Riluzole blocks human muscle acetylcholine receptors

Abstract

Riluzole, the only drug available against amyotrophic lateral sclerosis (ALS), has recently been shown to block muscle ACh receptors (AChRs), raising concerns about possible negative side-effects on neuromuscular transmission in treated patients. In this work we studied riluzole's impact on the function of muscle AChRs in vitro and on neuromuscular transmission in ALS patients, using electrophysiological techniques. Human recombinant AChRs composed of α(1)β(1)δ subunits plus the γ or ε subunit (γ- or ε-AChR) were expressed in HEK cells or Xenopus oocytes. In both preparations, riluzole at 0.5 μm, a clinically relevant concentration, reversibly reduced the amplitude and accelerated the decay of ACh-evoked current if applied before coapplication with ACh. The action on γ-AChRs was more potent and faster than on ε-AChRs. In HEK outside-out patches, riluzole-induced block of macroscopic ACh-evoked current gradually developed during the initial milliseconds of ACh presence. Single channel recordings in HEK cells and in human myotubes from ALS patients showed that riluzole prolongs channel closed time, but has no effect on channel conductance and open duration. Finally, compound muscle action potentials (CMAPs) evoked by nerve stimulation in ALS patients remained unaltered after a 1 week suspension of riluzole treatment. These data indicate that riluzole, while apparently safe with regard to synaptic transmission, may affect the function of AChRs expressed in denervated muscle fibres of ALS patients, with biological consequences that remain to be investigated.

Figure 1. Effect of riluzole on ACh-evoked whole-cell currents in HEK cells

A, typical whole-cell currents evoked by ACh (100 μm, horizontal bar) alone (C) or in the presence of riluzole (R; 0.5 μm, 30 s pre-treatment) in cells expressing the indicated AChR type. B, duration of pre-treatment with riluzole (0.5 μm) enhances its effect in cells expressing γ- or ɛ-AChRs. The integral of current response (Q_ACh or Q_ACh+riluzole) was calculated over the entire duration of ACh (100 μm) application; riluzole's effect was quantified by the ratio Q_ACh+riluzole/Q_ACh in each cell (4–10 for each point). Grey lines represent the exponential best fit of the experimental points. C, typical whole-cell currents evoked by ACh alone or plus riluzole 50 μm (as in A). D, currents as in A, with GDPβS (100 μm) included in the patch pipette. Notice that riluzole's effect is preserved. E, summary of riluzole's effect on ACh-evoked responses, quantified as in B. *Q_ACh+riluzole significantly different from Q_ACh (P < 0.001; Student's paired t test). In all panels, holding potential, −60 mV.

Figure 2. Inhibition of ACh-evoked currents by riluzole in oocytes

A, typical currents evoked by ACh (horizontal bars) alone (C) or together with riluzole (R; 50 μm for ɛ-AChR; 10 μm for γ-AChR; 120 s pretreatment for both) in two oocytes expressing AChRs as indicated. B, plot of relative amplitude of currents evoked by ACh (50 μm, −60 mV) plus riluzole at various concentrations (120 s pretreatment), expressed as a percentage of control current (ACh alone) in the same cell, for oocytes expressing γ-AChRs (□, 8/3) or ɛ-AChRs (◯, 10/3). IC₅₀ and n_H were respectively 8.0 ± 1.1 μm and 1.5 ± 0.2 (γ-AChR), and 55.6 ± 0.7 μm and 1.5 ± 0.3 (ɛ-AChR).

Figure 3. Effect of riluzole on unitary ACh-evoked currents in outside-out patches

A and B, unitary events evoked by ACh (100 nm, −80 mV) in outside-out patches expressing γ-AChRs (A) or ɛ-AChRs (B), sequentially exposed to ACh alone (control) or plus riluzole (0.5 μm), as indicated. Bottom panels represent the histograms of channel closed times, obtained in the recordings shown above, best fitted with four (γ-AChR) or three (ɛ-AChR) exponential components. Notice the lengthening of the closed times in the presence of riluzole, as compared to control conditions. Channel openings represented by downward deflections.

Figure 4. Effect of riluzole in cell-attached recordings

A, typical cell-attached recordings in two different HEK cells expressing γ-AChR, including in the patch pipette ACh (100 nm) alone (slope conductance, 31 pS) or plus riluzole (0.5 μm; slope conductance, 34 pS), as indicated. Bottom, the histograms of channel open durations, obtained in the same recordings, were best fitted by two exponential components, with time constants τ_op1 = 1.5 ms (33%) and τ_op2 = 9.5 ms (67%) for control; 1.1 ms (35%) and 7.4 ms (65%) in presence of riluzole. B, cell-attached recording of unitary events in two different patches on the same myotube from an ALS patient, using ACh (100 nm) alone or plus riluzole, as indicated. Channel conductance was 30 pS (control) or 34 pS (+ riluzole). Bottom, the corresponding histograms of channel opening durations, best fitted by two exponential components with τ_op1 = 1.4 ms (39%) and τ_op2 = 12.7 ms (61%) in control conditions, 1.8 ms (36%) and 11.7 ms (64%) in the presence of riluzole. In all panels, channel openings are represented by upward deflections; cells were bathed in standard external solution.

Figure 5. Effect of riluzole during fast ACh applications

A, superimposed currents evoked by sequential applications of ACh (100 μm, 100 ms) alone (C) or together with riluzole (R; 0.5 μm, 30 s pre-treatment) to outside-out patches from HEK cells expressing γ- or ɛ-AChRs, as indicated. B, bar graphs representing relative current integral during 10, 20 and 100 ms from ACh application. For γ-AChRs, riluzole-induced reduction of Q_ACh was significant already at 10 ms (*P < 0.002); for ɛ-AChR, current integral was significantly reduced only at 100 ms (*P = 0.02). In all panels, holding potential, −60 mV.

Figure 6. Riluzole has no effect on neuromuscular transmission in ALS patients

A and B, superimposed traces representing 10 CMAPs evoked by repetitive stimulation of the ulnar (A) and axillary nerve (B) in a patient, during riluzole treatment (left) and at the end of a 1 week medicament washout. Notice that CMAP amplitude remains constant in this individual. C, dispersion plot of the difference between CMAP amplitude evoked by the 1st and 5th stimulus, expressed as a percentage of the amplitude of the 1st response, for all 36 patients considered in the study. Boxes represent mean ± 95% confidence interval of all values. For both ulnar and axillary nerves, values were not modified by riluzole washout (P > 0.9).