The human adult subtype ACh receptor channel has high Ca²⁺ permeability and predisposes to endplate Ca²⁺ overloading

Abstract

Slow-channel congenital myasthenic syndrome, caused by mutations in subunits of the endplate ACh receptor (AChR), results in prolonged synaptic currents and excitotoxic injury of the postsynaptic region by Ca²⁺ overloading. The Ca²⁺ overloading could be due entirely to the prolonged openings of the AChR channel or could be abetted by enhanced Ca²⁺ permeability of the mutant channels. We therefore measured the fractional Ca²⁺ current, defined as the percentage of the total ACh-evoked current carried by Ca²⁺ ions (P_f), for AChRs harbouring the αG153S or the αV249F slow-channel mutation, and for wild-type human AChRs in which P_f has not yet been determined. Experiments were performed in transiently transfected GH4C1 cells and human myotubes with simultaneous recording of ACh-evoked whole-cell currents and fura-2 fluorescence signals. We found that the P_f of the wild-type human endplate AChR was unexpectedly high (P_f∼7%), but neither the αV249F nor the αG153S mutation altered P_f. Fetal human AChRs containing either the wild-type or the mutated α subunit had a much lower P_f (2–3%). We conclude that the Ca²⁺ permeability of human endplate AChRs is higher than that reported for any other human nicotinic AChR, with the exception of α7-containing AChRs (P_f > 10%); and that neither the αG153S nor the αV249F mutations affect the P_f of fetal or adult endplate AChRs. However, the intrinsically high Ca²⁺ permeability of human AChRs probably predisposes to development of the endplate myopathy when opening events of the AChR channel are prolonged by altered AChR-channel kinetics.

Although the permeability of the endplate acetylcholine receptor (AChR) channel to Ca²⁺ ions was demonstrated around three decades ago (Kusano et al. 1975; Bregestovski et al. 1979), the physiological role of the ensuing Ca²⁺ influx has not been established. However, several lines of information suggest that the Ca²⁺ influx through fetal and adult AChRs (γ-AChR and ɛ-AChR, respectively) probably plays a role in formation and function of the normal endplate. Ca²⁺ influx at the endplate might contribute to shape muscle spiking activity, by regulating the activation of depolarizing voltage-gated Na⁺ channels and of repolarizing Ca²⁺-dependent K⁺-channels (Allard et al. 1996; Young & Caldwell, 2005). Ca²⁺ entry through fetal AChR channels has also been noted to promote myotube survival (Bandi et al. 2005).

The Ca²⁺ permeability of endplate AChRs also plays a role in endplate diseases. Specifically, cationic overloading of the endplate leads to pathological alterations of endplate structure (Salpeter et al. 1979); however, this can be prevented by chelation of Ca²⁺ in the extracellular fluid (Leonard & Salpeter, 1979). This endplate myopathy has been repeatedly described in the slow-channel congenital myasthenic syndrome (SCCMS), a disease that arises from dominant point mutations in AChR subunits that prolong opening events of the AChR channel and decay of the synaptic current (reviewed by Engel et al. 2003a, b). Consistent with the pathological role of Ca²⁺ in the SCCMS, millimolar levels of Ca²⁺ accumulation has been demonstrated at endplates of transgenic mice (Gomez et al. 2002) and SCCMS patients (Engel et al. 1982, 2003b; Vohra et al. 2004).

The cytotoxic effects of excessive Ca²⁺ entry (referred to as excitotoxicity) have been well characterized when glutamate receptors, in particular the highly Ca²⁺-permeable NMDA receptors, are over stimulated by excessive synaptic release of glutamate (Rothman & Olney, 1995; Arundine & Tymianski, 2003). Indeed, for these receptors, the percent current carried by Ca²⁺ ions, termed fractional Ca²⁺ current (P_f), is as large as 12% (Burnashev et al. 1995). In the case of SCCMS, Ca²⁺ enters the endplate through the ɛ-AChR, whose P_f in rodents is about 4% (for review see Fucile, 2004), which is much smaller than the P_f of NMDA receptors. This suggests that the excitotoxic events in the SCCMS would not be adequately explained by the prolonged synaptic response to ACh unless it was also augmented by increased Ca²⁺ permeability of the mutant channel. Consistent with this assumption, in animal models of the SCCMS, endplate myopathy was observed chiefly when the prolonged synaptic currents were associated with an increased Ca²⁺ permeability of the mutant AChR channels (Gomez et al. 2002).

