INTRODUCTION

Top Abstract Introduction Procedures Results Discussion Conclusions References

Calcium release from internal stores contributes to the transient increments in intracellular free calcium concentration ([Ca²⁺]) that underlie many responses of excitable cells, such as synaptic plasticity and gene expression in neurons (Gosh and Greenberg, 1995), and contraction in skeletal and cardiac muscle (Meissner, 1994). Calcium is released from these stores through two separate pathways, the inositol 1,4,5-trisphosphate receptor channels (Furuichi et al., 1994) and the ryanodine-sensitive calcium release channels (RyR channels) (Meissner, 1994; Zucchi and Ronca-Testoni, 1997).

Different species and different excitable cells express diverse ryanodine receptor (RyR) isoforms (Furuichi et al., 1994; Ogawa, 1994; Sutko and Airey, 1996). In mammals, the RyR-1 isoform is the main isoform found in adult skeletal muscle cells (Furuichi et al., 1994), whereas RyR-2 is the main isoform of cardiac muscle cells (Coronado et al., 1994; Furuichi et al., 1994). Neurons contain the RyR-3 isoform in addition to the RyR-1 and RyR-2 isoforms (Giannini and Sorrentino, 1995). Small amounts of RyR-3 are also found in adult diaphragm muscle (Murayama and Ogawa, 1997). Amphibian, avian, and fish skeletal muscle cells express two different RyR isoforms,  and  (Ogawa, 1994; Sutko and Airey, 1996). The amphibian and avian  and  isoforms have a significant extent of homology with the RyR-1 and the RyR-3 mammalian isoforms, respectively (Ogawa, 1994; Oyamada et al., 1994; Ottini et al., 1996).

The physiological mechanisms of activation of these diverse RyR channels are not well understood at the present time (Furuichi et al., 1994; Giannini and Sorrentino, 1995; Melzer et al., 1995; Zucchi and Ronca-Testoni, 1997). Only for the cardiac muscle RyR-2 isoform is there general consensus that calcium is the physiological agonist that triggers calcium release from sarcoplasmic reticulum (SR) after plasma membrane depolarization (Melzer et al., 1995). Yet all single RyR channels studied so far show activation by µM cytosolic [Ca²⁺] (Coronado et al., 1994; Meissner, 1994; Melzer et al., 1995). In the particular case of the RyR-3 isoform, functionally expressed single RyR-3 channels from mammalian smooth muscle are activated by µM [Ca²⁺] (Chen et al., 1997), and ryanodine binding to RyR-3 obtained either from mammalian brain (Murayama and Ogawa, 1996) or from diaphragm muscle is calcium dependent (Murayama and Ogawa, 1997). Thus it is likely that calcium activation of channel activity, the underlying mechanism of calcium-induced calcium release, is a general feature of all the RyR isoforms present in excitable cells.

Different single-channel responses to changes in cis (cytosolic) [Ca²⁺] have been reported. Channels present in SR vesicles isolated from mammalian cardiac muscle show sigmoidal activation by cis µM [Ca²⁺]. Most reports show a lack of inhibition by increasing [Ca²⁺] up to 1-2 mM (Rousseau et al., 1986; Anderson et al., 1989; Holmberg and Williams, 1990; Chu et al., 1993). This type of calcium dependence will be named the C calcium dependence (for cardiac). Yet cardiac channel activity decreases after changing pCa from 8 to 4 or 3 in a time-resolved bilayer system (Schiefer et al., 1995). Furthermore, variable inhibition of cardiac channel activity recorded in steady-state conditions was reported when cis [Ca²⁺] was increased to 2 mM or beyond (Laver et al., 1995; Copello et al., 1997). These results suggest that cardiac RyR channels present more than one type of response to increases in cytoplasmic [Ca²⁺].

The ryanodine-sensitive calcium channels of mammalian skeletal muscle SR display a bell-shaped calcium dependence (which will be named the MS calcium dependence (for mammalian skeletal)) with activation by µM [Ca²⁺] (Smith et al., 1986) and inhibition by [Ca²⁺]  0.1 mM (Fill et al., 1990; Chu et al., 1993). However, recent studies have shown that these channels also display more than one type of calcium dependence (Copello et al., 1997). Single RyR channels obtained from amphibian, avian, and fish skeletal muscle display two types of calcium dependencies, C and MS (Bull and Marengo, 1993; Percival et al., 1994; O'Brien et al., 1995).

