Cyclic AMP-dependent Phosphorylation of an Immunoaffinity-purified Homotetrameric Inositol 1,4,5-Trisphosphate Receptor (Type I) Increases Ca2+ Flux in Reconstituted Lipid Vesicles*

We established a novel method to isolate a single type of inositol 1,4,S-trisphosphate receptor (IPsR) among the heterogeneous population of receptors to study the regulatory mechanism of Ca” release. We raised in the rabbit a polyclonal antibody against synthetic peptide corresponding to amino acids of type I IPsR (IPsR-I) is most abundant in cerebellum. We purified IPsR-I from a 1% 3-[(3-cholamido- propyl)dimethylammonio]- 1-propanesulfonic acid solubilized mouse cerebellar microsomal fraction by immunoaffinity chromatography on an anti-pep 6 an-tibody-Sepharose 4B column with specific elution by the pep 6 peptide (GHPPHMNVNPQQ) of the IPsR-I C

much of the IP3R structure. IP3R has a transmembrane domain near the C terminus with a large N terminus and short C terminus in the cytoplasmic compartment. The cloning of mouse, rat (6), Drosophila (7), Xenopus (8), and human cDNA showed that the general structure of the receptor is highly conserved. Reconstitution of the purified IP3R into planar lipid bilayer and functional expression of the receptor in L cells by transfecting a plasmid vector containing mouse IP3R cDNA showed that IP3R is a Ca2+ release channel (9,lO). IP3induced Ca2+ release (IICR) is modulated by various endogenous mediators such as ATP (9,11,12), GTP (13), fatty acids (14), and Ca2+ (15, 16). Phosphorylation is one of the very important mechanisms regulating cellular function. Cyclic AMP-dependent protein kinase (PKA) predominantly phosphorylates the receptor (17,18). IP3R contains two consensus amino acid sequences that fulfill the criteria for PKA action (5), and these were shown to be phosphorylated by PKA both in vitro (17) and in intact cells (18). Supattapone et al. (17) and Quinton and Dean (19) reported that phosphorylation by PKA of cerebellar membranes or platelet membranes substantially reduces the potency of IP3 in releasing Ca2+. However, there is much evidence that phosphorylation of the IP3R by PKA increases IICR in platelets (20), hepatocytes (21), and cerebellar membranes (22). The real nature of IP3R regulation by PKA has remained elusive.
These studies were accomplished by using a crude microsomal system. Phosphorylation by PKA may alter the components of the calcium regulation system such as the calcium pump or IP3 formation by phospholipase C in the membranes or some other regulatory proteins, making interpretation of the experimental results complicated and confusing. The heterogeneity of IP3R was reported to result from different genes (23,24), leaving the possibility that the effect of mediators on IICR might be different depending on the type of IP3R. Actually, one of the new types of receptor (type 11) lacks the consensus sequences needed for phosphorylation by PKA (23). Ferris et al. (25) recently purified IP3R from rat cerebellum by conventional methods and reconstituted them into lipid vesicles. However, the receptors were composed of heterogeneous types. In order to clarify the real function of PKA phosphorylation, it is crucial to purify a single type of receptor and reconstitute it into liposomes or a lipid bilayer to analyze the effect on Ca2+ release. We developed a novel purification method for the isolation of a single type of IP3R (IP3R-I) by immunoaffinity chromatography using a subtype-specific antibody and reconstituted it into lipid vesicles after dialysis. We found that the Ca2+-releasing activity of IP3R-I is enhanced by PKA phosphorylation. In addition, we characterized some biochemical properties of IP3R-I.

EXPERIMENTAL PROCEDURES
Materiuls--IP3 and CHAPS were obtained from Dojindo Laboratories (Kumamoto, Japan). [3H]IP3 and "Ca2+ were obtained from DuPont NEN. [y3'P]ATP (6000 Ci/mmol), ECL Western blotting system, and Hyper film-ECL were obtained from Amersham Corp. Ethylene glycol-bis(sulfosuccinimidy1 succinate) (S-EGS) was obtained from Pierce Chemical Co. Phosphatidylcholine and phosphatidylserine were obtained from Avanti Polar Lipids. Chelex 100 and Affi-Gel 10 were obtained from Bio-Rad. Sephadex G-50, CNBractivated Sepharose 4B, and protein A-Sepharose CL-4B were obtained from Pharmacia LKB Biotechnology Inc. CAMP-dependent protein kinase catalytic subunit (2500 units) and protein kinase inhibitor were obtained from Sigma. All of the other chemicals were of the highest purity commercially available.
