Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Tertiapin is a peptide composed of 21 amino acids (Fig. 1) that was initially isolated from venoms of the European honey bee, Apis mellifera, more than 20 years ago (Gauldie et al., 1976). This peptide has six positively charged residues in a molecule, four of which are clustered within the C-terminal half of the polypeptide chain. Tertiapin completely lacks negatively charged residues, and the lysine at the carboxyl end is in an amide form. Four cysteines in the peptide form two disulfide bonds. NMR spectroscopy revealed that tertiapin has a type-I reverse turn and an -helix, which are connected by a loop of an extended -sheet (Xu and Nelson, 1993). The three-dimensional structure of tertiapin is highly compact due to extensive side chain interaction. Different from other bee peptidyl toxins, such as apamin and mast cell-degranulating peptide, the biological target of tertiapin had not been known until Jin and Lu (1998) found that tertiapin potently blocks certain types of recombinant inwardly rectifying K⁺ (Kir) channels. Although there are many known peptidyl toxins blocking the voltage-dependent K⁺ channels (Adams and Swanson, 1996), tertiapin is the sole potent Kir channel blocker identified so far.

View larger version (5K): [in this window] [in a new window]   Fig. 1.   Amino acid sequence of tertiapin. The amino acid sequence of tertiapin is indicated with the single-letter code. The lines connecting cysteines represent the disulfide bond arrangement.

Kir channels are highly K⁺-selective ion channels that carry K⁺ ions from extracellular to intracellular space more efficiently than in the other direction (Nichols and Lopatin, 1997). This inwardly rectifying property allows the channels to fix the cell resting membrane potential toward the K⁺ equilibrium potential (E_K) (~85 mV under physiological condition) without preventing action potential generation (Hille, 1992). There are multiple distinct types of Kir channels that differ in the extent of inward rectification. Some of them completely shut outward K⁺ currents at the membrane potential more positive than E_K +60 mV (strong inward rectifiers), whereas others permit large outward currents even at E_K +200 mV (weak inward rectifiers). Kir channels are also diverse in the regulation of activity. Some Kir channels are constitutively active, but others are regulated by various intracellular signaling molecules, such as G proteins, nucleotides, protein kinase/phosphatase, and anchoring proteins (Yamada et al., 1998).

Kir channels play a particularly important role in such a cell type as cardiac myocytes, where the resting potential is ~E_K, whereas the action potential duration is as long as several hundred milliseconds. Cardiac myocytes possess three distinct types of Kir channels (Isomoto et al., 1997). The I_K1 channel is a constitutively active, strong inward rectifier and determines the basic electrophysiological property of working ventricular myocytes and some atrial myocytes. Nodal, atrial, and Purkinje myocytes have another inward rectifier, muscarinic K⁺ (K_ACh) channel, whose inward rectification is slightly weaker than that of the I_K1 channel (Yamada et al., 1998). K_ACh channels have a low basal activity and are activated by acetylcholine (ACh), adenosine, and sphingosine-1-phosphate. Receptors recognizing these substances activate the pertussis toxin-sensitive heterotrimeric G protein, which in turn directly activates the channels through its -subunit. This signaling pathway mediates the negative chronotropic and dromotropic effects caused by the substances. The third Kir channel is a weak inward rectifier, the ATP-sensitive K⁺ (K_ATP) channel (Terzic et al., 1995). This channel is completely inhibited by millimolar concentrations of intracellular ATP in normal cardiac myocytes. When the ATP level decreases on ischemia, the channel is activated and evokes K⁺ currents, which markedly shorten the action potential duration to attenuate contraction.

