Does the small conductance Ca 2 + -activated K + current I SK flow under physiological conditions in rabbit and human atrial isolated cardiomyocytes?

Background: The small conductance Ca 2 + -activated K + current (I SK ) is a potential therapeutic target for treating atrial fibrillation. Aim: To clarify, in rabbit and human atrial cardiomyocytes, the intracellular [Ca 2 + ]-sensitivity of I SK , and its contribution to action potential (AP) repolarisation, under physiological conditions. Methods: Whole-cell-patch clamp, fluorescence microscopy: to record ion currents, APs and [Ca 2 + ] i ; 35 – 37 ◦ C. Results: In rabbit atrial myocytes, 0.5 mM Ba 2 + (positive control) significantly decreased whole-cell current, from (cid:0) 12.8 to (cid:0) 4.9 pA/pF ( P < 0.05, n = 17 cells, 8 rabbits). By contrast, the I SK blocker apamin (100 nM) had no effect on whole-cell current, at any set [Ca 2 + ] i (~100 – 450 nM). The I SK blocker ICAGEN (1 μ M: ≥ 2 x IC 50 ) also had no effect on current over this [Ca 2 + ] i range. In human atrial myocytes, neither 1 μ M ICAGEN (at [Ca 2 + ] i ~ 100 – 450 nM), nor 100 nM apamin ([Ca 2 + ] i ~ 250 nM) affected whole-cell current (5 – 10 cells, 3 – 5 patients/ group). APs were significantly prolonged (at APD 30 and APD 70 ) by 2 mM 4-aminopyridine (positive control) in rabbit atrial myocytes, but 1 μ M ICAGEN had no effect on APDs, versus either pre-ICAGEN or time-matched controls. High concentration (10 μ M) ICAGEN (potentially I SK -non-selective) moderately increased APD 70 and APD 90 , by 5 and 26 ms, respectively. In human atrial myocytes, 1 μ M ICAGEN had no effect on APD 30 – 90 , whether stimulated at 1, 2 or 3 Hz (6 – 9 cells, 2 – 4 patients/rate). Conclusion: I SK does not flow in human or rabbit atrial cardiomyocytes with [Ca 2 + ] i set within the global average diastolic-systolic range, nor during APs stimulated at physiological or supra-physiological ( ≤ 3 Hz) rates.


Introduction
The small conductance Ca 2+ -activated K + current (I SK ) has been shown, in a variety of cell types and experimental conditions, to increase in amplitude upon increasing the intracellular Ca 2+ concentration ([Ca 2+ ] i ), e.g.[1][2][3][4][5][6][7].As such, I SK has the potential to change the cardiac action potential (AP) shape in response to physiological and/or pathophysiological changes in [Ca 2+ ] i .This, and reports that I SK may be more prominent in atria than ventricles, e.g.[8][9][10], has led to the consideration of I SK as a potential therapeutic target in the treatment of atrial fibrillation (AF) [11,12].In animal models of AF, I SK blockers have demonstrated efficacy in preventing new-onset AF, i.e. in physiologically normal (un-remodelled) atria [6,[13][14][15], as well as AF resulting from atrial remodelling by chronic atrial tachypacing [6,16]; although atrial remodelling can diminish I SK blocker effects [10,17].It is important, therefore, to establish both the [Ca 2+ ] i -sensitivity of atrial I SK and the contribution of I SK to atrial AP repolarisation, under physiological conditions, in un-remodelled atrium.
The physiological global average [Ca 2+ ] i has been measured (with [Ca 2+ ]-sensitive dyes and fluorescence microscopy) in atrial myocytes of various species, e.g.human: diastolic range 168-325 nM and systolic 413-600 nM [7,18,19]; rabbit: diastolic 170 nM, systolic 453 nM [20] and these values, at least systolic, are within the range of I SK EC 50 s above.Furthermore, since the Ca 2+ sensed by SK channels may be located in the sub-sarcolemmal space [21], which may reach 10-20 μM during systole [22], then it is reasonable to argue that I SK should activate, to some degree, under physiological conditions and could therefore influence AP shape and duration (APD) in un-remodelled atrium.Additionally, the supra-physiological rates from AF paroxysms could elevate [Ca 2+ ] i further (as suggested by studies which increased stimulation rate to 2-3 Hz in human atrial trabeculae [23,24]) with the potential, therefore, to enhance I SK blocker effects on APD.
