Calcium transport across the plasma membrane: stimulation by calmodulin.

Active transport of calcium into inside-out vesicles of red blood cell membranes was stimulated equally by (i) the purified protein activator of calcium-activated, magnesium-dependent adenosinetriphosphatase isolated from red cell hemolyzates and (ii) calmodulin, a protein activator of cylic nucleotide phosphodiesterase isolated from bovine brain. The results provide further evidence for the identity of red blood cell activator and calmodulin and show that this cytoplasmic protein may participate in the regulation of plasma membrane calcium transport.

Numerical simulation of assay of the calcium pump of intact red blood cells.

An assay of the Ca pump ATPase of intact human RBCs is described in a companion paper (Wu, L., Hinds, T. R. and Vincenzi, F. F. (1992) Biochim. Biophys. Acta 1106, 56-62). The assay is based on the rapid loss of ATP in RBCs that occurs when the cells are exposed to the ionophore, A23187, in the presence of Ca. An unexpected finding was that the initial loss of ATP follows pseudo-first-order kinetics. This was unexpected because the ATP content of RBCs is somewhat higher than the Km of the Ca pump for ATP. Thus, the initial loss of ATP would be expected to follow zero-order kinetics; at least if the Ca pump ATPase operated with Michaelis kinetics. We performed a series of computer simulations of the Ca pump ATPase to investigate the possible cause of the unexpected pseudo-first-order behavior. The results confirmed that the data can not be accounted for by Michaelis kinetics of the Ca pump ATPase. Possible effects of adenylate kinase were tested and were also not found to account for the pseudo-first-order behavior of an ATPase operating with Michaelis kinetics. The enzymatic properties of the Ca pump ATPase were re-examined. It was found that the Ca pump ATPase exhibits positive cooperativity toward ATP. The apparent cooperativity was 1.91. In simulations it was found that positive cooperativity of the Ca pump ATPase in the range of 1.5 to 2.0 could account for the pseudo-first-order behavior. Excellent fit of the simulation data to first-order behavior was true with or without any contribution from adenylate kinase. Rate constants of ATP loss were thus examined using cooperativity of 2.0. Over a wide range the rate constant of the loss of ATP was directly proportional to the assumed Vmax of the Ca pump ATPase, but only if the data were limited to loss of less than 67% of the initial ATP. It is suggested, therefore, that the rate constant for the initial loss of ATP in intact RBCs, as stimulated by the ionophore A23187, can be taken as a measure of the capacity of the Ca pump ATPase.

Enhancement of (Ca2+ + Mg2+)-ATPase activity of human erythrocyte membranes by hemolysis in isosmotic imidazole buffer. I. General properties of variously prepared membranes and the mechanism of the isosmotic imidazole effect.

1. Membranes prepared from human erythrocytes hemolyzed in isosmotic (310 imosM) imidazole buffer, pH 7.4, show enhanced and stabilized (Ca2+ + Mg2+)-ATPase activity compared with membranes prepared from erythrocytes hemolyzed in hypotonic (20 imosM) phosphate or imidazole buffer, pH 7.4. 2. Exposure of intact erythrocytes or well-washed erythrocyte membranes to isosmotic imidazole does not cause enhanced (Ca2+ + Mg2+)-ATPase activity. 3. Exposure of erythrocyte membranes, in the presence of isosmotic imidazole, to the supernatant of erythrocyte hemolysis or to a partially purified endogenous (Ca2+ + Mg2+)-ATPase activator, promotes enhanced (Ca2+ + Mg2+)-ATPase activity. Under appropriate conditions, NaCl can be shown to substitute for imidazole. The results demonstrate that imidazole does not act directly on the erythrocyte membrane but rather by promoting interaction between an endogenous (Ca2+ + Mg2+)-ATPase activator and the erythrocyte membrane.

Enhancement of (Ca2+ + Mg2+)-ATPase activity of human erythrocyte membranes by hemolysis in isosmotic imidazole buffer. II. Dependence on calcium and a cytoplasmic activator.

