Transmembrane Redox Sensor of Ryanodine Receptor Complex*

Inositol 1,4,5-trisphosphate receptors (IP3R) and ryanodine receptors (RyR) mediate the release of endoplasmic and sarcoplasmic reticulum (ER/SR) Ca2+stores and regulate Ca2+ entry through voltage-dependent or ligand-gated channels of the plasma membrane. A prominent property of ER/SR Ca2+ channels is exquisite sensitivity to sulfhydryl-modifying reagents. A plausible role for sulfhydryl chemistry in physiologic regulation of Ca2+ release channels and the fidelity of Ca2+release from ER/SR is lacking. This study reveals the existence of a transmembrane redox sensor within the RyR1 channel complex that confers tight regulation of channel activity in response to changes in transmembrane redox potential produced by cytoplasmic and luminal glutathione. A transporter selective for glutathione is co-localized with RyR1 within the SR membrane to maintain local redox potential gradients consistent with redox regulation of ER/SR Ca2+release. Hyperreactive sulfhydryls previously shown to reside within the RyR1 complex (Liu, G., and Pessah, I. N. (1994) J. Biol. Chem. 269, 33028–33034) are an essential biochemical component of a transmembrane redox sensor. Transmembrane redox sensing may represent a fundamental mechanism by which ER/SR Ca2+channels respond to localized changes in transmembrane glutathione redox potential produced by physiologic and pathophysiologic modulators of Ca2+ release from stores.

A change in cytosolic Ca 2ϩ concentration serves as a signal for modulating a wide range of cellular activities (1)(2)(3). A major mechanism for increasing cytosolic Ca 2ϩ includes release of Ca 2ϩ from internal stores (endoplasmic or sarcoplasmic reticulum, ER or SR) 1 via a genetic superfamily of Ca 2ϩ release channels including inositol 1,4,5-trisphosphate receptors (IP 3 R) and ryanodine receptors (RyR) (4 -6). A prominent functional property of all of these channels is exquisite sensitivity to reduction and oxidation by sulfhydryl reagents (7)(8)(9)(10)(11)(12). The functional consequences of sulfhydryl modification of RyRs in-clude phases of activation and inhibition, revealing that multiple classes of sulfhydryl groups residing on Cys residues of all three isoforms of RyR channel complexes are important for native functioning and subject to chemical modification (11,12). However, defining a role for sulfhydryl redox chemistry in RyR function has been controversial since the initial suggestion that sulfhydryl oxidation is a key step in channel activation (13). A plausible physiological role for redox control of ER/SR Ca 2ϩ release channels and its attendant mechanism has remained elusive.
It is known that glutathione (GSH) and glutathione disulfide (GSSG) constitute the major redox buffer system of skeletal muscle and many non-muscle cells (14,15). In the typical mammalian cell, the ratio of [GSH]/ [GSSG] in the cytosol is Ն30:1, thereby maintaining very reduced redox potential (RP) of approximately Ϫ220 mV (16). By contrast, the RP of the ER lumen is significantly more oxidized (approximately Ϫ180 mV) and is maintained with a 3:1 to 1:1 ratio of [GSH]/[GSSG] (16,17). Thus, the typical microsomal membrane within which the RyR and IP 3 R reside is normally subject to a large transmembrane RP difference of 40 -50 mV with the lumen much more oxidized than the cytosol (16,17).
To study redox regulation of RyR channel activity, the bilayer lipid membrane (BLM) preparation affords precise control of the redox state on both the cytoplasmic (cis) and luminal (trans) faces of the reconstituted channel by adjustment of the [GSH]/[GSSG] ratio to form varied redox potentials. In the present work, we provide direct evidence that RyR1 channel activity follows transmembrane redox potential. Chemical labeling studies with CPM indicate previously identified hyperreactive sulfhydryl moieties within the RyR1 complex (18,19) constitute an essential component of a unique transmembrane redox sensor.