Production of Antisera against a Synthetic ZP& C-terminal Peptide and Purified ZP& Protein-A peptide corresponding to the C-terminal region of the IPIR-I (amino acid residues 2736-2747 designated pep 6, GHPPHMNVNPQQ) were custom-synthesized on an Applied Biosystems synthesizer, model 430A. The peptide was conjugated to bovine serum albumin (BSA) via l-ethyl-3-(3-dimethylaminopro-py1)carbodiimide (EDC) as described by Richardson et al. (26). Briefly, 20 mg of pep 6 and 20 mg of BSA were dissolved in 5 ml of phosphate-buffered saline (PBS), pH 7.3, and 20 mg of EDC was added at 4 "C with constant stirring. The mixture was stirred overnight. The remaining unreacted EDC and pep 6 were separated by gel filtration on a Sephadex G-50 column equilibrated in 50 mM ammonium acetate. The fractions containing pep 6-BSA were collected, lyophilized, and then dissolved in PBS. New Zealand White rabbits were immunized by intradermal injection with a homogenate containing 1 ml of Freund's complete adjuvant and 1 ml of conjugate (200 pg of pep 6). Three weeks later, the rabbits were boosted with homogenate containing 1 ml of Freund's incomplete adjuvant and 1 ml of the conjugate (100 pg of pep 6). Antiserum was collected each week thereafter. Booster injection was performed every 2 weeks until the titer of the antiserum was saturated. Polyclonal antibody against conventionally purified IP3R from mouse cerebellum was raised in rabbits by the same injection schedule and by injecting the same amount of antigen as pep 6.
Preparation of Membranes-Adult ddY mice were anesthetized and then killed by decapitation, and the cerebella were dissected. The tissues were mixed with 9 volumes of the solution containing 0.32 M sucrose, 1 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride (PMSF), 10 p~ leupeptin, 10 NM pepstatin A, 1 mM 2-mercaptoethanol, and 5 mM Tris-HC1, pH 7.4, and were homogenized in a glass-Teflon Potter homogenizer with 10 strokes at 850 rpm. The homogenate was centrifuged at 1000 X g for 5 min at 2 "C, and the pellet was washed again under the same conditions. The combined supernatants were centrifuged at 105,000 X g for 60 min at 2 "C to precipitate the membrane fraction. The membrane fraction was resuspended in 1 mM EDTA, 1 mM 2-mercaptoethanol, 0.1 mM PMSF, 10 p M leupeptin, 10 p~ pepstatin A, and 50 mM Tris-HC1, pH 7.4 (buffer A) and then solubilized by the addition of 10% CHAPS to give final protein and detergent concentration of 3.0 mg/ml and 1.0%, respectively.
Zmmunoaffinity Purification of ZP& from Mouse Cerebella-The IgG fraction from the anti-pep 6 serum was purified by ammonium sulfate precipitation and dialyzed against 0.1 M NaHC03 and 0.5 M NaCl, pH 8.3, and then 10 mg of IgG were coupled to 1 ml of CNBractivated Sepharose 4B according to the manufacturer's protocol. To isolate IPsR, the membrane fraction was solubilized with buffer A containing 1% CHAPS and stirred for 30 min at 0 "C. The solution was centrifuged at 20,000 X g for 60 min at 2 "C, and the supernatant was incubated with the anti-pep 6 antibody-Sepharose 4B beads for 2 h at 4 "C under gentle stirring. The beads were then transferred into a column and washed with 50 ml of buffer A containing 1 mg of phosphatidylcholine/ml, and then IP3R was eluted by pep 6 at a concentration of 5 p~ in buffer A containing 1 mg of phosphatidylcholine/ml. IP3R was eluted by a batch method using 1 ml of eluting reagent/l ml of packed beads for 20 min at 4 "C under gentle stirring, and then eluting reagentbeads were centrifuged at 100 X g for 30 s. The supernatant was obtained as eluate. This elution step was repeated at least 5 times (eluates 1-5). The nonadsorbed fraction to the beads was reincubated with the regenerated beads to maximize binding of IP3R (cycles 1-3). The beads were regenerated with 1 M glycine and 1.5 M NaC1, pH 2.5. In each cycle IP3R was eluted by the method as described above.
Cross-linking of Denatured ZP& and Nondenatured ZPd2"Immunoaffinity-purified IP3R was denatured with 6 M urea, 20 mM dithiothreitol, and 1% SDS in buffer A containing 1% CHAPS for 5 min at 95 "C and dialyzed against PBS containing 1% CHAPS, 0.1 mM PMSF, 10 p~ leupeptin, and 10 p~ pepstatin A (buffer B) to remove Tris, urea, and dithiothreitol. The concentration of SDS was adjusted to 0.3% by dilution of the sample with PBS containing 0.1 mM PMSF, 10 pM leupeptin, and 10 p~ pepstatin A. The purified IP3R (nondenatured form) was dialyzed against buffer B. Aliquots (50 pl) of each solution were treated with 10 mM S-EGS. The reactions were allowed to proceed for 30 min at 0 "C and then were quenched by the addition of 10 pl of 50 mM Tris-HC1, pH 8.0. Then, 10 pl of sample was solubilized in agarose-PAGE buffer at a final concentration of 1% SDS, 1 mM EDTA, 5% 2-mercaptoethanol, 10 mM Tris-HC1, pH 8.0, 10% glycerol and heated in a boiling water bath for 3 min. The samples were analyzed with agarose-PAGE and immunoblotting (see below).
