The Bacterial Flagellar Rotary Motor
MotAB Proton Stators, Hook Universal Joints, and Run-and-Tumble Chemotactic Memory
For nearly a century, microscopists assumed bacteria swam by waving flexible tails like miniature eels or eukaryotic sperm. Howard Berg and Robert Anderson proved in 1973 that the bacterial flagellum has zero internal motor proteins along its shaft. Instead, it is a rigid helical propeller spun from its base by a true rotary bearing in the cell wall.
Michael Silverman and Melvin Simon made that nanoscale rotation visible to the human eye in 1974. Using anti-flagellin antibodies, they glued a single flagellar stub of an Escherichia coli cell onto a glass microscope slide. The motorized basal bearing spun the entire two-micrometer cylindrical bacterial body in circles at 10 revolutions per second.
Edward Purcell explained in his 1977 lecture Life at Low Reynolds Number why bacteria need a rotary corkscrew. For a two-micrometer cell swimming at 30 micrometers per second in water, the Reynolds number is $10^{-4}$—hydrodynamically identical to a human swimming through warm tar. Because inertia is zero, any symmetrical back-and-forth stroke moves net zero distance.
Driven by the inward flow of roughly 1,200 protons per revolution across a $150\text{-mV}$ membrane potential, the E. coli flagellar motor spins freely at 300 hertz ($18,000\text{ rpm}$). In the marine bacterium Vibrio alginolyticus, a sodium-ion-driven polar flagellar motor spins even faster. It reaches 1,700 hertz ($102,000\text{ rpm}$)—five times faster than a Formula One racing engine.
Passing through the inner membrane, peptidoglycan wall, and outer membrane via concentric protein bushings (the MS, P, and L rings), the 45-nanometer cytoplasmic C-ring rotor forms the base of the drive train. Up to eleven $\text{MotA}_5\text{MotB}_2$ stator complexes surround the rotor's outer rim. Each stator is bolted directly to the rigid peptidoglycan cell wall.
Cryo-EM structures solved at $3.1\text{ \AA}$ in 2020 by Susan Lea and Nicholas Taylor revealed how $\text{MotA}_5\text{MotB}_2$ converts proton flow into torque. The stator is itself a miniature rotary wheel. As protons neutralize conserved aspartates on the central $\text{MotB}_2$ post, the surrounding $\text{MotA}_5$ pentamer spins like a pinion gear against the large C-ring rotor.
Pushkar Lele and Howard Berg discovered in 2013 that the bacterial flagellar motor has an automatic mechanical transmission that recruits stators on demand. Spinning against near-zero load in water, the rotor holds only one or two MotAB stators. When external viscosity rises, higher mechanical torque stabilizes binding hooks so up to eleven stators snap into the ring.
Connecting the basal drive shaft to the 10-micrometer-long flagellar filament is a 55-nanometer curved tubular coupling built from 120 copies of the FlgE protein: the hook. Acting as a flexible universal joint, the hook bends rotary torque through 90 degrees. Six flagella sprouted around the cell's flanks can therefore swing backward into a single tail bundle.
Because an E. coli cell is only two micrometers long, concentration differences across its body length are buried in molecular shot noise, and Brownian motion knocks its heading 90 degrees off course every ten seconds. Douglas Macnab and Daniel Koshland proved in 1972 that bacteria navigate gradients by temporal comparison. They measure whether attractant concentration rises over time.
When all six motors spin counterclockwise, their left-handed helical filaments braid into a single bundle that pushes the bacterium in a straight run for one second. Reversing even one motor to clockwise rotation within a millisecond twists its filament into a right-handed curly state. That polymorphic flip yanks the bundle apart to tumble the cell onto a new heading.
Dennis Bray and Victor Sourjik demonstrated that the nose-mounted array of 4,000 chemoreceptor trimers acts as a cooperative molecular amplifier with a Hill coefficient of 15 to 20. When attractant levels drop, the receptor-coupled kinase CheA phosphorylates the diffusible messenger CheY into CheY-P. Binding of CheY-P to the FliM switch on the C-ring flips rotation clockwise.
How does the cell avoid saturating its chemoreceptors as it climbs across five orders of magnitude in nutrient concentration? The methyltransferase CheR steadily adds methyl groups to four glutamate side chains on the receptor tail while the methylesterase CheB-P removes them. That four-second biochemical memory continually resets the baseline to compare the present second against the past three.