Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

In severe ischemia and hypoxia, the ATP/ADP ratio decreases, thereby triggering the opening of K⁺ channels that are closed by intracellular ATP and opened by MgADP (Noma, 1983; Venkatesh et al., 1991; for review, see Gross and Auchampach, 1992). The opening of these ATP-sensitive K⁺ channels (K_ATP channels) clamps the cardiocyte at the potassium equilibrium potential and renders it nonexcitable. Whereas this may salvage ATP and preserve the structural integrity of the cell (Noma, 1983; Gross and Auchampach, 1992), it also increases the electrical heterogeneity of the heart and promotes reentry arrhythmias (Janse and Wit, 1989). In addition, the opening of K_ATP channels leads to the accumulation of extracellular K⁺ in the ischemic zone, depolarizes the cell, and induces cytotoxic Ca²⁺ entry (Wilde et al., 1990; Venkatesh et al., 1991). In accordance with these ideas, block of plasmalemmal K_ATP channels in the cardiac myocyte has been shown to protect against ischemia-induced ventricular fibrillation (for reviews, see Wilde, 1994; Gögelein et al., 1999).

In addition to their presence in cardiocytes, K_ATP channels are found also in the plasma membrane of other myocytes and, in particular, of pancreatic -cells, where they couple insulin release to the plasma glucose level (for review, see Ashcroft and Ashcroft, 1990). A further type of K_ATP channel has been demonstrated in the mitochondrial inner membrane of the cardiocyte (Inoue et al., 1991). In cardiac ischemia, this channel also opens, and this is thought to mediate cardiac preconditioning (for review, see Szewczyk and Marbán, 1999; Grover and Garlid, 2000), a process by which a brief period of "conditioning" ischemia protects the heart against longer ischemic periods (Murry et al., 1986). In addition, ischemic cardiocytes release adenosine, which opens K_ATP channels in coronary smooth muscle cells. This decreases the resistance of the coronary vascular bed, thus contributing to the autoregulation of coronary blood flow (Daut et al., 1990; Dart and Standen, 1993). Hence, a blocker of surface K_ATP channels of the cardiocyte, if it was to be therapeutically useful, should inhibit neither the mitochondrial K_ATP channel in the cardiocyte nor the surface K_ATP channels in coronary myocytes and pancreatic -cells.

HMR 1883 (1-[[5-[2-(5-chloro-o-anisamido)ethyl]-2-methoxyphenyl]sulfonyl]-3-methylthiourea) is a novel sulfonylthiourea that differs from glibenclamide, the prototype sulfonylurea blocker of the pancreatic -cell channel, in several aspects (Fig. 1; see also Gögelein et al., 1999). Experiments in isolated tissues and in animals in vivo have shown that these modifications lead to an unusual pharmacological profile (for review, see Gögelein et al., 1999). In contrast to glibenclamide, HMR 1883 has only a weak potency at pancreatic -cells but inhibits the K_ATP channel in the sarcolemma of the cardiocyte with micromolar potency; at 10 µM, the compound does not affect coronary flow (Gögelein et al., 1998), and it does not inhibit the mitochondrial K_ATP channel at 30 µM (Sato et al., 2000). However, little is known about the interaction of HMR 1883 with recombinant surface K_ATP channels and sulfonylurea receptors (SURs).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Structures of glibenclamide and HMR 1883. In HMR 1883, the benzamido moiety of glibenclamide is retained; however, the sulfonylthiourea part of HMR 1883 continues in the meta position of the central benzene ring. The pKa values of the nitrogen next to the sulfonyl group are 4.4 (HMR 1883) and 6.2 (glibenclamide), respectively (Gögelein et al., 1999).

