K2p3.1 protein is expressed as a transmural gradient across the rat left ventricular free wall

Abstract Introduction K2p3.1, also known as TASK‐1, is a twin‐pore acid‐sensitive repolarizing K+ channel, responsible for a background potassium current that significantly contributes to setting the resting membrane potential of cardiac myocytes. Inhibition of IK2p3.1 alters cardiac repolarization and is proarrhythmogenic. In this study, we have examined the expression of K2p3.1 and function of this channel in tissue and myocytes from across the left ventricular free wall. Methods and Results Using fluorescence immunocytochemistry, the expression of K2p3.1 protein in myocytes from the subendocardial region was found to be twice (205% ± 13.5%) that found in myocytes from the subepicardial region of the left ventricle (100% ± 5.3%). The left ventricular free wall exhibited a marked transmural gradient of K2p3.1 protein expression. Western blot analysis confirmed significantly higher K2p3.1 protein expression in subendocardial tissue (156% ± 2.5%) than subepicardial tissue (100% ± 5.0%). However, there was no difference in K2p3.1 messenger RNA expression. Whole‐cell patch clamp identified IK2p3.1 current density to be significantly greater in myocytes isolated from the subendocardium (7.66 ± 0.53 pA/pF) compared with those from the subepicardium (3.47 ± 0.74 pA/pF). Conclusions This is the first study to identify a transmural gradient of K2p3.1 in the left ventricle. This gradient has implications for understanding ventricular arrhythmogenesis under conditions of ischemia but also in response to other modulatory factors, such as adrenergic stimulation and the presence of anesthetics that inhibits or activates this channel.

activation at pH more than 7.8 but complete inhibition at pH values less than 6.4. [1][2][3] The pK of the channel is~7.3 with a Hill coefficient of~1.6, thus showing a steep sensitivity to pH across this key physiological and pathophysiological range. 1 Channel functionality of K 2p 3.1 can also be modulated by several other factors: inhibition occurs during hypoxia, in response to α adrenergic 4 and endothelin-1 receptor stimulation 5

and in
the presence of the endocannabinoid anandamide (or its stable analog methanandamide). 6,7 It also shows phosphorylation-dependent inhibition by protein kinase C 7 and inhibition by platelet activating factor. 8 In contrast, activation of K 2p 3.1 channels occurs in response to the key inhalational anesthetics halothane, sevoflurane, and isoflurane. 4,9 The broad spectrum of intrinsic and extrinsic factors that dynamically modulate this channel make this an important target of interest for understanding the response of the cardiac action potential to numerous stressors and conditions. K 2p 3.1 protein has been demonstrated to be present in both the atria and ventricles of the heart. 2,10,11 In myocytes derived from human atria samples, I K2p3.1 has been shown to contribute significantly to repolarization of the myocyte and constitutes approximately 40% of the background potassium current. 12 Inhibition or complete genetic knockout of K 2p 3.1 channels in the mouse heart causes a significant prolongation of the ventricular action potential duration (APD) (by approximately 17%) and an increase in QTc and QRS intervals (by approximately 30% each). 13,14 In ventricular myocytes, similarly to atrial myocytes, I K2p3.1 contributes to repolarization and shortening of the APD. 15,16 This impact has been demonstrated to vary between cells with Putzke et al, 16 suggesting this may be due to variation in the specific origin of the myocytes in such pooled cell preparations and their associated action potential characteristics. Within this study, they were able to model the impact of current through K 2p 3.1 using the "Pandit" mathematical model of the rat ventricular subepicardial myocyte, 17 identifying a predicted significant prolongation of the APD if this current were inhibited in myocytes from this region. 16 No previous investigations have, however, directly investigated potential heterogeneity of I K2p3.1 expression within the ventricular wall. Such heterogeneity would have implications for understanding the response of the heart to the numerous changes altering the activity of this channel and its associated current. Our hypothesis was, therefore, that a heterogeneous expression pattern of K 2p 3.1 and associated current exists within the left ventricular free wall. Such a difference would have the potential to impact the dispersion of repolarization across the wall of the heart in the event of changes in pH and many other physiological and pathological signals, predisposing to arrhythmias.

