| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1 Department of Pharmacology and Physiology, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowlegementsReferences |
|---|
(Received 7 March 2005;
accepted after revision 12 April 2005;
first published online 21 April 2005)
Corresponding author N. Shirokova: Department of Pharmacology and Physiology, UMDNJ, New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103, USA. Email: nshiroko{at}umdnj.edu
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowlegementsReferences |
|---|
In addition to the pivotal role in cell energy metabolism, mitochondria are also involved in other cellular processes of key importance. In particular, it has been suggested that these organelles participate in the control of intracellular Ca2+ homeostasis. It has been shown that mitochondrial Ca2+ uptake can modify the intracellular Ca2+ transients in a variety of cell lines and some tissues. On the other hand, [Ca2+] in the mitochondrial matrix ([Ca2+]m) controls the metabolism, as three major dehydrogenases of the mitochondrial tricarboxylic acid (TCA) cycle are Ca2+ sensitive. The increase in [Ca2+]m enhances the production of nicotinamide adenine dinucleotide (NADH), electron transport, proton leak, ATP synthesis and the production of reactive oxygen species (ROS) (for recent reviews see Duchen, 2000; Hajnóczky et al. 2000; Rizzuto et al. 2004) . All these Ca2+-dependent mechanisms, in turn, can exert positive and negative feedback effects on cytoplasmic Ca2+ signals. Therefore, calcium homeostasis, metabolism, and bioenergetics are intimately interconnected in living cells.
Release of Ca2+ from the sarcoplasmic reticulum (SR) is a required step in skeletal muscle excitation–contraction coupling (ECC). It is initiated during the depolarization of the transverse tubular membrane via an allosteric interaction between the voltage sensors of ECC (dihydropyridine receptors; DHPRs) and the associated SR Ca2+ release channels (ryanodine receptors; RyRs) (Schneider & Chandler, 1973; Ríos et al. 1993; Nakai et al. 1996). It has been proposed that the initial voltage-activated increase in local [Ca2+] at the triad opens the RyRs, which are not allosterically coupled with DHPRs, via Ca2+-induced Ca2+ release (CICR) (Ríos & Pizarro, 1988; Shirokova et al. 1996). However, the existence of CICR in mammalian skeletal muscle has been questioned. No Ca2+ sparks, the elementary events of CICR, were found in mammalian muscle cells during electrical stimulation (Shirokova et al. 1998; Conklin et al. 1999; Csernoch et al. 2004). This observation suggests the existence of some inhibitory mechanisms that suppress regenerative CICR in vivo.
In Shirokova et al. (1999), we proposed that a functional interaction between DHPRs and RyRs prevents RyRs from being activated by Ca2+. However, this idea was challenged by Kirsch et al. (2001), who demonstrated the abundance of sparks in mechanically skinned fibres, where the physical coupling between DHPRs and RyRs is presumably preserved (Lamb, 2002). In Isaeva & Shirokova (2003), we suggested that mitochondria, being in close proximity to Ca2+ release sites, are capable of interfering with Ca2+ released from the SR, thereby inhibiting CICR by a not yet established mechanism. It is conceivable that wash out of cytosolic constituents, which inevitably follows all chemical or mechanical skinning procedures, can impair mitochondrial function and relieve their inhibitory effect on CICR, thus allowing Ca2+ sparks to appear. The results we present in this paper provide further support for this hypothesis by correlating the appearance of Ca2+ sparks with the functional state of the mitochondria in skeletal muscle fibres of different metabolic strength.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowlegementsReferences |
|---|
Female rats (Sprague-Dawley, 175–200 g) and mice (Swiss Webster, 25–30 g) were killed by cervical dislocation under deep anaesthesia induced by intraperitoneal injection of sodium pentobarbital (100–200 mg (kg body weight)–1). Single fibres from rat extensor digitorum longus (EDL) or soleus muscles were manually dissected as described by García & Schneider (1993) and Shirokova et al. (1996). Single fibres from mouse flexor digitorum brevis (FDB) muscle were enzymatically dissociated by a procedure detailed in Wang et al. (1999). Fibres were transferred to an experimental chamber, pushed down against the coverslip floor of the chamber, permeabilized with saponin (as in Isaeva & Shirokova, 2003) and immersed into one of the ‘internal solutions’ (see below). The Institutional Animal Care and Use Committee at UMDNJ–New Jersey Medical School approved the use and killing method of all animals in this study.
Solutions
The L-glutamate internal solution contained (mM): potassium L-glutamate (140), Hepes (10), EGTA (0.5), sodium phosphocreatine (5), Mg-ATP (5) and CaCl2 (0.155) for nominal [Ca2+] of 
150 nM and [Mg2+] of 380 µM. The D-glutamate-based solution contained 140 mM of potassium D-glutamate, instead of potassium L-glutamate. In aspartate-based solution, potassium L-glutamate was replaced with potassium aspartate, and nominal [Ca2+] and [Mg2+] were adjusted to 150 nM and 380 µM, respectively. Dissociation constants were taken from the NIST Critically Selected Stability Constants of Metal Complexes Database 46 (US Department of Commerce, Technology Administration, NIST, Gaithersburg, MD, USA). Calculations were performed assuming 1 : 1 stoichiometry for Mg-ATP. For all solutions pH was adjusted to 7.0 and osmolality to 300 mosmol kg–1.