EXPERIMENTAL PROCEDURES
Preparation of SR Membranes--Sarcoplasmic reticulum membrane vesicles were prepared from back and hind limb skeletal muscles of New Zealand White rabbits according to the method of Saito et al. (20) with some modifications. During the SR preparation, GSH and GSSG were included in the homogenization buffer, and the glutathione RP was made to Ϫ220 mV, which mimics the typical cytoplasmic RP in vivo. The preparations were stored in 10% sucrose, 10 mM Hepes, pH 7.4, at Ϫ80°C until needed.
GSH and GSSG Stock Solutions-GSH was dissolved in degassed Hepes (20 mM) buffer, and the solution was adjusted to pH 7.0. Aliquots (ϳ0.2 ml) were transferred to vials and sealed after blowing with argon. The vials were stored at Ϫ20°C for no longer than 60 days. Once thawed and opened for use, the vial was discarded. GSSG solution was also made and stored in a similar manner except without degassing and argon protection. Samples of GSH and GSSG were taken from both sides of the BLM chamber at the end of channel recordings to verify that the initial redox potential did not change during the course of the experiments. Redox buffers were found to be stable for at least 1 h.
Transport Measurements by Light Scattering Techniques-Osmotically induced changes in microsomal vesicle size and shape (21) were monitored at 400 nm at a right angle to the incoming light beam, using a fluorimeter (F-2000, Hitachi). A decrease in light scattering reflected vesicular swelling, a consequence of osmotic changes as GSH and GSSG transported into the vesicle lumen. Briefly, SR vesicles (25 g/ml) were equilibrated in a hypotonic medium (5 mM K-PIPES, pH 7.0). The osmotically induced changes in light scattering were measured after the addition of a small volume (Ͻ10% of the total incubation volume) of concentrated and stock solutions of the compounds to be tested. No changes in light scattering were observed when an identical assay solution was diluted with the same volume of stock buffer lacking GSH or GSSH. Sucrose was used to test the selectivity of GSH-and GSSGinduced changes in light scattering because it was shown to be weakly permeable to SR membranes (21). Flufenamic acid, a known anion transport inhibitor (17), was used to test the specificity of GSH or GSSG permeation of junctional SR.
Preparation of GSH and GSSG Extracts from SR Lumen-SR vesicles (100 g) were incubated in 1 ml of solution A (100 mM KCl, 20 mM MOPS, 100 M CaCl 2 ), and an indicated combination of [GSH] and [GSSG] at 37°C for 15 min. The mixtures were then centrifuged at 16,000 ϫ g for 25 min at 4°C. Each supernatant was carefully decanted from its respective pellet, and the latter was rinsed three times with 1 ml of solution A. The pellets were resuspended and homogenized in 100 l of buffer containing 1 N potassium P i buffer, 0.5% CHAPS (w/v), and alamethicin (22) (0.1 mg/mg protein) and then incubated for 30 min at 37°C to permit the release of all vesicular glutathione. Following incubation, 100 l of 5% tricholoracetic acid in redox quenching buffer (20 mM HCl, 5 mM diethylenetriaminepentaacetic acid, 10 mM ascorbic acid) was added into the solubilization mixture and the extract centrifuged at 16,000 ϫ g for 25 min. The supernatant was collected, applied to a Microcon YM-3 (3000 molecular weight cut-off, Millipore) and centrifuged at 16,000 ϫ g for 1 h. The filtrate (ϳ200 l), which represented the total SR luminal glutathione (GSH ϩ GSSG), was analyzed for [GSH]/[GSSG] as described below.