| Acquisition of tissue and single myocytes
Male Wistar rats aged 5 to 6 months were humanely killed via concussion followed by cervical dislocation. All work was performed in accordance with the Animals (Scientific Procedures) Act 1986 and approved by local ethics committees. Each heart was dissected to remove the left ventricular free wall. The cross-section of the ventricular wall for immunocytochemistry was oriented in cryomedia and snap frozen in isopentane cooled by liquid nitrogen. For Western blot analysis and messenger RNA (mRNA) analysis the left ventricular free wall was dissected into the subepicardial and subendocardial layers (~30% of the total wall thickness from each surface), the midlayer and outer endothelium layers were discarded; regional tissue samples were frozen in liquid nitrogen and stored at −80°C until used.
The preparation of subepicardial and subendocardial single myocytes used a variation of a method previously described by Harrison et al. 18 Briefly, each heart was perfused via the aorta with a 4-(2-hydroxyethyl)-

| Electrophysiological recordings
The method for recording K2p3.1 currents was adapted from Besana et al. 19 Subepicardial or subendocardial myocytes were initially super- NaOH. Cells were held at −10 mV in the presence of the high K + solution and allowed to equilibrate for 6 minutes before measuring currents. R( + )-Methanandamide (Sigma-Aldrich), a selective inhibitor of the K 2p 3.1 channel, was dissolved in ethanol (5 mg/mL), and further diluted to a final working concentration of 10 µM. To selectively identify current through K 2p 3.1 channels, a 6 second voltage ramp protocol from −50 to 30 mV was performed in the presence and absence of the channel inhibitor with the difference current determined by subtraction.
Cell capacitance was assessed using 10 mV hyperpolarizing pulses from a holding potential of −80 mV. Recordings were made using an Axopatch 200B amplifier (Molecular Devices, San Jose, CA) with digital/analog conversion via a Digidata 1322 interface (Molecular Devices). During K 2p 3.1 current recordings whole-cell capacitance was compensated and series-resistance compensation was typically 70% to 75%. Data were acquired at 5 kHz with a low pass Bessel filter of 1 kHz. Electrophysiological data were analyzed using Clampfit 9 (Molecular Devices). Records were not corrected for the liquid junctional potential that was measured to be −9.8 mV. Peak current density comparisons were made at 30 mV. Current density measurements were expressed relative to the whole-cell capacitance, which was determined as the integral of the whole-cell capacitance transient. I K2p3.1 was taken as the methanandamide-sensitive current. protein (205% ± 13.5%) compared with the signal from myocytes from the subepicardial region (100% ± 5.3%) ( Figure 1C; n = 92 cells per group; P < 0.0001). This significant difference of immunofluorescence was maintained when the signal was expressed relative to cell cross-sectional area ( Figure 1D; n = 92 cells per group; P < 0.0001). As such, using this method total and density of K 2p 3.1 protein expression were determined to be higher in cells from the subendocardial region compared with those from the subepicardial region.

| Statistical analysis
3.2 | A transmural gradient of K 2p 3.1 protein expression across the left ventricular free wall Ventricular tissue exhibited the same pattern of labelled K 2p 3.1 protein as previously observed in myocytes with strong labelling of transverse striations across each myocyte and also present at the intercalated disc (Figure 2A). The regional distribution of immunofluorescent labelled K 2p 3.1 protein was determined across the left ventricular free wall. The profile of immunofluorescent labelled K 2p 3.1 protein intensity showed a progressive increase towards the endocardial surface. Immunofluorescence was averaged for each 20% interval distance along the profile from the epicardium for each section ( Figure 2B and 2C). Significantly greater fluorescence was detected in the most endocardial segment at 80% to 100% of the transmural distance (171% ± 7.7% endocardial layer) compared with the initial epicardial 0% to 20% of the distance (100% ± 5.3% epicardial layer). Similar significant differences were noted when the 0% to 20% interval distance was compared with the 60% to 70% distance interval (130% ± 6.4%) and 80% to 90% distance interval (166% ± 8.7%) ( Figure 2D) (n = 49 sections; one-way analysis of variance with Holm-Sidak comparisons *P < 0.001).  Figure 3A). We have previously shown this specific band to be absent when the anti-K 2p 3.1 antibody is competitively inhibited with the appropriate antigenic K 2p 3.1 peptide. 10 Furthermore, no bands were observed after omission of the primary anti-K 2p 3.1 antibody. The integrated signal from the labelled K 2p 3.1 protein was normalized to that for labelled desmin protein as identified per tissue sample lane. Our assessment of the relative density of labelled K 2p 3.1 protein showed that tissue from the subendocardial region possessed more than 50% more K 2p 3.1 protein than tissue from the subepicardial region (157% ± 2.6% vs 100% ± 5.0%, respectively; Figure 1A; n = 8; P < 0.0001).
In contrast, analysis of K 2p 3.1 mRNA expression found there was no significant difference in mRNA content of the myocardium

