Functional reconstitution of the bovine brain GABAA receptor from solubilized components

The GABAA/benzodiazepine receptor has been solubilized from membrane preparations of bovine cerebral cortex and has been reconstituted, in a functionally active form, into phospholipid vesicles. In preliminary experiments, the receptor was labeled with the photoactive benzodiazepine [3H]flunitrazepam prior to solubilization. A peptide of apparent molecular weight 53,500 was specifically labeled by this method, and this was used as a marker for the receptor during the reconstitution procedures. The labeled protein was solubilized with approximately 40% efficiency by 1% beta-octyl glucoside. Reconstitution was achieved by mixing the solubilized proteins with a 4:1 mixture of soybean asolectin and bovine brain phospholipids, followed by chromatography on Sephadex G-50-80 to remove detergent. The incorporation of the GABAA receptor into membrane vesicles has been verified by sucrose gradient centrifugation in which the [3H]-flunitrazepam-labeled peptide comigrated with [14C]phosphatidylcholine used as a lipid marker. Vesicles prepared without labeled markers retained the ability to bind both [3H]flunitrazepam and the GABA analogue [3H]muscimol. Furthermore, the binding parameters were very similar to those measured using native membrane preparations. A novel fluorescence technique has been used to measure chloride transport mediated by the GABAA receptor in reconstituted vesicles. Chloride influx was rapidly stimulated in the presence of micromolar concentrations of muscimol and was blocked by preincubation of the membranes with muscimol (desensitization). Flux was also blocked by pretreatment with the competitive GABAA receptor blocker bicuculline or with the noncompetitive GABAA receptor antagonist picrotoxin.     

Functional reconstitution of a GABAA receptor purified from bovine brain.

The GABAA/benzodiazepine receptor has been solubilized from bovine brain membranes and purified by benzodiazepine affinity chromatography. Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate revealed two major protein species of 53 and 56 kDa. The purified protein has been reconstituted, in a functionally active form, into phospholipid vesicles. Chloride flux responses of the reconstituted preparations were investigated in stopped-flow experiments by monitoring fluorescence changes of a chloride-sensitive dye trapped within the vesicles. Flux was rapidly stimulated by muscimol and this response was potentiated by diazepam and blocked by desensitization of the receptor and by preincubation with the channel blocker, picrotoxin.     

Functional reconstitution of the glycine receptor

The functional reconstitution of the chloride channel coupled glycine receptor is described. Glycine receptors were purified from the cholate extract of rat spinal cord membranes by affinity chromatography and incorporated into phospholipid vesicles by the addition of phosphatidylcholine and removal of detergent by gel filtration. The reconstituted vesicles showed the same polypeptide composition as the purified receptor (proteins of Mr 48,000 and 58,000). The pharmacological characteristics of the glycine receptor were also preserved in the proteoliposomes, as demonstrated by the displacement of [3H]strychnine binding by several glycinergic ligands and by photoaffinity labeling experiments. In order to observe functional responses (i.e., specific agonist-induced anion translocation), we have developed an assay based on the fluorescence quenching of an anion-sensitive entrapped probe, SPQ [6-methoxy-N-(3-sulfopropyl)quinolinium]. Reconstituted vesicles were loaded with the fluorescent probe during a freeze-thaw-sonication cycle in the presence of added liposomes containing cholesterol. In such a reconstituted system, glycine receptor agonists are able to increase the rate of anion influx into the vesicles. The action of agonists is blocked by the simultaneous presence of strychnine or other glycine antagonists. Our results show that the purified 48,000- and 58,000-dalton polypeptides reconstituted into phospholipid vesicles can bind ligands and promote specific ion translocation in a way similar to the glycine receptor in its native environment.     