However, reports of the contribution of I SK to APD are equivocal.In a study comparing animal species and cardiac chambers, apamin had no effect on intact tissue APD in ventricle of rat, dog or human, or in atrium of rat or dog [25].Absence of APD change [6], as well as shortening [26] or lengthening [27], by I SK inhibitors was reported in dog atrial myocytes or tissues.In human atrial myocytes, APD was increased [9,10,28,29], or unaffected [26], by I SK inhibitors.No rabbit atrial studies were found; a species also used for understanding AF mechanisms and their prevention [30].The reasons for such reported disparity amongst these studies are unclear, but could include the variety of species and cardiac preparations/chambers studied, experimental conditions such as [Ca 2+ ] i -buffering, stimulation rate, temperature, and also differing selectivities and/or concentrations of various I SK blockers used.
The bee venom toxin, SK channel pore blocker, apamin, is the conventionally used inhibitor of I SK (the apamin-sensitive K + current termed I KAS [4,5,7]), usually at 100 nM [4,5,7,25,26,28,29].This concentration should not affect Na + current (I Na ), transient outward K + current (I TO ), L-type Ca 2+ current (I CaL ), rapid (I Kr ) or slow (I Ks ) delayed rectifier K + currents, or inward rectifier K + current (I K1 ) [31] and will be used here for comparison, along with another SK pore blocker, ICAGEN, which showed high selectivity for I SK over I Na , I TO , I Kur (ultra-rapid delayed rectifier K + current), I CaL , I CaT (low voltage-activated Ca 2+ current), I Kr , I Ks , I K1 and I f (funny current) [10].
The aims of the present study, therefore, are to: 1) measure wholecell I SK in human and rabbit atrial cardiomyocytes, using the I SK blockers apamin and ICAGEN, with [Ca 2+ ] i measured and set to various levels in the global average diastolic-systolic range; 2) investigate whether I SK flows during the atrial AP, when sub-sarcolemmal [Ca 2+ ] i should exceed global average levels, at physiological and supraphysiological rates.

Patients, rabbits, ethics
Right atrial tissues were obtained from 29 adult patients who were undergoing cardiac surgery, predominantly for coronary artery bypass grafting.Procedures and experiments involving human atrial myocytes were approved by West of Scotland Research Ethics Service (REC: 17/ WS/0134).Written, informed consent was obtained from all patients.The investigation conformed to the principles outlined in the Declaration of Helsinki.All patients were in sinus rhythm on the day of surgery.See Table 1 for patients' clinical characteristics and drug treatments.Rabbits (n = 37; strain: New Zealand White; supplier: Envigo UK; sex: male; age (mean ± SE [range]): 21.1 ± 0.6 [14.0-30.4]weeks; weight: 3.00 ± 0.06 [2.12-3.88]kg; feeding: ad libitum) were humanely killed by intravenous injection of anaesthetic (100 mg/kg Na + -pentobarbital, via the left marginal ear vein) and removal of the heart, which was retrogradely perfused via the aorta before isolating cardiomyocytes.Procedures and experiments involving rabbit left atrial myocytes (UK Project Licence: 70/8835) were approved by Glasgow University Ethics Review Committee, and conformed to the guidelines from: Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes.

Data and statistical analysis
Data are expressed as means±SEM.Continuous data were compared using 2-sided, 2-sample Student's t-tests; paired or un-paired as appropriate.P < 0.05 was regarded as statistically significant.All statistical analyses were done using Graphpad Prism 7.00 or 8.00 software (Graphpad, San Diego, CA, USA).

An I SK blocker, apamin (by contrast with Ba 2+ ), had no effect on rabbit atrial whole-cell current over a range of measured [Ca 2+ ] i
In rabbit atrial isolated myocytes, Ba 2+ (0.5 mM), used as a positive control for K + current inhibition, markedly decreased whole cell current, as shown by the original current traces in Fig. 1Ai and average data in 1Aiii.The effect was rapid and reversible, peaking typically 30-60 s after starting Ba 2+ superfusion.The mean current density measured at − 115 mV, where current change was large but avoiding any artifact at the ramp start (− 120 mV), was significantly decreased, by 62% (Fig. 1Aiii).
