Intracellular Binding of Terfenadine Competes with Its Access to Pancreatic ß-cell ATP-Sensitive K+ Channels and Human ether-à-go-go-Related Gene Channels.

Most blockers of both hERG (human ether-à-go-go-related gene) channels and pancreatic ß-cell ATP-sensitive K+ (KATP) channels access their binding sites from the cytoplasmic side of the plasma membrane. It is unknown whether binding to intracellular components competes with binding of these substances to K+ channels. The whole-cell configuration of the patch-clamp technique, a laser-scanning confocal microscope, and fluorescence correlation spectroscopy (FCS) were used to study hERG channels expressed in HEK (human embryonic kidney) 293 cells and KATP channels from the clonal insulinoma cell line RINm5F. When applied via the pipette solution in the whole-cell configuration, terfenadine blocked both hERG and KATP currents with much lower potency than after application via the bath solution, which was not due to P-glycoprotein-mediated efflux of terfenadine. Such a difference was not observed with dofetilide and tolbutamide. 37-68% of hERG/EGFP (enhanced green-fluorescent protein) fusion proteins expressed in HEK 293 cells were slowly diffusible as determined by laser-scanning microscopy in the whole-cell configuration and by FCS in intact cells. Bath application of a green-fluorescent sulphonylurea derivative (Bodipy-glibenclamide) induced a diffuse fluorescence in the cytosol of RINm5F cells under whole-cell patch-clamp conditions. These observations demonstrate the presence of intracellular binding sites for hERG and KATP channel blockers not dialyzable by the patch-pipette solution. Intracellular binding of terfenadine was not influenced by a mutated hERG (Y652A) channel. In conclusion, substances with high lipophilicity are not freely diffusible inside the cell but steep concentration gradients might exist within the cell and in the sub-membrane space.

Multiple hERG channel blocking pathways: implications for macromolecules.

Numerous non-cardiovascular drugs have a potential to induce life-threatening torsades de pointes (TdP) ventricular cardiac arrhythmias by blocking human ether-à-go-go-related gene (hERG) currents via binding to the channel's inner cavity. Identification of the hERG current-inhibiting properties of candidate drugs is performed focusing on binding sites in the channel pore. It has been suggested that biologicals have a low likelihood of hERG current inhibition, since their poor diffusion across the plasma membrane prevents them from reaching the binding site in the channel pore. However, biologicals could influence hERG channel function by binding to 'unconventional' noncanonical binding sites. This Opinion gives an overview on noncanonical blockers of hERG channels that might be of relevance for the assessment of the possible torsadogenic potential of macromolecular therapeutics.

Human ether-a-go-go-related (HERG) gene and ATP-sensitive potassium channels as targets for adverse drug effects.

Torsades de pointes (TdP) arrhythmia is a potentially fatal form of ventricular arrhythmia that occurs under conditions where cardiac repolarization is delayed (as indicated by prolonged QT intervals from electrocardiographic recordings). A likely mechanism for QT interval prolongation and TdP arrhythmias is blockade of the rapid component of the cardiac delayed rectifier K+ current (IKr), which is encoded by human ether-a-go-go-related gene (HERG). Over 100 non-cardiovascular drugs have the potential to induce QT interval prolongations in the electrocardiogram (ECG) or TdP arrhythmias. The binding site of most HERG channel blockers is located inside the central cavity of the channel. An evaluation of possible effects on HERG channels during the development of novel drugs is recommended by international guidelines. During cardiac ischaemia activation of ATP-sensitive K+ (KATP) channels contributes to action potential (AP) shortening which is either cardiotoxic by inducing re-entrant ventricular arrhythmias or cardioprotective by inducing energy-sparing effects or ischaemic preconditioning (IPC). KATP channels are formed by an inward-rectifier K+ channel (Kir6.0) and a sulfonylurea receptor (SUR) subunit: Kir6.2 and SUR2A in cardiac myocytes, Kir6.2 and SUR1 in pancreatic beta-cells. Sulfonylureas and glinides stimulate insulin secretion via blockade of the pancreatic beta-cell KATP channel. Clinical studies about cardiotoxic effects of sulfonylureas are contradictory. Sulfonylureas and glinides differ in their selectivity for pancreatic over cardiovascular KATP channels, being either selective (tolbutamide, glibenclamide) or non-selective (repaglinide). The possibility exists that non-selective KATP channel inhibitors might have cardiovascular side effects. Blockers of the pore-forming Kir6.2 subunit are insulin secretagogues and might have cardioprotective or cardiotoxic effects during cardiac ischaemia.

