The AMPK Activator A769662 Blocks Voltage-Gated Sodium Channels: Discovery of a Novel Pharmacophore with Potential Utility for Analgesic Development

Abstract

Voltage-gated sodium channels (VGSC) regulate neuronal excitability by governing action potential (AP) generation and propagation. Recent studies have revealed that AMP-activated protein kinase (AMPK) activators decrease sensory neuron excitability, potentially by preventing sodium (Na+) channel phosphorylation by kinases such as ERK or via modulation of translation regulation pathways. The direct positive allosteric modulator A769662 displays substantially greater efficacy than other AMPK activators in decreasing sensory neuron excitability suggesting additional mechanisms of action. Here, we show that A769662 acutely inhibits AP firing stimulated by ramp current injection in rat trigeminal ganglion (TG) neurons. PT1, a structurally dissimilar AMPK activator that reduces nerve growth factor (NGF) -induced hyperexcitability, has no influence on AP firing in TG neurons upon acute application. In voltage-clamp recordings, application of A769662 reduces VGSC current amplitudes. These findings, based on acute A769662 application, suggest a direct channel blocking effect. Indeed, A769662 dose-dependently blocks VGSC in rat TG neurons and in Nav1.7-transfected cells with an IC50 of ~ 10 μM. A769662 neither displayed use-dependent inhibition nor interacted with the local anesthetic (LA) binding site. Popliteal fossa administration of A769662 decreased noxious thermal responses with a peak effect at 5 mins demonstrating an analgesic effect. These data indicate that in addition to AMPK activation, A769662 acts as a direct blocker/modulator of VGSCs, a potential mechanism enhancing the analgesic property of this compound.

Fig 1. A769662 applied acutely to rat TG neurons blocks AP firing evoked by ramp current stimuli.

(A) Representative action potential traces in response to ramp current stimuli before and after acute application of (5 sec) of A769662 (200 μM). Acute application of A769662 (n = 5) onto rat TG neurons significantly reduced numbers of action potentials (B) and reversed time-to-first action potential (AP) peak compared to control (n = 5). This effect was reversible with washout (n = 5). Differences in the mean numbers of action potentials among groups were analyzed by comparing the slopes and intercepts generated from linear regression Comparisons among groups for time to first spike were performed by two-way ANOVA. Colored stars denote significant effects compared to control group. * p < 0.05, ** p < 0.01 and *** p < 0.001.

Fig 2. PT1 and resveratrol inhibit NGF-induced hyperexcitability but have no acute effect on action potential firing.

(A) Patch clamp analysis of rat TG neurons cultured in the presence of NGF (50 ng/ml, n = 11) show an increase in the number of ramp evoked action potentials compared to control (n = 8) which is reversed by PT1 (100 μM; 1 hr, n = 11) and resveratrol (200 μM, 1 hr, n = 6). Colored stars denote significant effects compared to NGF group. * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001. (B) and (C) Acute application of PT1 (30 μM, n = 10; 100 μM, n = 8) demonstrates no influence on action potential firing compared to control. The number of ramp current evoked action potentials was normalized to control. (D) Acute application of other AMPK activators (resveratrol, 200 μM, n = 6; metformin, 20 mM, n = 6) does not block VGSCs.

Fig 3. A769662 dose-dependently blocks rat TG VGSC current amplitude.

(A) Representative Na⁺ current traces in TG neurons in the presence of A769662. Currents were elicited in TG neurons with a current step protocol initiated from -80 mV to 0 mV for 25 msec from a holding potential of -80 mV. (B) Acute application of A769662 reduced the Na⁺ current amplitude in TG neurons (n = 6, **** p < 0.0001). C) Data generated was fitted to a hill equation to plot the dose response curve for percent current amplitude block by A769662. A769662 inhibited Na⁺ current amplitude with an IC₅₀ of 10 μM. D) Peak current vs time plot of the effect of A769662 on Na⁺ current amplitude. Arrows correspond to A769662 (blue) and washout (red) in A) and B).

Fig 4. Lack of use-dependent effects of A769662 on rat TG neurons.

30 repetitive 25 ms depolarizing pulses of -10 mV were applied from a holding potential of -80 mV at 0.5Hz (A) and 5Hz (B) in the absence (●) and presence (■) of A769662 (200 μM). Peak current amplitude at each pulse was normalized by the peak current amplitude of the first pulse under each condition and plotted vs the pulse number. (C) Fraction of current at the 30^th depolarizing pulse at 0.5 Hz and 5 Hz in the absence and presence of A769662 (n = 5–15, two-way ANOVA, * p < 0.05, *** p <0.001).

Fig 5. A769662 blocks hNa_v1.7 currents in HEK cells.

(A) A step from -80 mV to 0 mV for 25 ms was used to elicit currents in hNa_v1.7-transfected HEK cells with or without A769662 application. (B) Acute application of A769662 (200 μM) reduced the Na⁺ current amplitude in hNa_v1.7-transfected HEK cells (n = 10, one-way ANOVA, **** p < 0.0001 C) Peak current vs time plot of the effect of A769662 on hNa_v1.7 currents. Arrows correspond to A769662 (blue) and washout (red) in A) and B).

Fig 6. A769662 showed no binding at the local anesthetic binding site.

(A, B) Nav1.7 current traces recorded under control (solid line), presence of 200μM A769662 (dash line), and washout (dot lines) conditions respectively from a cell expressing WT Nav1.7 channels (A) or a cell expressing F1737A/Y1744A mutant channels (B); (C,D) 200μM A769662 reduced Nav1.7 sodium channel currents from cells expressing WT channels (C) and cells expressing F1737A/Y1744A mutant channels (D), and washing out could partially reverse the inhibition effects for both WT and F1737A/Y1744A mutant channels.

Fig 7. A769662 exhibits local anesthetic effects in rats.

(A) Administration of A769662 (30 μg, 100 μg and 300 μg) into the popliteal fossa produces a dose-dependent nerve block by reversing the paw withdrawal latency to noxious thermal heat using the Hargreaves method compared to vehicle treated rats (n = 4–7, regular two-way ANOVA, * p < 0.05 and *** p < 0.001). (B) Paw withdrawal latency plotted as a function of A769662 or PT1 dose. Dose-response curve of A769662 was generated by fitting data to the Hill equation yielding an IC₅₀ of 90 μg.