Calcium-induced changes in permeability of dioleoylphosphatidylcholine model membranes containing bovine heart cardiolipin

At calcium concentrations up to about 4 mM a selective permeability increase of cardiolipin/dioleoylphosphatidylcholine (50:50, mol%) membranes for calcium and its chelator arsenazo III is observed. Under these conditions calcium does not occupy all the binding sites of cardiolipin at the membrane interface and no vesicle-vesicle interactions are found. Lowering of the cardiolipin content of the vesicles to 20 mol% extends the calcium concentration range in which a selective permeability for calcium and arsenazo III is appearing up to about 12 mM. We suggest that the observed selective permeability increase is caused by transient formation of inverted micellar structures in the membrane with cardiolipin as translocating membrane component for calcium and arsenazo III. At calcium concentrations of 4 mM and higher for 50 mol% cardiolipin-containing vesicles a general permeability increase is found together with calcium-cardiolipin binding in a 1:1 stoichiometry, vesicles aggregation and, above 8 mM of calcium, vesicle fusion. The loss of barrier function of the membrane under these conditions is correlated with vesicle aggregation and may be explained by a transition from a bilayer into a hexagonal HII organization of the phospholipids.     

Effect of lipid composition on the calcium/adenosine 5'-triphosphate coupling ratio of the Ca2+-ATPase of sarcoplasmic reticulum

The Ca2+-ATPase of sarcoplasmic reticulum was purified and depleted of proteolipids by solubilization in Triton X-100 and by fractionation on a DE-52 column. The protein reconstituted by deoxycholate-cholate dialysis at low lipid to protein ratios (2-5 mg of lipid/mg of protein), with either dioleoylphosphatidylethanolamine or monogalactosyldiglyceride, exhibited high initial rates of ATP-dependent Ca2+ uptake [300-900 nmol min-1 (mg of protein)-1] and coupling ratios (Ca2+ transported/ATP hydrolyzed) up to 1.2. Ca2+-ATPase reconstituted with lipids of increasing degrees of methylation (dioleoylphosphatidylethanolamine, dioleoylmonomethylphosphatidylethanolamine, dioleoyldimethylphosphatidylethanolamine and dioleoylphosphatidylcholine) or increasing degrees of glycosylation (monogalactosyldiglyceride and digalactosyldiglyceride) revealed a progressive decrease in both ATP-dependent Ca2+-uptake and coupling ratios. The rate and extent of Ca2+ uptake decreased as the dioleoylphosphatidylethanolamine/dioleoylphosphatidylcholine or monogalactosyldiglyceride/dioleoylphosphatidylcholine molar ratios in the reconstituted vesicles were reduced. Vesicles reconstituted with high molar ratios of dioleoylphosphatidylethanolamine/dioleoylphosphatidylcholine or monogalactosyldiglyceride/dioleoylphosphatidylcholine and at a high lipid to protein ratio became leaky and released the Ca2+ accumulated inside the vesicles when the temperature of the incubation mixture was increased (e.g., from 20 to 37 degrees C).(ABSTRACT TRUNCATED AT 250 WORDS)     

Oxidation of sulfhydryl groups and inhibition of the (Ca2+ + Mg2+)-ATPase by arsenazo III

In the presence of divalent cations, the metallochromic Ca2+ indicator arsenazo III is reduced by sulfhydryl groups to form an azo anion radical. Reduced arsenazo III is reoxidized back to its original state by oxygen. The formation of the arsenazo III azo anion radical in the presence of sarcoplasmic reticulum vesicles leads to the rapid inhibition of the (Ca2+ + Mg2+)-ATPase. These data indicate that several factors should be considered when arsenazo III is used as a Ca2+ indicator; (1) Functionally important sulfhydryl groups may be oxidized by arsenazo III; (2) the generation of free radicals by arsenazo III reduction may be toxic to the system being studied; (3) the absorbance spectrum of arsenazo III is altered when reduced by sulfhydryl groups.     

