Aminoglycoside antibiotics restore functional expression of truncated HERG channels produced by nonsense mutations

Aminoglycoside antibiotics restore functional expression of truncated HERG channels produced by nonsense mutations Yan Yao, MD,*a Siyong Teng, MD,*b Ning Li, MD,* Yinhui Zhang,* Penelope A. Boyden, PhD,† Jielin Pu, MD, PhD*b From the *Center for Arrhythmia Diagnosis and Treatment, Cardiovascular Institute and Fu Wai Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing, China, and †Department of Pharmacology, Columbia University, New York, New York. BACKGROUND Pharmacologic restoration of the trafficking defects of HERG missense mutations has been documented. However, whether correction of HERG nonsense mutations is possible is unknown. OBJECTIVE The purpose of this study was to investigate the effect of aminoglycoside antibiotics on the expression of nonsense mutants expected to produce truncated HERG channels. METHODS HERG channel and mutant currents were recorded by whole-cell patch clamp techniques. Pharmacologic rescue was applied by culturing the cells in 400 ␮g/mL G-418 or gentamicin for 24 hours. RESULTS Current densities were significantly reduced in cells expressing R1014X and W927X mutants compared to those of cells expressing wild-type (WT) HERG. R863X and E698X mutants failed to generate any typical HERG currents. Mean peak tail current density of R1014X mutant was significantly lower than that of WT (3.9 ⫾ 1.4 pA/pF, n ⫽ 8, vs 47.8 ⫾ 6.3 pA/pF, n ⫽ 12, P ⬍.05) and increased to 12.7 ⫾ 3.3 pA/pF (n ⫽ 7, P ⬍.05) and 18.3 ⫾

Introduction Long QT syndrome is an inherited arrhythmia syndrome characterized by a prolonged QT interval and recurrent ventricular arrhythmia leading to syncope or sudden cardiac death. Ten genes encoding ion channels or structure proteins associated with long QT syndrome have been identified.1–3 The human ether-à-go-go-related gene (HERG) encodes the ␣ subunit of the rapidly activating component of the delayed rectifier potassium channel current IKr. Mutations in HERG have been proven to result in type 2 long QT syndrome (LQT2) by a “loss-of-function” mechanism. Of the more than 200 HERG mutations linked to LQT2, 62% are misa

Dr Ms. Yao is a PhD student Dr Siyong Teng and Dr Jielin Pu contributed equally to this article. This work was funded by “973” National Basic Research Program of China (Program No. 2007CB512000, Project No. 2007CB512008 to Dr. Pu) and National Natural Science Foundation of China (No. 30571040 to Dr. Teng). Address reprint requests and correspondence: Dr. Jielin Pu, Center for Arrhythmia Diagnosis and Treatment, Cardiovascular Institute and Fu Wai Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, 167 Bei-Li-Shi Road, Beijing, 100037, China. E-mail address: [email protected]. (Received September 8, 2008; accepted January 11, 2009.) b

3.7 pA/pF (n ⫽ 8, P ⬍.05) after G-418 and gentamicin treatment. The voltage dependence of activation of R1014X was also restored after drug treatment. Furthermore, expression of full-length proteins for R1014X induced by drugs was detected by western blot and confocal imaging. Similar results were observed in W927X. For R863X and E698X, however, gentamicin treatment had no effect. In the cells cotransfected with WT/R1014X, gentamicin and G-418 demonstrated different results: gentamicin, but not G-418, increased the current density by 2.2-fold (n ⫽ 12, P ⬍.05). CONCLUSION The study findings provide proof of principle that interventions designed to read through premature stop mutations may at least partially reverse the LQT2 phenotype in vitro. KEYWORDS Long QT syndrome; HERG; Nonsense mutation; Aminoglycoside antibiotics (Heart Rhythm 2009;6:553–560) © 2009 Heart Rhythm Society. All rights reserved.

