Cellular physiology of cultured C. elegans mechanosensory neurons

In C. elegans the response to the gentle touch delivered by an eyelash is mediated by six specialized sensory neurons, the mechanosensory neurons. Analyses of mec mutants (mechanosensory abnormal) identified many genes that are important for the normal function of these neurons including those encoding the DEG/ENaC ion channel subunits mec-4 and mec-10, mec-7 and mec-12 that encode for b and a tubulins respectively, the collagen gene mec-5, the EGF-rich mec-9 and the stomatin-like gene mec-2. MEC-4 and MEC-10 are coexpressed exclusively in the six touch neurons and form a heteromeric channel postulated to constitute the core of a mechanosensory ion channel. Localized tension is postulated to be administered by tethering the extracellular MEC-4/MEC-10 channel domains to extracellular matrix proteins such as MEC-5 and MEC-9 and anchoring intracellular channel domains to a 15-protofilament microtubule network exclusively assembled in the touch receptor neurons by MEC-7 and MEC-12.

C. elegans has provided an extremely powerful model system for the identification of the molecular determinants of the mechanotransduction complex, but has had limitation for the direct electrophysiological characterization of the postulated mechanically gated ion channel complex. In addition, a recent report (Goodman et al., Nature 2002 415, 1039 - 1042) has shown that while functional MEC-4/MEC-10 channels can be reconstituted in Xenopus oocytes, in this expression system they are not mechanically gated possibly due to the lack of some important interacting proteins. Recently a method has been developed for the isolation and culture of C. elegans embryonic cells (Christensen et al., Neuron 2002 33: 503-514) that allows the application of standard electrophysiological techniques to isolated touch neurons.

Of the six touch neurons, four (ALML/R and PLML/R) develop during embryogenesis and are likely to be present among isolated embryonic cells.  We cultured embryonic cells isolated from Pmec-4::GFP-expressing worms. Within 24 h after isolation, GFP expression was detected in ~2.0% of cells in culture.  All cells expressing GFP developed only one neuronal process that increased in length with further days of culture. Cultured touch neurons expressed, as in vivo, acetylated a tubulin and underwent necrosis when expressing the deadly mec-4 mutant subunit harboring the A713V substitution (mec-4(d)). Cell death was rescued in cultured touch neurons by maneuvers known to rescue mec-4(d) induced cell-death in vivo, namely addition of amiloride or dantrolene to the culture media or the deletion of calreticulin (Xu K. et al., Neuron 2001 31:957-71). These results suggest that touch neurons in culture differentiate and behave similarly to what is observed in vivo.

In the attempt to identify and study the mechanosensitive ion channel complex we patched clamped in the cell-attached and inside-out modes cultured touch neurons.  We identified a strongly outwardly rectifying potassium channel that displays a conductance of ~13 pS and a stretch-activated sodium channel that functions independently of mec-4. We are presently carrying experiments to identify the molecular determinants of the stretch-activated channel and to determine the localization of MEC-4 channels within the touch neurons using specific antibodies.

We thank Dr. K. Strange for sharing with us preliminary observation on cultured touch neurons.