Antibody Purification by Affinity Chromatography-Peptide (10 mg) was coupled to 2 ml of Affi-Gel 10 in 0.1 M NaHC03, pH 8.0, with gentle agitation at 4 "C overnight. Unreacted coupling sites were then blocked by the addition of 0.1 ml of 1 M ethanolamine, pH 8.0, for 2 h at room temperature. Antibodies were first selected by passing 1.5 ml of antiserum, which was obtained from the rabbit injected with the pep 6-BSA conjugate, over 0.8 ml of a protein A-Sepharose CL-4B column that was equilibrated in PBS. The column was washed with PBS, and the bound antibody was eluted with 0.1 M citrate buffer, pH 2.5, containing 1 M NaCl, and then the eluate was dialyzed against PBS at 4 "C overnight. The antibody eluted from the protein A column was next subjected to peptide affinity column chromatography. Antibody in approximately 5 ml of PBS and peptide affinity gel (1 ml) were mixed and gently agitated at 4 "C overnight. This mixture was packed into a column. The column was washed with PBS, and the antibody was eluted with 0.1 M citrate buffer, pH 2.5, containing 1 M NaCl and dialyzed against PBS for three times every 8 h.
Immunoprecipitation-10 p1 of the pep 6 affinity-purified anti-pep 6 polyclonal antibody solutions (concentration of antibody is 150 pg/ ml) was added to the 100 pl of denatured IP3R solution (containing 0:3% SDS), which were then incubated for 2 h at 4 "C. 50 pl of protein A-Sepharose CL-4B suspension was added, and the sample tubes were gently rotated for 30 min at 4 "C. The protein A-Sepharose CL-4B particles were washed three times with 1 ml of buffer containing 50 mM Tris-HC1, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS. The pellets were mixed with 20 p1 of 4% SDS, 10% 2-mercaptoethanol, 20% glycerol, 0.125 M Tris-HC1, pH 6.8, and then were heated in a boiling water bath for 3 min. After centrifugation, the supernatants were subjected to 4-10% SDS-PAGE in the buffer system of Laemmli (27) and analyzed with immunoblotting using polyclonal antibody against conventionally purified IP3R protein.
Reconstitution of ZP& into Lipid Vesicle-Proteins in the immunoaffinity column eluates were mixed with phospholipid solution. Phosphatidylcholine and phosphatidylserine dissolved in chloroform were mixed at a concentration ratio of 3:l. The lipid mixture was dried to a thin film under a stream of nitrogen gas and desiccated to remove any residual traces of organic solvent for 20 min at room temperature. The lipid film was suspended at 1 mg/ml in buffer C (50 mM NaC1,50 mM KCl, and 20 mM Tris-HC1, pH 7.4) containing 1% CHAPS and the immunoaffinity-purified IP3R at a concentration of 5 pg/ml. The protein/lipid/detergent mixture (usually 1-ml volume) was incubated on ice for at least 20 min with occasional gentle swirling, and then dialysis was performed. The dialysis schedule followed Ferris et al. (25), namely a 1000-fold excess of buffer C containing 2 mM 2-mercaptoethanol, 1 mM EDTA, and 1 mM EGTA was changed every 8 h for 48 h. Finally, the dialysis buffer was changed to a 1000-fold excess of buffer C containing 2 mM 2-mercaptoethanol to remove EDTA and EGTA, and the dialysis was continued for an additional 24 h changing the buffer every 8 h. Proteoliposomes thus formed were collected by centrifugation at 100,000 X g for 20 min, and the resulting pellets were suspended with buffer C containing 2 mM 2-mercaptoethanol in the same volume of originally dialyzed sample. The suspension was subjected to '%a2+ uptake assay.
