Essays on Science and SocietyEPPENDORF ESSAY WINNER

Deconstructing C. elegans Sensory Mechanotransduction

See allHide authors and affiliations

Science  15 Oct 2004:
Vol. 306, Issue 5695, pp. 427-428
DOI: 10.1126/science.1105624

We as human beings depend greatly on the sensory neurons that govern our sense of touch. If such cells cease to function properly, we may lose the ability to respond to another's touch—say, as a dancer responds to a partner's lead. Additionally, we may be unable to respond to a more painful and potentially damaging event. Reduced touch sensation, or peripheral sensory neuropathy, is especially common in people with diabetes and is a significant contributing factor to lower-extremity amputations (1). Other mechanoreceptor neurons mediate equally vital sensory modalities: proprioception for balance and the control of internal organs, such as the bladder and kidney, and baroreception for homeostatic control of heart rate. Despite the importance of these mechanical senses, however, exactly how sensory cells detect the mechanical energy in a touch, the bend of a limb, or changes in blood vessel diameter remains a mystery.

Although electrical responses to mechanical stimuli were first measured in the 1920s (2, 3), surprisingly little is known about the protein machinery that converts mechanical energy into ionic currents in touch-sensitive neurons, and even less is known about how the individual protein components of this nanomachine operate. Research into the molecular basis of touch transduction lags behind research into other senses because, in many animals, sensory nerves that detect touch are scattered across the body and are deeply embedded in the skin—two properties that complicate traditional biochemical approaches. To circumvent these difficulties, we study touch sensation in the nematode worm Caenorhabditis elegans. This is the only animal for which we know the cellular anatomy of the entire nervous system. Compared with tens of thousands cutaneous sensory neurons in mammals, each worm has only six nerve cells that govern touch sensation along its body wall (4). Genetic analyses by Martin Chalfie and his colleagues revealed that the worm's sense of touch requires at least 12 specific proteins, encoded by the mec or mechanosensory abnormal genes [reviewed in (5)].

To understand how proteins identified by genetic screens contribute to mechanotransduction, my collaborators and I developed methods for in vivo recording from identified sensory neurons in C. elegans (6) and used these methods to record electrical responses to external force in C. elegans touch neurons (7). The initial experiments focused on a pair of neurons that sense gentle touch to the worm's tail. We found that receptor potentials in C. elegans touch neurons are reminiscent of the responses of vibration-sensitive Pacinian corpuscles measured 40 years ago in mammals (8), which suggests that aspects of mechanotransduction may be similar in nematodes and mammals (see the figure). Activation of mechanoreceptor currents (MRCs) in C. elegans touch neurons is extremely rapid: Current begins to flow within 1 ms of force application. It is the first step in transduction, preceding both membrane depolarization (7) and increases in somatic Ca2+ (9). Such latencies are nearly two orders of magnitude faster than those reported for Drosophila phototransduction (20 to 100 ms), the current record-holder for a second messenger-mediated G protein signaling cascade (10). Thus, external force might open ion channels directly rather than operating via a separate force receptor.

Responding to touch.

Comparision of mechanoreptor potentials in C. elegans touch receptor neurons (top) and mammalian Pacinian corpuscles (bottom). [Data in the bottom panel are adapted from (8)]

We are also working to deconstruct the mechanotransduction complex by asking how mutations in mec genes alter MRCs in vivo. The first mutations we studied eliminate or alter four membrane proteins (MEC-4, MEC-10, MEC-2, and MEC-6), which are believed to form the ion channel at the core of the transduction complex. Consistent with this idea, null mutations in mec-4, mec-2, and mec-6 abolish MRCs without affecting other ion currents, which indicates that these proteins are specifically required for the generation of MRCs and are likely to encode subunits of the ion channels that carry MRCs in vivo. Additionally, an allele of mec-10, which substitutes glutamate for a conserved glycine residue near the second transmembrane domain of MEC-10 (11), reduced but did not eliminate MRCs. This reduction appears to result from altered ion selectivity, as opposed to a genetic deletion of transduction channels. In short, all mutations that diminish touch sensation abolish or alter MRCs in vivo. Our findings link the application of external force to the activation of a molecularly defined sensory transduction channel.

In addition to analyzing electrical responses to external force in vivo, we have taken the first steps toward reconstituting this ion channel complex in Xenopus oocytes. So far, we have bypassed the need for mechanical gating by studying a constitutively active mutant channel (the “d” form). Coexpression of MEC-4d and MEC-10d produces a constitutively active current that, like native MRCs (7), is carried by Na+ and blocked by amiloride (12). By contrast with native MRCs, however, neither MEC-2 nor MEC-6 was required to produce detectable channel activity in oocytes (12, 13). Both accessory proteins increased activity of expressed MEC-4d/10d channels at least tenfold without inducing a detectable increase in surface expression of either MEC-4 or MEC-10 (12, 13), which suggests that MEC-2 and MEC-6 increase single-channel conductance or open probability. Preliminary studies of single MEC-4d/10d channels suggest that neither MEC-2 nor MEC-6 significantly increases single-channel conductance, however (14). Additional studies of both expressed and native channels are needed to clarify the function of these accessory proteins in mechanotransduction.

By demonstrating that native MRCs require intact copies of mec-4, mec-10, mec-2, and mec-6, we show that each of these four genes is required for the first step in mechanotransduction—namely, activation of sensory mechanotransduction channels. Such channels may be directly gated by mechanical energy, because MRCs can be detected within 1 ms of stimulation. Because both C. elegans touch receptor neurons and mammalian Pacinian corpuscles respond preferentially to changes in force (7, 8), we speculate that DEG/ENaC channels could be sensory mechanotransduction channels in nonciliated mechanoreceptor neurons in nematodes and mammals alike. These initial studies raise new questions, such as: How do touch receptor neurons detect changes in force while remaining insensitive to continuous force application? How is force transferred from the worm's cuticle to transduction channels? What determines sensitivity? A better understanding of the answer to this last question could lead to improved diagnosis and treatment of sensory neuropathy.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.

Navigate This Article