Detection of Proton Release from Cultured Human Myotubes to Identify Malignant Hyperthermia Susceptibility

The patients were selected randomly among the persons tested for MHS by IVCT during a 2-yr period. The ethics committee of the University of Ulm, Ulm, Germany, approved the procedures. Informed consent was obtained from the patients. Indications for MH testing were an adverse anesthetic event of the patient (three persons) or of one of his or her close relatives (three persons), membership of an established MHS family (eight persons), or isolated familial creatine kinase elevations (two persons). Eight of the patients were MHS and eight MHN according to the IVCT results. Mutation screening of the RyR1 gene was performed by restriction analysis or single strand conformation analysis (SSCA), 18checking for most of the known MH mutations including the 15 mutations recognized as causative in heterologous expression experiments. 8In two of the MHS patients, the G2434R mutation was identified; two further patients carried the R614C mutation. In the other subjects, no mutation could be detected.

The following solutions were used in dissection, culture, and experimental measurements (temperature, 37°C).

• Krebs-Ringer's solution: NaCl, 118 mm; KCl, 3.4 mm; MgSO⁴, 0.8 mm; KH²PO⁴, 1.2 mm; glucose, 11.1 mm; NaHCO³, 25.0 mm; CaCl², 2.5 mm; pH, 7.40.

• Dissociation solution: A Ca²⁺- and Mg²⁺-free phosphate-buffered saline solution (L1825;

Biochrom, Berlin, Germany) supplemented with 0.1% bovine serum albumin (BSA) (A4503; Sigma, Steinheim, Germany), 2.5 mg/ml Trypsin (0153–61; Difco, Sparks, MD), and 220 U/ml collagenase type A (C6885, Sigma).

Ham F12 solution bicarbonate-buffered basic medium (F0815, Biochrom).

Growth medium: Skeletal muscle cell growth medium kit (C-23060; PromoCell, Heidelberg, Germany) containing 5% fetal calf serum (FCS) and various growth factors.

Differentiation medium: Skeletal muscle cell differentiation medium kit (C-23161, PromoCell) containing no FCS and fewer growth factors.

Cytosensor^®test solution: A weakly pH-buffered basic nutrient F12 medium (N6760, Sigma) with equimolar substitution of Na-bicarbonate by NaCl (total osmolarity, 290 mOsm; pH, 7.35).

Stock solutions of the trigger reagents 4-CmC (10 mm) and caffeine (100 mm) were prepared in Cytosensor^®test solution or bi-distilled water (for IVCT). Halothane was applied using a vaporizer (Vapor 19.1; Draeger, Lübeck, Germany). Caffeine was purchased from Merck (Darmstadt, Germany), halothane from Zeneca (Plankstadt, Germany), and 4-CmC from Fluka (Neu-Ulm, Germany).

Testing for MHS was carried out according to the protocol of the European Malignant Hyperthermia Group. 6Briefly, biopsies were excised near the distal insertion of the m. vastus lateralis during regional anesthesia (3-in-1 block). Small muscle bundles were prepared, and mechanical tension was recorded while increasing concentrations of the test agents (halothane, caffeine) were applied. In most of the investigated biopsies, a dose–response curve for 4-CmC was recorded in addition. Halothane concentrations were monitored during the test by infrared spectroscopy (Iris; Draeger, Lübeck, Germany). In separate measurements, the concentrations of the test agents in the bath solutions were checked by gas chromatography (halothane) and by ultraviolet spectrophotometry (caffeine, 4-CmC).

MH-susceptible muscle showed abnormal contractures of 2 mN or more in at least one muscle specimen exposed to 0.44 mm halothane or less and  one exposed to 2 mm caffeine or less, whereas MHN muscle strips showed no abnormal response.