Measurement of GSH and GSSG Content-The GSH and GSSG content of each luminal extract was determined in a manner similar to the method of Senft et al. (23), using the fluorescence probe o-phthaldialdehyde (OPA) with a few modifications. To measure total glutathione (GSH ϩ GSSG), 25 l of each extract was added into 65 l of 1 N potassium P i buffer containing 100 M dithionite to fully reduce GSSG. The reaction mixture was incubated for 1 h at room temperature, followed by addition of 80 l of 0.1 N potassium P i . To quantify the total GSH product produced from the reduction reaction, 20 l of OPA (5 mg/ml) was introduced into the mixture and permitted to incubate at room temperature for 90 min. OPA fluorescence was measured with a fluorescence spectrophotometer (model F-2000, Hitachi) at excitation 365 nm/emission 430 nm. The portion of total glutathione attributable to GSH was analyzed in the manner described for total glutathione without the use of dithionite. To determine the original background content of luminal SR GSH and GSSG before the addition of exogenous glutathione, 100 g of SR was incubated with alamethicin (0.1 mg/mg SR protein) to release the glutathione to the solution bathing the SR vesicles. These values were used as background correction for determining the net exogenous GSH and GSSG transported into SR vesicles.
Single-channel Kinetics in BLM-Reconstitution of RyR1 and recording of channel activity were performed as previously reported (24) with some modifications. RyR1 channels were reconstituted into planar lipid bilayer (5:2 phosphatidylethanolamine:phosphatidylcholine, Northern Lipids Inc., 50 mg/ml in decane) by introducing SR vesicles to the cis chamber. The cis chamber contained 0.7 ml of 250 or 500 mM CsCl, 50 -200 M CaCl 2 ,and 10 mM Hepes, pH 7.4, whereas the trans side (virtually grounded) contained 50 or 100 mM CsCl, 50 -100 M CaCl 2 , and 10 mM Hepes, pH 7.4. Upon the fusion of SR vesicle into bilayer, cis chamber was perfused, and the cis/trans CsCl gradient were reversed. Single-channel activity was measured using a patch clamp amplifier (Dagan 3900) at a holding potential as indicted in each figure. The data were filtered at 1 kHz before being acquired at 10 kHz by a DigiData 1200 (Axon Instruments, Foster City, CA). The data were analyzed using pClamp 6 (Axon) and CDA 1.0 (provided by Dr. G. Liu) without additional filtering. The length of representative current traces was Ն1 s. Average P o was calculated from Ն1 min recording.

RESULTS AND DISCUSSION
Junctional SR membranes from rabbit skeletal muscle prepared in the presence of a [GSH]/[GSSG] redox buffer with a potential of Ϫ220 mV (calculated from the Nernst relationship (16)) resulted in a high percentage (43%) of reconstituted channels (n ϭ 102) exhibiting a characteristic low open probability (P o ) gating mode (Fig. 1A). The remaining channels (57%) exhibited a high P o gating mode. This observation was consistent with those of previous reports (25)(26)(27). Although the molecular details underlying low P o and high P o gating modes has remained unclear, the oxidation of one or more classes of sulfhydryl moieties was implicated in stabilizing these experimentally observed gating modes. Low P o and high P o gating modes represent stable RyR1 channel gating behaviors that have been observed in BLM experiments. These gating modes appeared to depend on the overall redox state of the receptor complex because low P o behavior was promoted by treatment with reducing agents, whereas high P o behavior was promoted by oxidizing agents. A characteristic of these gating behaviors was that once they were achieved, the channels did not require the continued presence of reducing or oxidizing equivalents to maintain their gating modes. This observation strongly suggested that low P o and high P o gating behavior may result from the breaking or forming intra-of inter-subunit disulfide bonds (25). This hypothesis was further supported by the observation that gating behaviors can be interconverted by the subsequent addition of reducing or oxidizing reagent. Although channels exhibiting low and high P o gating modes were found to be tightly responsive to transmembrane RP, the present study focuses primarily on channels exhibiting the low P o gating mode. To test the response of the RyR1 channel to cytosolic oxidation, the channel was challenged with a highly oxidized RP of Ϫ180 mV generated by adding [GSH]/[GSSG] at a ratio of 3:1 (total [glutathione] ϭ 4 mM) to the cis chamber of the BLM (Fig. 1B). Surprisingly, the channel showed a negligible change in gating activity (P o ϭ 0.011). However, immediately after the addition of 3:1 [GSH]/[GSSG] to generate the same redox potential of Ϫ180 mV on the luminal (trans) side of the channel, channel P o increased 13-fold. (mean P o rose from 0.011 to 0.138 based on 2-4 continuous minutes of record before and after setting the trans redox potential; Fig. 1C). If channel activation were simply the result of inclusion of an oxidizing potential of Ϫ180 mV on the luminal face of the channel, then the higher channel activity should persist after removal cytoplasmic RP. However, removal of the cytoplasmic RP by extensive perfusion of the cis chamber with an identical solution lacking glutathi- one resulted in a 13-fold decrease in channel P o (mean P o from 0.138 to 0.011; Fig. 1D). Responsiveness to transmembrane redox potential has been observed in 83 of 106 separate reconstitution experiments from junctional SR prepared with or without redox buffering in the initial steps (including channels exhibiting both low and high P o gating modes) and appears to represent a common feature of RyR1 channel regulation.
