Regulation of Cardiac IKs Potassium Current by Membrane Phosphatidylinositol 4,5-Bisphosphate*

Regulation of the slowly activating component of delayed rectifier K+ current (IKs) by membrane phospholipid phosphatidylinositol 4,5-bisphosphate (PtdIns-(4,5)P2) was examined in guinea pig atrial myocytes using the whole-cell patch clamp method. IKs was elicited by depolarizing voltage steps given from a holding potential of –50 mV, and the effect of various test reagents on IKs was assessed by measuring the amplitude of tail current elicited upon return to the holding potential following a 2-s depolarization to +30 mV. Intracellular application of 50 μm wortmannin through a recording pipette evoked a progressive increase in IKs over a 10–15-min period to 208.5 ± 14.6% (n = 9) of initial magnitude obtained shortly after rupture of the patch membrane. Intracellular application of anti-PtdIns(4,5)P2 monoclonal antibody also increased the amplitude of IKs to 198.4 ± 19.9% (n = 5). In contrast, intracellular loading with exogenous PtdIns(4,5)P2 at 10 and 100 μm produced a marked decrease in the amplitude of IKs to 54.3 ± 3.8% (n = 5) and 44.8 ± 8.2% (n = 5), respectively. Intracellular application of neomycin (50 μm) or aluminum (50 μm) evoked an increase in the amplitude of IKs to 161.0 ± 13.5% (n = 4) and 150.0 ± 8.2% (n = 4), respectively. These results strongly suggest that IKs channel is inhibited by endogenous membrane PtdIns(4,5)P2 through the electrostatic interaction with the negatively charged head group on PtdIns(4,5)P2. Potentiation of IKs by P2Y receptor stimulation with 50 μm ATP was almost totally abolished when PtdIns(4,5)P2 was included in the pipette solution, suggesting that depletion of membrane PtdIns(4,5)P2 is involved in the potentiation of IKs by P2Y receptor stimulation. Thus, membrane PtdIns(4,5)P2 may act as an important physiological regulator of IKs in guinea pig atrial myocytes.

The delayed rectifier K ϩ current (I K ) 1 is activated by membrane depolarization and thereby provides an outward current, which is essential for initiating phase 3 repolarization of the action potential in cardiac muscle. Two kinetically and phar-macologically distinct components of I K , I Kr (rapid) and I Ks (slow), have been identified in cardiac myocytes from various mammalian species (1,2) including humans (3). The KCNQ1 gene encodes the pore-forming ␣ subunit, K V LQT1, that assembles with an accessory ␤ subunit minK (I sK ) protein (encoded by KCNE1 gene) to produce the I Ks channel (4,5), whereas the HERG (human ether-á -go-go-related gene) product forms the pore-forming subunit of the I Kr channel (6,7). Mutations in any of these genes have been linked to long QT syndrome, an inherited cardiac arrhythmia characterized by abnormal ventricular repolarization and a high risk for sudden cardiac death (8).
Previous studies have demonstrated that both I Ks and I Kr represent relevant targets for the actions of autonomic neurotransmitters, hormones, intracellular messengers, and exogenous drugs. I Ks is modulated by protein kinase A, protein kinase C (PKC), and intracellular free Ca 2ϩ (9 -11), whereas I Kr is sensitive to inhibition not only by methanesulfonanilide drugs but also by a wide range of other medications including some anti-arrhythmic, anti-histamic, antibiotic, and psychoactive agents (8). These modulations of I Kr and I Ks profoundly affect the repolarization process of the cardiac action potential and thereby mediate the regulation of cardiac function by these intracellular signaling molecules or exogenous pharmacological compounds.
Stimulation of many G q -coupled receptors activates phosphoinositide-specific phospholipase C (PLC), leading to the hydrolysis of PtdIns(4,5)P 2 to form inositol 1,4,5-trisphosphate (InsP 3 ) and diacylglycerol. These two second messengers, respectively, activate InsP 3 receptors on intracellular Ca 2ϩ stores and PKC and thereby play a pivotal role in transmitting the extracellular signals to the cellular responses. In addition to serving as a precursor for the production of second messengers, membrane PtdIns(4,5)P 2 has been shown to directly interact with many ion channels and transporters. Hilgemann and Ball (12) revealed for the first time that the Na ϩ /Ca 2ϩ exchanger and ATP-sensitive K ϩ channels are positively regulated by endogenous PtdIns(4,5)P 2 in guinea pig cardiac cell membranes. Recent studies have demonstrated that PtdIns(4,5)P 2 modulates the function of a number of ion channels, including many inwardly rectifying K ϩ channels (13)(14)(15), the HERG K ϩ channel (16) and voltage-gated P/Q-type Ca 2ϩ channels (17).