Confocal imaging and image processing
Local changes in cytoplasmic [Ca2+] were measured with the fluorescent Ca2+ indicator fluo-3 (penta-potassium salt, 50 µM) added to internal solutions. Mitochondrial membrane potential was monitored with the potentiometric dye tetramethyl rhodamine ethyl ester (TMRE; 10 nM). A laser scanning confocal microscope Radiance 2000 (Bio-Rad, Hercules, CA, USA) connected to a Zeiss Axiovert 100 inverted microscope equipped with a x 63, 1.2 NA, water immersion lens (Zeiss Inc., Oberkochen, Germany) was used to acquire confocal images of fluo-3 or TMRE-related fluorescence. Fluo-3 was excited with the 488 nm line of an argon laser and TMRE was excited with the 543 nm line of a He–Ne laser. The emitted light was collected above 500 nm (fluo-3) or above 570 nm (TMRE). Fibres were imaged in the XY mode at 500 lines s–1. As a rule, series of 40 images (102.8 by 102.8 µm) were acquired at 1 Hz at random locations within fibres. Discrete events of Ca2+ release were identified with an automatic detection method (Cheng et al. 1999) modified for localization of events in XY images, as previously described in Isaeva & Shirokova (2003). Because XY images carry no information about event morphology, we call all of them Ca2+ sparks.
The two-photon confocal scanner Radiance 2001 MP (Bio-Rad, Hercules, CA, USA) attached to a Nikon Eclipse TE 2000 microscope equipped with a x 60, 1.2 NA, water immerssion objective (Nikon, Tokyo, Japan) was used to monitor local changes in mitochondrial NADH autofluorescence. For two-photon excitation, a Ti:Sapphire laser (MIRA 900F, Coherent, Santa Clara, CA, USA) pumped with a frequency-doubled (Nd:YVO4) solid-state diode laser (Verdi-10 W, Coherent) provided the 80 MHz mode-locked laser pulses with 730 nm wavelength. The emitted light between 410 and 490 nm was collected with a direct (non-descanned) detector. Series of 20 images (101.2 by 101.2 µm) were acquired every 5 min from a fixed location within the fibre.
Digital photometry
To estimate the Ca2+ content of the sarcoplasmic reticulum in permeabilized skeletal muscle fibres under different experimental conditions, we adapted a technique developed by Duke & Steele (1998). For this purpose, permeabilized segments of muscle fibres were mounted into a two-Vaseline gap chamber. The chamber was designed with the middle pool to function as a flow chamber that allowed for rapid changes of the solution. The middle pool, which usually corresponds to the voltage-clamped compartment, was connected to small inlet and outlet pools at both ends, where the perfusion and suction lines were attached (more details are in Shirokova & Ríos, 1996). Throughout the experimental protocol, the middle segment of the fibre was continuously perfused with internal solution containing the Ca2+ indicator fura-2 (potassium salt, 2 µM) at a rate of 0.5 ml min–1. Waste solution was collected continuously at the outlet pool. Solutions containing 20 mM caffeine were rapidly applied (2 ml min–1) for a duration of 2 s via a dedicated inlet pipette. To minimize contraction of fast-twitch fibres, 20 µM of N-benzyl-p-toluene sulphonamide (BTS; Cheung et al. 2002) was added. Both EDL and soleus fibres were stretched to about 3.5 µm per sarcomere length.
Dual-excitation recordings of intracellular Ca2+-related fluorescence in fura-2-loaded fibres were performed with a RatioMaster M-40 high-speed fluorescence photometer (PTI, Lawrenceville, NJ, USA) mounted on a Zeiss Axiovert 200 microscope (Zeiss Inc., Oberkochen, Germany) equipped with a quartz x 40, 1.25 NA, glycerol-immersed objective (Partec GmbH, Münster, Germany). The quartz objective is designed specifically to optimize fluorescence measurements with probes excitable in the UV range of light wavelengths. The fibre was illuminated with light of 340 and 380 nm at 50 Hz. The fluorescence emission in the centre section of fibre was detected through a rectangular pinhole (20 x 40 µm). Ca2+ transients are presented as the ratio of light intensities emitted above 500 nm. The SR Ca2+ load was estimated from the amplitude of caffeine-induced cytoplasmic Ca2+ transients.
Whole-cell NADH fluorescence signals were imaged with a CH1 CCD camera (Photometrics, Tucson, AZ, USA), as previously described by Hajnóczky et al. (1999) and also by Isaeva & Shirokova (2003). Fibres were excited at 360 nm and the light emitted above 420 nm was recorded. A single series of 300 images at 0.33 Hz was acquired in each experiment.
Statistics
Values are presented as means ± S.E.M., and n represents the number of analysed cells. Student's t test was used for comparing paired observations. P < 0.05 was considered significant.
Chemicals
Fluo-3 and fura-2 were obtained from Biotium (Hayward, CA, USA). MitoTracker Green FM and TMRE were purchased from Molecular Probes (Eugene, OR, USA). Other chemicals were from Sigma (St Louis, MO, USA).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowlegementsReferences |
|---|
Time course of appearance of Ca2+ sparks correlates with the fibre's mitochondrial content
In our previous studies (Isaeva & Shirokova, 2003) we noticed that, after fibres were permeabilized with saponin and immersed into the internal solution, it took some time for spontaneous Ca2+ sparks to appear. We speculated that the delay in the appearance of sparks is related to the degradation of cellular metabolic pathways. To test this hypothesis we designed a number of experiments that correlate the onset of the appearance of Ca2+ sparks with changes in cellular metabolism.