Differential expression and regulation of these Kir channels in cardiac myocytes are fundamental requirements for the physiological heart beat. However, the lack of specific blockers of these channels has impeded the direct demonstration of the role of the I_K1 and K_ACh channels in the cardiac physiology. Thus, we examined the effect of tertiapin on ion channel currents in cardiac myocytes. We show here that tertiapin potently and selectively blocks K_ACh channels. Tertiapin is, therefore, a useful pharmacological tool to identify the functional role of K_ACh channels in the cardiac physiology and pathophysiology.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Isolation. Male Japanese-White rabbits weighing 1.5 to 2.5 kg were treated with heparin sodium (200 U/kg b.wt.) injected via an ear vein. Approximately 20 min later, the rabbits were anesthetized with pentobarbital sodium (30 mg/kg b.wt.) applied through a vein in another ear. As soon as rabbits completely lost a nociceptive response, the heart was quickly removed and mounted on the Langendorff apparatus. The heart was immediately reperfused with the oxygenated control bathing solution (for composition, see later) through coronary arteries at 37°C. When the heart exhibited a stable sinus rhythm, the solution was switched to the oxygenated nominally Ca²⁺-free Tyrode's solution (for composition, see later). When the heart completely ceased contraction, the solution was exchanged to the same solution containing 0.04% (w/w) collagenase. After 5 to 10 min of digestion, the heart was perfused with 60 ml of KB solution (for composition, see later) and stored in the fresh KB solution at 4°C until use.

Electrophysiological Measurement. A small piece of tissue was excised from the digested myocardium with fine scissors and gently shaken in the control bathing solution in an experimental chamber mounted on the stage of a vertical microscope (Axiovert S100; Carl Zeiss, Jana, Germany). When myocytes dissociated from the shaken tissue stuck to the glass bottom of the chamber, the control bathing solution was perfused through a gravity-feeding apparatus connected to the chamber.

(1)

Solutions. The control bathing solution contained 136.5 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl₂, 0.53 mM MgCl₂, 5.5 mM glucose, and 5.5 mM HEPES (pH adjusted to 7.4 with NaOH). Nominally Ca²⁺-free Tyrode's solution was the same in composition as the control bathing solution except that no CaCl₂ was added. The KB solution contained 10 mM taurine, 10 mM oxalic acid, 70 mM glutamic acid, 25 mM KCl, 10 mM KH₂PO₄, 11 mM glucose, 0.5 mM EGTA, and 10 mM HEPES (pH adjusted to 7.3 with KOH). The internal solution contained 140 mM KCl, 2 mM MgCl₂, 5 mM EGTA, and 5 mM HEPES (pH adjusted to 7.4 with KOH). The concentration of free Mg²⁺ in this solution was calculated to be 1.4 mM. When nucleotides such as ATP and GTP were added to the internal solution, MgCl₂ was supplemented to maintain the free Mg²⁺ concentration. The K⁺ external solution contained 140 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, and 5 mM HEPES (pH adjusted to 7.4 with KOH).