We have previously shown that, in guinea pig cardiac myocytes, the stimulation of P2Y receptor (G protein-coupled ATP receptor) by extracellular ATP markedly enhances I Ks through a mechanism that appears to be independent of either the activation of PKC or the elevation of intracellular Ca 2ϩ (18,19). Moreover, it was recently demonstrated in guinea pig ventricular myocytes that the stimulation of P2Y receptor evokes a pronounced reduction in the activity of ATP-sensitive K ϩ channels through a PLC-induced depletion of membrane PtdIns(4,5)P 2 (20). Therefore, this study was designed to ex-* This work was supported by Grants-in-aid for Scientific Research 13670042 and 15590184 (to H. M.) from Japan Society for the Promotion of Science. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Isolation of Atrial Myocytes-All of the experiments conformed with
The Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publication number 85-23, revised 1996) and were approved by the institution's Animal Care and Use Committee. Single atrial myocytes were obtained from the heart of adult Hartley guinea pigs using an enzymatic dissociation procedure as described previously (21,22). The hearts were retrogradely perfused with nominally Ca 2ϩ -free Tyrode solution containing 0.4 mg ml Ϫ1 collagenase (Wako Pure Chemical Industries, Osaka, Japan) for 7-10 min through the coronary artery. After the enzyme treatment, single myocytes were dissociated and stored in a high K ϩ , low Cl Ϫ Kraftbrü he solution (22).
Whole-Cell Patch Clamp Technique and Data Analysis-Isolated atrial myocytes were current-and voltage-clamped using the whole-cell configuration of the patch clamp technique (27) with EPC-8 patch clamp amplifier (HEKA, Lambrecht, Germany). Patch electrodes were made from glass capillaries (outer diameter, 1.5 mm; inner diameter, 0.9 mm) using a Sutter P-97 microelectrode puller (Novato, CA), and the tips were then heat-polished with a microforge. These electrodes had a resistance of 1.5-2.5 megohms when filled with pipette solutions and allowed a relatively rapid diffusion of test reagents in pipette solution into the cell interior after rupture of the patch membrane. In all of the experiments, atrial myocytes were exposed to the extracellular solution for measuring I Ks prior to rupture of the patch membrane and the measurement of I Ks was initiated immediately after establishment of the whole-cell mode to study the time course of changes in the amplitude of I Ks during internal dialysis with various test reagents. All experiments were performed at 34 -36°C.
Current and voltage signals were filtered with a low pass 3-kHz filter (48 db/octave) and digitized at a sampling rate of 0.2-1.0 kHz using a 100-kHz analog-to-digital board installed in a personal computer (PC98RL, NEC, Tokyo, Japan). The inset in the figures represents membrane currents or potentials recorded at time points indicated by numerals on the graph, and the zero-current level is indicated to the left of current traces by a horizontal line and the zero-potential level is denoted by the dashed line.
All of the averaged data are presented as the mean Ϯ S.E. with the number of experiments given within parentheses. Statistical comparisons were made using either Student's paired or unpaired t test as appropriate, and differences were considered to be significant at p Ͻ 0.05.

Enhancement of I Ks by Intracellular Dialysis with
Wortmannin-It has been demonstrated that, when applied in micromolar concentrations, wortmannin inhibits the activity of phosphatidylinositol 4-kinase (PtdIns 4-kinase) as well as phosphatidylinositol 3-kinase (PtdIns 3-kinase), resulting in a substantial reduction in the cellular levels of both PtdIns(4,5)P 2 and phosphatidylinositol 4-phosphate (28 -30). To elucidate the action of endogenous membrane PtdIns(4,5)P 2 on I Ks , we first examined the effect of 50 M wortmannin on I Ks in guinea pig atrial myocytes. Fig. 1A demonstrates a representative result showing the time course of changes in the amplitude of I Ks recorded from an atrial myocyte dialyzed with a pipette solution containing 50 M wortmannin. Immediately after gaining access to the cell interior (rupture of the patch membrane), the membrane potential was held at Ϫ50 mV and was then repetitively (every 20 s) depolarized to a test potential of ϩ30 mV for 2 s to evoke I Ks . The amplitude of the I Ks tail current elicited upon repolarization to a holding potential of Ϫ50 mV, which reflects the degree of I Ks activation during the preceding depolarizing step to ϩ30 mV, was progressively increased as wortmannin was dialyzing into the cell and typically reached a maximum response ϳ10 -15 min after rupture of the patch membrane (Fig. 1A). The effect of internal application of wortmannin on I Ks was quantitatively assessed by normalizing the amplitude of I Ks tail current measured ϳ10 -15 min after establishing whole-cell mode with reference to that obtained shortly (Յϳ30 s) after whole-cell mode. In a total of nine myocytes, the addition of 50 M wortmannin to the pipette solution increased the amplitude of I Ks tail current to 208.5 Ϯ 14.6% within a period of 10 -15 min (see also Fig. 3).