| Regional differences in K 2p 3.1 current
We have determined a transmural gradient of protein expression but this does not automatically indicate a functional difference across the ventricular wall. Currents recorded using a slow ramp protocol (−50 to 30 mV over 6 seconds) showed a methanandamide-sensitive component to the current-voltage relationship (shown in Figure 4) typical of an open rectifying current with a reversal potential at −10 mV (−19.8 mV when corrected for the junction potential).
Current-voltage relationships in myocytes from both the subepicardial and subendocardial layers displayed similar rectification. Overall, a significantly larger current density was measured in myocytes from the subendocardial region (7.66 ± 0.53 pA/pF; n = 7) in comparison with myocytes from the subepicardial region (3.47 ± 0.74 pA/pF; n = 6; Figure 4E; *P < 0.001). On average the current density in myocytes from the subendocardial layer was more than double that identified in cells from the subepicardial region. The cardiac action potential displays characteristics that are distinct for each region of the healthy heart. 22 The key difference in the action potential profiles across the left ventricular wall is the rate of repolarization, which occurs more rapidly in the subepicardial region of the left ventricle than the subendocardial region, principally due to heterogeneity in expression of key potassium channels. 23 We have now identified significant differences in I K2p3.1, adding to our understanding of the heterogeneity of K + channel expression. 3,8 Previous measures by Putzke et al 16  Given that myocytes from the subendocardial region already have longer APD compared with those from the subepicardial region, inhibition of K 2p 3.1 channels is therefore likely to further exacerbate this difference increasing dispersion of repolarization.