We have described how RyR channels derived from rat brain cortex neurons exhibit three types of responses to changes in cis [Ca²⁺] (Marengo et al., 1996). In addition to the C and MS calcium dependencies, a third response, the low P_o calcium dependence, was observed with the highest frequency. This low P_o behavior is characterized by a bell-shaped calcium dependence with fractional open times (P_o) less than 0.1 in the [Ca²⁺] range 0.1 µM to 0.5 mM (Marengo et al., 1996). A calcium dependence with the same characteristics as the low P_o dependence has recently been reported for mammalian skeletal RyR channels (Copello et al., 1997).

In addition to cytosolic [Ca²⁺], many other agents modify RyR channel activity (Coronado et al., 1994; Meissner, 1994), among them, agents that modify sulfhydryl (SH) groups. Thus SH reagents (Abramson et al., 1995), free radicals (Stoyanovsky et al., 1994), and hydrogen peroxide (Favero et al., 1995) activate RyR channels incorporated into planar lipid bilayers. Heavy metals (Abramson et al., 1983; Trimm et al., 1986; Salama et al., 1992), mercurials (Bindoli and Fleischer, 1983), dithiodipyridines (Nagura et al., 1988; Prabhu and Salama, 1990; Donoso et al., 1997), and derivatives of nitric oxide (Stoyanovsky et al., 1997) trigger calcium release from isolated SR vesicles as well. Furthermore, SH reagents shift the calcium activation curve of ryanodine binding to the left (Stuart et al., 1992; Favero et al., 1995), and heavy metals enhance the calcium sensitivity of tension development in skinned fibers from rabbit psoas muscle (Salama et al., 1992). These findings suggest that SH reagents modify the activation of mammalian RyR channels by cytosolic [Ca²⁺].

In this work we studied in steady-state conditions the responses to changes in cis [Ca²⁺] of single RyR-channels present in SR isolated from rabbit or frog skeletal muscle, or from rabbit cardiac muscle. All native RyR channels studied displayed more than one calcium dependence, despite the channel isoform(s) present in the SR vesicles. Furthermore, we found that 1) oxidation of SH residues modified the calcium dependencies of all single RyR channels, regardless of their origin, and 2) reduction of channels with the C calcium dependence (native or oxidized) caused a decrease in channel activity. We propose that the oxidation state of the channel protein is a decisive factor in determining the calcium dependence of the channel activity exhibited by any given isoform.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion Conclusions References

Membrane isolation

Triad-enriched SR vesicles were isolated from fast skeletal muscles of rabbit (New Zealand) and frog (Caudiverbera caudiverbera), as reported elsewhere (Hidalgo et al., 1993). Endoplasmic reticulum vesicles from rat (Sprague-Dawley) brain cortex were obtained as described previously (Marengo et al., 1996). All membranes were isolated with or without 3-5 mM dithiothreitol (DTT) present throughout the isolation procedure, as indicated in the text. SR from rabbit cardiac muscle was obtained by using a modification of the procedure developed to isolate triads from skeletal muscle (Hidalgo et al., 1993). Briefly: the ventricles of one rabbit heart were cleaned of blood and connective tissue, and were placed in 5 volumes of ice-cold buffer (0.15 M KCl, 5 mM MgS0₄, 20 mM 3-[N-morpholino]propanesulfonic acid/Tris, pH 6.8, 1 µg/ml leupeptin, 1 µg/ml pepstatin A, 0.4 mM benzamidine, 1 mM phenylmethylsulfonyl fluoride, plus or minus 3 mM DTT). The tissue was finely minced and homogenized twice in a Polytron homogenizer (Heidolph Diax 600) for 45 s at 13,500 rpm. Cardiac SR vesicles sedimenting between 1,500 and 17,000 × g were collected by differential centrifugation and were resuspended in the same buffer used for homogenization. To remove contractile proteins, the suspension was made 0.6 M in KCl by the addition of solid salt, and the sediment obtained at 1500 × g was discarded. Cardiac SR vesicles were collected by sedimentation at 17,000 × g and were resuspended in a small volume of homogenization buffer plus 0.3 M sucrose. Small aliquots were quickly frozen in liquid N₂ and stored at 80°C.