We therefore wished to establish the P_f values of the wild-type human ɛ- and γ-AChRs, which have not been previously documented. We also measured the P_f of two well-characterized human slow-channel AChRs that harbour the αG153S or the αV249F mutation. In humans carrying αG153S or αV249F, single-channel recordings show that γ-AChR is also expressed at the endplate, accounting for 5–12% of all channel openings (Sine et al. 1995; Milone et al. 1997). Hence, we measured the fractional Ca²⁺ current carried by the mutant γ-AChR as well.

For the reliable determination of P_f (Zhou & Neher, 1993), the AChR channels must be the only pathway for Ca²⁺ entry because concomitant Ca²⁺ release from intracellular stores will lead to an overestimate of the Ca²⁺ signal and confound the results. Mature muscle fibres are not suitable for this purpose as full blockade of Ca²⁺ release from the sarcoplasmic reticulum is difficult to achieve. Moreover, the large size of mature muscle fibres is likely to prevent their successful voltage clamping, and activation of voltage-gated Ca²⁺ channels will invalidate the experiment. Small myotubes in culture are more amenable to P_f measurements, but they only express γ-AChR. We therefore performed experiments on transfected GH4C1 cells expressing either γ- or ɛ-AChRs and used cultured human myotubes to validate our expression system for γ-AChR.

We report that wild-type human ɛ-AChR has an unexpectedly high P_f, which is not affected by either the αG153S or the αV249F mutation. Thus, a constitutively high P_f of the mature human receptor predisposes the endplate to excitotoxic damage when the synaptic current is prolonged.

Methods

Cell cultures and transfection

Cultures were maintained in a humidified incubator with 5% CO₂ at 37°C. All media and supplements were purchased from Gibco/Invitrogen, except for epidermal growth factor (Sigma). For experiments, cells were plated onto 35-mm Petri dishes at a density of 10⁴ cells cm⁻². Rat pituitary GH4C1 cells were grown in HAM F10 nutrient mixture plus 10% fetal calf serum (FCS) and 1% penicillin-streptomycin.

Different combinations of the human wild-type α, β, δ, γ or ɛ subunits, or α subunits harbouring the αG153S or αV249F mutation, were expressed in GH4C1 cells by transient transfection using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions, to yield γ- or ɛ-AChR containing either the wild-type α subunit or one of the mutant α subunits. The other cDNAs used (mouse α, β, δ, ɛ or human α7) were transfected by the same procedures. For each subunit, 1 μg cDNA was added to each 35-mm Petri dish, and experiments were carried out 48–72 h after transfection. Human myotubes were obtained from satellite cells derived from human muscle specimens (Broccolini et al. 2004). The muscle specimens were obtained with informed written consent from patients without neuromuscular disease undergoing surgery. Cells were allowed to proliferate on collagen-coated Petri dishes, in Dulbecco's modified Eagle's medium (DMEM) supplemented with 15% FCS, 2% penicillin-streptomycin, 1% insulin-transferrin-selenium and 10 ng ml⁻¹ epidermal growth factor. Differentiation was induced at approximately 50% confluence by switching to a low-serum medium (DMEM plus 2% horse serum and penicillin-streptomycin). Due to the reduced physical contact between cells, many cells differentiated but remained mononucleated, as previously described (Krause et al. 1995).

Solutions and chemicals

Cells were superfused with a standard external medium containing (mm): NaCl 140, KCl 2.8, CaCl₂ 2, MgCl₂ 2, Hepes-NaOH 10 and glucose 10; pH 7.3. CaCl₂ was omitted in the Ca²⁺-free solution. For cell-attached experiments, patch pipettes were filled the same solution plus ACh (100 nm). For whole-cell recordings, patch pipettes were filled with a solution containing (mm): N-methyl-d-glucamine (NMDG) 140, Hepes-HCl 10, Fura-2 pentapotassium salt 0.25 and thapsigargin 0.001; pH 7.3. Calibration measurements of F/Q, which require that current is only due to the flux of Ca²⁺ ions, were performed using the following extracellular solution (mm): NMDG 130, CaCl₂ 10 and Hepes-HCl 10; pH 7.3. All chemicals were purchased from Sigma (USA), except for Fura-2 pentapotassium salt and the acetoxylmethyl ester (AM) form of Fura-2 (Molecular Probes).