Effects of fluoroquinolones on HERG channels and on pancreatic beta-cell ATP-sensitive K+ channels.

An inhibition of the cardiac rapid delayed rectifier K(+) current (I(Kr)) and of the ATP-sensitive K(+) (K(ATP)) current seems to be involved in the mechanisms of the cardiotoxic effects and the alterations in glucose homeostasis, respectively, induced by some fluoroquinolones. The aim of the present study was to compare the effects of fluoroquinolone derivatives on the pore-forming subunit of the cardiac I(Kr), which is encoded by human ether-a-go-go-related gene (HERG), and on the ATP-sensitive K(+) (K(ATP)) channel from the clonal insulinoma cell line RINm5F. Sparfloxacin blocked HERG currents half-maximally (IC(50) value) at a concentration of 33.2 microM, whereas norfloxacin and lomefloxacin each tested at a concentration of 300 microM inhibited HERG currents only by 2.8+/-3.6% and 12.3+/-4.7%, respectively. Four newly synthesized fluoroquinolone derivatives with either a p-fluoro-phenyl (compound C3) or an o-fluoro-phenyl (compound C4) substituent at position N(1) and an additional dimethylated piperazine ring (compounds C1 and C2) inhibited HERG currents by 7.3-14.7% at test concentrations of 100 microM. The rank order of potency for the inhibition of K(ATP) currents was C2>C1, C4, sparfloxacin>C3. In conclusion, the structural requirements for fluoroquinolones to inhibit I(Kr) currents and K(ATP) currents appear to differ. The amino group at position C(5) seems to be primarily responsible for the strong HERG current blocking property of sparfloxacin. In contrast, for the block of pancreatic beta-cell K(ATP) currents by fluoroquinolones the substituents at positions N(1), C(7) and C(8) all might play a role.

Fluorescence microscopy studies with a fluorescent glibenclamide derivative, a high-affinity blocker of pancreatic beta-cell ATP-sensitive K+ currents.

Hypoglycemic sulfonylureas (e.g. tolbutamide, glibenclamide) exert their stimulatory effects on pancreatic beta-cells by closure of ATP-sensitive K(+) (K(ATP)) channels. Pancreatic K(ATP) channels are composed of two subunits, a pore-forming inwardly rectifying K(+) channel (Kir6.2) subunit and a regulatory subunit (the sulfonylurea receptor of subtype 1 (SUR1)) in a (SUR1/Kir6.2)(4) stoichiometry. The aim of the present study was to characterize the interaction of green-fluorescent 3-[3-(4,4 difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-S-indacen-3-yl)propanamido] glibenclamide (Bodipy-glibenclamide) with pancreatic beta-cell K(ATP) channels using patch-clamp and fluorescence microscopy techniques. Bodipy-glibenclamide inhibited K(ATP) currents from the clonal insulinoma cell line RINm5F half-maximally at a concentration of 0.6nM. Using laser-scanning confocal microscopy Bodipy-glibenclamide was shown to induce a diffuse fluorescence across the RINm5F cell, but only about 17% of total Bodipy-glibenclamide-induced fluorescence intensity in RINm5F cells was due to specific binding to SUR1. Using fluorescence correlation spectroscopy, it could be demonstrated that the fluorescence label contributes to the protein binding and, therefore, possibly also to the non-specific binding of Bodipy-glibenclamide observed in RINm5F cells. Specific binding of Bodipy-glibenclamide to SUR1 in RINm5F cells might be localized to different intracellular structures (nuclear envelope, endoplasmic reticulum, Golgi compartment, insulin secretory granules) as well as to the plasma membrane. In conclusion, Bodipy-glibenclamide is a high-affinity blocker of pancreatic beta-cell K(ATP) currents and can be used for visualizing SUR1 in intact pancreatic beta-cells, although non-specific binding must be taken into account in confocal microscopy experiments on intact beta-cells.