Chemicals. Tertiapin was a kind gift from Peptide Institute Inc. (Osaka, Japan). Acetylcholine chloride, ATP (dipotassium salt), and GTPS were purchased from Sigma Chemical Co. (St. Louis, MO). Pentobarbital sodium was obtained from Dainippon Pharmaceutical Co. Ltd. (Osaka, Japan). Heparin sodium was obtained from Aventis (Frankfurt, Germany). Collagenase was obtained from Yakult (Tokyo, Japan). GTP and all other chemicals were purchased from Wako Pure Chemical Industries (Osaka, Japan).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Tertiapin on Ion Channel Currents in Rabbit Atrial Myocytes. We measured membrane currents from a rabbit atrial myocyte in the whole-cell configuration (Fig. 2A). From the holding potential of 40 mV, 500-ms voltage pulses between 100 and +40 mV were applied to the cell in a 10-mV increment every 5 s. In the control condition (a), the hyperpolarizing voltage steps to potentials more negative than 70 mV elicited inward I_K1 currents (Sakmann and Trube, 1984; Kurachi, 1985). The currents were promptly activated and then slightly decreased. On repolarization to 40 mV, a large inward Na⁺ current was evoked (Brown et al., 1981). When the membrane was depolarized to the potentials more positive than 40 mV, the slow inward Ca²⁺ (I_Ca) current appeared (McDonald et al., 1994), which was followed by gradually developing outward voltage-dependent K⁺ (I_K) currents (Sanguinetti and Jurkiewicz, 1990). Repolarization to 40 mV from these potentials elicited the outward I_K tail currents. The current-voltage (I-V) relationships measured at the beginning and the end of each step are shown in Fig. 2B (closed and open circles, respectively). At potentials negative to 40 mV, both the I-V curves inwardly rectified due to the I_K1 currents. At potentials positive to 40 mV, the instantaneous I-V curve exhibited the U shape with a peak at 0 mV reflecting the inward L-type I_Ca current. On the other hand, the steady-state I-V curve gradually increased in the positive direction as the membrane potential was depolarized. This was due to the I_K currents.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Effect of tertiapin on ion channel currents in a rabbit atrial myocyte. A, in the whole-cell configuration, membrane currents were measured from a rabbit atrial myocyte under the control condition (a), 80 s after the application of ACh (1 µM) (b), 60 s after the application of ACh plus tertiapin (100 nM) (c), and 73 s after the washout of ACh (d). The pipette was filled with internal solution containing 3 mM ATP and 100 µM GTP. Under each condition, 500-ms voltage pulses between 100 and +40 mV were applied from 40 mV in 10-mV increments every 5 s. Arrowheads and dotted lines indicate the zero current level. A leak and capacitative currents were not subtracted. B, I-V relationship measured from the experiment in A. Filled and open symbols indicate the current amplitudes at the beginning and end of each voltage step, respectively.

Effect of Tertiapin on K_ATP Channels in Rabbit Ventricular Myocytes. Figure 3 depicts the effect of tertiapin on another Kir channel in cardiac myocyte, the K_ATP channel. These traces were recorded from a ventricular myocyte. In the ventricular myocyte, there was a much larger I_K1 current than in atrial myocytes (cf. Fig. 2A, a). Under the control condition, the steady-state I-V curve (Fig. 3B, open circles) was in an N shape due to the strong inward rectification of I_K1 channels (Kurachi, 1985). Pinacidil (100 µM), an activator of K_ATP channels, induced large time-independent outward currents at potentials more positive than 70 mV (Fig. 3A, b) (Terzic et al., 1995). However, inward currents at potentials negative to 80 mV did not increase as prominently; this is because the inward rectification of K_ATP channels is so weak that the current exhibits even a slightly outward rectification in the presence of 5.4 mM extracellular K⁺ as expected from the constant field theory (Findlay, 1987). Therefore, the inward rectification of the steady-state I-V relationship became weaker after the application of pinacidil than the control (Fig. 3B, open squares). The I_Ca currents (roughly estimated as a difference between the instantaneous and the steady-state currents at 30 mV) became smaller after the application of pinacidil. This was probably due to the rundown of I_Ca channels (McDonald et al., 1994). Tertiapin (1 µM) did not significantly inhibit the pinacidil-induced currents (Fig. 3A, c, and B, triangles), whereas glibenclamide (1 µM), a specific inhibitor of K_ATP channels, strongly suppressed the K_ATP currents (Fig. 3A, d) (Terzic et al., 1995). Under this condition, the steady-state currents (Fig. 3B, open inverted triangles) at potentials negative to 80 mV were slightly smaller than those under the control condition. This is probably because of the weak inhibitory effect of tertiapin on I_K1 channel currents. In summary, tertiapin (1 µM) did not significantly inhibit K_ATP channels. This observation was confirmed in two additional experiments with different ventricular myocytes (not shown).

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Effect of tertiapin on KATP channel in a rabbit ventricular myocyte. A, whole-cell currents recorded from a rabbit ventricular myocyte under the control condition (a), 171 s after the application of pinacidil (100 µM) (b), 74 s after the application of pinacidil and tertiapin (1 µM) (c), and 148 s after the application of pinacidil, tertiapin, and glibenclamide (1 µM) (d). The same voltage pulses as in Fig. 2 were applied to the cell. A leak and capacitative currents were not subtracted. B, I-V relationship measured in A. Filled and open symbols indicate the current amplitudes at the beginning and the end of each voltage step, respectively.