GABA-stimulated 36Cl- influx into reconstituted vesicles with purified GABAA/benzodiazepine receptor complex

Solubilized and Purified gamma-aminobutyric acid (GABA)A receptors from membrane vesicles of the bovine cerebral cortex were reconstituted into phospholipid vesicles and 36Cl- influx into the vesicles was examined. GABA induced a significant stimulation of the 36Cl- influx into reconstituted vesicles with 1.5% CHAPS/0.15% asolectin solubilized receptor and flunitrazepam further enhanced the GABA-stimulated influx. The purification of GABAA/benzodiazepine receptor complex and Cl- channel solubilized by 1.5% CHAPS/0.15% asolectin from membrane vesicles was achieved by 1012-S affinity column chromatography. The reconstituted vesicles with the purified receptor complex and Cl- channel also exhibited GABA-stimulated 36Cl- influx. This GABA-stimulated influx of 36Cl- was also enhanced by flunitrazepam, while suppressed by bicuculline, a GABAA receptor antagonist. These results strongly suggest that GABAA receptor is directly coupled with Cl- channel, whereas benzodiazepine receptor may be functionally coupled with GABAA receptor and modulates the GABA-stimulated Cl- influx through GABAA receptor. The present results also indicate that the purified GABAA receptor complex is coupled with Cl- channel and possesses functional characteristics as GABAA receptor.     

Reconstitution of the rat liver vasopressin receptor coupled to guanine nucleotide-binding proteins

The V1 vasopressin receptor has been solubilized from rat liver membranes with the zwitterionic detergent 3-[(3-cholamidopropyl)dimethylammoniol]-1-propanesulfonate (CHAPS) and reconstituted into phospholipid vesicles. There is essentially complete solubilization of the receptor by 3% CHAPS at a protein concentration of 15 mg/ml. Reconstitution into soybean phospholipid vesicles is readily achieved either by gel filtration chromatography or by membrane dialysis. The binding of [3H]vasopressin to proteoliposomes is specific, saturable, reversible, and magnesium-dependent. In contrast, the detergent-soluble vasopressin receptor does not display specific binding. The apparent affinity of the reconstituted receptor for [3H]vasopressin is approximately 4-fold lower than that of the receptor in native membranes. In addition, the binding of [3H]vasopressin to reconstituted vesicles is not sensitive to 100 microM guanosine 5'-O-thiotriphosphate (GTP gamma S) as it is in native membranes. However, the apparent affinity of the reconstituted receptor for ligand approximates that of native membranes when membranes are prebound with vasopressin prior to solubilization and reconstitution into vesicles. Furthermore, vesicles reconstituted from membranes prebound with vasopressin show GTP gamma S sensitivity of [3H] vasopressin binding. This finding strongly suggests that vasopressin stabilizes a receptor-G-protein complex during solubilization. The rat liver vasopressin receptor is a glycoprotein, as shown by its specific binding to the lectin "wheat germ agglutinin." The vasopressin receptor can be reconstituted from the N-acetylglucosamine-eluted peak of a wheat germ agglutinin-Sepharose column, and [3H] vasopressin binding activity is purified 5-6-fold from membranes by this chromatographic procedure. The functionality of the partially purified receptor is indicated by its ability to bind ligand with high affinity and by its ability to functionally interact with a G-protein when vasopressin is bound prior to solubilization.     

Amoxapine inhibition of GABA-stimulated chloride conductance: investigations of potential sites of activity

Amoxapine inhibits GABA-stimulated chloride conductance by acting on the GABAA-receptor chloride-ionophore complex which can be studied using membrane vesicles prepared from rat cerebral cortex. Amoxapine produces a right shift in the GABA concentration-response curve for the stimulation of 36Cl- uptake into these vesicles with no apparent change in the maximum response. Schild analysis of these data gave a pA2 value of 5.52 with a slope of 0.79. Amoxapine inhibits the binding of the GABAA receptor selective antagonist [3H]SR 95531 with an IC50 value of 3.45 microM and a pseudo Hill coefficient of 0.83. In contrast, 10 microM amoxapine inhibits [3H]flunitrazepam binding by less than 25% while the benzodiazepine antagonist Ro 15-1788 reduces the amoxapine inhibition of GABA-stimulated chloride conductance only at high concentrations. These data suggest that amoxapine does not inhibit chloride conductance by acting as a benzodiazepine inverse agonist and either acts directly on the GABAA receptor as an antagonist or blocks GABA activity at a site closely coupled to it. The ability of amoxapine to inhibit GABA-stimulated chloride conductance is a likely explanation for its proconvulsant activity observed at high doses.     

