Functional properties and pharmacological inhibition of ASIC channels in the human SJ-RH30 skeletal muscle cell line

Abstract

The rhabdomyosarcoma cell line (SJ-RH30) exhibits both ultrastructural and electrophysiological hallmarks of mammalian skeletal muscle. We have used patch-clamp electrophysiology to study acid-gated currents in these cells. At a holding potential of −60 mV, rapid application of extracellular solutions of pH 6.5 produced inward current responses in ∼85% of cells. The amplitude of these responses exhibited a marked pH dependence. In addition, the kinetics of both activation and desensitization were faster at more acidic pH, whereas the deactivation rate was pH independent. Repeated applications of a pH 6.0 solution produced a tachyphalaxis that could be substantially attenuated by reducing the duration of the acid challenge and increasing the interstimulus interval. The current–voltage relationship of the acid-induced currents was linear at positive potentials but an area of negative slope conductance was observed in the negative potential range. This was not eliminated by removal of extracellular Ca²⁺, a manoeuvre which did, however, substantially increase current amplitude. Changing the transmembrane Na⁺ gradient altered the current–voltage relationship in a fashion commensurate with an underlying conductance predominantly permeable to Na⁺. Pharmacologically, acid-induced currents were blocked 84.4 ± 1.2% by 30 μm amiloride and 91.8 ± 3.0% by a 1 : 1000 dilution of Psalmopoeus cambridgei venom. Inhibition by both agents could be reversed by a short period of compound washout. By contrast, APETx2 had no significant effect on acid-evoked currents. These observations strongly suggest the acid-induced current is mediated by ASIC1 channels. In agreement with this, current responses of SJ-RH30 cells to a pH 6.0 challenge were greatly enhanced by extracellular lactate. These data demonstrate the presence of ASIC1 channels in a cell line with skeletal muscle characteristics. In addition, significant levels of ASIC1 and ASIC3 mRNA were found in skeletal muscle tissue samples. These findings are discussed with regard to the role such a conductance might play if present in skeletal muscle in vivo.

Due to the presence of charged amino acid side chains in their structure, most ion channels are affected in some way by changes in pH. There are, however, three classes of mammalian ion channel whose acid-sensitive phenotype has attracted particular attention in terms of their potential physiological function. These are (1) the acid-sensitive ion channels (ASICs), a class of amiloride sensitive Na⁺ channels activated by extracellular acidification (Waldmann et al. 1999; Kellenberger & Schild, 2002), (2) the caspsaicin or vanilloid receptor (TRPV1), a non-selective cation channel that can be both activated and modulated by acid as well as heat and capsaicin (Jordt et al. 2000; Gunthorpe et al. 2002), and (3) the TASK channels, a subgroup of the two-pore-domain background K⁺ channels that are inhibited by extracellular acidification (Mathie et al. 2003; Lesage, 2003). Notably, acid affects all of these channels to produce a depolarizing influence on the cells in which they reside.

Our work described here is specifically concerned with the ASIC channels. There are four genes for ASIC subunits in the human genome. These encode subunits with two predicted transmembrane domains which are thought to form tetrameric assemblies that can be either homomeric or heteromeric in nature (Waldmann, 2001; Kellenberger & Schild, 2002; Krishtal, 2003). Electrophysiological characterization of homomeric assemblies of the four ASIC channels has demonstrated that they each exhibit a characteristic biophysical behaviour with respect to gating properties and pH dependence. Pharmacologically, all homomeric and heteromeric ASIC channels examined are amiloride sensitive, although there is some subtype differentiation in terms of amiloride potency. Two toxins have been described as ASIC blockers: the first is psalmotoxin 1 (Escoubas et al. 2000), which selectively targets ASIC1a channels, and the second is an ASIC3-selective sea anemone-derived toxin APETx2 (Diochot et al. 2004). Numerous functional roles have been suggested and examined for ASIC channels, many of which have been recently reviewed by Krishtal (2003).