Intracellular pH changes induced by calcium influx during electrical activity in molluscan neurons

Simultaneous measurements of electrical activity and light absorbance have been made on nerve cell bodies from Archidoris monteryensis injected with indicator dyes. pH indicators, phenol red and bromocresol purple, and arsenazo III, which under normal conditions is primarily a calcium indicator have been employed. Voltage clamp pulses which induced calcium influx caused an absorbance decrease of the pH dyes indicating an internal acidification. The onset of the pH drop lagged the onset of Ca2+ influx by 200-400 ms, and pH continued to decrease for several seconds after pulse termination which shut off Ca2+ influx. Trains of action potentials also produced an internal pH decrease. Recovery of the pH change required periods greater than 10 min. The magnitude of the pH change was largely unaffected by external pH in the range 6.8-8.4. The voltage dependence of the internal p/ change was similar to the voltage dependence of calcium influx determined by arsenazo III, and removal of calcium from the bathing saline eliminated the pH signal. In neurons injected with EGTA (1-5 mM), the activity- induced internal Ca2+ changes were reduced or eliminated, but the internal pH drop was increased severalfold in magnitude. After the injection of EGTA, voltage clamp pulses produced a decrease in arsenazo III absorbance instead of the normal increase. Under these conditions the dye was responding primarily to changes in internal pH. Injection of H+ caused a rise in internal free calcium. The pH buffering capacity of the neurons was measured using three different techniques: H+ injection, depressing intrinsic pH changes with a pH buffer, and a method employing the EGTA-calcium reaction. The first two methods gave similar measurements: 4-9 meq/unit pH per liter for pleural ganglion cells and 13-26 meq/unit pH per liter for pedal ganglion cells. The EGTA method gave significantly higher values (20-60 meq/unit pH per liter) and showed no difference between pleural and pedal neurons.     

Induction of Ca2+ transport in liposomes by insulin.

The requirement of extracellular Ca2+ for insulin action has been indicated by past studies. With a view to understand the interaction of insulin with Ca2+ in the vicinity of the cell membrane, we have examined the ability of insulin and its constituent polypeptide chains A and B to translocate Ca2+ and Mg2+ across the lipid bilayer in two sets of synthetic liposomes. The first were unilamellar vesicles made of dimyristoylphosphatidylcholine and contained the Ca2+ sensor dye arsenazo III. Peptide-mediated Ca2+ and Mg2+ transport in these vesicles was monitored at 37 degrees C in a neutral buffer containing CaCl2 or MgCl2 using a difference absorbance method. In the second set, multilamellar vesicles of egg lecithin containing trapped fura-2 were employed and the cation transport was followed at 20 degrees C by fluorescence changes in the dye. Control experiments indicated that the hormonal peptides caused no appreciable perturbation of the vesicles leading to leakage of contents or membrane fusion. In both liposome systems, substantial Ca2+ and Mg2+ transport was observed with insulin and the B chain; the A chain was less effective as an ionophore. Quantitative analysis of the transport kinetic data on the B chain showed a 1:1 peptide-Ca2+ complex formed inside the membrane. In light of the available structural data on Ca2+ binding by insulin and insulin receptor, our results suggest the possibility of the hormone interacting with the receptor with the bound Ca2+.     

The stoichiometry of A23187- and X537A-mediated calcium ion transport across lipid bilayers

Initial rates of ionophore-mediated Ca2+ transport across egg phosphatidylcholine bilayers of large unilamellar vesicles were measured using the absorbance change of arsenazo III at 650 nm as an indicator of Ca2+ translocation. A23187 induced the movement of Ca2+ in a 2:1 ionophore: Ca2+ complex, whereas its methyl ester (CH3A23187) and X537A mediated Ca2+ movement in a 1:1 ionophore: Ca2+ complex. The relative potencies of these ionophores in transporting Ca2+ across lipid membranes were A23187 much greater than X537A greater than CH3A23187.     

Interaction of metallochromic indicators with calcium sequestering organelles.

Examples are presented of the interaction between cell organelles and metallochromic indicators used in the measurement of ionized Ca2+. Sarcoplasmic reticulum was found to sequester murexide type indicators along with Ca2+ in the presence of ATP, but not to sequester arsenazo III and antipyrylazo III. The presence of a permeable anion suppresses the sequestration of murexide type indicators by the sarcoplasmic reticulum. In the presence of ruthenium red, both rat liver and beef heart mitochondria release sequestered Ca2+ with arsenazo III, but not with murexide.     