RyR1 channels responded to local changes in transmembrane RP irrespective of the absolute concentration of glutathione used in the buffer. Both the cytoplasmic and luminal sides of the membrane were set to Ϫ180 mV using a variety of [GSH] and [GSSG] ranging from 0.1 to 4 mM (Fig. 2, top panel, conditions a-d). RyR1 channels exhibited P o values ranging from 0.014 to 0.084 in the presence of 10 M cis Ca 2ϩ and undefined RP (no GSH/GSSG). RyR1 channels responded with a 7-8-fold enhancement in gating activity once a symmetrical Ϫ180 mV transmembrane RP was established (Fig. 2, lower panel). In n ϭ 19 separate reconstitution experiments, channel activity was enhanced to the same degree (7-8-fold of the respective control activity in the absence of a defined transmembrane RP) and was independent of the absolute [GSH] and [GSSG] used to obtain a symmetrical Ϫ180 mV RP. These results suggest that RyR1 channel activity responds to changing transmembrane RP and may modify sensitivity to cytoplasmic Ca 2ϩ . Thus, one unique physiologic role of redox sensing afforded to ER/SR Ca 2ϩ channels is the continuous alignment of Ca 2ϩ -induced Ca 2ϩ release gain with small changes of transmembrane redox potential. High concentrations of either GSH or GSSG (Ն2 mM) added singly to the cytoplasmic side of the RyR1 channel resulted in persistent changes in channel gating behavior (data not shown) as previously reported (28). These effects are likely due to the extreme reduction and oxidation potentials that nonselectively alter protein thiols/disulfides and adversely impact a physiologically relevant transmembrane redox sensor (11,12,28).
In healthy non-muscle cells, the ER transmembrane RP gradient is maintained by one or more transporters that facilitate diffusion of either GSH or GSSG and the process is blocked by flufenamic acid (17). To establish the physiological relevance of a trans SR membrane redox sensor within the RyR1 complex, we determined whether or not junctional SR possesses a transport mechanism necessary to generate a transmembrane redox gradient. First, light scattering measured spectrally shows that transport of GSH or GSSG (10 mM) from the extravesicular space into the vesicle lumen (Fig. 3A). The rate of decline in fluorescence represents the uptake of GSH or GSSG into the SR vesicles (thereby decreasing light scattering). Based on the initial rate, junctional SR vesicles transported GSH ϳ5 times faster than GSSG, consistent with recent measurements performed with liver microsomes (17). Despite the higher initial rates of GSH over GSSG, the steady state capacity of SR vesicles for GSH was only 2-fold higher than GSSG within 10 min (⌬F ϭ 40 versus 80; Fig. 3A). Transport of glutathione is selective because the addition of equimolar sucrose to the transport medium produces only a very small change in light scattering in the same time frame as previously shown (21). Flufenamic acid is a known blocker of glutathione transport across ER membrane vesicles prepared from liver (17). Likewise, movement of GSH and GSSG across junctional SR was fully inhibited by 1 mM flufenamic acid (Fig. 3A). Second, we measured the transmembrane GSH and GSSG ratio at steady state (10 min) using o-phthaldialdehyde as a quantitative indicator (23). This method permits direct quantitative analysis of total luminal glutathione (GSH ϩ GSSG) and GSH. In this manner, the ratio of luminal GSH/GSSG can be measured in the presence of varying extravesicular RP Fig. 3B shows several ratios of [GSH]/[GSSG] added to the SR buffer to give extravesicular RP ranging from Ϫ231 mV to Ϫ180 mV. After a 10-min equilibration, the vesicles were extracted and the rela- , which causes vesicle shrinkage. The initial increase was