sense mutations, 30% are nonsense or frameshift mutations, and 8% are other mutations.4,5 Previous studies have demonstrated that most LQT2-linked missense mutations reduce functional HERG current by a “trafficking-deficient” mechanism.6 This trafficking defect could be rescued by either incubation at lower temperatures (27°C) or treatment with drugs such as HERG channel blockers (E4031, cisapride) and other compounds (thapsigargin).6 – 8 Successful rescue was evident by the expression of the complex-glycosylated mature form of the HERG protein and an increase in current density.6,9,10 Unlike missense mutations, nonsense mutations always introduce premature termination codons, giving rise to toxic truncated proteins that exert a dominant-negative effect on wild-type (WT) channel.11,12 Currently, no studies on whether these truncated HERG proteins can be pharmacologically rescued to express full-length functioning proteins have been reported. Aminoglycoside antibiotics are a group of drugs that are known to reduce the accuracy and fidelity of translation and to encourage the expression of functional proteins, a phenomenon observed in cystic fibrosis, Duchenne muscular dystrophy, and other inherited diseases resulting from nonsense mutations.13–15 Studies in heterologous expression

1547-5271/$ -see front matter © 2009 Heart Rhythm Society. All rights reserved.

doi:10.1016/j.hrthm.2009.01.017

554 systems, animal models, and clinical trials have demonstrated that aminoglycoside antibiotics can suppress the nonsense mutation and partially restore functional fulllength expression by permitting the readthrough of premature termination codons. Depending on the particular stop codon and its upstream and downstream sequences, readthrough efficacy induced by different antibiotics can vary broadly. In 1985, G-418 was shown to suppress a nonsense mutation and restore gene activity to almost 20% of WT levels.16 The effect of gentamicin on readthrough efficacy also has been well documented.13–15 In the present study, we tested the effect of G-418 and gentamicin on four nonsense mutations bearing premature termination codons. Because LQT2 is inherited in an autosomal dominant pattern, patients generally have a heterozygous genotype. Therefore, we also studied the effect of these drugs on heteromultimeric channels by cotransfection of WT and mutant HERG genes.

Heart Rhythm, Vol 6, No 4, April 2009

Materials and methods

mM): 120 KAsp, 20 KCl, 1 MgCl2, 5 HEPES, 10 EGTA, 2 Na2-phosphocreatine, and 4 K2ATP (pH 7.2 with KOH). pCLAMP 10.0 software (Axon Instruments, Foster City, CA) was used to record currents and analyze the data. Origin 6.0 (MicroCal, Northampton, MA, USA) was used for fitting current traces and generating graphs. All patch clamp experiments were performed at room temperature. No leak subtraction was applied during recording. Current densities were calculated by dividing the peak tail current by the cell capacitance. Voltage dependence of channel activation was obtained by fitting normalized curves with the Boltzmann function7: I/Imax ⫽ 1/(1 ⫹ exp[(V1/2 – V)/ k]), where I ⫽ measured tail current, Imax ⫽ maximal tail current, V ⫽ applied membrane voltage, V1/2 ⫽ voltage at which half of the channels are activated, and k ⫽ slope factor. Data are presented as mean ⫾ SEM. Student’s t-test and ANOVA coupled with a Student-Newman-Keuls test were used for statistical analysis. P ⬍.05 was considered significant.

Site-directed mutagenesis and constructs

Western blot analysis

WT-HERG cDNA and N470D-HERG cDNA were kindly presented by Dr. Zhengfeng Zhou (Oregon Health and Science University, Portland, OR, USA) and subcloned into pcDNA3 vector (Invitrogen, Carlsbad, CA, USA) through the BamHI and EcoRI sites. Four mutants were constructed by polymerase chain reaction– based mutagenesis strategy as described previously.4,17–19 All mutations were verified by sequencing and then subcloned into pcDNA3.1 vector through the HindIII and BamHI sites.

Membrane protein preparation and western blot experiments were performed as previously described.19 Equal amounts of protein were run in 8% SDS-polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane. The membranes were incubated overnight at 4°C with antibody against N-terminus or C-terminus of HERG channel protein (1:400 dilution; Alomone Laboratories, Jerusalem Israel). The antibody was visualized with a BCIP/NBT detection kit.