Skeletal muscle can experience quite profound use-dependent changes in both its intracellular and extracellular pH. During intensive exercise anaerobic metabolism and lactate production are engaged in muscle. As a consequence the pH of muscle in vivo can fall as low as 6.5 (Juel, 1998). The most apparent consequence of this pH change to a vigorously exercising individual is pain that emanates from active muscle. This is believed to reflect the activation of acid-sensitive channels on sensory neurones that innervate the muscle mass (Sluka et al. 2003). A related sensory mechanism may also play a role in the generation of the muscle pressor reflex (Li et al. 2004). Over and above this, changes in muscle lactate and pH have long been suggested to directly alter muscle function and contribute, along with a wide host of other factors, to muscle fatigue (see Fitts, 1994). Despite this long-standing view of the role of muscle acidification in fatigue, little, if anything, is known about acid-sensitive ion channels present in skeletal muscle. One reason for this lack of knowledge may arise from some of the experimental difficulties presented to the electrophysiologist by the physical nature of skeletal muscle.

In the adult, the contractile components of skeletal muscle consist of fibres made up of a highly ordered multicellular syncytium, typically containing a thousand or more diploid nuclei and sometimes reaching several centimetres in length. These fibres are interspersed with undifferentiated myoblast satellite cells which only initiate a programme of differentiation when called upon for muscle regeneration of expansion. The large size, surface area and conductance of most muscle fibres makes whole-cell voltage clamp analysis with the patch clamp method impractical, although two-electrode voltage clamp and Vaseline gap methods can be used. Faithful voltage clamp analysis of rapidly desensitizing ligand-gated channel responses is also complicated by the size of muscle fibres and the limit this places on solution exchange times.

To allow investigation of some potential characteristics of human muscle cell physiology we have recently been studying the SJ-RH30 cell line as a skeletal muscle surrogate. This cell line was established from the bone marrow of a child with rhabdomyosarcoma (a tumour of skeletal muscle progenitor cells (Merlino & Helman, 1999)) and exhibits certain ultrastructural elements of skeletal muscle differentiation. Furthermore, it functionally expresses both the skeletal muscle specific voltage-gated Na⁺ channel (NaV1.4) and the muscle type nicotinic receptor (α1) (D. P. Gitterman & A. D. Randall, unpublished observations). In this paper we describe the properties of the ASIC currents in this cell line and discuss how these may contribute to muscle cell physiology if their expression in the SJ-RH30 cells reflects the situation in native human skeletal muscle.

Methods

Cell culture

SJ-RH30 cells were obtained from the American Type Culture Collection and maintained at 37°C and 5% CO₂ in 50 ml flasks in a culture medium composed of RPMI-1640 medium (Invitrogen) supplemented with 10% fetal calf serum (Invitrogen) and 1 mm glutamine. They were passaged every 7 days, when 1 × 10⁶ cells were transferred to 50 ml of new culture medium. To prepare cells for experiments they were plated onto poly-l-lysine coated glass coverslips at a density of 5000 cells cm⁻² and maintained under identical conditions in the same culture medium. Electrophysiological recordings were made 24–96 h after plating.

Whole cell patch clamp electrophysiology

During whole cell patch clamp experiments, the recording chamber was perfused at a rate of 2 ml min⁻¹ with pseudo-physiological saline of the following composition (mm): 140 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 10 d-glucose, 10 HEPES at pH 7.3. As required batches of this solution were also prepared at different pH values, with pH adjustments being made with HCl. The pipette solution consisted of (mm): 140 CsCl, 4 MgCl₂, 10 EGTA, 10 HEPES, adjusted to pH 7.3 with CsOH. Electrodes filled with this solution had resistances of 3–6 MΩ; liquid junction potentials of 4.5 mV were not corrected for. Variants of this electrode solution in which either all or a defined portion of the CsCl was replaced with NaCl were prepared for current–voltage analysis.

In addition to the constant perfusion of the recording chamber, the recorded cell and its nearby environs were superfused using a commercially available rapid solution exchange system (SF-77B Perfusion Fast-Step, Warner Instruments, Hamden, CT, USA). This system enables changes to the solution bathing the cell to be made on an approximately 30 ms time scale. Solution exchanges made in this fashion were under control of the data acquisition software (pCLAMP 9, Axon Instruments, Union City, CA, USA), thus allowing precise control over both the timing and duration of solution changes. In all recordings, following establishment of the whole-cell configuration, the cell was lifted well clear of the underlying substrate. This manoeuvre was essential to ensure rapid solution exchange to the entire cell membrane.

