dashed lines indicate speeds before increase in viscous loads. (D) Steady-state torque–speed curves for CCW rotation of the bacterial flagellar motor (solid curve). Two load lines are shown (dashed lines), one for high loads (Left, with a steep slope) and one for low loads (Right, with near-horizontal slope). The square near 300 Hz (state a) indicates the torque delivered by a single stator element to the motor at low loads. The filled circles (from state b to state c) represent the driving torques for the corresponding speeds at which the 1-μm bead rotates in Fig. 1C. The dotted lines indicate the driving torque on the motor for stator elements ranging in number from 1 to 11.

Fig. 2.

Stator remodeling upon mechanical stimulus: (A) YFP–MotB speed increments observed during adaptation of a single tethered cell. (B) The corresponding raw TIRF images for each step in speed are shown. The corresponding motor intensity versus speed data, collected beginning at 0, 5, and 9 min (a, b, and c, respectively), were fit by a line. (C) The average of motor intensities minus the initial motor intensity (13 motors) versus the corresponding motor speeds. (D) Distribution of cell rotational speeds immediately upon tethering.

Fig. 3.

Remodeling of motors locked in the CW or CCW state: Stator remodeling following bead attachment in exclusively CCW rotating motors (A) and exclusively CW rotating motors (B). The black line is the mean value of the absolute speed for the interval between successive steps.

Fig. 4.

Kinetics of adaptation: (A) Adaptation in average CW_bias. The solid curves are exponential fits to the data points. The rate constants are k_cheRcheB ∼0.48 ± 0.00 min⁻¹ (nine motors) and k_WT ∼0.36 ± 0.01 min⁻¹ (13 motors). (B) Adaptation in motor speed, with speeds normalized by the speeds at steady state. The solid curves are exponential fits to the data points. The rate constants are ξ_cheRcheB ∼0.30 ± 0.01 min⁻¹ (nine motors) and ξ_WT ∼0.36 ± 0.01 min⁻¹ (13 motors).