followed immediately by a decrease in light scattering, which reflected vesicular swelling because of movement of GSH or GSSG from the extravesicular space into the vesicle lumen. For the trace labeled "F.A.," the SR vesicles were incubated with 1 mM flufenamic acid for 10 min prior to adding 10 mM GSH or GSSG. As a negative control for the specificity of transportable GSH/GSSG for the vesicles, 10 mM sucrose was added in the same buffer instead of glutathione. This experiment was repeated five times on three SR preparations with the same results. In B, measurement of the transmembrane GSH/GSSG ratio at steady state was performed using OPA as a quantitative indicator (23)  tive concentration of GSH and GSSG inside the vesicle lumen determined using OPA as described under "Experimental Procedures." [GSH] is determined directly by adding OPA, whereas total [glutathione] ([GSH] ϩ [GSSG]) was determined after reducing the sample with dithionite (23). Regardless of the extravesicular redox potential set experimentally, the luminal ratio of [GSH]/[GSSG] converged to 3:1 (Fig. 3B, fifth  column). These results reveal for the first time the ability of SR lumen to clamp the ratio of its [GSH]/[GSSG] within narrow limits, despite the presence varying cytoplasmic RP. A 3:1 ratio of [GSH]/[GSSG] is consistent with a significantly more oxidized microsomal lumen, as demonstrated previously (16,17). Taken together, the present results reveal that junctional SR membranes, such as ER membranes from non-muscle origin (17), possess a selective transporter for GSH and GSSG. Although a common feature of this microsomal transporter is a preference for GSH over GSSG (based on initial rates), steady state analysis reveals the ability of the ER/SR lumen to favor a 3:1 ratio of GSH/GSSG regardless of the cytosolic RP. A 3:1 GSH/GSSG is consistent with the observation that healthy cells maintain an oxidized luminal potential (ϳϪ160 to Ϫ180 mV) relative to the cytosol (ϳϪ220 mV) (16). How the ER/SR lumen maintains an oxidized potential despite the preference for transport of GSH is unclear. One possibility is that GSH is oxidized to GSSG within the ER/SR lumen and that the latter is preferentially retained (17). In support of this hypothesis, there is evidence that GSSG can be formed locally within the ER lumen, although the mechanism(s) remain obscure (17,29,30). The existence of a GSH/GSSG transporter co-localized with RyR1 within junctional SR membranes would be expected if transmembrane redox sensing is a significant physiologic modulator of RyR1 function.
The ability of the RyR1 complex to respond to transmembrane RP and the apparent ability of SR/ER to create a transmembrane redox gradient raises the possibility that RyR1 channel activity could be actively regulated by subtle localized changes in transmembrane RP Fig. 4 show the typical response of RyR1 channels to small changes in transmembrane RP In the absence of a defined RP, low P o channels exhibited infrequent gating activity even when cis Ca 2ϩ was present in an optimal range (30 -100 M) Ca 2ϩ (Fig. 4, A and C; top traces). Once the transmembrane RP was buffered to Ϫ220/Ϫ180 mV (cis/trans), the P o increased 2.7-fold from 0.019 to 0.052 (Fig.  4A, middle trace). However, upon adjusting cis to a more oxidized Ϫ180 mV, the channel P o immediately increased an additional 2.1-fold (Fig. 4A, bottom trace) for a total enhancement of nearly 6-fold. In this regard, RyR1 channels were under tight control by transmembrane RP, regardless of their gating mode (low versus high P o ). Fig. 4B shows an example of a high P o channel closely following as little as ϩ10 -20 mV incremental jumps in RP on the cis side relative to a fixed luminal RP of Ϫ180 mV.