Cell culture and transfection Human embryonic kidney 293 (HEK293) cells were cultured in minimal essential medium supplemented with 10% fetal bovine serum and 1% glutamine in a humidified 5% CO2 incubator at 37°C. HEK293 cells were transiently transfected with either 0.6 ␮g WT or an equal amount of mutant genes (Effectene Transfection Reagent, Qiagen, Hilden, Germany). For cotransfection experiments, cells were transiently cotransfected with 0.6 ␮g WT and 0.6 ␮g R1014X or R863X mutant genes to express the heteromultimeric channels. Green fluorescent protein (GFP) gene (0.2 ␮g) was cotransfected as an indicator. Twenty-four hours after transfection, G-418 (Promega, Madison, WI, USA) or gentamicin (Sigma, St. Louis, MO, USA) was added to the medium and cultured for 24 hours at a final concentration of 400 ␮g/mL. Cells were cultured in drug-free medium for 1 hour before patch clamp recordings.10

Patch clamping Membrane currents were recorded by whole-cell patch clamp using suction pipettes having a tip resistance of 2 to 5 M⍀ when filled with internal solution. Cells were superfused with Tyrode’s solution containing the following (in mM): 135 NaCl, 5.4 KCl, 1.8 CaCl2, 1 MgCl2, 0.33 NaH2PO4, 10 glucose, and 10 HEPES (pH 7.4 with NaOH). The internal pipette solution contained the following (in

Confocal imaging Cells were fixed with 4% paraformaldehyde for 20 minutes at room temperature, permeabilized with 0.2% Triton X-100 for 20 minutes, and then incubated in blocking solution (5% bovine serum albumin and 0.05% azide in phosphate buffered saline) for 30 minutes. Subsequently, cells were incubated in blocking solution with rabbit polyclonal antibody against HERG channel (1:400 dilution) overnight at 4°C. After rinsing out of phosphate buffered saline for three times, cells were incubated with anti-rabbit TRITC-conjugated secondary antibody for 30 minutes. Confocal images were obtained using a confocal laser scanning microscope (Zeiss LSM510, Jena Germany) and analyzed using LSM Image Browser software.

Results Nonsense mutations inhibited the functional expression of HERG channels The HERG channel sequence is composed of 15 exons on chromosome 7q35. All four nonsense mutations in this study are located in the C-terminus at exons 8, 10, 12, and 13, respectively (Figure 1A). Amino acids 667 to 1159 of the C-terminal region have been implicated in channel maturation, assembly, and trafficking, such as the cyclic nucleotide binding domain (cNBD), RXR signal, and tetramerizing coiled-coil domain (TCC) (Figure 1B).20,21

Yao et al

Aminoglycoside Antibiotics Restore Truncated HERG Channels

555

Figure 1 Schematic structure of HERG gene and its presumable truncations. A: Location of four mutations investigated in this study. The numbers within the boxes denote the exons, and the connecting horizontal lines between the boxes denote the introns. B: Schematic representation of the HERG gene with six transmembrane domains (S1–S6) and several important regions (PAS, cNBD, RXR signal, TCC domain, region between residues 866 – 899 and 1018 –1122). The presumable truncations are indicated by downward arrows.

Figure 2A shows representative current recordings from HEK293 cells transiently transfected with WT-HERG and four nonsense mutants (R1014X, W927X, R863X, E698X). In Figure 2B, the mean peak tail current densities of R1014X (3.9 ⫾ 1.4 pA/pF, n ⫽ 8, P ⬍.05) and W927X (11.6 ⫾ 2.4 pA/pF, n ⫽ 8, P ⬍.05) channels were significantly lower than that of WT channel (47.8 ⫾ 6.3 pA/pF, n ⫽ 12). R863X and E698X mutants generated currents comparable to endogenous currents. No tail currents were observed in at least eight cells, illustrating their inability to form functional channels. The voltage dependence of acti-

vation did not differ between WT-HERG (V1/2 ⫽ – 6.1 ⫾ 0.3 mV, k ⫽ 9.5 ⫾ 0.3) and W927X mutant channels (V1/2 ⫽ – 4.5 ⫾ 0.6 mV, k ⫽ 9.8 ⫾ 0.5), whereas for the R1014X mutant, the activation curve was negatively shifted with V1/2 ⫽ –12.1 ⫾ 1.8 mV (P ⬍.05) and k ⫽ 10.2 ⫾ 1.2 (P ⬎.05, Figures 3C and Figure 4C).

Aminoglycoside antibiotics enhanced the functional expression of mutant channels The effect of aminoglycoside antibiotics was first investigated on WT-HERG channel. Peak tail current densities of

Figure 2 Expression of wild-type (WT) and mutant HERG channels transfected in HEK293 cells. A: Representative families of current traces recorded from cells expressing WT and four mutant HERG channels as well as untransfected cells (control). Cells were held at – 80 mV and depolarized to voltages between – 60 and 50 mV for 4 seconds, and then clamped to –50 mV for 4 seconds. B: Current/voltage relationship of peak tail currents recorded from WT-HERG, R1014X, and W927X mutant channels.

