Optimisation and validation of a medium-throughput electrophysiology-based hERG assay using IonWorks HT.

INTRODUCTION: Regulatory and competitive pressure to reduce the QT interval prolongation risk of potential new drugs has led to focus on methods to test for inhibition of the human ether-a-go-go-related gene (hERG)-encoded K+ channel, the primary molecular target underlying this safety issue. Here we describe the validation of a method that combines medium-throughput with direct assessment of channel function. METHODS: The electrophysiological and pharmacological properties of hERG were compared using two methods: conventional, low-throughput electrophysiology and planar-array-based, medium-throughput electrophysiology (IonWorks HT). A pharmacological comparison was also made between IonWorks HT and an indirect assay (Rb+ efflux). RESULTS: Basic electrophysiological properties of hERG were similar whether recorded conventionally (HEK cells) or using IonWorks HT (CHO cells): for example, tail current V1/2 -12.1+/-5.0 mV (32) for conventional and -9.5+/-6.0 mV (46) for IonWorks HT (mean+/-S.D. (n)). A key finding was that as the number of cells per well was increased in IonWorks HT, the potency reported for a given compound decreased. Using the lowest possible cell concentration (250,000 cells/ml) and 89 compounds spanning a broad potency range, the pIC50 values from IonWorks HT (CHO-hERG) were found to correlate well with those obtained using conventional methodology (HEK-hERG)(r=0.90; p<0.001). Further validation using CHO-hERG cells with both methods confirmed the correlation (r=0.94; p<0.001). In contrast, a comparison of IonWorks HT and Rb+ efflux data with 649 compounds using CHO-hERG cells showed that the indirect assay consistently reported compounds as being, on average, 6-fold less potent, though the differences varied depending on chemical series. DISCUSSION: The main finding of this work is that providing a relatively low cell concentration is used in IonWorks HT, the potency information generated correlates well with that determined using conventional electrophysiology. The effect on potency of increasing cell concentration may relate to a reduced free concentration of test compound owing to partitioning into cell membranes. In summary, the IonWorks HT hERG assay can generate pIC50 values based on a direct assessment of channel function in a timeframe short enough to influence chemical design.

Stimulation of neuropeptide Y gene expression by brain-derived neurotrophic factor requires both the phospholipase Cgamma and Shc binding sites on its receptor, TrkB.

In PC12 cells, it has been previously reported that nerve growth factor stimulates neuropeptide Y (NPY) gene expression. In the current study we examined the signalling pathways involved in this effect by transiently expressing in PC12 cells the receptor (TrkB) for the related neurotrophin, brain-derived neurotrophic factor (BDNF). BDNF caused a 3-fold induction of luciferase expression from a transiently co-transfected plasmid possessing the firefly luciferase gene under the control of the NPY promoter. This effect of BDNF was completely blocked by either a Y484F mutation in TrkB (which blocks high-affinity Shc binding to TrkB) or by a Y785F substitution [which blocks the binding, phosphorylation and activation of phospholipase Cgamma (PLCgamma)]. Activation of the NPY promoter by neurotrophin-3 in PC12 cells overexpressing TrkC was also completely blocked by a naturally occurring kinase insert which prevents the high-affinity binding of Shc and PLCgamma. NPY promoter activation by BDNF was blocked by PD98059, suggesting a role for mitogen-activated protein kinase (MAP kinase). Stimulation of NPY gene expression by PMA, but not by BDNF, was blocked by Ro-31-8220, a protein kinase C inhibitor, excluding a role for this serine/threonine protein kinase in the effect of BDNF. In addition, BDNF did not cause an elevation in cytosolic Ca2+ concentration. Taken together, our results suggest that stimulation of the NPY promoter by BDNF requires the simultaneous activation of two distinct pathways; one involves Shc and MAP kinase, and the other appears to be PLCgamma-independent but requires an intact tyrosine-785 on TrkB and so may involve an effector of TrkB signalling that remains to be identified.

Direct inhibition of 5-hydroxytryptamine3 receptors by antagonists of L-type Ca2+ channels.