All recordings were made at room temperature (20–24°C) using an Axopatch 200B amplifier (Axon Instruments). Recordings were made from a total of 68 cells; mean membrane capacitance was 15.7 ± 0.6 pF and mean series resistance was 5.6 ± 0.3 MΩ. Data was analysed off-line using Clampfit 9 (Axon Instruments) and Origin 7.5 (OriginLab Corp., Northampton, MA, USA) analysis software. All data are presented as means ±s.e.m. unless otherwise stated. Statistically significant differences (P < 0.05 using a two-sample Student's t-test) are indicated by an asterisk. Amiloride and lactic acid were obtained from Sigma; Psalmopoeus cambridgei venom was obtained from Spider Pharm Inc. (Yarnell, AZ, USA) and APETx2 was custom synthesized.

TaqMan mRNA profiles

Poly A⁺ RNA from 20 tissues of four different individuals (two males, two females except prostate) was prepared, reverse transcribed and analysed by TaqMan quantitative PCR as previously described (Chapman et al. 2000). Briefly, 0.5–1 μg of poly A⁺ RNA was reverse transcribed using random priming and the cDNA produced was used to make up to 1000 replicate plates with each well containing the cDNA from 1 ng poly A⁺ RNA. TaqMan quantitative PCR (Applied Biosystems, Warrington, UK) was used to assess the level of each gene relative to genomic DNA standards. The data are presented as the means of mRNA copies detected per nanogram poly A⁺ RNA from four individuals ± s.e.m.(n = 4).

Gene-specific reagents for TaqMan analysis were as follows. ASIC 3: forward primer 5′-AGGACTCTCTCCGCCTCCCA-3′, reverse primer 5′-CATGTCCAGGATGTCAGGGC-3′, TaqMan probe 5′-CATGTCCAGGATGTCAGGGC-3′; ASIC 2: forward primer 5′-CATGTCCAGGATGTCAGGGC-3′, reverse primer 5′-CATGTCCAGGATGTCAGGGC-3′, TaqMan probe 5′-CATGTCCAGGATGTCAGGGC-3′; ASIC 1: forward primer 5′-CATGTCCAGGATGTCAGGGC-3′, reverse primer 5′-CATGTCCAGGATGTCAGGGC-3′, TaqMan probe 5′-CATGTCCAGGATGTCAGGGC-3′; Housekeeper GAPDH: forward primer 5′-CATGTCCAGGATGTCAGGGC-3′, reverse primer 5′-CATGTCCAGGATGTCAGGGC-3′, TaqMan probe 5′-CATGTCCAGGATGTCAGGGC-3′.

Results

pH dependence of response amplitude and kinetics

In ∼85% of cells tested, exposure to acidic solutions of pH 6.5 evoked robust inward currents at a holding potential of −60 mV. Both the kinetics and the amplitude of the observed current were dependent on the level of extracellular acidification employed. Changing the extracellular solution from its control pH of 7.3 to pH 6.5 for 1 s produced slowly activating currents which, having achieved their peak amplitude, exhibited little or no macroscopic desensitization. More acidic solutions applied for the same time produced more rapidly activating currents, which also desensitized in the continued presence of acid (Fig. 1A). A range of extracellular pH values from 7.0 to 4.5 was tested in eight cells. The resulting pH response relationship shows a steep pH dependence between 7.0 and 6.0 (Fig. 1B), followed by a decrease in the peak amplitude at pH values below 5.5. Currents deactivated rapidly following all levels of acid challenge.

Figure 1. Acid solution evokes pH-dependent inward currents in SJ-RH30 cells.

