Because ivermectin is such a widely used nematocidal drug, we were interested in whether and how ivermectin resistance might arise in treated populations. To address this question, we have been studying ivermectin resistance genes in C. elegans. We previously reported that one of the genes that confers resistance to ivermectin,
avr-15, encodes a subunit of a chloride channel that is expressed in the pharyngeal muscle and is gated by glutamate and ivermectin. But recessive loss-of-function mutations in
avr-15 only result in, at most, 2-fold resistance to ivermectin. What other genes could be contributing to ivermectin senstivity? Carl Johnson showed that more substantial resistance to ivermectin could be observed in worms that had mutations in both
avr-14 and
avr-15 even though the
avr-14 single mutant exhibits no resistance. To confirm and extend these results we isolated alleles of
avr-14 by mutating
avr-15 worms and selecting for survival from egg to adulthood in the presence of 10 ng/ml ivermectin. We isolated 3 alleles of
avr-14 all three of which had similar degrees of resistance to ivermectin (~20 fold). We mapped these alleles and found they mapped near a candidate gene,
gbr-2. Laughten et al., (Gene 201:119, 1997), in a search for cDNAs encoding chloride channel subunits, isolated two cDNAs that correspond to the alternatively-spliced gene products of
gbr-2 (transcripts "A" and "B"). To see whether
gbr-2=
avr-14, we rescued the ivermectin sensitivity of an
avr-14 ;
avr-15 double with a cosmid containing
gbr-2. We also sequenced our alleles of
avr-14 and found that all three had mutations in
gbr-2. Two of the strains have a V->E mutation in a residue of the extracellular amino terminal domain that is encoded by an exon common to both transcripts. It is not known why this mutated valine is important but it is conserved in all ligand-gated chloride channels. The other strain has a missense mutation that truncates the protein prior to the last transmembrane domain and is therefore likely null. However, it only affects the transcript A. To understand how
avr-14 makes worms sensitive to ivermectin: 1) We expressed both of the
avr-14 transcripts in Xenopus oocytes. Transcript B encoded a channel that is opened by both ivermectin and glutamate. Oocytes expressing the transcript A did not respond to any drug we tried. It is interesting that transcript A is affected by the missense mutation. 2) We made a GFP fusion that included the first 6 exons of
avr-14. It expressed in a subset of ring neurons and a couple neurons in the tail but not in the pharynx. Thus, it seems that ivermectin targets at least two chloride channels in two tissues:
avr-15 in the pharyngeal muscle and
avr-14 in the nervous system. The fact that ivermectin targets at least two members of a multi-gene family should make it less likely that ivermectin resistance will arise in treated parasites.