Purification and ultrastructural properties of the calcium accumulating membranes in isolated sarcoplasmic reticulum preparations from skeletal muscle

Structural and mechanistic implications of the amino acid sequence of calcium-transporting ATPases

Work is reviewed in which the amino acid sequences of two Ca2+-transporting ATPases of sarcoplasmic reticulum (SR) from slow (or cardiac) and fast skeletal muscle were determined from the nucleotide sequences of cloned cDNAs. Analysis of hydrophobicity and secondary structure, combined with the known shape derived from electron micrographs, leads to a model of five domains with functional implications. The major globular part of the molecule is in the cytoplasm and consists of one antiparallel and two parallel beta-sheet domains. One of the latter binds ATP, which, in the presence of Ca2+, phosphorylates an aspartic acid on the other domain. It is proposed that subsequent kinase-like movements are transmitted to the SR membrane via a penta-helical, calcium-binding stalk. The Ca2+ is first trapped and then translocated via the ten helices which constitute the transmembrane (channel) region. The difference in requirements for counter ions between the Ca2+- and Na+/K+-ATPases can be explained in terms of differing charge distributions in this channel.

Positions of the sites labeled by N-cyclohexyl-N'-(4-dimethylamino-1-naphthyl)carbodiimide on the (Ca2+ + Mg2+)-ATPase

N-Cyclohexyl-N'-(4-dimethylamino-1-naphthyl)carbodiimide (NCD-4) labels (Ca2+ + Mg2+)-ATPase at Ca2+-protectable sites, believed to be at or near the two Ca2+ binding sites on the ATPase, and at nonspecific sites. The labeled ATPase has been reconstituted into lipid bilayers containing phosphatidylethanolamine labeled with fluorescein isothiocyanate. The distance between NCD-4 and fluorescein groups was measured using Forster energy transfer and the NCD-4 labels were found to be approx. 20 A from the lipid/water interface suggesting that the Ca2+ binding sites on the ATPase are also 20 A from the lipid/water interface. Addition of vanadate causes no change in the efficiency of energy transfer, suggesting that the Ca2+ binding sites on the E1 conformation of the ATPase do not move significantly with respect to the lipid/water interface in the E1-E2 transition.

Distances between the functional sites of sarcoplasmic reticulum (Ca2+ + Mg2+)-ATPase and the lipid/water interface

Measurements of fluorescence energy transfer have been performed to determine the distance between the lipid-water interface and the ATP-binding site in the (Ca2+ + Mg2+)-ATPase from sarcoplasmic reticulum. The calculated distance between the donor, FITC bound to the protein (nucleotide binding-site marker), and the acceptor, rhodamine-5'-isothiocyanyldipalmitoylphosphatidylethanolamine (RITC-DPPE) incorporated in the membrane, was in the range of 34-42 A. In addition the distance between the high affinity Ca2+-binding sites and the lipid/water interface has been calculated by luminescence energy transfer from Tb3+ bound to the Ca2+ sites to RITC-DPPE included in the membrane, and it was approx. 10 A.

Two Ca2+ ATPase genes: homologies and mechanistic implications of deduced amino acid sequences

Rabbit genomic DNA contains two genes that encode Ca2+ ATPases of fast twitch and of slow twitch (and cardiac) sarcoplasmic reticulum, respectively. The deduced amino acid sequences of the products of the two genes are highly conserved in putative Ca2+ binding regions, in sectors leading from cytoplasmic domains into transmembrane domains, and in transmembrane helices. A transport mechanism is proposed in which Ca2+ binds to negatively charged groups on amphipathic stalk sectors, becoming occluded during enzyme phosphorylation by bound ATP. Rotation of the stalk sectors is induced as the energy in the phosphorylated enzyme (E1P) is utilized in conformational changes leading to the low energy form, E2P. Rotation leads to disruption of high affinity Ca2+ binding sites and release of Ca2+ into a charge-lined membrane channel. Ca2+ then traverses the membrane by exchange diffusion.

Amino-acid sequence of a Ca2+ + Mg2+-dependent ATPase from rabbit muscle sarcoplasmic reticulum, deduced from its complementary DNA sequence

We have cloned and sequenced complementary DNA encoding a Ca2+-ATPase of rabbit muscle sarcoplasmic reticulum. We propose a model of the protein which has 3 cytoplasmic domains joined to a set of 10 transmembrane helices by a narrow, penta-helical stalk. In this model, ATP bound to one cytoplasmic domain would phosphorylate an aspartate in an adjoining cytoplasmic domain, inducing translocation of Ca2+ from binding sites on the stalk.