1. Activity of the (Ca2+ + Mg2+)-ATPase of erythrocyte membrane may be enhanced by a cytoplasmic protein activator. The presence of Ca2+ is necessary for the ionic strength-dependent interaction between the erythrocyte membrane and the activator. This is true no matter the purity of activator (unfractionated hemolysis supernatant or partially purified activator) or the major source of ionic strength (imidazole or NaCl). 2. When the endogenous activator enhances (Ca2+ + Mg2+)-ATPase activity of the erythrocyte membrane, there is a physical association between activator and membrane. This association is not disrupted by a decrease in ionic strength to 0.005 but is reversed by exposure to 5 mM ethyleneglycol bis(beta-aminoethyl ether)-N,N'-tetraacetic acid. 3. Activator binding necessary for enhancement of (Ca2+ + Mg2+)-ATPase activity may occur during preparation of membranes or during incubation for assay of ATPase.

Assay of the Ca pump ATPase activity of intact red blood cells.

An assay for the Ca pump ATPase of intact human red blood cells (RBCs) was developed. The assay utilized a small volume (typically 10 microliters) of packed RBCs in 1 ml of a buffer of known composition. The assay was based on the exposure of intact RBCs to the ionophore, A23187, in the presence of Ca. Such exposure caused a rapid degradation of ATP in RBCs. This degradation process is modeled in a numerical simulation in a companion paper (Vincenzi, F. F. and Hinds, T. R. (1992) Biochim. Biophys. Acta 1105, 63-70). The loss of ATP followed pseudo-first-order kinetics, and the rate constants for ATP degradation was taken as a measure of the capacity of the Ca pump ATPase. A number of variables were examined to optimize the activity of the ATPase. These variables included the concentrations of Ca and A23187. Because A23187 can promote loss of cellular Mg, it was necessary to include MgCl2 in the incubation medium to optimize ATPase activity. Likewise, it was determined that inclusion of iodoacetic acid optimized the rate of ATP loss, presumably by preventing the resynthesis of ATP from ADP and inorganic phosphate. Cobalt inhibited the ionophore-dependent loss of ATP by apparent competition with Ca for binding to A23187. Results of many assays demonstrated substantial differences in the rate constant for ATP loss in RBCs from different individuals. RBCs were selected according to density. Density associated loss of Ca pump ATPase activity was observed both by the intact RBC assay, and by assay of Ca pump ATPase activity in saponin lysates of RBCs. The correlation coefficient between the two assays was 0.93. It is suggested that the rate constant for ATP loss in intact RBCs exposed to A23187 and Ca can be taken as a measure of the Ca pump ATPase activity. This may be useful when isolated membrane ATPase assays fail (e.g., dog RBCs). The intact cell assay can also be carried out on very small volumes of cells and may be of particular value when RBC volumes are limited.

Inhibition of the Ca pump of intact red blood cells by t-butyl hydroperoxide: importance of glutathione peroxidase.

Incubation of human red blood cells (RBCs) with t-butyl hydroperoxide (tBHP) resulted in inhibition of the Ca-pump ATPase. This was demonstrated using an assay of the Ca-pump ATPase activity in intact RBCs. In this assay, activity of the Ca-pump ATPase is expressed as the rate constant of the initial loss of ATP in RBCs exposed to Ca and A23187. Pseudo-first-order rate constants (Ca-pump ATPase rate constants) were lower in the presence of tBHP versus controls. Incubation of RBCs with tBHP resulted in both a time- and concentration-dependent inhibition of the Ca-pump ATPase (IC50 approximately 1 mM). Incubation of RBCs with tBHP also resulted in decreased oxyhemoglobin, increased methemoglobin and increased thiobarbituric acid reactive substances (TBARS). GSH levels were significantly lower in the presence of tBHP. GSH fell from a control value of 2.2 mmol/l RBC to 0.46 mmol/l RBC after incubation with 0.25 mM tBHP for 15 min. Both butylated hydroxytoluene and stobadine prevented the formation of TBARS and were partially effective in protecting the Ca-pump ATPase from tBHP-induced inhibition. Dithiothreitol was completely effective in preventing the tBHP-induced formation of TBARS as well as inhibition of the Ca-pump ATPase. However, when added after exposure to tBHP, dithiothreitol was unable to restore Ca-pump ATPase activity completely. An activity of dithiothreitol independent of enzymic thiol group reduction was apparent. In the presence of mercaptosuccinate, a potent inhibitor of glutathione peroxidase, the ability of dithiothreitol to protect the Ca-pump ATPase from tBHP-induced inhibition was abolished. Therefore, protection by dithiothreitol may be afforded by its ability to replenish GSH from oxidized glutathione, thus allowing glutathione peroxidase to metabolize tBHP. These results may be interpreted to suggest that inhibition of the Ca-pump ATPase in intact RBCs occurs as a result of tBHP-induced oxidant stress and subsequent lipid peroxidation which can be prevented by certain antioxidants including butylated hydroxytoluene, stobadine, and thiol-containing compounds such as dithiothreitol. These findings provide further insight into the mode of action of hydroperoxides and certain reactive oxygen species that have been implicated in oxidative stress associated with various pathological conditions. The importance of the GSH/glutathione peroxidase system in metabolizing organic hydroperoxides is also demonstrated.