Concentration-Dependent Inhibition of K_ACh and I_K1 Channels by Tertiapin. We next examined the concentration-dependent effect of tertiapin on K_ACh and I_K1 channels (Fig. 4). The upper row in Fig. 4A depicts a continuous recording of the whole-cell current of an atrial myocyte. Throughout the experiment, three consecutive voltage steps to 100, 40, and +10 mV were applied from the holding potential of 40 mV every 3 s (left bottom corner of Fig. 4A). The duration of each the step was 300 ms. The thick lines above the current trace indicate the perfusion protocol, according to which the experiment was divided into four parts (indicated by thin lines 1-4).

View larger version (42K): [in this window] [in a new window]   Fig. 4.   Experimental protocol to assess the concentration-dependent effect of tertiapin on KACh currents. A, top row shows a record of whole-cell currents of a rabbit atrial myocyte. Thick lines above the trace indicate the bath perfusion protocol, according to which the experiment is divided into four parts (indicated by thin lines 1-4). Throughout the experiment, the voltage pulse drawn at the bottom left corner was applied from 40 mV every 3 s. In this pulse protocol, the membrane potential was first hyperpolarized to 100 mV and then depolarized to 40 and finally to +10 mV. Each step was 300 ms in duration. An inward Na+ current is eliminated from the trace in the top row. Open and closed circles indicate the current amplitudes at the end of the steps to 40 and 100 mV, respectively. The time-dependent decrease in these current amplitudes in part 2 was fit by eqs. 2 and 3, respectively. The curves were extended to the end of part 3 and shown as lines f and g. Bottom row, fast speed records of the currents within the single set of voltage pulses applied at times a-e indicated in the top row. The Na+ current is not eliminated in these traces. B, the amplitudes of the ACh-induced currents at 40 () and 100 mV () in parts 2 and 3 were normalized to those predicted from the lines f and g and plotted against time after the application of ACh (see eqs. 4-7). Two horizontal lines in the graph (designated as 40 and 100 mV) indicate the mean of the normalized currents between 95 and 135 s, which were 0.64 and 0.60, respectively. C, comparison of waveforms of KACh currents in the presence and absence of tertiapin. Trace c  a was produced by subtraction of trace a from trace c, corresponding to the Ach-induced KACh current shortly before the application of tertiapin. Line g predicts that the current at time c would have decreased to 77% at time d if tertiapin had not been applied. Thus, the Ach-induced KACh current at time d in the absence of tertiapin is shown as trace 0.77c  a. B, this current is further decreased to 64% in the presence of tertiapin, which is shown as trace 0.64(0.77c  a) (dotted line). The real Ach-induced KACh current at time d in the presence of tertiapin was obtained by subtraction of trace e from trace d (trace d  e, solid line).

40

100

40

100

40(1)

100(1)

40

100

(2)

(3)

40

100

40

100

40(4)

100(4)

(4)

(5)

(6)

(7)

40

100

40

100

100

100

100

40

100

40

100

(8)

40

100

100(4)

100(1).

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Concentration-dependent effect of tertiapin on KACh and IK1 channels. Symbols and bars indicate the mean ± S.E. The number of observation at each point is indicated in parentheses. Open and closed circles indicate the normalized Ach-induced KACh currents at 40 and 100 mV, respectively. Continuous and dotted lines are the fit of the data at 40 and 100 mV by the Hill equation (eq. 8 in text), respectively. The dissociation constant and the Hill coefficient were, respectively, estimated as 7.4 nM and 0.81 at 40 mV and 8.4 nM and 0.82 at 100 mV. Squares represent the IK1 currents at 100 mV in the presence of tertiapin normalized to those in the absence of tertiapin.