Channel recording and analysis

Channel recording and analysis were performed as described previously (Bull and Marengo, 1994). Values of P_o were calculated from single-channel records lasting at least 180 s. All experiments were carried out at room temperature (22-24°C). The recording conditions were 40 mM Ca²⁺-HEPES, 15 mM HEPES/Tris, pH 7.4, in the trans compartment; 225 mM HEPES/Tris, pH 7.4, and variable [Ca²⁺] in the cis compartment. To set the desired cis [Ca²⁺], 0.5 mM total Ca²⁺ and sufficient N-(2-hydroxyethyl)-ethylenediamine-triacetic acid (HEDTA) or EGTA were added to the cis compartment. Resulting cis [Ca²⁺] values were routinely checked with a calcium electrode.

After channel incorporation into the bilayer and establishment of recording conditions (Bull and Marengo, 1994), oxidation was carried out by the addition of 2,2'-dithiodipyridine (DTDP) or thimerosal to the cis compartment. SH oxidation was stopped by removal of the nonreacted reagent through extensive perfusion of the cis compartment (5-10 times the cis volume) with a solution containing 225 mM HEPES/Tris (pH 7.4). Native or oxidized channels were treated with SH-reducing agents by essentially the same procedure. The cis [Ca²⁺] used during incubation with SH reagents is specified in the text. The pseudo-steady states reached during oxidation or reduction were defined as periods of time during which the slope of the change in P_o versus time, calculated in successive frames of 1.024 s, did not differ from zero.

Theoretical analysis of calcium dependence

To fit the experimental data, the following general function (Bull and Marengo, 1993; Marengo et al., 1996) was used for the three different calcium dependencies:

(1)

In this equation the parameter P_o,max corresponds to the theoretical P_o value of maximum activation by calcium; K_a and K_i correspond to the calcium concentrations for half-maximum activation and inhibition of channel activity, respectively; and n is the Hill coefficient for activation and inhibition. Data fitting yielded the particular cases described in detail in the Results section.

Materials

Lipids were obtained from Avanti Polar Lipids (Birmingham, AL). Protease inhibitors and other reagents were obtained from Sigma Chemical Co. (St. Louis, MO).

	RESULTS

Top Abstract Introduction Procedures Results Discussion Conclusions References

High-conductance calcium channels (~100 pS with Ca²⁺ as permeant ion) were obtained after fusion with planar lipid bilayers of vesicles isolated from the different excitable tissues used in this work. The addition of µM ryanodine (not shown) always induced the characteristic subconductance open state produced by this alkaloid (Rousseau et al., 1987).

Effect of cis [Ca²⁺] on the activity of native RyR channels from skeletal or cardiac muscle

Native RyR channels derived from rabbit (Fig. 1 A) or frog (Fig. 1 B) skeletal muscle SR presented the same three calcium dependencies: low P_o (solid diamonds), MS (open circles), and C (solid circles), displayed by the RyR channels of brain cortex neurons (Marengo et al., 1996). The native RyR channels of SR vesicles isolated from rabbit ventricular muscle cells presented the MS calcium dependence (Fig. 1 C, open circles), in addition to the most commonly reported C response (Fig. 1 C, solid circles). Channels with the low P_o behavior were not observed in cardiac SR vesicles (N = 19 experiments).

View larger version (16K): [in this window] [in a new window]   FIGURE 1   Effect of cis [Ca2+] on the activity of RyR-channels derived from muscle SR vesicles. Data obtained with RyR-channels from SR vesicles isolated with or without DTT are presented. The panels show the changes in Po as a function of cis [Ca2+]; values are given as mean ± SEM. For RyR channels that presented the low Po calcium dependence (), the solid line through the data was obtained by the best fit of the experimental points to Eq. 2, as defined in the text. For RyR channels that displayed the MS () or the C () calcium dependence, the lines through the experimental points represent the best fits of the data to Eqs. 1 or 3, respectively, as defined in the text. The best fits of the experimental points to these equations yielded the parameters given in Table 1. Data for MS and C calcium dependencies of SR from frog skeletal muscle, as well as theoretical curves for the C calcium dependence, were taken from Bull and Marengo (1993), and are included for comparison purposes.

