The short QT syndrome (SQTS) is a recently identified cardiac repolarisation disorder that is associated with shortened QT intervals on the ECG (typically to ~320 ms or less (1-3)) and poor rate adaptation of the QT interval. It is also associated with an increased incidence of both atrial and ventricular arrhythmias and of sudden death, in the absence of cardiac structural abnormalities (3, 4). The SQTS is genetically heterogeneous and thus far gain-of-function mutations have been reported to three cardiac potassium (K+) channel genes: KCNH2 (for SQT1), KCNQ1 (for SQT2) and KCNJ2 (for SQT3) (5-9). The first gene mutations to be identified in SQTS patients were to KCNH2 (hERG) (5). In two unrelated families showing a SQTS phenotype different KCNH2 missense mutations were identified that resulted in a common amino-acid substitution (asparagine lysine: N K) at position 588 in the hERG channel protein (5). A subsequent study identified the same mutation in a third SQTS family (6).

The N588 residue is located in the S5-Pore linker region of the hERG channel (alternative nomenclature Kv11.1, encoded by KCNH2 (10)). This region has been shown to be important in the rapid voltage-dependent inactivation gating that normally results in marked inward rectification of the steady-state current-voltage (I-V) relation for wild-type (WT) hERG current (IhERG) and, by extension, of the cardiac native rapid delayed rectifier K+ current, IKr (11-13). Conventional voltage-clamp measurements of N588K IhERG indicate that at both ambient (14) and physiological (15) experimental temperatures the N588K mutation shifts the process of inactivation to more positive potentials (shifting the voltage-dependent inactivation relation by +60 to +100 mV (15, 14, 16)), with the consequence that little rectification of the I-V relation occurs over physiologically relevant membrane potentials. Use of the action potential clamp (AP clamp) technique has shown that during ventricular AP waveforms N588K-hERG channels pass greater current earlier during the AP plateau phase (5, 14, 15). This would be anticipated to lead to increased IKr earlier during ventricular APs and thereby to ventricular AP shortening and hence QTc interval abbreviation, a notion that has been substantiated by ventricular AP simulation studies (9, 17, 18).

The profile of IKr/hERG during AP waveforms from different regions of the heart is known to differ (19). However, despite the fact that SQT1 appears to increase the risk of atrial fibrillation (1, 6, 20), at present there is no information on the effects of the N588K mutation on the profile of IhERG during atrial AP waveforms. The AP clamp technique permits recording of ionic currents during dynamic, physiological waveforms and can therefore take account of membrane potential ‘history’ (21, 22) in a way that is not possible with conventional voltage clamp. Due to the differing configurations of atrial and ventricular APs, the effects of the N588K mutation on IhERG cannot be assumed to be identical during the two AP types and the AP clamp offers a potential means of gaining insight into the effects of the N588K mutation on atrial repolarisation. Moreover, although a shortening of the effective refractory period resulting from accelerated repolarisation is likely to increase the likelihood of re-entrant arrhythmias (3, 4, 23), the overall mechanism(s) by which the N588K mutation increases arrhythmogenic risk are at present incompletely understood. In the case of SQT1, the results of ventricular AP clamp experiments performed at ambient temperature on heterologously expressed WT and N588K-hERG channels have suggested that increased heterogeneity of repolarisation between the APs from the ventricular myocardium and those from Purkinje fibres (14) may contribute to the substrate for ventricular arrhythmogenesis. This possibility arises from observations made at ambient temperature that, under AP clamp, in the case of the Purkinje fibre AP clamp, the current profiles for WT and N588K-hERG were almost identical (14) – most likely because the relatively low plateau phase of the Purkinje AP results in comparatively little inactivation of WT-hERG during the early phases of the AP. Feasibly, a differential effect of the N588K mutation during ventricular and Purkinje fibre APs could account for pronounced U waves observed on the ECG of some SQT patients (5, 14, 24). However, due to complex effects of temperature on IhERG kinetics, data obtained at ambient (room) temperature cannot readily be extrapolated to the situation at 37°C (25), so the relevance of this observation to the situation at mammalian body temperature remains to be verified.

Another possible way that the N588K-hERG mutation might affect susceptibility to arrhythmogenesis is by modifying the ability of WT-hERG/IKr to provide cardiac tissue with protection from premature beats: premature stimuli applied in AP clamp experiments have been shown to activate outward transient WT IhERG currents and these could protect native cardiac tissue against premature depolarisations (19). At present it is not known if the N588K mutation influences the response of IKr/IhERG to premature stimuli. Accordingly, the aims of this study were: (i) to characterise and compare at 37 °C the effects of the N588K mutation on IhERG during different cardiac APs (atrial, ventricular, Purkinje fibre), complementing these with in silico AP clamp simulations, based on experimental data and the known properties of WT and N588K IhERG (14-16) and (ii) to determine the effects of the N588K mutation on the pattern of IhERG observed during premature stimuli following AP commands.


METHODS
Maintenance of cell lines stably expressing wild-type hERG and N588K-hERG channels

Measurements of wild-type (WT) hERG and N588K-hERG current (IhERG) were made from Chinese Hamster Ovary (CHO) cells stably expressing either WT or N588K-hERG. The N588K mutation was engineered into the full length hERG transcript (which had been subcloned into a modified pcDNA3.0 vector at the HindIII and EcoRI restriction enzyme sites) using the Quikchange II XL site-directed mutagenesis kit (Stratagene) as previously described (15). The cells were passaged using enzyme-free cell dissociation solution (Chemicon International) and were plated onto small glass coverslips in 30 mm Petri dishes containing Kaighn’s modification of Ham’s F12-K medium (Gibco), supplemented with 10% Fetal bovine serum (Gibco) and 200 µg ml-1 gentamycin (Gibco). Cells were incubated at 37 °C for a minimum of 24 hours prior to any electrophysiological study.

Co-expression of MiRP1 with hERG

MinK Related Peptide 1 (MiRP1, encoded by the gene KCNE2) is a putative accessory sub-unit that has been suggested to associate with hERG to recapitulate channels that mediate native IKr (26, 27). A contribution of MiRP1 to native IKr is debatable however (28, 29), and it has been suggested that MiRP1 is unlikely to associate significantly with hERG in areas of the heart outside of the conduction system (30, 31). Nevertheless, for the sake of completeness, some AP clamp experiments with hERG and MiPR1 co-expression were performed in the present study. For these, a similar approach was taken to that used in (28, 32, 33), using the same transfection protocol we have described previously (15). Briefly, cells stably expressing WT or N588K-hERG were plated onto glass coverslips. 24 hours later the cells were incubated with 2 µg of MiRP1 (in pCI-neo, kindly donated by Dr S. Goldstein) and 0.5 µg green fluorescent protein (in pCMX; kindly donated by Dr Jeremy Tavare) using FuGene 6 transfection reagent (Roche diagnostics) and Optimem-1 medium (Gibco). After an incubation period of 5-6 hours the Optimem-1 medium was replaced with fresh F-12K medium. The cells were incubated at 37 °C for a minimum of 16 hours before any electrophysiological studies were performed (15). Co-expression of MiRP1 with hERG results in a left-ward shift in the voltage dependence of activation (28, 32) and this was used a marker of functional MiRP1 expression in cells demonstrating expression of GFP. Use of a series of conventional depolarising voltage commands from a holding membrane potential of -80 mV to a range of more positive membrane potentials (see (15)), allowed half-maximal activation voltage (V0.5) values to be assessed. For both WT and N588K-hERG expressing cell lines, MiRP1-transfected cells showed V0.5 values that were ~ -10 mV shifted compared to hERG alone, concordant with previous reports (28, 32).

Electrophysiological recordings

Glass coverslips onto which cells had been plated were placed in a 0.5 ml bath mounted on an inverted microscope (Nikon Diaphot) and continuously superfused (at 37°C) with Tyrode’s solution containing (in mM): 140 NaCl, 4 KCl, 2.5 CaCl2, 1 MgCl2, 10 Glucose and 5 HEPES (titrated to pH 7.45 with NaOH) (15, 34, 35). Patch pipettes (Corning 7052 glass, AM Systems) were pulled (Narishige PP830) and fire polished (Narishige MF83) to give final resistances of 1.5-3.5 M. The pipette dialysis solution contained (in mM): 130 KCl, 1 MgCl2, 5 EGTA, 5 MgATP and 10 HEPES (titrated to pH 7.2 with KOH) (15, 34, 35). Whole-cell patch clamp recordings were made using an Axopatch 1D amplifier (Axon instruments) and a CV-4 1/100 headstage. Typically, 80% of the pipette series resistance could be compensated. Data were recorded using a Digidata 1200B interface (Axon instruments) and stored on a Viglen computer. The data were sampled at 10 kHz, using a filter bandwidth of 2 kHz on the amplifier.