Surface K_ATP channels are composed of two types of subunits, inwardly rectifying K⁺ channels (Kir6.x) and SURs (for reviews, see Ashcroft and Gribble, 1998; Aguilar-Bryan and Bryan, 1999). The Kir6.x subunits form the pore of the channel. SUR is a member of the ATP-binding cassette protein superfamily (Ashcroft and Gribble, 1998; Aguilar-Bryan and Bryan, 1999) and carries binding sites for nucleotides (Ueda et al., 1999), for the sulfonylureas (Aguilar-Bryan et al., 1995), and for the K_ATP channel openers like pinacidil and its analog, P1075 (Hambrock et al., 1998; Schwanstecher et al., 1998). Two genes code for the SUR subtypes. SUR1 is mainly found in pancreatic -cells and in neurons (Aguilar-Bryan et al., 1995); SUR2 is in myocytes with one isoform (SUR2A) in skeletal and cardiac and the other (SUR2B) in smooth muscle (Inagaki et al., 1996; Isomoto et al., 1996). SUR1 has high affinity for the sulfonylureas and a low affinity for the K_ATP channel openers; the converse is true for the SUR2 isoforms (Hambrock et al., 1998; Schwanstecher et al., 1998; Dörschner et al., 1999; Hambrock et al., 1999; Russ et al., 1999).

It was the aim of this study to investigate the interaction of HMR 1883 and glibenclamide with the recombinant K_ATP channels and SURs in electrophysiological and radioligand binding studies; particular emphasis was given to the interaction with SUR2A and SUR2B. The results show that HMR 1883 exhibits a unique selectivity profile at the different SURs and that binding of HMR 1883 to SUR2 is more sensitive to inhibition by nucleotides than that of glibenclamide.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture, Transfection, and Membrane Preparation. Human embryonic kidney 293 cells were cultured as described previously (Hambrock et al., 1998) in minimum essential medium containing glutamine and supplemented with 10% fetal bovine serum and 20 µg/ml gentamycin. Cells were transfected with the pcDNA 3.1 vector (Invitrogen, Karlsruhe, Germany) containing the coding sequence of rat SUR1, murine SUR2A, or murine SUR2B [GenBank accession numbers L40624 (Aguilar-Bryan et al., 1995), D86037, and D86038, respectively (Isomoto et al., 1996)]. Cells stably transfected with these proteins were isolated in the presence of the antibiotic geneticin and expressed the different SUR subtypes at levels of 1.0 ± 0.2 pmol/mg of protein. For patch clamp experiments, cells were transiently transfected with SURx+Kir6.x [murine Kir6.1, D88159 (Yamada et al., 1997); murine Kir6.2, D50581 (Inagaki et al., 1995)] at a molar plasmid ratio of 1:1 using lipofectAMINE and OPTIMEM (Invitrogen) according to the manufacturer's instructions. pEGFP-C1 vector (CLONTECH, Palo Alto, CA), encoding for green fluorescent protein, was added for easy identification of transfected cells (Russ et al., 1999). Cells were allowed to express transfected DNA for 48 h and were then used for electrophysiological experiments.

Radioligand Binding Experiments. Membranes (final protein concentration 0.1-0.5 mg of protein/ml) were added to the incubation buffer (139 mM NaCl, 5 mM KCl, 5 mM HEPES, and 2.2 mM MgCl₂) supplemented with 1 mM Na₂ATP, the radioligand ([³H]glibenclamide = 1.0-1.6 nM or [³H]P1075 = 1.5-3 nM), and the inhibitor of interest at 37°C. In case that an ATP-regenerating system was coupled, creatine kinase (5 U/ml) and creatine phosphate (3 mM) were added in the presence of 10 mM Mg²⁺ as described in Hambrock et al. (1999). For low MgATP (3 µM), the incubation medium contained 3 µM ATP and 1 mM Mg²⁺. At equilibrium (SUR1 + [³H]glibenclamide, 15 min; SUR2A + [³H]P1075, 13 min; and SUR2B + [³H]P1075, 30 min), incubation was stopped by diluting 0.3-ml aliquots in triplicate into 8 ml of quench solution (50 mM tris-(hydroxymethyl)-aminomethane and 154 mM NaCl, pH 7.4) at 0°C. The solution was filtered over Whatman GF/B filters and filters washed twice with 8 ml of quench solution. Nonspecific binding was determined in the presence of 1 µM (unlabeled) glibenclamide (SUR1) or 10 µM P1075 (SUR2) and did not exceed 20% of total binding with the exception of the [³H]P1075 binding experiments in the presence of low ATP (3 µM) where it reached up to 40% of total binding.