Measurement of %a2+ Uptake into Vesi~les--'~Ca~+ uptake assay was carried out by modifying the methods of Ferris et al. (25) and Nunoki et al. (28). Calcium influx was measured by incubating 25 p1 of reconstituted vesicles diluted with 75 pl of buffer C containing 2 mM 2-mercaptoethanol, 2 pCi of "%a2+, and monoclonal antibody (mAb) 18A10 or mAb 10A6 when needed. After incubation for 1 min at 30 "C, various concentrations of IP3 were added. The Caz+ channel activity was measured under countertransport conditions in which -'%a2+ was transported into liposome when the channel was opened by IP3. After another incubation for 1 min at 30 "C, the influx reaction was stopped with 300 p1 of buffer C supplemented with 0.3 mM CaC12, 5 mM MgS04, 1 mM CdC12, and 100 pg/ml heparin (stop buffer). This 400-pl mixture of vesicles and stop buffer was immediately loaded onto a 3-ml Chelex 100 column (Bio-Rad) preequilibrated with buffer containing 200 mM sucrose, 20 mM Tris-HC1 (pH 7.4 at 25 "C), and 0.3% BSA. The column was immediately washed with 5 ml of the buffer containing 200 mM sucrose, 20 mM Tris-HC1, pH 7.4, to wash the vesicles and the intravesicular "Ca2+ through the column. The "Can+ content of the vesicles was measured with a liquid scintillation counter.
Cross-linking of Proteoliposomes Reconstituted with IPJ?-lOO pl of the proteoliposomes was centrifuged at 100,000 X g for 20 min at 2 "C, and the resulting pellet was resuspended in 50 pl of 50 mM sodium phosphate, pH 8.0. The suspension was treated with various concentrations of cross-linker, S-EGS. The reactions were allowed to proceed for 30 min at 0 "C and then were quenched by the addition of 10 p1 of 50 mM Tris-HCI, pH 8.0. After that, 10 pl of sample was solubilized in agarose-PAGE sample buffer at a final concentration of 1% SDS, 1 mM EDTA, 5% 2-mercaptoethanol, 10 mM Tris-HC1, pH 8.0, 10% glycerol and heated in a boiling water bath for 3 min. The samples were analyzed with agarose-PAGE and immunoblotting.
Agarose-PAGE, SDS-PAGE, and Immunoblotting-Electrophoresis in a 1.75% polyacrylamide and 0.5% agarose slab gel was carried out by the method of Kiehm and Ji (29). The gels were subjected to immunoblotting as described elsewhere (9).

Phosphorylntion of Reconstituted IPa by PKA-The reconstituted
IPsRs were phosphorylated by incubation for 30 min at 37 "C with PKA in a reaction mixture that contained 10 mM MgC12, 50 mM NaCI, 50 mM KCI, 20 mM Tris-HCI, pH 7.4,2 mM 2-mercaptoethanol, 50 p~ ATP, and 4.8 pl of 50 units/pl PKA catalytic subunit solution in a final volume of 1.0 ml. After 30 min proteoliposome solutions were centrifuged at 100,000 X g for 20 min at 2 "C to remove ATP, MgCl2, and PKA. The resulting pellets were suspended in buffer C with 2 mM 2-mercaptoethanol. To examine the autophosphorylation of IP3R under the above conditions and for the control experiment, proteoliposomes were incubated without PKA in a reaction mixture that contained 10 mM MgC12, 50 m M NaCI, 50 mM KCI, 20 mM Tris-HCI, pH 7.4,2 mM 2-mercaptoethanol, 50 p~ ATP, 2.56 mg of sucrose, and 4.8 p1 of 6 mg/ml dithiothreitol in a final volume of 1.0 ml. Sucrose and dithiothreitol were added in order to adjust the conditions with PKA solution. Centrifugation was performed at 100,000 x g for 20 min at 2 "C. The resulting pellets were suspended in buffer C with 2 mM 2-mercaptoethanol. Then the samples were subjected to ''Ca*+ uptake assay.
Determination of Stoichiometry of Phosphorylation-The phosphorylation of the reconstituted IP3R was carried out according to the procedure described above, using 5 p~ [y3'P]ATP (3 pCi/nmol). Samples were analyzed by SDS-PAGE; immunoblotting by mAb 18A10 as described above and radioactivity corresponding to the 250-kDa band were analyzed by a Fuji image analyzer (Fuji, Tokyo). Protein concentrations were determined by the method of [3H]IP3 binding assay with purified IP3R as standard.
Immunohistochemistry-After ether anesthetization, an adult ddY mouse was killed by injecting the solution containing 4% paraformaldehyde and 0.1 M sodium phosphate buffer, pH 7.4, via the left ventricle and washing out from the right atrium, and then the brain was dissected and embedded in paraffin. Paraffin-embedded samples were cut sagittally into 6-pm thick sections that were then deparaffinized and equilibrated in PBS. The sections were sequentially incubated as follows: (i) in 0.05% Triton X-lOO/PBS for 10 min, (ii) in 0.5% skim milk/PBS for 30 min, (iii) in purified IgG or pep 6 affinity-purified antibody against IP3R at the concentration of 10 pg/ ml for 60 min, (iv) in biotinylated anti-rabbit IgG solution for 30 min, (v) in avidin-biotin-peroxidase complex (ABC) solution for 30 min, and (vi) in 0.1% diaminobenzidine, 0.02% hydrogen peroxide, 0.01 M imidazole/PBS. Biotinylated anti-IgG and ABC solutions were from Vectastain ABC kit.
