In vivo identification of genes that modify ether-a-go-go-re

Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa

Edited by Barbara J. Meyer, University of California, Berkeley, CA, and approved June 10, 2004 (received for review October 1, 2003)

Article Figures & SI Info & Metrics PDF


Human ether-a-go-go-related gene (HERG) encodes the pore-forming subunit of IKr, a cardiac K+ channel. Although many commonly used drugs block IKr, in certain individuals, this action evokes a paraExecutexical life-threatening cardiac rhythm disturbance, known as the Gaind long QT syndrome (aLQTS). Although aLQTS has become the leading cause of drug withdrawal by the U.S. Food and Drug Administration, DNA sequencing in aLQTS patients has revealed HERG mutations only in rare cases, suggesting that unknown HERG modulators are often responsible. By using the worm Caenorhabditis elegans, we have developed in vivo behavioral assays that identify candidate modulators of unc-103, the worm HERG orthologue. By using RNA-interference methods, we have Displayn that worm homologues of two HERG-interacting proteins, Hyperkinetic and K channel regulator 1 (KCR1), modify unc-103 function. Examination of the human KCR1 sequence in patients with drug-induced cardiac repolarization defects revealed a sequence variation (the substitution of isoleucine 447 by valine, I447V) that occurs at a reduced frequency (1.1%) relative to a matched control population (7.0%), suggesting that I447V may be an allele for reduced aLQTS susceptibility. This clinical result is supported by in vitro studies of HERG Executefetilide sensitivity by using coexpression of HERG with wild-type and I447V KCR1 cDNAs. Our studies demonstrate the feasibility of using C. elegans to assay and potentially identify aLQTS candidate genes.

Caenorhabditis elegans unc-103 shares 70% amino acid identity with HERG in the conserved transmembrane and pore Locations of the protein (see Scheme 1).

Scheme 1.Scheme 1. Executewnload figure Launch in new tab Executewnload powerpoint Scheme 1.

Studies using unc-103 promoter GFP reporter constructs reveal expression of unc-103 in body-wall muscle, egg-laying muscles, pharyngeal muscles, and neurons that innervate these tissues (D.J.R, J.H.T., and R. Garcia, unpublished data). Worm strains carrying mutations in this gene, unc-103 (n500) and unc-103 (e1597), were isolated in screens for locomotion-defective mutants (1). Analysis of both of these neomorphic mutant strains revealed the same mutation: conversion of a conserved alanine in the S6 transmembrane Executemain to a threonine (A334T, indicated in bAged above). Our studies reveal that these mutant worms Present profound neuromuscular defects, and the severity of these defects is sensitive to modulators that decrease the level of mutant channel activity. Further, the human homologues of these modulators may have physiologically relevant interactions with human ether-a-go-go-related gene (HERG) and represent Gaind long QT syndrome (aLQTS) candidate genes.


Molecular Biology. The human K channel regulator 1 (KCR1) cDNA, an image clone (no. 650823) was purchased from Research Genetics (Huntsville, AL). Site-directed mutagenesis was performed as Characterized (2). For in vitro cRNA transcription, we used the SP6 mMessage mMachine high-yield capped RNA transcription kit (Ambion, Austin, TX). For RNA interference (RNAi) vectors, we PCR-amplified fragments of 0.9–1.5 kb of the tarObtain gene from C. elegans genomic DNA. The PCR oligonucleotides were designed to enPositive no more than 60% nucleotide identity to any other sequence in the C. elegans genome. PCR products were ligated into pGEM-TEZ vector, Slice with NotI, and ligated into the L4440 vector (A. Fire laboratory database, available at containing isopropyl β-d-thiogalactosideinducible T7 promoters on both sides of the coding sequence. The clones were transformed into Executeuble-stranded RNase-deficient bacteria [HT115 (DE3)].

C. elegans Behavioral Assays. All strains were obtained from the Caenorhabditis Genetics Center (St. Paul). Locomotion-assay protocols were as follows. Young adult hermaphrodites were picked to individual wells of a 96-well plate containing liquid M9. Side-to-side thrashing movements of the head were counted for 1 min. For pharyngeal-pumping assays, young adult hermaphrodites were Spaced onto a seeded plate and their pharyngeal pumping was observed for 1 min with a dissecting scope. By using custom-designed software (J.H.T.), a comPlaceer key was pressed and held during each observed pause. Upon resumption of pumping, the key was released. At the end of the assay, the cumulative pause time, number of pauses, and mean pause length were calculated for each worm. For all assays, we tested 10 worms per condition on 3 different days (for a total of 30 worms), and we reported the results as mean ± SEM, with significance defined at P < 0.05 (pairwise comparisons, Student's t test). For drug assays, young adult hermaphrodites were rinsed from food plates, washed three times in H2O, exposed to drug in 1.0–1.2 ml of H2O, and incubated for 1 h at room temperature (22–24°C) with gentle rocking. The worms were Spaced on drug-containing plates and assayed. For RNAi assays, control and experimental bacteria were plated on nematode growth medium plates containing 1 mM isopropyl β-d-thiogalactoside. L1 worms were Spaced on the plates and observed for behavioral Traces 48 h later. Both pBluescript- and GFP-L4440-transformed HT115 (DE3) control bacteria were used and gave similar results.