Patch-Clamp Experiments. The patch-clamp technique was used in the whole-cell and inside-out configuration. Transfected human embryonic kidney 293 cells showing green fluorescent protein fluorescence were chosen. Patch pipettes were drawn from borosilicate glass capillaries (GC 150, Harvard Apparatus, Edenbridge, UK) and heat polished using a horizontal microelectrode puller (Zeitz, Augsburg, Germany). For inside-out patches, bath and pipette were filled with a high K⁺-Ringer solution containing (in mM): KCl, 142; NaCl, 2.8; MgCl₂, 1; CaCl₂, 1; D(+)-glucose, 11; and HEPES, 10, titrated to pH 7.4 with NaOH at 22°C. After filling with buffer, pipettes had a resistance of 1 to 1.5 M. After excision of the patch, the pipette was moved in front of a pipe with a high K⁺-EGTA-Ringer solution containing (in mM): KCl, 143; MgCl₂, 0.85; CaCl₂, 1; ethylene glycol-bis-(2-aminoethyl ether)-N,N,N',N'-tetraacetic acid, 5; D(+)-glucose, 11; and HEPES, 10, titrated to pH 7.2 with NaOH at 22°C. HMR 1883, glibenclamide, and ATP (free Mg²⁺ was kept constant) were dissolved as described below and added to the pipe solution. Patches were clamped to 50 mV. All responses were normalized to the effect of 1 µM glibenclamide (=100% block).

Data Analysis. Concentration dependencies were analyzed by fitting the logistic form of the Hill equation,
to the data. n (= n_H) is the Hill coefficient, x the concentration of the compound under study, and K is the midpoint of the curve with px = logx and pK = logK. The dependence of the midpoint of an inhibition curve (IC₅₀ value) on the concentration of the radioligand, L, was calculated according to the equation (Cheng and Prusoff, 1973),
where K_i is the inhibition constant and K_D the equilibrium dissociation constant of the radioligand.

Chemicals. [³H]P1075 (specific activity 4.5 TBq (117 Ci) mmol¹) was purchased from Amersham Pharmacia Biotech (Freiburg, Germany) and [³H]glibenclamide [specific activity 1.85 TBq (50 Ci) mmol¹] from DuPont NEN (Bad Homburg, Germany). The reagents and media used for cell culture and transfection were from Invitrogen (Eggenstein, Germany). Na₂ATP and Li₂GDP were from Roche Molecular Biochemicals (Mannheim, Germany); creatine kinase, creatine phosphate, glibenclamide, and noradrenaline from Sigma (Deisenhofen, Germany). HMR 1883 was the kind gift of Aventis (Frankfurt, Germany) and P1075 of Leo Pharmaceuticals (Ballerup, Denmark). K_ATP channel inhibitors were dissolved in dimethyl sulfoxide/ethanol (1:1) and further diluted with the same solvent or with incubation buffer. In binding studies, the final solvent concentration in the assays was always below 0.3%, in electrophysiological studies  0.1%.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Binding of HMR 1883 and Glibenclamide to Recombinant SURs. The interaction of HMR 1883 and glibenclamide with SUR1 was studied in [³H]glibenclamide competition assays. In the presence of MgATP (1 mM), HMR 1883 inhibited specific [³H]glibenclamide binding completely and in a monophasic manner (n_H ~ 1) with an inhibition constant (K_i value) of 63 µM (Table 1). The K_i value (= K_D) for glibenclamide was 1.5 nM. In the absence of MgATP, both inhibition curves were shifted to the left by a factor of about 7 (Table 1).