Rapid incorporation of the solubilized dihydropyridine receptor into phospholipid vesicles

We describe the rapid incorporation of the CHAPS solubilized dihydropyridine receptor into phospholipid vesicles. A series of sucrose gradient sedimentation experiments demonstrate that the (+)-[3H]PN200-110-labeled dihydropyridine receptor is associated with lipid vesicles following detergent removal by Extracti-gel chromatography. Solubilization of the receptor results in a loss of (+)-[3H]PN200-110 binding affinity relative to that observed in native membranes; the high affinity binding of (+)-[3H]PN200-110 can be restored upon reincorporation of the receptor into phospholipid vesicles. Similarly, the incorporation of the receptor restores its stability to incubation at 37 degrees C relative to that of the detergent solubilized receptor, thereby mimicking the properties of the membrane bound form of the receptor. The dissociation rate of (+)-[3H]PN200-110 from the reconstituted receptor is shown to be allosterically regulated by verapamil and diltiazem, indicating that the binding sites for these calcium antagonists have been inserted along with the dihydropyridine receptor into phospholipid vesicles. The results presented in this report, thus demonstrate the successful reconstitution of the dihydropyridine receptor into phospholipid vesicles by a variety of criteria. The reconstitution method described here is rapid and efficient, and should now facilitate structure-function studies of this receptor and its interrelationships with other regulatory components of the voltage-sensitive calcium channel system.     

Characterization of the solubilized and reconstituted ATP-dependent vesicular glutamate uptake system

We have previously provided evidence for ATP-dependent glutamate uptake into synaptic vesicles, and, based upon the unique properties of the vesicular uptake system, we have proposed that the vesicular glutamate translocator plays a crucial role in selecting glutamate for neurotransmission. In this study, we have solubilized the vesicular glutamate uptake system, proposed to consist of at least a glutamate translocator and a proton pump Mg-ATPase, from rat brain synaptic vesicles, and reconstituted the functional ATP-dependent glutamate uptake system into liposomes. The glutamate uptake in the reconstituted system is dependent upon ATP, markedly potentiated by low millimolar concentrations of chloride and inhibited by agents known to dissipate electrochemical proton gradients. Moreover, it exhibited low affinity for glutamate (Km = 2 mM), yet high specificity for glutamate; thus, it did not recognize aspartate and other agents known to interact with glutamate receptors. These properties are indistinguishable from those observed in intact synaptic vesicles. The solubilized functional components of the glutamate uptake system, alone or as a complex, have been estimated to have a Stokes radius in the range of 69 to 84 A. The reconstitution experiments described here provide a functional assay for the solubilized vesicular glutamate uptake system and represent an initial step towards the purification of the glutamate translocator.     

Isolation and reconstitution of the N-formylpeptide receptor from HL-60 derived neutrophils

A multifunctional receptor for N-formylpeptides exists on the membranes of neutrophils. This receptor has now been isolated from neutrophils derived from HL-60 promyelocytic leukemia cells. After solubilization by Nonidet-P40 and purification by affinity chromatography and HPLC the isolated receptor was reconstituted into egg phosphatidylcholine vesicles by SM-2 Bio-Bead removal of the Nonidet-P40. Analysis of the affinity and selectivity of the receptor was done by direct binding of two high-affinity ligands, formyl-Met-Leu-[3H]Phe-OH and formyl-Nle-Leu-Phe-[3H]Tyr-OH. The data suggest that the receptor can be isolated and reconstituted without apparent alteration of its binding affinity and selectivity, and that there appear to be no co-factors or subunits upon which these binding characteristics are dependent.     

