Differential modulation of inositol 1,4,5-trisphosphate receptor type 1 and type 3 by ATP

Cell Calcium (2000) 27 (5), 257–267 © 2000 Harcourt Publishers Ltd

Research

doi: 10.1054/ceca.2000.0121, available online at http://www.idealibrary.com on

Differential modulation of inositol 1,4,5-trisphosphate receptor type 1 and type 3 by ATP K. Maes, L. Missiaen, P. De Smet, S. Vanlingen, G. Callewaert, J. B. Parys, H. De Smedt Laboratorium voor Fysiologie, K U Leuven Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium

Summary Binding of ATP to the inositol 1,4,5-trisphosphate receptor (IP3R) results in a more pronounced Ca2+ release in the presence of inositol 1,4,5-trisphosphate (IP3). Two recently published studies demonstrated a different ATP sensitivity of IP3-induced Ca2+ release in cell types expressing different IP3R isoforms. Cell types expressing mainly IP3R3 were less sensitive to ATP than cell types expressing mainly IP3R1 (Missiaen L, Parys JB, Sienaert I et al. Functional properties of the type 3 InsP3 receptor in 16HBE14o- bronchial mucosal cells. J Biol Chem 1998; 273: 8983–8986; Miyakawa T, Maeda A, Yamazawa T et al. Encoding of Ca2+ signals by differential expression of IP3 receptor subtypes. EMBO J 1999; 18: 1303–1308). In order to investigate the difference in ATP sensitivity between IP3R isoforms at the molecular level, microsomes of Sf9 insect cells expressing full-size IP3R1 or IP3R3 were covalently labeled with ATP by using the photoaffinity label 8-azido[α-32P]ATP. ATP labeling of the IP3R was measured after immunoprecipitation of IP3Rs with isoform-specific antibodies, SDS-PAGE and Phosphorimaging. Unlabeled ATP inhibited covalent linking of 8-azido[α-32P]ATP to the recombinant IP3R1 and IP3R3 with an IC50 of 1.6 µM and 177 µM, respectively. MgATP was as effective as ATP in displacing 8-azido[α-32P]ATP from the ATP-binding sites on IP3R1 and IP3R3, and in stimulating IP3-induced Ca2+ release from permeabilized A7r5 and 16HBE14o- cells. The interaction of ATP with the ATP-binding sites on IP3R1 and IP3R3 was different from its interaction with the IP3-binding domains, since ATP inhibited IP3 binding to the N-terminal 581 amino acids of IP3R1 and IP3R3 with an IC50 of 353 µM and 4.0 mM, respectively. The ATP-binding sites of IP3R1 bound much better ATP than ADP, AMP and particularly GTP, while IP3R3 displayed a much broader nucleotide specificity. These results therefore provide molecular evidence for a differential regulation of IP3R1 and IP3R3 by ATP. © 2000 Harcourt Publishers Ltd

INTRODUCTION Inositol 1,4,5-trisphosphate (IP3) is a second messenger used by most cell types to induce Ca2+ release from internal stores [1]. IP3 binds at the N-terminus of the IP3 receptor (IP3R), while the channel region is located at the C-terminus. There is a large transducing domain between the IP3-binding region and the channel region [2], containing interaction sites for several modulators of IP3-induced Ca2+ release such as Ca2+, calmodulin, Received 21 January 2000 Revized 3 May 2000 Accepted 3 May 2000

Correspondence to: Karlien Maes, Laboratorium voor Fysiologie, K U Leuven Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium. Tel.: +32 16 345736; Fax.: 32 16 345991; E-mail: [email protected]

FKBP12, kinases, phosphatases and ATP (reviewed in [1] and [3]). The effects of ATP on IP3-induced Ca2+ release are concentration-dependent. Submillimolar ATP concentrations stimulate the IP3R by binding specifically to an ATP-binding site on the IP3R [4–8], while higher ATP concentrations inhibit IP3-induced Ca2+ release by competing with IP3 for the IP3-binding site [6–11]. The IP3Rs are encoded by three different genes, resulting in the existence of IP3R1, IP3R2 and IP3R3 (reviewed in [1] and [3]). Recent studies of cell types expressing a single IP3R subtype demonstrated functional differences in the regulation of IP3-induced Ca2+ release by IP3 and several modulators like ATP, Ca2+ and thimerosal. In the case of ATP, the EC50 for the stimulatory effect of ATP on IP3-induced Ca2+ release was more than 10-fold higher in IP3R3-expressing 16HBE14o- cells (341 µM) than in A7r5 cells (32 µM), which mainly express IP3R1 257