In a separate set of experiments, the amplitude of I Ks tail current, measured using the same voltage-clamp protocol (2-s depolarization to ϩ30 mV from a holding potential of Ϫ50 mV applied at a 20-s interval), was found to gradually decrease to 88.5 Ϯ 2.1% (n ϭ 11) of initial magnitude over the same time period (ϳ10 -15 min) when the cell was dialyzed with control pipette solution (see also Figs. 3 and 5A). Furthermore, the time course of changes in I Ks tail current was not appreciably affected by internal dialysis with the PtdIns 3-kinase inhibitor LY-294002 (31) at 10 M (decrease to 89.4 Ϯ 2.5% of initial values, n ϭ 5, Fig. 3) or wortmannin at 50 nM (decrease to 87.1 Ϯ 2.6% of initial values; n ϭ 7, Fig. 3), a concentration that was demonstrated to fully inhibit PtdIns 3-kinase with a relatively small effect on PtdIns 4-kinase (28). Similarly, the myosin light chain kinase inhibitor ML-7 (5 M) also did not mimic the stimulatory action of 50 M wortmannin on I Ks over the same time period (decrease to 88.2 Ϯ 2.1% of initial values, n ϭ 5, Fig. 3). Taken together, it seems most reasonable to assume that a progressive increase in I Ks observed during internal dialysis with 50 M wortmannin arises primarily from the inhibition of PtdIns 4-kinase, but not from the inhibition of the PtdIns 3-kinase or myosin light chain kinase, by this compound.
To evaluate whether the voltage dependence of I Ks activation is affected by internally applied wortmannin, depolarizing voltage steps to various test potentials between Ϫ40 and ϩ50 mV were applied from a holding potential of Ϫ50 mV shortly (within ϳ1 min) and approximately 15 min after a rupture of the patch membrane with pipette solution containing 50 M wortmannin (Fig. 1B). I Ks recorded within ϳ1 min of patch rupture (Fig. 1B, left panel) was expected to be little, if any, affected by internally applied wortmannin and was therefore regarded as representing control traces. In this example, I Ks , which was recorded ϳ15 min after patch rupture (Fig. 1B, right panel), exhibited a steady-state response to wortmannin, which was confirmed by monitoring the changes in the amplitude of I Ks in response to repetitive depolarizing voltage steps to ϩ30 mV (data not shown). Fig. 1C shows I-V relationships for I Ks tail current recorded under these two conditions (open circles, within ϳ1 min after patch rupture; filled circles, ϳ15 min after patch rupture). The amplitude of I Ks tail current was increased with internal dialysis of 50 M wortmannin by a factor of more than two at each test potential. The smooth curves through the data points ( Fig.  1C) represent the least-squares fit to a Boltzmann equation shown in Equation 1, where I K,tail,max is the fitted maximal tail current amplitude, V1 ⁄2 is the half-maximal voltage, V m is the test potential, and k is the slope factor. The V1 ⁄2 and k values were 5.5 Ϯ 2.4 and 11.5 Ϯ 1.4 mV, respectively, for the data obtained within ϳ1 min after patch rupture and 3.3 Ϯ 1.0 and 11.6 Ϯ 1.8 mV for the data recorded ϳ15 min after patch rupture (n ϭ 4), thus showing that the voltage dependence of I Ks activation was not significantly affected during internal dialysis with 50 M wortmannin.
Effects of Anti-PtdIns(4,5)P 2 Monoclonal Antibody, PtdIns-(4,5)P 2 , and PtdIns(4,5)P 2 plus Wortmannin on I Ks -To assess whether a possible reduction in the level of membrane PtdIns(4,5)P 2 associated with an addition of wortmannin (50 M) resulted in a marked increase in I Ks (Fig. 1), we examined the effects of the addition of anti-PtdIns(4,5)P 2 monoclonal antibody (32) and exogenous PtdIns(4,5)P 2 on I Ks in guinea pig atrial myocytes (Fig. 2). Previous patch clamp experiments have clearly demonstrated that anti-PtdIns(4,5)P 2 monoclonal antibody added to cytoplasmic solutions in either inside-out or whole-cell mode can prevent the PtdIns(4,5)P 2 interaction with several types of ion channel proteins by specifically binding to endogenous PtdIns(4,5)P 2 (14 -17). Fig. 2A demonstrates a representative example of the I Ks response to internal application of the 40-fold diluted anti-PtdIns(4,5)P 2 monoclonal antibody. The amplitude of I Ks tail current elicited upon repolarization to Ϫ50 mV following 2-s depolarization to ϩ30 mV was progressively increased to 198.4 Ϯ 19.9% (n ϭ 5, Fig. 3) of its initial magnitude over a period of 10 -15 min after patch rupture. In contrast, the addition of control mouse IgG to the pipette solution did not appreciably affect the time course of changes in the amplitude of I Ks tail current (decrease to 87.9 Ϯ 2.6% of initial values after 15 min of whole-cell mode, n ϭ 5, Fig. 3). These results indicate that blocking the action of endogenous PtdIns(4,5)P 2 leads to a marked potentiation of I Ks .