In the first set of experiments, we examined whether the time course of Ca2+ spark appearance correlates with the metabolic capacity of the particular cell type. In this study we used muscles that contain fibres with different mitochondrial content and subcellular localization of the organelles. Rat soleus muscle contains mostly type I cells (e.g. Ariano et al. 1973). EDL skeletal muscle of rat contain 60% of type IIa fibres (Ariano et al. 1973). About 70% of the fibres from mouse FDB muscle are of the IIX type (Gonzalez et al. 2003). The oxidative capacity of these fibres increases in the order: IIX, IIa and I. In particular, the relative volume of mitochondria in soleus type I fibres is almost two times larger than that in EDL type IIa cells (Eisenberg, 1983). The larger the mitochondrial content, the more mitochondria are generally targeted to the I-bands, close to the junctional SR, where they are able to interfere with Ca2+ release (Ogata & Yamasaki, 1985).
Because each muscle contains fibres of different types, we first tested whether fibres we selected from rat EDL and soleus muscles during the dissection procedure indeed differed in mitochondrial content. To visualize mitochondria, cells were permeabilized and immersed into L-glutamate solution. The mitochondrial NADH autofluorescence was imaged with a two-photon excitation laser-scanning confocal microscope 10 min after fibre permeabilization (top panels in Fig. 1A and B). For a more quantitative analysis, mitochondria were identified with an automatic digital image processing algorithm similar to that used for spark detection. The detected mitochondria are shown as binary masks in two middle panels. The ‘mitochondrial content’ was estimated as the ratio (r) of the number of pixels occupied by mitochondria to the total number of pixels occupied by the fibre. The ratio was significantly larger in soleus fibres (Fig. 1C). Obviously, the analysis we used here is not very precise, as the algorithm is somewhat sensitive to the average NADH signal intensity. However, similar results were also obtained with the potential sensitive dye TMRE (lower two panels) and with the mitochondria-targeted fluorescent indicator MitoTracker Green FM (data not shown).
|
|
|
|
|
Temporal onset of Ca2+ sparks depends on the SR Ca2+ load
Several lines of evidence suggest that luminal [Ca2+] regulates the activity of RyR Ca2+ release channels. In particular, it has been shown that SR Ca2+ overload stimulates SR Ca2+ release in cardiac muscle cells by increasing Ca2+ spark frequency and also their magnitude (DelPrincipe et al. 1999; Lukyanenko et al. 2001). Studies on skeletal muscle also suggest the existence of some, although not well defined yet, luminal mechanisms that regulate Ca2+ release from the SR (e.g. Donoso et al. 1995; Zhou et al. 2004). The average Ca2+ content of the SR is known to be different between fibre types (e.g. Fryer & Stephenson, 1996). This can affect the temporal onset of Ca2+ sparks. Therefore, it is important to monitor the Ca2+ loading state of the SR in our experiments.
We probed the SR Ca2+ load in permeabilized EDL and soleus muscle fibres by brief applications of caffeine. On each experimental day, both types of fibres were obtained from the same animal and studied in parallel. Figure 5A and B shows representative recordings of the 340 : 380 fura-2 fluorescence ratio from saponin-permeabilized fibres. In these experiments, cells were continuously perfused with L-glutamate solution containing 2 µM fura-2. Caffeine (20 mM) was briefly (2 s) applied at 5 min intervals (indicated by arrows). Caffeine application produced a transient increase in fluorescence ratio due to release of Ca2+ from the SR. Figure 5C shows two superimposed traces of caffeine-induced Ca2+ transients recorded at about 30 min after permeabilization in EDL and soleus fibres. It can be seen that the decaying phases of the transients were markedly different, being much slower in the soleus than in the EDL fibre. This may reflect differences in the molecular makeup of the system that is responsible for the removal of Ca2+ from the cytoplasm, following its release from the SR. In particular, compared to EDL muscle fibres, soleus cells contain virtually no parvalbumin, a soluble myoplasmic Ca2+ and Mg2+ binding protein (Celio & Heizmann, 1982). The amplitude of the caffeine-induced Ca2+ transients (
R) was determined at different times after fibre permeabilization. It was considered to be an estimate of SR Ca2+ content. This assumption is also supported by the observation that under our conditions, more prolonged application of caffeine did not increase the amplitude of cytosolic Ca2+ transients. Figure 5D presents averaged data compiled from 16 EDL and 18 soleus cells. It shows that in both fibre types, refilling of the SR with Ca2+ was complete and reached a steady-state within 20 min after the cells were permeabilized and immersed in the internal solution with slightly elevated [Ca2+] (150 nM). A subtle decrease in the amplitude of the caffeine-induced Ca2+ transients after 40 min of incubation indicates a partial decline in SR Ca2+ load towards the end of such prolonged experiments.