Inhibition of basal and calmodulin-activated Ca2+-pump ATPase by fractionated compound 48/80.

Compound 48/80 (48/80), a mixture of polycationic compounds was fractionated using affinity chromatography on calmodulin-Sepharose. Unfractionated 48/80 and various fractions were tested for their potential inhibitory effects on ATPase activities of isolated human red blood cell membranes. ATPase activities tested included: Mg2+-ATPase, the Na+/K+-pump ATPase, and the Ca2+-pump ATPase in both its basal (calmodulin-independent) and calmodulin-activated state. Neither 48/80 nor its various fractions were very potent or efficacious inhibitors of the Mg2+-ATPase or the Na+/K+-pump ATPase. In agreement with previous reports, 48/80 was found to be an inhibitor of the calmodulin-activated Ca2+-pump ATPase. By contrast, we found that unfractionated, as well as some fractionated, material inhibited both the basal (calmodulin-independent) and calmodulin-activated Ca2+-pump ATPase activity. A fraction designated as Fraction III bound to calmodulin-Sepharose in the presence of Ca2+ and low salt and was eluted in the absence of Ca2+ and 0.15 M NaCl. By gel filtration, Fraction III had an apparent average molecular weight of 2064 (1320 for unfractionated material). Fraction III was the most potent inhibitor of the Ca2+-pump ATPase with IC50 values for the basal and calmodulin-activated forms of the enzyme of 0.6 and 1.2 micrograms/ml, respectively. Inhibition by Fraction III was cooperative with n apparent values of 2.4 and 5.7, respectively, for the basal and calmodulin-activated forms of the enzyme. Thus, binding of 48/80 constituents to calmodulin can not fully account for the observed data. Direct interaction of 48/80 constituent(s) with the enzyme and/or the lipid portion of the membrane is suggested.

Calmodulin pharmacology.

Calmodulin (CaM) is a major intracellular receptor for Ca2+. CaM is thus a crucial receptor to consider in pharmacological modification of cellular activity. Potential mechanisms by which drugs may modify CaM effectiveness are considered in the context of its interaction with Ca2+ and in turn with its various effectors. Some examples of established drug mechanisms are considered. A wide range of chemical compounds representing diverse pharmacological classes are anti-CaM under some conditions. No simple relationships have been established between molecular level events and therapeutic applicability of anti-CaM compounds.

The effect of ETH 1001 on ion fluxes across red blood cell membranes.

The calcium selective ionophore, ETH 1001, and the divalent cation ionophore, A23187, promoted Ca2+ flux across RBC membranes under various experimental conditions. ETH 1001 did not promote the passive movement of Mg2+ whereas A23187 did. The results confirm the potential application of ETH 1001 as a Ca2+ selective ionophore for biological membranes.

Purified red blood cell Ca2+-pump ATPase: evidence for direct inhibition by presumed anti-calmodulin drugs in the absence of calmodulin.

A variety of presumed anti-calmodulin (anti-CaM) drugs was tested for their potential inhibitory effects on the isolated, purified and reconstituted Ca2+-pump ATPase of human red blood cell membranes. Anti-CaM drugs inhibited the Ca2+-pump ATPase both in the absence and presence of added CaM. Qualitatively similar inhibition was observed in two different ATPase assay systems. In asolectin vesicles in the absence of added CaM trifluoperazine (TFP), N-(6-aminohexyl)-5-chloro-1-naphthalene- sulfonamide (W-7), vinblastine, dibucaine, imipramine, propranolol and dimethylpropranolol (UM-272) were all inhibitory. Potency of anti-CaM drugs was generally greater on the enzyme reconstituted in asolectin vesicles than on the enzyme reconstituted in phosphatidylcholine vesicles, either in the presence or absence of CaM. The results emphasize that anti-CaM drugs have actions other than to bind to CaM. Possible direct interaction of amphipathic cationic anti-CaM drugs with the Ca2+-pump ATPase and/or its lipid environment is suggested.