We then tested the effects of exogenous PtdIns(4,5)P 2 at concentrations of 10 and 100 M on I Ks . The amplitude of I Ks tail current in myocytes loaded intracellularly with 10 and 100 M PtdIns(4,5)P 2 was monotonically declined to 54.2 Ϯ 3.8% (n ϭ 5) and 44.8 Ϯ 8.2% of initial magnitude (n ϭ 5; Figs. 2B and 3), respectively. Thus, I Ks was found to be significantly decreased by exogenously applied PtdIns(4,5)P 2 , consistent with the view that I Ks is highly sensitive to inhibition by PtdIns(4,5)P 2 .
To confirm that a marked potentiation of I Ks induced by 50 M wortmannin (Fig. 1) was indeed evoked by substantial depletion of membrane PtdIns(4,5)P 2 , we examined the effect on I Ks of wortmannin applied together with exogenous PtdIns(4,5)P 2 (100 M) to the cell (Fig. 2C). Under such a condition, the amplitude of I Ks tail current was found to be monotonically reduced to 38.9 Ϯ 8.4% of initial magnitude (n ϭ 6), which is very similar to the degree of reduction observed in the presence of exogenous PtdIns(4,5)P 2 (100 M) alone (Fig. 3), thus showing that the concomitant presence of exogenous PtdIns(4,5)P 2 almost totally disrupts the ability of wortmannin to potentiate I Ks . Therefore, it seems conceivable that a marked potentiation of I Ks in the presence of wortmannin (50 M) was mediated primarily through substantial decrease of membrane PtdIns(4,5)P 2 . These results provide functional evidence to support the view that endogenous membrane PtdIns(4,5)P 2 produces a tonic and potent inhibitory action on I Ks in guinea pig atrial myocytes. We also examined the effect of the PLC inhibitor U-73122 on I Ks in guinea pig atrial myocytes. The addition of 10 M U-73122 to the pipette solution did not discernibly affect the time course of changes in I Ks over a 10 -15 min after establishment of whole-cell mode (decrease to 85.1 Ϯ 4.6% of initial values, n ϭ 5; Fig. 3), thus suggesting that the inhibition of the basal activity of PLC does not lead to a potentiation of I Ks in guinea pig atrial myocytes. This result appears to be consistent with a previous study demonstrating that basal PtdIns(4,5)P 2 levels are not appreciably affected by PLC inhibition with 20 M U-73122 in guinea pig atria (33).
To test whether the effectors downstream of PtdIns(4,5)P 2 are involved in regulation of I Ks , the effect of inositol polyphosphates InsP 3 , InsP 5 , and InsP 6 was examined by directly adding to the pipette solution. None of these inositol polyphosphates at a concentration of 50 M was found to affect the time course of changes in I Ks after establishment of whole-cell mode (decrease to 94.2 Ϯ 6.3, 90.2 Ϯ 9.9, and 83.4 Ϯ 11.7% of initial values by InsP 3 , InsP 5 and InsP 6 , respectively, n ϭ 5-7, Fig. 3), thus suggesting that I Ks regulation by PtdIns(4,5)P 2 is not mediated through the action of InsP 3 , InsP 5 , or InsP 6 .
Involvement of Anionic Head of PtdIns(4,5)P 2 in the Enhancement of I Ks -It has been demonstrated that the polyvalent cation neomycin and trivalent cation aluminum bind with a high affinity to a negatively charged head group on PtdIns(4,5)P 2 and thereby disrupt electrostatic interactions of membrane PtdIns(4,5)P 2 with ion channel and transporter proteins (12)(13)(14). If endogenous PtdIns(4,5)P 2 in the plasma membrane produces an inhibitory action on I Ks channels through such an electrostatic mechanism, neutralizing the negative charges in PtdIns(4,5)P 2 should result in an enhancement of I Ks . To test this assumption, we examined the effects of internal application of neomycin (50 M) or AlCl 3 (50 M) on I Ks . As demonstrated in Fig. 4, A and C, the amplitude of I Ks elicited by 2-s depolarizations to various test potentials between Ϫ40 and ϩ50 mV was significantly enhanced by the addition of either neomycin or AlCl 3 to the pipette filling solution. Intracellular application of neomycin (50 M) and AlCl 3 (50 M) resulted in an enhancement of I Ks tail current by 61.0 Ϯ 13.5% (n ϭ 4) and 50.0 Ϯ 8.2% (n ϭ 4), respectively, when evaluated by the changes in the amplitude elicited upon repolarization to Ϫ50 mV following 2-s depolarization to ϩ30 mV. As illustrated in Fig. 4, B and D, the voltage dependence of I Ks activation was not significantly shifted by either neomycin or AlCl 3 . In the neomycin application group, V1 ⁄2 and k values averaged 7.8 Ϯ 2.7 and 11.8 Ϯ 1.6 mV, respectively, for the data obtained within ϳ1 min after patch rupture and 6.9 Ϯ 2.5 and 12.1 Ϯ 1.9 mV for the data recorded ϳ10 min after patch rupture (n ϭ 4). In AlCl 3 application group, V1 ⁄2 and k values were 8.1 Ϯ 2.8 and 12.1 Ϯ 1.8 mV, respectively, for the data obtained within ϳ1 min after patch rupture and 6.7 Ϯ 1.8 and 12.0 Ϯ 1.6 mV for the data recorded ϳ10 min after patch rupture (n ϭ 4). Dialysis of the cell interior with neomycin or AlCl 3 was thus found to be effective at increasing the amplitude of I Ks without appreciably affecting the voltage dependence of current activation.