Effects of pH, temperature, and calcium concentration on the stoichiometry of the calcium pump of sarcoplasmic reticulum

Coupling of Ca2+ transport to ATP hydrolysis by isolated skeletal muscle sarcoplasmic reticulum vesicles has been investigated by means of ATP pulse methods. The stoichiometric amounts of Ca2+ transported per pulse of ATP were measured by Ca2+-stat methods, using either a Ca2+ electrode or arsenazo III as end point detectors, or by means of 45CaCl2. Maximum coupling ratios (Ca2+/ATP), of 1.82 +/- 0.13 occurred at pH 6.8, 25 degrees C, and in the presence of saturating Ca2+ concentrations. Ca2+/ATP values decreased at alkaline pH, with an apparent pK alpha of 7.9. The coupling ratio was unaltered between 6 and 30 degrees C, but decreased to 0.4 at 42 degrees C. Uncoupling by alkaline pH and high temperatures was reversible. The coupling process was Ca2+-dependent, with a K0.5 value for Ca2+ of 0.12 microM and a Hill coefficient of 2.0. Ca2+ ions, which were transported into vesicles under conditions resulting in low coupling ratios, were retained as the calcium oxalate precipitate, following complete hydrolysis of substrate. Passive Ca2+ efflux and Ca2+ exchange, were independent of pH. The observed variations in Ca2+/ATP ratio cannot readily be explained on the basis of a pump-leak model. Rather, the Ca2+-ATPase appears to be capable of pumping Ca2+ ions, under physiological conditions, with variable stoichiometry that is dependent upon its thermodynamic loading.     

The entrapment of the Ca2+ indicator arsenazo III in the matrix space of rat liver mitochondria by permeabilization and resealing. Na+-dependent and -independent effluxes of Ca2+ in arsenazo III-loaded mitochondria.

The permeabilization-resealing technique [Al-Nasser & Crompton, Biochem. J. (1986) 239, 19-29] has been applied to the entrapment of arsenazo III in the matrix compartment of rat liver mitochondria. The addition of 10 mM-arsenazo III to mitochondria permeabilized with Ca2+ partially restores the inner-membrane potential (delta psi) and leads to the recovery of 3.9 nmol of arsenazo III/mg of protein in the matrix when the mitochondria are washed three times. The recovery of entrapped arsenazo III is increased 2-fold by 4 mM-Mg2+, which also promotes repolarization. ATP with or without Mg2+ decreased arsenazo III recovery. Under all conditions, less arsenazo III than [14C]sucrose is entrapped, in particular in the presence of ATP. The amount of arsenazo III entrapped is proportional to the concentration of arsenazo III used as resealant, and is equally distributed between heavy and light mitochondria. Arsenazo III-loaded permeabilized and resealed (PR) mitochondria develop delta psi values of 141 +/- 3 mV. PR mitochondria retain arsenazo III and [14C]sucrose for more than 2 h at 0 degrees C. At 25 degrees C, and in the presence of Ruthenium Red, PR mitochondria lose arsenazo III and [14C]sucrose at equal rates, but Ca2+ efflux is more rapid; this indicates that Ca2+ is released by an Na+-independent carrier in addition to permeabilization. The Na+/Ca2+ carrier of PR mitochondria is partially (60%) inhibited by extramitochondrial free Ca2+ stabilized with Ca2+ buffers; maximal inhibition is attained with 2 microM free Ca2+. A similar inhibition occurs in normal mitochondria with 3.5 nmol of matrix Ca2+/mg of protein, but the inhibition is decreased by increased matrix Ca2+. The data suggest the presence of Ca2+ regulatory sites on the Na+/Ca2+ carrier that change the affinity for matrix free Ca2+.     

Ca2+ transport studied with arsenazo III in Tetrahymena microsomes. Effects of calcium ionophore A23187 and trifluoperazine

Transport of Ca2+ in microsomal membrane vesicles of the Tetrahymena has been investigated using arsenazo III as a Ca2+ indicator. The microsomes previously shown to carry a Mg2+-dependent, Ca2+-stimulated ATPase (Muto, Y. and Nozawa, Y. (1984) Biochim. Biophys. Acta 777, 67-74) accumulated calcium upon addition of ATP and Ca2+ sequestered into microsomal vesicles was rapidly discharged by the Ca2+ ionophore A23187. Kinetic studies indicated that the apparent Km for free Ca2+ and ATP are 0.4 and 59 microM, respectively. The Vmax was about 40 nmol/mg protein per min at 37 degrees C. The calcium accumulated during ATP-dependent uptake was released after depletion of ATP in the incubation medium. Furthermore, addition of trifluoperazine which inhibited both (Ca2+ + Mg2+)-ATPase and ATP-dependent Ca2+ uptake rapidly released the calcium accumulated in the microsomal vesicles. These observations suggest that Tetrahymena microsome contains both abilities to take up and to release calcium and may act as a Ca2+-regulating site in this organism.     