Since I SK is [Ca 2+ ] i -sensitive, before repeating these experiments with the conventional I SK blocker apamin, pipette solutions with three different [Ca 2+ ] were prepared, anticipated to approximate 100 nM, 300 nM and 500 nM free [Ca 2+ ], i.e. substantially overlapping the global average diastolic and/or systolic [Ca 2+ ] i range.Each [Ca 2+ ] was measured optically using Fura-2, and calculated firstly by titrating against EGTA-buffered solutions, then by replacing EGTA with the faster binding buffer BAPTA as would be required in the myocytes.The resulting average free [Ca 2+ ] with BAPTA, as measured in glass tubes, was 82 nM, 248 nM and 447 nM, i.e. lower (and within 18%) than the respective anticipated values (Fig. 1Bi).The [Ca 2+ ] i subsequently measured (with Fura-2) in rabbit atrial isolated myocytes using these Ca 2+ /BAPTA solutions in the patch pipettes, averaged 99 nM, 361 nM, and 453 nM, i.e. within 20% of the anticipated values (Fig. 1Bii).
In different groups of myocytes, using these same Ca 2+ /BAPTA solutions and repeating the voltage clamp experiment in Fig. 1A, except with apamin (100 nM) instead of Ba 2+ , it was shown (Fig. 1C) that apamin had no significant effect on whole-cell current at any [Ca 2+ ] tested (i.e. over the [Ca 2+ ] i range of ~100 to 450 nM), as shown by the representative traces (panels i, iii & v) and average data (ii, iv & vi).

Lack of substantial or consistent effect of the I SK blocker ICAGEN on rabbit atrial whole-cell current
In rabbit atrial myocytes, the I SK blocker ICAGEN, superfused at 1 μM (a concentration deemed selective for I SK ), under the same conditions, including the same pipette [Ca 2+ ] range, as for the above apamin experiments, had no substantial or consistent effect on whole cell current: see Fig. 2, panels (i) for representative traces, and panels (ii) for the average data.There was no significant effect of ICAGEN when using pipette [Ca 2+ ] of either ~82 nM or ~ 447 nM (Fig. 2A and C); with [Ca 2+ ] ~ 248 nM, there was a small increase in whole cell current, by 15%.

Human atrial whole-cell current: absence of effect of I SK blockers
In human atrial isolated myocytes, whereas 0.5 mM Ba 2+ (positive control) markedly and significantly decreased whole-cell current (Fig. 3Ai and ii), the I SK blocker ICAGEN (1 μM) had no significant effect on whole-cell current whether using pipette [Ca 2+ ] of ~82 nM (Fig. 3Bi and ii), ~248 nM (iii & iv), or ~ 447 nM (v & vi).Apamin (100 nM) was also tested, using pipette [Ca 2+ ] ~ 248 nM, and this I SK blocker also had no significant effect (Fig. 3 Ci and ii).

Effect of I TO /I Kur blocker 4-AP, but not I SK blocker ICAGEN, on action potentials in rabbit atrial myocytes
To test whether I SK flows in the rabbit atrial isolated myocytes during the action potential (AP), i.e. when local (sub-sarcolemmal) [Ca 2+ ] i is expected to exceed global average levels, drugs were superfused during the continuous recording of APs, which were stimulated at 1 Hz.First, the K + channel blocker 4-AP at 2 mM (inhibits I TO and I Kur ) was tested (Fig. 4A).The representative superimposed AP traces in Fig 4Ai show , probably from rundown of I CaL .Therefore, a separate group of contemporaneous, time-matched controls (TMC), was used (Fig. 4C) to compare the degree of change in APD in the presence of ICAGEN with that in its absence over a matched period, in order to avoid missing a potential small APD-lengthening effect of ICAGEN which would otherwise be obscured by such APD rundown.The TMCs showed a small degree of APD-shortening at APD 30 and APD 70 (Fig. 4C), and there was no significant difference in the mean degree of APD change at any measured level of repolarisation between the ICAGEN and TMC groups (Fig. 4C).