|
The experiments shown in Fig. 6 were designed to assess how SR Ca2+ load affects the onset of Ca2+ sparks. For this, EDL cells were incubated in L-glutamate solution supplemented with 1 µM 2',5'-di(tert-butyl)-1,4-benzohydroquinone (TBQ) or 1 µM cyclopiazonic acid (CPA), two different potent inhibitors of SR Ca2+-ATPase. The drugs, at these concentrations, did not completely deplete the SR but reduced caffeine-induced Ca2+ transients (e.g. SR Ca2+ load) by more than 50% (R was reduced from 0.64 ± 0.09 to 0.29 ± 0.03, n
= 6, by TBQ; and from 0.35 ± 0.03 to 0.15 ± 0.02, n
= 5, by CPA). The examination of the temporal onset of spontaneous Ca2+ sparks under conditions of reduced SR Ca2+ load revealed that the delay to the first detection of Ca2+ sparks and the time to the maximal frequency of sparks were significantly prolonged (52 ± 3.7 min and 81 ± 2.4 min, n
= 5, correspondingly in the presence of TBQ, and 54 ± 5.1 min and 85 ± 4.5 min, n
= 5, in CPA). Therefore, the slower temporal onset of Ca2+ sparks in soleus fibres in respect to EDL cells cannot be explained by the observed difference in luminal SR [Ca2+].
|
It has been shown that cytoplasmic redox state, mainly determined by the ratio of GSSG/GSH, tightly influences the activity of RyR (Feng et al. 2000). In mammalian muscle cells, the cytoplasmic GSSG/GSH ratio is about 1 : 30 (5 mM total). Thus, under normal physiological conditions RyRs are functioning in a reduced cytosolic environment. According to Feng et al. (2000), oxidation of the cytoplasm enhances the activity of RyRs. It is conceivable that the washout of GSH from the cytosol may underlie the appearance of sparks in skinned cells. However, this possibility is less likely because the addition of 5 mM GSH into the L-glutamate solution, in order to compensate for cytosol oxidation during the washout, did not have a significant effect either on the time course of appearance of sparks, or on their maximal frequencies. In seven EDL muscle fibres studied in the presence of GSH, time to the first spark recording, time to the maximal frequency of sparks and the maximal frequency of sparks were 15.6 ± 3.4 min, 30.0 ± 2.8 min, and 2.3 ± 0.3 x 104 µm–2 s–1, respectively. The numbers are not significantly different from that obtained in eight fibres studied in parallel in L-glutamate (18.2 ± 2.8 min, 33.6 ± 3.1 min, and 2.5 ± 0.3 x 104 µm–2 s–1, respectively).
Mitochondrial substrates delay the appearance of Ca2+ sparks
Results of the experiments illustrated in Figs 2–4 suggested the involvement of mitochondria in the development of spontaneous Ca2+ activity in permeabilized skeletal muscle cells. In the next set of experiments, we further examine this possibility. For this purpose, skeletal muscle fibres of different metabolic strength (FDB, EDL and soleus) were immersed into internal solutions based on L- or D-glutamate or aspartate. While L-glutamate is a substrate of the TCA cycle, D-glutamate is not. Whereas the addition of substrates into the internal solution is supposed to stimulate mitochondrial metabolism, their removal would slow it down. Aspartate, being a product of the TCA cycle, is expected to inhibit oxidative metabolism even more than D-glutamate.
The inset in Fig. 7A shows representative changes in NADH fluorescence recorded in EDL cells when L-glutamate-based solution was subsequently replaced with one based on D-glutamate and aspartate. These measurements provided us with information about the redox state of the mitochondrial NAD system, which in turn reflects mitochondrial metabolism. Images were acquired every 10 s. Whole-cell NADH fluorescence was measured with digital photometry as described in Methods. It was spatially averaged over the region of interest corresponding to the fibre. The signal is represented in arbitrary units on the plot. The NADH fluorescence slowly decreased when the fibre was incubated in L-glutamate solution. It diminished more rapidly after L-glutamate solution was replaced with one based on D-glutamate. Replacement of D-glutamate solution with one based on aspartate led to a dramatic oxidation of the mitochondrial NADH pool, indicating a massive impairment of the mitochondrial metabolism under this experimental condition. Addition of 20 µM rotenone, an inhibitor of NADH dehydrogenase, partially restored NADH fluorescence. Similar results were obtained in four EDL fibres.
|
Mitochondrial redox state correlates with Ca2+ spark activity
The experiment illustrated in the inset of Fig. 7A revealed that mitochondrial NADH fluorescence gradually declined after fibre permeabilization. It also showed that the rate of the NADH oxidation depends on the composition of solutions the fibres were bathed in. The rate of decay of the NADH signal appeared to be the slowest in substrate-rich L-glutamate solution and the fastest in aspartate solution. This correlates with the time course of Ca2+ spark development under similar experimental conditions.