Inhibitory Effect of Tertiapin on K_ACh Channels Is Not Mediated by Muscarinic Receptors. One possible explanation for the selective inhibitory effect of tertiapin on K_ACh channels might be that tertiapin impaired activation of K_ACh channels through an antagonistic effect on muscarinic receptors. This possibility is, however, denied in Fig. 6. Here, the whole-cell current was measured in an atrial myocyte with a pipette containing 10 µM GTPS, a hydrolysis-resistant analog of GTP. GTPS can fully activate K_ACh channels in the absence of any receptor agonists (Ito et al., 1991; Kurachi et al., 1996; Yamada et al., 1998). From the holding potential of 40 mV, the voltage pulse shown in Fig. 4A was applied every 5 s. This current trace started ~2 min after formation of the whole-cell configuration. In this particular experiment, GTPS already activated K_ACh channels to a stable level from the very beginning of the recording (Fig. 6, A and B, trace a). Tertiapin (100 nM) suppressed the GTPS-induced K_ACh currents to leave the inactivating I_K1 current at 100 mV (Fig. 6B, trace b). Subtraction of the current before the application of tertiapin from that after the application of tertiapin yielded a tertiapin-sensitive current (Fig. 6B, trace a  b). This current exhibited the typical slow relaxation of K_ACh channels. Similar results were obtained in two other experiments (not shown). Thus, tertiapin inhibited K_ACh channels in a receptor-independent manner.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Effect of tertiapin on KACh channel currents activated by GTPS. A, a whole-cell current recorded from an atrial myocyte with a pipette containing 3 mM ATP and 10 µM GTPS. The same voltage pulse as in Fig. 4A was applied from 40 mV every 5 s. , current amplitude at the end of the 300-ms voltage step to 100 mV. The inward Na+ current is removed from the trace. This recording started ~2 min after brake-in, at which KACh channels were already activated to a stable level. B, left, current recorded during the single set of voltage pulses at time a (solid line) and b (dotted line) in A. The inward Na+ current is not removed. Right, a tertiapin-sensitive current was obtained by subtracting the trace b from the trace a.

Effect of Tertiapin on K_ACh Channels at Single-Channel Level. Figure 7A shows the single-channel recordings of K_ACh channels in the cell-attached configuration with the pipette filled with K⁺ external solution containing 145 mM KCl and 1 µM ACh with or without 30 nM tertiapin. The cell membrane potential was zeroed by bath perfusion of the internal solution, whereas the patch membrane potential was changed by altering the pipette potential. In the absence of tertiapin (left), numerous brief inward K_ACh currents appeared at potentials negative to E_K (~0 mV under this condition) (Sakmann et al., 1983). At potentials positive to E_K, only small outward currents were occasionally seen due to the inward rectification of the channel (Yamada and Kurachi, 1995). In the presence of tertiapin (right), the amplitude of single-channel K_ACh currents was not significantly different from the control at each membrane potential. Figure 7B depicts the single-channel I-V relationship of K_ACh channels in the absence (open circles) and presence (closed circles) of 30 nM tertiapin (n = 3 for each point). The linear fit of the averaged data gave the estimate of the single K_ACh channel conductance of 44.2 pS in the absence of tertiapin (dotted line) and 46.9 pS in the presence of tertiapin (solid line). Thus, tertiapin did not reduce the single-channel conductance of K_ACh channels.

View larger version (46K): [in this window] [in a new window]   Fig. 7.   Effect of tertiapin on the single-channel conductance and open times of KACh channels. A, in the cell-attached configuration, single KACh channel currents of atrial myocytes were recorded at indicated patch membrane potentials. Patch pipettes contained 145 mM K+ and 1 µM ACh with (left) or without (right) 30 nM tertiapin. These two sets of the current traces were obtained from different atrial myocytes. Arrowheads indicate the zero current level in each trace. B, relationship between the membrane potential (Vm) and the single-channel amplitude (i) in the absence () and the presence () of 30 nM tertiapin. Symbols and bars indicate the mean ± S.E. of three experiments at each point. Lines are the linear fit of the data in the absence and presence of tertiapin with a slope of 44.2 pS (dotted line) and 46.9 pS (solid line), respectively. C, open-time histograms constructed from experiments shown in A at 80 mV in the absence (top) and presence (bottom) of 30 nM tertiapin. Because the patches were clearly not one-channel patches, the histograms were constructed from the arbitrarily selected portions of the records in which no simultaneous openings of more than one channel were observed. The relative number of events longer than 20 ms is summed in the last bin. The histograms were fit by the following biexponential functions: top graph, y = 0.14 exp(t/0.6) + 0.029 exp(t/5.6); bottom graph, y = 0.10 exp(t/2.0) + 0.0079 exp(t/10.4), where y is the relative number of events, and t is the open time in ms.