Homopentameric complexes of either the A or As subunit of the 5-hydroxytryptamine3 receptor form Ca(2+)-permeable channels that can be activated by the selective agonist 1-(m-chlorophenyl)-biguanide (mCPBG). In both N1E-115 neuroblastoma cells and human embryonic kidney 293 cells stably expressing the 5-HT3 receptor As subunit, (+)-verapamil, (-)-verapamil, diltiazem, and nimodipine caused reversible and concentration-dependent (IC50 = 2.5-6.5 microM) inhibition of the increases in cytosolic [Ca2+] evoked by mCPBG. In voltage-clamped human embryonic kidney 293 cells stably expressing the 5-HT3 receptor As subunit, similar concentrations of the Ca2+ channel antagonists (IC50 = 3.0-6.8 microM) accelerated the rate at which 5-HT-evoked currents decayed without affecting the amplitude of the peak current. In equilibrium competition binding assays to membranes from Sf9 cells infected with the 5-HT3 receptor As subunit, [3H]mCPBG and [3H]granisetron were displaced by (+)-verapamil, (-)-verapamil, and diltiazem; (+)-verapamil was approximately 10-fold more potent than (-)-verapamil and approximately-30-fold more potent than diltiazem. Nimodipine neither displaced [3H]granisetron binding nor affected its displacement by diltiazem and (+)-verapamil. The stereoselectivity of verapamil binding, which contrasts with the similar potency of each isomer in functional assays, was maintained when the incubations were performed at 20 degrees or when an antagonist of the 5-HT3 receptor, [3H]granisetron, was used as the radioligand. The interaction between verapamil and either [3H]mCPBG or [3H]granisetron binding was not competitive. We conclude that the inhibition of [3H]mCPBG binding by diltiazem and verapamil is mediated by a site that is distinct from both the agonist-binding site and from the site through which nimodipine inhibits 5-HT3 receptor function. Our results provide evidence for allosteric regulation of agonist binding to 5-HT3 receptors and the first example of a ligandgated ion channel whose function is directly inhibited by members of all three major classes of L-type Ca2+ channel antagonists.

Ca2+ permeability of cloned and native 5-hydroxytryptamine type 3 receptors.

We have used single-cell imaging of fura-2-loaded cells to examine the Ca2+ signals evoked by activation of 5-hydroxytryptamine type 3 (5-HT3) receptors in undifferentiated N1E-115 neuroblastoma cells and in human embryonic kidney (HEK) 293 cells transfected with either of the two cloned 5-HT3 receptor subunits. The selective 5-HT3 receptor agonist 1-(m-chlorophenyl)-biguanide (mCPBG) caused a concentration-dependent increase in the cytoplasmic Ca2+ concentration ([Ca2+]i) in N1E-115 cells and in HEK 293 cells transfected with either the 5-HT3 A subunit or the 5-HT3 As subunit. In each case, the [Ca2+]i rise was steeply dependent on the mCPBG concentration (nH = 2-4) and abolished by removal of extracellular Ca2+ or addition of ondansetron. Pretreatment of N1E-115 cells with thapsigargin, caffeine, and ryanodine to deplete intracellular Ca2+ stores had no effect on the mCPBG-evoked Ca2+ signals, indicating that they result entirely from stimulated Ca2+ entry. The steep concentration-effect curves therefore are not a consequence of amplification of Ca2+ influx by Ca(2+)-induced Ca2+ release from intracellular stores and probably reflect cooperative activation of 5-HT3 receptors by mCPBG. Depolarization of transfected HEK 293 cells with medium containing increased K+ concentrations invariably failed to evoke an increase in [Ca2+]i, confirming the absence of voltage-gated Ca2+ channels and indicating that the mCPBG-evoked rise in [Ca2+]i results from Ca2+ permeation of 5-HT3 receptors. However, in N1E-115 cells and transfected HEK 293 cells, both extracellular Na+ and K+ substantially inhibited the Ca2+ influx evoked by activation of 5-HT3 receptors, possibly by inhibition of agonist binding or by competition with Ca2+ for permeation of the channel. We conclude that 5-HT3 receptors are Ca2+ permeant, that the Ca2+ influx is sufficient to generate a significant rise in [Ca2+]i, and that, because the A and As subunits behave similarly, conflicting electrophysiological analyses of Ca2+ currents cannot be explained by differences between these two subunits.

BQ-123, cyclo(-D-Trp-D-Asp-Pro-D-Val-Leu), is a non-competitive antagonist of the actions of endothelin-1 in SK-N-MC human neuroblastoma cells.

SK-N-MC cells, derived from a human neuroblastoma, respond to endothelin (ET) peptides with an increase in the free intracellular calcium concentration. The response is biphasic, with the secondary plateau phase being abolished or reduced by removal of extracellular Ca2+ or by the presence of 100nM nitrendipine. Restoration of Ca2+ to the bathing solution in cells stimulated by ET-1 in the absence of Ca2+ caused the plateau to reappear. The order of potency of ET family peptides was ET-2 greater than or equal to sarafotoxin S6b greater than or equal to ET-1 much greater than ET-3, suggesting that ETA receptors mediate the response. Sarafotoxin S6c and the C-terminal hexapeptide endothelin (16-21) were inactive in these cells. [Ala1,3,11,15]ET-1, a linear analogue of ET-1 which has been suggested to be a selective ETB receptor agonist, was a weak competitive antagonist of the actions of ET-1 in these cells. However, BQ-123, recently introduced as a selective and competitive antagonist at ETA receptors, was a potent non-competitive antagonist of ET-1 giving a 50% reduction in the maximum response at 6nM.