A, sample traces showing whole cell currents recorded in response to 1 s applications of a range of acid solutions at pH 7.0–4.5. B, the pH response curve shows a steep relationship between pH and activation of the channel. Overlaying the currents also clearly highlights the differences in kinetics to different pH challenges. C, left panel, 10–90% rise time (inset is the same graph with an expanded y-axis). D, the decay constant of a single exponential fitted to the desensitizing phase of the current response. Both graphs demonstrate the more rapid activation and inactivation of responses as acidity increases

The kinetics and pH dependence of the current responses were further analysed by calculating both the 10–90% rise time of the current activation phase (Fig. 1C) and the decay time constant of the desensitizing phase (Fig. 1D). With respect to rise time, responses to pH 6.5 activated roughly 7 times more slowly (576.0 ± 33.5 ms, n = 8) than those at more acid levels (pH 6.0 and below). There was a further trend toward more rapid activation with decreasing pH, but no differences were significant between solutions ranging from pH 6.0 to pH 4.5 (88.7 ± 6.3 ms to 51.3 ± 5.6 ms, n = 7–8, Fig. 1C inset). In order to determine the decay constant of the response, a single exponential was fitted to the desensitizing phase of the current. This was not possible for currents evoked by pH 7.0 and 6.5 as they showed little or no time-dependent decay (see Fig. 1A). The relationship between desensitization time constant and extracellular pH for pH 6.0–4.5 is plotted in Fig. 1D and illustrates the speeding of response desensitization at more acidic extracellular pH values.

Tachyphalaxis of acid responses

In an initial series of experiments, SJ-RH30 cells voltage-clamped at −60 mV were challenged at 30 s intervals with 1 s applications of acidic (pH 6) extracellular solution. Using this protocol, responses exhibited a time-dependent decline in amplitude, decreasing to 57.8 ± 3.2%(n = 6) of the original amplitude by the 10th acid challenge and to 40.0 ± 4.4%(n = 6) by the 20th acid challenge (Fig. 2A). In an effort to minimize run-down, the duration of exposure to pH 6 was reduced to 300 ms and the interval between applications was increased to 2 min. In this case responses only decreased to 87.7 ± 3.2%(n = 5) of control by the 5th and to 80.5 ± 5.9%(n = 5) of control by the 10th challenge (Fig. 2B). The observed decline in current amplitude therefore appears to reflect the extent of agonist exposure, and thus seemingly represents a desensitization–resensitization cycle, rather than a loss of channel function as a consequence of prolonged whole-cell recording (i.e. run-down).

Figure 2. The decrease in current amplitude in response to repeated acid challenges is most likely due to desensitization rather than whole cell run-down.

A, responses to repeated 1 s (indicated by the bar) applications of pH 6.0 at 30 s intervals. The numbers above the bars indicate that responses shown are to the 1st, 10th and 20th application of acid. There is a clear decrease in current amplitudes which is shown in the summary graph in the right panel: responses have decreased to roughly 40% of control after 20 challenges over a 10 min period. B, by contrast, when the duration of pH 6.0 applications is reduced to 300 ms and the interval between them is increased to 2 min, currents are significantly more sustained. Typical responses to the 1st, 5th and 10th challenge are shown in the left panel. As seen in the graph on the right, more than 80% of the response remains after 10 applications.

Current–voltage relationship of acid-gated responses

The decline in the acid-induced current in response to repeated agonist responses described above means that repeated applications do not produce reproducible current amplitudes. For this reason, current–voltage relationships for acid-gated currents could not be accurately determined with repeated acid challenges made at different holding potentials. Instead a voltage ramp was used during a relatively sustained portion of the current. Responses to pH 6.5 were chosen for this analysis as they exhibit the least macroscopic desensitization. A 200 ms voltage ramp from −100 mV to +100 mV was applied during the steady state of the response to an application of pH 6.5. A control ramp was then run at pH 7.3 which was subtracted from the former to yield the net acid-gated current (Fig. 3A). As can be seen in the current–voltage relationships pooled from seven experiments (Fig. 3B), the I–V relationship was linear at positive potentials but exhibited a decline in slope conductance at negative membrane potentials. The shape of the I–V relationship was not due to a voltage-dependent block by Ca²⁺ ions at negative potentials, since although Ca²⁺ removal substantially enhanced current amplitude at −60 mV (3.54 ± 0.28 times control value, n = 7), the shape of the I–V relationship remained very similar (Fig. 3C).