Purification of IPa with Conventional Method and Measurement of I3H]IP3 Binding-The methods are described elsewhere (4).

Immumaffinity Purification of I P a -I Homotetramer from
Mouse Cerebellum"IP3R heterogeneity derived from different genes has been reported in rat and mouse. We originally cloned cerebellar IP&-1 (5), and Siidhof et al. ( complete sequence of rat IP3R-111. Recently our group cloned the whole cDNA of human IP3R-11 and IP3R-111 (53). A comparison of the sequences for the three receptors (I, 11, and 111) revealed the most C-terminal regions of the receptors are different, suggesting that each region should exhibit different antigenicity. In order to purify the functional IP3R-I alone, we have developed a novel method for the rapid, gentle purification of the receptor using immunoaffinity chromatography. A polyclonal antibody against synthetic peptide corresponding to the most C-terminal amino acid, 2736-2747 (designated pep 6) of IP3R-I, was raised and used for immunoaffinity chromatography. IP3R was solubilized from crude mouse cerebellar membrane fraction in 1% CHAPS at pH 7.4, and the fraction was incubated with the affinity Sepharose beads to which the anti-pep 6 antibody was coupled. The beads were rinsed to remove unbound protein, as described under "Experimental Procedures." For the specific elution of IP3R-I from the beads without losing the receptor activity, pep 6 was used instead of the denaturing conditions, because we have previously observed that the binding activity of IP3R is irreversibly inactivated once they are exposed to conditions including low pH and chaotropy even if IP3R was rapidly brought to a neutral pH (data not shown). Another polyclonal antibody against conventionally purified IP,R was raised in the rabbit to detect whole types of the receptor. Anti-purified IP3R polyclonal antibody (polyAb-whole) probably crossreacts to whole types of the receptor (IP3R-II, -111, or -IV), because these IP3Rs display a high degree of homology in the membrane-spanning domain, N-terminal IP3 binding domain, and some midparts of the receptor (5, 23, 24). Fig. 1 shows the SDS-PAGE analysis of solubilized microsomal fraction (lane 1 ) and eluate by pep 6 after immunoaffinity chromatography using anti-pep 6 antibody (lane 2). As the concentration of immunoaffinity-purified IP3R is very low, we concentrate it about five times to detect IP3R stained with Coomassie Blue (Fig. 1). By this concentration process the protein must be degraded a little. When the initial sample was loaded on a gel and was stained with silver stain, the lower molecular bands were not visible (data not shown). Therefore, the lower molecular weight bands in Fig. 1, lane 2, are degraded protein of The eluate is immunoreactive with mAb 10A6 and polyAb-whole (Fig. 2), indicating that this purified protein by one step is IP&. It was found that 30% (= 100 -(57.1 + 12.9)) of the applied [3H]IP3 binding sites were reproducibly adsorbed to the immunoaffinity column, and at least about 8% (= 1.1 + 1.7 + 2.2 + 2.0 + 1.8) of them were eluted by pep 6 at the first purification cycle (Table I) (Table I). In the third purification cycle 3.3% of initially applied [3H]IP3 binding sites were flow-through from the column. To characterize the population of the affinity-purified IP3R among the whole subtypes on the cerebellar membranes, we analyzed the adsorbed (the eluate 3 in the first cycle) and the nonadsorbed (the flowthrough fraction in the third cycle) IP3R. Each fraction having the same amount of [3H]IP3 binding activity was applied to the gel (Fig. 2), followed by immunoblotting with various antibodies to IP&. The fraction that is not adsorbed to the anti-pep 6 antibody column hardly showed immunoreactivity with mAb 10A6 as well as anti-pep 6 antibody against IP3R-I t -but showed an immunoreactivity with polyAb-whole (Fig. 2), indicating that polyAb-whole reacts to other types of receptor. The nonadsorbed IP3R reacting to the immunoaffinity beads is probably the candidate of type I1 and/or type I11 receptor.

4.0)) readsorption to the beads
Since the IP3R is composed of tetramer (9), there arises an important question whether the immunoaffinity-purified IP3R-I is a homotetramer of type I or a heterotetramer composed of subunits of other types of the receptor. To investigate the subunit composition of the immunoaffinity-purified receptor, we denatured the IP3R to monomer (Fig. 3A) and immunoprecipitated the monomers with anti-pep 6 antibody against type I receptor in the same condition of denaturation (see "Experimental Procedures"). We analyzed by immunoblotting with polyAb-whole and got a positive reaction with the immunoprecipitated monomer but almost no reaction with the supernatant fraction (Fig. 3B), suggesting that immunoaffinity-purified receptor is composed of homotetramer of type I subunit. From the present data, we can conclude that IP3R purified by the present method is composed to the same subunit rather than other subunits coming from different genes. Fig. 4 provides a comparison of anti-pep 6 and polyAbwhole for immunohistochemical analysis of the IP3R. A section of the cerebellum from a 22-day-old mouse was stained by each polyclonal antibody. The soma and the dendritic arborization of Purkinje cells were stained in a way similar to our previous immunohistochemical data obtained with monoclonal antibodies (30,31).