High concentration ICAGEN prolonged action potentials in rabbit atrial myocytes
The I SK blocker ICAGEN was tested at high concentration (10 μM), for the dual purpose of investigating potential non-selective effects on atrial APs and, by replicating the AP recording conditions as in Fig. 4 (in which 1 μM ICAGEN was tested), confirming the ability of this drug to exert any electrophysiological effects in these isolated myocytes under the same conditions; thus adding qualification of the earlier absence of effects at the lower (I SK -selective) concentration.The mean APD 70 and APD 90 after ICAGEN were not significantly different from the pre-ICAGEN values (Fig. 5D), despite a small increase in 6 of 7 cells.However, when comparing against the TMCs, thus negating the compounding effects of APD rundown, a small but significant increase in mean APD 70 and APD 90 by ICAGEN was revealed, of 5 and 26 ms, respectively (Fig. 5E).

Absence of effect of I SK blocker (1 μM ICAGEN) on human atrial action potentials, stimulated at various rates
In human atrial myocytes stimulated at physiological rate (1 Hz), 1 μM ICAGEN had no significant effect on AP configuration (Fig 6Ai ) or duration (APD 30-90 ), whether measured against pre-ICA values (Aii) or against TMCs (Aiii).Supra-physiological rate stimulation, i.e. with the potential to elevate [Ca 2+ ] i and thus I SK , was then used.Fig. 6B and C show that 1 μM ICAGEN had no significant effect on AP configuration or duration at either 2 Hz or 3 Hz stimulation.

Discussion
The study of I SK blocker effects on whole-cell currents first required a positive control for verification that, under our experimental conditions and recording protocols, changes in a K + current of potentially small magnitude (e.g.~1-2 pA/pF [4,5,21,29,38]) would be detectable.Our results with the K + channel blocker Ba 2+ [8,36], provided such verification; with acute, marked, stable and reversible current inhibition consistently observed in both rabbit and human atrial myocytes.For the subsequent investigation of effects of apamin on whole-cell current at different [Ca 2+ ] i s, we chose the Ca 2+ buffer BAPTA (over EGTA) to set and keep constant the pipette and myocyte [Ca 2+ ] i , owing to its fast binding kinetics and low pH-sensitivity [21,39].The amount of CaCl 2 added to the fixed, saturating concentration (5 mM) of BAPTA to achieve the desired pipette [Ca 2+ ] values (100-500 nM) was determined by first measuring the relationship between [Ca 2+ ] and variable CaCl 2 with fixed and saturating (10 mM) EGTA, since EGTA has high and accurately known purity [40]; using the ratiometric dye Fura-2.It was important to then verify (measure) the [Ca 2+ ] in the BAPTA-buffered pipette solutions, and ultimately the [Ca 2+ ] i in the isolated myocytes.We found the mean measured pipette [Ca 2+ ] values to be moderately lower than the desired values, by a maximum of 18%, and with low inter-sample variation at each [Ca 2+ ].Disparity between calculated and measured free [Ca 2+ ] is well recognised, and can be substantial [39], and emphasises the need to verify the [Ca 2+ ] by measurement.We then measured the myocyte global [Ca 2+ ] i resulting from dialysis with these solutions, in groups of rabbit atrial cells, which indicated that average [Ca 2+ ] i in the subsequent apamin experiments should range ~ 100-450 nM (i.e.within 20% of the desired values, albeit with relatively larger inter-cell variation; see Fig 1Bii); and close to the physiological global average diastolic-systolic range [7,[18][19][20], and the [Ca 2+ ] i EC 50 s for I SK in isolated myocytes [5][6][7].