Involvement of PtdIns(4,5)P 2 Depletion in Extracellular ATPinduced I Ks Potentiation-The results so far represented strongly indicate that plasma membrane PtdIns(4,5)P 2 exerts a FIG. 2. Effects of anti-PtdIns(4,5)P 2 monoclonal antibody, PtdIns(4,5)P 2 or PtdIns(4,5)P 2 plus wortmannin on I Ks . Atrial myocytes loaded with the 40-fold diluted anti-PtdIns(4,5)P 2 monoclonal antibody (A), 100 M PtdIns(4,5)P 2 (B), or 50 M wortmannin plus 100 M PtdIns(4,5)P 2 (C) through a recording pipette were depolarized every 20 s from a holding potential of Ϫ50 to ϩ30 mV for 2 s. The amplitude of the tail current elicited upon return to the holding potential was plotted as a function of time after gaining access to the cell interior. potent inhibitory action on I Ks and that disruption of PtdIns(4,5)P 2 interaction with I Ks channels results in a marked increase in I Ks in guinea pig atrial myocytes. It has previously been demonstrated in guinea pig cardiac myocytes that the activation of P2Y receptor stimulates PLC and thereby depletes plasma membrane PtdIns(4,5)P 2 that is required for maintaining the activity of the ATP-sensitive K ϩ channels (20). We next checked whether the possible reduction of PtdIns(4,5)P 2 levels mediates potentiation of I Ks associated with P2Y receptor stimulation (18,19). As demonstrated in Fig.  5A, the amplitude of I Ks current usually underwent some rundown (10 -20% of initial magnitude, see also Fig. 3) over the first 5-10 min following establishment of whole-cell mode with control pipette solution. After this current rundown was allowed to reach a steady-state level, the cell was then exposed to 50 M ATP, which typically evoked an increase in amplitude of I Ks by 103.5 Ϯ 10.6% (n ϭ 5; Fig. 5D) when evaluated by monitoring the changes in tail current amplitude elicited upon repolarization to Ϫ50 mV following 2-s depolarization to ϩ30 mV. In guinea pig atrial myocytes, increasing the concentration of ATP above 50 M caused no further increase in I Ks (18), showing that 50 M ATP evokes a maximal enhancement of I Ks .
Previous whole-cell patch clamp experiments have demonstrated that suppression of the muscarinic K ϩ channels by ␣ 1 -adrenergic agonist phenylephrine is greatly attenuated when PtdIns(4,5)P 2 is included in the pipette solution (34), thus indicating that substantial reductions of plasma membrane PtdIns(4,5)P 2 associated with receptor-mediated PLC activation can be effectively compensated by exogenous PtdIns(4,5)P 2 applied through a recording pipette. Therefore, we checked whether extracellular ATP at a maximally effective concentration (50 M) can increase the amplitude of I Ks in atrial myocytes loaded with 100 M PtdIns(4,5)P 2 (Fig. 5B) where membrane PtdIns(4,5)P 2 level is expected to be kept relatively stable during PLC activation. In these myocytes, as expected, a progressive and marked decrease in the amplitude of I Ks was consistently observed following rupture of the patch membrane. The effect of bath application of ATP was then tested after the declining response reached a steady-state level (10 -15 min after whole-cell mode, see also Fig. 2B). In a total of five myocytes loaded with 100 M PtdIns(4,5)P 2 , ATP at 50 M increased the amplitude of I Ks tail current only by 16.1 Ϯ 9.9% (Fig. 5D), which is significantly smaller than the control response. The stimulatory effect of extracellular ATP was thus largely abolished by the exogenously applied PtdIns(4,5)P 2 . This result strongly suggests that a reduction of PtdIns(4,5)P 2 levels in the plasma membrane is primarily involved in the potentiation of I Ks associated with the stimulation of P2Y receptor.