Modulation of channel P o is independent of the direction in which the transmembrane RP is altered (reduction 3 oxidation or oxidation 3 reduction). For example, once the transmembrane RP was set to symmetrical Ϫ180 mV, the P o increased 4.2-fold compared with the initial P o under undefined RP (Fig.  4C, compare top and middle traces). Subsequent reduction of cis to Ϫ210 mV reduced P o 2-fold. These results reveal that RyR1 channels tightly follow the cis/trans RP in both oxidizing and reducing directions. Similarly, the P o of low P o RyR1 channels closely followed as little as Ϫ10 mV incremental jumps in RP on the cis side relative to a fixed trans of Ϫ180 mV in transmembrane RP (Fig. 4D). In separate experiments, the cytoplasmic RP was kept fixed at Ϫ160 mV and the luminal RP incrementally titrated from Ϫ160 to Ϫ190 mV in Ϫ5 to Ϫ10 mV increments (n ϭ 4, data not shown). As the RP difference across the BLM increased, channel P o increased accordingly (227% enhanced P o at Ϫ160 mV cis versus Ϫ190 mV trans). These results reveal that once RP is set on both the cis and trans sides, RyR1 channel activity tightly follows incremental changes in RP on either side of the membrane. Tight regulation of gating activity by transmembrane RP may be likened to the tight voltage regulation observed with voltage-gated ion channels, therefore representing a fundamentally new mode in channel regulation.
The response of the RyR1 channels to a physiologically relevant redox difference on cytoplasmic and luminal sides of the membrane indicates that an essential component of the redox sensing mechanism may span the ER/SR membrane and may involve the redox state of essential Cys moieties. The redox difference across the ER/SR membrane may therefore provide the driving force for electron transfer reactions that modify the gating kinetics in response to physiological ligands. From what has been known for many redox-sensitive biochemical processes, including targets of transcription factors, antioxidants, cytokines, as well as ion channels/transporters, cell growthrelated genes, kinases, phosphatases, etc., electron flow through CH 2 -SH moieties of conserved Cys residues within proteins account for their redox-sensing properties (14). An essential component of RyR1 redox sensing may involve the transfer of electrons among two or more closely spaced Cys moieties within the channel complex. The RyR1 complex has FIG. 4. RyR1 channel activity closely follows the transmembrane redox potential. Following channel reconstitution into BLM, the activity of RyR1 channels was recorded before and after the RP was incrementally varied. A, channel activity was monitored during a stepwise change in transmembrane RP, from a reduced cis of Ϫ220 mV to a more oxidized cis (Ϫ180 mV). In this experiment, the trans RP remained constant at Ϫ180 mV. B, a separate high P o channel obtained from the same SR preparation showing response to stepwise changes in ϩ10 -20 mV increments (trans was constant at Ϫ180 mV). C, a typical response of a low P o channel to changes in transmembrane RP in the opposite direction from an oxidized potential of Ϫ180 mV to a more reduced potential of Ϫ210 mV. D, a separate low P o channel obtained from the same SR preparation showing the response to stepwise changes in Ϫ10 mV increments (trans was constant at Ϫ180 mV). RP been shown to possess a small number of highly reactive sulfhydryl moieties in which chemical reactivity appears to dramatically increase with decreasing P o by the presence of mM Mg 2ϩ and/or nM Ca 2ϩ (18,19,24). Consistent with previous reports (18,19), the fast labeling kinetics of these hyperreactive thiols by the fluorescent maleimide CPM (7-diethylamino-3-(4Ј-maleimidylphenyl)-4-methylcoumarin) took place under conditions that favor RyR1 channel closure (e.g. 10 mM Mg 2ϩ ). The hyperreactive SR thiols rapidly formed thioether adducts with 0.2-1 pmol of CPM/g of SR protein (t1 ⁄2 of 11.67-15.37 s; n ϭ 4, data not shown). To determine whether chemical modification of hyperreactive Cys moieties alter gating behavior, channels were arylated with CPM under conditions known to promote selective labeling of only hyperreactive Cys residues and the functional consequences determined in the BLM. Fig.  5, A and C, reveals that short-term (Յ5 min) exposure of a channel to 40 and 50 nM CPM, respectively, caused negligible changes of P o (Fig. 5, A and C) and mean open dwell time (Fig.  5B) (n ϭ 8). Reconstitution of RyR1 channel from such specifically CPM-relabeled SR vesicles also demonstrated unchanged channel gating behavior (n ϭ 13, data not shown). However, prolonged incubation of CPM produced additional nonspecific interactions with less reactive Cys residues and thus led to a time-dependent irreversible inactivation of the channel ( Fig.  5C; n ϭ 3). These results reveal that formation of thioether adducts with the most reactive (hyperreactive) Cys residues within the RyR1 complex does not alter overt aspects of channel gating behavior.