Localization of ATP-dependent calcium transport activity in mouse pancreatic microsomes

Electron-dense deposits representing calcium oxalate crystals which result from ATP-dependent calcium uptake have been localized within vesicles of of a heavy microsomal fraction prepared from mouse pancreatic acini. In the absence of either ATP or oxalate, no electron-dense deposits could be observed. By subfractionation of microsomes on discontinuous sucrose gradients, it could be shown that the highest energy-dependent calcium transport activity was associated with the rough endoplasmic reticulum. In rough microsomes, the 45Ca2+-uptake measured was 7 times greater than that of smooth microsomes in the presence of ATP and oxalate and about 3 times greater in he presence of ATP alone. When ribosomes were released from the rough endoplasmic reticulum vesicles by treatment with KCl in the presence of puromycin, the stripped microsomes showed a 40% increase in the specific 45Ca2+-uptake activity measured in he presence of ATP and oxalate and an increase of 80 to 90% in the presence of ATP alone. From these results it can be concluded that the calcium transport activity of microsomes prepared from mouse pancreatic acini is located predominantly in the rough endoplasmic reticulum membrane.

Preparation and characterization of muscarinic-acetylcholine-receptor-enriched membranes from pig atria

A procedure was developed for the large scale preparation of membranes from pig atria which are enriched 10-13 fold in the muscarinic acetylcholine receptor. The procedure involved differential centrifugation and sucrose-gradient centrifugation in solutions containing 150 mM-NaClO4 and 5 mM-EDTA to minimize membrane aggregation. The final membrane preparation bound about 1.1 pmol of L-quinuclidinyl benzilate/mg of protein. Comparable results were obtained with either fresh or frozen tissue. About the same yield (120 pmol of L-quinuclidinyl benzilate sites/100 g of tissue) and specific activity of membranes were obtained from different regions of the atria. The final preparation was stable at -80 degrees C in buffered sucrose solutions. The membranes appeared mostly as sheets or fragments and partly as closed vesicles in the electron microscope and were heterogeneous in isopycnic Percoll gradients. Marker enzyme studies showed that the receptor was enriched in parallel with the plasma membrane markers guanylate cyclase (particulate form) and (Na+ + K+)-activated ATPase. Some contamination by mitochondrial outer and endoplasmic reticulum membranes was evident from the distribution of monoamine oxidase and glucose-6-phosphatase activity, but the preparation was largely free of sarcoplasmic reticulum, mitochondrial inner, and lysosomal membranes.

Freeze-fracture study of water-soluble, standard proteins and of detergent-solubilized forms of sarcoplasmic reticulum Ca2+-ATPase

Conventional freeze-fracturing electron microscopy was used to study water-soluble proteins and different forms of Ca2+-ATPase-detergent complexes. Freeze-fracture images of solutions containing proteins larger than myoglobin showed the presence of distinct, randomly dispersed particles on smooth fracture surfaces. The distribution of sizes of these particles was closely to Gaussian, with a mean size which was correlated to the Stokes diameter. Monomeric Ca2+-ATPase from sarcoplasmic reticulum, solubilized by deoxycholate or a non-ionic detergent, showed a bimodal distribution of particle sizes. Even more complex distributions were found for dimeric and trimeric preparations of Ca2+-ATPase. The results can be interpreted on the assumption that the Ca2+-ATPase molecule is elongated, with an overall length of about 110 A and a width in its largest part of about 75 A. It is concluded on the basis of the presented results that freeze-fracture electron microscopy can be successfully used for morphological studies of protein molecules in solution.

Isopycnic-zonal centrifugation of plasma membrane, sarcoplasmic reticular fragments, lysosomes, and cytoplasmic proteins from phasic skeletal muscle