556

Heart Rhythm, Vol 6, No 4, April 2009

Figure 3 Effect of aminoglycoside antibiotics on R1014X mutant channel. A, B: Examples of current recordings and current/voltage relationship obtained from cells transfected with R1014X mutant channel in the presence of G-418 or gentamicin. C: Activation curves of wild-type (WT) and R1014X mutant channels. Mean peak tail current densities were calculated and plotted against the test potential, then fitted to a Boltzmann function. The leftward shift on the activation curve of R1014X mutant channel was corrected by drug treatment.

WT channels were 41.2 ⫾ 5.4 pA/pF (n ⫽ 7) before and 46.3 ⫾ 5.8 pA/pF n ⫽ 7 after 24 hours of the combined G-418 and gentamicin treatment and not significantly different from that of untreated WT channel (47.8 ⫾ 6.3 pA/pF, n ⫽ 12, both P ⬎.05). To test whether functional expression of nonsense mutants could be pharmacologically rescued, cells transfected with R1014X were incubated in culture medium containing 400 ␮g/mL G-418 or gentamicin for 24 hours. As shown in Figure 3A and 3B, peak tail current density was significantly increased after G-418 (12.7 ⫾ 3.3 pA/pF, n ⫽ 7, P ⬍.05) or gentamicin treatment (18.3 ⫾ 3.7 pA/pF, n ⫽ 8, P ⬍.05) compared to R1014X alone. In addition, the leftward shift of the R1014X channel activation curve was corrected by drug treatment (Figure 3C), indicating that the electrophysiologic properties of the mutant channel were partially restored. The average values of V1/2 and k were –3.5 ⫾ 0.6 mV and 10.8 ⫾ 0.6 in G-418 –treated cells (n ⫽ 7), and – 4.8 ⫾ 0.7 mV and 10.3 ⫾ 0.5 in gentamicin-treated cells (n ⫽ 8). Similar results were observed in the W927X mutant (Figures 4A and 4B). Peak tail current density was enhanced from 11.6 ⫾ 2.4 pA/pF (n ⫽ 8) to 19.5 ⫾ 2.7 pA/pF (n ⫽ 7, P ⬍.05) after gentamicin treatment. The voltage dependence of activation did not differ among WT, W927X, and gentamicin-treated W927X mutant channel (Figure 4C). Interestingly, gentamicin had no effect on the upstream

mutations R863X and E698X, as no tail current could be observed after drug treatment (data not shown).

Nonsense mutations produced truncated channel proteins and changed distribution pattern Expression of WT and mutant HERG channel proteins was analyzed by western blotting using a HERG N-terminal antibody (Figure 5A). Two bands at around 135 kDa and 155 kDa were observed in the WT-HERG lane, representing the core-glycosylated immature form and the complex-glycosylated mature form of HERG channel proteins, respectively. Both R1014X and R863X expressed truncated HERG proteins. R1014X showed two bands at 120 kDa and 140 kDa, signifying the presence of core and complex glycosylation. However, R863X only showed a single 90kDa band, indicating that the trafficking defect may have inhibited the acquisition of complex glycosylation. The missense mutation N470D, previously proven to have a trafficking-deficient phenotype,6,10 also demonstrated a single band, at about 135 kDa. To test subcellular localization of WT and mutant HERG channels, confocal imaging was performed using an Nterminal antibody. As shown in Figure 5B, no fluorescence was detected in untransfected HEK293 cells. In cells expressing WT-HERG, strong red fluorescence signals were visualized predominantly in the plasma membrane. Cells transfected with R1014X and W927X mutants demonstrated a fluorescence distribution pattern similar to WT-

Yao et al

Aminoglycoside Antibiotics Restore Truncated HERG Channels

557

Figure 4 Rescue of HERG current in cells expressing W927X mutant by gentamicin. A: Representative current recordings of W927X mutant channel with or without gentamicin. B: Mean peak tail current densities of R1014X and W927X mutant channels were normalized to WT-HERG channel. After treated by gentamicin, the proportion of peak tail current densities of mutant channels to that of WT-HERG channel was significantly enhanced (from 8% to 38% in R1014X mutant channel and from 24% to 41% in W927X mutant channel). C: Voltage dependence of activation was not much different among WT-HERG, W927X alone, and W927X with gentamicin treatment. *P ⬍.05.