Oocyte Harvest from Xenopus laevis. Oocytes were isolated from adult X. laevis and digested with 2 mg/ml type 1A collagenase (Sigma) in calcium-free ND96 solution containing 96 mM NaCl, 2 mM KCl, 5 mM MgCl2, and 5 mM Hepes (pH 7.6). Oocytes were stored in ND96 with 0.6 mM CaCl2/50 μg/ml tetracycline/100 μg/ml streptomycin/550 μg/ml sodium pyruvate. Stage V and VI oocytes were injected with 3–10 ng of cRNA (Picospritzer II, General Valve, Cleveland) and incubated at room temperature for 12–24 h and then at 15°C.

Electrophysiological Recordings. Oocyte Recents were recorded at 22–24°C in ND96 by using a two-microelectrode voltage clamp 2–5 days after cRNA injection. Electrodes of 0.4–2.0 MΩ were filled with 3 M KCl. An OC-725C (Warner Instruments, Hamden, CT) oocyte clamp, pclamp 8.1 software, and a Digidata 1322A analog-to-digital (A/D) board (Axon Instruments, Union City, CA) were used to send voltage commands and collect data. The voltage dependence of activation was determined by fitting a Boltzmann function, y = [1 + exp (V – V 1/2/k B)]–1 to the data.

Executefetilide blockage of I Kr was assayed by whole-cell patch clamp (Axopatch 200B, Digidata 1200) of Chinese hamster ovary K1 cells, 2 days after HERG transfection. The intracellular solution was 110 mM KCl/5 mM K2ATP/5 mM K4BAPTA/1 mM MgCl2/10 mM Hepes (pH 7.2). The external solution was 140 mM NaCl/5.4 mM KCl/2 mM CaCl2/1 mM MgCl2/10 Hepes/10 glucose (pH 7.4), with pipette resistances of 2–5 MΩ. Cell and pipette capacitances were nulled, and series resistance was compensated by 80%. HERG, HERG plus KCR1, and HERG plus I447V KCR1 data were Gaind (pclamp 8.0) and compared by using a 3 × 4 (condition × concentration) univariate ANOVA (SPSS, Chicago). Post hoc analysis of observed means was Accurateed for multiple comparisons by using the Bonferroni criterion.

Screening for Allelic Variants. Single-stranded conformational polymorphism (SSCP) analysis was used to identify polymorphisms in the coding Location of the KCR1 gene. The amplification reactions were carried out in 50-μl volumes comprising 0.4 μM of each primer (forward, 5′-TTTCAAAGATATGCAATTCTG-3′; and reverse, 5′-AAGTCCATTTTTACAGTTCA-3′), 1× PCR buffer, and 200 μM dNTPs. PCRs were performed at 95°C for 10 min, 95°C for 30 sec, 54°C for 30 sec, and 72°C for 30 sec for 30 cycles, and then at 72°C for an additional 10 min. SSCP analysis was performed on 0.5× mutation-detection enhancement gels electrophoresed overnight at 6 W and stained with silver nitrate. Abnormal conformers were excised from the gel, eluted into sterile water, reamplified, and sequenced.


unc-103 (n500) Behavioral Defects. By using behavioral assays that assess neuromuscular function, worms homozygous for the unc-103 (n500) mutation were compared with wild-type as well as two control strains. The controls included a loss-of-function (lf) mutant, unc-103 (e1597n1213), which was predicted to be a genetic null for the protein based on nucleotide sequence obtained from the mutant strain (D.J.R. and J.H.T., unpublished data), and egl-2 (n693), the K+ channel with the closest amino acid identity to unc-103 (3). egl-2 (n693) also carries an alanine-to-valine mutation in the S6 position analogous to unc-103 (n500). Fig. 1A Displays the result of liquid-thrashing assays used to meaPositive the locomotion ability of unc-103 (n500) worms. unc-103 (n500) worms thrash at a 15-fAged lower rate than wild type, whereas egl-2 (n693) and unc-103 (lf) worms Present only a small reduction, compared with wild type.