Action potential clamp waveforms

Ventricular and atrial action potentials (AP’s) used as voltage-clamp commands were generated using the ten Tusscher et al. human ventricular cell model and Nygren et al. human atrial cell model (36, 37). The ventricular AP waveform used was a human epicardial AP with a duration at 50 and 90% of repolarisation (APD50 and APD90) of 270 and 319 ms, respectively. The atrial AP waveform had an APD50 and APD90 of 42 and 266 ms. In the absence of a viable human Purkinje fibre model, the Purkinje fibre AP waveform used was the same canine Purkinje cell AP used previously by Cordeiro and colleagues (14) with APD50 and APD90 values of 170 and 360 ms respectively. N588K-hERG exhibits a modest alteration to Na:K permeability and a positive shifted IhERG reversal potential (14, 15). Conductance-voltage (G-V) relations for N588K-hERG negative to ~ -70 mV exhibited high levels of noise and therefore G-V plots for N588K-hERG shown in Figs. 1-3 were cut-off negative to ~-70 mV (15). For experiments involving paired ‘premature’ stimulation (19), a ventricular AP command protocol was applied in which a premature AP command waveform was superimposed 100 ms before the APD90 of an initial AP command. For subsequent applications of the protocol the premature AP command waveform was applied in 10 ms increments, until it was applied 190 ms after the APD90 of the initial AP command. A similar approach was adopted for experiments in which ‘premature’ stimuli were applied following an initial atrial or Purkinje fibre AP command waveform.

For some ventricular AP clamp experiments an abbreviated ventricular AP command was used, in order to approximate accelerated ventricular repolarisation in the SQTS. Rate corrected QT (QTc) intervals as short as 248 ms have been reported (38, 39), compared to normal QTc intervals of ~450 ms or less (40), representing QTc-shortening of up to ~45%. Accordingly, the ventricular AP command was abbreviated (manually, using Microsoft Excel), to give APD50 and APD90 values of 138 ms and 171 ms respectively, representing a ~46% shortening of APD90. In contrast to the situation regarding ventricular repolarisation, no clinical or animal model data are currently available regarding the occurrence or extent of atrial or Purkinje fibre AP shortening in SQT1. Therefore abbreviated AP commands were not incorporated into paired AP stimulation experiments using these AP commands. Protocols were converted for use in Clampex 8. Start-to-start intervals of 3 seconds between successive applications of protocols were used, in order to facilitate deactivation between successive commands. All current traces were corrected for leak-current online, using an interspersed P/4 protocol (15, 21).

Simulations

As the AP clamp approach takes into account of membrane potential ‘history’ (21, 22) in a way that is not possible with conventional voltage clamp, AP clamp computer simulations were also performed to gain additional insight into the relative impact of the N588K mutation on IhERG during the different AP waveforms. For these, ventricular, atrial, and Purkinje fibre waveforms were used as voltage commands in silico AP clamp simulations that were based on our data and the known properties of WT and N588K IhERG (14-16).

WT IhERG was best described by a formulation used previously by Noble and colleagues (22) to incorporate mono-exponential activation and bi-exponential deactivation. The model equations for wild type IhERG channel are:



Where: gKr1 and gKr2 are the conductance values of the components that combine to produce net IhERG, with their states being described by xr1 and xr2 respectively (22, 41); r is a rectification factor that reproduces voltage-dependent rectification of IhERG that arises due to current inactivation; V is the membrane potential and EK is the potassium equilibrium potential. Details of parameter values are listed in (22).

As the original Noble et al. (22) equations were developed for guinea-pig ventricular myocytes, in the present study, some of the parameters of the equations were adjusted to recapitulate human ventricular IKr/IhERG, as presented by ten Tusscher et al. (36). The resultant gKr1 and gKr2 are 0.018 and 0.011 nS/pF respectively and the equation for xr2 becomes:



We have previously shown that, at 37°C, the N588K mutation produces a profound positive shift in voltage-dependent inactivation kinetics of hERG (15, 16), without a concomitant alteration to voltage-dependent activation (15). We have also shown that, over the range of physiologically relevant voltages, electrophysiological changes in SQT1 are reasonably well recapitulated by removing voltage-dependent inactivation from IKr/IhERG (17). Consequently, in this study changes in IhERG due to the N588K mutation (the ‘SQT1’ condition) were simulated by setting ‘r’ to 1, without any accompanying alteration to the voltage-dependence of IhERG activation.

AP clamp simulations were performed with the same AP commands to those used for in vitro experiments, without changes to gKr1 and gKr2 (therefore using a constant baseline current density). As a consequence of this, any observed differences in amplitude and time-course between WT IhERG (control) elicited by ventricular, atrial, and Purkinje fibre AP commands would be attributable solely to differences between the voltage command waveforms. Similarly, any observed differences in elicited currents between WT and SQT1 simulations with a given command profile would be attributable solely to the loss of IhERG voltage-dependent inactivation incorporated into the SQT1 current formulation.

Data analysis and presentation

Data were analysed using Clampfit 8 (Axon Instruments), Excel 2002 and Prism v3 (Graphpad Inc) software. The data are presented as the mean ± standard error of the mean. Statistical analysis was carried out using a Student’s t-test or a two-way analysis of variance (ANOVA) with Bonferroni post-test using Prism v3 (Graphpad Inc). P values of less than 0.05 were taken as being statistically significant.


RESULTS
IhERG during human ventricular AP clamp

For comparison with current profiles during AP voltage commands, Figures 1Ai and Aii show representative WT and N588K-hERG current (IhERG) records elicited by a conventional voltage-clamp protocol (comprised of a 2s square command from -80 mV to +20 mV, followed by a repolarising step to -40 mV to observe IhERG tails; cf (16, 35)). The current shown in Fig. 1Ai displays characteristic properties of WT IhERG – with a large, progressively deactivating current tail following repolarisation to -40 mV at the end of the +20 mV step (16, 35). Fig.1Aii shows that, for N588K IhERG the current during the +20 mV step was much larger than the subsequent tail current on repolarisation to -40 mV. This is consistent with the known attenuated-inactivation properties of N588K-hERG and is similar to previously published profiles of N588K IhERG during conventional voltage-clamp protocols at 37°C (15, 16, 35). Figs. 1B and 1C show representative data from ventricular AP clamp experiments. Figs. 1Bi and Bii show representative WT and N588K-hERG current (IhERG) records superimposed on the applied ventricular AP command waveform. WT IhERG exhibited a small increase in current during the upstroke of the AP, which peaked during the repolarisation phase. In contrast, there was a marked rise in outward current for N588K IhERG following depolarisation, due to previously reported attenuated inactivation (5, 15, 14), resulting in a dome shaped current (Fig. 1Bii). These findings are consistent with current profiles at 37°C observed previously using a guinea-pig ventricular AP command (15). Plots of the instantaneous I-V relationship for WT and N588K-hERG during the repolarisation are shown in Figs. 1Ci and Cii (traces from each of 7 experiments were normalised to peak current and are shown overlaid). WT and N588K IhERG during repolarisation peaked at -37.6 ± 1.7 and +24.6 ± 0.6 mV respectively, representing a +62 mV shift in peak repolarising current (n = 7 cells for each; P < 0.001, unpaired t test). To take into account the changes in driving force during the AP, the instantaneous conductance-voltage (G-V) relationships were plotted as shown in Figs. 1Di and Dii (15, 19, 21). For WT-hERG, conductance increased as the membrane potential followed the course of repolarisation (Fig. 1Di (21)). The G-V profile for N588K-hERG was consistent with that observed previously during a guinea pig AP command (15), with conductance rising quickly, reaching a maximum between +20 and +40 mV, and remaining relatively constant during the remainder of the repolarisation (Fig. 1Dii, (15)). Whilst the role in native IKr of the putative accessory subunit (MiRP1) is a matter of debate (26, 28), and MiRP1 may be unlikely to play a significant role in channels mediating IKr outside of the conduction system (31), for the sake of completeness experiments similar to those shown in Fig. 1 were also performed using hERG + MiRP1 co-expression. The IhERG profiles during the ventricular AP waveform for WT and N588K-hERG in the presence of MiRP1 were similar to those seen for hERG alone (not shown), with peak repolarising current occurring at -42.8 ± 1.2 mV and +24.3 ± 0.6 mV respectively for WT and N588K IhERG (a +67 mV shift in peak repolarising current; P < 0.001; n = 9 and 8 cells for WT and N588K-hERG respectively).