The Properties of I P a Reconstituted into Lipid Vesicles-
The immunoaffinity-purified IP3R was reconstituted into lipid vesicles by the dialysis method. Before establishing our assay system, we tested various reconstitution methods. We initially tested the conditions of various phospholipid concentrations. Under the 10 mg/ml phospholipid with the protein/ lipid ratio of 1:200, the conditions gave little amplitude of 45Ca2+ influx into proteoliposomes in response to the addition of Ips. Subsequent micrographic observations revealed that proteoliposomes were formed as multilamella (data not  into proteoliposomes in the presence of 250 nM IP3. The specific 45CaZ+ influx, the difference between the "Ca2+ influx in the presence and absence of IP3, reaches a plateau at 45 s. The amount of "Ca" influx is expressed relative to the value of 45Ca2+ influx in the presence of IP3 at 45 s. The background leak of "Ca" into the vesicles at time 0 in the absence of IP3 was between 1500 and 3000 cpm. The data were corrected for the contribution of this value. And the maximal specific '%a2+ influx is between 2000 and 3000 cpm. shown). Sephadex chromatography (32) was also tested. Electron micrographs showed that the formed vesicles were too small to be reconstituted with the IP3R tetramer molecule (data not shown). Several detergents were tested. In the case of the octyl glucoside dilution method, [3H]IP3 binding activity was irreversibly inhibited by the detergent. The present protocol using immunoaffinity-purified IP3R consistently gave high amplitude of 45Ca2+ transport.
IPsR-formed proteoliposomes, in which the phospholipid concentration is 1 mg/ml and the immunoaffinity-purified IPaR/lipid ratio is 1:100, resulted in IP3-induced influx of "Ca2+ into proteoliposomes. Fig. 5 shows the time course of "Ca2+ influx into proteoliposomes in the absence or presence of 250 nM IP3, which gives half-maximal response for 45Ca2+ influx. Under our present assay conditions, the influx increased to a plateau by 45 s to 1 min accompanied by rapid "Ca" uptake within 10 s (Fig. 5). An identical result was obtained in the presence of 10 PM IP3. The time course of 45Ca2+ influx is similar to that of Bourguignon et dl. (33); however, it is different from that of Ferris et al. (25), who obtained a full plateau by 4-8 s, presumably due to the difference of IP3R protein concentration in the proteoliposomes. Under our present conditions the protein concentration is 10 times less than Ferris et al. (25). Fig. 6A shows the dose-response curve for IP3-induced 45Ca2+ influx. Half-maximal IP3-stimulated 4sCa2+ influx required 200 nM IP3 (Fig.  6A), a slightly higher value than Ferris et al. (25) reported.
In the previous study, we found that mAb 18A10, which recognizes the epitope close to the putative Ca2+ channel region at the C terminus of IP3R-1, is highly useful as a specific functional inhibitor of IICR as demonstrated in cerebellar microsomes (34) and hamster eggs (35). mAb MA10 partially blocks IICR from cerebellar microsomes (34), suggesting that microsomes contain not only type I but also other types of receptor. To verify that purified and reconstituted IP3R has the pharmacological properties of the IP3R-I, the effects of mAb 18A10 were examined. mAb 18A10 (5 pg/ml) almost completely blocked the IP3-induced 45Ca2+ influx, as shown in Fig. 6B, suggesting that the purified and reconstituted IP3R was IP3R-I. In contrast, mAb 10A6 did not inhibit IP3-induced 45Ca2+ influx into proteoliposomes (Fig. 6B). Thus, these data indicate that a single subtype of IP3R (IP3R-I) is reconstituted into liposomes, enabling us to elucidate the functional characteristics of IP&-I.