Transport of Ca2+ by diltiazem across the lipid bilayer in model liposomes.

The calcium channel antagonist diltiazem was examined for its ability to translocate Ca2+ from an aqueous medium to the nonpolar lipid milieu. We monitored the spectral changes caused by the drug-mediated cation transport at 37 degrees C in unilamellar vesicles made of dimyristoyl phosphatidylcholine (DMPC) and containing the calcium-sensitive dye arsenazo III trapped inside. Vesicle leakage or membrane fusion caused by diltiazem was assessed by the use of vesicles containing fluorescent indicators. These effects were, however, found to be insignificant compared with ion transport. The transport was negligible at temperatures below the liquid crystalline to gel transition temperature of DMPC indicating a carrier mechanism of ion transport. A quantitative analysis of the transport kinetics indicated that a 1:2 Ca(2+)-drug complex is formed inside the lipid. The calcium ionophoretic ability of diltiazem, combined with other related data, suggests a possible role for Ca2+ in the conformation of the drug in the lipid membrane milieu and in its interaction with the calcium channel.     

Isolation of molecules recognized by monoclonal antibodies and antisera: the solid phase immunoisolation technique

Coupling of Ca2+ transport to ATP hydrolysis in isolated sarcoplasmic reticulum vesicles has been studied following pulsed additions of either ATP or Ca2+. ATP was infused as a pulse into medium, whose free Ca2+ concentration was maintained constant at saturating levels by a calciumstat procedure, using either a Ca2+-selective electrode or the spectrophotometric arsenazo III technique as Ca2+ indicators. The low ATP levels virtually exclude contributions by "basal" ATPase activity. Passive leakage of Ca2+, monitored after an ATP pulse, does not contribute more than 5% to subintegral coupling ratios. Pulsed additions of Ca2+ were made into medium. containing saturating concentrations of ATP, whose hydrolysis was monitored by a pH-stat procedure. Ca2+-stimulated hydrolysis continued until all the Ca2+ was transported into the vesicles. Values for the coupling ratio, Ca2+/ATP, of 1.82 +/- 0.12 and 1.79 +/- 0.15 were obtained by the ATP- and Ca2+-pulse methods, respectively.     

The influence of pH on the absorption spectrum of arsenazo III.

The absorption spectrum of arsenazo III in media containing K+, Mg2+ and Ca2+ is sharply influenced by pH in the range of 7.5--5.0. The effect of pH is particularly pronounced in the wavelength range 532--602 nm due to the large pH dependence of the dissociation constant of Mg-arsenazo III complex. Therefore absorption changes at these wavelengths during muscle contraction cannot be used as reliable indicators of free ionized Ca2+ concentration in the cell. The effect of pH is less pronounced, but still noticeable at the wavelength pairs 575--650 or 660--685 nm. Multiple layers of muscle cells grown on polystyrene coils permit measurement of absorption changes of arsenazo III, introduced into the cells, by equilibration with 0.5 mM arsenazo III under routine culture conditions. The absorbance changes recorded at 660--685 nm are probably related to changes in intracellular free Ca2+ concentration.     

Na+-Ca2+ exchange in squid optic nerve membrane vesicles is activated by internal calcium

The role of intracellular Ca2+ as essential activator of the Na+-Ca2+ exchange carrier was explored in membrane vesicles containing 67% right-side-out and 10% inside-out vesicles, isolated from squid optic nerves. Vesicles containing 100 microM free calcium exhibited a 2-fold increase in the initial rate of Na+i-dependent Ca2+ uptake as compared with vesicles where intravesicular calcium was chelated by 2 mM EGTA or 10 mM HEDTA. The activatory effect exerted by intravesicular Ca2+ on the reverse mode of Na+-Ca2+ exchange (i.e. Na+i-Ca2+o exchange) is saturated at about 100 microM Ca2+i and displays an apparent K 1/2 of 12 microM. Intravesicular Ca2+ produced activation of Na+i-Ca2+i exchange activity rather than an increase in Ca2+ uptake due to Ca2+-Ca2+ exchange. The presence of Ca2+i was essential for the Na+i-dependent Na+ influx, a partial reaction of the Na+-Ca2+ exchanger. In fact, the Na+ influx levels in vesicles loaded with 2 mM EGTA were close to those expected from diffusional leak while in vesicles containing Ca2+i an additional Na+-Na+ exchange was measured. The results suggest that in nerve membrane vesicles Ca2+ at the inner aspect of the membrane acts as an activator of the Na+-Ca2+ exchange system.     