We also checked whether extracellular ATP can further enhance I Ks in wortmannin-treated myocytes in which membrane PtdIns(4,5)P 2 level is expected to be substantially reduced (30). As demonstrated in Fig. 5C, extracellular ATP potentiated the amplitude of I Ks tail current by 27.7 Ϯ 12.4% (n ϭ 6; Fig. 5D) in myocytes loaded with 50 M wortmannin, thus showing that stimulatory effect of extracellular ATP was partly but not totally abolished in the presence of wortmannin. Assuming that endogenous membrane PtdIns(4,5)P 2 was reduced to some extent by treatment with wortmannin (30), this result was also consistent with the view that extracellular ATP potentiates I Ks through a reduction of membrane PtdIns(4,5)P 2 .

Effect of I Ks Potentiation by Extracellular ATP on Action Potentials in Guinea Pig Atrial
Myocytes-It has previously been demonstrated that bath application of ATP at micromolar concentrations significantly shortens the action potential duration (APD) in guinea pig atrial myocytes (35,36). It is assumed that the APD shortening by extracellular ATP is primarily mediated through its stimulatory action on two distinct K ϩ channels, namely, I Ks (18,19) and the muscarinic K ϩ channels (35)(36)(37)(38). Whereas potentiation of I Ks by extracellular ATP remains rather stable over a period of at least 5 min during exposure to the agonist, activation of I K,ACh by micromolar concentrations of ATP is characterized by its rapid decay to the pre-agonist (control) levels within 1-2 min despite the continued presence of the agonist (36 -38). It was previously suggested that this transient nature of I K,ACh activation during exposure to extracellular ATP is due to a dual action of ATP, namely, a rapid activation via a membrane-delimited mecha-

FIG. 3. Summary of I Ks response to internal dialysis with various test reagents. I Ks was repetitively activated by 2-s depolarization to
ϩ30 mV from a holding potential of Ϫ50 mV immediately after patch rupture with a control pipette solution or the pipette solution containing various test reagents as indicated. The amplitude of tail current elicited upon return to the holding potential at a steady-state response was normalized with reference to that obtained briefly after patch rupture. The columns and bars denote the means Ϯ S.E., respectively. **, p Ͻ 0.01 compared with control group. nism involving a pertussis toxin-sensitive G protein G K and a gradual inhibition via a pertussis toxin-insensitive G protein (36 -38).
To evaluate the contribution of I Ks potentiation by extracellular ATP to the shortening of APD in guinea pig atrial myocytes, the effect of extracellular ATP on APD measured at 90% repolarization (APD 90 ) was compared in the absence and presence of the selective I Ks blocker chromanol 293B. Fig. 6, A and  B, demonstrates representative examples for the changes in APD 90 during exposure to 10 M ATP under control conditions and in the presence of 293B at 30 M, a concentration reported to largely abolish cardiac I Ks with little if any effect on other membrane currents (23,24). Under control conditions (n ϭ 5), bath application of 10 M ATP rapidly and markedly decreased APD 90 from a control value of 66.1 Ϯ 5.2 to 19.9 Ϯ 1.6 ms within ϳ30 s of the application, which was then gradually attenuated to a steady level of 35.4 Ϯ 2.9 ms in the continued presence of the agonist (Fig. 6, A and C). In a separate set of experiments (n ϭ 5), atrial myocyte was initially exposed to 30 M 293B, which significantly prolonged APD 90 to 86.8 Ϯ 5.1 ms, presumably through an almost total inhibition of I Ks (Fig.  6, B and D). Further addition of 10 M ATP in the presence of 30 M 293B rapidly shortened APD 90 to 19.6 Ϯ 2.2 ms, which was consistently followed by a gradual recover to the steadystate level (87.8 Ϯ 6.9 ms) close to the pre-agonist levels (86.8 Ϯ 5.1 ms) within 1-2 min (Fig. 6, B and D). Thus, with the blockade of I Ks , the APD shortening evoked by extracellular ATP was characteristically transient. These observations suggest that, whereas the transiently activated I K,ACh during exposure to extracellular ATP (36 -38) mainly contributes to APD shortening observed within ϳ1 min of ATP application (initial phase), APD shortening which remains thereafter (late phase) can be primarily ascribed to the potentiation of I Ks (Fig. 6A). Therefore, we would suggest that the I Ks potentiation by extracellular ATP underlies a significant shortening of APD observed at the late phase of ATP application (ϳ46% shortening by 10 M ATP, Fig. 6C). DISCUSSION The present result that I Ks in guinea pig atrial myocytes is markedly potentiated by the application of anti-PtdIns(4,5)P 2 monoclonal antibody but is reduced by the addition of exogenous PtdIns(4,5)P 2 (Fig. 2, A and B) strongly suggests that endogenous membrane PtdIns(4,5)P 2 exerts an inhibitory action on I Ks . Polyvalent cations (such as neomycin) that preferentially neutralize