258 K Maes, L Missiaen, P De Smet, S Vanlingen, G Callewaert, J B Parys, H De Smedt

[12]. IP3-induced Ca2+ release in genetically engineered DT40 B cells that express a single IP3R subtype was also found to respond differently to ATP. In IP3R1-expressing cells, the rate of Ca2+ release was enhanced by ATP with an EC50 of 0.39 mM, while IP3R3-expressing cells were much less sensitive and IP3R2-expressing cells were insensitive to ATP [13]. The apparent difference in ATP sensitivity for the three IP3R isoforms is compatible with the presence of different putative ATP-binding sites in their primary sequence. One glycine-rich region with putative ATP-binding properties is conserved in all three isoforms whereas another cluster is only found in IP3R1 [2,14–16]. We have previously shown that both glycinerich domains, expressed as bacterial fusion proteins, could bind ATP with high affinity[17]. ATP binding to IP3R1 is found to be highly selective for adenine nucleotides (ATP > ADP > AMP > GTP), both in biochemical and functional assays [4–6,8,17,18]. There are controversial results regarding the effect of the ATP species; in some studies [4,5,7,8] MgATP was as effective as free ATP in stimulating IP3R1, while in other studies it was only 50% as effective [6] or ineffective [19]. Caffeine, an adenine-containing compound, is known to interact with the IP3R1. Millimolar concentrations of caffeine inhibit the IP3R1 [20–26]. This inhibition was prevented by adenine nucleotides [24,26]. The relative potencies of various nucleotides for counteracting the inhibition by caffeine [26] and for activating the stimulatory ATP-binding sites [4–7,18] of IP3R1 were the same. We therefore suggested that caffeine interacted with the ATP-binding site of IP3R1[17]. The aim of this study was to elucidate the difference in ATP sensitivity between IP3R1 and IP3R3 at the molecular level. In addition, we wanted to compare the nucleotide specificities of both IP3R isoforms. We also investigated a possible difference between free ATP and MgATP in stimulating IP3-induced Ca2+ release. Finally, we looked for a possible interaction of caffeine with the ATP-binding sites of IP3R1 and IP3R3. In order to investigate the properties of a single IP3R subtype, we expressed full-size IP3R1 and IP3R3 in Sf9 insect cells using the baculovirus expression system. We used the photoaffinity label 8-azido[α-32P]ATP to link ATP covalently to the ATP-binding site(s) of the IP3R isoforms. We have found that the IC50 for inhibiting the photoaffinity labeling was 100-fold higher for IP3R3 than for IP3R1. ATP also inhibited IP3 binding to the N-terminal 581 amino acids of both isoforms, but the IC50 values for this interaction were much higher than those for competing with the photoaffinity labeling site. The ATPbinding site on IP3R3 had a much broader nucleotide specificity than the binding site on IP3R1. Although IP3R1 and IP3R3 appeared to be differentially regulated by ATP, they could both use MgATP as effectively as free Cell Calcium (2000) 27(5), 257–267

