Sulfhydryl oxidation induces rapid calcium release from sarcoplasmic reticulum vesicles.

Micromolar concentrations of cupric ion (Cu2+) and mercaptans such as cysteine, cysteamine, and homocysteine trigger large and rapid Ca2+ release from skeletal muscle sarcoplasmic reticulum (SR) vesicles. At the concentrations used, Cu2+ alone does not induce Ca2+ release nor does cysteine alone; both are required to induce Ca2+ release from SR. Cu2+ is known to catalyze the autooxidation of cysteine to its disulfide form cystine; Cu2+/mercaptan-induced Ca2+ release appears to be caused by Cu2+-catalyzed formation of a mixed disulfide between the exogenous mercaptan and a critical sulfhydryl on a transmembrane protein. In the oxidized state the SR is highly permeable to Ca2+. Supporting evidence for this interpretation is as follows. The order of Ca2+-releasing reactivity of the mercaptans is the same as the order in which these compounds undergo oxidation to disulfide forms in the presence of Cu2+. Ca2+ efflux induced by cysteine and Cu2+ can be reversed by the addition of the disulfide reducing agent dithiothreitol. Hypochlorous acid and plumbagin, both potential sulfhydryl oxidants, induce rapid Ca2+ efflux from SR vesicles; in addition, Cu2+, which catalyzes H2O2 oxidation of cysteine, enhances H2O2-induced release. Oxidation-induced Ca2+ release from SR can be partially reversed or blocked by ruthenium red or the local anesthetics procaine and tetracaine. The Ca2+ efflux rates are strongly Mg2+ dependent and are significantly higher in heavy SR than in light SR. These data suggest that the Ca2+ efflux thus induced is via the "Ca2+ release channel" and that the oxidation state of a critical sulfhydryl group on this protein may be the principal means by which the Ca2+ permeability of the SR is regulated in vivo.

Micromolar concentrations of cupric ion (Cu2+) and mercaptans such as cysteine, cysteamine, and homocysteine trigger large and rapid Ca2+ release from skeletal muscle sarcoplasmic reticulum (SR) vesicles. At the concentrations used, Cu2+ alone does not induce Ca2+ release nor does cysteine alone; both are required to induce Ca2+ release from SR. Cu2+ is known to catalyze the autooxidation of cysteine to its disulfide form cystine; Cu2+/mercaptan-induced Ca2+ release appears to be caused by Cu2+-catalyzed formation of a mixed disulfide between the exogenous mercaptan and a critical sulfhydryl on a transmembrane protein. In the oxidized state the SR is highly permeable to Ca2+. Supporting evidence for this interpretation is as follows. 1) The order of Ca2+-releasing reactivity of the mercaptans is the same as the order in which these compounds undergo oxidation to disulfide forms in the presence of Cuz+. 2) Ca2+ efflux induced by cysteine and Cu2+ can be reversed by the addition of the disulfide reducing agent dithiothreitol.
3) Hypochlorous acid and plumbagin, both potential sulfhydryl oxidants, induce rapid Ca2+ efflux from SR vesicles; in addition, Cu2+, which catalyzes H202 oxidation of cysteine, enhances H202-induced release. Oxidation-induced Ca2+ release from SR can be partially reversed or blocked by ruthenium red or the local anesthetics procaine and tetracaine. The Ca2+ efflux rates are strongly Mg2+ dependent and are significantly higher in heavy SR than in light SR. These data suggest that the Ca2+ efflux thus induced is via the "CaZ+ release channel" and that the oxidation state of a critical sulfhydryl group on this protein may be the principal means by which the Ca2+ permeability of the SR is regulated in vivo.
The sarcoplasmic reticulum (SR)' is known to play a major role in the regulation of intracellular Ca2+ concentration in muscle and thereby in the generation of force. In skeletal, and in some striated muscles, the rapid movement of Ca2+ into the cytosol has been attributed mainly to efflux of Ca2+ from the SR triggered by an action potential across the T-tubule.
The removal of Caz+ from the cytosol, leading to relaxation, is via a Ca2+ pump which is integral to the SR membrane (1).