Reconstitution of pure acetylcholine receptor in phospholipid vesicles and comparison with receptor-rich membranes by the use of a potentiometric dye

Acetylcholine receptor, isolated in Triton X-100 on a cobra alpha-neurotoxin affinity column was incorporated into unilamellar phospholipid vesicles by a detergent depletion method using Amberlite XAD-2. Vesicles of an average diameter of 25 nm were formed, as verified by freeze-fracture electron microscopy and gel filtration. 85 to 95% of the alpha-bungarotoxin binding sites of the reconstituted acetylcholine receptor were oriented towards the outside of the vesicles. In the reconstituted receptor one molecule of residual Triton X-100 per 2.5 alpha-bungarotoxin binding sites on the receptor molecule could be assessed. The reconstituted protein was not accessible to papain digestion, whereas the pure acetylcholine receptor, solubilized by Triton X-100 was split into smaller polypeptides under the same condition. Reconstituted acetylcholine receptor and receptor-rich membranes did not exhibit the same behavior as measured by use of a potentiometric dye. This is interpreted as an irreversible alteration of at least 95% of the receptors purified in the presence of Triton X-100. Furthermore, it could be shown that the fluorescence intensity changes induced by carbamylcholine in receptor-rich membranes did not reflect ion fluxes, but conformational changes of the protein or a displacement of the dye from the protein.     

Stimulation of 32Pi transport into human erythrocyte ghosts and reconstituted vesicles by Mg2+ and hemoglobin

Human erythrocyte ghosts catalyze a low rate of 32Pi uptake. A severalfold stimulation of 32Pi uptake was observed after exposure of the membranes to an erythrocyte lysate or to hemoglobin in the presence of Mg2+. Ghosts prepared from erythrocytes that had been exposed to 10 microM 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid showed a marked reduction in 32Pi uptake. Reconstitution of membranes with added phospholipids by freezing and thawing, by octylglucoside dilution or by cholate dialysis, yielded vesicles that catalyzed 32Pi uptake. When membranes were incubated with hemoglobin and Mg2+ prior to reconstitution, the rate of uptake was increased severalfold. The inhibition of hemoglobin and Mg2+-dependent uptake of 32Pi by chloride suggests that the transport in the reconstituted vesicles is catalyzed by the classical inorganic anion transporter.     

Glutamate uptake by brain synaptic vesicles. Energy dependence of transport and functional reconstitution in proteoliposomes

The dependence of glutamate uptake on ATP-generated proton electrochemical potential was studied in a highly purified preparation of synaptic vesicles from rat brain. At low chloride concentration (4 mM), the proton pump present in synaptic vesicles generated a large membrane potential (inside-positive), associated with only minor acidification. Under these conditions, the rate of L-[3H]glutamate uptake was maximal. In addition, L-glutamate induced acidification of the vesicle interior. D-Glutamate produced only 40% of the effect, and L-aspartate or gamma-aminobutyric acid produced less than 5%. The initial rate of glutamate-induced acidification increased with increasing glutamate concentration. It was saturable and showed first-order kinetics (KM = 0.32 mM). Correspondingly, L-glutamate induced a small reduction in the membrane potential. The rate of ATP hydrolysis was unaffected. In comparison, glutamate had no effect on acidification or membrane potential in resealed membranes of chromaffin granules. At high chloride concentration (150 mM), the vesicular proton pump generated a large pH difference, associated with a small change in membrane potential. Under these conditions, uptake of L-[3H]glutamate by synaptic vesicles was low. For reconstitution, vesicle proteins were solubilized with the detergent sodium cholate, supplemented with brain phospholipids, and incorporated into liposomes. Proton pump and glutamate uptake activities of the proteoliposomes showed properties similar to those of intact vesicles indicating that the carrier was reconstituted in a functionally active form. It is concluded that glutamate uptake by synaptic vesicles is dependent on the membrane potential and that all components required for uptake are integral parts of the vesicle membrane.     