HERG cells. In contrast, cells transfected with R863X and E698X mutants displayed a restricted perinuclear distribution. This was consistent with the western blot experiments. When channel proteins were transported to the cell surface, two protein bands were shown, as seen in cells expressing R1014X channel. When mutant proteins with trafficking defect were retained within the endoplasmic reticulum, only one protein band could be detected, as seen in cells expressing R863X channel.

Aminoglycoside antibiotics induced expression of full-length HERG channel proteins When a HERG C-terminal antibody, which is incapable of recognizing truncated channel proteins, was used, no protein bands were observed in cells transfected with R1014X mutant (Figure 5C). After treatment with G-418 or gentamicin, however, two protein bands were detected, implying translational readthrough for full-length HERG channel proteins. This was accordant with the gentamicin-enhanced current density observed above (Figures 3A and 3B). In contrast, the trafficking defect of N470D apparently could not be rescued, as only one protein band was visible following treatment (Figure 5C). Also, gentamicin could not induce any western band in the R863X mutant when a C-terminal antibody was used (data not shown). These observations were further supported by the confocal images shown in Figure 5D. In HEK293 cells expressing the R1014X mutant, gentamicin induced red fluores-

cence signals when HERG C-terminal antibody was used. However, no fluorescence signals were observed before and after gentamicin treatment in R863X mutant.

Aminoglycoside antibiotics diminished the dominant-negative effect of nonsense mutants One of the mechanisms causing LQT2 is the “dominantnegative” effect, in which functional expression of WTHERG is reduced by more than half when mutant subunits assemble into heteromultimeric channels. To study the interaction between WT and mutant channels, the same amounts of WT and R1014X cDNA, or WT and R863X cDNA, were transfected into HEK293 cells. In Figures 6A and 6B, the tail current density of WT/R1014X channels was less than half of that from WT channel alone (13.1 ⫾ 2.2 pA/pF, n ⫽ 8, vs 47.8 ⫾ 6.3 pA/pF, n ⫽ 12, P ⬍.05), indicative of “dominant-negative” inhibition. However, this phenomenon was absent in WT/R863X channels, which produced approximately 60% of WT channel current density (28.1 ⫾ 3.4 pA/pF, n ⫽ 7, vs 47.8 ⫾ 6.3 pA/pF, n ⫽ 12, P ⬍.05). Figure 6D shows the western blot results of cells cotransfected with WT and mutant cDNAs. Two bands representing the 135-kDa immature form and the 155-kDa mature form of HERG were observed in all three lanes. We next investigated the effects of aminoglycoside antibiotics on heteromultimeric channels. As shown in Figures 6A and 6B, gentamicin enhanced the expression of WT/ R1014X channels by increasing the average tail current

558

Heart Rhythm, Vol 6, No 4, April 2009

Figure 5 Western blot analysis and confocal imaging for wild-type (WT) and mutant HERG channels. A: Western blot of WT and mutant HERG channels was performed using HERG N-terminal antibody. B: Images taken from HEK293 cells transiently transfected with WT, R1014X, W927X, R863X, and E698X mutants as well as from untransfected cells (as control). For each mutant, a phase contrast photomicrograph (right) and confocal imaging using anti-HERG N-terminal antibody (left) are shown. Calibration bar ⫽ 5 ␮m. C: Western blot analysis of cells expressing R1014X mutant channel in the absence and presence of G-418 or gentamicin. N470D missense mutant channel was used to test the effect of aminoglycoside antibiotics on defective trafficking. D: Subcellular distribution of HEK293 cells transfected with WT-HERG, R1014X, and R863X mutants before and after gentamicin treatment. C-terminal antibody was used. Calibration bar ⫽ 20 ␮m.

density 2.2-fold (29.2 ⫾ 3.7 pA/pF, n ⫽ 12, P ⬍.05). However, current density remained at the same level following G-418 treatment (13.7 ⫾ 1.8pA/pF, n ⫽ 8, P ⬎.05). Of note, G-418 and gentamicin treatment did not shift the activation curve of WT/R1014X channels (Figure 6C).