Specificity of nonadrenergic imidazoline binding sites in insulin-secreting cells and relation to the block of ATP-sensitive K(+) channels.

To characterize the specificity of nonadrenergic imidazoline binding sites of insulin-secreting HIT cells, competitive binding of insulinotropic imidazolines and quinine was measured and compared with the effect of these compounds on native K(ATP) channels and with a heterologously expressed variant of the pore-forming subunit (Kir6.2 deltaC26). There were two nonadrenergic imidazoline binding sites for [(3)H]clonidine with K(d) values of 61 nM and 4.5 microM, respectively. Quinine reduced specific binding incompletely (73%) with K(i) values of 75 nM and 133 microM. Clonidine, N-allyl-clonidine (alinidine), and quinine inhibited native K(ATP) channels as well as Kir6.2deltaC26 channels. Coexpression of Kir6.2deltaC26 and SUR1 (the regulatory subunit of K(ATP)) did not increase the potency of quinine. There are nonadrenergic imidazoline binding sites in insulin-secreting HIT cells which also recognize quinine. One of these sites is Kir6.2, the pore-forming subunit of the K(ATP) channel.

Effects of lomefloxacin and norfloxacin on pancreatic beta-cell ATP-sensitive K(+) channels.

In patients administered lomefloxacin alterations in blood glucose concentrations have been observed in some cases and lomefloxacin has previously been shown to augment insulin release from rat pancreatic islets at micromolar concentrations. The aim of the present study was to compare the effects of two structurally related fluoroquinolones, lomefloxacin and norfloxacin, on ATP-sensitive K(+) (K(ATP)) currents from the clonal insulinoma cell line RINm5F using the whole-cell configuration of the patch-clamp technique. The application of lomefloxacin concentration-dependently blocked K(ATP) currents from RINm5F cells with a half-maximally inhibitory concentration of 81 microM, whereas the application of norfloxacin (at concentrations up to 300 microM) had only minor effects on K(ATP) currents. Block of pancreatic beta-cell K(ATP) currents could be mediated by interaction of lomefloxacin either with the regulatory subunit (SUR1) or with the pore-forming subunit (Kir6.2). We favour the latter hypothesis, since some fluoroquinolones have recently been shown to block the pore-forming subunit of the cardiac rapid delayed rectifier K(+) current I(Kr) (which is encoded by HERG (human ether-a-go-go-related gene)). Thus, as demonstrated for cardiac HERG channels in previous studies and for pancreatic beta-cell K(ATP) channels in the present study, fluoroquinolones differ markedly in their potencies to inhibit K(+) channel activity.

Interactions at human ether-à-go-go-related gene channels.

Several noncardiovascular drugs have the potential to induce Torsades de Pointes cardiac arrhythmias via blockade of the rapid component of the cardiac delayed rectifier K(+) current (I(Kr)), which is encoded by human ether-à-go-go-related gene (hERG). The aim of the present study was to characterize possible interactions between terfenadine, binding to a site located inside the central cavity, and the following substances with various binding sites: dofetilide, fluvoxamine, chlorobutanol, and a hERG-specific toxin isolated from scorpion venom (CnErg1). The whole-cell configuration of the patch-clamp technique was employed on hERG channels stably expressed in human embryonic kidney 293 cells. Terfenadine does not interact with dofetilide or fluvoxamine at hERG channels. Slight subadditive inhibitory effects on hERG peak tail currents were observed when terfenadine and CnErg1 were administered in combination. Terfenadine and chlorobutanol synergistically inhibit hERG peak tail currents and enhance each other's inhibitory effect in a concentration-dependent way. In conclusion, terfenadine interacts with CnErg1 and chlorobutanol, but not with dofetilide or fluvoxamine, at hERG channels. It is shown that interactions between chlorobutanol and a hERG channel blocker binding inside the central cavity (terfenadine) produce synergistic effects on hERG currents.