Figure 3. The current–voltage relationship is unaffected by removal of extracellular calcium.

A, stimulus waveform and sample traces recorded at pH 7.3 and pH 6.5 in the same cell. The voltage protocol is as follows: hold at −60 mV, drop to −100 mV for 10 ms, ramp to +100 mV over 200 ms and return to −60 mV. The current versus voltage plots are very similar either in the presence (B) or absence (C) of extracellular calcium.

In Fig. 3B and C the acid-activated current was inward-going at all potentials between −100 mV and +100 mV. Notably, our standard electrode solution which was used for these recordings contains no Na⁺ ions. When we replaced intracellular CsCl with NaCl, the current–voltage relationship reversed close to zero as would be expected for a pore predominantly permeable to Na⁺, while an intermediate level of intracellular Na⁺ (14 mm) produced a reversal potential of ∼+40 mV (data not shown).

Antagonism by amiloride and Psalmopoeus venom

ASICs and other structurally related Na⁺-selective channels are sensitive to the diuretic drug amiloride. We therefore examined the ability of this compound to block our acid-mediated responses in SJ-RH30 cells. Following an initial determination of the response to pH 6.0, a 2 min preincubation with 30 μm amiloride produced an 84.4 ± 1.2%(n = 7) inhibition of subsequent responses to pH 6.0 (Fig. 4A). Furthermore, the acid response recovered almost completely (93.3 ± 3.4% of control) after a 5 min wash-out of the drug. Although a well characterized ASIC antagonist, amiloride possess little subtype selectivity and also has actions at other sodium-selective channels. In contrast, psalmotoxin 1 (PcTx 1), a toxin purified from the venom of the tarantula species Psalmopoeus cambridgei, has been shown to be selective for homomeric ASIC1a over all other ASIC channels (Escoubas et al. 2000). Although we were unable to obtain the purified toxin, we tested the effect of a highly diluted solution of raw spider venom. At a 1 in 1000 dilution this venom inhibited current responses to pH 5.5 by 91.8 ± 3.0%(n = 7). After a 5 min wash-out, 80.9 ± 1.4% of the response had recovered (Fig. 4B). SJ-RH30 cells are also known to express ASIC3 mRNA. In order to determine whether the current component resistant to Psalmopoeus venom is mediated by ASIC3, we tested the effect of APETx2. This toxin has been shown to be selective for ASIC3 over other ASIC subunits (Diochot et al. 2004). At a concentration of 1 μm, which we have shown to inhibit recombinant human ASIC3 (authors' unpublished observations), APETx2 had little effect on responses to pH 6.0 producing a mean reduction of 5.0 ± 3.4% (n = 3, Fig. 4C).

Figure 4. Amiloride and Psalmopoeus cambridgei venom potently block ASIC currents in SJ-RH30 cells, while APETx2 is without effect.

In each case, a representative experiment is shown in the left panel. Summary graphs on the right confirm the significant reduction of current amplitudes by amiloride and Psalmopoeus venom. A, 30 mm amiloride inhibits ASIC responses by 84.4 ± 1.2%; they recover to 93.3 ± 3.4% of control upon wash-out. B, in response to Psalmopeus venom (1 : 1000), currents are reduced by 90.8 ± 3.0% and return to 80.9 ± 1.4% after wash-out. By contrast, APETx2 had no significant effect on acid-evoked responses. C, at a concentration of 1 μm, the toxin reduced currents to 95.0 ± 3.8% of control (91.4 ± 5.4% of control on wash-out response).