Passive Ca transport in human red blood cell ghosts measured with entrapped arsenazo III

The rate of Ca influx into ghosts containing arsenazo III changes with time, being most rapid during the first 5 min after Ca is added to the outside and declining thereafter. The rate of Ca influx is a nonlinear function of extracellular Ca and plateaus as the latter is increased above 1 mM. The rate of Ca influx was measured as a function of the transmembrane gradients of Na and K and changes in the permeability of the membrane to K and Cl produced by valinomycin and SITS (4-acetamido-4'-isothiocyano-stilbene-2-2'-disulfonic acid), respectively. Changes in the rate of Ca influx are consistent with expected effects of these treatments on the membrane potential. Oligomycin (10 micrograms/ml) and quinidine (1 mM) inhibit the rate of Ca uptake by inhibiting Ca-induced changes in the K permeability. At constant membrane potential, furosemide produced a slight (15%) consistent increase in Ca uptake. Other experiments show that resealed ghosts are heterogeneous in their passive permeability to Ca and that A23187 can be used to effectively eliminate such differences. The results of this paper show that resealed human red cell ghosts containing arsenazo III can be used to continuously monitor intracellular free Ca and to study the factors that influence the permeability of the red cell membrane to Ca.     

Changes of intracellular Ca++ as measured by arsenazo III in relation to the K permeability of human erythrocyte ghosts.

A Ca-sensitive dye, arsenazo III, has been incorporated into resealed human erythrocyte ghosts and calibrated to monitor continuously micromolar concentrations of intracellular ionized Ca ([Ca++]i). When the external concentration of Ca is much greater than [Ca++]i, [Ca++]i increases because of a net balance between Ca influx and efflux. Dynamic changes in [Ca++]i regulate K efflux, which in turn may influence the rate of Ca influx. A procedure for purifying arsenazo III is also described.     

Relation between the passive transport of calcium into vesicles of the myocardial sarcolemma and membrane potential

The effect of membrane potential on the passive 45Ca2+ uptake by cardial sarcolemmal vesicles was investigated. Membrane potentials were generated by the K+ gradient in the presence of valinomycin and were measured using fluorescent dye diS-C3-(5). It was shown that the 45Ca2+ influx into vesicles increased twice after membrane depolarization. Evaluation of the 45Ca2+ influx over a wide range of membrane potentials produced a profile similar to that of current-voltage relationships for single calcium channels in isolated cardiomyocytes. Passive 45Ca2+ transport was inhibited by 1 mM Cd2+ and Co2+. It is suggested that the voltage-dependent Ca2+ influx into vesicles occurs through Ca2+-channels.     

Dissociation of calcium from the phosphorylated calcium-transporting adenosine triphosphatase of sarcoplasmic reticulum: kinetic equivalence of the calcium ions bound to the phosphorylated enzyme.

