Flagella are rotating protein filaments, roughly 20 nanometers wide, that propel bacteria, archaea, and eukaryotic sperm cells through liquid. A bacterial flagellum assembles itself from about 30 protein types, spins at up to 1,700 revolutions per second, and switches direction in milliseconds. This explainer covers how the motor works, what it costs the cell, how chemotaxis steers it, and where the standard textbook picture breaks down.
The rotary motor and its stator rings
A bacterial flagellum has three functional modules. The first is a reversible rotary motor embedded in the cell envelope. The second is a curved universal joint called the hook. The third is a long helical filament that acts as a propeller.
The motor itself is a ring of membrane proteins. In Escherichia coli and Salmonella, the stator units MotA and MotB form channels that conduct protons. Each stator complex sits around the rotor and generates torque by a power-stroke mechanism coupled to proton flow. The rotor consists of the MS ring, the switch complex, and the C ring. The C ring is the steering wheel: it binds phosphorylated CheY, the response regulator of the chemotaxis system, and that binding flips the motor's rotation direction.
The energy source differs by domain. Bacteria and archaea use ion-motive force across the membrane, usually protons in bacteria and often sodium ions in marine species. Eukaryotic flagella do not rotate at all. They bend, driven by dynein motors sliding microtubules past each other inside a 9+2 axoneme. So the word flagellum describes two unrelated machines that converged on the same name.
Assembly is ordered and extracellular. Flagellin subunits travel through a narrow channel inside the growing filament and polymerize at the tip, not the base. A filament can grow several micrometers per minute. This tip growth means the newest material is at the far end and the oldest is at the base.
Quantified context: speed, efficiency, and cost
Rotation rates range from about 100 turns per second in some marine bacteria to roughly 1,700 in E. coli. Filament length runs 5 to 20 micrometers, several times the cell body. Swimming speed is typically 20 to 50 micrometers per second, or about 30 body lengths per second.
Efficiency is modest but competitive. A bacterial flagellum converts ion flow to mechanical work with roughly 1 percent efficiency by some estimates, yet it outperforms diffusion by orders of magnitude for finding nutrients. The hydrodynamic cost is real: a cell turning thousands of flagella expends a large fraction of its energy budget on motility.
| Feature | Bacterial flagellum | Archaeal flagellum | Eukaryotic flagellum |
|---|---|---|---|
| Motor type | Rotary, ion-driven | Rotary, ATP-driven | Bending, dynein-driven |
| Filament protein | Flagellin | Archaeal flagellin | Tubulin |
| Typical diameter | About 20 nm | About 10 nm | About 200 nm |
| Rotation or bending | Rotation | Rotation | Bending |
| Energy source | Proton or sodium gradient | ATP | ATP |
| Steering | Chemotaxis via Che proteins | Limited data | Calcium and cAMP signaling |
| Assembly | Tip growth, extracellular | Tip growth | Intracellular |
The table matters because textbooks often treat all three as one structure. They are not homologous in any simple sense.
Chemotaxis: how a cell steers
Chemotaxis is a temporal comparison, not a spatial map. A bacterium cannot sense which direction a nutrient lies in. Instead, it measures whether the concentration is rising or falling over the past second or so, then biases its next swim.
When a cell swims up a gradient, the chemotaxis system suppresses tumbling. The result is longer runs and fewer turns. The motor keeps turning counterclockwise, the flagellar bundle stays coherent, and the cell moves forward. When the gradient weakens, phosphorylated CheY rises, binds the C ring, and the motor briefly reverses or pauses. The bundle flies apart, the cell tumbles, and it picks a new heading.
The temporal window is short. Adaptation takes seconds, mediated by methylation of the receptor complex. A cell effectively compares current concentration with a memory of roughly the last few seconds. This lets it climb gradients spanning several orders of magnitude without saturating.
The counterintuitive detail: a bacterium cannot swim straight. It runs and tumbles in a biased random walk, and the bias is the entire navigation strategy. There is no compass and no map. Steering emerges from statistics.
Edge cases and when the standard model fails
Several organisms break the textbook rules.