The presence of DTT as a reducing agent during isolation of SR vesicles affected the probability of finding the different calcium dependencies. Whereas SR vesicles isolated from cardiac muscle with DTT yielded with comparable frequency native channels with the MS or the C calcium dependencies (N = 14 experiments), cardiac vesicles isolated without DTT revealed only channels with the C calcium dependence (N = 5). Skeletal SR vesicles isolated in the presence of DTT lacked channels with C calcium dependence (N = 14, rabbit; N = 11, frog); this calcium dependence was only observed in channels from skeletal SR vesicles isolated without DTT (N = 3, rabbit; N = 37, frog). A detailed description of each calcium dependence follows.

Low P_o calcium dependence

(2)

MS calcium dependence

C calcium dependence

(3)

Sulfhydryl oxidation modified the activity of single RyR-channels from brain cortex neurons, skeletal, or cardiac muscle

The observation that SR vesicles isolated with or without DTT presented different calcium dependencies suggests that SH reagents modify the channel response to changes in cis [Ca²⁺]. To test this hypothesis, the effects of two different SH-oxidizing reagents, DTDP and thimerosal, on channels from skeletal and cardiac muscle and from brain cortex neurons were investigated. DTDP reacts with free SH groups, forming a covalent disulfide bond, and this reaction can be reversed by reducing agents; thimerosal, in contrast, reacts irreversibly with SH groups (Brocklehurst, 1979).

The addition of DTDP (N = 35 experiments) or thimerosal (N = 10 experiments) to the cis compartment caused an increase in P_o in all native channels that spontaneously displayed either the low P_o or the MS behavior, as detailed below. With a delay that depended on channel previous activity, new pseudo-steady-state P_o values, lasting several minutes, became apparent. These new pseudo-steady states had P_o values significantly higher (p < 0.001) than the preceding stationary P_o values, and the slope of the change in the new P_o values versus time was equal to zero (see Experimental Procedures). No changes in channel P_o were observed after the addition of water-soluble thimerosal to the trans (luminal) compartment (N = 4), indicating that the SH residues involved in the observed P_o changes were only accessible to the cis solution. Long incubations or the addition of high concentrations of SH reagents caused a reduction in the activity of all channels tested, rendering them insensitive to the cis addition of caffeine, ATP, and calcium (data not shown); these results suggest that extensive oxidation caused irreversible loss of channel activity. The RyR channels from cardiac muscle were more susceptible to irreversible inhibition by SH reagents than channels from brain or skeletal muscle, as described further in the text.

Oxidation of RyR channels from brain cortex neurons

View larger version (30K): [in this window] [in a new window]   FIGURE 2   Time course of activation of single RyR-channels from brain cortex neurons (A) or mammalian skeletal muscle (B and C) during sustained incubation with DTDP or thimerosal. The values of channel Po, calculated in successive periods of 1.024 s, are displayed as a function of time. (A) Effects of DTDP on the activity of a single RyR channel from rat brain cortex that displayed the low Po calcium dependence. At the arrow, 50 µM DTDP was added to the cis compartment that contained 30 µM [Ca2+]. Representative current traces of 10-s duration, obtained at 0 mV, are displayed in the inset for the intervals marked I, II, and III above the x axis; the average Po value for each interval is also indicated. The channel opened upward with an amplitude of 3.1 pA. (B) Effects of thimerosal on the activity of a single RyR channel from rabbit skeletal muscle that displayed the low Po calcium dependence. At the arrow, 200 µM thimerosal was added to the cis compartment that contained 30 µM [Ca2+]. Representative current traces of 10-s duration obtained at 0 mV during intervals I, II, and III are displayed in the inset; the average Po values for all periods are given. The channel opened upward with an amplitude of 3.1 pA. (C) Effects of DTDP on the activity of a single RyR channel from rabbit skeletal muscle that displayed the MS calcium dependence. At the arrow, 100 µM DTDP was added to the cis compartment that contained 30 µM [Ca2+]. Representative current traces of 10-s duration obtained at 0 mV during intervals I and II are displayed in the inset; the average Po values for both periods are given. The channel opened upward with an amplitude of 3.2 pA.