The internalization of 45Ca by the calcium-transporting ATPase into sarcoplasmic reticulum vesicles from rabbit muscle was measured during a single turnover of the enzyme by using a quench of 7 mM ADP and EGTA (25 degrees C, 5 mM MgCl2, 100 mM KCl, 40 mM MOPS.Tris, pH 7.0). Intact vesicles containing either 10-20 microM or 20 mM Ca2+ were preincubated with 45Ca for approximately 20 s and then mixed with 0.20-0.25 mM ATP and excess EGTA to give 70% phosphorylation of Etot with the rate constant k = 300 s-1. The two 45Ca ions bound to the phosphoenzyme (EP) become insensitive to the quench with ADP as they are internalized in a first-order reaction with a rate constant of k = approximately 30 s-1. The first and second Ca2+ ions that bind to the free enzyme were selectively labeled by mixing the enzyme and 45Ca with excess 40Ca, or by mixing the enzyme and 40Ca with 45Ca, for 50 ms prior to the addition of ATP and EGTA. The internalization of each ion into loaded or empty vesicles follows first-order kinetics with k = approximately 30 s-1; there is no indication of biphasic kinetics or an induction period for the internalization of either Ca2+ ion. The presence of 20 mM Ca2+ inside the vesicles has no effect on the kinetics or the extent of internalization of either or both of the individual ions. The Ca2+ ions bound to the phosphoenzyme are kinetically equivalent. A first-order reaction for the internalization of the individual Ca2+ ions is consistent with a rate-limiting conformational change of the phosphoenzyme with kc = 30 s-1, followed by rapid dissociation of the Ca2+ ions from separate independent binding sites on E approximately P.Ca2; lumenal calcium does not inhibit the dissociation of calcium from these sites. Alternatively, the Ca2+ ions may dissociate sequentially from E approximately P.Ca2 following a rate-limiting conformational change. However, the order of dissociation of the individual ions can not be distinguished. An ordered-sequential mechanism for dissociation requires that the ions dissociate much faster (k greater than or equal to 10(5) s-1) than the forward and reverse reactions for the conformational change (k-c = approximately 3000 s-1). Finally, the Ca2+ ions may exchange their positions rapidly on the phosphoenzyme (kmix greater than or equal to 10(5) s-1) before dissociating. A Hill slope of nH = 1.0-1.2, with K0.5 = 0.8-0.9 mM, for the inhibition of turnover by binding of Ca2+ to the low-affinity transport sites of the phosphoenzyme was obtained from rate measurements at six different concentrations of Mg2+.     

Effects of alloxan and streptozotocin on calcium transport in isolated mouse liver mitochondria

The effect of alloxan and streptozotocin on the fluxes of Ca2+ in isolated mouse liver mitochondria was studied with dual wave-length spectrophotometry, using antipyrylazo III as metallochromic indicator. Streptozotocin had no effect on Ca2+ uptake, whereas alloxan inhibited the initial rate and extent of Ca2+ influx in a way dependent on the duration of preincubation, and occurrence of Pi in the reaction mixture. A rapid release of Ca2+ followed upon addition of either FCCP or alloxan after the reaction had been started. When added to preloaded mitochondria, alloxan induced a concentration dependent release of Ca2+. The data suggest that alloxan induces an initial release of mitochondrial Ca2+, which is followed by inhibition of Ca2+ influx. The initial release may be due to uncoupler activity induced by alloxan, and the inhibition of Ca2+ influx may be a consequence of inhibited Pi transport.     

Oxidation induced by phthalocyanine dyes causes rapid calcium release from sarcoplasmic reticulum vesicles

The copper containing phthalocyanine dyes, alcian blue, copper phthalocyanine tetrasulfonic acid, and Luxol fast blue MBSN are found to induce rapid calcium efflux from actively loaded sarcoplasmic reticulum (SR) vesicles. Alcian blue (5 microM), with 1 mM free Mg2+ triggered Ca2+ efflux at rates greater than 20 nmol/mg of SR/s. As in the case of Ca2+ efflux induced by calcium, heavy metals, or SH oxidation with Cu2+/cysteine, efflux induced by phthalocyanines is also stimulated by adenine containing nucleotides and inhibited by millimolar Mg2+ and submicromolar ruthenium red (RR). In addition, analogs of RR, such as hexamminecobalt(III) chloride or hexammineruthenium(III) chloride also inhibit Ca2+ efflux but are effective at somewhat higher concentrations (approximately 50 microM). Calcium release stimulated by phthalocyanines is specific for SR derived from the terminal cisternae region rather than longitudinal SR. Preincubation of alcian blue with the reducing agents, sodium dithionite, dithiothreitol, or cysteine causes complete loss of Ca2+ release activity from SR vesicles. Reoxidation of the alcian blue leads to return of the Ca2+ release activity of the phthalocyanine dye. The copper containing phthalocyanine dyes appear to cause rapid Ca2+ release from SR vesicles by oxidizing sulfhydryl groups associated with the calcium release channel. Moreover, phthalocyanines appear to act by oxidizing a pair of neighboring sulfhydryls to a disulfide because subsequent additions of the reducing agent dithiothreitol promote the closure of the Ca2+ channel and calcium re-uptake.