We found that apamin clearly had no effect on whole-cell current, over the voltage range tested (− 120 to +60 mV), at any of the three [Ca 2+ ] i used, indicating that I SK was not activated under these conditions in rabbit atrial myocytes.No previous studies of apamin on rabbit atrial whole-cell current could be found, but there are several atrial or ventricular studies in other species [4,7,9,25,29,38,[41][42][43].Six of those [4,7,9,29,38,43] measured an apamin-sensitive (subtraction) current, often of relatively low density (e.g.~1-2 pA/pF [4,29,38,43]), to study effects of various interventions on this current.In the other three, a statistical comparison of apamin vs control was provided, which showed no significant effect in rat or dog ventricular myocytes [25], and a small inhibitory effect (maximum 14%) in mouse [41] and human [42] atrial myocytes.To further investigate whole-cell I SK (or its potential absence of activation) in rabbit atrial myocytes, we used the more recently developed, potent and putatively selective small molecule SK channel inhibitor ICAGEN [10].Using the same conditions as for apamin, including pipette [Ca 2+ ]s, we found ICAGEN to have no effect on whole- cell current, at either the highest or lowest [Ca 2+ ] i .At the intermediate [Ca 2+ ] i , there was a statistically significant effect of ICAGEN at − 115 mV, albeit it small (15%); and since it was an increase (rather than a decrease as would result from any K + current inhibition), does not indicate that I SK was active.In a detailed ion current selectivity profiling study of ICAGEN using a heterologous expression system (HEK293 and CHO cells) [10], the IC 50 for I SK2 and I SK3 (SK2 and SK3 being the most abundantly expressed isoforms in human atrium [10]) was 0.3 μM and 0.5 μM, respectively.The ICAGEN concentration used in the present study (1 μM) was ~2-3-fold greater, re-enforcing, along with our apamin data, the absence of I SK flowing in these rabbit atrial myocytes.No reported studies of ICAGEN on currents in rabbit atrial cells could be found.
In the human atrial myocytes, we also found ICAGEN (1 μM) to have a clear absence of effect on whole-cell current, over the voltage and [Ca 2+ ] i ranges tested.In the only other study using ICAGEN in human atrial myocytes [10] (in which pipette free [Ca 2+ ] was calculated as 300 nM, and high [K + ] o (20 mM) was used), a small but significant decrease in whole-cell current (by ~9%) was produced by 1 μM ICAGEN.The IC 50 of this effect was marginally lower (at 143 nM) than in the heterologous expression system.This further supports the use of 1 μM ICAGEN for I SK inhibition, and our data indicate that, as with the rabbit atrial cells, I SK did not flow in the human atrial myocytes, with [Ca 2+ ] i up to ~450 nM.
During the cardiac AP, the [Ca 2+ ] in the sub-sarcolemmal space may reach 10-20 μM [22], i.e. ~1-2 orders of magnitude greater than either the reported calculated [Ca 2+ ] i EC 50 s for I SK activation [1][2][3][4][5][6][7] or the highest measured pipette [Ca 2+ ] used here.Moreover, SK channels may preferentially sense this localised Ca 2+ source, arising both from I CaL [21] and sarcoplasmic reticulum [44].Therefore, a higher degree of I SK activation (and thus responses to apamin and/or ICAGEN) might be expected during the recording of APs (with minimal [Ca 2+ ] i -buffering, as used here: 0.1 mM EGTA) than of whole-cell-current with strongly buffered and constant [Ca 2+ ] i.Nevertheless, we found that, by marked contrast with K + current inhibition using 4-AP (I TO /I Kur blocker) which rapidly and substantially prolonged the APs, ICAGEN at 1 μM had no significant effect on APD at either 30, 70 or 90% repolarisation in the rabbit atrial myocytes, thus indicating absence of I SK activation in these cells.
We considered it prudent to check for any effects of ICAGEN at a 10fold higher concentration, i.e. with the potential to partially inhibit I TO (IC 50 : 21 μM [10]) and possibly also I Kr (IC 50 : >32 μM [10]) and thus to prolong APD.In this case, we found a significant, albeit moderate, increase in both APD 70 and APD 90 by ICAGEN (10 μM) in rabbit atrial cells.Whilst we cannot exclude the possibility of a contribution from a potentially stronger I SK inhibition with 10 μM ICAGEN, the more selective 1 μM is already 2-7 times higher than reported IC 50 s [10], and since prolongation of both APD 70 and APD 90 results from partial I TO inhibition in rabbit atrial cells under similar conditions [45], the possibility that the APD prolongation by 10 μM ICAGEN resulted from partial I TO inhibition also cannot be excluded.No previous studies of effects of ICAGEN on APs in rabbit atrial myocytes or tissues could be found, but there are two studies on rat atrial tissues, in which high concentration ICAGEN was tested (10-30 μM) and found to prolong APD 90 [46,47]; consistent with the present (10 μM ICAGEN) data.Furthermore, apamin (100 nM) has also been studied in rat atrial tissues, and had no effect on APD [25], in line with our lower concentration (1 μM) ICAGEN results.