In the following series of experiments, we applied two-photon fluorescence microscopy to monitor local changes in NADH fluorescence. EDL and soleus muscle fibres were permeabilized and exposed to either L-glutamate or aspartate solutions. The NADH signal was recorded every 3 min from the same confocal plane within the fibre during 1 h. Figure 8A shows three images of NADH fluorescence acquired right after EDL fibre permeabilization (3 min) and also 30 and 60 min after the cell had been incubated in L-glutamate solution. The fluorescence signal emitted by NADH molecules within several small groups of mitochondria was determined on every image and then averaged. The values were normalized to the NADH value at the beginning of the experiment. They are plotted in Fig. 8B against time (filled circles). The decay of NADH fluorescence in this and six other EDL cells was fitted with a single exponential function (line on the plot) yielding an average 
of decay of 14.8 ± 0.81 min. The time course of decay was significantly slower in soleus fibres studied in the same way (
= 21.1 ± 1.72 min, n
= 9). When fibres of both types were examined in aspartate solution, the NADH signal had diminished significantly faster compared with that in L-glutamate (EDL:
= 6.3 ± 0.64 min, n
= 11; soleus:
= 7.4 ± 0.61 min, n
= 9). In general, the temporal onset of Ca2+ sparks in these fibre types correlated well with the rate of decay of mitochondrial redox potential (and mitochondrial metabolism) of the respective fibre.
|
There are many mechanisms by which mitochondria can influence intracellular Ca2+ homeostasis: via ATP synthesis, accumulation of Ca2+, generation of superoxide radicals, etc. Disruption of mitochondrial functions due to a collapse of the mitochondrial redox potential can subsequently alter the pattern of intracellular Ca2+ signalling.
One obvious pathway through which mitochondria can affect cytosolic Ca2+ signals is via generation of ATP. Reduction in metabolically produced ATP can, in general, affect the function of two major molecules involved in intracellular Ca2+ homeostasis: RyR Ca2+ release channels and SR Ca2+-ATPases. However, this mechanism is unlikely to influence the temporal onset of sparks because: (1) all bathing solutions contained 5 mM ATP to minimize the contribution of metabolically generated ATP; and (2) incubation of EDL cells in L-glutamate solution with 2.5 µM oligomycin, an inhibitor of the F1/F0 ATP synthase, did not modify the time course of spark appearance. In seven fibres studied in the presence of oligomycin, time to the first spark recording, time to the maximal frequency of sparks and the maximal frequency of sparks were 21.2 ± 1.4 min, 35.1 ± 2.1 min, and 1.1 ± 0.1 x 104 µm–2 s–1, respectively. The numbers are not significantly different from that obtained in 15 fibres studied in parallel in L-glutamate (19.1 ± 1.7 min, 32.1 ± 2.2 min, and 1.1 ± 0.2 x 104 µm–2 s–1, respectively).
Previously we have suggested that mitochondrial Ca2+ uptake plays a significant role in the modulation of spontaneous Ca2+ sparks in permeabilized mammalian muscle fibres (Isaeva & Shirokova, 2003). We have shown that oxidation of the mitochondrial NADH pool is accompanied by a decrease in the driving force for Ca2+ uptake via the electrogenic Ca2+ uniporter. It is possible that a gradual impairment of mitochondrial Ca2+ uptake after fibre permeabilization and wash out of the cytosol may influence the temporal pattern of Ca2+ spark appearance. Unfortunately, most of the mitochondrial Ca2+ uniporter inhibitors are not specific and are likely to affect cytosolic Ca2+ transients by multiple interconnected pathways (e.g. via direct inhibition of Ca2+ release channels, generation of ROS, etc.). In addition, Ru360, the most specific blocker of the uniporter, appears to have a poor stability (Wang & Thayer, 2002; also Dr A. P. Thomas, personal communication). We were not able to prevent oxidation of Ru360 in long-lasting experiments. Therefore, up to date, we have no adequate experimental tools to explore to what extent mitochondrial Ca2+ buffering is involved in unmasking sparks in skinned preparations. Consequently, we cannot rule out the mitochondrial Ca2+ uptake as the mechanism that is at least in part responsible for the phenomena described above.
Another possible way for mitochondria to affect cytoplasmic Ca2+ signals is via generation of ROS. During aerobic respiration to generate ATP in mitochondria, leakage of electrons regularly produces mitochondrial superoxide anions that are later reduced to H2O2 by manganese superoxide dismutase. H2O2 appears to be one of the primary messenger molecules in the redox signalling pathway (e.g. reviewed by Reid, 2001; Brookes et al. 2004). It can travel somewhat to interact with remote targets, even though its life time is short because of its rapid metabolism. However, since the SR lies very close to mitochondria, it may well be influenced by ROS released by these organelles. Every cell has a set of potent ROS scavengers to prevent oxidative damage. As one of the major targets for ROS, mitochondria have their own defence instrument. Because catalase is not detected in the mitochondria of mammalian tissues except heart (Radi et al. 1991), mitochondrial glutathione peroxidase plays a key role in metabolizing hydrogen peroxide. Therefore, reduced glutathione (GSH), required for the activity of mitochondrial glutathione peroxidase, is the potential free radical scavenger. GSH is synthesized in the cytosol and transported into mitochondria (Griffith & Meister, 1985; Martensson et al. 1990). However, oxidized glutathione disulphide (GSSG) in the mitochondria cannot be exported back to the cytosol (Olafsdottir & Reed, 1988) for conversion to GSH. Therefore, mitochondrial NADPH is a required reducing equivalent for the regeneration of GSH from GSSG. A reduction of the mitochondrial redox potential can consequently lead to a depletion of the mitochondrial GSH pool, weakening the mitochondrial defence against ROS and finally increasing the leak of ROS to the cytoplasm. This, in turn, can have effects on the RyRs and result in Ca2+ sparks.