However, despite extensive studies both in vitro and in uiuo, the underlying mechanism responsible for Ca2+ release from the SR in skeletal and striated muscle is largely unknown.
Several hypotheses have been put forth in attempts to link the action potential's arrival at the T-tubule to Ca2+ release from the SR: 1) Ca2+-induced Ca2+ release (2-5); 2) depolarization of the SR membrane (6,7); 3) changes in pH (8); 4) voltage-dependent charge displacement within the T-tubular membrane (9), which may be responsible for the movement of feet (10) or bridges (11) at the triadic junction spanning the T-tubule/SR gap; and 5 ) Ca2+ release induced by inositol 1,4,5-triphosphate (12, 13). The issue is not resolved and the physiological significance of the methods used to trigger Ca2+ release from SR remains controversial. The proposed mechanisms have been reviewed recently by A. Martonosi (14).
The theories mentioned above have one thing in common.
Efflux of Ca2+ from the SR is presumed to be mediated by a Ca2+ channel distinct from the Ca2+ pump (Ca2+,Mg2"ATPase). In most studies, Ca2+ release has been described phenomenologically in terms of the agents which induce the changes in Ca2+ permeability, with little or no discussion on a molecular level as to the mechanisms by which these agents regulate the opening and closing of the "Ca2+-release channel." We have previously reported that micromolar concentrations of heavy metal ions ( i e . Cu2+, H$+, Ag+, Cd2+, or Zn2+) trigger the rapid release of Ca2+ by apparently binding to a sulfhydryl group on an integral membrane protein of S R vesicles (15); the potency of these heavy metals to induce Ca2+ release from SR was found to be similar to their relative binding affinities to sulfhydryl groups. Upon actively accumulating Ca2+ in the presence of ATP, Ag+ (5-10 pM) induces rapid Ca2+ release (-58 nmol of Ca2+/mg of protein/s) from heavy SR vesicles (16). Ag+-induced release was examined as a function of pH, M$+, and ionic strength; in light, heavy, and intermediate SR; and in the presence of known blockers of Ca2+ efflux. The data strongly suggest that Ag+ acts at the physiological site of Ca2+ release (16). Other investigators have reported that less reactive sulfhydryl reagents like organic mercurials (50-100 p M ) also induce ca2+ release from SR (17,18).
Also, we (15) and other investigators (19) have shown that cupric phenanthroline, which is known t o catalyze the air oxidation of sulfhydryls to disulfides (ZO), can induce Ca2+ release from the SR, though it has been suggested (19) that the mode of action of cupric phenanthroline may involve cross-linking of the Ca2+,Mg2"ATPase and not the oxidation of sulfhydryls on a separate Ca2+ channel, as we have suggested (15).
Clearly, heavy metals do not play a role in physiological Ca2+ release from SR. However, given that they seem to strongly interact with the Ca2+ release channel, they are useful as probes of the functions of this channel and especially the role, if any, of this critical sulfhydryl in the regulation of the Ca2+ permeability of the SR in uiuo. In the present report we show that cysteine and other biologically common mercaptans, in the presence of Cu2+, induce large and rapid increases in the Ca2+ permeability of the SR. Cu2+ is known to catalyze the oxidation of cysteine to its disulfide form, cystine (21,22). We have found that the relative potency of the mercaptans to cause release (in the presence of Cu2+) parallels the rate at which these compounds undergo Cu2+-catalyzed oxidation to disulfides. Hypochlorous acid (23) and plumbagin (24) permeability of the SR induced by sulfhydryl oxidation may be 1) coincidental with no physiological significance or 2) an indication that sulfhydryl oxidation is the chemical reaction which links the depolarization of the T-tubule membrane with Ca2+ release from the SR or 3) an indication that this critical sulfhydryl is involved in regulation of the Ca2+ channel though its normal function does not involve undergoing oxidation. A preliminary report on the work presented here has appeared in abstract form (27).