Fig. 1. Profile of IhERG during rectangular and ventricular AP voltage commands A. Upper traces show example current profiles of WT (Ai) and N588K IhERG (Aii) elicited by conventional rectangular voltage command protocol shown in lower traces. The voltage-step to +20 mV to activate IhERG was preceded by a brief (50 ms) step from -80 mV to -40 mV to monitor instantaneous leak/endogenous current, which was minimal in both Ai and Aii. B. Upper traces show example current profiles of WT (Bi) and N588K IhERG (Bii) elicited by the ventricular AP voltage command waveform (shown overlaid). C. Instantaneous I-V relationships overlaid for 7 cells each for WT (Ci) and N588K-hERG (Cii). D. Representative instantaneous G-V plots for WT (Di) and N588K-hERG (Dii). Arrows in C and D denote direction of membrane potential repolarisation during the AP command.

IhERG profile during atrial and Purkinje fibre AP commands

Some patients with SQT1 have been shown to have atrial fibrillation associated with shortened effective refractory periods (3), which is highly suggestive of abbreviated atrial repolarisation. Yet, to date there has been no investigation as to the effects of the N588K mutation on the profile of IhERG during atrial APs. To address this issue, we applied atrial AP command waveforms to cells expressing WT or N588K-hERG and recorded the resultant profile of IhERG, shown in Figs. 2Ai and Aii. WT IhERG showed a dome shaped profile (Fig. 2Ai) as reported previously (19). The current profile for N588K-hERG also displayed a dome shaped profile during the atrial AP, with the current peaking slightly earlier than WT IhERG, and IhERG decreasing earlier (Fig. 2Aii). Plots of the normalised instantaneous I-V relationship during atrial repolarisation for WT and N588K-hERG are shown in Figs. 2Bi and Bii (traces from 5 and 7 experiments normalised to peak current are shown overlain for WT and N588K-hERG respectively). The maximum IhERG during repolarisation was shifted by ~+15 mV: with maximal WT IhERG occurring at -42.0 ± 1.3 mV (n = 5 cells), and that for N588K-hERG occurring at -26.8 ± 0.7 mV (n = 7 cells; P < 0.001). Example traces of the instantaneous G-V relationship for WT and N588K-hERG during atrial repolarisation are shown in Figs. 2Ci and Cii. WT-hERG conductance increased as the cell membrane followed the course of repolarisation, reaching a maximum ~-60 mV (Fig. 2Ci). The conductance of N588K-hERG increased earlier than WT-hERG, reaching a maximum between -20 and -40 mV, then levelled out during the remainder of repolarisation (Fig. 2Cii). IhERG profiles for WT and N588K-hERG in the presence of MiRP1 co-expression (not shown) were similar to those for hERG alone, with maximal repolarising current for WT-hERG + MiRP1 occurring at -44.7 ± 1.3 mV and for N588K-hERG + MiRP1 occurring at -28.5 ± 0.8 mV (n = 5 for each; representing a +16 mV shift in maximal repolarising current; P < 0.001).

Fig. 2. Profile of IhERG during an atrial command A. Example current profiles of WT (Ai) and N588K IhERG (Aii) overlaid on atrial AP voltage command waveform. B. Instantaneous I-V relationships overlaid for 5 and 7 cells each for WT (Bi) and N588K-hERG (Bii). C. Representative instantaneous G-V plots for WT (Ci) and N588K-hERG (Cii). Arrows in B and C denote direction of membrane potential repolarisation during the AP command.

Fig. 3 shows representative current traces of WT (Fig. 3Ai) and N588K-hERG (Fig. 3Aii) recorded during a PF AP command waveform. WT IhERG increased progressively, reaching a maximum during phase 3 of the AP. In contrast, N588K IhERG increased earlier following the AP upstroke and reached a maximum during phase 2 of the AP, with the subsequent decline in current occurring earlier than for WT-hERG. Figs. 3Bi and Bii show plots of the instantaneous I-V relationship for WT and N588K-hERG respectively (traces of 8 and 10 experiments normalised to the peak current are shown overlain for WT and N588K respectively). WT IhERG reached a maximum at -45.6 ± 1.6 mV (n = 8 cells), while N588K IhERG reached a maximum much earlier at -14.8 ± 0.3 mV (n = 10 cells), representing a ~+30 mV shift in peak repolarising current (P < 0.001). Example traces of the instantaneous G-V relationships are shown in Figs. 3Ci and Cii for WT and N588K-hERG respectively. WT-hERG conductance increased as the membrane potential followed the course of repolarisation, reaching a maximum during the end stages of repolarisation (Fig. 3Ci). In the case of N588K-hERG, the conductance increased rapidly just after the peak of the AP, reaching a maximum between -20 and 0 mV, then remained relatively constant as the membrane potential followed the course of repolarisation (Fig. 3Cii). In additional experiments with MiRP1 co-expression, WT-hERG + MiRP1 IhERG peaked at -43.2 ± 0.8 mV (n = 8 cells) and for N588K-hERG + MiRP1 IhERG peaked at -13.0 ± 0.3 mV (n = 8 cells; again a ~+30 mV shift in peak repolarising current; P < 0.001).

Fig. 3. Profile of IhERG during a Purkinje fibre command A. Example current profiles of WT (Ai) and N588K-hERG (Aii) overlaid on Purkinje fibre AP voltage command waveform. B. Instantaneous I-V relationships overlaid for 8 and 10 cells each for WT (Bi) and N588K-hERG (Bii). C. Representative instantaneous G-V plots for WT (Ci) and N588K-hERG (Cii). Arrows in B and C denote direction of membrane potential repolarisation during the AP command.

Relative effects of the N588K mutation on peak repolarising current during different AP types

Figs. 1-3 show that the N588K mutation affected the overall profile of IhERG during each of the AP command waveforms used, though to differing extents. This was examined further by comparing peak repolarising current density for the differing AP waveforms, for both WT and N588K-hERG. Comparisons were made between AP waveforms for each of WT and N588K rather than between the two channel types, since it could not be ruled out that differences in current magnitude between WT and mutant channels could reflect (wholly or in part) different functional expression levels of the WT and mutant channel cell lines. Comparisons of current during the different APs were made in both the absence and presence of MiRP1.

Fig. 4A shows that maximal WT IhERG density during the ventricular AP command was significantly greater (by 2.9-fold and 1.7-fold respectively) than during atrial and Purkinje AP command waveforms. Fig. 4B shows corresponding data for N588K-hERG, with differences between peak IhERG density during the ventricular AP and that during atrial and Purkinje fibre APs being augmented (to 6.6-fold and 2.6 fold respectively). Figs. 4C and 4D show similar data for WT and N588K-hERG in the presence of MiRP1. Differences in peak IhERG density between the different APs showed similar trends to those seen in the absence of MiRP1 for both WT and N588K-hERG, although the 1.2-fold difference in WT peak IhERG density between ventricular and Purkinje fibre APs did not attain significance, compared to the 3.1-fold difference between current during ventricular and atrial APs. For N588K IhERG, relative differences in peak current density (ventricular versus atrial of 6.7 fold and ventricular versus Purkinje of 2.6 fold) were similar to those observed in the absence of MiRP1. The most noteworthy observation here relates to the comparison between ventricular and Purkinje fibre APs – since both of these cell types contribute to overall repolarisation of the ventricles – the results of which indicate that ratios in peak repolarising IhERG of ventricular versus Purkinje fibre APs were augmented for N588K hERG.