Modulation of gastric mucosal calcium channels activity by platelet-derived growth factor.

A dihydropyridine-sensitive gastric mucosal calcium channels were isolated from the solubilized epithelial cell membranes by affinity chromatography on wheat germ agglutinin. The channels following labeling the calcium antagonist receptor site with [3H]PN200-100 were reconstituted into phospholipid vesicles which exhibited active 45Ca2+ uptake as evidenced by La3+ displacement assays. The uptake of calcium was independent of sodium and potassium gradients indicating the electroneutral nature of the process. The channels responded in a dose dependent manner to dihydropyridine calcium antagonist, PN200-110, which at 0.5 microns exerted maximal inhibitory affect of 66% on 45Ca2+ uptake, while a 52% enhacement in 45Ca2+ uptake occurred with a specific calcium channel activator, BAY K8644. On platelet-derived growth factor (PDGF) binding in the presence of ATP, channel protein showed an increase in tyrosine phosphorylation of 55 and 170 kDa calcium channel proteins. Such phosphorylated channels following reconstitution into vesicles displayed a 78% greater 45Ca2+ uptake. The results demonstrate the importance of PDGF in the regulation of gastric mucosal calcium uptake.     

Reconstitution of neurotoxin-stimulated sodium transport by the voltage-sensitive sodium channel purified from rat brain

Incorporation of the saxitoxin receptor of the sodium channel solubilized with Triton X-100 and purified 250-fold from rat brain into phosphatidylcholine vesicles is described. Fifty to 80% of the saxitoxin receptor sites are recovered in the reconstituted vesicles (KD = 3 nM). Unlike the detergent-solubilized saxitoxin receptor, the reconstituted saxitoxin binding activity is stable to incubation at 36 degrees C. Approximately 75% of the reconstituted saxitoxin receptor sites are externally oriented and 25% are inside-out. The initial rate of 22Na+ uptake into reconstituted vesicles is increased up to 3- to 4-fold by veratridine with a K0.5 of 11 microM. Seventy per cent of this increase is blocked by external tetrodotoxin (TTX) with a Ki of 10 nM. All of the veratridine-stimulated 22Na+ uptake is blocked when TTX is present on both sides of the vesicle membrane, or when tetracaine is added to the external medium. The apparent binding constants for veratridine, saxitoxin, and TTX are essentially identical to those in intact rat brain synaptosomes. The results demonstrate reconstitution of sodium transport, as well as neurotoxin binding and action, from substantially purified sodium channel preparations.     

The tricarboxylate carrier from rat liver mitochondria. Purification, reconstitution and kinetic characterization

The tricarboxylate carrier from rat liver mitochondria has been purified and reconstituted into phospholipid vesicles. Its activity has been characterized by both a radioactive citrate uptake assay and a coupled enzymatic assay. A Km of 40 microM and a Vmax of 1.56 mumol x min-1 x mg-1 have been determined for the carrier. Cholesterol levels of between 5-10% of total lipid content are shown to cause a decrease in carrier activity.     