Two or 3 h after surgery, excised muscle tissue sections (50–100 mg) were cut in small pieces in Krebs-Ringer's solution and transferred to dissociation solution. The enzymatic dissociation (30 min at 37°C) was stopped by F12 solution containing 10% FCS. After filtration (50 and 20 μm pore size) and centrifugation (100 g , 5 min), the pellet containing the satellite cells was resuspended in growth medium. Satellite cells were cryopreserved using 10% DMSO. They were rethawed for culture and seeded onto micronets (Capsule cups; Molecular Devices, Sunnyvale, CA) for proliferation. Staining with crystal violet (according to Molecular Devices instructions) showed a cell number of about 50 myotubes per micronet. The medium was exchanged every 2 days. After 4–6 days in culture, the growth medium was exchanged for differentiation medium to induce fusion and differentiation. Experiments were carried out on days 4–6 after induction of differentiation. Microscopic inspection showed that growth and density of cells on the micronets was similar to standard cultures in Petri dishes. Petri dish cultures were always run in addition to the micronet cultures to check for reasonable growth (cell viability).

The Cytosensor^®microphysiometer was used to detect proton secretion from the cultured myotubes. Each measurement unit of the instrument contains a silicon chip with a pH-sensitive surface (fig. 1A) on the bottom of a perfusable microchamber. Light pulses from an infrared light-emitting diode (LED) induce transient currents driven by a variable voltage. Because a change in the proton concentration of the bath alters the surface potential at the solution interface, a pH change shifts the current–voltage relation. The device was calibrated automatically at the start of each experiment. Test measurements with solutions of different pH showed that the voltage shift was proportional to pH (approximately 60 mV/pH unit) in the pH range 6–8 and that changes were highly reproducible.

The measuring compartment of the microchamber containing the cells had a volume of 1.4 μl and was perfused discontinuously with a weakly buffered culture medium (see above) at a rate of 50 μl/min and a temperature of 37°C. Figure 1Bshows four consecutive perfusion cycles. The perfusion time of each cycle was 36 s. During this time (pump on), the chamber pH increased because of equilibration with the bath. On stopping the perfusion (pump off, 24-s interval), the biosensor voltage decreased (i.e. , pH decreased) because protons that were extruded by the cells accumulated in the chamber. One digitally stored raw data point (pH measurement) was acquired per second. The mean time derivative of a 20-s interval within each 24-s, pump-off interval was calculated by linear regression to determine the rate (Rate 1, Rate 2, and so on) of proton extrusion. 16,19The instrument allowed accurate measurements of the acidification rate in the milli-pH units/min range. The recorded signal amplitudes were of the order of 0.01–0.1 pH units/min. The timing of the perfusion cycle was a compromise between temporal resolution and the time required to determine the acidification rate in the pump-off interval. Pump cycle control, data acquisition, and rate calculation were carried out with the system's microcomputer (Macintosh Power PC 7600/132) and the Cytosoft^®program supplied with it. Further analysis was carried out off-line using Microsoft Excel spreadsheets.

At the beginning of the third pump cycle in figure 1B, the flow system was switched from a drug-free solution to a solution containing 600 μm 4-CmC. It can be seen that in the following stop interval, the pH decreased more rapidly, indicating a higher rate of proton secretion (Rate 3, Rate 4) and, hence, metabolic activity. In initial experiments, we applied caffeine as in the IVCT to activate myotubes. Caffeine produced variable responses and mostly a decrease rather than an increase in the proton release rate. Also, other studies indicate that caffeine is not an ideal tool to study Ca²⁺turnover in cultured cells because of unspecific effects. 9,20Because the membrane-permeant cAMP derivative dibutyryl cyclic AMP (D0627, Sigma) showed similar inhibition of proton extrusion (two experiments at 500 μm and 1 mm), we tentatively attributed the observed behavior to the known inhibitory effect of caffeine on phosphodiesterases and discontinued experiments with this drug.

The acidification rates were normalized to the rate at the beginning of the experiment, i.e. , before adding 4-CmC (predrug condition) to account for differences in rate resulting from varying numbers of myotubes in the samples. Measurements were included in the analysis when an increase in rate by 50% of the predrug baseline value was reached at any of the applied concentrations. Myotube cultures from 20 patients were measured using the described protocol; 4 patients were excluded because responses remained below threshold.