| DISCUSSION
F I G U R E 3 Quantification of K 2p 3.1 protein and mRNA expression. A, An example of a Western blot showing the identified specific single band labelled for K 2p 3.1 protein (and desmin protein band for normalization). Significantly more K 2p 3.1 protein was expressed in tissue from the subendocardial than the subepicardial layer (mean ± SEM; n = 8; Student's t test *P < 0.0001). B, No significant difference in the amount of K 2p 3.1 mRNA as expressed relative to GAPDH mRNA was identified between regions (mean ± SEM; n = 10). GAPDH, glyceraldehyde 3-phosphate dehydrogenase; mRNA, messenger RNA Inhibition of I K2p3.1 is known to promote spontaneous activity and prolong the cardiac action potential, leading to early after depolarizations (EADs) and spontaneous ectopic activity creating a substrate for ventricular arrhythmia. 8 Factors that modulate K 2p 3.1 can arise as a result of an ischaemic insult with local accumulation of hydrogen ions being the most notable acute inhibitor under such conditions. 3,19,24 The high sensitivity of K 2p 3.1 to extracellular pH in the physiological range leads to potentially complete current inhibition in the event of a regional or global ischemic insult. Even modest acidosis under low-flow or complete ischemia will significantly reduce the current through this channel prolonging the APD and predisposing to EADs. Also associated with ischaemic insults is inflammation and the rapid activation of leukocytes that release platelet activating factor (PAF). 25 This release of PAF in the heart is associated with APD prolongation and EADs and arrhythmias. 26 The mechanism for this action is by inhibition of K 2p 3.1 resulting in action potential abnormalities as previously demonstrated in mouse and canine ventricular myocytes. 8,19,26 We now propose this effect is compounded by the heterogeneous nature of the expression of K 2p 3.1 channels, which, therefore, is likely to create a marked increase in action potential heterogeneity during and immediately post myocardial infarction. An issue further compounded by the fact acidosis and ischemia is often itself be heterogeneous in nature with the subendocardial region commonly showing greater changes and frequency of issues than the subepicardial region.
While the rat heart differs considerably from the human in terms of size, rate and shape of the cardiac action potential, it remains one of the most widely used models used to study cardiac function, and, in particular, cardiac heterogeneity. Qualitatively, there are similarities in apparent electrical heterogeneity across the left ventricular F I G U R E 4 K 2p 3.1 Current-voltage relationship in the left ventricular free wall. Subepicardial and subendocardial layer-derived myocytes were whole-cell voltage-clamped. A, Cells were voltage-ramped from −50 to 30 mV over 6 seconds with a holding potential of −10 mV (upper line, black), whole-cell currents were recorded (lower line, red). B, Current-voltage relationship in the presence (met) and absence (control) of a specific K 2p 3.1 inhibitor methanandamide. C, Example of the methanandamide-sensitive current-voltage relationship from a single ventricular myocyte. D, Current-voltage relationship from epicardial and endocardial myocytes. E, Average currents at 30 mV, taken as the peak outward current, were shown in the endocardial myocyte to be double that measured in the epicardial myocyte (Students t test, *P < 0.001; Epi n = 7; Endo n = 6) wall between rats and humans with differences in action potential duration between the subendocardial and subepicardial region present in both species, and in each case associated with regional differences in potassium currents. A key difference between rats and humans is the duration that potassium currents act in terms of impacting repolarization. The rat action potential is quick (several to 10 seconds of msec in duration depending on heart rate) and this means only rapidly acting currents with a relatively large capacity for ion flux can significantly alter action potential duration. In contrast, the longer human action potential (100 seconds of msec in duration) gives scope for smaller and slower activating fluxes to influence action potential duration. As such, the impact of I K2p3.1 and moderation of this component could potentially be greater in humans than the rat heart. A potential indicator of this proposed role is the fact that the volatile anesthetic sevoflurane, known to activate K 2p 3.1 channels, precipitates increased dispersion of repolarization in the human heart, an indicator of increasing action potential heterogeneity across the ventricular wall. 27 Such an effect may be due, at least in part, to a comparable heterogeneous distribution of K 2p 3.1 channel expression in the human heart to that which we have now documented in the rat, although future analysis of human ventricular samples would be required to confirm this. K 2p 3.1 channels have previously had limited investigation with respect to clinical issues and outcomes in the ventricle of the heart but the heterogeneity of K 2p 3.1 expression and activity has implications for understanding pharmacology of several commonly used agents. One of the most commonly used class III antiarrhythmia drugs, amiodarone has an inhibitory effect on K 2p 3.1, 28 as do α1 adrenergic agonists. 4 The heterogeneous impact and prolonging effect on the cardiac APD under some conditions may be antiarrhythmogenic, but the implications are that by extension of the effects seen with PAF and other blockers they could also be proarrhythmogenic. Indeed, even amiodarone has been shown to be proarrhythmic in several cases (eg, Stanton et al 29 ), although its polygenic actions complicating direct interpretation of mechanism actions on K 2p 3.1 could be key. In contrast to the numerous agents that can inhibit K 2p 3.1, inhalational anesthetics such as sevoflurane and isoflurane at clinically relevant concentrations can activate human K 2p 3.1 causing significant APD shortening. As mentioned above such effects appear to have a heterogeneous impact, which fits with the current data and has the potential to contribute to arrhythmia formation. 4,27 Such impacts are, however, complex to interpret, particularly under conditions of cardiac surgery and where other interactions with local and circulating inhibitory factors are occurring.

| CONCLUSION
In conclusion, a transmural gradient of K 2p 3.1 protein expression exists across the left ventricular wall with expression increasing from the epicardial to the endocardial surface. This transmural gradient of K 2p 3.1 function is likely to contribute to electrical stability of the left ventricular wall, while being highly susceptible to physiological and pharmacological modulation that can promote arrhythmogenesis.