To test whether ROS indeed can account for the development of Ca2+ sparks in permeabilized mammalian muscle cells, we first studied the time course of spark appearance under experimental conditions where exogenous scavengers chelated ROS. Figure 9A and B illustrates the experiments in which EDL muscle fibres were exposed to aspartate-based internal solutions, and to aspartate solution supplemented with 600 U ml–1 of superoxide dismutase (SOD) or with a combination of 600 U ml–1 of SOD and 500 U ml–1 of catalase. Fibres were studied in parallel under all three conditions. The temporal onset of Ca2+ spark appearance in aspartate was similar to that listed in Table 1 (time to the first spark detection and time till the maximal frequency of sparks was reached were 7.9 ± 1.01 min and 11.4 ± 1.80 min, n = 7, respectively). SOD alone substantially inhibited Ca2+ sparks and delayed their temporal onset (corresponding times were 11.0 ± 1.00 min and 18.0 ± 1.22, n = 5). Two scavengers, used in combination, further suppressed sparks and significantly delayed their appearance compared with that in control (times were 11.3 ± 1.25 and 18.3 ± 1.17 min, n = 4).
|
Finally, we tested whether the introduction of ROS into the bathing solution produces an effect opposite to that of scavengers (Fig. 9E and F). H2O2 at 50 µM added to the aspartate solution somewhat increased the maximal Ca2+ spark frequency, but only slightly accelerated their appearance. It is possible that the changes in the temporal onset of sparks were difficult to detect because the onset is very fast in aspartate. Interestingly, the effects of H2O2 appeared to be concentration dependent. While 50–200 µM H2O2 promoted sparks, 10 mM H2O2 completely inhibited them (data not shown).
Figure 10 illustrates the experiments with ROS scavengers, SOD and catalase, carried out in EDL muscle cells in D-glutamate (Fig. 10A and B) or in L-glutamate (Fig. 10C and D) containing internal solutions. In D-glutamate, SOD and SOD in combination with catalase dramatically reduced the Ca2+ spark frequency and also substantially delayed the spark appearance. Time to the first spark detection and time when the maximal frequency of sparks was recorded were 23.8 ± 1.25 min and 33.8 ± 1.57 min, n = 8, and 36.7 ± 6.01 min and 46.7 ± 1.67 min, n = 3, in SOD and SOD and catalase, respectively. These times were significantly longer than those recorded in the control experiments (no scavengers added) carried out in parallel (13.3 ± 2.04 min and 23.9 ± 2.32 min, n = 8). Similar results were obtained in L-glutamate. Addition of scavengers not only decreased the spark frequency, but also delayed their temporal onset. Times to the first spark detection and to their maximal frequency were 28.8 ± 1.25 min and 36.3 ± 2.39 min, n = 4, and 37.7 ± 5.05 min and 46.7 ± 2.89 min, n = 3, in SOD and SOD and catalase, respectively, and corresponding times recorded in the control group of experiments were 18.8 ± 1.25 min and 30.0 ± 1.64 min, n = 8).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowlegementsReferences |
|---|
Ca2+ sparks in skeletal muscle cells
The discovery of Ca2+ sparks in cardiac myocytes by Cheng et al. (1993) constituted an important advance in our understanding of the control of Ca2+ release in muscle cells. Ca2+ sparks are often referred to as elementary events of CICR. Infrequent appearance of Ca2+ sparks in adult mammalian skeletal muscle fibres was considered as an indication of a less prominent role of CICR in mammalian ECC, compared to amphibian muscle. Several explanations for this difference have already been considered but most have been rejected. One possibility to account for the inhibited CICR in mammals was envisaged in the molecular makeup of the muscle fibres themselves. Most adult mammalian muscles express only the RyR1 isoform, whereas amphibian skeletal muscle and embryonic mammalian muscle, in which sparks were found, express two RyR isoforms, RyR 
and RyR ß or RyR1 and RyR3, respectively (for reviews see Sutko & Airley, 1996; Franzini-Armstrong & Protasi, 1997). Naturally, it has been suggested that RyR3 may be necessary for skeletal muscle to produce Ca2+ sparks. However, this hypothesis did not sustain experimental challenge: spontaneous Ca2+ sparks were detected in developing skeletal muscle from RyR3-knockout mice (Shirokova et al. 1999; Conklin et al. 2000). In addition, in developing skeletal muscle, Ca2+ sparks were found to be spatially segregated and mutually exclusive with the homogeneous Ca2+ release induced by membrane depolarization. This observation led to the suggestion that an allosteric negative feedback interaction between DHPRs and RyRs prevents CICR (Shirokova et al. 1999). However, the latter hypothesis was again questioned by the work of Kirsch et al. (2001) who demonstrated the abundance of sparks in mechanically skinned muscle cells. Our recent studies (Isaeva & Shirokova, 2003) added a new twist to this complicated story. Our results suggested that mitochondria, being in close proximity to the Ca2+ release sites, also play an important role in inhibiting CICR.
The present results on the temporal onset of the appearance of Ca2+ sparks in different fibre types lend further support for the notion that spontaneous Ca2+ activity in permeabilized cells is suppressed by mitochondria. The time lag between fibre permeabilization and first recordings of Ca2+ sparks suggests that a gradual degradation of physiological processes within the cell accounts for the appearance of sparks. The match between a fibre's mitochondrial content and the temporal onset of sparks, the relationship between oxidative strength of the bathing solution and spark appearance, and the correlation between spark development and the dissipation of the mitochondrial redox potential indicate that disruption of mitochondrial functions during the cytosol washout can be one of the possible reasons for the relief of CICR inhibition.