Electrophysiology

Whole-cell currents were recorded using borosilicate glass patch pipettes (3–6 MΩ tip resistance) connected to an Axopatch 200A amplifier (Molecular Devices, Union City, CA, USA). Data were recorded and analysed using pCLAMP8 software (Molecular Devices). Whole-cell capacitance and patch series resistance (5–15 MΩ) were estimated from slow-transient compensations, with a series-resistance compensation of 70–90%. Cells were voltage clamped at a holding potential (HP) of −70 mV (unless otherwise indicated) and continuously superfused using a gravity-driven perfusion system with independent tubes for standard and agonist-containing solutions, positioned 50–100 μm from the patched cell. A fast exchanger system (RSC-100, BioLogique, France) allowed complete solution exchange in less than 50 ms. Data are given as means ± s.e.m. Two data sets were considered statistically different when P > 0.05 (ANOVA test).

Single-channel data were recorded using Sylgard-coated borosilicate glass pipettes (tip resistance, 3–4 MΩ), with an Axopatch 200B amplifier. Data were filtered at 4 kHz and sampled at 20 kHz, using pClamp9 software (Molecular Devices). Analysis was performed with a 50% threshold criterion using pClamp9, omitting events shorter than 0.2 ms. Slope conductance was calculated by linear fitting of the unitary amplitudes recorded at different pipette potentials (at least three for each cell). The critical time used to identify a burst (determined for each cell from the closed-time distribution) ranged from 1 to 3 ms, as previously reported (Fucile et al. 2002). The distribution of burst durations was fitted (by pClamp 9 routines) with the sum of two (or three) exponential components, with time constraints τb1, τb2 (and τb3), respectively.

P_f determination

The methods to determine P_f (the proportion of whole-cell current carried by Ca²⁺) have been fully reported previously (Fucile et al. 2000). Fluorescence determinations were made using a conventional fluorescence microscopy system composed of an upright microscope (Axioskop 2, Zeiss, Germany), a digital 12-bit cooled camera (SensiCam, PCO, Germany) and a monochromator (Till, Germany). The system was driven by Axon Imaging Workbench software (Molecular Devices). All optical parameters and the setting of the digital camera (exposure time, 25 ms; 4 × 4 binning) were maintained throughout all measurements to avoid non-homogeneous data. Records of fluorescent signals and membrane currents were synchronized by a digital signal from the imaging computer triggering the start of electophysiological recording. Images were acquired and stored on a PC (Dell, USA), then analysed off-line. The changes in intracellular calcium ([Ca²⁺]_i) were measured at a single excitation wavelength (380 nm), to achieve a higher time resolution, and expressed as the ratio of time-resolved fluorescence variation to the basal fluorescence (ΔF/F).

Petri dishes were incubated at 37°C with the membrane-permeant dye Fura-2 AM (4 μm) for 45 min in DMEM. The AChR-expressing cells were initially identified as those generating a fluorescence transient in response to a brief application of nicotine. These cells were selected for whole-cell recordings, and filled with Fura-2 pentapotassium salt through the patch pipette. As high basal [Ca²⁺]_i alters P_f values (S.F. and F.E., unpublished observations), the ratiometric dye was chosen to carefully assess [Ca²⁺]_i prior to P_f measurements. Cells with high basal [Ca²⁺]_i (fluorescence at 340 nm (F₃₄₀)/fluorescence at 380 nm (F₃₈₀) ratio values, > 2) or low basal F₃₈₀ values (< 200 arbitary units (a.u.)), indicating poor loading, were discarded. Determinations were carried out after obtaining a stable value of basal fluorescence, using the same a.u. setting throughout all the experiments. The ratio between the fluorescence increase (F, in a.u.) and the total charge entering the cell (Q, in nC) was defined as:

where I_Nic was the nicotine-evoked whole-cell current. Each F/Q point was obtained by measuring the charge entering the cell at each fluorescence acquisition time. The F/Q ratio values (in nC⁻¹) used in determining P_f were represented by the slopes of the linear regression best fitting the F versus Q plots. We used only F/Q points measured immediately after the onset of the ACh-induced response that exhibited a linear relationship, indicating that the Ca²⁺-buffering capability of Fura-2 was not saturated. P_f was determined by normalizing the F/Q ratio obtained in standard medium (F/Q_standard) to the ratio obtained in calibration medium (F/Q_calibration) containing Ca²⁺ as the only permeant ion (Ca²⁺ medium), according to the equation:

The F/Q_calibration values obtained in this work were 8.0 ± 0.4 nC⁻¹ (mean ± s.e.m., n = 6 cells) for GH4C1 cells and 3.0 ± 0.4 nC⁻¹ (n = 4) for human myotubes.

Results

Electrophysiological characterization

The pituitary cell line GH4C1 is a reliable mammalian expression system for neuronal AChRs (Fucile, 2004). To confirm that it also represents an adequate expression system for muscle γ- and ɛ-AChR, we first tested whether the kinetic properties of AChR channels expressed in these cells are comparable to those observed in muscle cells. As we aimed to measure the Ca²⁺ permeability of γ- and ɛ-AChRs containing either wild-type or mutant α subunits, we considered several subunit assemblies.

In cells expressing wild-type ɛ-AChR, ACh (100 nm) evoked unitary events with a slope conductance of 50.1 ± 3.5 pS (n = 3) and burst duration of 2.5 ms. Under our experimental conditions, most bursts consisted of single openings (see Fig. 1A left), although we cannot rule out the possibility that brief openings, below our detection threshold, were also present. The distribution of burst durations was adequately fitted by two exponential components in all cells (Fig. 1B left). For the wild-type γ-AChR, single-channel conductance was 30.5 ± 2.1 pS (n = 5) and the burst duration was about 6 ms. Again, most bursts comprised a single opening (Fig. 1A right), and the distribution of burst durations was fitted by two exponential components (Fig. 1B right).

Figure 1. Single-channel currents of wild-type human ɛ- and γ-AChRs.

A, typical examples of single-channel openings (upward) recorded in the cell-attached configuration from GH4C1 cells transiently transfected with wild-type αβɛδ or αβγδ AChR subunit cDNAs, as indicated. ACh concentration in the recording pipette was 100 nm. Channel conductance was 46.4 pS (ɛ-AChR) and 30.5 pS (γ-AChR). Traces filtered at 2 kHz for display purposes. B, distribution of burst durations (calculated with critical τ = 1 ms) from the same recordings, best-fitted by the sum of two exponential components (dotted lines). Time constants (weight) of the fitting distributions for ɛ-AChR were: τ_b1 = 0.55 ms (43%), τ_b2 = 2.7 ms (57%); and for γ-AChR were: τ_b1 = 0.4 ms (18%), τ_b2 = 6.8 ms (82%).

The αV249F or αG153S mutations had no significant effect on conductance of either the ɛ- or γ-AChR channel (n = 3 cells for each mutation), but significantly prolonged the burst duration and altered its distribution so that it was now best-fitted with three exponential components in all the tested cells. In agreement with previous reports (Sine et al. 1995; Milone et al. 1997), the burst duration of ɛ-AChRs containing the αG153S mutation was about 13 ms, and of ɛ-AChRs harbouring the αV249F mutation was between 35 and 80 ms (Fig. 2). For γ-AChRs, the burst duration was about 40 ms in AChRs carrying either the αG153S or the αV249F mutation (Fig. 2).

Figure 2. Single-channel currents of mutant human ɛ- and γ-AChRs.

