Power source for the motor

The movement of protons across the M-ring/Mot complex from the periplasm into the cytoplasm inside the
cell, provides the energy required to drive the rotation of the flagellum. In some bacteria that thrive in very
salty conditions (halophiles or 'salt-lovers') such as bacteria in salt lakes, may use sodium ions (Na+)
instead of protons, but the end result is the same since both protons and sodium ions carry a single unit of
positive electric charge. However, ions will only undergo net diffusion down an electrochemical gradient. A
chemical gradient occurs when the concentration of a chemical is higher in one region that in another
and when the regions are joined the ions will tend to flow from the region of high concentration to the
region of low concentration (i.e. down the chemical gradient).

A note concerning diffusion. This is a net movement that results from diffusion. Diffusion is the random
movement of molecules, atoms or ions (or other particles) which occurs as a result of thermal energy,
which is really the kinetic energy of movement of the particles. Collisions between particles ensure that the
particles change direction frequently, giving them a random motion overall - that is they are equally likely
to be moving in any direction. If a gas of particles is injected into a container (for example hydrogen
sulphide molecules entering a room when its glass phial container is smashed) then this random motion
will, over time, cause the particles to spread out and fill the room - the smell of the hydrogen sulphide
(rotten eggs) has drifted or diffused throughout the room. When two regions of differing concentration are
brought into contact (such as across a porous membrane) then on average more molecules will diffuse
from the region of high concentration into the region of low concentration and so the concentration will
average out over both regions and then the movement of particles from one region into the other will, on
average, equal the movement in the opposite direction.

Ions are subject to another 'force', other than that due to chemical concentration, that causes their net
movement and this is the
electrical gradient. Since protons carry positive electric charge and like electric
charges repel whilst opposite electric charges attract, protons will move from a region of greater positive
electric charge (which repels them) into a region of relative negative electric charge (which attracts them).
Note that, as for chemical concentration, it is the relative differences or gradient that determines the
driving force. Thus charged particles are subject to both a chemical gradient and an electrical gradient,
what we call an
electrochemical gradient. If these gradients are equal and opposite then no net
movement results, but if both push in the same direction then more rapid movement results.

Thus, to achieve a constant flow of protons from the periplasm into the cell cytoplasm the bacteria must
maintain an electrochemical gradient across the cell membrane. Many bacteria establish both a favourable
chemical gradient (a higher concentration of protons in the periplasm than in the cytoplasm) and a
favourable electrical gradient (the periplasm is more positively charged than the cytoplasm). (Though
some appear to utilise only the chemical concentration gradient). Several factors contribute to these
gradients, but the most important is the active pumping of protons from the cytoplasm into the periplasm.
As the protons flow back in they are pumped back out and recycled. Now, pumping protons against an
electrochemical gradient, that is into the periplasm from the cytoplasm, requires energy (it is like pushing
an object uphill against its natural tendency to roll back down) . This energy ultimately comes from the food
nutrients that the bacteria oxidise (burn) as fuel. You don't get something for nothing! To swim a bacteria
needs energy from an external source. For example, sugars taken into the cell are oxidised and some of
the energy released (about 40%) is converted into chemical energy (stored in molecules called ATP) and
the rest is lost as heat and sound (a 'wastage' that is inevitable in any machine, though in living systems
some of this heat is again tapped as chemical energy as it makes molecules more reactive by making them
move faster). So, the proton pumps use up ATP and the energy released from it to pump protons from the
cytoplasm into the periplasm. This will be covered in more detail in a forth-coming link. The protons then
flow back into the cytoplasm, when allowed to do so, such as when proton-conducting channels open up in
the Mot proteins in the flagella motor - causing electricity to flow through the motor and causing it to rotate.

Note how living systems, like all machines, convert energy from one source to another and use it to do
work. They do not make their own energy, merely convert it, just as a power station does. In this case
chemical energy is converted into an electrochemical gradient and then into the force of flowing protons
(electric current) and then into rotatory mechanical energy which propels the cell. The force produced by
the protons flowing across the cytoplasmic membrane (containing the Mot and M-ring) is called the
motive force (PMF)

How does the PMF cause rotation of the rotor - the mechanism of flagella rotation?

Several models have been proposed over the years, but of these only one model type remains popular,
but the exact details have not yet been elucidated. However, it is important to look at other models too,
especially since they may be of importance in other systems and the actual flagella motor may use a
combination of more than one of these models.

Molecular cross-bridges

Berg and Brown (1972) proposed a model in which the periphery of the M-ring was linked to the cell
envelope by up to five cross-bridges, similar to those in skeletal muscle. These bridges would detach and
reattach at a different site, bringing about rotation. More specifically, Lauger (1988) proposed a similar
mechanism in which protons bring about conformational changes/movements in bridges between the M
and S rings, as shown below:
Models of flagella rotation - how does the rotor work?
This model is applicable to the linear motion (as opposed to rotary motion) of many cells and cell
components in eukaryotic (non-bacterial) cells, including animal cells such as muscle cells. Lauger
adapted these models to work in a rotary system.