Fig. 4. Current density plots during AP commands A. Maximal WT IhERG density during ventricular, atrial and Purkinje fibre AP command waveforms (n = 7, 5 and 8 cells respectively). B. Maximal N588K IhERG density during ventricular, atrial and Purkinje fibre AP command waveforms ( n = 7, 7 and 10 cells respectively). C. Maximal WT + MiRP1 IhERG density during ventricular, atrial and Purkinje fibre AP command waveforms (n = 9, 5 and 8 cells respectively). D. Maximal N588K + MiRP1 IhERG during ventricular, atrial and Purkinje fibre AP command waveforms (n = 8, 5 and 8 cells respectively). In ‘A’-‘D’ * denotes statistical significance of P < 0.05; # denotes statistical significance of P < 0.01; § denotes statistical significance of P < 0.001.

To enable comparison between currents elicited by the different AP waveforms for WT and mutant conditions, we performed in silico AP clamp simulations, using the same AP voltage commands as used to obtain the in vitro experimental data described in Figs. 1-4. The results of our AP clamp simulations are shown in Fig. 5. Figs. 5Ai, Bi and Ci respectively show the simulated profiles of WT (‘Control’) IhERG during ventricular, atrial and Purkinje fibre AP commands, whilst Figs. 5Aii, Bii and Cii show comparable currents under simulated SQT1 conditions. For each AP waveform, the simulated current profiles were similar to those observed experimentally. Additionally, one advantage of the simulations is that the consistent use of similar baseline conductance values made it possible to compare directly the resulting current amplitudes between SQT1 and WT IhERG for each AP profile. Thus, Figs. 5Ai and 5Aii show both an alteration to the overall profile of IhERG during the ventricular AP command in SQT1 (similar to that shown in Fig. 1) and also a marked increase in peak current amplitude (the SQT1 peak current was 4.87-fold that in control). As predicted from the in vitro current measurements, simulated WT IhERG during the atrial AP command (Fig. 5Bi) was markedly smaller than that during the ventricular AP command. Nevertheless, for the atrial AP command the peak SQT1 IhERG amplitude was increased - to 2.17-fold the WT IhERG amplitude. For the Purkinje fibre AP, peak WT IhERG during the AP command (Fig. 5Ci) was somewhat reduced compared to that during the ventricular AP, whilst in simulated SQT1 conditions, peak current was 2.87-fold that in the control condition. Collectively, the results of our AP clamp simulations are consistent with the trend indicated by the in vitro AP clamp experiments: under SQT1 conditions repolarising IhERG was increased during each AP type studied, with the greatest effect seen for the ventricular AP waveform.

Fig. 5. AP clamp experiment simulations. A. Shows simulated wild-type (‘Control’, Ai) and SQT1 mutation condition currents (‘SQT1’, Aii) elicited by the overlaid ventricular AP command waveform. B. Shows simulated wild-type (‘Control’, Bi) and SQT1 (‘SQT1’, Bii) currents elicited by the overlaid atrial AP command waveform. C. Shows simulated wild-type (‘Control’, Ci) and SQT1 (‘SQT1’, Cii) currents elicited by the overlaid Purkinje fibre AP command waveform.

Effect of premature pulses following the ventricular action potentials on WT and N588K-IhERG

It has been suggested previously (19) that the hERG/IKr channel may play a protective role in countering the effects of premature membrane depolarisation. This is because, due to comparatively rapid recovery from inactivation and slow deactivation, a proportion of hERG channels are able to pass rapid, transient outward currents in response to premature stimuli, even following complete AP repolarisation (19). At present, the effects of the N588K mutation on the response of hERG to premature stimuli are not known. We investigated this by using paired ventricular AP commands under AP clamp (see Methods), as performed previously for WT-hERG by Lu et al (19). Figs. 6Ai and Aii show representative traces of WT and N588K IhERG during paired AP clamp commands (for clarity only 11 of the 30 paired waveforms are shown), with the voltage protocol shown as an inset. Considering WT-hERG, it can be seen that at the upstroke of the premature AP command a rapid outward current transient occurred; this was followed by a sustained current which was similar in profile to the current seen during the first AP command (Fig. 6Ai, see also (19)). The outward current transients for WT-hERG (which were normalised to the maximal outward current transient recorded (19, 32)) increased in amplitude the later the premature AP command waveform occurred, reaching a maximum 30 ms after the APD90, before decreasing at longer time periods (Fig. 6Bi; n = 7 cells), presumably due to progressive effects of IhERG deactivation, and showing a profile similar to that reported previously (19). Fig. 6Aii shows representative current traces for N588K-hERG during the same protocol, with mean normalised data (n = 7 cells) plotted in Fig. 6Bi, for direct comparison with the corresponding data for WT-hERG. The profile of N588K-hERG outward current transients differed from that of WT-hERG, with normalised outward current transient amplitude being maximal early during the protocol (up to 10 ms after the APD90, labelled ‘t1’), but decreasing in amplitude more extensively than WT-hERG between 30-70 ms after the APD90 (Fig. 6Bi, labelled ‘t2’). Fig. 6Bii shows similar data for WT-hERG (n = 7 cells) and N588K-hERG (n = 6 cells) during the abbreviated ventricular AP command waveform (see Methods). The time-course of peak outward current transients for WT-hERG was similar to that for the longer AP command waveform, with the peak current occurring 10 ms after the APD90 (Fig. 6Bii). N588K-hERG outward current transients increased slightly prior to the APD90, reaching a maximum 20 ms before the APD90 (Fig. 6Bii, labelled ‘t3’). As in the case for the longer AP command waveform (Fig. 6Bi), there was a time-period (from 20-120 ms after APD90 of the first AP command, labelled ‘t4’) during which N588K-hERG normalised outward current transient amplitudes were smaller than those of WT-hERG (Fig. 6Bii).

Fig. 6. Profile of WT and N588K IhERG during premature ventricular AP commands A. Representative current traces of WT (Aii) and N588K-hERG (Aii) during paired ventricular AP command waveforms elicited by the protocol shown in the figure inset. B. Plots of WT and N588K-hERG normalised outward current transients during paired ventricular AP command waveforms (Bi; n = 7 cells each) and shortened AP command waveforms (Bii; n = 7 and 6 for WT and N588K-hERG respectively). C. Comparison of normalised outward current transients of WT-hERG during a longer duration paired ventricular AP command waveform (n = 7 cells), against outward current transients for N588K-hERG during a shortened paired ventricular AP command waveform (n = 6 cells), plotted against interpulse interval (Ci). Comparison of WT-hERG normalised outward current transients during longer and shorter duration paired ventricular AP command waveforms (Cii; n = 7 cells). In ‘B’ and ‘C’ ‘*’ denotes statistical significance of P < 0.05; # denotes statistical significance of P < 0.01; § denotes statistical significance of P < 0.001.

In order to approximate a comparison between a ‘normal’ response of hERG to premature stimulation and the response of N588K-hERG in the setting of SQT1 (one that would be associated with abbreviated APs) we compared the profile of outward current transients for WT-hERG shown in Fig. 6Bi with those for N588K-hERG for the abbreviated AP command shown in Fig. 6Bii: for this the respective normalised plots against time of outward current transient amplitude were overlaid Fig. 6Ci). It can be seen that in the period starting from 100 ms before the APD90 to 10 ms after the APD90 (labelled ‘t5’), N588K-hERG exhibited greater outward current transients than did WT-hERG (which might be anticipated to result in an increased protective role of N588K-hERG in suppressing effect of premature depolarising stimuli during this time period; P < 0.001, two-way ANOVA). However in the period from 20-120 ms after the APD90 (labelled ‘t6’), the normalised N588K-hERG outward current transients were much smaller in amplitude than those of WT-hERG (Fig. 6Ci; P < 0.05 two-way ANOVA). For purposes of comparison Fig. 6Cii shows normalised plots of WT-hERG outward current transients during the ‘normal’ and abbreviated (‘short’) AP command waveforms. As might be anticipated, the relationships for the two AP waveforms were similar in profile, with the maximal outward current transient occurring earlier (labelled ‘t7’) and the decline in current transient amplitude commencing earlier for the shorter AP command. Consequently, there was a period (labelled ‘t8’), over which normalised outward current transients for WT-hERG were significantly smaller in amplitude following the abbreviated AP compared to the ‘normal’ AP command. Collectively, these observations raise the possibility that at some time-points after the APD90, accelerated ventricular AP repolarisation may be associated with altered ability of hERG to counter potentially arrhythmogenic premature membrane depolarisation.