Comparison of the effects of methadone and heroin on human ether-à-go-go-related gene channels.

Torsades de pointes (TdP) is a life-threatening form of ventricular arrhythmia that occurs under conditions of delayed cardiac repolarization indicated by prolonged QT intervals in ECG recordings. The main mechanism of QT prolongation and TdP is block of the rapid component of the cardiac delayed rectifier K(+) current (I(Kr)), which is encoded by hERG (human ether-à-go-go-related gene). The opioid agonist methadone has previously been demonstrated to inhibit hERG currents, and there are reports of serious cardiac arrhythmias and deaths from TdP and ventricular fibrillation in patients taking methadone. The aim of the present study was to compare the effects of the opioid agonists methadone and heroin (3,6-diacetylmorphine) on hERG currents stably expressed in human embryonic kidney (HEK 293) cells using the whole-cell configuration of the patch-clamp technique. Both methadone and heroin inhibit hERG currents in a concentration-dependent manner. The following values were calculated for IC(50) (concentration causing half-maximal inhibition) and n (the Hill coefficient): 4.8 microM and 0.9 for methadone, 427 microM and 0.7 for heroin. In conclusion, the potency for block of hERG currents is about 100-fold lower for heroin when compared to methadone.

Report and recommendations of the workshop of the European Centre for the Validation of Alternative Methods for Drug-Induced Cardiotoxicity.

Cardiotoxicity is among the leading reasons for drug attrition and is therefore a core subject in non-clinical and clinical safety testing of new drugs. European Centre for the Validation of Alternative Methods held in March 2008 a workshop on "Alternative Methods for Drug-Induced Cardiotoxicity" in order to promote acceptance of alternative methods reducing, refining or replacing the use of laboratory animals in this field. This review reports the outcome of the workshop. The participants identified the major clinical manifestations, which are sensitive to conventional drugs, to be arrhythmias, contractility toxicity, ischaemia toxicity, secondary cardiotoxicity and valve toxicity. They gave an overview of the current use of alternative tests in cardiac safety assessments. Moreover, they elaborated on new cardiotoxicological endpoints for which alternative tests can have an impact and provided recommendations on how to cover them.

Electrophysiological and fluorescence microscopy studies with HERG channel/EGFP fusion proteins.

HERG (human ether-a-go-go-related gene) encodes the Kv11.1 protein alpha-subunit that underlies the rapidly activating delayed rectifier K+ current (IKr) in the heart. Alterations in the functional properties or membrane incorporation of HERG channels, either by genetic mutations or by administration of drugs, play major roles in the development of life-threatening torsades de pointes cardiac arrhythmias. Visualization of ion channel localization is facilitated by enhanced green fluorescent protein (EGFP) tagging, but this process can alter their properties. The aim of the present study was to characterize the electrophysiological properties and the cellular localization of HERG channels in which EGFP was tagged either to the C terminus (HERG/EGFP) or to the N terminus (EGFP/HERG). These fusion constructs were transiently expressed in human embryonic kidney (HEK) 293 cells, and the whole-cell patch-clamp configuration and a confocal laser scanning microscope with primary anti-HERG antibodies and fluorescently labeled secondary antibodies were used. For EGFP/HERG channels the deactivation kinetics were faster and the peak tail current density was reduced when compared to both wild-type HERG channels and HERG/EGFP channels. Laser scanning microscopic studies showed that both fusion proteins were localized in the cytoplasm and on discrete microdomains in the plasma membrane. The extent of labeling with anti-HERG antibodies of HEK 293 cells expressing EGFP/HERG channels was less when compared to HERG/EGFP channels. In conclusion, both electrophysiological and immunocytochemical studies showed that EGFP/HERG channels themselves have a protein trafficking defect. HERG/EGFP channels have similar properties as untagged HERG channels and, thus, might be especially useful for fluorescence microscopy studies.