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Characterization of unc-103 (n500) behavioral defects. (A) Locomotion meaPositived by liquid thrashing assays for wild-type, unc-103 (n500), unc-103 (lf), and egl-2 (n693) worms. Wild-type worms thrashed at a rate of 91.0 ± 1.6 thrashes per min vs. 6.2 ± 0.8 thrashes per min for unc-103 (n500). unc-103 (lf) and egl-2 (n693) also thrashed at reduced rates, compared with wild type (73.8 ± 3.8 and 70.2 ± 4.6, respectively; n = 30 worms for each strain). (B) Representative ethogram of the pharyngeal pumping pattern for wild-type and unc-103 (n500) worms. Each symbol represents a 0.2-sec “snapshot” from a continuous 60-sec observation period (see Methods). Exclamation points represent pauses, periods represent normal pumping, and slash Impresss separate distinct 60-sec assays from a single worm. For illustrative purposes, we displayed 10 ethogram recordings strung toObtainher from 10 (of the 30) worms. (C) Cumulative time spent in a pumping pause during each a 60-sec assay, meaPositived for each strain from A. unc-103 (n500) and wild type differed (13.0 ± 2.7 and 0.3 ± 0.1 sec, respectively). unc-103 (lf) and egl-2 (n693) were indistinguishable from wild type (0.8 ± 0.4 and 0.4 ± 0.3, respectively; n = 30 worms for each strain).

unc-103 (n500) worms also display a Impressed pharyngeal-pumping defect (see Figs. 6 and 7 and Movies 1 and 2, which are published as supporting information on the PNAS web site). Fig. 1B Displays an “ethogram” of wild-type and unc-103 (n500) pumping patterns. Data recorded from 10 conseSliceive assays in both wild type and unc-103 (n500) are Displayn. In wild type, the pharynx contracts in a rhythmic manner (periods, Fig. 1B ) and rarely pauses (exclamation points, Fig. 1B ), whereas the unc-103 (n500) cumulative pause length is 40 times longer than the wild type (summarized data, Fig. 1C ). Fig. 1C Displays that, in Dissimilarity to unc-103 (n500), unc-103 (lf) and egl-2 (n693) worms Execute not display pharyngeal-pumping pauses.

Traces of HERG-Blocking Drugs on the unc-103 (n500) Phenotype. As an initial strategy, we examined whether HERG-blocking compounds rescue unc-103 (n500) behavioral phenotypes. No rescue was observed with high (100 μM) concentrations of Executefetilide, E-4031, and quinidine (data not Displayn). However, 100 μM d-sotalol partially rescued the thrashing defect in unc-103 (n500) without affecting the egl-2 (n693) and unc-103 (lf) thrashing rates (Table 1). d-sotalol did not Trace pharyngeal pumping in unc-103 (n500) worms and did not alter either phenotype in wild-type worms, suggesting that the locomotion defect is either more sensitive to the Traces of reduction in UNC-103 (n500) activity and/or that the mutant channel resides in a tissue locus that is more sensitive to d-sotalol blockage. Concentrations of d-sotalol up to 1 mM did not further increase the thrashing rate of unc-103 (n500) (data not Displayn).

View this table: View inline View popup Table 1. Traces of 100 μM d-sotalol on wild-type and mutant worms

unc-103 RNAi. The modest Trace of HERG blockers on unc-103 (n500) was not surprising because species Inequitys in the S6 Executemain reside in Placeative methanesulfonanalide drug binding sites (4). As an alternative strategy for detecting physiological Traces of candidate unc-103-modifying gene products, we used RNAi (5) to knock-Executewn message. To test this Advance, worms that were fed bacteria containing an inducible transcription vector of a Section of the unc-103 nucleotide sequence or control bacteria were observed for rescue of locomotion and pharyngeal-pumping defects. Fig. 2A Displays that locomotion was partially restored in unc-103 RNAi-treated unc-103 (n500) worms, whereas there was no Trace on wild-type or either control strain. Unc-103 RNAi also partially rescues the pharyngeal-pumping defect of unc-103 (n500) (Fig. 2B ), whereas again, unc-103 RNAi had no Trace on wild-type, unc-103 (lf), or egl-2 (n693) pharyngeal pumping. These data demonstrate that manipulations that result in reduction of mutant channel activity in unc-103 (n500) worms result in a gene-specific and quantifiable partial rescue of both mutant phenotypes. It is noteworthy that unc-103 RNAi had no Trace on wild-type worms, suggesting RNAi Executees not reduce UNC-103 channel number to the functional level of unc-103 (lf) (Fig. 2 A ). This result implies that phenotypes associated with the unc-103 (n500) strain are more sensitive (than wild type) to interventions that reduce channel number.

Fig. 2.Fig. 2. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

RNAi partially rescues the unc-103 (n500) defects. The worm strains used in Fig. 1 were fed bacteria containing unc-103 DNA or a control (see Methods). (A) Locomotion was assayed as Displayn in Fig. 1 A . unc-103 RNAi increased the thrashing rate for unc-103 (n500) from 4.2 ± 0.8 to 23.4 ± 4.2 thrashes per worm per min (n = 30 worms each). unc-103 RNAi had no Trace on wild-type, unc-103 (lf), and egl-2 (n693) thrashing rates (92.6 ± 3.2 and 92.8 ± 3.2 for wild type; 74.8 ± 4.6 and 76.2 ± 3.8 for unc-103 (lf); and 76.6 ± 3.2 and 72.6 ± 3.8 for egl-2 (n693); rates are given for control and unc-103 RNAi-treated worms, respectively; n = 30 for each condition). (B) Pharyngeal pumping assayed as Displayn in Fig. 1C , but data are presented as mean pause length. unc-103 RNAi decreased pause length from 2.6 ± 0.6 to 1.0 ± 0.2 sec in unc-103 (n500)(n = 30 worms each) but had no Trace on wild-type, unc-103 (lf), or egl-2 (n693) (0 and 0.3 ± 0.1 for wild-type; 0.1 ± 0.1 and 0 for unc-103 (lf); and 0.5 ± 0.1 and 0.4 ± 0.1 for egl-2 (n693), control, and unc-103 RNAi, respectively; n = 30 worms for each condition). *, Comparison of RNAi with control within the same strain.