Discussion In this study, we first compared the electrophysiologic expression of WT-HERG with four HERG nonsense mutations that produce C-terminus truncations of varying lengths. HERG-like current could be elicited in both R1014X and W927X mutants, but tail current was lost when the more upstream mutations R863X and E698X were used. These data were consistent with results observed in other truncated or deleted C-terminal constructs.17,19,22–25 Aydar and Palmer22 reported that a minimum of 881 residues is necessary for channel expression and that truncation of 311 or more residues from the C-terminus will result in a nonfunctional HERG channel, whereas truncation of 215 or fewer residues has no discernable effects on channel activity. We observed a similar trend in our experiments. R1014X and W927X mutants have a C-terminal truncation of 146 and 233 amino acids and gave rise to HERG-like currents. On the other hand, R863X and E698X mutants resulted in truncation of 297 and 462 amino acids and

disrupted IKr. This suggests that HERG functional expression is dependent on the length of the C-terminal truncation. It has been reported that a 104-amino-acid C-terminal domain (1018-1122) is critical for faithful recapitulation of IKr.18 A segment within this region identified as the TCC domain (1036-1074) has been shown to be the dominant factor driving subunit assembly; none of the mutant channels lacking TCC was functionally active.20 However, in our study, both R1014X and W927X mutants, presumably without the TCC domain, still generated typical IKr, suggesting that the TCC domain may not be essential for assembly. Another segment of interest is the amino acid triplet RXR (position 1005-1007), an endoplasmic reticulum retention signal that reduces cell trafficking when exposed.26 Although RXR signal is retained in the R1014X mutant, two protein bands could be detected, suggesting R1014X does not result in a trafficking defect. The mutants studied responded variably to aminoglycoside antibiotics treatment. These drugs partially restored electrophysiologic expression for R1014X and W927X mutants but had no effect on R863X and E698X mutants. Akhavan et al21 reported that residues 860-899 are indispensable for endoplasmic reticulum exit and normal stability of the channel protein. Point mutations or premature termination mutations within this domain can disrupt sur-

Yao et al

Aminoglycoside Antibiotics Restore Truncated HERG Channels

559

Figure 6 Aminoglycoside antibiotics diminished the dominant-negative effect of nonsense mutants. A: Representative current traces recorded for WT-HERG, WT/R863X, and WT/R1014X. Dominant negative effect was observed on WT/R1014X channels. B: Current density of WT/R1014X was significantly increased with gentamicin treatment but not with G-418. C: No significant changes in activation curves were observed with drug treatment. D: Western blot experiments of cells expressing WT-HERG, WT/R1014X, and WT/R863X channels. Two bands were detected in all lanes, representing a 135-kDa immature form and 155-kDa mature form of HERG proteins. *P ⬍.05; †P ⬎.05.

face delivery through defective trafficking. It is conceivable that because both R863X and E698X lack C-terminal residues 860-899, they may have difficulty exiting the endoplasmic reticulum and thus cannot be pharmacologically rescued. Furthermore, truncated channel proteins are more rapidly degraded by the nonsense-mediated mRNA decay mechanism.27 The role of cyclic nucleotide binding domain (cNBD, amino acids 750 – 870) in HERG channel also has been elucidated in terms of channel trafficking.23 Deletion of any structural motifs along the cNBD domain can disrupt channel trafficking, which probably provides another explanation for the molecular defect and drug response in mutants R863X and E698X. Although both G-418 and gentamicin increased the functional expression of R1014X mutant, they generated different effects on WT and heteromeric WT/R1014X HERG channels, indicating that different aminoglycoside antibiotics may induce various effects on a certain kind of HERG channel. Similar results has been observed in mutations of other genes, such as the cystic fibrosis transmembrane receptor (CFTR) gene in cystic fibrosis, ataxia–telangiectasia mutated (ATM ) gene in ataxia–telangiectasia, and arginine vasopressin receptor V2 (AVPR2) gene in nephrogenic diabetes insipidus.13–15,28,29 It has been suggested that aminoglycosides-induced readthrough of human premature termination codons mutations may have therapeutic potential for nearly one third of all genetic disorders.28 However, it is

