Action of MK-7264 (gefapixant) at human P2X3 and P2X2/3 receptors and in vivo efficacy in models of sensitisation

Abstract

Background and purpose: The P2X3 receptor is an ATP-gated ion channel expressed by sensory afferent neurons and is used as a target to treat chronic sensitisation conditions. The first-in-class, selective P2X3 and P2X2/3 receptor antagonist, the diaminopyrimidine MK-7264 (gefapixant), has progressed to Phase III trials for refractory or unexplained chronic cough. We used patch clamp to elucidate the pharmacology and kinetics of MK-7264 and rat models of hypersensitivity and hyperalgesia to test its efficacy on these conditions.

Experimental approach: Whole-cell patch clamp of 1321N1 cells expressing human P2X3 and P2X2/3 receptors was used to determine mode of MK-7264 action, potency, and kinetics. The analgesic efficacy was assessed using paw withdrawal threshold and limb weight distribution in rat models of inflammatory, osteoarthritic, and neuropathic sensitisation.

Key results: MK-7264 is a reversible allosteric antagonist at human P2X3 and P2X2/3 receptors. Experiments with the slowly desensitising P2X2/3 heteromer revealed concentration- and state-dependency to wash-on, with faster rates and greater inhibition when applied before agonist compared to during agonist application. The wash-on rate (τ value) for MK-7264 at maximal concentrations was much lower when applied before compared to during agonist application. In vivo, MK-7264 displayed efficacy comparable to naproxen in inflammatory and osteoarthritic sensitisation models and gabapentin in neuropathic sensitisation models, increasing paw withdrawal threshold and decreasing weight-bearing discomfort.

Conclusions and implications: MK-7264 is a reversible and selective P2X3 and P2X2/3 antagonist, exerting allosteric antagonism via preferential activity at closed channels. Its efficacy in rat models supports its clinical investigation for chronic sensitisation conditions.

Figure 1

MK‐7264 activity at human P2X2, P2X3, and P2X2/3 receptors. (a–c) Representative concatenated traces of whole‐cell patch‐clamp recordings for human P2X3 (a), P2X2/3 (b), and P2X2 (c) receptor currents in 1312N1 stably transfected cells. Fully recovered P2X3 currents were evoked by 10‐μM α,β‐meATP applied for 1 s every 4 min and 10‐μM α,β‐meATP applied for 2 s at 30‐s intervals for P2X2/3 currents. P2X2 currents were evoked by 10‐μM ATP every 4 min. MK‐7264 was applied 4 min, 30 s, and 4 min prior to agonist application for P2X3 (1 μM), P2X2/3 (1 μM), and P2X2 (10 μM) receptors, respectively. Agonist application is indicated by the perfusion bar. Currents shaded in grey are in the presence of MK‐7264. (d) MK‐7264 concentration–dependent inhibition curve at human P2X3 (closed circles) and P2X2/3 receptors (open circles; N = 5). Data sets were compared with an F test and found to be significantly different from one another (F = 11.74, P = .018). (e) Effect of MK‐7264 on α,β‐meATP concentration–response curve at P2X3 receptors. Currents are in the presence of 300‐nM MK‐7264 (open circles) or vehicle control (closed circles; N = 6). MK‐7264 was applied for 4 min prior to agonist application. EC₅₀ values were compared with a paired sample t test and found to be significantly different (P = .018). The holding potential was −80 mV for all experiments

Figure 2

MK‐7264 antagonism has different on rates when applied before or during agonist application at the human P2X2/3 receptor. (a) Representative trace showing effect of a 30 s pre‐incubation with 3‐μM MK‐7264 on currents evoked by 10‐μM α,β‐meATP at human P2X2/3 receptors. Agonist application is indicated by the perfusion bar. Currents shaded in grey are in the presence of MK‐7264. (b) The effect of varying the pre‐incubation time of MK‐7264 (3 μM) on its antagonism of P2X2/3 responses (10‐μM α,β‐meATP, 2 s; closed circles; N = 7), and the effect of wash‐off time on response recovery following maximal antagonism (open circles). Time constants (τ values) were calculated from single exponential curve fits. Data were normalised to the control response prior to MK‐7264 application. (c) Representative trace showing effect of rapid co‐application of agonist and MK‐7264. Responses were evoked with 10‐μM α,β‐meATP for 2 s followed by co‐application with 3‐μM MK‐7264 or vehicle. (d) Relationship between magnitude of P2X2/3 antagonism and wash‐on time during α,β‐meATP application (open circles) and before application (closed circles). Data were normalised to response prior to MK‐7264 application (closed circles) or the magnitude of response at the same time point in the presence of vehicle (open circles; N = 5). Data sets were compared with a one‐way ANOVA (F = 45.95, P < .05) followed by Tukey's post hoc test