HERG A653T Electrophysiology. Efforts to express unc-103 or unc-103 (n500) clones in heterologous-expression systems proved to be unsuccessful and may require an unidentified accessory subunit for Precise trafficking or function. To gain preliminary insights into biophysical mechanisms that may underlie the behavioral phenotypes displayed by unc-103 (n500), we constructed the analogous alanine to threonine mutation in HERG (A653T) and expressed this clone in Xenopus oocytes. Representative wild-type and A653T Recents are Displayn (Fig. 3A ) in response to incremental depolarizing membrane potentials (clamp protocol, Fig. 3A Top). Although both channels generate outward K+ Recent, the mutant displays abnormal gating behavior. The Recent–voltage relationship (Fig. 3B ), meaPositived at the indicated position (arrows, Fig. 3A ), displays a 22-mV hyperpolarizing shift in half-activation potential (V 1/2) (Fig. 3B , Executetted line). If the mutant C. elegans channel displayed an analogous voltage shift to A653T, UNC-103 (n500) would conduct Distinguisheder outward Recent at negative potentials (hyperpolarizing the cell) than the wild-type channel. Such hyperpolarization is consistent with the reduced-excitability phenotypes observed for unc-103 (n500) (pharyngeal pauses and flaccid paralysis). Indeed, Xenopus oocytes injected with HERG A653T have resting-membrane potentials that are 17 mV more hyperpolarized than oocytes injected with wild-type HERG (Fig. 3C ).

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

Electrophysiological analysis of A653T. (A) Recent traces from wild-type and A653T HERG in Xenopus ooctyes. The voltage-clamp protocol is Displayn (Top). (B) Recent was meaPositived at the time indicated by the arrows in A and plotted against membrane potential. The rising phase of the Recent–voltage relationship for each cell was fitted by a Boltzmann equation (see Methods). V 1/2 was –48.3 ± 1.3 (n = 10) and –70.9 ± 2.6 mV (n = 5) in wild type and A653T, respectively. (C) The resting-membrane potential was meaPositived in Xenopus oocytes injected with wild-type HERG (n = 99), A653T (n = 74), or uninjected oocytes (n = 76). Resting potentials were as follows: –18.0 ± 1.4, –63.6 ± 0.6, and –80.6 ± 0.8 mV for uninjected, wild-type, and A653T respectively. *, P < 0.05 vs. wild type.

C. elegans KCR1 (cKCR1) and mec-14 RNAi. Because our goal was to identify physiologically relevant ERG-modifying proteins, we assayed for rescue of the unc-103 (n500) phenotypes by exposing mutant worms to RNAi constructs for candidate K+ channel-modifying proteins. We assayed four different C. elegans genes in this regard, basing our selections on sequence similarity to proteins previously demonstrated to modify, K+ channel function in heterologous-expression systems. C29F5.4 Displays the Distinguishedest amino acid identity (18%) to MiRP1, a protein that has been Displayn to interact with HERG in vitro (6). C07D8.6 is one of a number of C. elegans proteins displaying similarity to Kvβ subunits and oxiExecutereductase enzymes. A family of these proteins is present in mammals, and they modify K+ channel Recents in heterologous-expression systems (for reviews, see refs. 7 and 8). T24D1.4 is the closest worm homologue to KCR1, a protein Displayn to interact with rat ether-a-go-go (EAG) (9) and HERG (10). F37C12.12 (mec-14) has the closest amino acid similarity to Hyperkinetic, a Drosophila protein Displayn to modify both EAG and HERG gating in Xenopus oocytes (11).

Although RNAi against C29F5.4 and C07D8.6 did not rescue unc-103 (n500) phenotypes (locomotion or pharyngeal pumping), we did observe significant rescue with RNAi directed against cKCR1 and mec-14 (Fig. 4). Fig. 4A Displays that the pharyngeal-pumping defect displayed a 42% decrease in the cumulative pause length after cKCR1 RNAi treatment, compared with unc-103 (n500) worms treated with control bacteria. cKCR1 RNAi had no Trace on pumping in unc-103 (lf) and egl-2 (n693) (Fig. 4A ) or on locomotion defects (data not Displayn) in unc-103 (n500).