In the human atrial myocytes, stimulated at physiological rate (1 Hz), our data indicate that I SK did not flow during the AP since ICAGEN, at the I SK -selective 1 μM, had no significant effect on APD 30-90.In a single previous human atrial study using ICAGEN [10], myocytes were stimulated at 0.5 Hz, and 1 μM ICAGEN caused a significant (though moderate: ~13%) increase in APD 90 .A higher concentration (10 μM) was used in intact atrial muscle strips in the same study, which prolonged APD 90 at 1 Hz stimulation.Other I SK blockers have also been studied in human atrial myocytes, with 100 nM apamin producing either no significant effect on APD 50 or APD 90 (in atrial cells from patients with heart failure [26]) or an increase (of ~30% [9], ~18% [28], or ~ 11% [29]).An AP-clamp study in human atrial myocytes demonstrated a lack of apamin (100 nM)-sensitive current flowing during the course of the AP [42] whereas, in the same study, the non-peptidic SK pore blocker UCL-1684 [48] decreased outward, repolarising, current.Finally, NS8593 (a negative allosteric I SK modulator) caused a moderate (~13%) increase in APD 90 in human atrial cells (stimulated at 0.5 Hz) [10], although non-selectivity of this drug for I SK was noted [10,41].The variety of I SK blockers and protocols used and AP responses observed so far in human atrial myocytes, would suggest the utility of a future study to directly compare such drugs in these cells including, along with apamin and ICAGEN, UCL-1684, the negative allosteric I SK modulator AP14145 [3], and also the I SK activator NS309 [49].
Increasing the AP firing rate (as would occur with new-onset, or paroxysmal, AF) has the potential to enhance I SK activation (and thus APD-responses to I SK inhibition), if it were to increase [Ca 2+ ] i sufficiently.In human atrial isolated trabeculae, increasing the stimulation rate from 1 Hz to 2 Hz caused an ~1.6-fold increase in systolic [Ca 2+ ] i [23]; although only diastolic [Ca 2+ ] i was increased (at 3 Hz) in another study [24].Our data, in human atrial isolated myocytes, indicate that increasing the stimulation rate from 1 Hz to 2 Hz or 3 Hz did not significantly activate I SK during the AP, since 1 μM ICAGEN had no significant effect on APD 30-90 .This is in agreement with a human atrial study using apamin [26]; which had no significant effect on APs whether stimulated at 0.5, 1 or 2 Hz.With high concentration ICAGEN (10 μM) in human atrial tissues [10], APD 90 was increased at 1 Hz stimulation, but not at 3 Hz.Furthermore, in rat atrial tissues, 20 μM and 30 μM ICAGEN prolonged APD 90 at 5 Hz stimulation, but the effect was not potentiated by increasing the rate to 9 Hz or 11 Hz [46].Rate-dependent effects of I SK inhibition were also investigated in dog and rat ventricular muscles (0.3-3 Hz); apamin having no effect on AP configuration at any rate [25].In a recent study in human atrial myocytes, 5 Hz stimulation for 10 mins promoted localisation of SK2 channels to the sarcolemma, increased I SK and shortened APD; effects attenuated by apamin [29].
Elevated atrial [Ca 2+ ] i can also result from acute adrenergic stimulation; shown (using isoprenaline [50,51]) to reveal I SK which was absent in the un-stimulated cells.Furthermore, chronic hyper-adrenergic drive (e.g.associated with ventricular hypertrophy [50]) may enhance I SK , by PKA-dependent phosphorylation and altered channel rectification [50,52].Neither acute nor chronic adrenergic activation were studied here, but their complex signalling pathways, potentially involving a wide range of I SK -associated multi-protein complexes (e.g.calmodulin and various protein kinases and phosphatases), warrant future investigation.