Results are presented as means and SD. Differences in location between groups were evaluated in an exploratory data analysis using exact two-sided Wilcoxon rank sum tests. P  values less than 0.05 were interpreted as indicating relevant differences. Because of the exploratory nature of the comparison, no adjustment for multiple testing was carried out. P  values were calculated using StatXact^®, Version 5 (Cytel Software, Cambridge, MA).

Figure 1Cshows the time-resolved change in the acidification rate of primary cultured human myotubes caused by 600 μm 4-CmC. Each measurement point corresponds to a single pump cycle of the perfusion system. It can be seen that the proton secretion rate increases rapidly from baseline to a peak and then decreases to a lower steady level. After removing the stimulus, the rate recovered close to the baseline value.

The 4-CmC effect on Ca²⁺release is attributed to a direct action on the Ca²⁺release channel (RyR1) of the SR, 11and the threshold for activation of RyR1 by 4-CmC has been shown to be decreased in preparations with MH mutations. 11,12Because the increase in the proton secretion rate, shown in figure 1C, likely results from the activated cell metabolism in response to the 4-CmC–triggered release of Ca²⁺, differences in the dose dependence of the acidification response between normal and MH myotubes should also be expected.

Figure 2illustrates the acidification rate responses of an MHS (fig. 2A) and an MHN (fig. 2B) culture at different 4-CmC concentrations. The drug concentration was gradually increased from a minimum value of 10 μm to 600 μm. At the end of the experiment, a considerably higher concentration (1,200 μm) was applied, which caused highly variable reactions with occasional signs of damage. Therefore, this concentration was not included in the analysis. Each drug application consisted of a 10-min exposure with the respective concentration and a 10-min washout with 4-CmC–free solution (fig. 2C). The ordinate shows the change in acidification rate as a percentage of the initial (predrug) basal rate. The responses were qualitatively similar. A step change in 4-CmC concentration was answered by a rapid increase in the acidification rate and a subsequent rapid decrease. The rate remained increased, although at a lower level, during the time of the stimulus and decreased to below that of baseline after return to the control solution.

Figure 3shows the relative change in acidification rate as a function of 4-CmC concentration using mean values obtained from cultures of eight MHS patients (filled symbols). For comparison, the concentration dependence obtained from cultures of eight MHN subjects is shown (open symbols).

In figure 3A, the total change in acidification rate is shown, whereas figure 3Bshows only the amplitude of the rapid response reached immediately on application of 4-CmC. Figure 3Cshows the change in baseline rate, measured immediately before the stimulation. In the MHN cultures, the baseline rate remained essentially constant, whereas in the MHS cultures, it gradually increased. In all three plots, larger mean values of the acidification rates could be observed for all concentrations in the MHS group. The best discrimination between MHS and MHN cells is evident from the total increase in rate (fig. 3A), i.e. , taking into account amplitude and baseline. The increase in the total rate for MHS cells was about fourfold, at a concentration of 200 μm 4-CmC. Exploratory statistical comparison of the data was carried out using the nonparametric exact Wilcoxon rank sum test. At 9 of 11 concentrations shown, P  values for comparison of the total change in acidification rate were below 0.05, showing a clear difference between the two groups.

All biopsies used for this investigation had shown unambiguous results in the European IVCT test, i.e. , the concentration thresholds for contractures of muscle fiber bundles triggered by halothane and caffeine were either normal (MHN) or clearly shifted to lower concentrations (MHS). In addition, 4-CmC–induced force responses were measured in 13 of 16 biopsies. The results agreed with the results obtained using classic IVCT agents. Figure 4Ashows the mean dose–response curves for these biopsies. On average, a contracture force comparable with the one reached at 150 μm 4-CmC in the MHN biopsies was already obtained at only 25 μm in the MHS fiber bundles.

Molecular Devices, Ismaning, Germany) data given as 4-CmC concentrations that produce a suprathreshold drug response (threshold definitions, 50% increase above baseline value of acidification rate in the Cytosensor^®test [ordinate] and 2 mN contracture force in the IVCT [abscissa]). The dashed lines indicate the concentrations that permitted full separation of the MHS and MHN groups in either test (150 μm and 75 μm, respectively).