Typical examples of single-channel openings recorded in the cell-attached configuration from GH4C1 cells transiently transfected with αV249Fβɛδ, αG153Sβɛδ, αV249Fβγδ or αG153Sβγδ AChR subunit cDNAs, as indicated. Under each trace is the corresponding distribution of burst durations, best-fitted by the sum of three exponential components. ACh concentration in the recording pipette was 100 nm. Channel conductance and time constants of the burst fitting components (weight) were: 45.4 pS, τ_b1 = 0.3 ms (18%), τ_b2 = 2.7 ms (19%), τ_b3 = 183 ms (62%) (αV249Fβɛδ; critical τ = 2 ms); 34.4 pS, τ_b1 = 0.6 ms (25%), τ_b2 = 1.8 ms (25%), τ_b3 = 64 ms (50%) (αV249Fβγδ; critical τ = 3 ms); 50.6 pS, τ_b1 = 0.2 ms (29%), τ_b2 = 2.9 ms (23%), τ_b3 = 16.8 ms (47%) (αG153Sβɛδ; critical τ = 1 ms); 26.3 pS, τ_b1 = 0.3 ms (30%), τ_b2 = 6.6 ms (23%), τ_b3 = 41.7 ms (47%) (αG153Sβγδ; critical τ = 1 ms). Traces filtered at 2 kHz for display purposes.

As P_f determinations are best performed using nicotine as an agonist (see below), we examined the properties of I_Nic in GH4C1 cells transiently transfected with either human wild-type or mutant α subunits together with complementary wild-type β, δ and γ or ɛ subunits. No response was observed in non-transfected cells, indicating that GH4C1 native cells do not express functional nicotinic AChRs. Nicotine (100 μm) elicited inward currents with peak amplitudes between 0.3 and 4 nA for all the AChR isoforms examined, indicating equivalent expression levels (Table 1). Upon prolonged transmitter application, I_Nic decayed exponentially as a result of receptor desensitization. The decay was best-fitted to a single-exponential function for wild-type γ- and ɛ-AChR (Table 1). Enhanced desensitization has been reported on the basis of single-channel data for ɛ-AChRs containing either the αV249F or the αG153S mutation (Sine et al. 1995; Milone et al. 1997). Consistent with this enhanced desensitization, the decay of I_Nic was faster in cells expressing ɛ-AChR or γ-AChR that harboured the mutant α subunit (compare Figs 3 and 4 with Fig. 5). These data show that γ- and ɛ-AChRs containing the αV249F or αG153S mutations retain their main functional characteristics in GH4C1 cells.

Table 1.

Functional parameters of nicotine-evoked whole-cell responses

AChR subunits	No. of cells	Peak amplitude (nA)	Decay (s)	Pf (%)
WT αβγδ	8	1.2 ± 0.3	2.0 ± 0.5	2.9 ± 0.2
WT αβɛδ	8	1.2 ± 0.3	2.1 ± 0.8	7.2 ± 1.1
αV249F βɛδ	6	1.3 ± 0.5	0.48 ± 0.06*	7.6 ± 0.6
αG153S βɛδ	7	1.3 ± 0.2	0.5 ± 0.1*	7.1 ± 0.5
αV249F βγδ	7	1.7 ± 0.5	0.35 ± 0.05*	3.3 ± 0.2
αG153S βγδ	5	2.3 ± 0.5	0.8 ± 0.2*	2.9 ± 0.6

Peak amplitude of inward currents elicited by nicotine application (100 μm; duration, 1 s), expressed as mean ± s.e.m., in the indicated number of cells. Current decay was best-fitted by a single exponential with the indicated time constants.

^*Significantly different from the corresponding wild-type value (P < 0.01).

Figure 3. Ca²⁺ permeability of human γ-AChR.