MATERIALS AND METHODS
Preparation of SR Vesicles-Rabbit skeletal sarcoplasmic reticulum vesicles were prepared according to the method of MacLennan (28). Light and heavy SR (LSR and HSR) vesicles were prepared by further fractionation of the crude SR preparation on a sucrose step gradient as described previously (16 = 15 or 20 pM as indicated), and the differential absorption changes of the dye were continuously moni-tored. ATP (0.5 mM) was added to initiate active Ca2+ uptake, and the time course of Ca2+ uptake was recorded until Ca" loading was completed (as judged by a leveling-off of the dye signal). Ca2+ release was induced by sequentially adding various concentrations of cysteine (or other mercaptans) then Cu2+ or by adding a sulfhydryl oxidizing agent. The maximum efflux rates were calculated by determining the slope of the efflux curve at its steepest phase. The Ca2+ efflux rate is expressed in units of nmol of Ca2+ effluxed per s per mg of SR protein or as a rate constant in s-I.
Effects of pH and ionic Strength-Cu'+/cysteine-induced Ca2+ release was measured as a function of pH and ionic strength using passively loaded SR vesicles. For pH-dependent Ca2+ efflux, SR vesicles were Ca2+ loaded by incubating SR vesicles (10 mg of protein/ ml) in a medium containing (in mM): 100 KCI, 0.8 MgCI,, 0.8 CaCI2, 10 HEPES, and 10 PIPES at pH values ranging from 6.75 to 8.0. For measurements of Ca2+ efflux as a function of ionic strength the vesicles were suspended at 10 mg of protein/ml in solutions containing (in mM): 20 HEPES, 1 CaCI,, 1 MgC12, pH 7.25, and various concentrations of KC1 and sucrose to obtain a final osmolarity of approximately 225 mOSM. In both studies, the vesicles were incubated on ice for 12-18 h to allow complete equilibration of free Ca" concentration across the SR membrane. The passively loaded SR vesicles were rapidly diluted by a factor of 40 in similar but Ca2+-free media containing 0.1 mM arsenazo I11 at room temperature. The differential absorption of the Ca2+ indicator was continuously monitored to follow the passive efflux of Ca2+ from the lumen of the SR (at 1 mM Ca2+) to the extravesicular medium (at 25-35 pM Ca2+). Sequential additions of cysteine and then CuCI2 initiated a rapid rise (under usual conditions) of the Ca" dye signal (Ca" efflux) which gradually leveled off after a period of time. Upon completion of the experiment, the Ca2+ ionophore A23187 was added (0.5 pg/ml) to abolish any residual Ca2+ gradient across the SR membrane and thus to determine the total internal Caz+ content of the vesicles. At this point, a known amount of Ca" was added in order to provide an internal calibration of the Ca2+ dye response for that particular experiment. The rate constant of efflux was calculated from the rate of Ca2+ efflux (in nmol of Caz+/ mg of protein/s) divided by the total internal Ca2+ loading of the vesicles (in nmol of Ca2+/mg of protein); percentage of Caz+ effluxed was calculated by dividing the amount of Ca2+ effluxed due to Cu2+/ cysteine by the total internal Ca2+ loading of the vesicles. Points plotted are the mean f S.E. of four runs.
Measurement of Sulfhydryl Oxidation Rates with DTNB-The reagent DTNB (or Ellman's reagent) has been used extensively to quantitatively determine sulfhydryl concentration at pH above 8.0 (32). DTNB provides a convenient optical assay for studying sulfhydryl group oxidation since it does not interact with disulfides and because it exhibits a large change in absorption at 410 nm upon reacting with sulfhydryls. Rates of sulfhydryl oxidation of cysteine, cysteamine and homocysteine (at room temperature) were measured in the presence of Cu2+ using DTNB with the following protocol. An aliquot of a solution (0.9 ml) containing 100 mM KC1, 5 mM HEPES, and either 1 or 10 mM MgC12, pH 7.0 (Solution A) was mixed with 0.1 ml of a solution B containing 1 mM DTNB dissolved in 120 mM Tris-HC1, pH 8.5. The final pH of the mixture (A + B) was 8.1. The absorbance at 410 nm of this mixture provided the "reference" baseline absorption against which other mixtures were compared. The sulfhydryl compound (30 pM) to be tested was added to 0.9 ml of solution A plus 2,4, or 6 p~ Cu". At chosen time intervals, 0.1 ml of solution B was added, and the absorption of the new mixture was recorded. The concentration of nonoxidized sulfhydryl groups left in the medium was computed by comparing the absorption of the reaction mixture with that of a control reaction mixture which did not contain Cuz+. The resulting plot of remaining free " S H groups versus time was linear, as expected (this reaction is zero order with respect to [cysteine] (22)); the rate of sulfhydryl oxidation was computed from the slope of this line.