Structure of IPa in Proteoliposomes-The liposomes reconstituted with IP3R were cross-linked with S-EGS. The cross-linked products of the IP3R were detected by immunoblotting using mAb 18A10 (Fig. 7). Four protein bands (I, 11, 111, and IV) corresponding to monomer, dimer, trimer, and tetramer of IP3R were detected. Lanes 1 and 2 are cross-linked products on the gel under the reducing conditions, and the IP3 ( n u ) FIG. 6. Concentration response relationships for IPs-induced "Ca" influx. A, vesicles reconstituted with purified IPSR were assayed for IP3-induced 45Ca2+ influx as described under "Experimental Procedures." Various concentrations of IP3 were added as indicated in the figure. "Ca2+ influx was measured, and specific counts/min were calculated as the difference between the 'Ta2+ influx in the presence and absence of IP3. Data were normalized to the maximal response observed at 10 p~ IP3 from eight experiments like the one in B. B, inhibition of IP3-induced 45Ca2+ influx into proteoliposomes by mAb 18A10. Reconstituted IP3R was preincubated with mAb 18A10 (5 pglml, 0) or mAb 10A6 (5 pg/ml, A) or without antibody (0) for 10 min at 30 "C, and then IP3-induced "Ca2+ influx was assayed and specific counts/min were calculated.  ( l a n e 2), and 0 mM (lanes 3   and 4 ) , solubilized with agarose-PAGE sample buffer under the reducing conditions in the presence (lanes [1][2][3] or nonreducing conditions in the absence ( l a n e 4 ) of 2-mercaptoethanol, and then subjected to agarose-PAGE. The cross-linked IP3R protein was visualized by immunoblotting with mAb 18A10. I, 11, 111, and IV indicate the positions of monomeric, dimeric, trimeric, and tetrameric forms of IP3R, respectively. tetrameric form was the major product at the cross-linker concentration of 10 mM (lane 1 ), indicating that reconstituted IP3R is composed of tetramer.
Next, the orientation of IP3R on proteoliposomes was investigated. About 50% of immunoaffinity-purified IP3R was incorporated into proteoliposomes and had the same orientation as in uiuo, with about 90% of the [3H]IP3 binding site facing outward. This was demonstrated by [3H]IP3 binding study in the presence and absence of detergents. This was similar to the result Huganir and Racker (36) observed in acetylcholine receptor reconstitution experiments. Furthermore, the result that mAb 18A10 inhibited IPS-induced 45Ca2+ influx into reconstituted liposomes when the antibody was added to the outside (Fig. 6B) suggests that most of the Cterminal region, which is the epitope of mAb 18A10, is exposed outside. Thus, these characteristics of the reconstituted IPsR in the tetrameric structure and transmembrane topology are similar to those of native IP3R in the cell (6, 9,37).
Stimulation of Reconstituted IP& Activity by PKA Phosphorylation-Yamamoto et al. (18) have previously reported that purified IP3R can be phosphorylated by PKA. Supattapone et al. (17) have shown that phosphorylation by PKA of cerebellar membranes substantially reduces the potency of IPS in releasing 45Ca2+, whereas Volpe and Alderson-Lang (22) reported enhanced Ca2+ release activity in canine cerebellar membranes. We examined the effect of phosphorylation using the purified IP3R incorporated into liposomes. Addition of the catalytic subunit of PKA to the proteoliposomes resulted in pronounced incorporation of 32P into the 250-kDa protein corresponding to reconstituted IP3R (Fig. 8). Incorporation of radioactivity into the major band provides near stoichiometric levels of phosphorylation with about 0.8 mol of phosphate/ mol of IP3R. Although autophosphorylation of the IP& was reported by Ferris et al. (38), under our present assay conditions, 30 min of incubation with 50 p M ATP and 10 mM MgC12 without PKA at 37 "C, incorporation of 32P into the IP3Rreconstituted lipid vesicles is not observed (Fig. 8, lane 1 ).
Next, in order to determine whether receptor phosphorylation by PKA alters the calcium channel function of IPS, we examined the effect of phosphorylation on IPS-induced 45Ca2+ influx into the proteoliposomes. Fig. 9 shows the time course of 46Ca2+ uptake into phosphorylated or nonphosphorylated proteoliposome. Both the initial velocity and extent of 45Ca2+ influx into proteoliposomes are enhanced by phosphorylation of IP3R (Fig. 9). We and Ferris et al. previously showed that ATP increases IPS-induced Ca2+ flux using the planar lipid bilayer (9) or liposome system (12). Since under the present The enhancement of "Ca2+ influx into the phosphorylated proteoliposomes can be expressed as a fraction of the value of the "Ca2+ influx into the unphosphorylated proteoliposomes in response to 10 PM IP3 at 45 s by three independent experiments. In each experiment, measurements vary by less than IO%, and "Can+ influx into phosphorylated proteoliposomes is increased. These data are normalized, and results f S.D. are illustrated. The background leak of "Can' into the vesicles at time 0 in the absence of IP3 was between 3200 and 3800 cpm. The data were corrected for the contribution of this value, and the maximal specific "Can+ influx into the phosphorylated or nonphosphorylated proteoliposomes is between 3000 and 4000 cpm. conditions ATP is removed by centrifugation after the PKA reaction is finished, the increasing effect of 45Ca2+ influx substantially reflects PKA activity on the IP3R. The final extent of uptake of 45Ca2+ into phosphorylated proteoliposomes in response to various IPS concentrations was also increased about 20% compared with the control experiment ( Fig. 10). In contrast, the stimulation effect was not observed after proteoliposomes were incubated with either PKA or ATP alone and was prevented after proteoliposomes were incubated with Walsh peptide, an inhibitor of PKA at a concentration of 5 p~ in the presence of both PKA and ATP (Fig. 11). We evaluated the influence of phosphorylation on the binding properties of the IP3R. No change in the [3H]IP3 binding is apparent upon phosphorylation, suggesting that phosphorylation by PKA does not have an influence on the IP3 binding region (data not shown). These data allow us to conclude that the phosphorylation of IP3R-I by PKA significantly increased IP3-induced ''CaZ+ flux.