important to recognize that readthrough efficiency may be influenced by many factors: the characteristics of three stop codons, the local nucleotide context, the effect of nonsensemediated mRNA decay, the toxic effects of different aminoglycoside antibiotics, and other unknown factors involved in readthrough regulation.27,30 It is generally believed that aminoglycoside antibiotics readthrough stop codons with dramatically different efficiencies, and UAA is less sensitive to aminoglycosides than UAG and UGA.30 Interestingly, a nonsense mutation (E375X) in the KCNA5 gene encoding Kv1.5, a voltage-gated ultrarapid delayed rectifier potassium channel (IKur), was found in a patient with idiopathic atrial fibrillation.31 The mutant disrupted IKur, which is vital for atrial repolarization, exerted a dominant-negative effect on the WT channel perhaps underlying the arrhythmia. Rescue was applied with 1 mg/mL gentamicin. As a consequence, functional expression of E375X and WT/E375X were significantly restored, similar to the results of our study. This was the first study that linked the effect of aminoglycoside antibiotics with an arrhythmia-related mutant gene, validating the clinical potential to treat cardiac arrhythmia syndromes caused by nonsense mutations.

Study limitations and conclusion The findings of this study provide proof of principle that interventions designed to read through premature stop mu-

560 tations may at least partially reverse the LQT2 phenotype in vitro. However, although certain nonsense mutations can be pharmacologically rescued, others cannot be. This will limit the use of these drugs in clinical trials. The consequences of premature stop codons in the natural state and after aminoglycosides treatment are not identified. Further studies are needed to elucidate how aminoglycoside antibiotics change stop codons at the molecular level and the exact outcome brought about by readthrough at the protein level. The plasmids in this study were transfected into HEK293 cells, a noncardiac heterologous expression system. In order to further establish the feasibility of pharmacologic correction in cardiomyocytes, studies using the native cardiomyocyte system (i.e., neonate mouse cardiomyocytes) should be performed.

Acknowledgments We thank Professor Shiqiang Wang (Beijing University, Beijing, China) for help with using the confocal laser scanning microscope. We also thank Dr. Zhengfeng Zhou (Oregon Health and Science University, Portland, OR, USA) for providing plasmids and Yaoli Pu (Duke University, Durham, NC, USA) for help with editing and reviewing this manuscript.

References 1. Modell SM, Lehmann MH. The long QT syndrome family of cardiac ion channelopathies: a HuGE review. Genet Med 2006;8:143–155. 2. Vatta M, Ackerman MJ, Ye B, et al. Mutant caveolin-3 induces persistent late sodium current and is associated with long-QT syndrome. Circulation 2006;114: 2104 –2112. 3. Medeiros-Domingo A, Kaku T, Tester DJ, et al. SCN4B-encoded sodium channel beta4 subunit in congenital long-QT syndrome. Circulation 2007;116:134 – 142. 4. Splawski I, Shen J, Timothy KW, et al. Spectrum of mutations in long-QT syndrome genes: KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation 2000;102:1178 –1185. 5. Tester DJ, Cronk LB, Carr JL, et al. Allelic dropout in long QT syndrome genetic testing: a possible mechanism underlying false-negative results. Heart Rhythm 2006;3:815– 821. 6. Anderson CL, Delisle BP, Anson BD, et al. Most LQT2 mutations reduce Kv11.1 (hERG) current by a class 2 (trafficking-deficient) mechanism. Circulation 2006;113:365–373. 7. Delisle BP, Anderson CL, Balijepalli RC, et al. Thapsigargin selectively rescues the trafficking defective LQT2 channels G601S and F805C. J Biol Chem 2003;278:35749 –35754. 8. Rossenbacker T, Mubagwa K, Jongbloed RJ, et al. Novel mutation in the Per-Arnt-Sim domain of KCNH2 causes a malignant form of long-QT syndrome. Circulation 2005;111:961–968. 9. Rajamani S, Anderson CL, Anson BD, et al. Pharmacological rescue of human K(⫹) channel long-QT2 mutations: human ether-a-go-go-related gene rescue without block. Circulation 2002;105:2830 –2835. 10. Gong Q, Anderson CL, January CT, et al. Pharmacological rescue of trafficking defective HERG channels formed by coassembly of wild-type and long QT mutant N470D subunits. Am J Physiol Heart Circ Physiol 2004;287:H652– H658.