Decreased calcium pump adenosine triphosphatase in red blood cells of hypertensive subjects.

Several operationally defined adenosine triphosphatase (ATPase) activities were determined in vitro in red blood cell lysates of normotensive or hypertensive humans: Mg2+-ATPase, Na+,K+-ATPase, and Ca2+ pump ATPase, the latter in the calmodulin-activated and basal states. Basal Ca2+ pump ATPase was defined as the Ca2+-activated ATPase resistant to 10(-4) M trifluoperazine. Subjects were part of a double-blind study in which treatment was divided into several phases: baseline (4 weeks), placebo or calcium (1 g elemental calcium/day, 8 weeks), placebo washout (4 weeks), placebo or calcium (1 g elemental calcium/day, 8 weeks). Irrespective of the phase of treatment, the basal Ca2+ pump ATPase activity in red blood cell lysates of 36 hypertensive subjects was significantly less than that in lysates from 18 normotensive subjects. Other ATPase activities did not differ significantly, although all ATPases tended to be decreased in hypertension. The data are consistent with previous reports of altered membrane Ca2+ binding and transport in hypertension, but the precise changes are not elucidated.

Inhibition by activated neutrophils of the Ca2+ pump ATPase of intact red blood cells.

Human neutrophils, activated by phorbol myristate acetate in the presence of intact red blood cells (RBCs), caused inhibition of the Ca2+ pump ATPase of the RBCs and fragmentation of the enzyme as well as other membrane proteins. Inhibition of the Ca2+ pump ATPase of intact RBCs was directly related to the neutrophil concentration and the time of incubation. Ca2+ pump ATPase activity was partially protected by the addition of exogenous glutathione-glutathione peroxidase, but not by superoxide dismutase. The addition of sodium azide, a potent inhibitor of endogenous RBC catalase, enhanced inhibition of the Ca2+ pump ATPase of intact RBCs. Examination by SDS-polyacrylamide gel electrophoresis of membrane proteins isolated from RBCs preincubated with activated neutrophils showed gross changes in banding patterns as compared to controls. Thus, a significant amount of methemoglobin appeared to be associated with the membrane proteins, and, in general, protein bands appeared to be more diffuse and less defined than proteins in control lanes. In addition, there was an increase in the low molecular weight protein bands. Using a monoclonal antibody to the Ca2+ pump ATPase, it was shown that the 140 kDa band representing the Ca2+ pump ATPase decreased, with concomitant appearance of two low molecular weight bands running at 8.2 and 6.8 kDa in the membrane proteins from RBCs preincubated with activated neutrophils. The data are interpreted to suggest that inhibition of the Ca2+ pump ATPase in intact RBCs under these conditions occurred as a result of: neutrophil-derived superoxide, dismutation of superoxide, to H2O2, diffusion of H2O2 into RBCs, a Fenton type reaction between oxyhemoglobin, and H2O2 producing hydroxyl radical and/or a ferryl radical capable of promoting protein fragmentation of RBC membrane proteins, including the plasma membrane Ca2+ pump ATPase.

Stereochemical analogs of a muscarinic, ganglionic stimulant. 2. Cis and trans olefinic, epoxide, and cyclopropane analogs related to 4-[N-(3-chlorophenyl)carbamoyloxy]-2-butynyltrimethylammonium chloride (McN-A-343).

Preparation of analogs of 4-[N-(3-chlorophenyl) carbamoyloxy]-2-butynyltrimethylammonium chloride [1 (McN-A-343)], cis- and trans-4-[N-(4-chlorophenyl)carbamoyloxy]-2-butenyltrimethylammonium iodides (5 and 6), and the corresponding epoxides and cyclopropanes is reported. Pharmacological testing for ganglion-stimulating activity demonstrated that the trans olefin 6 and trans epoxide 8 have properties similar to 1, while the trans cyclopropane analog 10 was inactive. All cis compounds were inactive. The muscarinic ganglion-stimulating properties of the active compounds are interpreted in terms of similar fit at the receptor level by the alkyltrimethylammonium ion and the ether oxygen 5.7 A distant, as well as an electron-rich center midway between groups in the form of a double bond or unshared electron pairs. Comparison of smooth muscle and ganglion-stimulating properties of the compounds showed that trans epoxide 8 was the most selective for muscarinic ganglionic sites.