ATP and in both cases caffeine could compete with ATP binding. MATERIALS AND METHODS Materials 8-azido[α-32P]ATP (2 mCi/ml, 12 Ci/mmol) was purchased from ICN Pharmaceuticals Inc. (Costa Mesa, CA, USA). 45CaCl2 (2.2 mCi/ml, 134 µg Ca2+/ml), [3H]IP3 (10 µCi/ml, 30 Ci/mmol), the anti-mouse and anti-rabbit alkaline phosphatase-coupled secondary antibodies and the Vistra ECF substrate were obtained from Amersham Pharmacia Biotech AB (Uppsala, Sweden). Protein A-Sepharose beads, heparin-agarose beads, polyethylene glycol, γ-globulins and phosphatidylcholine were from Sigma (St Louis, MO, USA). IP3 was from Roche Molecular Biochemicals (Basel, Switzerland). CHAPS was from Pierce (Rockford, IL, USA). Microsome preparations, antibodies and Western blotting Microsomes of rabbit cerebellum, 16HBE14o- cells and Sf9 insect cells were prepared as described by Parys et al. [27], Sienaert et al. [28] and Yoneshima et al. [29], respectively. The polyclonal antibody against the C-terminus of IP3R1 (Rbt03) and the mouse monoclonal antibody against an N-terminal epitope of human IP3R3 (MMAtype3) (Transduction Laboratories, Lexington, KY, USA) were characterized previously [27,28,30,31]. The various microsomal preparations were analyzed on 3–12% Laemmli-type gels and transferred to ImmobilonP (Millipore Corp., Bedford, MA, USA). Immunodetection of the proteins on the transfers was exactly as previously described [31,32]. Expression of IP3R1 and IP3R3 in insect Sf9 cells The full-length mouse IP3R1 was expressed in insect Sf9 cells as described previously [32]. Recombinant rat IP3R3 was expressed in a similar way. The 5′-untranslated region of the original pCB6+IP3R3 clone [33] was removed and replaced by a Kozak sequence (GCCGCC) by PCR-amplification of the 5′-terminal part from the start codon up to the AflII restriction site (nucleotide 1691). The resulting fragment was cloned into the EcoRIand AflII-digested baculovirus (Autographa californica) transfer vector pVL1393 (Invitrogen Corp, Carlsbad, CA, USA). Subsequently, the remaining part of the IP3R3 cDNA was digested by AflII and NotI and subcloned in the pVL1393 construct. Subsequent production and purification of recombinant viruses and production of recombinant IP3R3 protein in Sf9 cells was identical as for IP3R1 [32]. © Harcourt Publishers Ltd 2000

Differential modulation of IP3R by ATP 259

Photoaffinity labeling, solubilization and immunoprecipitation of IP3R 50 µl of a microsomal fraction (10 mg protein/ml) was mixed with 50 µl of the irradiation buffer containing 20 µM of 8-azido[α-32P]ATP (3 µCi), 100 µM 2-mercaptoethanol, 0.15 M NaCl, 10 mM MgCl2 and 50 mM Tris-HCl, pH 7.4. Varying concentrations of unlabeled nucleotides were added as indicated in the legends. The solutions were irradiated on ice for 2.5 min with UV light (5 × 8 W) in the Stratalinker® UV Crosslinker (Stratagene, La Jolla, CA, USA) at 10 cm distance. 300 µl of a solution containing 10 mM dithiothreitol, 0.5 M NaCl, 1 mM EDTA and 50 mM Tris-HCl, pH 7.4 was added and the samples were centrifuged for 17 min at 35 700 g at 4°C. The pellets were solubilized (at 5 mg/ml) for 1.5 h in buffer containing incubation buffer (50 mM Tris-HCl, pH 7.4, 200 mM NaCl, 0.2 mM PMSF, 0.83 mM benzamidine and 10 mM 2-mercaptoethanol) with addition of 77 nM aprotinin, 1.1 µM leupeptin, 0.7 µM pepstatin A, 2.5% CHAPS, 1% phosphatidylcholine and 0.025% NaN3. After centrifugation for 17 min at 35 700 g at 4°C, the supernatants (100 µl) were incubated overnight at 4°C with 15 µl Rbt03 antibody against IP3R1 or 15 µl MMAtype3 antibody against IP3R3 and 600 µl incubation buffer. Subsequently, the samples were incubated with 100 µl of 50% Protein A-Sepharose beads for another 3 h at 4°C. The beads were washed two times with wash buffer containing the incubation buffer with addition of 0.5% CHAPS, 0.2% phosphatidylcholine and 0.005% NaN3. Finally, the beads were resuspended in 100 µl of the wash buffer and 50 µl of 15% SDS, 575 mM sucrose, 325 mM Tris-HCl, pH 6.8 and 715 mM 2-mercaptoethanol. The samples were heated at 95°C for 5 min and the supernatants were analyzed by SDS-PAGE on a 3–12% linear gradient and transferred to Immobilon-P. The blot was exposed to a Phosphor Screen (Molecular Dynamics, Sunnyvale, CA, USA) and scanned with the Storm 840 PhosphorImager, equipped with the ImageQuaNT 4.2 software (Molecular Dynamics, Sunnyvale, CA, USA). Expression of Lbs-1 and Lbs-3 in Escherichia coli, preparation of soluble fraction and partial purification The expression of the recombinant N-terminal 581 amino acids of IP3R1 and IP3R3 (Lbs-1 and Lbs-3, respectively) was essentially as described [32]. Preparation of the soluble fraction of Escherichia coli and partial purification of the recombinant proteins on a heparin-agarose column were described previously [32]. IP3-binding assay [3H]IP3 binding was performed in 100 µl of a solution containing 50 mM Tris-HCl, pH 7.4, 50 mM NaCl, 1 mM © Harcourt Publishers Ltd 2000