Premature stimulation experiments were also performed using paired atrial (Fig. 7A) and Purkinje fibre (Fig. 7B) AP commands. Example current traces for WT and N588K-hERG during paired atrial AP command waveforms are shown in Figs. 7Ai and 7Aii respectively. It can be seen that for WT-hERG, the outward current transients initially increased with increasing delay between the first and premature AP command waveforms, before decreasing at longer intervals (presumably due to deactivation of IhERG, (19)). For N588K-hERG, the outward current transients were maximal for the first paired AP command waveforms and started to decrease in amplitude the later the premature AP command waveform was applied. The outward current transients for WT and N588K-hERG were normalised to the maximal outward current transient observed in each cell, pooled and then plotted against the time at which the premature AP command waveform was applied (Fig. 7Aiii). It can be seen that at almost all of the time periods investigated, the normalised outward current transients for N588K-hERG (n = 7 cells) were smaller in amplitude than for WT-hERG (n = 6 cells). For paired Purkinje fibre AP commands, it can be seen that for WT-hERG, as the delay between the first and premature AP command waveforms increased, the outward current transients increased in amplitude before subsequently decreasing (presumably due to IhERG deactivation see (19)). For N588K-hERG, the outward current transients only increased in amplitude very slightly before subsequently decreasing in amplitude as the delay between the first and premature AP command waveform increased. Plots of the normalised outward current transients against the time at which the premature AP command waveform was applied are shown in Fig. 7Biii. Prior to APD90 (labelled ‘a’ in Fig. 7Biii), normalised outward current transients for N588K- hERG (n = 6 cells) were larger in amplitude compared to WT-hERG (n = 7 cells; P < 0.05, two-way ANOVA). However, when the premature AP command waveform was applied after the APD90 (labelled ‘b’ in Fig. 7Biii), the normalised outward current transients for N588K-hERG were smaller in amplitude than for WT-hERG between 10 ms after the APD90 to 120 ms after the APD90 (Fig. 7Biii, P < 0.05, two-way ANOVA). Collectively, the results shown in Figure 7 raise the possibility that the N588K mutation may also be associated with an altered ability of hERG to counter potentially arrhythmogenic premature membrane depolarisations following atrial or Purkinje fibre waveforms.

Fig. 7. Profile of WT and N588K IhERG during premature atrial and Purkinje fibre AP commands A. Representative current traces of WT (Ai) and N588K-hERG (Aii) during paired atrial AP command waveforms elicited by the protocol shown in the figure inset. (Aiii) shows plots of normalised outward current transients during paired atrial AP commands waveforms (n=6 WT; n=7 N588K). * denotes statistical significance of P < 0.05; # denotes statistical significance of P < 0.01. B. Representative current traces of WT (Bi) and N588K-hERG (Bii) during paired Purkinje fibre AP command waveforms elicited by the protocol shown in the figure inset. (Biii) shows plots of normalised outward current transients during paired atrial AP commands waveforms (n = 7 and 6 cells for WT and N588K-hERG respectively). * denotes statistical significance of P < 0.05; # denotes statistical significance of P < 0.01; § denotes statistical significance of P < 0.001. Points ‘A’ and ‘B’ represent time periods during which normalised outward current transients for N588K-hERG were significantly larger ‘a’ and smaller ‘b’ than for WT-hERG.

For comparison with the results presented in Figs. 6 and 7, we also compared the time-course of outward IhERG transients using conventional ‘square pulse’ commands (protocol shown as an inset to Fig. 8A). The membrane potential was depolarised to +40 mV (a potential within the ventricular AP overshoot/plateau range) for 500 ms. It was then repolarised to -40 mV for increasing periods of time (between 2 and 20 ms). This repolarisation potential was chosen as WT IhERG is maximal at ~-40 mV during the different AP commands (whereas the N588K mutation altered the timing of peak repolarising current during different AP commands to different extents). Membrane potential was then again depolarised to +40 mV for 100 ms. Representative current traces during this protocol for WT IhERG and N588K IhERG are shown in Fig. 8Ai and 8Aii respectively (dotted line indicates zero current; indicating that, as anticipated from the known voltage-dependent inactivation kinetics for N588K IhERG (15, 14), greater inactivation of IhERG occurred during the initial +40 mV step for WT-hERG). For each of WT and N588K-hERG (5 and 7 cells respectively), the time-course of recovery of outward transient amplitude was determined by normalising peak outward current amplitude during the third step to that for the maximal outward current transient amplitude observed. The resulting data were plotted against repolarising step duration and fitted with a one phase exponential association as shown in Fig. 8B. The relation for N588K IhERG rose more rapidly than that for WT IhERG with respective time-constant (react) values of 1.8 (95% CI of 1.6-2.1 ms) and 0.8 ms (95% CI of 0.65-0.91 ms) respectively (P < 0.001, unpaired t-test).

Fig. 8. WT and N588K IhERG transients during a square pulse protocol A. Example current profiles of WT (Ai) and N588K IhERG (Aii) elicited by the voltage protocol shown in the figure inset. B. Plots of recovery of WT and N588K IhERG transient amplitude against time. For each cell studied, the peak currents during the third pulse were normalised to the maximum currents observed during the third step. Data were fitted with a one phase exponential association giving react values of 1.8 and 0.8 ms for WT and N588K IhERG respectively (n = 5 and 7 cells respectively). *** denotes statistical significance of P < 0.001.

DISCUSSION
Novelty and principal findings of this study

Effects of the N588K-hERG mutation on IhERG kinetics during conventional rectangular voltage commands have been studied previously at both ambient and physiological temperatures (5, 14, 15). However, although some data on the effect of this gain-of-function mutation on current during ventricular (5, 15, 14), and Purkinje (14) AP waveforms have been obtained previously: (i) ours is the first study of N588K hERG to utilise human AP command waveforms and does so at a physiological recording temperature (contrasting with (5, 14)); (ii) it is the first to report the effect of the N588K mutation on the profile of IhERG during atrial APs; (iii) it is the first to compare between ventricular and Purkinje fibre AP commands the effects of the SQT1 mutation on IhERG at physiological temperature; (iv) it is also the first to use a mathematical simulation approach to make parallel comparisons of IhERG elicited by the three waveforms in the setting of SQTS. The principal findings of this study are that at 37°C: (i) in addition to its effects on IhERG profile during ventricular AP commands, the N588K mutation also alters the IhERG profile during atrial and Purkinje AP waveforms, though does so to differing extents for the differing AP configurations; and (ii) the profile of outward IhERG transients elicited by premature depolarising stimuli is altered for N588K-hERG. Several aspects of our findings merit discussion.

Effects of N588K on the profile of IhERG during ventricular APs

Although extensive IhERG activation can occur at ventricular AP plateau voltages (21), rapid voltage-dependent inactivation of WT IhERG normally limits current size early following the initial ventricular AP overshoot (21, 41). Rapid recovery from inactivation as the AP voltage subsequently declines means that the voltage-dependence of WT IhERG during ventricular AP repolarisation is similar to that of the fully-activated current-voltage (I-V) relation, as the membrane potential scans the ‘open channel’ I-V relation (21, 42). Thus as long as there is sufficient time for IhERG to activate at plateau voltages, similar WT IhERG voltage-dependencies might be anticipated during ventricular AP commands taken from different species with ventricular APs possessing a high plateau phase (for example, see (21) for data obtained using both guinea-pig and rabbit AP commands). Accordingly, WT IhERG during human ventricular AP commands in the present study had a very similar voltage-dependence to that observed at 37°C previously with native or simulated guinea-pig ventricular AP commands (15, 19, 21).

IhERG activation is able to occur as normal for N588K-hERG, but due to the relative lack of channel inactivation the current profile peaks much earlier during the AP and then follows the course of membrane repolarisation (15, 14). The effects of the N588K mutation on the voltage-dependence of IhERG during ventricular APs seen here are quantitatively similar to those observed using a guinea-pig ventricular AP command at 37°C (15) and qualitatively similar to those seen at ambient temperature using canine AP commands (5, 14). Although our ventricular AP clamp data are suggestive of a significant increase to IhERG amplitude as well as an altered time-course during ventricular APs for N588K-hERG, due to potential differences in functional channel expression in WT and N588K-hERG expressing cell lines, caution is required in making any direct quantitative comparison between WT and N588K current magnitudes from these experiments. However, the simulations in Figs. 5Ai and Aii utilised similar baseline conductance values to one another and so are not subject to this limitation. These data indicate that, solely as a result of loss of voltage-dependent inactivation/rectification, the maximal outward IhERG during the AP command in the SQT1 condition approached five-fold that during the control (WT IhERG) condition. Thus, the N588K mutation can be expected to greatly increase the magnitude and contribution of IhERG/IKr to ventricular AP repolarisation. Additionally, as indicated by the G-V plots in Fig. 1 of this study (see also (15)), N588K-hERG contributes outward conductance more consistently throughout ventricular AP repolarisation than does WT-hERG. The net result of these changes to IKr in SQT1 is abbreviated ventricular repolarisation and associated shortening of the QT interval (2, 5, 6).