Cytomembrane streaming

Adam (1977) imagined that the M-ring subunits resembled the blades of a paddle wheel. These blades
are angled such that lipids flowing towards the M-ring slide past one side of the ring while becoming
caught by paddles on the other (see figure below). Lipid streaming can be caused by addition of lipids to
membrane on one side and removal of lipids from the other side. A change in the direction of rotation
could occur by conformational changes in the subunit blades or by a change in direction of membrane
streaming. Membrane streaming has not been directly demonstrated in bacteria, but may be involved in
gliding motility. Note that the rotor has been divided into 16 subunits, which is in accord with
observations which show that the motor (M-ring) is indeed divided into about 16 sectors (and the S-ring
appears to have 17 components).
When fluid flows past a solid object, the fluid near to the object is dragged back and slowed, forming a
boundary layer. In real fluids, there is also a microscopically thin layer of fluid that is stationary and
bonded to the solid surface with layers of fluid further out moving progressively faster as they slide past
one another (forming the boundary layer which is thicker for rougher surfaces), until the bulk velocity is
reached far from the boundary. This layer of stationary fluid produces a no-slip boundary condition.
However, sometimes it is useful to model a fluid as exerting no friction and sliding past solid surfaces
unrestrained. What we have in this model is such an ideal perfect slip boundary condition on one side of
the paddle-wheel (which is a good approximation) and a no-slip boundary condition on the other side
(meaning that the fluid is held back more than is usual for a smooth surface due to the projecting
paddles). These approximations make such systems easier to model mathematically, but should not be
taken too literally.

Osmoelectric / electrokinetic motors

This type of model has been described by Glagolev and Skulachev (1978) and Lauger (1988), but the
one detailed here was put forward by Mitchel (1984). The figure below shows a theoretical turnstile
molecular motor. Molecules Sp (electrically positively charged stator subunit) and Sn (negatively
charged stator subunit) represent the stator (S-ring) on either side of the rotor, R (rod). Protons (from
outside the cell) bind a negatively charged site on Sp, making it positively charged. This attracts a
vacant negatively charged proton-binding site on R, which rotates to pick-up the proton from Sp. The R
site becomes positively charged and, as another proton occupies the Sp site, is attracted to the vacant
negatively charged proton binding site on Sn. R rotates again, giving its proton to Sn which passes it to
the cytosol and the whole cycle repeats. However, in such a system, the motor would be equally likely to
rotate either way if it wasn't for asymmetry introduced into the system which biases the direction of
rotation. This asymmetry could result from less electrical resistance to proton flow in one direction or by
a difference between the amount of the negative charge (n) on Sn and the positive charge (p) on Sp.

An array of such stators could occur in a ring (S-ring) above a rotary ring with multiple proton binding
sites (M-ring) (fig.10c). The M-ring would be connected to the rod, causing it to rotate (note friction is
negligible due to the small size of the components). Signals could bring about a change in rotation by
reversing proton flow through the S-ring subunits, so that Sp becomes Sn and Sn becomes Sp.
Note that more than one proton may be transferred to each Sp and removed from each Sn during each
stage of rotation, indeed it has been estimated that 1040 protons are required for each revolution of the

Dielectric motors

Small spherical or cylindrical particles rotate when immersed in liquids and subjected to electromagnetic
fields. When a potential difference occurs across such a particle the surfaces become charged and
mutual repulsion/attraction causes rotation, as shown below:
Current flows (as opposite charges attract), resetting the charges as shown below:
Thus, the particle continues to rotate. A potential difference would be needed within the membrane,
either side of the basal complex, not across it and so is not simply the result of a membrane proton
potential. Any asymmetry in membrane structure could create a charge difference across the motor,
such as protons flowing through a proton channel on one side of the rod, as shown below:
Such channels could be spaced at intervals in a ring around the rod, such as the M-ring (see the figure
below). In this case, the M-ring might not connect to the rod and could remain stationary. Any of the
rings could contribute in generating torque.
As yet, there is no evidence (that I am aware of, but I am still researching this topic) as to which model is
most likely, though the osmoelectric-type seems most popular. Mathematical treatment enables most of
these models to fit observed kinetics of the bacterial motor.

Role of the Mot and Switch Complexes

The proteins MotA and MotB are essential for energy transduction during flagella rotation. Mot mutants are
unable to rotate their flagella, but otherwise appear normal. Rings of cyto- membrane particles, large
enough to accommodate M-rings, have been observed in some bacteria, but are absent in Mot mutants
et al., 1988). Mot proteins probably form part of the basal complex. MotA and MotB together
form a proton channel
(MotA contributes most to the channel). Various receptors sense environmental
signalling and intracellular signals operate a "switch" which controls the direction of flagella rotation. MotA
and MotB are thought to make up the stator, anchored in the inner membrane and periplasm and
enclosing the M ring.