Homogenates of the posterior latissimus dorsi muscle, a phasic muscle, were fractionated by a one-step zonal centrifugation technique into four major organelle populations and cytoplasmic constituents. These were: (1) Plasma membrane fragments with a modal equilibrium density of 1.10 and containing 5'-nucleotidase, alkaline phosphodiesterase, p-nitrophenylphosphatase and acid phosphatase (beta-glycerophosphate was used as the substrate). (2) Sarcoplasmic reticular fragments which could be further subdivided into calcium transport vesicles, with a model equilibrium density of 1.16, that exhibited calcium uptake; K+-ATPase; leucyl-bet-naphthylamidase; acid phosphodiesterase; acid phosphatase (using cytidine monophosphate as the substrate); and sarcoplasmic reticular lysosomes, with a model equilibrium density of 1.18, possessing dipeptidyl-aminopeptidase II, cathepsin D, alpha-glucosidase, N-acetyl-beta-glucosaminidase, and NADH oxidase activity. (3) Mitochondria with a modal equilibrium density of 1.21. (4) Catalase-containing vesicles with a modal equilibrium density of 1.22; and cytoplasmic constituents (modal density of 1.25) with phosphorylase, pyruvate kinase, myosin-ATPase, aldolase, and protein and RNA content. The purity of these organelles was equal to or better than previous efforts, with a 30-fold purification achieved for 5'-nucleotidase and alkaline phosphodiesterase. These results lend support to the hypothesis that the sarcoplasmic reticulum of phasic muscle, in addition to its specialized role in excitation-contraction coupling, represents a multifunctional membrane system, and that, similar to the smooth endoplasmic reticulum of other cells, it includes some membrane-bound lysosomal enzymes and NADH oxidase.

Functional heterogeneity of the sarcoplasmic reticulum within sarcomeres of skinned muscle fibers

Precipitation of Ca oxalate in the sarcoplasmic reticulum of chemically skinned rabbit psoas fibers caused an increase in light scattering which was proportional to the amount of Ca accumulated per unit fiber volume. The increase in scattering was used to measure net accumulation rates and steady-state Ca capacities of the sarcoplasmic reticulum in single fibers. The data obtained were qualitatively and quantitatively similar to those reported for isolated vesicle preparations. Under conditions in which Ca was not depleted from the medium, Ca accumulation was linear with time over much of its course. Steady-state capacities were independent of the Ca concentration; uptake rates were half-maximal at 0.5 microM Ca++ and saturated above about 1.0 microM. Both rate and capacity varied with the oxalate concentration, being maximal at oxalate concentrations greater than or equal to mM and decreasing in proportion to one another at lower concentrations, with a threshold near 0.25 mM. At the lower loads, electron micrographs showed many sarcoplasmic reticulum elements empty of precipitate alongside others that were full, whereas virtually all were filled in maximally loaded fibers. These data indicate that the Ca oxalate capacity of each fiber varies with the number and volume of elements in which Ca oxalate crystals can form at a given oxalate concentration, and that individual regions of the sarcoplasmic reticulum within each sarcomere differ in their ability to support Ca oxalate precipitation. Our working hypothesis is that this range in ability to form Ca oxalate crystals involves differences in ability to accumulate and retain ionized Ca inside the sarcoplasmic reticulum.

Mass isolation of cell surface membrane fragments from pigeon heart

Cell surface membrane fragments were isolated and purified by successive rate zonal and isopycnic centrifugation of calcium oxalate-loaded pigeon heart microsomes in sucrose density gradients. The most highly purified cell membrane fraction sediments at a buoyant density of 1.105 g/ml. Some of the membrane pieces are present as open fragments and leaky vesicles, while others form tightly sealed vesicles of both inside-in and inside-out membrane orientation. The pigeon heart cell membrane preparation exhibits high (Na+ + K+ + Mg2+)-ATPase and adenylate cyclase activities. Additional activity of these enzymes is uncovered by sodium dodecyl sulfate and alamethicin, respectively. Electron microscopic inspection of the cell surface membrane preparation revealed (a) a predominance of thick-walled vesicles with smooth surfaces on negative staining and (b) binding of concanavalin A to the bulk of isolated membrane pieces following their incubation with the lectin.

Cholesterol in muscle membranes

The composition of skeletal muscle microsomes is reviewed. Evidence for the involvement of cholesterol in the transport of calcium by vesicles derived from the sarcoplasmic reticulum is considered. Results obtained by non aqueous extractions of skeletal muscle microsomes, and by use of the cholesterol analogue 20, 25 diazacholesterol indicate that cholesterol is not involved in calcium transport by vesicles of sarcoplasmic reticulum origin. Use of density perturbation procedures indicating that cholesterol is present in muscle membranes other than those of the sarcoplasmic reticulum involved in calcium transport is discussed. The distribution of membranal cholesterol in muscle is compared to that in other tissues.

