High-throughput cell-based assays for identifying antagonists of multiple smoking-associated human nicotinic acetylcholine receptor subtypes

Abstract

There is substantial evidence that in addition to nicotine, other compounds found in tobacco smoke significantly influence smoking behavior. Further, recent years have seen an explosion in the availability of non-combusted products that deliver nicotine, such as e-cigarettes and "home-brew" vaping devices that are essentially unregulated. There are many thousands of compounds in tobacco smoke alone, and new products are constantly introducing new compounds. Uncovering which of these compounds are active, across multiple smoking-relevant subtypes of the nicotinic acetylcholine receptor (nAChR) that influence tobacco/nicotine addiction, requires a high-throughput screening (HTS) approach. Accordingly, we developed a panel of HTS-friendly cell-based assays, all performed in the same cellular background and using the same membrane potential dye readout, to measure the function of the α3β4-, α4β2-, and α6β2-nAChR subtypes. These subtypes have each been prominently and consistently associated with human smoking behavior. We validated our assays by performing pilot screening of an expanded set of the Prestwick FDA-approved drug library. The screens displayed excellent performance parameters, and moderate hit rates (mean of 1.2% across all three assays) were achieved when identifying antagonists (chosen since effects of endogenous antagonists on consumption of nicotine/tobacco products are under-studied). Validation rates using an orthogonal assay (⁸⁶Rb⁺ efflux) averaged 73% across the three assays. The resulting panel of assays represents a valuable new platform with which to screen and identify nAChR subtype-selective compounds. This provides a resource for identifying smoking-related compounds in both combusted and non-combusted tobacco products, with potential relevance in the search for additional smoking-cessation therapies.

Keywords: Cell-based screening; Membrane potential assays; Nicotinic acetylcholine receptor.

Fig. 1.

Nicotine induction verification in membrane potential assay. SH-EP1 cells expressing nAChR subtypes (A) α3β4, (B) α4β2, or (C) α6/3 β2 β3 were seeded at 7000 cells per well and treated with a 12-point, 1:3 serial dilution from 100 μM of (−)-nicotine. Concentration-dependent response curves were achieved in three independent experiments on different days. Each data point represents mean ± SD (four replicates per condition/concentration). Curve fitting was performed using nonlinear regression four-parameter methods from GraphPad Prism 7, and EC₅₀ values are summarized in (D).

Fig. 2.

Validation of known nAChR antagonists. SH-EP1-α4 β2-nAChR cells were treated in a concentration-dependent manner with three known nicotine antagonists, (A) Dh βE, (B) hexamethonium, or (C) mecamylamine, in the presence of 500nM of nicotine. Data are shown for various incubation times, and all data were normalized to DMSO control. (D) Cells expressing different subtypes were treated with a range of mecamylamine concentrations serially diluted from 100 μM to 0.56 nM in the presence of nicotine at the pre-determined EC₉₀ value (20 μM for α3β4 and 500 nM for α4β2 and α6/3 β2 β3). All data are presented as means +/− SD of quadruplicate wells (n = 4). Curves were fitted using nonlinear regression four parameter methods from GraphPad Prism 7 and IC₅₀ values (M) are summarized as table inset for each antagonist at individual time point and/or receptor.

Fig. 3.

HTS results for SH-EP1-α4 β2-nAChR. Raw fluorescent signals acquired during two independent HTS runs were normalized to positive-signal (nicotine EC₉₀) and zero-signal (DMSO) internal plate controls. The normalized fluorescent signals (NFS) were analyzed for run-to-run reproducibility and selection of hits to be advanced in orthogonal assays. (A, B) Scatter plots of the NFS of two independent HTS runs in SH-EP1 cells expressing human α4β2, evaluating assay reproducibility. Please note that data from internal controls were plotted on a separate panel (A), adjacent to those from test compounds (B), to avoid extensive overlap and thus crowding. HTS results were highly reproducible as evaluated for (A) internal positive controls (antagonist mecamylamine in EC₉₀ of nicotine, yellow) and negative controls (DMSO in EC₉₀ of nicotine, blue) (R² = 0.987) and for (B) test compounds from library (green triangles, R² = 0.781). (C, D) Occurrence histograms capturing the distribution of NFS in test wells for run 1 (C, red) and run 2 (D, blue) respectively. Hit selection thresholds (mean – 3 SD) are marked as vertical lines on either plot. Selected hits are represented as green histogram bars. (E,F) Scatterplots of the NFS as a function of the test compounds for HTS run 1 (E) and run 2 (F) are presented, each point represents a single compound. Please note that internal control values are not shown on these panels since they would reduce clarity. The no antagonist control value (baseline) is, by definition, 1 in these plots while the full agonist control value is, by definition, 0. The plots show clusters of signals at baseline (NFS = 1) for all inactive compounds. The hit selection threshold for this run (mean – 3 SD) is marked as a horizontal dotted grey line. Red dots with NFS values below the selection threshold indicate the overlapping hits from both runs, which were further tested in orthogonal assays.

Fig. 4.

Validation and specificity testing of HTS antagonist hits using an orthogonal assay. SH-EP1 cells stably expressing human α3β4-, α4β2-, or α6/3 β2 β3-nAChR were loaded with ⁸⁶Rb⁺, pretreated for 5 min with compounds identified as antagonist hits in HTS, then exposed to an EC₉₀ concentration of carbamylcholine for 3 min in the presence of the test compound. Atropine (1.5 μM) was included in all wells to block the potential function of muscarinic acetylcholine receptors that might otherwise be stimulated by carbamylcholine. (A) All plates were normalized to internal maximum and minimum controls, as described in the Materials and Methods section. The no-antagonist control values for each assay were used to define 0% inhibition of carbamylcholine-induced activity, while the full-antagonist control values were used to define 100% inhibition. Activity measured at α3β4-nAChR is shown with blue symbols, that at α4β2-nAChR with red symbols, and that at α6/3 β2 β3-nAChR with green symbols. Dashed lines are color-matched to data symbols and designate the 3 X SD thresholds above the no-antagonist controls, used to determine validation of antagonist activity in each of the assays. Compounds are grouped according to their validated activity across the three nAChR subtypes addressed by this study. (B) Venn Diagram presents overlap of validated hits across the α3β4-, α4β2-, and α6/3 β2 β3-nAChR subtypes.