We previously showed that
avr-15 mutant worms lack neurotransmission by the M3 pharyngeal motor neuron. Furthermore, we were able to demonstrate that whereas depolarized pharyngeal muscles from wild type animals respond to the iontophoretic application of glutamate by hyperpolarizing, pharyngeal muscles from
avr-15 mutants do not (Avery et al., WBG 13(4),72; J. Dent, 1995 International Worm Meeting, 69). These results suggested that
avr-15 is necessary postsynaptically, in pharyngeal muscle, for the expression of a glutamate-gated chloride channel that mediates M3 neurotransmission. One possibility was that
avr-15 codes for a component of the channel itself. Cully et al. (Nature 371,707-711) cloned two subunits of a glutamate-gated chloride channel, GluCla and GluClb, from C. elegans. GluCla encodes a subunit that, as a homomer, forms an ivermectin-gated chloride channel. Given
avr-15's role in ivermectin sensitivity (see Avery et al., WBG 13(4),72) and glutamatergic neurotransmission, one possibility was that
avr-15 coded for GluCla. To test this hypothesis, we probed a YAC grid with GluCla. Although the map position of GluCla was not consistent with it being encoded by
avr-15, we were able to use GluCla to identify a closely related gene whose map position is consistent with that of
avr-15. We isolated a cDNA encoding this related gene and found that the predicted protein was 85% identical to GluCla at the amino acid level and ~45% identical to GluClb. These homologies indicate that the new subunit and GluCla define a subclass of channel subunits and therefore we designated the new subunit GluCla2 and the original GluCla we call GluCla1. To demonstrate that GluCla2 is encoded by
avr-15, we showed that a cosmid containing GluCla2 restores M3 neurotransmission when transformed into
avr-15 worms. We also found that the
ad1051 allele of
avr-15 contains a nonsense mutation early in the open reading frame, which indi cates that
ad1051 is likely to be a null allele. Finally, we were able to restore M3 neurotransmis sion to
avr-15 mutants by transforming them with an
avr-15 cDNA driven by the
myo-2 promoter, which should drive expression specifically in the pharyngeal muscle. Thus, expression of the receptor postsynaptically in pharyngeal muscle is, as we predicted, sufficient to restore M3 neu rotransmission. In addition, we transformed the
myo-2::
avr-15 construct into an ivermectin resis tant
avr-14;
avr-15 double mutant strain and found that the transformed worms became sensitive to ivermectin. Thus, ivermectin's effect on the pharyngeal muscle, mediated by AVR-15/ GluCla2, is sufficient to kill C. elegans.