A, left-hand traces show simultaneous recordings of whole-cell current (top) and Ca²⁺ transient (bottom), induced by nicotine (100 μm, horizontal thin line), in a GH4C1 cell transiently transfected with wild-type αβγδ AChR subunit cDNAs. Holding potential, −70 mV. At the excitation wavelength of 380 nm, [Ca²⁺]_i increase corresponds to a downward deflection of fluorescence intensity. Traces are aligned and share the same temporal scale. Middle panel shows typical epifluorescence image of a GH4C1 cell dialysed with Fura-2 through the patch pipette (halo on the right). Contrast was adjusted to increase the visibility of nucleus. Graph on the right shows linear relationships between ΔF/F and Q obtained from the same cell. For this example, the P_f value, calculated by normalizing the slope obtained in A (standard medium) to the mean slope obtained from calibration experiments (dashed line), is 3.2%. B, left-hand traces show simultaneous recordings of whole-cell current (top) and Ca²⁺ transient (bottom), induced by nicotine (100 μm, horizontal thin line), from a human myotube. Holding potential, −70 mV. Traces are aligned and share the same temporal scale. Middle panel shows typical fluorescence image of a binucleated human myotube dialysed with Fura-2. Graph on the right shows linear relationship between ΔF/F and Q obtained from the cell shown in B. For this example, the P_f value, calculated by normalizing the slopes obtained in B (standard medium) with respect to the mean slope obtained from calibration experiments (3.07 nC⁻¹; dashed line), is 2.9%.

Figure 4. Ca²⁺ permeability of human, mouse and chimeric ɛ-AChRs.

A, simultaneous recordings of whole-cell current (top) and Ca²⁺ transient (bottom) induced by nicotine (100 μm, horizontal thin line) in GH4C1 cells transiently transfected with wild-type human αβɛδ (left), mouse αβɛδ (middle) and human αβδ plus mouse ɛ (right) AChR subunit cDNAs. Holding potential, −70 mV. Traces are aligned and share the same temporal scale. B, linear relationships between ΔF/F and Q obtained from the cells shown in A. For these examples, the P_f values, calculated by normalizing the slopes obtained in A (standard medium) with respect to the mean slope obtained from calibration experiments (8.0 ± 0.4 nC⁻¹, dashed line) are 7.5% for human αβɛδ (^), and 3.4% and 3.5% for mouse αβɛδ (▵) and for human αβδ-mouse ɛ (□), respectively.

Figure 5. Ca²⁺ permeability of human mutant AChRs.

A, simultaneous recordings of whole-cell current (top) and Ca²⁺ transient (bottom), induced by nicotine (100 μm, horizontal thin line), in GH4C1 cells transiently transfected with αV249Fβɛδ (left) and αG153Sβɛδ (right) AChR subunit cDNAs. Holding potential, −70 mV. Traces are aligned and share the same temporal scale. B, simultaneous recordings of whole-cell current (top) and Ca²⁺ transient (bottom), induced by nicotine, in GH4C1 cells transiently transfected with αV249Fβγδ (left) and αG153Sβγδ (right) AChR subunit cDNAs. Holding potential, −70 mV. Traces are aligned and share the same temporal scale. C, linear relationships between ΔF/F and Q obtained from the cells shown in A and B. For these examples, the P_f values, calculated by normalizing the slopes obtained in A and B (standard medium) with respect to the mean slope obtained from calibration experiments (8.0 ± 0.4 nC⁻¹, dashed line), are 7.7% (αV249Fβɛδ, ▵), 7.6% (αG153Sβɛδ, □), 3.7% (αV249Fβγδ, ▴), and 3.4% (αG153Sβγδ, ▪).

Control measurements of Ca²⁺ mobilization

As the fractional Ca²⁺ current of AChRs describes the percentage of the ACh-evoked current carried by Ca²⁺ ions, it is crucial that Ca²⁺ ions enter the cytoplasm only through the AChR channel and not from any other source. Accordingly, in P_f measurements, we used nicotine instead of ACh to prevent the activation of putative native muscarinic AChRs that might mobilize Ca²⁺ from intracellular stores. Moreover, we depleted intracellular stores by intracellular dialysis (via the patch pipette) of the Ca²⁺-ATPAse inhibitor thapsigargin, and prevented their refill by omitting ATP from the intracellular solution. The small size of GH4C1 cells ensured adequate control of the intracellular milieu (Fig. 3A).