Oxidation of RyR channels from mammalian or amphibian skeletal muscle

Oxidation of RyR channels from cardiac muscle

View larger version (28K): [in this window] [in a new window]   FIGURE 3   Time course of the effects of DTDP or thimerosal on single cardiac RyR channels. The values of channel Po, calculated in successive periods of 1.024 s, are displayed as a function of time. (A) Effect of DTDP on a single native RyR channel with the C calcium dependence. At the arrow, 50 µM DTDP was added to the cis compartment that contained 30 µM [Ca2+]. Representative current traces of 10-s duration, obtained at 0 mV, are displayed in the inset for the intervals marked I and II above the x axis; the average Po values for each interval are also indicated. The channel opened upward with an amplitude of 3.0 pA. (B) Effect of thimerosal on a single native RyR channel with the MS calcium dependence. At the arrow, 10 µM thimerosal was added to the cis compartment that contained 500 µM [Ca2+]. Representative current traces of 10-s duration obtained at 0 mV before and after the addition of thimerosal are displayed in the inset; the average Po values for both periods are given. The channel opened upward with an amplitude of 2.9 pA.

Effects of SH oxidation on the Ca²⁺ dependence of RyR channels from brain cortex neurons or skeletal or cardiac muscle

As a next step in the characterization of the effects of oxidation on channel activity, SH oxidation was stopped during the incubation period with DTDP or thimerosal, and the effect of changing cis [Ca²⁺] on the activity of the oxidized channels was studied.

RyR channels from brain cortex neurons

View larger version (13K): [in this window] [in a new window]   FIGURE 4   Effect of cis [Ca2+] on the activity of oxidized RyR channels from brain or muscle. The panels show the changes in Po as a function of cis [Ca2+] for RyR channels from brain cortex neurons (A), rabbit skeletal muscle (B), and rabbit cardiac muscle (C). Values are given as mean ± SEM. The number of channels studied for each calcium dependence (, MS; , C) is given in the text. The segmented lines represent the best fits obtained for the native channels, with the parameters given in Table 1.

View larger version (39K): [in this window] [in a new window]   FIGURE 5   Effect of sequential incubation with DTDP on the calcium dependence of a native neuronal RyR-channel. The three top current records indicate that this native channel displayed the low Po calcium dependence. After 7 min of oxidation with 200 µM DTDP, the reaction was stopped (see text), and the calcium dependence of the channel was characterized as MS, as demonstrated by the middle set of current records. After a second incubation with DTDP, the channel displayed the C behavior, as shown in the lower set of records.

RyR channels from mammalian or amphibian skeletal muscle

View larger version (43K): [in this window] [in a new window]   FIGURE 6   Effect of incubation with DTDP on the calcium dependence of a native rabbit skeletal RyR channel. The three top current records indicate that this native channel displayed the low Po calcium dependence. After incubation with 200 µM DTDP for 12 min, the channel displayed the C behavior, as shown in the lower set of records.

RyR channels from cardiac muscle

View larger version (47K): [in this window] [in a new window]   FIGURE 7   Effect of incubation with DTDP on the calcium dependence of a native cardiac RyR channel. The three top current records indicate that this native channel displayed the MS calcium dependence. After incubation with 10 µM DTDP for 2 min, the channel adopted the C behavior, as shown in the lower set of records.

Reversibility of the effects of SH oxidation

To further test the effects of changing the redox state of the channels on channel activity, we investigated whether SH-reducing agents 1) modified the calcium dependencies of native channels and 2) reversed the changes in channel response to cis [Ca²⁺] produced by oxidation.

