<p>OPEN WIDE:</p>

© 2002 Nature Publishing Group

MscL has one of the widest channel openings. Here transmembrane (TM) segments are in the open state. The side view is shown in relation to a hypothetically distorted bilayer. (Reprinted with permission

Stress – the bane of modern existence. Even cells have to deal with it, in its mechanical forms, at least. Osmotic pressure and shear forces from the environment signal dangerous situations that threaten the integrity of the cell membrane. Membrane channels sense and respond to these signals allowing cells to cope. In complex organisms, specialized cells go beyond mere coping, turning the signals into interpretable sensations such as touch and hearing. In recent years, biologists have discovered a wide array of mechanosensitive channels that mediate responses to physical forces, including members of at least four major protein families.

The challenge now is to separate the wheat from the chaff and find...


Paul Blount, University of Texas Southwestern Medical Center, Dallas, studies mechanosensitive channels in bacteria. In simple organisms, the channels' mission is simple: Ensure survival by responding to osmotic stress. Escherichia coli contains three such channels, MscL for large conductance, MscS for small conductance, and MscK, the potassium-gated channels.1 Crystallization of MscS and MscL reveals a simple monomeric structure containing just two membrane-spanning helices. The MscL channel is built from five monomers arranged in a circle, while MscS contains six.

The MscL channel possesses one of the widest openings of any known channel: a whopping 25 Å, large enough to accommodate small proteins. According to Blount, MscL acts as an "emergency relief valve." When cells experience severe osmotic shock, the channel opens wide, rapidly dumping cell contents to relieve membrane tension and prevent cell bursting. Researchers are rapidly closing in on the conformational changes that allow MscL to open. Eduardo Perozo, at the University of Virginia, and coworkers have shown that the transmembrane helices rotate, move away from the center, and tilt substantially upon opening.2

But understanding these conformational changes does not explain how changes in membrane tension trigger channel openings. Blount describes two possible models. First, the channels might "directly sense physical perturbations in the membrane through lateral pressures," perhaps through physical interactions with the lipids themselves. Alternatively, there is the "trapdoor" mechanism, in which something other than the membrane (e.g., a cytoskeletal element) physically pulls on the channel to open the pore. Purified MscL still opens in response to membrane tension after cell-free reconstitution in artificial lipid membranes, supporting the first model for this protein. "I think what's going on is that there's an energy that must be overcome, so the channel gates before you tear the membrane," says Blount. "But, the mechanisms may be different among the different types," he adds.


In more complex organisms, the functions of mechanosensitive channels go far beyond simple osmotic stress responses. In hearing and touch, mechanical stresses result in mechanotransduction: the conversion of mechanical energy, such as membrane tension, into a chemical signal, ultimately leading to neuronal firing. At least three eukaryotic channel families contain members that respond to mechanical stress: the transient receptor potential (TRP) channels, the two-pore (2P) domain potassium channels, and the degenerin channels.

The TRP family would seem the most likely candidates as primary sensory mechanotransducers. This large, diverse family contains members implicated in almost every form of sensation, including temperature, touch, vision, and taste. In particular, the hunt is on for a TRP channel that acts as the primary mechanotransducer in mammalian hearing, according to David Corey of Harvard Medical School. In the ear, hair cells convert sound waves into mechanical motion via the opening of this channel. In Drosophila, TRP channels known as nompC and nanchung possess the characteristics required of the putative hair-cell channel, and mutants for nanchung are completely deaf. But extensive searches have turned up only one vertebrate homolog for these channels, a nompC expressed primarily in the lateral line organ of zebra-fish, casting doubts on its role in hearing.3

"There has to be another channel in mammals," says Corey. He says he favors another TRP channel, mucolipin 3, which is highly expressed in hair cells. Genetic screening of deafness syndromes in mice has turned up several mucolipin 3 mutants. But, these mice also have abnormal inner ear development, making it impossible to determine whether mucolipin 3 is the primary mechanotransducer, or simply an important supporting player in hair-cell development.


TRP channels remain the most obvious candidates, though there is another ubiquitous mechanically gated family for which there is "no strong information on function," says Lazdunski, who studies the 2P domain potassium channels. These channels are an interesting byproduct of the genomic revolution. Previously, new channels were identified by their cellular activities, and molecular identities were revealed later by cloning. The 2P domain channels were first identified in silico by their homology to voltage-gated potassium channels (which contain only one P domain, and since their cloning, the major effort has been to discover their functions.

<p>TRP-ED UP:</p>

© 2003 Nature Publishing Group

TRP channels are used in organisms ranging from bacteria to higher plants and animals. The cytoplasmic ends of the S6 helices form the lower gate which opens and closes to regulate cation entry. The selectivity filter may also gate. The S1-S4 domain may flex relative to S5-S6 in response to stimuli (S5 not shown), but the paucity of positively charged arginines in TRP S4 helices indicates weak voltage sensitivity. The figure emphasizes the diversity of TRP cytoplasmic domains. (Reprinted with permission, D.E. Clapham, Nature, 426:517–24, 2003.)

Three of the approximately 14 2P domain channels discovered thus far, TREK-1, TREK-2, and TRAAK, are mechanically gated.4 According to Lazdunski, all three are expressed abundantly in brain, but are also found in many other tissues. TREK-1 in particular is "all over the place," he says, including in kidney and small intestine, where, he hypothesizes, it could play a role in osmosensation.

Lazdunski's group has studied the effects of membrane-altering lipids on 2P domain channel opening and concluded that these channels are gated by the direct sensing of membrane tension. For example, arachidonic acid (AA), which intercalates preferentially into the outer leaflet of the lipid bilayer, makes the membrane more convex, and this activates 2P domain channel opening. Chlorpromazine, which produces membrane concavity or "cupping," reverses AA-mediated channel opening. But Lazdunski does not discount cytoskeletal influences as well. He maintains that 2P domain channel gating is "regulated by the cytoskeleton," since the pressure dependence of these channels changes substantially when they are detached from it.

The physiological roles of the mechanically gated 2P domain channels are still decidedly murky, raising an important caveat. While mechanical gating of these and other channels can be demonstrated clearly in vitro, few have been extensively studied in vivo. Miriam Goodman at Stanford University, who studies mechanically gated channels in Caenorhabditis elegans, says that channels may be "stretch-sensitive" but not necessarily "mechanosensitive" in their cellular context. Blount concurs, noting that we may be seeing "what they can do, not necessarily what they are doing."


The overabundance of mechanosensitive channels in complex organisms might lead one to hope for a simpler situation in C. elegans. But according to Goodman, the nematode contains more than 100 TRP and 2P domain channels, not to mention members of yet another family, the degenerins. These mechanically gated channels have homologs in both flies and vertebrates, including the well known mammalian epithelial sodium channel (ENaC) family, leading to speculation that undiscovered isoforms could play major mechanotransduction roles in higher organisms as well.5

Degenerins are so named because their mutation leads to the degeneration of specific neuronal cells. The cells swell and lyse, Goodman says, suggesting a role in osmosensation. Mutations in other worm degenerins cause a loss of touch sensitivity. Mutations in genes encoding specialized microtubules and extracellular matrix proteins also affect touch sensitivity, which leads Goodman to favor the trapdoor gating mechanism. Yet she notes, " [The] two mechanisms need not be mutually exclusive."

It is clear that researchers interested in mechanosensation will have plenty to do for some time to come, and it's likely that many channels have yet to be characterized, according to Blount. "Mechanosensitive channels are almost everywhere you look for them."

Megan M. Stephan megan.stephan@cox.net is a freelance writer in Cheshire, Conn.

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?