With these precautions, nicotine failed to evoke Ca²⁺ transients from transfected cells superfused with a Ca²⁺-free external solution or from non-transfected cells exposed to the standard medium (data not shown). Thus, the experimental conditions excluded release of residual Ca²⁺ from internal stores. In transfected cells in the presence of extracellular Ca²⁺, nicotine application evoked a Ca²⁺ transient (Figs 3–5). The delay between current onset and the beginning of fluorescence change was below the sampling interval of our optical system, ruling out a long-delayed metabotropic cause of Ca²⁺ influx. Our experimental conditions prevented Ca²⁺ extrusion out of the cytoplasm, into intracellular organelles or to the extracellular medium, so that no recovery of [Ca²⁺]_i to the basal levels took place (see Figs 3–5). Considered together, these findings indicate that the enhanced [Ca²⁺]_i was due to Ca²⁺ entry from the extracellular medium through bona fide exogenous AChRs.

P_f determinations for wild-type γ-AChRs

For P_f determinations, nicotine-evoked Ca²⁺ and current responses were simultaneously recorded, and the F/Q ratios calculated. In cells expressing the wild-type γ-AChR, the P_f was around 3% (Fig. 3A and Table 1), a value comparable to that found in rodent γ-AChR (Ragozzino et al. 1998). To validate our heterologous cell-expression system, we measured the P_f of human γ-AChR in cultured primary myotubes that express γ-AChR soon after differentiation is induced (Krause et al. 1995). To ensure a valid space clamp and a complete intracellular dialysis through the patch pipette, only small myotubes with one to three nuclei were selected (Fig. 3B). These cells had a P_f value of 2.7 ± 0.6% (n = 6; Fig. 3B), similar to that found in transfected cells expressing the wild-type γ-AChR (Table 1).

P_f determinations for wild-type ɛ-AChRs

In cells expressing human wild-type ɛ-AChR (see Fig. 4A), normalization to calibration values yielded P_f values of ∼7% (Table 1), a value much higher than in rodents (Fucile, 2004). To rule out experimental artifacts, we repeated the P_f measurement for mouse and human wild-type ɛ-AChR in parallel. Consistent with the initial data, we found a P_f of 4.1 ± 0.4% (n = 5) in cells expressing mouse ɛ-AChR (Fig. 4B), indicating that the high fractional Ca²⁺ current of human ɛ-AChR is genuine. To further confirm that our experimental conditions were similar to those of our previous work, we repeated the determination of the P_f for the homomeric human α7-AChR in GH4C1 cells and found a P_f value of 10.9 ± 1.9% (n = 3, data not shown), which is similar to that previously measured in GH4C1 cells (11.4%; Fucile et al. 2003), or in rat native neurons (9.5%; Fucile et al. 2005).

The human ɛ subunit appears to be a crucial determinant of the P_f value, as no difference was observed between mouse and human γ-AChRs. To test this point directly, we produced a chimeric receptor by transfecting the human α, β and δ subunits along with the mouse ɛ subunit in GH4C1 cells. In these cells, nicotine elicited inward currents with a mean amplitude of 640 ± 180 pA (n = 7). The measured P_f value was 4.1 ± 0.3% (n = 5, Fig. 4B), confirming that the major determinant of the increased Ca²⁺ permeability of human ɛ-AChRs is indeed the ɛ subunit.

P_f determinations for mutant ɛ- and γ-AChRs

To evaluate whether mutations leading to SCCMS could increase the P_f of muscle AChRs, the measurements of this parameter were repeated for ɛ- and γ-AChRs comprising the mutant αV249F or αG153S subunits. However, neither the αV249F nor the αG153S mutations altered the P_f values of ɛ-AChR or γ-AChR as compared to the corresponding wild-type receptor (Fig. 5 and Table 1). Therefore, these slow-channel mutations had no significant effect on the Ca²⁺ permeability of the AChR channels.

Discussion

We report in this work that the P_f of the wild-type human ɛ-AChR is above 7%, which is higher than that of neuronal AChRs, with the notable exceptions of α7-AChR and the α9-containing AChR (Fucile, 2004; Fucile et al. 2006). The Ca²⁺ permeability of the human ɛ-AChR is significantly higher that than of mouse ɛ-AChR, whereas the Ca²⁺ permeability of the human γ-AChR is slightly, but not significantly, higher than that of mouse γ-AChR (see Fucile, 2004). Concurrent control studies on mouse ɛ-AChR and α7-AChR yielded P_f values consistent with previously published results (Ragozzino et al. 1998; Fucile et al. 2003), confirming the validity of the determinations. Interestingly, these findings contradict the general view that the Ca²⁺ permeability of neuronal AChRs is much larger than that of the muscle-type AChRs (see for instance Vernino et al. 1992).