Taken together, the present data indicate, in line with some of the studies above as discussed, absence of I SK activation in un-remodelled, non-adrenergically-stimulated, atrial myocytes, including at supraphysiological stimulation rates.However, it should be appreciated that substantial changes in I SK , and in myocyte AP-responses to I SK blockers, may result from pathological atrial remodelling; another aspect currently under debate.For example, chronic AF in patients has been associated with either increased [7,29,53] or decreased [10,38] atrial I SK or I KAS , and either potentiated [29] or attenuated [10] APD-responses to I SK blockers; and chronic atrial tachy-pacing in dogs increased atrial I SK and APD-responses to I SK inhibition [6].Atrial [29] or pulmonary vein [54] tachy-pacing studies have revealed the rapid onset of remodelling of I SK and associated proteins, e.g.within hours [54] or even minutes [29]; and also highlight both the functional importance of SK channel location (with remodelling promoting trafficking from perinuclear regions to the plasmalemma) and the complexity of the associated signalling pathways.Remodelling from heart failure or hypertrophy features increased ventricular I SK [5,50,55], whereas in atrium, both I SK responses [26] and tissue expression of SK channel protein [29] were unaffected; either by heart failure or by heart failure with concurrent AF.
Study limitations: we used isolated myocytes and, whilst allowing ion current and AP data to be compared under similar experimental conditions, the possibility of damage to certain ion channels (potentially including SK) by the requisite cell isolation enzymes, cannot be excluded.However, countering that possibility, both absence [6,25] and presence [10,27] of I SK blocker effects on APD were reported whether in intact tissues (not requiring enzymes) or in isolated myocytes.Also, we did not perform immuno-detection of SK channels in our myocytes.However, SK channel proteins have previously been shown, using immuno-histochemical staining techniques, to occur in the plasmalemma of human atrial myocytes, whether the cells were enzymaticallyisolated [29,42] or fixed in intact tissue sections [10,29,38], as well as other cell types as above [25,50,51,54].Finally, we used the rupturedpatch configuration, whereas several studies have used perforatedpatch [9,25,26,28] which, by limiting intracellular dialysis may help retain cytoplasmic (and potentially I SK -sensitive) signalling molecules.However, since both absence [25,26] and presence [9,28] of effect of I SK blockers on both whole cell current and APs were reported in these perforated-patch studies, this suggests the patch configuration also should not be a confounder.
In conclusion, the present findings add weight to the argument that substantial I SK does not flow in rabbit or human atrial isolated cardiomyocytes under physiological conditions.

that 4 -
AP substantially prolonged the APD, particularly around phase 1 (~APD 20 -APD 50 ) and the AP plateau (~APD 50 -APD 80 ).Fig 4Aii shows the typical time-course of this effect; with the onset of increase in APD 30 occurring within 30 s of starting 4-AP superfusion, and the peak drug effect occurring ~30-45 s later.Corresponding average APD data (panel iii) confirmed a marked and significant increase by 4-AP, in APD 30 (by 70%) and in APD 70 (by 28%); and with no significant effect on APD 90 .By contrast with this positive control, ICAGEN (1 μM) did not significantly alter the AP waveform, as shown by the AP traces in Fig 4Bi, by the representative time course of change in APD 30 , APD 70 and APD 90 in Fig 4Bii, and by the corresponding mean APD data in Fig 4Biii.The control APD may undergo a small degree of time-dependent shortening during the course of an experiment, often around the AP plateau region, e.g.APD 70 (see Fig 5Bii) Fig. 5A shows the representative effect of 10 μM ICAGEN on the AP waveform: APDprolongation around phase 2-3, i.e. ~APD 70 -APD 90 .The time course of this effect was moderately slower than for 4-AP, as shown for APD 70 (Fig 5Bi) and APD 90 (Ci).The increase in APD 70 with ICAGEN occurred despite the corresponding continuous decrease in the TMC (Fig 5Bii).