View larger version (44K): [in this window] [in a new window]   Fig. 8.   Effect of tertiapin on KACh channels in the outside-out and inside-out configurations. A, single-channel recording in the outside-out patch membrane of an atrial myocyte. A patch pipette was filled with the internal solution containing 3 mM ATP and 100 µM GTP. The external side of the patch membrane was perfused with the K+ external solution containing the agents indicated above the current trace. The membrane potential was 60 mV. The two left graphs under the current trace are the amplitude histograms constructed from parts a and b indicated by lines under the current record. The bin width was 0.01 pA. The vertical dotted line indicates the zero current level. The first four peaks in graph a and the first five peaks in graph b were fit by Gaussian curves. The individual Gaussian curves and the sum of them are indicated by lines in the graphs. The estimated single-channel amplitude is indicated in each histogram. The right graph under the current trace shows power density spectra, which were constructed from parts d and e and subtracted with that in the absence of ACh. The lines indicate the fit of the data with the sum of three Lorentzian functions (eq. 1). The corner frequencies were 4.44, 24.6, and 272 Hz in d and 4.97, 30.8, and 241 Hz in e. The corresponding zero-frequency asymptotes were 1.02, 0.258, and 0.00961 pA2s in d and 0.449, 0.0569, and 0.00388 pA2s in e. B, single-channel recording in the inside-out patch membrane of an atrial myocyte. A patch pipette contained the K+ external solution with 1 µM ACh, whereas the internal side of the patch membrane was perfused with the internal solution containing the agents indicated above the trace. The membrane potential was 60 mV. The single-channel amplitude and NPo were measured in the periods indicated by lines under the trace and are shown beneath the lines.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have shown that tertiapin potently suppressed ACh-induced K_ACh channel currents in rabbit cardiac myocytes (K_d ~ 8 nM) (Figs. 2, 4, and 5). Tertiapin also inhibited I_K1 channels, but by only less than 10% at 1 µM (Fig. 5). Thus, tertiapin was at least 100 time more potent in inhibiting K_ACh than I_K1 channels. The other ion channels, including K_ATP channels, were virtually insensitive to the toxin (Figs. 2 and 3). Tertiapin inhibited K_ACh channels in a voltage- and receptor-independent manner without affecting the slow relaxation (Figs. 4-6). At the single-channel level, tertiapin reduced NP_o without significantly affecting the single-channel conductance or the fast open-closed kinetics (Figs. 7 and 8; Table 1). Therefore, tertiapin is likely to cause a slow block on K_ACh channels, consistent with the high affinity of the toxin for the channels (Hille, 1992).