Are the hyperreactive thiols associated with the RyR1 complex an essential component of the transmembrane redox sensor? Fig. 6 addresses this important question by comparing the responses of two separate channels to parallel changes in transmembrane RP without and with chemical modification of the hyperreactive Cys residues with CPM. Both channels responded strongly to the instillation of a symmetric Ϫ180 mV transmembrane RP and were negatively regulated by a Ϫ220/ Ϫ180 mV cis/trans gradient (compare Fig. 6, A and B, traces  2-4). The channel in panel B, after exposure to 20 nM CPM for 120 s, failed to respond to a Ϫ180/Ϫ180 mV symmetric cis/ trans redox gradient (Fig. 6B, compare traces 6 and 7), whereas the control channel maintained its redox-sensing properties (Fig. 6A, compare traces 6 and 7). Despite the loss of transmembrane redox sensing, the CPM-modified channel, as predicted, maintained unchanged gating behavior and sensitivity to 10 M ryanodine (Fig. 6B, short trace, labeled Ry). In separate experiments, RyR1 channels had been reconstituted in BLM after hyperreactive thiols were specifically arylated by CPM under the conditions described above. It was found that these pretreated channels gated normally but lacked sensitivity to transmembrane RP changes (n ϭ 3, data not shown).
Our findings reveal the existence of a transmembrane redox sensor within a microsomal Ca 2ϩ channel and represent the first direct evidence linking transmembrane RP with a specific biochemical mechanism regulating microsomal Ca 2ϩ transport. Small changes in localized RP appear to dramatically influence the RyR1 activity regardless of cytoplasmic [Ca 2ϩ ]. Redox sensing is therefore likely to have significant regulatory impact on Ca 2ϩ -induced Ca 2ϩ release. Considering the broad distribution of IP 3 R and RyR receptors, redox control of microsomal Ca 2ϩ release channels may represent a fundamental mechanism by which mammalian cells regulate Ca 2ϩ signaling and homeostasis in response to localized changes in redox potential. Hyperreactive thiols within the channel complex are an essential component of a transmembrane redox sensor, which is likely to contribute important regulatory functions during normal intracellular signaling (mediated for example by nitric oxide) (31-  3 and 4). CMP (40 nM) was introduced into the cis chamber for 5 min (traces 5 and 6) and then removed by extensive perfusion of the chamber (traces 7 and 8). P o was analyzed and is denoted above the representative traces. Mean open time and the curve fit ( 1 and 2 , best fit with double exponential) of the channel before and after CPM treatment are displayed in B. In C, channel activity was initially recorded in the presence of 2 mM cis Mg 2ϩ before and after a 40-s exposure to 40 nM CPM (ϩCPM). The cis chamber was extensively perfused with buffer containing 100 M Ca 2ϩ but lacking Mg 2ϩ and CPM. Channel gating was continuously followed after a second exposure to 50 nM CPM (ϩϩCPM) on the cis side for Ͼ30 min. Prolonged arylation of protein thiols with CPM resulted in an irreversible inhibition of channel activity. . In the last step, 10 M ryanodine (Ry) was included in cis (short trace). This is a representative of three experiments with similar results. 34) and may be involved in mediating changes in Ca 2ϩ signaling during oxidative stress (24).