Donepezil inhibits neuromuscular junctional acetylcholinesterase and enhances synaptic transmission and function in isolated skeletal muscle

Abstract

Background and purpose: Donepezil, a piperidine inhibitor of acetylcholinesterase (AChE) prescribed for treatment of Alzheimer's disease, has adverse neuromuscular effects in humans, including requirement for higher concentrations of non-depolarising neuromuscular blockers during surgery. Here, we examined the effects of donepezil on synaptic transmission at neuromuscular junctions (NMJs) in isolated nerve-muscle preparations from mice.

Experimental approach: We measured effects of therapeutic concentrations of donepezil (10 nM to 1 μM) on AChE enzymic activity, muscle force responses to repetitive stimulation, and spontaneous and evoked endplate potentials (EPPs) recorded intracellularly from flexor digitorum brevis muscles from CD01 or C57BlWld^S mice.

Key results: Donepezil inhibited muscle AChE with an approximate IC₅₀ of 30 nM. Tetanic stimulation in sub-micromolar concentrations of donepezil prolonged post-tetanic muscle contractions. Preliminary Fluo4-imaging indicated an association of these contractions with an increase and slow decay of intracellular Ca²⁺ transients at motor endplates. Donepezil prolonged spontaneous miniature EPP (MEPP) decay time constants by about 65% and extended evoked EPP duration almost threefold. The mean frequency of spontaneous MEPPs was unaffected but the incidence of 'giant' MEPPs (gMEPPs), some exceeding 10 mV in amplitude, was increased. Neither mean MEPP amplitude (excluding gMEPPs), mean EPP amplitude, quantal content or synaptic depression during repetitive stimulation were significantly altered by concentrations of donepezil up to 1 μM.

Conclusion and implications: Adverse neuromuscular signs associated with donepezil therapy, including relative insensitivity to neuromuscular blockers, are probably due to inhibition of AChE at NMJs, prolonging the action of ACh on postsynaptic nicotinic acetylcholine receptors but without substantively impairing evoked ACh release.

Keywords: Alzheimer's disease; acetylcholine; acetylcholinesterase; anticholinesterase; endplate potential; muscle contraction; neuromuscular block; neuromuscular junction.

FIGURE 1

Donepezil inhibits muscle acetylcholinesterase (AChE) and antagonises non‐depolarising neuromuscular block. (a) Inhibition of mouse muscle AChE activity with increasing concentration of donepezil based on thiocholine production (see Methods). Each data point is the average of triplicate measurements from one homogenate prepared from 2 to 5 mice. The IC₅₀ based on the sigmoidal best fit to the data was estimated to be about 30–50 nM but inhibition was evidently incomplete up to concentrations of 10 μM. (b) Similar analysis from homogenates of mouse neocortex (mean of triplicate measurements of one mouse brain per point). The IC₅₀ was about 10–20 nM and inhibition was virtually complete in 100 nM donepezil. (c) Continuous recording of flexor digitorum brevis (FDB) muscle twitch amplitudes before and after adding a submaximal blocking dose of atracurium (ATc; 2 μM) to the recording chamber. Donepezil (100 nM, DPZ) partly antagonised the partial neuromuscular block. Twitch tension was restored after washing both compounds from the recording chamber with normal mammalian physiological saline (MPS). (d–g) Trains‐of‐four twitch responses from the recording shown in (c) on an expanded time scale; in MPS (d), after adding atracurium (e), following further addition of donepezil (f) and after restoring normal MPS (g).

FIGURE 2

Donepezil prolongs tetanic muscle force. (a–c) Isometric tension recordings from isolated flexor digitorum brevis (FDB) muscles during repetitive stimulation of the tibial nerve supply for 2 s at 20 Hz. Each panel shows the tension response in mammalian physiological saline (MPS) and 1–1.5 h after adding donepezil (DPZ) at the concentrations indicated. In control solution, tetanic muscle responses relaxed promptly at the end of the stimulus train but relaxation was progressively delayed (‘aftercontraction’) after adding donepezil. (d) Neostigmine (NEO; 100 nM), a carbamate anti‐AChE, had similar effects on relaxation of muscle tetani evoked by tibial nerve stimulation. (e, f) Time course of development and decay of aftercontractions after adding donepezil (e) or neostigmine (f) at the range of concentrations indicated, expressed as a percentage of the total area under the curve (AUC) during and after tetanic stimulation. Each point represents the mean ± SEM of recordings from 5 to 6 mice. The lowest to highest lines represent data in MPS (0), 10 nM, 100 nM and 1 μM donepezil (e), or MPS, 10 nM, 100 nM and 10 μM neostigmine (f). (g) Data summarising fractional AUC measurements for aftercontractions recorded 1 h after adding either donepezil or neostigmine at the concentrations indicated. Each point represents data from one nerve‐muscle preparation (N = 5–6 mice). Bars indicate mean ± 95% confidence intervals. *P < 0.05, significantly different from MPS; paired Wilcoxon test.