Jin and Lu (1998) found that tertiapin selectively blocks certain types of recombinant Kir channels expressed in Xenopus oocytes. Kir channels are formed by four subunits, each of which has two transmembrane segments (Ho et al., 1993; Kubo et al., 1993a; Yang et al., 1995). At least 13 distinct genes for Kir subunits have been identified and classified into six groups according to their primary structure (Kir1.x-Kir6.x; Doupnik et al., 1995). A homotetramer of Kir1.1 (ROMK1) subunits is a weak inward rectifier involved in the electrolyte metabolism in renal tubules but not expressed in the heart (Ho et al., 1993). Kir2.x (IRKx) subunits form as a homotetramer the strong inward rectifier in heart, muscle, and brain (Kubo et al., 1993; Morishige et al., 1994; Takahashi et al., 1994). Kir3.x (GIRKx) subunits make up G protein-gated Kir channels in heart, brain, and endocrine organs by forming either a homotetrameric or heterotetrameric complex (Kubo et al., 1993b; Krapivinsky et al., 1995; Yamada et al., 1998). Tertiapin blocks the homomeric Kir1.1 channel and the heteromeric Kir3.1/Kir3.4 channel with a K_d value of ~2 and ~8 nM, respectively, but blocks Kir2.1 channels by only less than 10% at 1 µM (Jin and Lu, 1998). These results are very consistent with what we found in the present study. The Kir3.1/Kir3.4 channel has been considered the cardiac K_ACh channel based on biochemical and functional studies (Krapivinsky et al., 1995). The equivalent potency of tertiapin (K_d ~ 8 nM) for the native K_ACh and the recombinant Kir3.1/Kir3.4 channels strongly supports this hypothesis. I_K1 channels were as insensitive to tertiapin as the Kir2.1 channel. However, the native I_K1 channel is more similar to the Kir2.2 than the Kir2.1 channel in single-channel conductance and kinetics (Takahashi et al., 1994; Isomoto et al., 1997; Matsumoto et al., 1999). Therefore, the sensitivity of the Kir2.2 channel to tertiapin also must be examined. The cardiac K_ATP channel is a hetero-octamer composed of four Kir6.2 subunits and four sulfonylurea receptors, ATP-binding cassette protein (Inagaki et al., 1996; Okuyama et al., 1998). The present study indicates that Kir6.2 channels are also insensitive to tertiapin (Fig. 3).

A Kir subunit has a domain referred to as a P or an H5 region between the two transmembrane segments (Kubo et al., 1993a; Isomoto et al., 1997). This region forms a K⁺ channel pore (Hartmann et al., 1991; Yellen et al., 1991). A site-directed mutagenesis on the P region of Kir1.1 subunit identified amino acids involved in the tertiapin binding (Jin and Lu, 1998). These residues are mainly located both amino and carboxyl terminal to the K⁺ channel signature sequence that forms the K⁺ selectivity filter (Heginbotham et al., 1994). A recent crystallographic analysis of the three-dimensional structure of a voltage-dependent K⁺ channel indicates that these two regions form the turret and the vestibule base, respectively, within the outermost part of the channel pore (Doyle et al., 1998). Although no structural information is available on the tertiapin binding site or sites in the Kir3.1/Kir3.4 channel, what is predicted from the structural analyses agrees with the observed voltage- and receptor-independent effect of extracellular tertiapin (Figs. 5 and 6) and the ineffectiveness of intracellular tertiapin on K_ACh channels (Fig. 8B). One amino acid (F137) in the P region of Kir3.1 may be responsible for the slow relaxation of K_ACh channels (Kofuji et al., 1996). The corresponding residue in Kir1.1 subunit (S135) does not significantly contribute to the tertiapin interaction (Jin and Lu, 1998), which is consistent with the lack of the effect of tertiapin on the slow relaxation kinetics (Fig. 4C). Although we could not analyze in detail the single-channel kinetics of K_ACh channels, tertiapin seems to cause a slow block on the channels (Hille, 1992). The state dependence of the tertiapin block is unknown, but K_ACh channels do not have to open to interact with tertiapin because K_ACh channels became refractory to ACh (1 µM) when pretreated with 100 nM tertiapin (not shown). Detailed structural and functional analyses of the interaction of tertiapin and K_ACh (or Kir3.1/Kir3.4) channels will delineate the pore structure and the mechanism of the gating of the heterotetrameric ion channel.

To summarize, tertiapin potently and specifically blocks K_ACh channels in cardiac myocytes. Tertiapin can therefore be a powerful pharmacological tool to identify the functional role of K_ACh channels in the cardiac physiology and pathophysiology.