Effects of N588K on the profile of IhERG during Purkinje fibre and atrial APs

A recent study by Gaborit and colleagues (43) of relative expression of a range of ion channel mRNAs in different regions of non-diseased human heart has reported hERG to be the most highly expressed K+ channel in ventricle and Purkinje fibres and the second most highly expressed (after Kv1.5) in atria. Notably, uniform hERG expression levels were observed in the three regions (43). On this basis it is reasonable to conclude that the N588K mutation could affect the contribution of IKr to AP repolarisation in all three regions. However, the electrophysiological consequences of the mutation are unlikely to be equivalent for each region, due to a fast phase 1 repolarization and absence of a high plateau phase in atrial and Purkinje fibre APs. Indeed, WT IhERG/IKr profile during atrial and Purkinje AP or AP-like voltage commands differs from that during ventricular APs (14, 19, 44, 45). The importance of the plateau phase in dictating IKr/IhERG magnitude and repolarisation profile is further illustrated: (i) by the fact that it has been shown that the magnitude of WT IhERG during a Purkinje fibre AP command can be increased when the AP plateau is artificially elevated to more positive voltages (45), and (ii) by the observation that the magnitude of WT (Control) IhERG in our simulated AP clamp experiments (Fig. 5) differed between the different AP command waveforms used.

The profile of WT IhERG during Purkinje fibre AP commands at 37 °C in the present study is broadly similar to that reported by Lu and colleagues (19) for heterologously expressed hERG and for native canine Purkinje fibre IKr during a descending ramp protocol (46). The use of the same Purkinje fibre AP waveform as used to study N588K-hERG previously at ambient temperature allows direct comparison between our study and that of Cordeiro et al (14). Comparatively little WT IhERG was observed during a Purkinje fibre AP command at ambient temperature and little difference was seen when N588K hERG was studied (14). By contrast, in the present study, we found that the N588K mutation did alter the voltage-dependence and magnitude of IhERG during a Purkinje fibre AP command, but that this occurred to a lesser extent than for the ventricular AP command. Notably, the difference in peak repolarising current density between ventricular and Purkinje waveforms was greater for N588K than for WT IhERG. This occurred whether hERG alone was studied or whether it was co-expressed with MiRP1. Although the functional significance of MiRP1 to native IKr remains a matter for debate, regional comparisons in the canine heart suggest that MiRP1 (at both mRNA and protein levels) is expressed most highly in Purkinje fibres (30). It is noteworthy, therefore, that although MiRP1 co-expression had comparatively little effect on the overall timing of WT and N588K IhERG during the different AP waveforms, when hERG was co-expressed with MiRP1 in the present study (paralleling co-expression in the earlier study by (14)) the effect of the N588K mutation on the difference in maximal IhERG density during repolarisation between ventricular and Purkinje fibre APs was greater than for hERG expressed alone. The simulation data presented in Fig. 5 are concordant with the experimental data in Fig. 3 in showing a clear augmentation of IhERG during the Purkinje fibre AP in the setting of SQT1, whilst they also demonstrate clearly that the extent of the increase is smaller than during a ventricular AP waveform. Collectively, our data provide the first evidence that, at mammalian physiological temperature, repolarising IhERG/IKr in SQT1 is likely to be greater than normal in both ventricular and Purkinje fibre APs, whilst at the same time heterogeneity in repolarising IhERG/IKr between ventricular cells and Purkinje fibres also is likely to be augmented.

The WT IhERG profile during the atrial AP command seen in this study resembles closely that reported previously for heterologously expressed WT hERG by Lu et al (19). Additionally, the voltage at which peak repolarising current occurred during atrial AP commands is similar to that reported by Muraki et al (44) for rabbit atrial IKr elicited by triangular ‘AP-like’ commands. The atrial AP command used here lacked a high plateau phase and the magnitude of WT currents was much smaller than for a ventricular AP waveform applied to the same cells, as observed previously by Lu et al (19). Nevertheless, the N588K mutation still resulted in a modest positive voltage shift in peak repolarising IhERG and contribution of outward (repolarising) conductance earlier during the AP command. Simulation of the effects of SQT1 on IhERG during an atrial AP command waveform showed more than a doubling of maximal IhERG amplitude during AP repolarisation. In the clinical setting, patients with SQT1 have been found to have shortened atrial effective refractory periods (ERPs), which is consistent with accelerated atrial repolarisation (1-4). Our observations provide an explanation for an underlying basis to atrial ERP shortening, in providing the first demonstration that IKr is likely to contribute greater repolarising current earlier during the atrial AP in the setting of SQT1. It is worth noting that there is experimental evidence for the existence of different AP morphologies in human atrium (e.g. (47, 48)). To take account of this, in additional experiments (not shown), we applied an atrial AP command generated using the Courtemanche et al (49) model, which differed in profile from the Nygren et al (37) AP shown in Fig. 2. Maximal IhERG during repolarisation then occurred at -41.44 ± 1.1 mV (n = 8 cells) for WT IhERG and -12.18 ± 0.4 mV (n = 7 cells) for N588K IhERG, representing a +29 mV shift in peak repolarising current (compared to ~+15 mV with the Nygren AP waveform). Peak IhERG during Courtemanche AP repolarisation was also 2.3-fold greater than that for the Nygren AP in for WT-hERG and 3.0-fold greater for N588K-hERG. These observations indicate that the effect of the N588K mutation can vary with atrial AP morphology, and (as with Purkinje-fibre/ventricular waveform differences), this may augment heterogeneity of atrial repolarisation and therefore, potentially, of refractoriness.

Physiological relevance of differential effects of the N588K mutation under AP clamp

Abbreviated ventricular repolarisation is a signal feature of the SQTS and the most marked effect of the N588K mutation under AP clamp occurs for ventricular AP commands. Our data are consistent with a less extensive change to IKr during atrial than ventricular APs, but support an earlier and greater contribution of the current to atrial repolarisation, which would lead to AP abbreviation and may help explain abbreviated atrial ERPs seen in SQT1 patients (1-4), thereby disposing SQT1 patients towards re-entry and increased susceptibility to atrial fibrillation (1, 6, 20). Under the conditions of this study the effect of the N588K mutation on IhERG during Purkinje fibre APs was more marked than during atrial APs, but less extensive than during ventricular APs. That this was evident not only during in vitro measurements, but also in computer simulations, indicates the profound influence of membrane potential history (and hence AP morphology) on overall IhERG profile. It is notable that SQTS patients exhibit prominent U waves (5, 24) and completion of ventricular repolarisation in SQTS patients (measured from the ECG T wave) has recently been suggested to be disassociated from the end of mechanical systole in SQT patients, whilst the U wave shows a close correlation, consistent with a mechanoelectrical basis for the U wave (50). However, our findings support the notion (proposed previously on the basis of a relative lack of effect of N588K on IhERG profile during Purkinje fibre APs at room temperature (14)) that heterogeneity between bulk ventricular and Purkinje fibre repolarisation is likely to be exacerbated in the setting of SQT1. This could contribute both to U wave formation and increased dispersion of repolarisation and heterogeneity of refractoriness between Purkinje fibres and ventricular tissue. In turn, such changes would be anticipated to contribute to a substrate for ventricular arrhythmia in SQT1.