EDTA, 10 mM 2-mercaptoethanol, 6 µg of partially purified Lbs-1 or Lbs-3, varying [ATP] and 5 nM [3H]IP3. Non-specific binding was determined in the presence of 10 µM unlabeled IP3. The solutions were incubated for 30 min at 0°C. 10 µl of γ-globulins (20 mg/ml) and 110 µl of 10% polyethylene glycol in 50 mM Tris-HCl, pH 7.4, 1 mM EDTA and 10 mM 2-mercaptoethanol were subsequently added. The mixtures were quickly filtered through glass fiber filters and washed using a Combi Cell Harvester (Skatron, Lier, Norway). Activity on the filters was quantified with a Beckman β-scintillation counter (LS 6500, Beckman Instruments Inc., Fullerton, CA, USA). 45

Ca2+ fluxes

A7r5 smooth muscle cells and 16HBE14o- bronchial epithelial cells were cultured as described in [4] and in [28], respectively. 45Ca2+ fluxes on permeabilized cells were studied on a thermostatically controlled plate at 25°C. The culture medium was aspirated and replaced by 1 ml of permeabilization medium containing 120 mM KCl, 30 mM imidazole-HCl, pH 6.8, 2 mM MgCl2, 1 mM ATP, 1 mM EGTA and 20 µg/ml saponin. The saponincontaining solution was removed after 10 min and the cells were washed once with a similar saponin-free solution. 45Ca2+ uptake into the non-mitochondrial Ca2+ stores was accomplished by incubation for 45 min in 1 ml of loading medium containing 120 mM KCl, 30 mM imidazole-HCl, pH 6.8, 5 mM MgCl2, 5 mM ATP, 0.44 mM EGTA, 10 mM NaN3 and 100 nM free 45Ca2+. After this phase of 45Ca2+ accumulation, the monolayers were incubated in 1 ml of efflux medium containing 120 mM KCl, 30 mM imidazole-HCl, pH 6.8, 1 mM EGTA and 2 µM thapsigargin. Additions of IP3, Ca2+, Mg2+, ATP and other nucleotides (ADP, AMP and GTP) are indicated in the legend to Figure 4 and Table 3. Because of chelation of Mg2+ by ATP, the actual free [Mg2+] and free [ATP] were calculated by the Maxchelator software (C. Patton, Stanford University, Stanford, CA, USA). The efflux medium was replaced each 2 min. At the end of the experiment the 45 Ca2+ remaining in the stores was released by incubation in 1 ml of 2% (w/v) SDS for 30 min.