Intracellular Ca2+ signalling and mitochondria
Mitochondria exert a multifactorial influence on a variety of cell functions. First, these organelles produce ATP through oxidative metabolism to supply energy. Second, mitochondria can accumulate Ca2+ whenever the local cytoplasmic [Ca2+] rises above a critical ‘set point’. Third, the mitochondrial respiratory chain is a major site for generation of superoxide radicals. Fourth, mitochondria can disrupt the fine balance between reduced and oxidized forms of cellular major redox couples, GSSG/GSH and NAD+/NADH, thus disturbing the cytosolic redox potential and consequently the activity of various redox-sensitive molecules. Most likely mitochondria have many more cellular and signalling functions, also depending on the cell type. Each of these functions is influenced by and can in turn influence the intracellular Ca2+ homeostasis.
It has been demonstrated in a large number of functional studies that mitochondrial Ca2+ uptake can participate in shaping global and local Ca2+ signals in a variety of cell types. For example, inhibiting the mitochondrial Ca2+ uniporter increased the frequency of spontaneous Ca2+ sparks in cardiac muscle cells (Pacher et al. 2002) and in mammalian skeletal muscle fibres (Isaeva & Shirokova, 2003), and, on the other hand, inhibited Ca2+ sparks in smooth muscle cells (Cheranov & Jaggar, 2004). This suggests that, at least in cardiac and skeletal muscle, mitochondrial Ca2+ uptake normally leads to suppression of RyR channel activity. Unfortunately, we were not able to test here to what extent the gradual decline of Ca2+ uptake via the Ca2+ uniporter, resulting from the oxidation of the mitochondrial NADH pool and consequent depolarization of the mitochondrial membrane, is responsible for the delayed development of sparks. So, this possibility is yet to be investigated.
The reduction of the mitochondrial redox potential not only leads to a lower driving force for the Ca2+ uniporter, but also results in a partial oxidation of the mitochondrial GSH pool. This, in turn, reduces the mitochondrial protection against ROS generated during oxidative phosphorylation, and presumably increases the escape of ROS (mostly H2O2) into the cytosol. Although the effect of ROS on SR Ca2+ release in skeletal muscle is still controversial (Brotto & Nosek, 1996; Andrade et al. 1998; Plant et al. 2002; Posterino et al. 2003), it has been shown that H2O2 increases the activity of RyRs incorporated into artificial bilayers (Boraso & Williams, 1994; Favero et al. 1995) and impairs the function of the Ca2+-ATPase in SR vesicle preparations (Xu et al. 1997). In addition, Posterino et al. (2003) found that treatment with H2O2 markedly potentiated caffeine-induced Ca2+ release in mechanically skinned skeletal muscle fibres isolated from rats. In our experiments, ROS scavengers (SOD and catalase) dramatically altered the pattern of appearance of Ca2+ sparks by reducing their maximal frequency and delaying their appearance. In contrast, H2O2 in the micromolar range promoted sparks and sped up their appearance. These data indicate ROS can be responsible for unmasking sparks in skinned preparations. They also suggest that weakening of the mechanisms protecting the cell against oxidation results from cytosol washout after permeabilization. Apparently, when ROS production exceeds the cell's capacity to chelate free radicals, CICR inhibition can be removed and sparks can appear. In addition, based on the experiments with 10 mM H2O2, we can also speculate that accumulation of ROS in the cytosol above a certain level may be responsible for complete shutdown of spark activity seen later on in the experiments.
In summary, our results show that the temporal onset of the appearance of spontaneous Ca2+ sparks in permeabilized mammalian muscle cells is closely associated with the redox state of mitochondria. We suggest that a misbalance between mitochondrial ROS generation and the cell's ability to counteract and neutralize oxidative stress is most likely responsible for the collapse of the physiological CICR inhibition and the subsequent manifestation of Ca2+ sparks. Out data also indicate that under physiological conditions (i.e. before skinning) mitochondrial ROS production and cellular ROS chelating are well balanced and in equilibrium, which keeps the number of spontaneous Ca2+ sparks minimal. However, we cannot exclude the possibility that other unknown diffusible factors produced in mitochondria (Ørtenblad & Stephenson, 2003), mitochondrial Ca2+ uptake, or some rearrangement in the structural environment of RyR channels caused by permeabilization and cellular swelling can also contribute to the described phenomena.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowlegementsReferences |
|---|
Ariano MA, Armstrong RB & Edgerton VR (1973). Hindlimb muscle fiber populations of five mammals. J Histochem Cytochem 21, 51–55.[Abstract]
Boraso A & Williams AJ (1994). Modification of the gating of the cardiac sarcoplasmic reticulum Ca2+-release channel by H2O2 and dithiothreitol. Am J Physiol 267, H1010–H1016.[Medline]
Brookes
PS, Yoon
Y, Robotham
JL, Anders
MW
&
Sheu
SS (2004). Calcium, ATP, and ROS: a mitochondrial love-hate triangle. Am J Physiol Cell Physiol
287, C817–C833.