View larger version (36K): [in this window] [in a new window]   Fig. 2.   Binding of HMR 1883 and glibenclamide to SUR subtypes and effect of MgATP. Data are presented as percentage of specific binding (% BS). a, inhibition of [3H]glibenclamide ([3H]GBC) binding to SUR1. Experiments were performed in a buffer without Mg2+ in the presence of 1 mM EDTA (MgATP) or in the presence of 2.2 mM Mg2+ and 1 mM Na2ATP (+MgATP); [3H]glibenclamide was ~1 nM. Means ± S.E.M. from three to four experiments are shown. The pKi values obtained from the fit of eq. 1 in Materials and Methods to the data and corrected according to eq. 2 are listed in Table 1; Hill coefficients were ~1. b and c, inhibition of [3H]P1075 binding to SUR2A/2B. ATP concentrations are indicated; free Mg2+ was ~1 mM except on the presence of ATP-regenerating system (ARS) where ATP = 1 mM, Mg2+ = 10 mM, creatine kinase = 5 U/ml, creatine phosphate = 3 mM. The pKi values derived from the Hill fits shown are listed in Table 1. Hill coefficients were ~1 with the exception of the HMR 1883 curves at SUR2B in the presence of 1 mM ATP ± ARS (Table 1); n = 3 to 4.

Interaction with Recombinant K_ATP Channels. The inhibition of the recombinant cardiac K_ATP channel Kir6.2/SUR2A by HMR 1883 and glibenclamide is shown in Fig. 3. The upper panel illustrates the current after patch excision into nucleotide-free solution, run down of the current, inhibition by MgATP, and refreshment upon washout of the nucleotide. HMR 1883 (10 µM) inhibited the current by 57%, and glibenclamide, at the saturating concentration of 1 µM, by 67%. For evaluation of the concentration-inhibition curves (Fig. 3, lower panel), the glibenclamide-sensitive fraction of the current (mean, 54% ± 3%; n = 49) was set to 100% in each patch. HMR 1883 and glibenclamide inhibited the Kir6.2/SUR2A channel with IC₅₀ values of 0.38 and 0.026 µM, respectively and Hill coefficients close to 1 (Table 2).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Inhibition by HMR 1883 and glibenclamide of Kir6.2/SUR2A and Kir6.2/SUR2B currents in inside-out patches at 50 mV. Upper panel, original recording from a patch containing Kir6.2/SUR2A channels in symmetrical high K+ buffer. After excision of the patch into nucleotide-free solution, the current was subject to rundown. Exposure to MgATP (1 mM) led to total inhibition (8-pA leak current indicated by the dotted line), and after washout, some refreshment was observed. Application of 10 µM HMR 1883 inhibited the current by 57% as indicated by the comparison of the lower line with the extrapolated current in the absence of inhibitor (upper line). After washout and application of MgATP (1 mM), 1 µM glibenclamide inhibited the current by 67%. Lower panel, concentration-inhibition curves of Kir6.2/SUR2A and 2B currents by glibenclamide (n = 4-5) and HMR 1883 (n = 3-12). The glibenclamide (1 µM)-sensitive fraction of the ATP-inhibited current, IGBC, was renormalized to 100%. Asterisks indicate significant difference between the block of Kir6.2/SUR2A and Kir6.2/SUR2B channels, respectively, by given concentrations of HMR 1883. For parameters of the Hill fit to the data, see Table 2.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Inhibition of Kir6.1/SUR2B current (IKATP) by HMR 1883 and glibenclamide measured in the whole-cell configuration at 60 mV. Upper panel, original current trace. After dialysis with 1 mM GDP in the presence of 0.3 mM ATP and 2 mM Mg2+, a current developed that was inhibited by cumulative addition of HMR 1883 to the bath. The lower panel shows the concentration-response curves for glibenclamide and HMR 1883 [data for glibenclamide from Russ et al. (1999); n = 4-17 for HMR 1883]. Note the significantly steeper curve for HMR 1883 with Hill coefficient 1.7 (Table 2).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Interaction with SUR1. In this study we have compared the interactions of HMR 1883 and glibenclamide with different SURs and recombinant cardiovascular K_ATP channels in radioligand binding and electrophysiological experiments. The results show that the structural differences between the two compounds result in markedly different selectivity profiles. The largest difference was seen in the affinity to SUR1, the SUR in the pancreatic -cell. In the [³H]glibenclamide inhibition experiments, HMR 1883 was ~4 × 10⁴ weaker than glibenclamide, both in the presence and absence of MgATP. In reasonable agreement with this result, about 2000-fold higher concentrations of HMR 1883 than of glibenclamide are required to depolarize RINm5F cells (Gögelein et al., 1998), a cell line derived from a rat pancreatic -cell tumor.