Some bacteria, including Spirochetes, keep their flagella inside the periplasm and rotate the whole cell body instead of an external filament. Vibrio cholerae and other species carry multiple flagella with distinct roles: a single polar sheathed flagellum for swimming and lateral flagella for surface movement. Magnetotactic bacteria align their swimming using magnetosomes, adding a geomagnetic sense on top of chemotaxis. Some pathogens use a type III secretion system that evolved from the flagellar export apparatus, blurring the line between motility and virulence machinery. Eukaryotic sperm flagella bend rather than rotate, and defects in dynein arms cause primary ciliary dyskinesia, a human disease.
The professional caveat: flagellar gene expression is tightly regulated and often switched off when motility is costly or unnecessary. A laboratory strain grown in rich broth may show few flagella even though the species is motile. Absence of flagella in a microscopy image is not evidence that the organism cannot make them.
Misconceptions worth correcting
The most common error is treating the flagellum as a simple propeller. It is a self-assembling nanomachine with its own export pathway, a universal joint, and a steering system. Removing any of about 30 genes can abolish function.
A second misconception is that flagella always increase fitness. Motility costs energy. In a nutrient-rich, well-mixed environment, nonmotile mutants often outgrow motile ones because they skip the cost. Flagella are an investment that pays off when gradients exist.
A third error is conflating bacterial and eukaryotic flagella. They differ in protein composition, diameter, energy source, and mechanism. Calling them the same structure is like calling a propeller and a rowing oar the same invention.
A fourth is assuming faster rotation always means faster swimming. Bundle formation matters more. A cell with many flagella that coordinate into a single bundle swims efficiently. A cell whose flagella fail to bundle wastes energy churning fluid.
Practical actions: measuring and interpreting motility
Confirm motility with a soft-agar swarm assay rather than a wet mount alone; swarming indicates functional flagella and chemotaxis, while Brownian motion does not. For microscopy, use phase contrast at 400x or higher; flagella are about 20 nm and below the resolution limit, so you see bundles, not individual filaments, unless you stain or use fluorescence. If you need to count flagella, stain with a flagellar stain or use fluorescent labeling; simple light microscopy will not resolve single filaments. In environmental samples, do not conclude a species is nonmotile from one preparation; flagellar expression depends on growth phase, temperature, and nutrient status. When testing antimicrobials that target motility, verify stasis versus killing separately, because paralyzed flagella can resume rotation after the agent washes out.
Warning flags: a swarm plate with no swarm does not prove nonmotility without a positive control. A "motile" result on a wet mount can be fluid flow, not swimming. Do not extrapolate a lab strain's behavior to a clinical isolate without checking expression conditions.
Limitations of what the evidence shows
Many core mechanisms are well established: the rotary motor, proton coupling, the chemotaxis signaling cascade, and tip assembly. Others remain active research. The precise torque-generation stroke of the stator, the energetics of motor switching, and the coordination of multiple flagella into a bundle are still debated at the quantitative level.
Numbers like "1,700 revolutions per second" come from specific strains under specific conditions and should not be generalized as species constants. Efficiency estimates vary widely by method and definition. Comparative genomics suggests flagellar systems have been lost and reacquired many times, so any single model organism is a partial guide.
Finally, the health-relevant claims are limited. Motility contributes to colonization and biofilm formation in some pathogens, but flagella are not universally required for virulence, and some pathogens shut them down during infection. The mechanistic science is deep; the clinical generalizations are narrower than headlines suggest.
The bottom line on flagella
Flagella are among the most efficient rotary machines known in biology, built from about 30 proteins and powered by ion gradients in bacteria and archaea or by dynein in eukaryotes. They range from 100 to 1,700 revolutions per second, cost real energy, and steer through a biased random walk rather than a map. The standard textbook picture gets the core right but misses the exceptions: periplasmic spirochete flagella, sheathed polar filaments, and the entirely separate eukaryotic axoneme. Treat motility as a regulated, conditional investment rather than a fixed trait, and check expression conditions before drawing conclusions from a single image.