Reagents-The buffers HEPES and PIPES and the sulfhydvl reducing agent dithiothreitol ( D m ) were purchased from Research Organics (Cincinnati, OH). Imidazole was purchased from Aldrich.

Hydrogen peroxide ( H z~z ) ,
MgCl,, and CaCI2 were purchased from J. T. Baker Chemical CO. ATP and the Ca2+ ionophore A23187 were purchased from Behring Diagnostics. Vacuum-distilled hypochlorous acid (HOC1) was a gift of Dr. J. Michael Albrich of the Oregon Graduate Center. The enzymes superoxide dismutase and catalase and all other reagents were purchased from Sigma.

" S H Oxidation Induces
Ca2' Release from SR Fig. 1, the differential absorption changes ( A A = A675 -Am) of arsenazo I11 were used to monitor ATP-dependent Ca2+ uptake followed by Cu2+/cysteine-induced Ca2+ release from SR vesicles. Two equal additions of Ca2+ to the SR reaction mixture produced equal changes in AA verifying the linearity of the dye response in the medium. An addition of ATP initiated active Ca2+ uptake by SR vesicles as inferred by the decrease in AA. Sequential additions of Cu2+ (2 PM) and then cysteine (10 PM) induced a partial but rapid release of the accumulated Ca2+, recorded by a rapid rise in AA. Reversing the order of cysteine and Cu2+ additions did not alter the rate or extent of Ca2+ release. However, the addition of either reagent alone or of a premixed solution of Cu2+ and cysteine to Ca2+-loaded SR vesicles did not induce Ca2+ release. Indeed, after a few minutes a mixture of Cu2+ and cysteine in these proportions contains little or no reduced cysteine because of the Cu2+-catalyzed oxidation of cysteine to form cystine (21,22). Addition of oxidized cysteine (cystine) did not induce Ca2+ release from SR (data not shown). Upon completion of Ca2+ efflux (indicated by a leveling-off of the absorption changes of the Ca2+ indicator), subsequent addition of 2 P M ruthenium red resulted in the active reuptake of the released Ca2+. Alternatively, ruthenium red added at the beginning of the experiment or just after ATP-dependent Ca2+ uptake blocked cysteine-induced Ca2+ release (not shown). Similarly, other inhibitors of Ca2+ release like procaine (10 mM) or tetracaine (1 mM) blocked or reversed cysteine-induced Ca2+ release (not shown). At the end of each experimental run, the Ca2+ ionophore A23187 was added to release all the Ca2+ stored in the lumen of the SR and to verify that the total Ca2+ released (after A23187) was equal to the total Ca2+ taken up by the vesicles. Under the present experimental conditions, the total Ca2+ actively sequestered by the SR vesicles was consistently in the range of 100 f 10 nmol of Caz+/mg of protein.

Sulfhydryl Oxidation with Mercaptans Induces Ca2+ Release from SR Vesicles-In
The experiment shown in Fig. 1 was repeated with cysteine Two aliquots of 10 p~ CaC1, were then added, and the difference in absorbance between 675 and 685 nm was monitored as a measure of external [Ca2+]. Uptake was initiated by addition of 0.5 mM ATP (free I M P ] -0.6 mM). When uptake was judged to be complete, small amounts of Cu2+ and cysteine were added sequentially, triggering Ca2+ release. Ca2+ release rates were calculated from the steepest slope of the release curve. Addition of 2 PM ruthenium red (RR) caused immediate reuptake of 90% of the released Ca2+.