Comparing the results of individual measurements, there was a clear correlation between the strength of the responses in the IVCT and in the acidification assay. As threshold for the acidification response, we chose an increase by 50% of the initial proton secretion rate. Figure 4Bshows on the ordinate for each tested subject the concentration at which acidification rate first exceeded threshold, i.e. , reached 150% of the initial basal rate. This concentration value is plotted against the concentration at which contracture force in the IVCT first exceeded 2 mN.

The histogram of figure 5shows the number of subjects who reached the given acidification threshold plotted versus  the 4-CmC concentration (logarithmic scale). This again underlines the complete separation of the two groups. According to figure 5, the optimal concentration range of 4-CmC to distinguish the MHS phenotype from a normal response with the Cytosensor^®test is between 150 and 200 μm. At 150 μm, all MHS patients and none of the MHN subjects had reached threshold.

In this first study to apply microphysiometry to cultured skeletal muscle, we recorded extracellular acidification of human myotubes with a sensitive pH-metric biosensor. Extracellular acidification is an integral consequence of cellular activity. Regardless of the pathway, cellular activation of mammalian cells goes along with an increased energy demand associated with ATP regeneration by anaerobic glycolysis or oxidative respiration. 16The main metabolites of these events, lactic acid and carbon dioxide, lead to intracellular acidification, which in turn causes efflux of protons to the extracellular environment. 21Therefore, increase of the energy metabolism of the activated cell usually leads to a higher rate of extracellular acidification.

Here, acidification responses of the myotubes were triggered by superfusion with 4-CmC, a substance that leads to highly specific activation of the ryanodine receptor 11and releases Ca²⁺from isolated SR vesicles. 12In skeletal myotubes, 4-CmC has been shown to trigger large intracellular Ca²⁺transients. 22The ATP requiring processes that are likely to be primarily influenced by a 4-CmC–induced Ca²⁺release are the active Ca²⁺transport into the SR and out of the cell and the actin-modulated myosin ATPase. Therefore, the acidification signal can easily be traced back to the disturbance in Ca²⁺turnover caused by 4-CmC. This is in agreement with our finding that MHS, which is known to modify the Ca²⁺release and force response to 4-CmC, also modified the acidification response in a similar way.

On application of 4-CmC, a phasic concentration-dependent increase in the proton secretion rate was observed. The effect could be divided into an initial peak that was followed by a plateau for the duration of the 4-CmC stimulus and an undershoot below the baseline when the chemical stimulus was removed. The limited time resolution of about 1 min of the Cytosensor^®microphysiometer may even have underestimated the phasic nature of the acidification response. A phasic–tonic response of the acidification rate as observed in the experiments is not uncommon in microphysiometric measurements of plasmalemmal receptor activation. In myotubes it may be a direct reflection of the intracellular Ca²⁺transients. The decrease after the initial peak may be the result of a corresponding decrease of Ca²⁺release caused by Ca²⁺-dependent inactivation of the ryanodine receptor 23–25or by depletion of the SR. On the other hand, the undershoot at the end of the signal is difficult to explain in this way and would be more compatible with two antagonistic events affecting the release of protons. An inhibitory process may persist for some time after the end of the stimulus and lead to a transient decrease in the acidification rate. This may also explain a net decrease in the acidification rate that was observed rarely in cultures exhibiting subthreshold responses as seen with caffeine and dibutyryl cAMP (see Materials and Methods). The latter may indicate a possible involvement of cAMP in inhibiting proton extrusion. Net proton consumption rather than production may also occur if creatine phosphate instead of carbon sources is used to replenish ATP. 21Finally, the phasic behavior of the 4-CmC response may also result from sarcolemmal changes that affect the mechanisms of outward flux of intracellularly generated protons in muscle. 26,27Owicki and Parce 21demonstrated that a simple transport model can generate an overshoot of extracellular acidification rate even though the intracellular pH decreases monotonically.