Altered timing of the response of IhERG to premature depolarising stimuli

When paired ventricular AP waveforms were applied, the N588K mutation had a dual effect on the profile of outward IhERG transients elicited by the second, premature stimulus. Initially (at time points preceding ~ APD90) larger outward current transients were observed for N588K-hERG, which can be attributed due to a combination of the channel’s attenuated inactivation and faster recovery from inactivation at a given potential. However, there was a subsequent time period where normalised N588K-hERG transient amplitude was smaller than that for WT IhERG. This period was larger when comparisons took into account differences in APD in order to approximate the situation in SQT1 (Fig. 6Ci). WT-hERG channels can proceed directly from an inactivated state to the closed state either directly or via the open state (19), and due to altered voltage-dependence of inactivation (14, 15) a smaller proportion of N588K-hERG channels experience the inactive state at potentials over the AP range. Thus, whilst the deactivation rate of IhERG in conventional voltage-clamp experiments has been reported not to differ significantly at 37°C between WT and N588K-hERG (15), the onset of deactivation at potentials during a dynamic AP command may be able to commence earlier for N588K-hERG. Moreover, as maximal outward transient amplitude occurs at earlier time-points for N588K-hERG than for WT-hERG, a decline in the normalised outward transient amplitude can be anticipated to occur earlier for mutant than WT channels. Potential differences in functional channel expression between WT and N588K-hERG precluded direct comparison of absolute outward current transient amplitudes from the two channel types in these experiments, though (using a similar approach that employed by Lu et al (32)) we found the ratio of maximal transient amplitude to peak repolarising current was greater for WT (6.4, n = 7 cells) than for N588K-hERG (1.6, n = 7 cells).

When paired atrial AP commands were studied (Fig. 7A), there was less normalised N588K IhERG compared to WT IhERG at all premature depolarisation intervals examined. When paired Purkinje fibre AP’s were studied (Fig. 7B), there was a dual effect of the N588K mutation, as was the case when normal and shortened premature ventricular AP’s were applied. There was a period prior to the APD90 where N588K-hERG would be more effective at suppressing premature depolarisations than WT-hERG. However, there was also a time period where N588K-hERG would be less effective at suppressing premature depolarisations. Thus, the response of IhERG to premature stimulation was altered by the N588K mutation for both atrial and Purkinje fibre APs; such changes may be particularly significant for Purkinje fibres where repolarisation is largely dependent on IKr (51-53).

An obvious question that arises from these data is what the possible functional consequences might be of the altered response of N588K-hERG to premature depolarising stimuli? It might be anticipated that, at time-points where premature stimuli elicit larger outward transients, hERG can offset more extensively effects of premature membrane potential depolarisation and therefore that decreases in IhERG transients increase arrhythmic risk. However, as highlighted by Lu et al (32), the integrated response of cardiac tissue to premature beats also depends on the behaviour of other ion channel types, and particularly sodium channel reactivation. It is feasible, therefore, that increases in hERG conductance during a premature stimulus could also be proarrhythmic, by antagonising Na+ channel driven depolarisation as occurs for loss of function Na+ channel mutations (32, 54). Thus the overall consequences of altered responses of IhERG to premature stimuli are likely to depend on the combined impact of changes to transient timing and amplitude and, in principle, it is feasible that alterations to transient timing and current amplitude can have either synergistic or antagonistic effects to one another. However, it is notable that, as shown in Fig. 6Cii, even when WT IhERG was considered alone, ventricular APD shortening resulted in a left-ward shift of the relationship between outward current transients and time, resulting in reduced outward transients for some time-points late in ventricular AP repolarisation. Thus, APD shortening per se is likely to influence the response of IKr to premature depolarisation. This situation may be exacerbated when N588K-hERG is considered and, coupled with decreased ERP in SQT1 (3, 6), lead to increased vulnerability to premature arrhythmogenic depolarisation. As discussed below, further future investigation, in silico, is warranted to examine this possibility.


CONCLUSIONS
At present there are no phenotypically accurate animal models incorporating gene mutations seen in human SQT patients (the closest model thus far reported utilised a hERG channel activator that increases conductance without altering inactivation kinetics (55, 56)). Therefore investigation of the likely consequences of SQT mutations currently requires alternative approaches. The value of the AP clamp technique is that in contrast to conventional voltage-clamp, it allows ionic current profiles during physiological waveforms to be investigated directly and takes account of membrane potential ‘history’ in a dynamic fashion that is not otherwise possible (21, 22). Whilst potential limitations of the technique are that data obtained with particular AP waveforms provide only a ‘snap-shot’ of how the channel of interest may behave and do not take into account other ionic conductances present in native tissue, AP clamp data can be highly valuable in providing insight into physiological ionic current profiles, particularly when considered together with parallel simulations, as employed here. The data in this study suggest that in SQT1 repolarising IKr is altered during atrial, Purkinje and ventricular APs with the greatest effect on the ventricles. The altered response of N588K-hERG to premature stimulation may act in concert with accelerated repolarisation to influence the ability of ventricular tissue to propagate premature beats. Our findings can now be utilised together with prior, conventional voltage clamp data (14, 15) to produce and validate further simulations of N588K-hERG in cell and tissue models. Although computational models of IKr where a rectification factor is removed or altered to mimic attenuation of inactivation can provide insight into the effect of the N588K mutation (Fig. 5 and see also (17, 18)), in order to replicate and examine the consequences of our findings regarding premature stimulation, AP and tissue simulations incorporating continuous-time Markov models of IKr (cf (19)) are likely to be required.

Acknowledgments: Support from the British Heart Foundation (PG/04/090, PG/06/139, PG/06/147, FS/08/021) is gratefully acknowledged.