Addition of the Ca2+ ionophore A23187 caused complete release of the remaining internal Ca*+. and the structurally related mercaptans cysteamine and homocysteine. In Fig. 2, the Cu2+ concentration was kept constant at 2 p M while the concentrations of cysteine, cysteamine, and homocysteine were varied over a wide range of values. For each experimental run, the rate of Ca2+ efflux induced by the addition of Cu2+ and the mercaptan was calculated from the maximum slope of the efflux trace, and the rates of Ca2+ efflux were plotted as a function of mercaptan concentration. The relationship between rate of efflux and mercaptan concentration exhibits several interesting features. 1) At its optimal concentration, cysteine is more reactive than cysteamine, and both are much more effective in triggering release than homocysteine. (Note the change in scale between Fig. 2, A and B.) 2) Ca2' efflux rates first increase with increasing mercaptan concentration, reaching a maximum rate, then decrease with further increase in mercaptan concentration. 3) Cu2+/mercaptan-induced Ca2+ release is inhibited by excess mercaptan (-10-fold or more over [Cu2+]). Thus, the mercaptans act as both agonists (at low concentrations) and as inhibitors (at high concentrations) of Ca2+ release. These features are consistent with Cu2+-catalyzed formation of mixed disulfide bonds between the thiol group on the added cysteine and a critical thiol group on the Ca2+ channel. For fixed protein and Cu2+ concentrations, at low cysteine concentrations the ratio of cysteine:Cu2+:channe1 appears to favor formation of mixed disulfides, while at higher cysteine concentrations the relatively small amount of Cu2+ spends more of its time catalyzing the formation of cystine and less time catalyzing mixed disulfide formation. On the other hand, the Cu2+-catalyzed oxidation of cysteine is known to be approximately first order with respect to Cu2+. Assuming that the mechanism of mixed disulfide formation is the same as that of cystine formation, a plot of Ca2+ efflux rate uersw [Cu"] should be roughly linear. In Fig. 3 this prediction is borne out; Ca2+ efflux rates uersus [Cu"] are plotted at two fixed concentrations of cysteine, 5 and 20 PM; the protein concentration was 0.25 mg/ml in each case. The curve is linear over a range of Cu2+ concentration from zero up to equimolarity with the added cysteine. The rate at which mercaptans autooxidize in the presence of Cu2+ parallels their ability to induce Ca2+ release from SR, as shown in Tables I and 11. This is a noteworthy observation, but we emphasize that in the first case (autooxidation of mercaptans, Table I), the reaction is between two identical molecules, while in the second (Ca2+ efflux induced by Cu2+ and mercaptan) Ca2+ efflux rates are used to assess relative rates of reaction between dissimilar molecules. For Table I, the rate of sulfhydryl oxidation of cysteine, cysteamine and homocysteine was measured with DTNB as described above. Cysteine and cysteamine self-oxidized rapidly compared to homocysteine, while glutathione oxidation (not shown) was too slow to detect over the time scale of these experiments. As shown in Table 11, glutathione is also ineffective at inducing Ca2+ release when in the presence of Cu2+, while cysteine, Cysteamine, and homocysteine are effective and to degrees comparable to their autooxidation reactivity. The efflux experiments summarized in Table I1 were performed essentially in the same manner as those of Figs. 1 and 2, the exception being that mercaptans other than cysteine were added at the concentrations indicated. The chemical reactivity of sulfhydryl groups is known to be enhanced by the close proximity of an amine; thus, it is not surprising that cysteine and cysteamine are more reactive than homocysteine (homocysteine's hydrocarbon chain is one carbon longer than those of cysteine and cysteamine). Penicillamine's sulfhydryl is attached to a tertiary carbon, which is also known to greatly When uptake was complete, 5 or 20 WM cysteine was added, followed by various amounts of CuC1,. Release rates were calculated from the steepest portion of the efflux curve. reduce reactivity (33). Consistent with sulfhydryl oxidationinduced Ca2+ release, the oxidized forms of cysteine and cysteamine were not effective at inducing Ca2+ release either in the presence or absence of Cu2+, nor was methionine, the S-methyl derivative of homocysteine. The sulfur atom of methionine can form strong complexes with heavy metal ions but this configuration does not generally lead to oxidation (33).