RESULTS Expression of IP3R1 and IP3R3 in Sf9 insect cells Microsomes of Sf9 insect cells either transfected with pVL1393-IP3R1 (Sf9-IP3R1) or pVL1393-IP3R3 (Sf9-IP3R3) expressed a protein immunoreactive to antibody Rbt03 (Fig. 1A, lane 2) or to antibody MMAtype3 (Fig. 1B, lane 3), respectively. Under identical conditions, no IP3R1 (Fig. 1A, lane 1) or IP3R3 (Fig. 1B, lane 1) were detected with the isoform-specific antibodies in control pVL1393 Cell Calcium (2000) 27(5), 257–267

260 K Maes, L Missiaen, P De Smet, S Vanlingen, G Callewaert, J B Parys, H De Smedt

A

Fig. 1 Expression of mouse IP3R1 and rat IP3R3 in Sf9 insect cells. (A) Microsomes of Sf9 insect cells transfected with pVL1393 (lane 1, 10 µg), of Sf9 insect cells transfected with pVL1393-IP3R1 containing the complete coding sequence of mouse IP3R1 (lane 2, 10 µg), of Sf9 insect cells transfected with pVL1393-IP3R3 bearing the complete coding sequence of rat IP3R3 (lane 3, 10 µg), and rabbit cerebellar microsomes (lane 4, 10 µg) were separated by SDS-PAGE and transferred to Immobilon-P. The blot was probed with the Rbt03 polyclonal antibody (dilution 1/10,000) and the position of IP3R1 is indicated (arrow). Only the relevant part of the blot is shown. (B) Microsomes of Sf9 insect cells transfected with pVL1393 (lane 1, 10 µg), of Sf9 insect cells transfected with pVL1393-IP3R1 containing the complete coding sequence of mouse IP3R1 (lane 2, 10 µg), of Sf9 insect cells transfected with pVL1393-IP3R3 bearing the complete coding sequence of rat IP3R3 (lane 3, 10 µg), of 16HBE14o- cells (lane 4, 200 µg) and of A7r5 cells (lane 5, 200 µg) were separated by SDS-PAGE and transferred to Immobilon-P. The blot was probed with MMAtype3 (dilution 1/1000) against IP3R3 and the position of IP3R3 is indicated (arrow). Only the relevant part of the blot is shown.

transfected Sf9 cells. IP3R1 was not detected in Sf9 cells transfected with pVL1393-IP3R3 (Fig. 1A, lane 3) and IP3R3 was not detected in Sf9 cells transfected with pVL1393-IP3R1 (Fig. 1B, lane 2). The expressed IP3R1 migrated on SDS-PAGE with the same molecular mass as observed for IP3R1 from rabbit cerebellar microsomes and the density per mg protein amounted to 2.5 times the value found in cerebellum (Fig. 1A, lane 4). The expressed IP3R3 migrated on SDS-PAGE with almost the same molecular mass as observed for IP3R3 from A7r5 or 16HBE14o- microsomes and the density per mg protein amounted to >50 times the value found in either of those two microsomal preparations (Fig. 1B, lane 4 and 5, respectively). Photoaffinity labeling of recombinant IP3R1 and IP3R3 with 8-azido[α-32P]ATP Sf9-IP3R1 and Sf9-IP3R3 microsomes were incubated with 8-azido[α-32P]ATP, UV irradiated and their Cell Calcium (2000) 27(5), 257–267

B Fig. 2 Inhibition of 8-azido[α-32P]ATP labeling of IP3R1 and IP3R3 by different [ATP]. (A) Microsomes of Sf9 insect cells were incubated with 20 µM 8-azido[α-32P]ATP in the presence of the indicated concentrations of unlabeled ATP. The microsomes were irradiated with UV light for 2.5 min and solubilized by CHAPS. IP3R1 and IP3R3 were subsequently immunoprecipitated with isoform-specific antibodies. After SDS-PAGE and blotting, labeled IP3Rs were visualized using the Storm 840 PhosphorImager (Molecular Dynamics). The details of the photoaffinity labeling of the IP3Rs were described in Materials and Methods. The relevant parts of the blots are shown. (B) The extent of photoaffinity labeling of the IP3Rs was quantified using the ImageQuaNT software. The percentage inhibition of labeling of IP3R1 (▲) and IP3R3 ( ) was plotted against the concentration of unlabeled ATP. Increasing [ATP] inhibited photoaffinity labeling of IP3R1 and IP3R3 with IC50 values of 1.6 µM and 177 µM, respectively. Each result is the mean of three independent experiments, each performed in duplicate. The standard deviations were always smaller than 10%.