REFERENCES

1. Gussak I, Brugada P, Brugada J et al. Idiopathic short QT interval: a new clinical syndrome? Cardiology 2000; 94: 99-102.
2. Gussak I, Brugada P, Brugada J, Antzelevitch C, Osbakken M, Bjerregaard P. ECG phenomenon of idiopathic and paradoxical short QT intervals. Card Electrophysiol Rev 2002; 6: 49-53.
3. Gaita F, Giustetto C, Bianchi F et al. Short QT Syndrome: a familial cause of sudden death. Circulation 2003; 108: 965-970.
4. Schimpf R, Wolpert C, Gaita F, Giustetto C, Borggrefe M. Short QT syndrome. Cardiovasc Res 2005; 67: 357-366.
5. Brugada R, Hong K, Dumaine R et al. Sudden death associated with short-QT syndrome linked to mutations in HERG. Circulation 2004; 109: 30-35.
6. Hong K, Bjerregaard P, Gussak I, Brugada R. Short QT syndrome and atrial fibrillation caused by mutation in KCNH2. J Cardiovasc Electrophysiol 2005; 16: 394-396.
7. Bellocq C, van Ginneken AC, Bezzina CR et al. Mutation in the KCNQ1 gene leading to the short QT-interval syndrome. Circulation 2004; 109: 2394-2397.
8. Hong K, Piper DR, Diaz-Valdecantos A et al. De novo KCNQ1 mutation responsible for atrial fibrillation and short QT syndrome in utero. Cardiovasc Res 2005; 68: 433-440.
9. Priori SG, Pandit SV, Rivolta I et al. A novel form of short QT syndrome (SQT3) is caused by a mutation in the KCNJ2 gene. Circ Res 2005; 96: 800-807.
10. Tamargo J, Caballero R, Gomez R, Valenzuela C, Delpon E. Pharmacology of cardiac potassium channels. Cardiovasc Res 2004; 62: 9-33.
11. Liu J, Zhang M, Jiang M, Tseng GN. Structural and functional role of the extracellular S5-P linker in the HERG potassium channel. J Gen Physiol 2002; 120: 723-737.
12. Torres AM, Bansal PS, Sunde M et al. Structure of the HERG K+ channel S5-P extracellular linker: role of an amphipathic alpha-helix in C-type inactivation. J Biol Chem 2003; 278: 42136-42148.
13. Jiang M, Zhang M, Maslennikov IV et al. Dynamic conformational changes of extracellular S5-P linkers in the hERG channel. J Physiology-(London) 2005; 569: 75-89.
14. Cordeiro JM, Brugada R, Wu YS, Hong K, Dumaine R. Modulation of IKr inactivation by mutation in KCNH2: a link to arrhythmogenesis in short QT syndrome. Cardiovasc Res 2005; 67: 498-509.
15. McPate MJ, Duncan RS, Milnes JT, Witchel HJ, Hancox JC. The N588K-HERG K+ channel mutation in the ‘short QT syndrome’: mechanism of gain-in-function determined at 37°C. Biochem Biophys Res Commun 2005; 334: 441-449.
16. McPate MJ, Duncan RS, Hancox JC, Witchel HJ. Pharmacology of the short QT syndrome N588K-hERG K+ channel mutation: differential impact on selected class I and class III antiarrhythmic drugs. Br J Pharmacol 2008; 155: 957-966.
17. Zhang H, Hancox JC. In silico study of action potential and QT interval shortening due to loss of inactivation of the cardiac rapid delayed rectifier potassium current. Biochem Biophys Res Commun 2004; 322: 693-699.
18. Weiss DL, Seemann G, Sachse FB, Dossel O. Modelling of short QT syndrome in a heterogeneous model of the human ventricular wall. Europace 2005; 7(Suppl 2): 105-117.
19. Lu Y, Mahaut-Smith MP, Varghese A et al. Effects of premature stimulation on HERG K+ channels. J Physiol 2001; 537: 843-851.
20. Bjerregaard P, Gussak I. Short QT syndrome. Ann Noninvasive Electrocardiol 2005; 10: 436-440.
21. Hancox JC, Levi AJ, Witchel HJ. Time course and voltage dependence of expressed HERG current compared with native “rapid” delayed rectifier K+ current during the cardiac ventricular action potential. Pflugers Arch 1998; 436: 843-853.
22. Noble D, Varghese A, Kohl P, Noble P. Improved guinea-pig ventricular cell model incorporating a diadic space, IKr and IKs, and length- and tension-dependent processes. Can J Cardiol 1998; 14: 123-134.
23. Bjerregaard P, Jahangir A, Gussak I. Targeted therapy for short QT syndrome. Expert Opinion Ther Targets 2006; 10: 393-400.
24. Antzelevitch C. Cellular Basis for the Repolarization Waves of the ECG. Ann N Y Acad Sci 2006; 1080: 268-281.
25. Vandenberg JI, Varghese A, Lu Y et al. Temperature dependence of human Ether-a-go-go Related Gene (hERG) K+ currents. Am J Physiol Cell Physiol 2006; 291: C165-C175.
26. Abbott GW, Sesti F, Splawski I et al. MiRP1 forms IKr potassium channels with HERG and is associated with cardiac arrhythmia. Cell 1999; 97: 175-187.
27. Sesti F, Abbott GW, Wei J et al. A common polymorphism associated with antibiotic-induced cardiac arrhythmia. Proc Nat Acad Sci USA 2000; 97: 10613-10618.
28. Weerapura M, Nattel S, Chartier D, Caballero R, Hebert TE. A comparison of currents carried by HERG, with and without coexpression of MiRP1, and the native rapid delayed rectifier current. Is MiRP1 the missing link? J Physiol 2002; 540: 15-27.
29. Hancox JC, McPate MJ, El HA, Zhang YH. The hERG potassium channel and hERG screening for drug-induced torsades de pointes. Pharmacol Ther 2008; 119: 118-132.
30. Pourrier M, Zicha S, Ehrlich J, Han W and Nattel S. Canine ventricular KCNE2 expression resides predominantly in Purkinje fibers. Circulation Research 2003; 93: 189-191.
31. Sanguinetti MC and Tristani-Firouzi M. hERG potassium channels and cardiac arrhythmia. Nature 2006; 440: 463-469.
32. Lu Y, Mahaut-Smith MP, Huang CL and Vandenberg JI. Mutant MiRP1 subunits modulate HERG K+ channel gating: a mechanism for pro-arrhythmia in long QT syndrome type 6. J Physiol 2003; 551: 253-262.
33. Friederich P, Solth A, Schillemeit S, Isbrandt D. Local anaesthetic sensitivities of cloned HERG channels from human heart: comparison with HERG/MiRP1 and HERG/MiRP1 T8A. Br J Anaesth 2004; 92: 93-101.
34. Ridley JM, Milnes JT, Duncan RS et al. Inhibition of the HERG K+ channel by the antifungal drug ketoconazole depends on channel gating and involves the S6 residue F656. Febs Letters 2006; 580: 1999-2005.
35. McPate MJ, Duncan RS, Witchel HJ, Hancox JC. Disopyramide is an effective inhibitor of mutant HERG K+ channels involved in variant 1 short QT syndrome. J Mol Cell Cardiol 2006; 41: 563-566.
36. Ten Tusscher KH, Noble D, Noble PJ, Panfilov AV. A model for human ventricular tissue. Am. J Physiol Heart Circ Physiol 2004; 286: H1573-H1589.
37. Nygren A, Fiset C, Firek L et al. Mathematical model of an adult human atrial cell - The role of K+ currents in repolarization. Circ Res 1998; 82: 63-81.
38. Bjerregaard P, Gussak, I. Atrial fibrillation in the setting of familial short QT interval. Heart Rhythm 2004; 1: S165.
39. Wolpert C, Schimpf R, Veltmann C, Giustetto C, Gaita F, Borggrefe M. Clinical characteristics and treatment of short QT syndrome. Expert Rev Cardiovasc Ther 2005; 3: 611-617.
40. Yap YG, Camm AJ. Induced QT prolongation and torsades de pointes. Heart 2003; 89: 1363-1372.
41. Hancox JC, Witchel HJ, Varghese A. Alteration of HERG current profile during the cardiac ventricular action potential, following a pore mutation. Biochem Biophys Res Commun 1998; 253: 719-724.
42. Clay JR, Ogbaghebriel A, Paquette T, Sasyniuk BI, Shrier A. A quantitative description of the E-4031-sensitive repolarization current in rabbit ventricular myocytes. Biophys J 1995; 69: 1830-1837.
43. Gaborit N, Le Bouter S, Szuts V et al. Regional and tissue specific transcript signatures of ion channel genes in the non-diseased human heart. J Physiol (London) 2007; 582: 675-693.
44. Muraki K, Imaizumi Y, Watanabe M, Habuchi Y, Giles WR. Delayed Rectifier K+ Current in Rabbit Atrial Myocytes. Am J Physiol-Heart Circ Physiol 1995; 269: H524-H532.
45. Dumaine R, Cordeiro JM. Comparison of K+ currents in cardiac Purkinje cells isolated from rabbit and dog. J Mol Cell Cardiol 2007; 42: 378-389.
46. Gintant GA. Characterization and functional consequences of delayed rectifier current transient in ventricular repolarization. Am J Physiol-Heart Circ Physiol 2000; 278: H806-H817.
47. Wang ZG, Fermini B, Nattel S. Delayed Rectifier Outward Current and Repolarization in Human Atrial Myocytes. Circ Res 1993; 73: 276-285.
48. Benardeau A, Hatem SN, RuckerMartin C et al. Contribution of Na+/Ca2+ exchange to action potential of human atrial myocytes. Am J Physiol-Heart Circ Physiol 1996; 271: H1151-H1161.
49. Courtemanche M, Ramirez RJ, Nattel S. Ionic mechanisms underlying human atrial action potential properties: insights from a mathematical model. Am J Physiol 1998; 275: H301-H321.
50. Schimpf R, Antzelevitch C, Haghi D et al. Electromechanical coupling in patients with the short QT syndrome: Further insights into the mechanoelectrical hypothesis of the U wave. Heart Rhythm 2008; 5: 241-245.
51. Cordeiro JM, Spitzer KW, Giles WR. Repolarizing K+ currents in rabbit heart Purkinje cells. J Physiol (London) 1998; 508: 811-823.
52. Varro A, Balati B, Iost N et al. The role of the delayed rectifier component IKs in dog ventricular muscle and Purkinje fibre repolarization. J Physiol (London) 2000; 523: 67-81.
53. Lu H, Vlaminckx E, Van De Water A, Gallacher D. Both beta-adrenergic receptor stimulation and cardiac tissue type have important roles in elucidating the functional effects of IKs channel blockers in vitro. J Pharmacol Toxicol Methods 2005; 51: 81-90.
54. Balser JR. Sodium “channelopathies” and sudden death - Must you be so sensitive? Circulation Res 1999; 85: 872-874.
55. Zhou J, Augelli-Szafran CE, Bradley JA et al. Novel potent human ether-a-go-go-related gene (hERG) potassium channel enhancers and their in vitro antiarrhythmic activity. Mol Pharmacol 2005; 68: 876-884.
56. Patel, C, Antzelevitch, C. Cellular basis for arrhythmogenesis in an experimental model of the SQT1 form of the short QT syndrome. Heart Rhythm 2008; 5:585-590.