Native RyR channels from skeletal muscle

View larger version (27K): [in this window] [in a new window]   FIGURE 8   Effect of SH-reducing agents on RyR-channel activity. (A) Time course of changes in Po after the addition of 20 mM GSH (arrow) to a native RyR channel from frog skeletal muscle in 0.5 mM cis [Ca2+], previously characterized as displaying the C calcium dependence. The values of Po, calculated in successive periods of 1.024 s, are displayed as a function of time. Representative current traces (10 s) obtained from the intervals marked I and II, as well as the average Po values for these entire periods, are displayed in the inset. The channel opened upward with an amplitude of 3.1 pA. (B) Time course of the effect of GSH on the activity of a native rabbit cardiac RyR channel, in 0.5 mM cis [Ca2+], with spontaneous C behavior. The values of Po, calculated in successive periods of 1.024 s, are displayed as a function of time. Representative current records (10 s) obtained from the intervals marked I and II, as well as the average Po values for these entire periods, are displayed in the inset. The channel opened upward with an amplitude of 3.2 pA. (C) Time course of changes in Po after the addition of 10 mM -mercaptoethanol (arrow) in 10 µM [Ca2+] to a neuronal RyR channel, which, after previous treatment with DTDP, had acquired the C calcium dependence. The values of Po, calculated in successive periods of 1.024 s, are displayed as a function of time. Representative current traces (10 s) obtained at 0 mV from the periods marked as I and II, as well as average Po values for these entire periods, are displayed in the inset. The channel opened upward with an amplitude of 3.0 pA.

Reduction of native cardiac RyR channels

Reduction of oxidized neuronal RyR channels

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion Conclusions References

Calcium dependencies of native RyR channels

The results presented in this work show that native RyR channels from mammalian and amphibian skeletal muscle exhibited the same three different calcium dependencies, C, MS, and low P_o, previously described for RyR channels from brain cortex neurons (Marengo et al., 1996). In addition, we found that native RyR channels from cardiac SR vesicles (N = 19) presented only two calcium dependencies, MS and C. These results suggest that cardiac RyR channels either do not present the low P_o behavior, or do so with low frequency (<5.3%).

In our previous work we proposed tentatively that each calcium dependence exhibited by native channels from brain (Marengo et al., 1996) or frog skeletal muscle (Bull and Marengo, 1993) reflected the calcium dependence of each RyR isoform present in these tissues. However, the findings described in this work do not support that proposition. Channels from amphibian skeletal SR, which has two RyR isoforms, presented three different calcium dependencies. Likewise, channels with two or three different calcium dependencies, respectively, were found in mammalian cardiac and skeletal SR vesicles, which have only one RyR isoform. In agreement with these results, additional calcium dependencies that differ from the most commonly reported behaviors have been observed for RyR channels from cardiac muscle (Schiefer et al., 1995) or mammalian skeletal muscle (Coppello et al., 1997). Thus the presence of only one RyR isoform does not define a single calcium dependence for the corresponding RyR channel activity.

The above findings indicate that other factors determine the calcium dependence of the different RyR channels. Previous results (Stuart et al., 1992; Favero et al., 1995; Salama et al., 1992) suggested that SH reagents modify the activation of RyR channels by cytosolic [Ca²⁺]. Accordingly, we studied the effects of SH reagents on channel activity, and we investigated whether the different channel responses to changes in cis [Ca²⁺] were affected by SH reagents.

Channel activity increased in response to treatment with SH-oxidizing reagents

We found that SH-oxidizing reagents significantly increased the activity of the different native RyR channels incorporated in lipid bilayers, as evidenced by a significant increase in P_o. These results confirm and extend to other tissues (brain) and species (frog) the previous findings reported for RyR channels from mammalian skeletal and cardiac muscle (Stoyanovsky et al., 1994; Favero et al., 1995). Moreover, the increase in RyR channel activity was produced stepwise, so that depending on the basal level of activity of the channel, one or two steps of increase in P_o were observed after oxidation. We have also confirmed previous results, indicating that long incubations with SH reagents produced an irreversible inhibition of channel activity (Abramson et al., 1995).

Channel responses to changes in cis [Ca²⁺] depended on the redox state of the RyR channels

The observation that oxidation or reduction of SH residues modified the type of response of the RyR channels to changes in cis [Ca²⁺] is the most significant and novel finding of the present study. In general, regardless of their origin (skeletal or cardiac muscle, brain cortex), extensive oxidation led to either C behavior or to channel inactivation, whereas reduction induced lower channel activity and inhibition by 0.5 mM [Ca²⁺]. These modifications in channel behavior could be effected either by the addition of oxidizing or reducing agents directly to the cis chamber during channel recording, or by the addition of DTT during all steps of membrane isolation. The calcium dependencies obtained after SH oxidations were comparable to those of the native channels. This concordance suggests that SH oxidation produced the same channel states found in native conditions.