the anionic head of PtdIns(4,5)P 2 were previously demonstrated to antagonize the actions of PtdIns(4,5)P 2 on ion channels or transporters (12)(13)(14), suggesting a role for electrostatic interactions of PtdIns(4,5)P 2 with ion channel or transporter proteins. In fact, site-directed mutagenesis studies on inwardly rectifying K ϩ channel subunits have identified several sites (basic residues) that are important for interaction with membrane PtdIns(4,5)P 2 , particularly within the proximal C terminus (13,14). The present finding that the addition of neomycin or aluminum results in a marked potentiation of I Ks (Fig. 4) also supports the view that an inhibitory action of PtdIns(4,5)P 2 is mediated through electrostatic interactions with basic residues in I Ks channel proteins. Whereas amino acid sequences of I Ks channels expressed in guinea pig heart are presently unknown, similarities in electrophysiological and pharmacological properties of I Ks between humans and guinea pigs (24) may indicate a high sequence homology of the channel protein between these two species. It should be noted that there are some basic residues cluster in the cytoplasmic faces of the C terminus in human KvLQT1 (e.g. 225 KVQQKQRQKHFNR 237 and 282 KPKKSVVVKKKKFKLDK 298 ) (4, 5), which might be candidates for sites of interaction with PtdIns(4,5)P 2 . Previous studies have provided strong evidence to suggest that membrane PtdIns(4,5)P 2 is fundamental for the activation of inwardly rectifying K ϩ channels (12)(13)(14)(15) and some of voltage-dependent K ϩ channels (16,39,40). The present study identifies the presence of the K ϩ channel that is inhibited by membrane PtdIns(4,5)P 2 , and thus, it appears that PtdIns(4,5)P 2 acts as both stimulatory and inhibitory factors in the regulation of K ϩ channel function.
It has been demonstrated in bovine adrenal glomerulosa cells that PtdIns 4-kinase, which regulates the synthesis of hormone-sensitive pools of polyphosphoinositides, is susceptible to inhibition by micromolar concentrations of wortmannin (28). This wortmannin-sensitive isoform of PtdIns 4-kinase (PtdIns 4-kinase ␤) has been demonstrated to be widely expressed in a variety of tissues including the heart (29). The present result that the stimulatory action of 50 M wortmannin on I Ks is completely abolished by the concomitant presence of exogenous PtdIns(4,5)P 2 ( Fig. 2C and 3) strongly suggests that potentiation of I Ks evoked by wortmannin arises primarily from a reduction in the membrane PtdIns(4,5)P 2 levels, probably because of the inhibition of PtdIns 4-kinase. It has been shown in SH-SY5Y human neuroblastoma cells that the blockade of PtdIns 4-kinase by 10 M wortmannin alone results in a substantial reduction in the level of PtdIns(4,5)P 2 as well as phosphatidylinositol 4-phosphate in the plasma membrane under basal conditions (30). Furthermore, incubation of these cells with 10 M wortmannin for 10 min markedly attenuates transient peaks of both InsP 3 accumulation and intracellular Ca 2ϩ elevation evoked by subsequent stimulation of M 3 -muscarinic receptor. These data support the view that micromolar concentrations of wortmannin not only inhibits the replenishment of PtdIns(4,5)P 2 following PLC-coupled receptor-mediated depletion (30,41,42) but also substantially reduces membrane PtdIns(4,5)P 2 pool even under basal conditions through the inhibition of wortmannin-sensitive isoform of PtdIns 4-kinase (30). The present finding that extracellular ATP is still effective at potentiating I Ks in wortmannin-treated atrial myocytes (Fig.  5C) may reflect partial (but not total) depletion of membrane PtdIns(4,5)P 2 associated with the inhibition of I Ks .
Whereas most metabotropic P2Y receptors are coupled to PLC through the pertussis toxin-insensitive G protein G q , leading to the hydrolysis of PtdIns(4,5)P 2 to yield diacylglycerol and InsP 3 (43), an enhancement of I Ks through P2Y receptor stimulation in guinea pig atrial myocytes was not appreciably affected by the presence of either 10 M H-7 or 20 mM BAPTA (1,2-bis(2-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid), thus suggesting that activation of PKC and elevation of intracellular Ca 2ϩ are not primarily involved (18,19). It has recently been shown in guinea pig ventricular myocytes that extracellular ATP depletes membrane PtdIns(4,5)P 2 through the stimulation of the G q -PLC pathway, leading to reductions of the activity of ATP-sensitive K ϩ channels (20). The present result that the addition of exogenous PtdIns(4,5)P 2 greatly prevents the stimulatory effect of extracellular ATP on I Ks (Fig. 5B) is consistent with the view that this stimulatory action of ATP is mediated through reductions in membrane PtdIns(4,5)P 2 . Thus, alterations of membrane PtdIns(4,5)P 2 levels via P2Y receptor stimulation appear to represent an effective signaling mechanism for regulation of I Ks in guinea pig cardiac myocytes.