In addition to Cu*+/mercaptan-induced Ca2+ release from SR, we have observed Ca2+ release induced by sulfhydryl oxidants, summarized in Table 111. Plumbagin most likely acts by oxidizing neighboring sulfhydryls to disulfides (24) and although less certain, it is probable that hypochlorous acid (HOC1) acts in the same way. Hydrogen peroxide (H202) is a well-known thiol oxidant whose activity as a thiol oxidant is greatly stimulated by the presence of Cu2+ ion (34); thus, the stimulation of H202-induced release with Cu2+ again implies sulfhydryl oxidation.
If Ca2+ release is caused by the oxidation of a critical sulfhydryl to a disulfide, then a disulfide reducing agent should in principle reverse the effect of Cu2+ and cysteine. In Fig. 4, SR vesicles were suspended in a standard KC1 buffer with 1 mM MgCl,, loaded with 20 p~ ca2+ using 1 mM M2+-

TABLE I Oxidation of mercaptans in the presence of Cu2'
The mercaptans listed were suspended at 30 WM in 100 mM KC1,5 mM HEPES, and either 1 or 10 mM MgCl,, pH 7.0 (0.9-ml total volume). 2 PM CuC1, was added and then at the times indicated, 0.1 ml of 120 mM Tris-HCl, 1 mM DTNB, pH 8.5, was added (final pH, 8.1). The concentration of unoxidized " S H groups in solution is proportional to the absorbance of light by DTNB at 410 nm; absorbance at 410 nm was monitored in a Beckman model DU-7 spectrophotometer. The rate of oxidation of " S H groups was calculated from the slope of the resulting absorbance uersw time plot. Clearly, the Cu2+-catalyzed oxidation of cysteine leads to Ca2+ release from SR. The oxidation of cysteine to cystine also leads to the formation of H2O2, presumably via superoxide radical. As noted in Table 11, cystine does not induce Ca2+ release from SR nor does H20, even at concentrations of 50 WM, whether in the presence of Cu2+ or not (data not shown). Superoxide dismutase catalyzes the aqueous disproportionation of superoxide radical to H202 and molecular oxygen, while catalase catalyzes the aqueous reduction of H202 to H20. Neither of these enzymes had any discernible effect on Cu2+/cysteine-induced Ca2+ efflux rates, even at molar enzyme concentrations equal to the concentration of Cu2+ (data not shown). Thus, the intermediates and side products of Cu2+-catalyzed autooxidation of cysteine to cystine ( i e . H2O2 and superoxide radical) and the end product, free cystine in solution, can be ruled out as releasing agents. All of the data  taken together strongly support the conclusion that the Ca2+ releasing activity of Cu2+ and cysteine involves the formation of a mixed disulfide between a sulfhydryl on the Ca2+ channel and the sulfhydryl on the added cysteine. Properties of Cu"/Cysteine-induced Ca2+ Release-Ag' and Ca2+-induced Ca2+ release rates have been reported to be strongly affected by Mg2+, pH, and ionic strength; also, the rates of Ca2+ release induced by these two methods are greatly dependent upon where the SR vesicles were derived from (Le. from longitudinal SR (LSR) or from terminal cisternae (HSR)). We measured Ca2+ release rates induced by Cuz+/ cysteine as a function of these parameters and found strong similarities between Cu2*/cysteine-induced Ca2+ release and Ca2+ release induced by these other methods. The rate of Ca2+ efflux induced by Cu2+/cysteine is plotted as a function of total Mg" in the medium in  Ca" release activity of " S H group oxidants SR vesicles were suspended at 0.25 mg of protein/ml in 100 mM KCI, 20 mM HEPES, and 3 mM MgC12. 20 p~ CaCl2 was added followed by 0.5 mM ATP (free [M%+] -2.5 mM). Upon completion of Ca2+ uptake, addition of either 50 p~ plumbagin or 100 p~ HOCl induced Ca2+ release; addition of 2 mM HzOz induced release which was stimulated by prior addition of 2 p M Cu2+. Release rates were calculated as described previously.  mM at time of DTT addition). Efflux was initiated by sequential addition of 10 p~ cysteine and 2 p~ CuClZ, and after completion of efflux, 2 mM DTT was added, resulting in slow reuptake of -80% of the released Ca2+. mM MgC12. Ca2+ uptake was initiated by addition of 0.5 mM ATP and then when the uptake was complete, 10 PM cysteine and 2 WM CuC12 were added to induce release. The ratio of ATP to ADP at the time of addition of Cu2+ and cysteine was about 4:l so that the free Mg2f concentration was about 0.2, 0.6,2.6,4.6, 7.1, and 9.6 mM, respectively. The rate and extent of Ca2+ release was inhibited with increasing free M e , a characteristic similar to Ca2+-induced Ca2+ release in SR vesicles (26) and in skeletal muscle fibers (36). We also measured the rate of Cu2+-catalyzed cysteine oxidation at high (10 mM) and low (1 mM) M P concentration, and we found a slight stimulation of the oxidation rate at the higher Mg2+ concentration (Table I). The lower rates of efflux at higher [Mg2+] thus seem to be due to inhibition of Ca2+ transport across the membrane and not inhibition of sulfhydryl oxidation.