The SH groups that are important in determining the calcium dependence of RyR channels are exposed to the cytosolic side of the bilayer, because thimerosal, a water-soluble SH reagent, increased channel P_o only when added to the cis compartment. The most likely target of SH modification are the many cysteine residues present in the large cytoplasmic domain of all the RyR isoforms characterized to date (Takeshima et al., 1989; Nakai et al., 1990; Otsu et al., 1990; Zorzato et al., 1990; Hakamata et al., 1992). Previous studies indicate that RyR channels from mammalian skeletal muscle have several highly reactive SH groups, and that conditions that favor channel closing increase by ~8-fold RyR channel labeling by a fluorescent thiol oxidizing reagent (Liu et al., 1994). These results indicate that these RyR channels have more free SH groups amenable to oxidation in the closed state. The changes in calcium dependence described in this work may involve molecular rearrangement of channel-protein segments produced by oxidation of specific SH residues, which somehow increase the calcium affinity of cytoplasmic activating sites and hinder calcium binding to the inhibiting sites. Thus when all specific SH residues are reduced, the channel would exhibit the low P_o calcium dependence. Partial oxidation of these residues would give rise to the MS calcium dependence. Further oxidation would induce C calcium dependence, and extensive oxidation would induce the irreversible loss of channel activity. According to this view, the fact that cardiac RyR channels did not exhibit low P_o behavior may be attributed to partial oxidation of the putative specific SH residues, even in vesicles isolated with DTT, because the cardiac channels were highly susceptible to oxidation. Alternatively, the RyR-2 isoform may lack the SH residues that, when reduced, give rise to low P_o behavior, because not all SH residues are conserved among isoforms (Takeshima et al., 1989; Nakai et al., 1990; Otsu et al., 1990; Hakamata et al., 1992).

However, as described by Liu et al. (1994), several other SR proteins, including triadin, were labeled with the fluorescent thiol-oxidizing reagent used by these authors. The extent of labeling of these proteins also increased when channels were closed by the addition of RyR channel blockers. We cannot rule out the possibility that RyR channels are incorporated into the planar lipid bilayers with other SR proteins, such as triadin, and that the redox state of these other proteins determines RyR channel calcium dependence. Yet, in the absence of direct experimental demonstration for the incorporation of such a complex into the bilayer, we favor the simpler interpretation that the redox state of the RyR channel protein itself conditions channel response to changes in cis [Ca²⁺]. Experiments with the purified channel protein should help settle this issue.

Physiological and pathological implications of the present results

An important observation of the present work is that SH oxidation increased channel activity as well as the sensitivity to activation by calcium of the RyR channels from the three excitable tissues studied. As a consequence, oxidative stress should enhance calcium-induced calcium release in neurons and muscle. A concomitant increase in free radicals and cytoplasmic [Ca²⁺] has been observed in conditions such as apoptosis (Wood and Youle, 1994), hypoxia (Bonfoco et al., 1995), reperfusion after ischemia (Kaneko et al., 1994), malignant hyperthermia (Duthie and Arthur, 1993), and neurodegenerative diseases (Bonfoco et al., 1995). If free radicals promote the oxidation of SH residues in the RyR channels of excitable cells, calcium-induced calcium release would be enhanced. As a consequence, an increase in cytoplasmic [Ca²⁺] would take place, as observed in these conditions.

	CONCLUSIONS

Top Abstract Introduction Procedures Results Discussion Conclusions References

From the results reported in this work, it is proposed that the redox state of specific SH groups present in the cytoplasmic domain of RyR channels controls their responses to changes in cis [Ca²⁺]. Channels with the low P_o behavior would have these SH groups in the reduced state. The addition of SH-oxidizing reagents would promote the rapid oxidation of these SH residues, producing a change from the low P_o to the MS behavior. In the continuous presence of SH-oxidizing reagents, additional SH groups would react, causing the observed change to C behavior. Because only this second oxidation reaction with DTDP was reversed by SH-reducing reagents, presumably the most highly reactive SH groups involved in the low P_o-to-MS transition undergo irreversible oxidation in vitro.