Stereochemical analogs of a muscarinic, ganglionic stimulant. 3. 2,3-Substituted bicyclo(2.2.1)hept-5-enes and -heptanes related to 4-(N-(3-chlorophenyl)carbamoyloxy)-2-butynyltrimethylammonium chloride (McN-A-343).

Preparation of analogs of 4-[N-(3-chlorophenyl)carbamoyloxy]-2-butynyltrimethylammonium chloride (1, McN-A-343), the isomeric 2-trimethylammoniomethyl-3-[N-(4-chlorophenyl)carbamoyloxymethyl]bicyclo [2.2.1]hept-5-ene iodides (10-13), and the corresponding -bicyclo[2.2.1]heptane iodides (14-17) are reported. None of the compounds demonstrated ganglion-stimulating activity similar to 1 or antagonized the effects of 1.

Effects of hemicholinium and bretylium on the release of autonomic transmitters in the isolated sino-atrial node.

1. In the isolated, spontaneously beating, sino-atrial node of the rabbit selective electrical excitation of intranodal autonomic nerve fibres results in a biphasic chronotropic response. This chronotropic response (negative followed by positive chronotropism) is due to the release of the autonomic transmitters (acetylcholine and noradrenaline, respectively) from intranodal nerve fibres.2. In the presence of 2 x 10(-4) g/ml hemicholinium, the negative chronotropic (cholinergic) response is abolished while the positive chronotropic (adrenergic) response is unaltered.3. In the presence of 5 x 10(-6) g/ml bretylium, the positive chronotropic response is abolished while the negative chronotropic response is little affected.4. After blockade of the negative chronotropic response by hemicholinium, bretylium abolishes the remaining positive chronotropic response. The effect of bretylium is not altered in the presence of hemicholinium.5. Considering currently accepted mechanisms of action for hemicholinium and bretylium, the results of these experiments do not lend support to the cholinergic link hypothesis of adrenergic neuro-effector transmission.

Ion transport ATPases as targets for free radical damage. Protection by an aminosteroid of the Ca2+ pump ATPase and Na+/K+ pump ATPase of human red blood cell membranes.

Preincubation of red blood cell membranes in the presence of ferrous sulfate and EDTA resulted in both a concentration- and time-dependent inhibition of the Na+/K+ pump ATPase, basal Ca2+ pump ATPase, and the calmodulin- (CaM) activated Ca2+ pump ATPase. The IC50 for all three ATPases was approximately 2.5 x 10(-5) M iron. The addition to membranes of ferrous iron and EDTA in an approximately 1:1 ratio resulted in conversion to the ferric iron form in several minutes. However, inhibition of the ion pump ATPases and cross-linking of membrane proteins occurred over the course of several hours. The time course of formation of thiobarbituric acid-reactive substances (TBARS) closely paralleled inhibition of the ion pump ATPases. Inhibition of the ion pump ATPases was prevented by the addition of deferoxamine or superoxide dismutase but not by mannitol, or catalase. Both butylated hydroxytoluene and tirilazad mesylate (U74006F) prevented the formation of TBARS, limited the inhibition of the ion pump ATPases, and reduced cross-linking of membrane proteins. These data may be interpreted to suggest that inhibition of ion pump ATPases in plasma membranes may occur as a result of iron-promoted formation of superoxide and subsequent lipid peroxidation, which can be prevented by free-radical scavengers including butylated hydroxytoluene and U74006F.

Inhibition of Ca2+-pump ATPase and the Na+/K+-pump ATPase by iron-generated free radicals. Protection by 6,7-dimethyl-2,4-DI-1- pyrrolidinyl-7H-pyrrolo[2,3-d] pyrimidine sulfate (U-89843D), a potent, novel, antioxidant/free radical scavenger.

Preincubation of red blood cell (RBC) membranes with a model system known to generate reactive oxygen species (ROS) and free radicals (200 microM ferrous sulfate and 200 microM EDTA, Fe2+/EDTA) resulted inhibition of the Na+/K+ -pump ATPases was also associated with membrane protein crosslinking and lipid peroxidation, the latter as monitored by the formation of thiobarbituric acid reactive substances (TBARS). Inhibition of the ion transport ATPases, protein cross-linking and formation of TBARS were prevented by U-89843D in a concentration-dependent manner, with half-maximal protection seen at 0.3 microM. U-89843D was more potent than the classical antioxidant butylated hydroxytoluene. Neither U-89843D nor the solvent DMSO had any effect on the assay of TBARS. U-89843D exerted only minimal inhibitory activity on ATPase activities. Thus, U-89843D was potent in vitro in preventing a variety of membrane-damaging reactions mediated by ROS. It is suggested that protection of membranes from ROS-mediated damage is of potential usefulness in the prevention and treatment of certain disease processes.