Reagent [CU'+] Oxidation
To examine Ca2+ release from SR as a function of pH and ionic strength, SR vesicles (10 mg of protein/ml) were passively loaded with Ca2+ by incubation overnight in 0.8 mM Ca2+ and various concentrations of sucrose, KCl, buffers, and pH (see figure captions). The vesicles were diluted by a factor of 40 (final [protein] = 0.25 mg/ml) into Ca2+ free but otherwise identical media containing 0.1 mM arsenazo 111. Ca2+ efflux was initiated by addition of 10 NM cysteine and 2 pM Cu2+ and monitored spectrophotometrically as described previously. The rate constants (efflux rate in nmol of Ca2+/s/mg of protein divided by Ca2+ loading in nmol of Ca2+/mg of protein) and extent (%) of Cu2+/cysteine-induced Ca2+ release as functions of pH and ionic strength are plotted in Figs. 6 and 7. The rate of Ca2+ release has a broad maximum at around pH 7-7.25 and decreases with decreasing ionic strength; both characteristics are similar to Ca2+ release triggered by Ag+ (16). The rate constant reported at pH 7.25,100 mM KC1,0.8 mM MgC12, corresponds to an efflux rate of -1.6 nmol of Ca2+/s/mg of protein.
The rate of Ca2+ release induced by Cu2+/cysteine is several times (-4-5 times)  were calculated by dividing the Ca2+ efflux rate (in nmol of Ca2+/s/mg of protein) by the total Ca2+ loading of the vesicles (in nmol of Ca2+/mg of protein). from HSR was 9.8 nmol of Ca2+/mg of protein/s while the rate of efflux from LSR was about 2.3 nmol of Ca2+/mg of protein/s.

DISCUSSION
With the aid of the metallochromic indicator arsenazo 111, we have shown that various mercaptans in the presence of Cu2+ induce rapid Ca2+ release from SR. We have confirmed these results with Millipore filtration techniques using 45Ca2+ as a tracer, with the metallochromic indicator antipyrylazo 111, and with a Ca2+-sensitive electrode (data not shown). Ca2+ release from SR induced by Cu2+/cysteine shares many similarities with Ag+-and Ca2+-induced Ca2+ release; all three methods of inducing Ca2+ release are inhibited by procaine, tetracaine, and ruthenium red. Also all three show similar responses to free M e , pH, and ionic strength. In addition, Cu2+/cysteine-induced Ca2+ release is significantly faster in HSR than LSR, another common characteristic of Ag+-and Caz+-induced Ca2+ release.
Ag+ and the other heavy metals appear to induce Ca2+ release from SR by binding to a critical sulfhydryl group which is somehow closely linked to the site of Ca2+ release in uiuo, but the Cu2+/mercaptan effect seems to be due to Cu2+catalyzed cross-linking of the mercaptan's sulfhydryl group with this criticai protein sulfhydryl. Arguments supporting this oxidative mechanism of Ca2+ release are as follows. 1) Ca2+ release induced by Cu2+ and cysteine is reversed by addition of an excess of the disulfide reducing agent DTT (Fig. 4). 2) The rates of Cu2+-catalyzed oxidation of mercaptans (Table I) parallel the rates o f Ca2+ release induced by these mercaptans (Table 11). 3) The Ca2+ release rate induced by Cu2 and cysteine appears to be first order with respect to Cu2+ (Fig. 3 ) , as expected for Cu2+-catalyzed oxidation (21, 22). 4) Hypochlorous acid (23) and plumbagin (24), both potential sulfhydryl oxidants, induce Ca2+ release from SR in the absence of Cu2+, while Cu2+, which enhances the rate of " S H Oxidation Induces ea2+ Release from SR HzOz oxidation of cysteine, stimulates H,Oz-induced release (Table 111).