Binding Studies with SUR2 Subtypes. The interaction of the two compounds with the SUR2 subtypes was studied using the opener [³H]P1075 as the radioligand (Bray and Quast, 1992; Quast et al., 1993); [³H]glibenclamide binding to SUR2 in membranes cannot be reliably measured due to insufficient affinity and rapid kinetics of the radioligand (Dörschner et al., 1999; Russ et al., 1999). The binding sites of openers and sulfonylureas at SUR have been proposed to be distinct but to be located in close neighborhood (Ashfield et al., 1999; Uhde et al., 1999), and binding studies have shown that the respective binding sites are linked by a negative allosteric interaction (Bray and Quast, 1992; Schwanstecher et al., 1992). This may be the reason for glibenclamide inhibiting [³H]P1075 binding to SUR2B with a potency considerably lower than the true binding affinity (~30-fold; Russ et al., 1999; Dörschner et al., 1999), and this may be similar for SUR2A (Hambrock et al., 1999). It therefore appears that the [³H]P1075 inhibition assay quantitatively underestimates the affinity of glibenclamide and HMR 1883 binding to SUR2. However, the potency ratio of different sulfonylureas and their derivatives is correctly reflected as are coupling effects of nucleotides and of coexpression with Kir6.1 (Quast et al., 1993; Hambrock et al., 1999; Russ et al., 1999).

Electrophysiological Experiments. In inside-out patches and in the absence of nucleotides, HMR 1883 inhibited the recombinant cardiac (Kir6.2/SUR2A) and the nonvascular smooth muscle K_ATP channel (Kir6.2/SUR2B) with IC₅₀ values of 0.38 and 1.2 µM, respectively. In contrast, glibenclamide was more potent and did not discriminate between these channels (IC₅₀ = 26-27 nM). Quantitatively, HMR 1883 was 15-fold weaker than glibenclamide at the SUR2A-containing channel and 44-fold weaker at the SUR2B-containing channel; in addition, the Kir6.2/SUR2B inhibition curve was surprisingly flat (n_H = 0.6). When the two compounds were compared at the recombinant vascular K_ATP channel, Kir6.1/SUR2B, in the whole cell configuration and in the presence of nucleotides, HMR 1883 was ~100-fold less potent than glibenclamide and the HMR 1883 inhibition curve was steep (n_H = 1.7).

Selectivity of HMR 1883 for the Cardiac over the Vascular K_ATP Channel. An important point for the intended therapeutic application of HMR 1883 is the selectivity of the compound for the cardiac over the vascular K_ATP channel. The present data give only qualitative indications. From the [³H]P1075 inhibition experiments at high MgATP one obtains potency differences of 4.5 and 3.1 in the presence and absence of the ATP-regenerating system. In inside-out patches and in the absence of nucleotides, the potency difference of HMR 1883 inhibiting the Kir6.2/SUR2A (cardiac) and the Kir6.2/SUR2B (nonvascular smooth muscle) channel was 3-fold; however, due to the flat inhibition curve of the SUR2B-containing channel, up to 10 times higher concentrations of HMR 1883 are required to produce substantial (~80%) block. Since coexpression of SUR2B with Kir.6.x affects glibenclamide potency in a manner depending on the Kir6.x subtype (Hambrock et al., 2001), the situation may again be different when potency of the two blockers at Kir6.2/SUR2A channel is compared with the recombinant vascular K_ATP channel, Kir6.1/SUR2B. In any event, experiments in Langendorff-perfused rabbit hearts (Gögelein et al., 1998) and in conscious dogs with a healed myocardial infarction which were subjected to acute coronary artery occlusion (Billman et al., 1998) have shown that HMR 1883, at concentrations sufficient to close the cardiocyte channel, did not affect the channel in coronary myocytes. This indicates a sufficient selectivity of the compound in these more "realistic" models where the channel is in its native environment and in contact with physiological (or pathophysiological) levels of nucleotides.