Brotto
MA
&
Nosek
TM (1996). Hydrogen peroxide disrupts Ca2+ release from the sarcoplasmic reticulum of rat skeletal muscle fibers. J Appl Physiol
81, 731–737.
Celio MR & Heizmann CW (1982). Calcium-binding protein parvalbumin is associated with fast contracting muscle fibres. Nature 297, 504–506.[CrossRef][Medline]
Cheng
H, Lederer
WJ
&
Cannell
MB (1993). Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science
262, 740–744.
Cheng
H, Song
LS, Shirokova
N, González
A, Lakatta
EG, Ríos
E
&
Stern
MD (1999). Amplitude distribution of calcium sparks in confocal images: theory and studies with an automatic detection method. Biophys J
76, 606–617.
Cheranov
SY
&
Jaggar
JH (2004). Mitochondrial modulation of Ca2+ sparks and transient KCa currents in smooth muscle cells of rat cerebral arteries. J Physiol
556, 755–771.
Cheung A, Dantzig JA, Hollingworth S, Baylor SM, Goldman YE, Mitchison TJ & Straight AF (2002). A small-molecule inhibitor of skeletal muscle myosin II. Nat Cell Biol 4, 83–88.[CrossRef][Medline]
Conklin
MW, Ahern
CA, Vallejo
P, Sorrentino
V, Takeshima
H
&
Coronado
R (2000). Comparison of Ca2+ sparks produced independently by two ryanodine receptor isoforms (type 1 or type 3). Biophys J
78, 1777–1785.
Conklin
MW, Barone
V, Sorrentino
V
&
Coronado
R (1999). Contribution of ryanodine receptor type 3 to Ca2+ sparks in embryonic mouse skeletal muscle. Biophys J
77, 1394–1403.
Csernoch
L, Zhou
J, Stern
MD, Brum
G
&
Ríos
E (2004). The elementary events of Ca2+ release elicited by membrane depolarization in mammalian muscle. J Physiol
557, 43–58.
DelPrincipe F, Egger M & Niggli E (1999). Calcium signalling in cardiac muscle: refractoriness revealed by coherent activation. Nat Cell Biol 6, 323–329.
Donoso
P, Prieto
H
&
Hidalgo
C (1995). Luminal calcium regulates calcium release in triads isolated from frog and rabbit skeletal muscle. Biophys J
68, 507–515.
Duchen
MR (2000). Mitochondria and calcium: from cell signalling to cell death. J Physiol
529, 57–68.
Duke AM & Steele DS (1998). Effects of cyclopiazonic acid on Ca2+ regulation by the sarcoplasmic reticulum in saponin-permeabilized skeletal muscle fibres. Pflugers Arch 436, 104–111.[CrossRef][Medline]
Eisenberg BA (1983). Quantitative ultrastructure of mammalian skeletal muscle. In Handbook of Physiology, section 10, Skeletal Muscle, ed. Peachey LD, p. 95 American Physiological Society, Bethesda, MD, USA.
Favero
TG, Zable
AC
&
Abramson
JJ (1995). Hydrogen peroxide stimulates the Ca2+ release channel from skeletal muscle sarcoplasmic reticulum. J Biol Chem
270, 25557–15563.
Feng W, Liu G, Allen PD & Pessah IN (2000). Transmembrane redox sensor of ryanodine receptor complex. Biol Chem 275, 35902–35907.[CrossRef]
Franzini-Armstrong
C
&
Protasi
F (1997). Ryanodine receptors of striated muscles: a complex channel capable of multiple interactions. Physiol Rev
77, 699–729.
Fryer MW & Stephenson DG (1996). Total and sarcoplasmic reticulum calcium contents of skinned fibres from rat skeletal muscle. J Physiol 493, 357–370.[Medline]
García
J
&
Schneider
MF (1993). Calcium transients and calcium release in rat fast-twitch skeletal muscle fibres. J Physiol
463, 709–728.
Gonzalez
E, Messi
ML, Zheng
Z
&
Delbono
O (2003). Insulin-like growth factor-1 prevents age-related decrease in specific force and intracellular Ca2+ in single intact muscle fibres from transgenic mice. J Physiol
552, 833–844.
Griffith
OW
&
Meister
A (1985). Origin and turnover of mitochondrial glutathione. Proc Natl Acad Sci U S A
82, 4668–4672.
Hajnóczky G, Csordas G, Madesh M & Pacher P (2000). The machinery of local Ca2+ signalling between sarco-endoplasmic reticulum and mitochondria. J Physiol 15, 69–81.
Hajnóczky
G, Hager
R
&
Thomas
AP (1999). Mitochondria suppress local feedback activation of inositol 1,4,5-trisphosphate receptors by Ca2+. J Biol Chem
274, 14157–14162.
Isaeva EV & Shirokova N (2003). Metabolic regulation of Ca2+ release in permeabilized mammal
), 17 EDL (•) and 17 soleus (
) fibres. Left vertical axis corresponds to EDL and soleus cells, right axis corresponds to FDB cells. B, cumulative frequencies of events. C, changes in resting fluorescence within fibres.
) show averaged frequency of sparks recorded in 5 fibres at different times after they were immersed into solution containing 1 µM TBQ. Circles (•) represent control data taken from 
, n
= 5). D, cumulative frequencies of sparks in control and in CPA.
, n
= 5), or SOD and catalase (