FIGURE 3

Donepezil prolongs miniature endplate potentials (MEPPs) and endplate potentials (EPPs). (a–d) Examples of averaged MEPPs (a, c) and single EPPs (b, d) from recordings of individual muscle fibres made in control mammalian physiological saline (MPS) and donepezil (1 μM), as indicated. Red superimposed lines in (a, c) indicate non‐linear least squares single exponential fits to the averaged MEPP decays. The single exponential decay time constant (τ) of the averaged MEPP in these two fibres is indicated. (e, f) Donepezil concentration–response data (blue points) and non‐linear least squares fits (red curves) for MEPP decay time constant (g) and EPP duration (h). Each point represents mean data from 9 to 30 muscle fibres (N = 3–5 mice). Error bars are 95% confidence intervals based on numbers of mice.

FIGURE 4

Donepezil increases incidence and magnitude of gMEPPs. (a, b) Slow time‐base recordings of miniature endplate potentials (MEPPs) from a muscle fibre in mammalian physiological saline (MPS) (a) and, in a different fibre, in 1 μM donepezil (b). In both these examples, MEPPs are interspersed with ‘giant’ MEPPs (gMEPPs) more than twice the median MEPP amplitude. Markings in red represent spontaneous (MEPP) events as detected by Minianalysis software. (c, d) Examples of gMEPPs, corresponding to those indicated by arrows in (a, b) and shown on a faster time‐base. The example shown in (d), recorded in 1 μM donepezil, evidently comprises distinctive steps, suggesting spontaneous but desynchronised release of several vesicular quanta of acetylcholine (ACh) from the motor nerve terminal supplying this neuromuscular junction (NMJ). Note differences in voltage calibration in (a, b) and in (c, d). (e, f) MEPP amplitude histograms from the recordings illustrated by traces shown in (a–d), emphasising the magnitude of gMEPPs recorded in donepezil. Note differences in scale of the abscissa in (f). (g) Plot of gMEPP amplitudes recorded in 30 s recordings of spontaneous activity in control MPS (CON) and in increasing concentrations of donepezil. Each point represents one gMEPP and error bars are means ± SD (n = 3–10 muscle fibres per mouse; N = 3–5 mice at each concentration). The largest gMEPPs were observed in 100 nM to 1 μM donepezil. (h) Mean MEPP frequency (excluding gMEPPs) overall was unaffected by increasing concentrations of donepezil. Each point represents the mean frequency based on MEPP recordings from flexor digitorum brevis (FDB) muscles dissected from one mouse and error bars indicate the mean ± SEM of these measurements (n = 3–10 muscle fibres per mouse; N = 3–5 mice).

FIGURE 5

Endplate potential (EPP) quantal content is unimpaired by donepezil. (a) Mean miniature EPP (MEPP) amplitudes, corrected to a resting membrane potential of −70 mV. Each point represents the mean of recordings from 3 to 10 muscle fibres per mouse. There was no significant effect of donepezil over the range of concentrations indicated. (b) Mean amplitudes of the first EPP recorded in trains of 30, evoked at 1 Hz. Each point is the average of recordings from 3 to 10 muscle fibres per mouse, after adjusting EPP amplitudes for non‐linear summation and correcting to a resting membrane potential of −70 mV (see Methods). (c) Quantal content of the first EPP in each recording, obtained directly by dividing corrected EPP amplitude by the corresponding corrected mean MEPP amplitude (excluding giant MEPPs [gMEPPs]). Each point represents the estimate of quantal content from one muscle fibre. Bars represent mean ± SD. (d) Mean quantal content of initial EPPs recorded from preparations dissected from each mouse. Each point represents the mean quantal content from one mouse. There was no significant effect of donepezil on mean quantal content. (e) EPP depression calculated from the ratio of the fourth EPP to the first (EPP_1–4) in trains of 30 EPPs evoked at 1 Hz. Each point represents data from one muscle fibre. Bars indicate mean ± SD. (f) Depression of EPP quantal content based on ratio of mean quantal contents of the first four EPPs (m_1–4) to the mean quantal content of EPPs from the 10th to the last EPP (normally the last 20 EPPs) in the train (m₁₀₊). Each point is the mean depression per mouse based on recordings from 3 to 10 muscle fibres in flexor digitorum brevis (FDB) preparations from one mouse. Error bars show the means of these values with 95% confidence intervals.