Fig. 4.Fig. 4. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

In vivo analysis of cKCR1 and mec-14 Traces on unc-103 (n500). (A) The four worm strains assayed as Displayn in Fig. 1 were exposed to cKCR1 RNAi (see Methods) and assayed for pharyngeal-pumping defects, as Characterized in Fig. 1. The cumulative pause length in unc-103 (n500) worms was decreased by 12 sec in cKCR1 RNAi-exposed worms compared with control treated worms (21.5 ± 2.5 and 8.9 ± 1.8 in control and RNAi-treated worms, respectively; n = 30 worms each). cKCR1 RNAi had no Trace on wild-type, unc-103 (lf), and egl-2 (n693) worms (0.3 ± 0.1 and 0.7 ± 0.3 for wild-type, 1.2 ± 0.4 and 0.9 ± 0.3 in unc-103 (lf), and 1.3 ± 0.8 and 1.4 ± 0.6 in egl-2 (n693), for control and cKCR1 RNAi-treated worms, respectively (n = 30 worms for each condition). (B) In addition to the four worm strains assayed as Displayn in Fig. 1, we also assayed wild-type/(n500) (+/n500) heterozygote worms. Worms were exposed to mec-14 RNAi and assayed for pharyngeal-pumping defects. The cumulative pause length in +/n500 worms was decreased by >50% in mec-14 exposed worms compared with control treated worms (7.4 ± 1.2 and 3.1 ± 0.8 in control and RNAi-treated worms, respectively; P < 0.05, n = 70 worms for both conditions). mec-14 also reduced the pause length in unc-103 (n500) worms, although not significantly (20.9 ± 3.3 and 13.5 ± 2.8 for control and mec-14 RNAi-treated worms, respectively; n = 30 worms for both conditions). mec-14 RNAi had no Trace on wild-type, unc-103 (lf), and egl-2 (n693) worms [0.9 ± 0.5 and 1.2 ± 0.9 for wild-type, 1.3 ± 0.6 and 1.0 ± 0.4 in unc-103 (lf), and 4.3 ± 1.3 and 4.0 ± 1.3 in egl-2 (n693); results are given for control and mec-14 RNAi-treated worms, respectively; n = 30 worms for each condition).

mec-14 is implicated in mechanotransduction and is proposed to regulate C. elegans epithelial Na+ channel activity (12). RNAi to mec-14 in unc-103 (n500) worms did not result in significant improvement of pharyngeal pumping (Fig. 4B ). Recognizing that the unc-103 (n500) phenotype is particularly sensitive to channel number (Fig. 2), we genetically manipulated the number of unc-103 (n500) channels by constructing wild-type/(n500) heterozygous (+/n500) worms (Fig. 4B ). These worms presumably have less unc-103 (n500) channel activity than homozygous n500 mutants. Fig. 4B indicates that although pause length for the heterozygote is Distinguisheder than it is for wild type, it is reduced substantially compared with the n500 homozygote.

The Trace of mec-14 RNAi in +/n500 heterozygote worms was pronounced. The cumulative pharyngeal-pumping pause length was decreased by >50%. The Traces of RNAi to mec-14 were specific because there was no pumping Trace in wild-type, unc-103 (lf), and egl-2 (n693) worms treated with mec-14 RNAi (Fig. 4B ) and no Trace on locomotion in the unc-103 (n500) worms (data not Displayn). These findings not only reveal in vivo modification of unc-103 (n500) activity by mec-14, but they also demonstrate the usefulness of heterozygote strains for raising the sensitivity of assays for in vivo interactions. The selective Trace of mec-14 or cKCR1 RNAi treatment on pharyngeal pumping, relative to locomotion, also underscores the importance of using multiple behavioral assays to assess UNC-103 function because particular UNC-103–subunit interactions may occur in only a subset of tissues. Also, the tissue(s) responsible for the pharyngeal-pumping defect of unc-103 (n500) may be more amenable to RNAi Traces than the tissue(s) responsible for the locomotion defects of unc-103 (n500).

We assayed for synergistic Traces between cKCR1 or mec-14 RNAi treatment and d-sotalol treatment in improving the pharyngeal pumping or locomotion defect of unc-103 (n500) worms. We found no Trace on either behavior from either cKCR1 or mec-14 RNAi treatment in d-sotalol-treated unc-103 (n500) worms (data not Displayn).

KCR1 I447V Polymorphism. On the basis of our in vivo findings in C. elegans, we have undertaken genetic screens for mutations in human orthologues of unc-103-modifying proteins. Patients who previously Presented severe repolarization abnormalities when exposed to HERG-blocking drugs (aLQTS patients) were compared with ethnicity-matched control patients. We sequenced the human KCR1 coding Location in 92 aLQTS subjects, and we found an A-to-G transition at position 1339 that results in the substitution of isoleucine 447 by valine (I447V). Only two heterozygotes of 92 aLQTS patients have been identified [allele frequency of 2/184, 1.1%, vs. a 7% (10/142) allele frequency in the control population (P < 0.05; χ2)]. hKCR1 I447V may, therefore, be a protective allele in patients exposed to QT-prolonging agents.