A "switch complex" may be part of the basal complex (and is thought to be made from FliM, FliN and FliG
proteins). This switch reverses the rotation between clockwise (CW) and counterclockwise (CCW) rotation.
These components are common to all bacterial flagella and are believed to be major components of the
motor. Freeze-fracture electron micrographs show 16 stud-like particles comprising the M-ring and 17
comprising the S-ring. Block and Berg (1984) used Mot mutants of
Escherichia coli. Plasmids bearing the
Mot genes under the control of a lac promoter were introduced into the mutants. Addition of a lac inducer
caused expression of these Mot genes and flagella rotation was slowly restored. As the flagella "warmed-
up" they increased speed in 16 discrete steps. This suggests that each M-ring subunit contributes in the
generation of torque (rotary force). The M-ring is, therefore, considered to be the flagella motor. In many
models the M-ring generates torque by its interaction with the stationary S-ring (acting as a stator). As the
M-ring rotates so it exerts force on the S-ring, which is attached to the cell, and the whole cell spins in the
opposite direction to the flagellum. It is assumed that the M-ring is connected to the rod and causes it and
the filament to rotate. The filament generates useful thrust (mutants with straight filaments rotate but
remain stationary).

MotA-MotB complexes probably form the proton channels and torque generators of the motor. These are
integral membrane proteins. FliG, FliM and FliN are cytosolic proteins required for energisation, swithing
and assembly of the flagellum (Schuster, 1994). Evidence suggests that these proteins form a structural
complex with FliF and FliG in a 1:1 stoichiometry and FliM and FliN also in a 1:1 stoichiometery. (The
stoichiometry is the chemical proprtion, i.e. the relative numbers of molecules of different chemicals bound
together in a complex, so a 1:1 stoichiometry means that the two molecules are present in equal numbers).

Update (Aug 2018) - a Refined Model

A revised model has been developed in part by Paul et al. (2010),  Paul and Blair (2006) and Dyer et al.
. In this model, FliN, FliM and FliG make-up the C-ring in the cytoplasm (this is supported to some
extent by mapping the proteins to the electron density of the flagella apparatus as seen in cryo-electron
microscopy). FliG has charged amino acid residues (near its C-terminus) which interact with proteins
entering the Mot stator complex to generate torque.
FliG is considered part of the rotor itself (along with
FliM and FliN). It is known that FliG contains charged residues involved in torque generation, but the
precise mechanism remains unknown
. Protons flow through MotA/B and are thought to bind to Mot,
changing its conformation and possibly causing it to push electrostatically on FliG
(Armitage and Berry,
2010). There are about 26 FliG molecules in each C-ring. FliN is hypothesised to act as scaffolding for
FliM and FliG.

Phosphorylated CheY, the signal to switch flagellar rotation sense in
Escherichia coli (from the CCW
default sense to CW to cause a tumble and change in orientation) binds FliM to change the interaction of
FliG with Mot (Dyer
et al. 2009).

Furthermore, the
c-di-GMP binding protein YcgR is thought to act as a brake upon binding to the
intracellular signalling molecule c-di-GMP (a cyclic molecules formed from two GMP or guanosine
monophopshate molecules) which has numerous signalling functions in prokaryotes. In their model, the
activated YcgR then binds to FliM, part of the switch complex, and to FliG to alter the conformation of FliG
and move its charged residues away from MotA to act as a brake on flagellar rotation. One would have to
consider screening of the charges and what sort of conformational change or movement would be
sufficient to lower the FliG-MotA charge interactions significantly.

Clockwise (CW) or Counterclockwise (CCW) rotation?

Different bacterial species may use flagella in different ways. However, the first example we shall consider
is the bacterium
Escherichia coli which inhabits the human large intestine. This bacteria has six flagella
scattered over its surface (what we call a
peritrichous arrangement). When most of these flagella are
rotating CCW the filaments come together to form a bundle at one end of the cell (which becomes the 'tail'
end) and the bacterium swims in a smooth straight line. However, periodically, the flagella switch to CW
rotation, which causes the bundle to fly apart into single flagella radiating in different directions, causing
the cell to tumble on the spot and randomly change its direction. CCW rotation is resumed and the cell
again swims in a straight line. These responses are a crucial part of the behaviour of these bacteria in
locating nutrient sources by a process called

Download an illustrated essay on bacterial motility and navigation in pdf format: Prokaryotes_motility.

See a worked solution to the
proton turbine model

Click here to learn about chemokinesis...
See a worked solution to the proton turbine model

Armitage, J. P. and R. M. Berry, 2010. Time for Bacteria to Slow down.
Cell 141: 24-26.

Dyer, C. M., A. S. Vartanian, H. Zhou, and F. W. Dahlquist, 2009. A Molecular Mechanism of Bacterial
Flagellar Motor Switching.
J Mol Biol. 2009 388: 71–84.

Paul, K., V. Nieto, W. C. Carlquist, D. F. Blair and R. M. Harshey, 2010. The c-di-GMP binding protein YcgR
controls flagellar motor direction and speed to affect chemotaxis by a "backstop brake" mechanism.
Molecular cell 38: 128-139.

Paul, K and D.F. Blair, 2006. Organization of FliN Subunits in the Flagellar Motor of Escherichia coli. J.
Bacteriology Apr. 2502–2511.

... more references still to be added ...