Increased cytosolic free calcium in red blood cells is associated with essential hypertension in humans.

Intracellular free calcium (Cai2+) was measured in human red blood cells (RBCs) using the Ca(2+)-sensitive fluorescent probe, fluo-3. Fresh RBCs were loaded with the acetoxymethyl ester of fluo-3 (fluo-3-AM) in vitro. Following incubation at 37 degrees C for 60 min, cells were separated from nonincorporated fluo-3-AM and fluo-3. Fluorescence was quantified with excitation and emission wavelengths of 506 nm and 526 nm, respectively. The relatively long wavelength characteristics of fluo-3 avoid much of the interference from hemoglobin that makes agents such as quin-2 and fura-2 of limited value in RBCs. A protocol with corrections for quenching and for extracellular fluo-3 was employed for each sample. Cai2+ in red blood cells from 15 normotensive volunteers averaged 134 +/- 12 nmol/L. In RBCs from 30 hypertensive patients who were not treated pharmacologically, Cai2+ was 317 +/- 32 nmol/L (P < .05). In a group of 27 hypertensives who were treated with various drugs, RBC Cai2+ was 221 +/- 27 nmol/L. Measurements were also performed in RBCs from 16 patients before and at least 2 weeks following treatment with one or more antihypertensive agents. Reduction of RBC Cai2+ from 342 +/- 47 to 252 +/- 39 (P < .001) was observed in 16 patients whose blood pressure fell with treatment. Overall, there was a correlation between RBC Cai2+ and diastolic blood pressure (r = 0.413, P < .01, d.f. = 71). Thus, the present results confirm an association between elevated RBC Ca2+ and essential hypertension.(ABSTRACT TRUNCATED AT 250 WORDS)

Pharmacological modification of the Ca2+-pump ATPase activity of human erythrocytes.

The Ca2+-pump ATPase of human RBC membranes appears to be exquisitely sensitive to a variety of amphipathic molecules. The acidic protein calmodulin (CaM) activates the enzyme some three- to fivefold with an apparent Kd of approximately 1-5 nM. A variety of other amphipathic anions, such as acidic phospholipids, free fatty acids, and anionic detergents, are less potent and in some cases less efficacious than CaM, but also activate the enzyme. Similar results have been observed for other CaM-dependent enzymes, and it is suggested that these agents mimic CaM in a general, but rather nonspecific, fashion. Activation of the human RBC Ca2+-pump ATPase by CaM or other amphipathic anions can be selectively antagonized by a wide variety (structurally and pharmacologically) of amphipathic cations. There is no simple relationship between antagonism of CaM in vitro and the general systemic pharmacology of these drugs. The only common feature of such drugs is that they are amphipathic cations. Neutral molecules such as saponin exerted neither CaM-like activity nor CaM antagonism. Great caution is urged in the inferential use of presumed anti-CaM drugs to study biological systems.

Calmodulin and the plasma membrane calcium pump.

The data summarized and presented in this paper are consistent with the interpretation that CaM participates in the regulation of the plasma membrane calcium pump. Certain drugs, such as phenothiazines can antagonize CaM. Ca2+ loading of RBCs promotes CaM binding to RBC membranes and results in decreased responsiveness of the [Ca2+ + Mg2+)-ATPase to CaM. The latter effect may be mediated by a Ca2+ activated transglutaminase. Activation of (Ca2+ + Mg2+)-ATPase by CaM in vitro was shown not to be instantaneous, probably because of slow binding. CaM binding to isolated RBC membranes exhibits a Ca2+ dependence that is similar to that for activation of the (Ca2+ + Mg2+)-ATPase, and CaM binding does not decrease at high [Ca2+]s. Calculations based on assumed values for RBC [Ca2+], [CaM], and binding affinities of Ca2+ for CaM and CaM(Ca2+)n for the Ca2+ pump ATPase resulted in the tentative conclusion that most pump sites are occupied by CaM in the RBC in vivo. This conclusion, and the relatively slow time course of the CaM effect on (Ca2+ + Mg2+)-ATPase prompt us to suggest that, for all practical purposes, CaM is a subunit of the Ca2+ pump ATPase in vivo.