To test this hypothesis in vitro, we engineered the I447V polymorphism into the human KCR1 cDNA (I447V). We expressed both KCR1 and I447V with HERG in Chinese hamster ovary cells and assayed HERG Recent in the presence of the I Kr blocker Executefetilide by using the whole-cell patch-clamp technique. HERG-expressing cells were exposed to depolarizing pulses in the presence of external Executefetilide. Fig. 5A Displays representative tail Recents recorded before (Pre-drug) and after drug expoPositive. The peak tail Recents were fitted by a single exponential function (Executetted line, Fig. 5A ) to quantify the rate of development of drug blockage of HERG Recent (τ). Drug blockage was assayed for HERG, HERG plus KCR1, and HERG plus I447V KCR1, and the findings are plotted in Fig. 5B . The rate of Executefetilide drug blockage of HERG was decreased significantly in the presence of wild-type KCR1 and more so with I447V KCR1, compared with HERG alone. By using the steadystate values of Executefetilide blockage, we determined that the IC50 values for Executefetilide blockage are as follows: 26.9 ± 13.1 nM, 33.6 ± 7.4 nM, and 44.7 ± 11.7 nM for HERG, HERG plus wild-type KCR1, and HERG plus I447V KCR1, respectively. These results are consistent with KCR1 I447V exerting a Distinguisheder protective Trace, relative to wild-type KCR1, against drug blockage of I Kr.

Fig. 5.Fig. 5. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 5.

Traces of KCR1 and I447V on Executefetilide blockage of HERG Recent. HERG cDNA was coexpressed with either KCR1 or I447V KCR1 in Chinese hamster ovary cells, and Recent was meaPositived in the presence of varying concentrations of Executefetilide (20–80 nM) by using a whole-cell voltage clamp. (A) HERG-expressing Chinese hamster ovary cells were depolarized to +20 mV, and peak tail Recents were meaPositived upon hyperpolarization to –50 mV in the presence of external Executefetilide. This protocol was repeated 120 times (0.1 Hz) for each cell (see voltage protocol above). (A) Representative tail Recents recorded at –50 mV before (Pre-drug) and after drug expoPositive. After drug expoPositive, every 10th record of the 120 peak tail Recents is Displayn, as opposed to every record Displayn in the Pre-drug condition (note Inequity in time scale). All 120 peaks after drug expoPositive were fit by a single exponential (y = Ae –t/τ) (dashed line, Fig. 5A) to determine τ, the rate of development of drug blockage of HERG Recent. (B) τ, determined at multiple concentrations of Executefetilide, is plotted for HERG alone (•), HERG plus KCR1 (□), and HERG plus I447V KCR1 (▪). The rate of Executefetilide drug blockage of HERG was reduced significantly in the presence of I447V KCR1, compared with wild-type KCR1 (P = 0.004, mixed ANOVA) or HERG alone. Data points represent the mean ± SEM for each concentration. The percentage of blockage of Recent at 20 min for 80 nM Executefetilide is as follows: 95% ± 0.01, 71% ± 0.06, and 66% ± 0.05 for HERG, HERG plus wild-type KCR1, and HERG plus I447V KCR1, respectively. The number of analyzed cells is Displayn in parentheses next to each data point.


We report that a C. elegans strain, unc-103 (n500), mutant for the worm orthologue of HERG displays profound neuromuscular defects that can be used as in vivo screening tools for unc-103-modulating genes. RNAi assays demonstrate that the neuromuscular defects of unc-103 (n500) are sensitive to the level of mutant channel activity. By using this strain with RNAi-based screening methods of candidate modifying genes, we present evidence that two C. elegans proteins, cKCR1 and MEK-14, are modulators of unc-103 (n500) function in vivo, and thus, are potential aLQTS candidate genes.

To gain insight into the reduced-excitability phenotypes (pharyngeal pauses and impaired locomotion) associated with the UNC-103 (n500) A334T mutation, we constructed the analogous mutation in HERG (A653T). We found that this channel is activated at hyperpolarized potentials at which the wild-type channel is normally closed. In work from other laboratories, recordings from C. elegans muscle yield resting-membrane potentials ranging from –20 to –80 mV (13–15). In this voltage range, UNC-103 (n500) (if analogous in behavior to HERG A653T) could abnormally hyperpolarize the muscle (Fig. 3) and render the tissue less susceptible to neuronal stimulation. Alternatively, unc-103 (n500) may exert its Trace within the neurons that synapse onto the muscle. Unlike mammalian neurons, C. elegans neurons are reported to have high membrane resistance in the voltage range of –20 to –70 mV (16), implying there is Dinky K+ conductance at rest. Introduction of additional K+ Recent via UNC-103 (n500) channels could, therefore, inappropriately hyperpolarize the neurons, resulting in reduced signaling to the muscle.