Heart Rhythm, Vol 6, No 4, April 2009 11. Paulussen A, Yang P, Pangalos M, et al. Analysis of the human KCNH2 (HERG) gene: identification and characterization of a novel mutation Y667X associated with long QT syndrome and a non-pathological 9 bp insertion. Hum Mutat 2000;15:483. 12. Sasano T, Ueda K, Orikabe M, et al. Novel C-terminus frameshift mutation, 1122fs/147, of HERG in LQT2: additional amino acids generated by frameshift cause accelerated inactivation. J Mol Cell Cardiol 2004;37:1205–1211. 13. Wilschanski M, Yahav Y, Yaacov Y, et al. Gentamicin-induced correction of CFTR function in patients with cystic fibrosis and CFTR stop mutations. N Engl J Med 2003;349:1433–1441. 14. Wagner KR, Hamed S, Hadley DW, et al. Gentamicin treatment of Duchenne and Becker muscular dystrophy due to nonsense mutations. Ann Neurol 2001; 49:706 –711. 15. Sangkuhl K, Schulz A, Rompler H, et al. Aminoglycoside-mediated rescue of a disease-causing nonsense mutation in the V2 vasopressin receptor gene in vitro and in vivo. Hum Mol Genet 2004;13:893–903. 16. Burke JF, Mogg AE. Suppression of a nonsense mutation in mammalian cells in vivo by the aminoglycoside antibiotics G-418 and paromomycin. Nucleic Acids Res 1985;13:6265– 6272. 17. Teng S, Ma L, Dong Y, et al. Clinical and electrophysiological characterization of a novel mutation R863X in HERG C-terminus associated with long QT syndrome. J Mol Med 2004;82:189 –196. 18. Kupershmidt S, Snyders DJ, Raes A, et al. A K⫹ channel splice variant common in human heart lacks a C-terminal domain required for expression of rapidly activating delayed rectifier current. J Biol Chem 1998;273:27231–27235. 19. Gong Q, Keeney DR, Robinson JC, et al. Defective assembly and trafficking of mutant HERG channels with C-terminal truncations in long QT syndrome. J Mol Cell Cardiol 2004;37:1225–1233. 20. Jenke M, Sanchez A, Monje F, et al. C-terminal domains implicated in the functional surface expression of potassium channels. EMBO J 2003;22:395– 403. 21. Akhavan A, Atanasiu R, Shrier A. Identification of a COOH-terminal segment involved in maturation and stability of human ether-a-go-go-related gene potassium channels. J Biol Chem 2003;278:40105– 40112. 22. Aydar E, Palmer C. Functional characterization of the C-terminus of the human ether-à-go-go-related gene K(⫹) channel (HERG). J Physiol 2001;534:1–14. 23. Akhavan A, Atanasiu R, Noguchi T, et al. Identification of the cyclic-nucleotidebinding domain as a conserved determinant of ion-channel cell-surface localization. J Cell Sci 2005;118:2803–2812. 24. Bhuiyan ZA, Momenah TS, Gong Q, et al. Recurrent intrauterine fetal loss due to near absence of HERG: clinical and functional characterization of a homozygous nonsense HERG Q1070X mutation. Heart Rhythm 2008;5:553–561. 25. Christe G, Theriault O, Chahine M, et al. A new C-terminal hERG mutation A915fs⫹47X associated with symptomatic LQT2 and auditory-trigger syncope. Heart Rhythm 2008;5:1577–1586. 26. Kupershmidt S, Yang T, Chanthaphaychith S, et al. Defective human ether-ago-go-related gene trafficking linked to an endoplasmic reticulum retention signal in the C terminus. J Biol Chem 2002;277:27442–27448. 27. Gong Q, Zhang L, Vincent GM, et al. Nonsense mutations in hERG cause a decrease in mutant mRNA transcripts by nonsense-mediated mRNA decay in human long-QT syndrome. Circulation 2007;116:17–24. 28. Lai CH, Chun HH, Nahas SA, et al. Correction of ATM gene function by aminoglycoside-induced read-through of premature termination codons. Proc Natl Acad Sci U S A 2004;101:15676 –15681. 29. Schulz A, Sangkuhl K, Lennert T, et al. Aminoglycoside pretreatment partially restores the function of truncated V(2) vasopressin receptors found in patients with nephrogenic diabetes insipidus. J Clin Endocrinol Metab 2002;87: 5247–5257. 30. Manuvakhova M, Keeling K, Bedwell DM. Aminoglycoside antibiotics mediate context-dependent suppression of termination codons in a mammalian translation system. RNA 2000;6:1044 –1055. 31. Olson TM, Alekseev AE, Liu XK, et al. Kv1.5 channelopathy due to KCNA5 loss-of-function mutation causes human atrial fibrillation. Hum Mol Genet 2006;15:2185–2191.