Adenosine triphosphate–regenerating reactions are thought to be the main causes for the characteristic detrimental events during a MH crisis, in particular the strong metabolic acidosis. Because cellular proton excretion is tightly coupled to energy consumption and replenishment by the cellular metabolism, we investigated the question as to whether the rate of extracellular acidification caused by the chemically stimulated cultured myocytes can be used to distinguish MHS from MHN preparations. The results show that the extent of the acidification response is in fact significantly different for the MHS myotubes.

The currently used test procedure for MH diagnosis (i.e. , IVCT) has been optimized to warrant high sensitivity (i.e. , very low number of false-negative results) at an acceptable specificity. 6For phenotyping to support genetic linkage analysis, a more specific test would be desirable. In particular, the search for alternative MH loci that show no linkage to the currently known loci requires precise phenotyping of MHS. 5Higher specificity (lower rate of false-positive results) can be gained by improving the diagnostic procedure. For instance, attempts have been made to include new test drugs (4-CmC, ryanodine) in the IVCT. 13,28A noninvasive method of MH testing, using nuclear magnetic resonance (NMR) spectroscopy to assess parameters of muscle energy metabolism, has been considered 29,30but failed to compete with the IVCT (for a recent summary, see Argov et al.  31).

The microphysiometry procedure presented here is a promising new approach to MH testing based on cell metabolism. However, the effect measured (acidification) is also many steps removed from Ca²⁺release. Therefore, in addition to diseases affecting Ca²⁺turnover (e.g. , central core disease, Brody disease, Duchenne muscular dystrophy), all disorders that may affect energy metabolism (e.g. , mitochondrial myopathies) and proton handling can be expected to show changes in acidification rate. Thus, a test based on detection of proton release is likely to have similar problems of specificity as with the IVCT, which is known to show abnormal responses in patients with certain myopathies. On the other hand, it may be of additional help to study the cellular pathophysiology in such disorders. It may also be of value to detect physiologic changes in organisms expressing genetically altered muscle proteins.

Even though it may be applied to isolated muscle fibers or fiber bundles, the method seems particularly suited for work on cell cultures that are not easily accessible to contraction or Ca²⁺measurements. A diagnostic test based on cultures derived from satellite cells would be less invasive than the standard IVCT because the necessary material can be acquired by needle biopsies. 14In addition, repetitive tests are feasible because cultured myocytes can be stored in liquid nitrogen for further experiments unlike the mature muscle fiber preparations that can only be studied within 6 h after taking the biopsy. Also, cultured cells can more easily be sent to distant locations for testing. Recently, it has been reported that the skeletal muscle isoform of the ryanodine receptor (RyR1) is also present in B-lymphocytes, 32opening the interesting possibility to perform a physiologic diagnostic test on blood cells to screen for forms of MHS that are linked to RyR1 mutations. We are currently exploring the use of microphysiometry for this purpose. 33 

In conclusion, this study describes a new approach to monitor Ca²⁺-induced metabolic changes in muscle cells and its use for diagnosing MHS. It may also be of benefit to solve questions regarding other human myopathies that show defects in Ca²⁺or H⁺handling or in energy metabolism.

The authors thank Frank Hafner, Ph.D., Application Scientist, and Kurt Grocki, Ph.D., Sales Manager, Molecular Devices (Ismaning, Germany), for constant support and for providing extensive time to test the Cytosensor^®microphysiometer in the laboratory. The authors also thank Ursula Mohr, Technician, and Erhard Schoch, Medical Engineer, (Department of Applied Physiology, University of Ulm, Ulm, Germany) for expert technical help, and the staff of the Rehabilitationskrankenhaus Ulm for carrying out the muscle biopsies. The authors thank Aisling Ryan, M.B., M.R.C.P.I. (Medical Doctor, Department of General Physiology, University of Ulm), for valuable suggestions on the manuscript, and Martina Kron, Ph.D. (Assistant Professor of Statistics, Department of Biometry and Medical Documentation, University of Ulm), for help with the statistical analysis.