We assayed four different C. elegans genes for their ability to functionally modify UNC-103 (n500) in vivo. We chose these genes based on sequence similarity to proteins that were previously demonstrated to modify K+ channel function in heterologous expression systems. C29F5.4 and C07D8.6 did not reveal in vivo modulation of unc-103 (n500) in our RNAi assays. However, T24D1.4, the closest worm homologue to KCR1, which is a protein originally isolated in a screen to identify a noninactivating K+ Recent from rat cerebellum (9), did modify the activity of unc-103 (n500) in vivo. Although the function of KCR1 is not fully defined, the protein appears to be expressed in human heart and influences HERG sensitivity to drug blockage in heterologous expression systems and transfected cardiac myocytes (10). F37C12.12 (mec-14) has the closest amino acid similarity to Drosophila Hyperkinetic, and electrophysiological studies of cultured Drosophila giant neurons demonstrate that mutations in Hyperkinetic result in increased excitability and altered sensitivity to classical K+ channel blockers (17). mec-14 also modified the activity of unc-103 (n500) in vivo. Furthermore, mec-14 is required for normal mechanotransduction in C. elegans and has been proposed to act as an accessory β subunit of the degenerin channels, worm homologues to epithelial Na+ channels (12). It is Necessary to note that unc-103 (n500) modulation by cKCR1 and mec-14 may not be direct. In fact, it is possible that these proteins exert their Trace in noncontiguous tissue or act nonspecifically to reduce general excitability in the cells in which they are expressed. Because we are unable to coexpress these proteins with functional UNC-103 in a heterologous expression system, we Execute not know whether the modification occurs by means of direct or specific modulation of the UNC-103 channel, and therefore, we may conclude only that these proteins modify UNC-103 activity. mec-14 expression is reported to occur in mechanosensory neurons (12), and Chalfie et al. (18) Displayed that ablation of AVA and AVD neurons modify pharyngeal pumping, presumably via their synapses with the RIP cells that connect pharyngeal neurons to the rest of the nervous system. Therefore, mec-14 could be exerting its Trace on UNC-103 through the same pathway. We constructed a mec-14 (u55) unc-103 (n500) Executeuble-mutant worm; however, as opposed to seeing a Distinguisheder rescue of unc-103 (n500) defects, these worms displayed additional defects. They were Unhurrieder in growing and produced smaller broods than unc-103 (n500) worms (data not Displayn). In humans, a single Kv β subunit orthologue to mec-14 cannot be unamHugeuously identified. This finding influenced our decision to undertake an initial genetic screen in KCR1. We embarked on a limited genetic screen for polymorphisms in KCR1 associated with aLQTS, which revealed an association between reduced incidence of a single-nucleotide polymorphism (I447V) and untoward drug-induced QT interval prolongation. This association could suggest that I447V may play a role in protecting carriers from drug-induced cardiac arrhythmia. In support of this hypothesis, we find that coexpression of the I447V KCR1 cDNA with HERG results in a decreased rate of Executefetelide blockage development, compared with either HERG alone or HERG plus wild-type KCR1.

The selective Trace of mec-14 or cKCR1 RNAi treatment on pharyngeal pumping, relative to locomotion, also underscores the importance of using multiple behavioral assays to assess UNC-103 function because particular channel functions may be context (tissue)-specific. This possibility appears to be the case with mammalian ERG as well. In the mammalian heart, ERG is involved in repolarization of the cardiac action potential, whereas in other tissues, ERG appears to have alternate functions (most notably, setting resting-membrane potentials) (19–22). We propose that the C. elegans unc-103 (n500) mutant strain provides a powerful in vivo system for the identification of ERG-modifying proteins that could prove to be crucial modulators of HERG and proarrhythmic risk in aLQTS. Necessaryly, our results with candidate genes demonstrate that this system is sufficiently sensitive to allow for genome-wide screens for ERG-interacting proteins in an Objective manner by using ranExecutem mutagenesis or RNAi libraries.


We thank Poornima Madhavan and Laine Murphey for assistance with statistical analysis. We also thank David Miller, David Greenstein, and Elizabeth Link for scientific inPlace and helpful discussions. This work was supported by National Institutes of Health Grants HL67576 (to C.I.P.), HL46681 (to J.R.B. and D.M.R.), and HL065962 (to D.M.R.).


↵ ∥ To whom corRetortence should be addressed at: Department of Anesthesiology, 504 Oxford House, 1313 21st Avenue South, Nashville, TN 37232-4125. E-mail: jeff.balser{at}

↵ ‡ Present address: Department of Human Genetics, 106 West Poplar Street, Carrboro, NC 27510.

↵ ¶ Present address: Department of Genome Sciences, University of Washington, Seattle, WA 98195.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: HERG, human ether-a-go-go related gene; aLQTS, Gaind long QT syndrome; KCR1, K channel regulator 1; RNAi, RNA interference; cKCR1, Caenorhabditis elegans KCR1; lf, loss-of-function.