An issue of interest is the role of redox reactions in excitation-contraction coupling in skeletal muscle. We have shown that formation of a mixed disulfide induces Ca2+ release from SR. However, the means we have used to reduce this disulfide, thus closing the channel ( i e . with DTT), is too slow to be of physiological significance. If sulfhydryl redox reactions control the opening and closing of the CaZc channel, these reactions would almost certainly be enzymatically catalyzed. The activities of several nontransport proteins are known to be regulated by the oxidation states of sulfhydryl pairs which can switch from the dithiol to the disulfide state and back on a time scale of a few milliseconds. For example, lipoamide dehydrogenase is regulated by intracellular NAD(H) (37) and glutathione reductase is regulated by intracellular NADP(H) (37). Alternatively, Robillard and Konings (38) describe a dithiol/disulfide exchange reaction which may be the means of regulating the activities of a number of microbial membrane transport systems. In view of the parallels between Ca2+-and oxidation-induced Ca2+ release and the fact that the molecular mechanism underlying Ca2+-induced Ca2+ release is unknown, we suggest that the underlying mechanism of Ca2+-induced Ca2+ release may involve oxidation or some other perturbation of this critical sulfhydryl. For example, Ca2+ may regulate an enzyme responsible for catalyzing sulfhydryl redox reactions linked to Ca2+ release or, if Ca2+-induced Ca2+ release involves direct binding of Caz+ to the Ca2+ channel, this critical sulfhydryl may be one of the functional groups involved in gating the channel, whether it undergoes oxidation or not.
We have emphasized an interpretation of our results in which Cuz+ acts as a catalyst in the formation of a mixed disulfide between the exogenous mercaptan and an essential sulfhydryl on the Ca2+ channel. We note, however, that the Cu2+-catalyzed oxidation of cysteine does transiently produce Cu+. Under aerobic conditions, however, Cavallini et al. (21) place an upper limit on the number of copper atoms in the +1 state at 20% while Zwart et al. ( 2 2 ) estimate this number to be more like 2%. The +1 oxidation state of copper is reportedly more stable in an aerobic environment when the Cu+ ion is at the center of a coordination complex between two or more high affinity ligands (such as sulfhydryl groups) (39). However, given the low initial concentration of Cu2+ and the even lower Cu+ concentration, Cu+ binding or complex formation would seem an unlikely mechanism even without the data supporting the oxidation hypothesis ( i e . the cysteine dependence in Fig. 2, the reversal with DTT, and plumbagin-, HOCl-, and Cu2+/HZO2-induced release). Another possible explanation for the effects of Cu2+ and cysteine would be that Cu2+ catalyzes a thiol/disulfide exchange reaction between the cysteine thiol and a protein disulfide (40), though the oxidative mechanism again seems more likely.
Whether or not sulfhydryl oxidation is engaged in the in vivo gating of the Caz+ channel of the SR, it is interesting to note that the mercaptan penicillamine (which is either a sulfhydryl oxidant or reductant, depending on conditions) has been used to slow down the onset of muscular dystrophy in chickens (41). Also, elevated levels of the Ca2+ have been implicated in the onset of this disease (42). We suggest that, in this clinical context in the absence of Cu2+, penicillamine may protect the same critical sulfhydryls involved in Cu2+/ cysteine-induced release from pathological oxidation; the result of this protection would be maintenance of normal resting-state Ca2+ permeability levels for the membrane and thus lower cytosolic Ca2+ concentrations. Seen in this light, oxidation-induced Ca2+ release from SR may hold important clues for understanding muscular dystrophy in addition to understanding the functions of normal muscle.