Reconstitution of membrane proteins into giant unilamellar vesicles via peptide-induced fusion

In this work, we present a protocol to reconstitute membrane proteins into giant unilamellar vesicles (GUV) via peptide-induced fusion. In principle, GUV provide a well-defined lipid matrix, resembling a close-to-native state for biophysical studies, including optical microspectroscopy, of transmembrane proteins at the molecular level. Furthermore, reconstitution in this manner would also eliminate potential artifacts arising from secondary interactions of proteins, when reconstituted in planar membranes supported on solid surfaces. However, assembly procedures of GUV preclude direct reconstitution. Here, for the first time, a method is described that allows the controlled incorporation of membrane proteins into GUV. We demonstrate that large unilamellar vesicles (LUV, diameter 0.1 microm), to which the small fusogenic peptide WAE has been covalently attached, readily fuse with GUV, as revealed by monitoring lipid and contents mixing by fluorescence microscopy. To monitor contents mixing, a new fluorescence-based enzymatic assay was devised. Fusion does not introduce changes in the membrane morphology, as shown by fluorescence correlation spectroscopy. Analysis of fluorescence confocal imaging intensity revealed that approximately 6 to 10 LUV fused per microm(2) of GUV surface. As a model protein, bacteriorhodopsin (BR) was reconstituted into GUV, using LUV into which BR was incorporated via detergent dialysis. BR did not affect GUV-LUV fusion and the protein was stably inserted into the GUV and functionally active. Fluorescence correlation spectroscopy experiments show that BR inserted into GUV undergoes unrestricted Brownian motion with a diffusion coefficient of 1.2 microm(2)/s. The current procedure offers new opportunities to address issues related to membrane-protein structure and dynamics in a close-to-native state.     

Functional incorporation and preferred orientation of phytohemagglutinin receptor glycoproteins in phospholipid vesicles

Porcine lymphocyte Phaseolus vulgaris phytohemagglutinin (PHA) receptor glycoproteins purified by affinity chromatography have been reassembled into vesicles made of phosphatidylcholine and phosphatidylserine by detergent (dodecyltrimethylammonium bromide) dialysis. The receptor glycoproteins were incorporated into the lipid vesicles in a nonselective manner with a yield of 65-70%. Vesicles containing the glycoproteins were sealed as evidenced by their impermeability to calcium ions, using quin 2 trapped inside the vesicles. The vesicles were agglutinated by PHA, suggesting that the saccharidic moiety of the reconstituted glycoproteins was, at least in part, oriented towards the extravesicular medium. This observation was further supported by the fact that the vesicles bound 125I-labeled PHA in a specific and saturable manner. At maximum amount of lectin bound, a ratio of 1.01 +/- 0.05 microgram of PHA per microgram glycoprotein incorporated was measured. When the binding data were analyzed by Scatchard plot, a downward concave profile was observed, suggestive of a positive cooperativity at low concentrations of lectin. The orientation of the reconstituted lectin receptor glycoproteins was determined by proteolytic treatments of labeled glycoproteins. The combined action of trypsin and chymotrypsin released, in the 120,000 X g supernatant, approximately 80% of label when 125I-tagged PHA receptor glycoproteins were incorporated into the vesicles. When the oligosaccharidic moieties of the receptor glycoproteins were specifically labeled, the simultaneous action of the two enzymes released approximately 70% of tritium labeling present in the reconstituted system. Taken together, these results suggest that the reconstituted PHA receptors are preferentially oriented into the phospholipid vesicles. The reconstituted PHA receptor glycoproteins competed effectively with cellular receptors in the assay of PHA-induced porcine lymphocyte activation. A 50% inhibition of [3H]thymidine incorporation was observed when 1 microgram of glycoproteins in vesicles was added to the cultured cells, whereas vesicles alone had no effect at this (equivalent) concentration.     

Cerebellar GABAB receptors modulate function of GABAA receptors.