Potentiation of ASIC currents by lactate

A recent study of the acid-sensitive ion channels of dorsal root ganglion (DRG) neurones reported a significant enhancement of responses to acid in the presence of 15 mm lactic acid. Given the central role of lactate in muscle physiology, we attempted to reproduce this finding on the ASIC channels present in the SJ-RH30 cell line using two different protocols. In the first, cells were switched from control solution (pH 7.3) to pH 6.5, then to pH 6.5 with 15 mm lactate, back to pH 6.5 and finally to control solution (Fig. 5A). Transferring the cells from pH 6.5 to pH 6.5 + lactate produced a sharp increase in the current, following which the responses exhibited a degree of macroscopic desensitization not apparent in the current prior to lactate addition. Upon returning the cell to pH 6.5 in the absence of lactate, the current once again increased and slowly desensitized until the cell was transferred back to pH 7.3. The initial lactate-induced increase in current was 340 ± 28%(n = 5). In the second protocol, cells were simply transferred from pH 7.3 either to pH 6.5 or to pH 6.5 + 15 mm lactate (Fig. 5B). Using this approach, current potentiation was not as dramatic as above, but still significant (69 ± 17%, n = 6). This protocol, however, clearly demonstrated the more than 6-fold faster activation kinetics of currents recorded in the presence of lactate (10–90% rise time in pH 6.5 = 621.4 ± 76.6 ms, in pH 6.5 + lactate = 94.7 ± 6.7 ms, all n = 6).

Figure 5. Lactate significantly potentiates ASIC currents under physiologically acid conditions; this effect was shown using two different protocols.

A, protocol 1: as shown in the sample trace, cells were exposed to pH 6.5 for 300 ms, followed by 15 mm lactate at pH 6.5 for 300 ms, followed by a return to pH 6.5 for a final 300 ms. The pooled data are shown in the summary graph. Current amplitudes increase to 440.5 ± 28.0% of control upon application of lactate, decrease to 317.2 ± 31.9% during the continued presence of lactate and rise again to 468.2 ± 26.5% when the cell is returned to pH 6.5 control solution. B, protocol 2: here cells are transferred from pH 7.3 to a pH 6.5 control solution. After several minutes' recovery, they are then transferred from pH 7.3 to pH 6.5 containing 15 mm lactate. Using this approach amplitudes are potentiated to 169.0 ± 17.3% of control.

Levels of ASIC subunit expression

Our findings may have far-reaching implications for the role of ASIC channels in the regulation of skeletal muscle function. In order to assess the potential relevance of our findings to the situation in vivo, we determined the level of mRNA expression of all three ASIC subunits in skeletal muscle as well as a range of other tissues. The data are summarized in Fig. 6. As previously reported by others, expression of all three ASIC subtypes was far greater in brain than any other tissue. Consequently, we expressed the level of expression in other tissues as a percentage of that found in brain. Expression of ASIC2 was fairly limited in the periphery, with low level expression in a limited number of tissues. There was no detectable ASIC2 in skeletal muscle. Expression of ASIC1 in skeletal muscle is 2.1 ± 0.4% of that in brain, a level comparable to several other tissues. Expression of ASIC3 in skeletal muscle was quite substantial, amounting to 16 ± 2% of brain levels, one of the highest levels observed in peripheral tissues.

Figure 6. TaqMan mRNA profile shows significant ASIC expression in skeletal muscle.

Data are normalized to values obtained in brain, which shows the highest level of expression for all three ASIC subunits. The black bar represents skeletal muscle in each graph. The letters on the x-axis represent the following tissues: A, brain, pool of 18 major regions; B, pituitary; C, heart; D, lung; E, liver; F, fetal liver; G, kidney; H, skeletal muscle; I, stomach; J, small and large intestine; K, spleen; L, blood leukocytes; M, macrophage; N, adipose; O, pancreas; P, prostate; Q, placenta; R, cartilage; S, whole bone; T, bone marrow.

Discussion

We have shown that the SJ-RH30 human rhabdomyosarcoma cell line, which expresses a number of properties reminiscent of skeletal muscle, possesses endogenous acid-gated currents. Literature comparisons of the biophysical and pharmacological properties of these currents indicate that they probably arise from channels containing ASIC1a subunits, possibly in a homomeric assembly.