Isolation of calcium pump system and purification of calcium ion-dependent ATPase from heart muscle

The procedure for the isolation of the highly active fraction of sarcoplasmic reticulum from pigeon and dog hearts is described. The method is based on the partial loading of heart microsomes with calcium and oxalate ions and the precipitation of loaded vesicles in sucrose and potassium chloride concentration gradients. Preparations obtained possess high activity of Ca2+-dependent ATPase and are also able to accumulate up to 10 mumol Ca2+ per mg protein. Purification of sarcoplasmic reticulum membranes is accompanied by a decrease in concentration of cytochrome a+a3 and an increase in the content of [32P]phosphoenzyme. The basic components in "calcium-oxalate preparation" from hearts are proteins with molecular weights of about 100000 (Ca2+-dependent ATPase) and 55000 Calcium-oxalate preparation from pigeon hearts was used for subsequent purification of Ca2+-dependent ATPase. Specific activity of purified enzyme from pigeon hearts is 12-16 mumol Pi/min per mg protein. Enzyme activity of purified Ca2+-dependent ATPase is inhibited by EGTA and is not sensitive to azide, 2,4-dinitrophenol and ouabain. The data obtained demonstrate the similarity of calcium pump systems and Ca2+-dependent ATPases isolated from heart and skeletal muscles.

Separation and characterisation of tryptic fragments from the adenosine triphosphatase of sarcoplasmic reticulum

The two halves of the ATPase, M, 115,000, from sarcoplasmic reticulum produ-ed by limited trypsin treatment have been purified in sodium dodecylsulphate. The fragment of Mr60,000 has been purified by electrophoresis on cellulose acetate slabs and that of Mr 55,000 by gel filtration. The two halves of the 60,000 Mr fragment (Mr33,000 and 24,000) produced by more extensive trypsin treatment have also been purified by gel filtration in sodium dodecylsulphate. The sum of the amino acid analyses of the constituent tryptic fragments is in good agreement with that for the whole ATPase. The amino acid compositions of the two halves of the ATPase were strikingly similar. N-terminal analysis shows that the ATPase and its constituent tryptic polypeptides all possess a single N-terminal alanine implying no further cleavage of the polypeptide by trypsin. Attempts to solubilize selectively the tryptic fragments from the membrane by a variety of denaturing and solubilising agents under a variety of conditions have proved unsuccessful, suggesting that the interaction between the tryptic polypeptides is stronger than between the lipid and the protein. The possibility that the interaction between the tryptic polypeptides includes disulphide bonding has been eliminated.

Changes in the activity of sarcoplasmic reticulum fragments and actomyosin isolated from skeletal muscle of thyroxine-treated cats

1. Sarcoplasmic reticulum fragments (SRF) and actomyosin were isolated from skeletal muscle of cats treated either with thyroxine or with placebo tablets, for 5-16 months.2. Ca(2+) uptake and binding by SRF, measured in the presence and absence of oxalate respectively, were reduced by thyroxine treatment.3. Actomyosin from the thyroxine-treated animals underwent ATP-induced syneresis at a faster rate than did that from the controls.4. Syneresis of the actomyosin preparations from controls and treated animals was inhibited to the same extent by EGTA when rates of syneresis were made the same by adjustment of the KCl concentration in the assay media. In contrast, at any given KCl concentration, syneresis of actomyosin from thyroxine-treated animals was inhibited to a lesser extent by EGTA.5. At the end of the isolation procedure, the amount of Ca(2+) remaining in the actomyosin suspension was similar for both treated and control animals.6. It was concluded that the effect of thyroxine on skeletal muscle may be the result of an action on both the sarcoplasmic reticulum and the contractile protein complex.

Sarcolemmal and sarcoplasmic reticular ATPase activities in the failing canine heart

The specific activity of both the sarcoleminal Na + -K + -activated and the sarcoplasmic reticular ATPase systems was studied in dogs with right ventricular hypertrophy and congestive heart failure induced by progressive pulmonary artery stenosis. In these enzyme preparations, mitochondrial or cross contamination of either of the cellular fractions studied did not appear to be a factor. In the sarcoplasmic reticular fraction, ouabain was without effect and succinic dehydrogenase assay showed less that 10% of the activity observed in pure mitochondrial preparations. In the sarcolemmal fraction, sodium azide did not significantly inhibit the ATPase activity, but ouabain inhibited 91% of it. The activity of the sarcolemmal ATPase system of the faihng heart did not differ from that of the control; however, the activity of failing heart sarcolemmal preparations was found to be only 61% inhibited by 10 -4 M ouabain, and the activity of control preparations was 91% inhibited. The sarcoplasinic reticular ATPase activity from right ventricles of dogs with congestive heart failure was significantly less ( P <0.001) than the respective control activity (average 15.9 compared to 28.1 µmole hr -1 mg -1 ), whereas the activity from left ventricles of failing hearts did not differ significantly from the respective control value. These findings suggest that a possible explanation for the diminished contractility of the chronically failing heart lies within the subcellular enzymatic ion-regulating systems.