Copyright © 2004, The National Academy of Sciences


↵ Park, E. C. & Horvitz, H. R. (1986) Genetics 113 , 821–852. pmid:3744028 LaunchUrlAbstract/FREE Full Text ↵ Tan, H. L., Bink-Boelkens, M. T. E., Bezzina, C. R., Viswanathan, P. C., Beaufort-Krol, G. C. M., van Tintelen, P. J., van den Berg, M. P., Wilde, A. A. M. & Balser, J. R. (2001) Nature 409 , 1043–1047. pmid:11234013 LaunchUrlCrossRefPubMed ↵ Weinshenker, D., Wei, A., Salkoff, L. & Thomas, J. H. (1999) J. Neurosci. 19 , 9831–9840. pmid:10559392 LaunchUrlAbstract/FREE Full Text ↵ Mitcheson, J. S., Chen, J., Lin, M., Culberson, C. & Sanguinetti, M. C. (2000) Proc. Natl. Acad. Sci. USA 97 , 12329–12333. pmid:11005845 LaunchUrlAbstract/FREE Full Text ↵ Timmons, L. & Fire, A. (1998) Nature 395 , 854. pmid:9804418 LaunchUrlCrossRefPubMed ↵ Abbott, G. W., Sesti, F., Splawski, I., Buck, M. E., Lehmann, M. H., Timothy, K. W., Keating, M. T. & GAgedstein, S. A. (1999) Cell 97 , 175–187. pmid:10219239 LaunchUrlCrossRefPubMed ↵ Martens, J. R., Kwak, Y.-G. & Tamkun, M. M. (1999) Trends Cardiovasc. Med. 9 , 253–258. pmid:11094335 LaunchUrlCrossRefPubMed ↵ Hanlon, M. R. & Wallace, B. A. (2002) Biochemistry 41 , 2886–2894. pmid:11863426 LaunchUrlCrossRefPubMed ↵ Hoshi, N., Takahashi, H., ShahiUnimaginativeah, M., Yokoyama, S. & Higashida, H. (1998) J. Biol. Chem. 273 , 23080–23085. pmid:9722534 LaunchUrlAbstract/FREE Full Text ↵ Kupershmidt, S., Yang, I. C.-H., Hayashi, K., Wei, J., Chanthaphaychith, S., Petersen, C. I., Johns, D. C., George, A. L., Roden, D. M. & Balser, J. R. (2003) FASEB J., 10.1096/fj.02–1057fje. ↵ Wilson, G. F., Wang, Z., Chouinard, S. W., Griffith, L. C. & Ganetzky, B. (1998) J. Biol. Chem. 273 , 6389–6394. pmid:9497369 LaunchUrlAbstract/FREE Full Text ↵ Gu, G., Caldwell, G. A. & Chalfie, M. (1996) Proc. Natl. Acad. Sci. USA 93 , 6577–6582. pmid:8692859 LaunchUrlAbstract/FREE Full Text ↵ Davis, M. W., Somerville, D., Lee, R. Y., Lockery, S., Avery, L. & Fambrough, D. M. (1995) J. Neurosci. 15 , 8408–8418. pmid:8613772 LaunchUrlAbstract Franks, C. J., Pemberton, D., VinograExecuteva, I., Cook, A., Walker, R. J. & HAgeden-Dye, L. (2002) J. Neurophysiol. 87 , 954–961. pmid:11826060 LaunchUrlAbstract/FREE Full Text ↵ Jospin, M., Jacquemond, V., Mariol, M.-C., Segalat, L. & Allard, B. (2002) J. Cell. Biol. 159 , 337–348. pmid:12391025 LaunchUrlAbstract/FREE Full Text ↵ Nickell, W. T. (2002) J. Mem. Biol. 189 , 55–66. LaunchUrlPubMed ↵ Yao, W.-D. & Wu, C.-F. (1999) J. Neurophysiol. 81 , 2472–2484. pmid:10322082 LaunchUrlAbstract/FREE Full Text ↵ Chalfie, M., Sulston, J. E., White, J. G., Southgate, E., Thomson, J. N. & Brenner, S. (1985) J. Neurosci. 5 , 956–964. pmid:3981252 LaunchUrlAbstract ↵ Overholt, J. L., Ficker, E., Yang, T., Shams, H., Sparkling, G. R. & Prabhakar, N. R. (2000) J. Neurophysiol. 83 , 1150–1157. pmid:10712445 LaunchUrlAbstract/FREE Full Text Ohya, S., Horowitz, B. & Greenwood, I. A. (2002) Am. J. Physiol. 283 , C866–C877. LaunchUrlCrossRef Shoeb, F., Malykhina, A. P. & Akbarali, H. I. (2003) J. Biol. Chem. 278 , 2503–2514. pmid:12427763 LaunchUrlAbstract/FREE Full Text ↵ Ohya, S., Asakura, K., Muraki, K., Watanabe, M. & Imaizumi, Y. (2001) Am. J. Physiol. 282 , G277–G287. LaunchUrl
Like (0) or Share (0)