Interactions between GABAA and GABAB receptors were studied using muscimol-stimulated uptake of 36Cl- by membrane vesicles from mouse cerebellum. Baclofen inhibited muscimol-stimulated 36Cl- uptake and this action was more pronounced with longer flux times (30 vs. 3 s) and after predesensitization of GABAA receptors. Baclofen also inhibited 36Cl- flux by cortical membranes but was more effective with cerebellar preparations. The action of baclofen was stereoselective, calcium-dependent, and blocked by the GABAB receptor antagonist 2-OH-saclofen. It was mimicked by GTP-gamma-S but not by GDP-beta-S, which suggests that baclofen may be acting via a G protein. The action of baclofen was inhibited by U73122, an inhibitor of phospholipase C. However, the potassium channel blockers tetraethylammonium or Ba2+ did not affect the action of baclofen. The results show that activation of GABAB receptors can inhibit the function of GABAA receptors and suggest that this action involves either a nondesensitizing subtype of GABAA receptor or the rate or recycling of desensitized to nondesensitized receptors. We speculate that this action of baclofen results from activation of phospholipase C and phosphorylation of a subtype of GABAA receptor by protein kinase C.     

Novel ATP-dependent calcium transport component from rat liver plasma membranes. The transporter and the previously reported (Ca2+-Mg2+)-ATPase are different proteins

An ATP-dependent calcium transport component from rat liver plasma membranes was solubilized by cholate and reconstituted into egg lecithin vesicles by a cholate dialysis procedure. The uptake of Ca2+ into the reconstituted vesicles was ATP-dependent and the trapped Ca2+ could be released by A23187. Nucleotides, including ADP, UTP, GTP, CTP, GDP, AMP, and adenyl-5'-yl beta, gamma-imidophosphate, and p-nitrophenylphosphate did not substitute for ATP. The concentration of ATP required for half-maximal stimulation of Ca2+ uptake into the reconstituted vesicles was 6.2 microM. Magnesium was required for calcium uptake. Inhibitors of mitochondrial calcium-sequestering activities, i.e. oligomycin, sodium azide, ruthenium red, carbonyl cyanide p-trifluoromethoxyphenylhydrazone, and valinomycin did not affect the uptake of Ca2+ into the vesicles. In addition, strophanthidin and p-chloromercuribenzoate did not affect the transport. Calcium transport, however, was inhibited by vanadate in a concentration-dependent fashion with a K0.5 of 10 microM. A calcium-stimulated, vanadate-inhibitable phosphoprotein was demonstrated in the reconstituted vesicles with an apparent molecular weight of 118,000 +/- 1,300. These properties of Ca2+ transport by vesicles reconstituted from liver plasma membranes suggest that this ATP-dependent Ca2+ transport component is different from the high affinity (Ca2+-Mg2+)-ATPase found in the same membrane preparation (Lotersztajn, S., Hanoune, J. and Pecker, F. (1981) J. Biol. Chem. 256, 11209-11215; Lin, S.-H., and Fain, J.N. (1984) J. Biol. Chem. 259, 3016-3020). When the entire reconstituted vesicle population was treated with ATP and 45Ca in a buffer containing oxalate, the vesicles with Ca2+ transport activity could be separated from other vesicles by centrifugation in a density gradient and the ATP-dependent Ca2+ transport component was purified approximately 9-fold. This indicates that transport-specific fractionation may be used to isolate the ATP-dependent Ca2+ transport component from liver plasma membrane.     

Purification and reconstitution of the Ca2+-ATPase from plasma membrane of pig erythrocytes.

The Ca2+-ATPase from plasma membranes of pig erythrocytes was purified by mixed micelle gel chromatography (Wolf, H.U., Diekvoss, G., and Lichtner, R. (1977) Acta Biol. Med. Germ. 36, 847-858). The enzyme was activated at high concentrations of Tween 20 (10 mg/ml) or by appropriate mixtures of Triton X-100 and phospholipids. It was highly unstable in the absence of Ca2+ and activator protein. The Ca2+-ATPase was incorporated into liposomes by freeze-thaw sonication. After removal of non-ionic detergent by passage through a phenyl Sepharose 4B column, the reconstituted vesicles catalyzed a rapid ATP-dependent uptake of Ca2+. Modulator protein from brain substituted for the natural activator protein and stimulated Ca2+ uptake in reconstituted vesicles.