Figure 3

MK‐7264 antagonism of human P2X3 receptor is associated with an onset of slowed desensitisation. (a) Representative trace showing a diminished peak response and appearance of slow desensitisation are associated with MK‐7264 antagonism. (b) Representative superimposed trace showing time‐dependent recovery of P2X3 fast desensitisation following MK‐7264 (300 nM) wash‐off. Agonist applied every 4 min. (c) Relationship between MK‐7264 concentration and the rate of desensitisation. Time constants (τ values) were derived from single exponential curves fitted to the desensitisation phase 250–750 ms after the peak response. The τ values were compared with a Kruskal–Wallis ANOVA (P < .05) followed by Dunn's post hoc test; N = 6. (d) The change in shape of the current following MK‐7264 antagonism quantified as a fractional response (residual current, 1 s after onset of agonist response over peak current) and compared with a Kruskal–Wallis ANOVA followed by Dunn's post hoc test. *P < 0.05 versus control response; N = 7

Figure 4

MK‐7264 efficacy in a rat model of inflammatory hyperalgesia. CFA hind paw injection model of inflammatory hyperalgesia. (a) Effect of a single p.o. dose of vehicle, naproxen, or MK‐7264 on paw withdrawal threshold determined by Von Frey filament testing (N = 10). (b) Effect of single p.o. dose of vehicle, naproxen, or MK‐7264 on postural equilibrium (weight bearing) of inflamed versus non‐inflamed hindlimb (N = 10). Naproxen servesd as a positive control. Sodium carboxymethyl cellulose is the vehicle control. Drug treatments were compared with a two‐way repeated measures ANOVA followed by Fisher least significant difference post hoc test. *P < .05 versus vehicle control

Figure 5

MK‐7264 efficacy in a rat model of osteoarthritic pain. MIA hind knee injection model of osteoarthritic pain. (a) Effect of single p.o. dose of vehicle and naproxen and a range of MK‐7264 doses on paw withdrawal threshold determined by Von Frey filament testing (N = 10). (b) Effect of single p.o. dose of vehicle and naproxen or a range of MK‐7264 doses on postural equilibrium (weight bearing) of injected versus non‐injected hindlimb (N = 10). Drugs were administered 13 days following MIA injection. Naproxen served as a positive control. Sodium carboxymethyl cellulose is the vehicle control. Drug treatments were compared with a two‐way repeated measures ANOVA followed by Fisher least significant difference post hoc test.*P < 0.05 versus vehicle control

Figure 6

MK‐7264 efficacy in a rat spared nerve model of neuropathic pain. Sciatic nerve injury model induced by axotomy of peroneal and tibial nerves. Phase I experiments represent single p.o. dosing 10 days after surgery. Following completion of the Phase I dosing regimen, Phase II experiments were undertaken using p.o. dosing every 12 hr for 5.5 days to test for drug tolerance. (a, b) Effect of Phase I dosing (a) and Phase II dosing (b) regimens on paw withdrawal threshold determined by Von Frey filament testing (N = 10). (c, d) Effect of Phase I dosing (c) and Phase II dosing (d) regimens on postural equilibrium (weight bearing) of injected versus non‐injected hindlimb (N = 10). Gabapentin served as a positive control. Sodium carboxymethyl cellulose is the vehicle control. Drug treatments were compared with a two‐way repeated measures ANOVA followed by Fisher least significant difference post hoc test.*P < 0.05 versus vehicle control

Figure 7

Plasma concentration of MK‐7264 following p.o. dosing in rats. Quantification of MK‐7264 plasma concentration following p.o. dosing with 10 (open circles), 20 (closed circles), and 60 (squares) mg·kg⁻¹ at 0, 1, and 3 hr after dosing (N = 5). Sampling from orbital blood