DISCUSSION
Biochemical and pharmacological analyses have suggested the heterogeneity of IP3R in both IP3 binding and Ca2+ was assayed, and specific counts/min were calculated. The representative data are shown. This experiment has been conducted three times, and measurements varied less than 10%. releasing activities (39,40). The heterogeneity of IP3R has been reported to come from different genes (23,24) or alternative splicing from a single gene (41,42), demonstrating diversity at the IP3 binding, regulatory, and Ca2+ releasing regions. Type I1 IP3R, the cDNA of which was recently cloned, has no consensus sequence of phosphorylation sites (23). IP3R mRNA detected by polymerase chain reaction technique from any tissue is composed of multiple types of receptors (43). To characterize in more detail the properties of each it is important to isolate a single type of the receptor rather than to use the heterogeneous mass of receptors. Conventional purification methods are mainly dependent on heparin-agarose and lentil lectin or concanavalin A column. Conventionally purified IP3R is actually the sum of heterogeneous IP3Rs. Although each IP3R is highly homologous, the sequence of the most C terminus differs from receptor to receptor. We prepared an antibody against the synthetic peptide corresponding to type I IP3R C terminus, which differs in sequence from any other receptor. By affinity purification with peptide elution, we could obtain rapid isolation of the native single type of IP3R. The experiment of immunoprecipitation of denatured IP3R revealed that the immunoaffinity-purified IP3R is composed of the same subunit of IP3R-I. This is therefore the first study to purify a single type of IP3R from a single gene. However, under these conditions alternative spliced forms are included.
We obtained mAb 18.410, which blocks the Ca2+ channel of the IP3R in the cerebellar microsomal fraction (34) and egg (35). When we compared the inhibition kinetics of mAb MA10 in Ca2+ release in proteoliposomes with the microsomal fraction, we found a great difference. Namely, mAb MA10 inhibited Caz+ release from microsomes at a low concentration of IP3 but did not suppress Ca2+ release at higher IP3 concentrations, at which maximal Ca2+ release occurred (34). However, in the present experiment, under the liposome system containing only type I IP3R, mAb 18A10 exclusively inhibited IP3-induced Ca2+ influx into reconstituted vesicles even at a higher concentration of IP3 (Fig. 6B). These differences can be explained by the heterogeneity of IP3Rs in the microsomal fraction obtained from the cerebellum. In the immunohistochemical study of the mouse, cerebellar IP3R-I detected by anti-pep 6 antibody is stained similarly to whole IP3Rs detected by polyclonal antibody against conventionally purified IP3R (Fig. 4), suggesting that different subtypes of the receptors localize on the same intracellular Ca2+ pool. mAb HA10 recognizes the C terminus of IP3R-I and cannot block the Ca2+ release of other IP3Rs from several gene products. In the microsome when the concentration of IP3 is increased, the Ca2+ channels of other types of IP3Rs that are probably situated on the same Ca2+ storage pool with IP3R-I adjacent to each other then become open, probably due to a difference in affinity for IPS, and allow maximal Ca2+ release. The concentration of IP3 in resting and stimulated cells is within a 0.1-10 p~ range (44). Recently, Khodakhah and Ogden (45) reported that in single Purkinje cells the IP3 concentration required to mobilize Ca2+ was more than 5 p~, which was 10-30-fold greater than in astrocyte and peripheral tissues. Therefore, the concentration of IPS that we tested in these experiments (0.1-10 pM is related to the physiological range. Phosphorylation by PKA of ligand-gated or voltage-gated ion channels, including glutamate receptors, nicotinic acetylcholine receptors, y-aminobutyric acid A receptors, sodium channels, or voltage-sensitive calcium channels, is a major mechanism in the regulation of their function (28,(46)(47)(48)(49)(50)(51). IP3R is phosphorylated by PKA. After the initial observation of Supattapone et al. (17) that PKA inhibits IICR from the rat cerebellar microsomal fraction with preloading of Ca2+, Volpe and Alderson-Lang (22) reinvestigated the effect of