The pH response relationship for the acid-gated current in SJ-RH30 cells is similar to those previously reported for ASIC1 currents. Thus, very little if any current was observed at pH 7.0 rising to maximal current between pH 6.0 and 5.5 (Gunthorpe et al. 2001; Chu et al. 2002). Furthermore, the various kinetic parameters of the currents observed at different pH values are not dissimilar to those reported previously for native and recombinant ASIC1 channels. (Gunthorpe et al. 2001; Chu et al. 2001; Hesselager et al. 2004). Considering the pH dependence and kinetics of the ASIC responses, it is important to note that although a pH of 6.0 or less is required to produce the peak current response, one would predict that the charge transfer produced for more maintained pH changes (i.e. several seconds) would be greatest for lesser acidifications (e.g. pH 6.5). This is the case because much slower desensitization is observed at these less extreme levels of pH. In the physiological realm, therefore, weak acidifications of < 1 pH unit from rest, which are commonly produced in muscle, may elicit more significant changes to membrane potential, conductance and transmembrane ionic equilibria.

Numerous previous studies of ASIC channels have described antagonism by the diuretic amiloride, which also blocks the related ENaC and degenerin Na⁺ channels. Although we have not constructed a full concentration–inhibition curve, the substantial degree of block we observed at 30 μm amiloride is in line with that obtained previously for ASIC1 (Gunthorpe et al. 2001; Askwith et al. 2004; Gao et al. 2004). More convincing evidence for a central role of ASIC1 in the acid-gated currents of SJ-RH30 cells comes from our experiments with Psalmopoeus venom, the source of the toxin psalmotoxin 1, which is reported to be a selective antagonist of ASIC1a (Escoubas et al. 2000). The profound reversible inhibition of the acid-gated current we observed with a 1 : 1000 dilution of this venom indicates a central role for ASIC1a-containing channels. Transcpritome analysis has shown significant levels of ASIC3 mRNA in this cell line. It is therefore possible that the portion of current resistant to blockade by Psalmopoeus venom may be mediated by an ASIC3 channel. To test this hypothesis we investigated the effect of the ASIC3-selective toxin APETx2 (Diochot et al. 2004). A lack of significant inhibition, however, demonstrates that ASIC3 subunits do not contribute to observed acid-evoked currents.

As reported for ASIC channels in sensory neurones, extracellular lactate elicited a substantial enhancement of the acid response of SJ-RH30 cells. The nature of this enhancement appeared to represent a shift in the pH dependence of gating to a more alkaline pH, since it speeded both response activation and desensitization. This is in agreement with the hypothesis for lactate action proposed by Immke & McCleskey (2001) who suggest that lactate chelates extracellular Ca²⁺ and thus reduces a Ca²⁺-mediated antagonism of a site underpinning acid-mediated channel activation. In agreement with this hypothesis, we found that removal of extracellular Ca²⁺ enhanced responses to pH 6.5 by some 250%. This action of lactate on a membrane conductance of a muscle-related cell may be of particular interest given the production of lactate by active skeletal muscle and the role it may play in fatigue (see below).

The slight rebound in the response after returning to lactate-free solution can be explained through faster repriming of the channel. It has been shown that recovery from desensitization can be accelerated by raising the calcium concentration (Immke & McCleskey, 2003). Thus, transferring the cell back into lactate-free solution results in a raising of the extracellular level of free Ca²⁺ and a consequent change in the state distribution of the channel favouring a greater proportion of open channels. An alternative mechanism for the rebound observed immediately following lactate removal is that, in addition to enhancing open probability via Ca²⁺ chelation, lactate directly blocks the pore and this latter effect reverses more rapidly following lactate removal.

To our knowledge, this is the first description of an acid-sensing channel in either skeletal muscle or skeletal muscle-related cells. Although ASIC channels have been mentioned in the context of skeletal (Sluka et al. 2003), smooth (Yiangou et al. 2001; Holzer, 2003) and cardiac muscle before (Sutherland et al. 2000), this has always been with reference to expression in sensory neurones that innervate the muscle rather than myocytes themselves. Such studies are usually concerned with the role of ASICs in muscle-related nociception, although one recent paper has suggested a role in a different sensory process, namely the muscle pressor reflex (Li et al. 2004). In addition, good evidence for expression of the related degenerin channels has been published for C. elegans muscle (Liu et al. 1996; Garcia-Anoveros et al. 1998).

