The K+ channel superfamily includes recently discovered subunits with four TM (TM) domains and two pore segments called TWIKs. Two major unanswered questions about TWIK subunits are: how do they form channels (e.g. are other subunits required?), and once formed, what are the properties of these channels? The cloning of
sup-9 (Perez de la Cruz, 1997 IWM) provided the first clue:
sup-9 is a TWIK, and based on gene interactions, SUP-9 subunits may form heterotrimeric channels with UNC-93 and SUP-10 (Greenwald and Horvitz, 1980; Levin and Horvitz, 1993). But what of other TWIKs? This question is intriguing because the sequencing consortium revealed about 50 genes for TWIK subunits in worms. We have recently found mutations in two more TWIK subunits that may help answer these questions. As previously reported (Johnstone, et al. 1998 WCWM), we found that
unc-110 encodes a TWIK subunit.
unc-110 (
e1913 )/+ mutants have a flaccid, rubber-band-Unc phenotype, as does a partial intragenic revertant,
e1913 sa859 , which is homozygous viable. Translational
unc-110 :: gfp fusions express GFP only in body-wall muscle. We hypothesize that
e1913sd causes inappropriate opening of the UNC-110 K+ channel in body wall muscles, which would oppose muscle excitation (see related abstract, Kunkel et al.). A putative null mutant with a nonsense mutation in exon 2 (
sa954 ) appears grossly wild-type. We find no genetic interactions between
unc-110 and either
unc-93 or
sup-10 (data not shown), suggesting that UNC-110 does not require the same auxilliary subunits as SUP-9 to form channels. Instead, we noticed an overlooked family of over a dozen
unc-93 -like genes in the genome. In order to identify genes that interact with
unc-110 we screened 105,000 haploid genomes for supression of the Unc phenotype of
sa859 . We have recently found that
egl-23 also encodes a TWIK. Both dominant
egl-23 alleles,
n601 and
n2579 , reduce egg-laying and enteric muscle contraction. A putative null mutant with a 213 bp deletion starting in exon 2 is grossly wild type.
egl-23 does not appear to interact genetically with the
unc-93 group, and we have screened 25,000 haploid genomes for suppression of the Egl phenotype.