As our measurements were performed in transfected GH4C1 cells, it is questionable whether ɛ-AChR has the same high Ca²⁺ permeability in native muscle fibres. It was previously shown that P_f values of the rat α7-AChR, which is highly permeable to Ca²⁺, were similar in GH4C1 cells and in native preparations (Fucile et al. 2003, 2005). Here we show that the P_f for human γ-AChR is similar in cultured human muscle cells and in GH4C1 cells. Moreover, single-channel properties of the ɛ-AChR expressed in GH4C1 cells are similar to those in human muscle fibres (Sine et al. 1995; Milone et al. 1997), and the single-channel properties of wild-type γ-AChR expressed in GH4C1 cells are similar to those in human myotubes (F.G. and A. Di Castro, unpublished observations). Finally, the kinetic properties of the ɛ-AChR mutants harbouring the αG153S or the αV249F are similar in GH4C1 cells and muscle fibres (Sine et al. 1995; Milone et al. 1997). These data confirm that GH4C1 cells represent a faithful expression system for nicotinic receptors.

Ca²⁺ influx at the endplate is held responsible for the endplate myopathy of the SCCMS. This is a dominantly inherited disorder of neuromuscular transmission in which point mutations in genes encoding subunits of the AChR produce abnormally long-lasting synaptic currents, resulting in excessive Ca²⁺ entry and in excitotoxic damage of the postsynaptic region (Engel & Sine, 2005). Staining with dyes that are sensitive to Ca²⁺ in the millimolar range (Bodensteiner & Engel, 1978) has demonstrated Ca²⁺ accumulation at human slow-channel endplates (Engel et al. 1982, 2003b; Vohra et al. 2004) and in mouse models of the disease (Gomez et al. 2002). Caspase activation, triggered by enhanced [Ca²⁺]_i has also been detected at the endplate of SCCMS patients (Vohra et al. 2004), providing a possible link between Ca²⁺ influx and focal damage at the endplate. Our data show that the P_f values observed for γ- or ɛ-AChRs harbouring the mutant αV249F or αG153S subunits were comparable to the corresponding control values, indicating that these mutations only affect channel opening kinetics, with endplate currents decaying up to 10 times more slowly in patients with SCCMS than in controls (Sine et al. 1995; Milone et al. 1997).

Although the details of the pathological process remain to be elucidated, the high P_f of human ɛ-AChR is a likely predisposing factor for the endplate myopathy, because in mice that have a lower P_f than humans the histological damage is only detected when the endplate currents are much more prolonged than in most SCCMS patients (Gomez et al. 2002). Further evidence that long openings of AChR channels must be coupled to a high Ca²⁺ influx to be toxic comes from the observation that developing mouse muscle fibres express γ-AChR at the endplate that has 3- to 5-fold longer openings than the ɛ-AChR (Grassi et al. 1998). The long-lasting γ-AChR openings are present for about 2 weeks postnatally (Villarroel & Sakmann, 1996; Grassi & Degasperi, 2000), although at a low and rapidly decreasing density. In this context, far from being toxic, γ-AChR channels are required for a correct synapse formation, a role in which they cannot be replaced by ɛ-AChR (Koenen et al. 2005). Thus, appearance of the endplate myopathy is dependent on strongly prolonged openings of the highly Ca²⁺-permeable mutant ɛ-AChR channels.

It is striking that human and mouse γ-AChRs are similar in their Ca²⁺ permeability, whereas that of the ɛ-AChR is markedly different between the two species. Here we unequivocally show that the relatively large P_f of the human ɛ-AChR is uniquely determined by the human ɛ subunit, as AChR containing human αβδ plus mouse ɛ subunits has a P_f identical to that of the mouse ɛ-AChR. The molecular determinants underlying the high Ca²⁺ permeability of the human wild-type ɛ-AChR await further investigation.