Comparison of the effects of metoclopramide and domperidone on HERG channels.

Torsades de pointes (TdP) is a potentially fatal form of ventricular arrhythmia that occurs under conditions where cardiac repolarization is delayed (as indicated by prolonged QT intervals from electrocardiographic recordings). A likely mechanism for QT prolongation and TdP is blockade of the rapid component of the cardiac delayed rectifier K(+) current (I(Kr)), which is encoded by HERG (human ether-a-go-go-related gene). The gastroprokinetic agent cisapride is a potent blocker of HERG currents and serious cardiac arrhythmias and deaths from TdP and ventricular fibrillation have been reported in patients taking cisapride. The aim of the present study was to compare the effects of the gastroprokinetic agents domperidone and metoclopramide on HERG channels transiently expressed in human embryonic kidney (HEK 293) cells using the whole-cell configuration of the patch-clamp technique. Both domperidone and metoclopramide concentration-dependently blocked HERG currents, and the following values were calculated for IC(50) (the concentrations causing half-maximal inhibition) and n (the Hill coefficient): 57.0 nmol/l and 0.99 for domperidone, 5.4 micromol/l and 0.95 for metoclopramide. The observation that the extent of block of HERG currents by domperidone increased at more positive membrane potentials whereas block of HERG currents by metoclopramide displayed a smaller degree of voltage dependency seems to indicate that domperidone and metoclopramide have distinct binding sites on HERG channels. In conclusion, the potency for block of HERG currents is about 100-fold lower for metoclopramide when compared to domperidone.

Mechanism of the insulin-releasing action of alpha-ketoisocaproate and related alpha-keto acid anions.

Alpha-ketoisocaproate directly inhibits the ATP-sensitive K(+) channel (K(ATP) channel) in pancreatic beta-cells, but it is unknown whether direct K(ATP) channel inhibition contributes to insulin release by alpha-ketoisocaproate and related alpha-keto acid anions, which are generally believed to act via beta-cell metabolism. In membranes from HIT-T15 beta-cells and COS-1 cells expressing sulfonylurea receptor 1, alpha-keto acid anions bound to the sulfonylurea receptor site of the K(ATP) channel with affinities increasing in the order alpha-ketoisovalerate < alpha-ketovalerate < alpha-ketoisocaproate < alpha-ketocaproate < beta-phenylpyruvate. Patch-clamp experiments revealed a similar order for the K(ATP) channel-inhibitory potencies of the compounds (applied at the cytoplasmic side of inside-out patches from mouse beta-cells). These findings were compared with the insulin secretion stimulated in isolated mouse islets by alpha-keto acid anions (10 mM). When all K(ATP) channels were closed by the sulfonylurea glipizide, alpha-keto acid anions amplified the insulin release in the order beta-phenylpyruvate < alpha-ketoisovalerate < alpha-ketovalerate approximately alpha-ketocaproate < alpha-ketoisocaproate. The differences in amplification apparently reflected special features of the metabolism of the individual alpha-keto acid anions. In islets with active K(ATP) channels, the first peak of insulin secretion triggered by alpha-keto acid anions was similar for alpha-ketoisocaproate, alpha-ketocaproate, and beta-phenylpyruvate but lower for alpha-ketovalerate and insignificant for alpha-ketoisovalerate. This difference from the above orders indicates that direct K(ATP) channel inhibition is not involved in the secretory responses to alpha-ketoisovalerate and alpha-ketovalerate, moderately contributes to initiation of insulin secretion by alpha-ketoisocaproate and alpha-ketocaproate, and is a major component of the insulin-releasing property of beta-phenylpyruvate.