However, it was previously demonstrated in guinea pig ventricular myocytes that endothelin enhances I Ks via PLC-mediated PKC activation and intracellular Ca 2ϩ mobilization (44). It was also shown that ␣ 1 -adrenceptor-induced potentiation of I Ks is mediated through PKC activation in the same cell types (45). Thus, P2Y receptors, ␣ 1 -adrenergic receptors, and endothelin receptors seemed to be all coupled to the G q -PLC pathway to potentiate I Ks , but signaling molecule associated with P2Y receptors appeared to be distinct from that associated with ␣ 1 -adrenergic receptors or endothelin receptors. Although the precise mechanism for this difference remains unknown, it was recently demonstrated in mouse atrial myocytes that the muscarinic K ϩ current is greatly reduced by the membrane PtdIns(4,5)P 2 depletion induced by ␣ 1 -adrenergic receptors (through PLC activation) but is not significantly affected by the PtdIns(4,5)P 2 reduction evoked by PLC-coupled M 1 (or M 3 /M 5 ) muscarinic receptors (46). This observation strongly suggests that the muscarinic K ϩ channel is preferentially co-localized with ␣ 1 -adrenergic receptor but not with M 1 (or M 3 /M 5 ) muscarinic receptor. Because the concentration of PtdIns(4,5)P 2 has been suggested to change locally within the plasma membrane (47), PtdIns(4,5)P 2 metabolism coupled to P2Y receptor It has been demonstrated in guinea pig cardiac myocytes that raising the intracellular Ca 2ϩ concentrations over 0.1 M evokes significant increase in the amplitude of I Ks (9,11). This intracellular Ca 2ϩ -dependent enhancement of I Ks has been assumed to play an important role in limiting an amount of Ca 2ϩ entry in Ca 2ϩ -overloaded myocardium. However, the cellular mechanism underlying this Ca 2ϩ -dependent enhancement of I Ks has yet to be fully characterized. It was previously shown that PLC in cardiac membranes is activated by intracellular free Ca 2ϩ at concentrations of Ն0.5 M (12). Thus, the potentiation of I Ks and activation of PLC are evoked by intracellular free Ca 2ϩ within a relatively similar concentration range. Therefore, it is reasonable to assume that an elevation of intracellular Ca 2ϩ levels within physiological range (Յϳ1-2 M) is accompanied by the activation of PLC, which should result in substantial depletion of membrane PtdIns(4,5)P 2 and subsequent potentiation of I Ks . Xie et al. (48) have suggested that intracellular Ca 2ϩ -induced rundown of the ATP-sensitive K ϩ channels is evoked by the Ca 2ϩ -dependent activation of PLC and resultant depletion of membrane PtdIns(4,5)P 2 (48). Alternatively, it is probable that intracellular Ca 2ϩ interrupts the electrostatic interaction between membrane PtdIns(4,5)P 2 and I Ks channels (48), which is again expected to result in potentiation of I Ks . Thus, it is of interest to elucidate whether such a PtdIns(4,5)P 2 -dependent mechanism is indeed involved in the intracellular Ca 2ϩ -evoked I Ks potentiation in cardiac myocytes.
In recent years, it has been shown that the KCNQ1/KCNE1 heteromeric channel, the molecular constituents of cardiac I Ks (4,5), is activated by exogenously applied PtdIns(4,5)P 2 (40,49) and is drastically suppressed by the PtdIns(4,5)P 2 -neutralizing polyvalent cations (40) as well as the receptor-mediated PLC activation (50). These observations appear to be apparently in contrast with the present results. To date, there has been little convincing evidence supporting the view that the I Ks channel is inhibited by the receptor-mediated activation of PLC in native cardiac myocytes (44). It may be probable that the native and recombinant ion channels behave differently to various exper- were applied as marked by horizontal bars. C and D, summarized data for changes in APD 90 by exposure to 10 M ATP in the absence (C, n ϭ 5) and presence (D, n ϭ 5) of 30 M 293B. APD 90 was measured during exposure to ATP at initial (within 30 s of application) and late (ϳ 2-3 min after application) phases. The columns and bars denote means Ϯ S.E., respectively. APD 90 at both initial and late phases of ATP application is significantly shorter compared with control (p Ͻ 0.01, C). APD 90 during exposure to 293B alone is significantly longer compared with control (p Ͻ 0.05) but is similar (p ϭ 0.69) to that at late phase of ATP application in the presence of 293B (D). imental stimuli. It has recently been shown that the application of 20 M PtdIns(4,5)P 2 increases the recombinant KCNQ2/3 heteromeric current but that even a higher concentration (100 M) produces at best a weak effect on the native M current (51). Alternatively, under the present experimental conditions, membrane PtdIns(4,5)P 2 level might change within a relatively modest range when various compounds are applied through a recording pipette in the whole-cell configuration and the present results do not necessarily rule out the possibility that I Ks can be inhibited by a drastic decrease in membrane PtdIns(4,5)P 2 levels, as expected for the stimulation of heterologously expressed G q -PLC-linked membrane receptors.