In light of the significant roles of pH changes and lactic acid in normal muscle function, our observations of ASIC responses in SJ-RH30 cells may be of considerable interest if they translate to the physiology of native skeletal muscle. Lactate produced during intense muscle activity can accumulate both inside the muscle cell and in the interstitial space. Associated with this are changes in extracellular pH. Several studies have addressed the role lactate and associated pH changes play inside muscle cells. Despite, historical views to the contrary (see for example Fitts, 1994) most of these studies have failed to uncover substantial effects of intracellular lactate or acidification on force production at near physiological temperatures (for review see Westerblad & Allen, 2003), and at least one study has suggested that lactate may protect against fatigue (Nielsen et al. 2001).

Less consideration has seemingly been given to potential effects of lactate and acid on the extracellular side of the muscle fibre. In strenuous exercise it has been reported that lactate may reach 20 mm in blood (Fitts, 1994). If ASIC channels are present on muscle cells, the combined effects of quite mild acidification and such levels of lactate would be expected to produce some degree of channel activation and a resultant depolarization and increase in membrane conductance. Tonic ASIC activation could also potentially produce a rise in both intracellular Na⁺ and Ca²⁺ (since, in addition to their predominant Na⁺ permeability, ASIC channels exhibit a measurable Ca²⁺ permeability (Yermolaieva et al. 2004). The outcomes produced by each of these features of the ASIC response will depend on the magnitude of the activated ASIC conductance and the corresponding level of other active conductances within the muscle cell membrane. Thus, a mild depolarization will reduce the size of the end plate potential required to bring the muscle cell to action potential threshold thereby potentially increasing force for a given motor neurone input. More substantial ASIC-mediated depolarization might produce spontaneous action potentials or even sufficient Na⁺ channel inactivation to generate ‘depolarization block’ of muscle excitability. ASIC-mediated increases in membrane conductance could shunt the muscle EJP, thus making it less effective, although this would require the activated conductance to be of the same order of magnitude as the resting membrane conductance to produce a substantial effect. Through a simple additive effect, ASIC triggered rises in intracellular Ca²⁺ ([Ca²⁺]_i) could enhance tension generation triggered by subsequent muscle action potentials. More substantial increases in [Ca²⁺]_i might produce effects on the sacroplasmic reticulum per se or even on the contractile apparatus. Indeed the precise effects produced are likely to reflect the proximity of any ASIC channels to the SR membrane. ASIC-mediate rises in intracellular sodium will reduce the Na⁺ driving force and may thereby alter action potential properties. Furthermore, via the Na⁺,K⁺-ATPase, rises in Na⁺ may increase cellular ATP usage and consequently augment metabolic acidosis.

We feel that our observations on the SJ-RH30 rhabdomyosarcoma indicate that a detailed investigation of ASIC expression and function in native skeletal muscle preparations may be warranted. As a first step in this direction we used TaqMan mRNA quantification methods to investigate ASIC expression in adult human muscle (n.b. SJ-RH30 is a human cell-line). We found evidence for expression of both ASIC1 and ASIC3 mRNA although no evidence for the presence of ASIC2 transcripts. Although these data provide some evidence for a potential physiological expression of lactate-sensitive ASIC channels in skeletal muscle cells, the presence of ASIC3 does not parallel the pharmacology of ASIC responses in SJ-RH30 cells where the response was dominated by an ASIC-1 type pharmacology. One possibility is that the tumour cell line we have worked with is more reminiscent of a muscle satellite cell than an adult myocyte. Certainly the level of differentiation of a tumour cell such as an SJ-RH30 and a myoblast may be more equivalent. Were ASICs to be found on myoblasts it is tempting to hypothesize that they may contribute to the triggering of the differentiation or division of such cells in response to muscle activity and/or damage, both of which are likely to result in greater production of lactate and acid.

Whatever cell type SJ-RH30 cells best represent, our observations of a lactate-modulated acid-induced current in these cells, combined with the TaqMan data from human muscle samples, suggests that a search for similar channels in native muscle tissue is well warranted. This may present certain methodological problems for the reasons we have described above, although examination of preparations such as primary cultures of neonatal muscle cells (Neville et al. 1997) may be a useful first step.