•

solubilized IP3R was immunoprecipitated with isoformspecific antibodies. The photoaffinity labeling of IP3R1 and IP3R3 with 8-azido[α-32P]ATP was specific since it was completely abolished in the presence of an excess unlabeled ATP (5 mM, Fig. 2). To determine the affinity of the ATP-binding site, we performed the labeling in the presence of various concentrations of unlabeled ATP, ranging from no addition to 10 mM (Fig. 2). The amount of photoaffinity labeling was quantified using the ImageQuaNT software and the percentage inhibition of the labeling (mean value for three experiments) was plotted against the concentration of unlabeled ATP (Fig. 2B). The IC50 for inhibiting the photoaffinity labeling was 1.6 ± 0.2 µM for IP3R1 and 177 ± 28 µM for IP3R3. © Harcourt Publishers Ltd 2000

Differential modulation of IP3R by ATP 261

Table 1 8-azido[α-32P]ATP labeling of recombinant IP3R1 and IP3R3 in the presence of IP3, caffeine or Mg2+. The extent of photoaffinity labeling of IP3R1 and IP3R3 was measured in the presence of IP3 (100 µM), caffeine (50 mM), or MgCl2 (10 mM) plus 10 µM (IP3R1) or 100 µM (IP3R3) ATP, quantified using the ImageQuaNT software and expressed as a percentage of the control value. The control value was the extent of photoaffinity labeling in the absence of added cold nucleotide. Each result is the mean (±SD) of three independent experiments, each performed in duplicate. The details of the photoaffinity labeling of the IP3Rs were described in Materials and Methods.

Table 3 Nucleotide dependence of IP3-induced Ca2+ release from A7r5 cells and 16HBE14o- cells. Non-mitochondrial Ca2+ stores, loaded to steady state with 45Ca2+, were incubated in efflux medium for 10 min, at which time 1.5 µM IP3 and 30 µM (A7r5) or 300 µM (16HBE14o-) ATP, ADP, AMP or GTP were added for 2 min. Concentrations of the nucleotides were chosen to be close to the EC50 value of ATP for stimulating IP3-induced Ca2+ release in A7r5 and 16HBE14o- cells, respectively. Each result is the mean (± SD) of two independent experiments each performed in duplicate. % stimulation

% of control value IP3R1 IP3 100 µM Caffeine 50 mM MgCl2 10 mM

97 ± 4 36 ± 4 110 ± 10

IP3R3 104 ± 5 28 ± 6 93 ± 8

ATP ADP AMP GTP

A7r5

16HBE14o-

100 107.0 ± 9 56.7 ± 4.6 28.5 ± 2.1

100 52.4 ± 3 44.7 ± 5.2 90 ± 10.3

Table 2 Nucleotide specificity of recombinant IP3R1 and IP3R3. The extent of photoaffinity labeling of IP3R1 and IP3R3 was measured in the presence of various nucleotides (ATP, ADP, AMP and GTP), quantified using the ImageQuaNT software and expressed as percentage inhibition of photoaffinity labeling. Each result is the mean (±SD) of three independent experiments, each performed in duplicate. The details of the photoaffinity labeling of the IP3Rs were described in Materials and Methods. % inhibition of photoaffinity labeling

ATP ADP AMP GTP

10 µM 100 µM 1 mM 10 µM 100 µM 1 mM 100 µM 1 mM 100 µM 1 mM

IP3R1

IP3R3

91 ± 0.3 100 100 86 ± 2 89 ± 2.3 100 74 ± 3.4 93 ± 1.2 51 ± 4.7 87 ± 0.2

29 ± 3.2 51 ± 4.5 91 ± 0.7 18 ± 7.4 45 ± 3.1 68 ± 0.8 40 ± 3.6 59 ± 3.2 55 ± 2.7 90 ± 0.7

Inhibition of IP3 binding by ATP Since ATP competes with IP3 for the IP3-binding site [6–11], we wondered whether the observed photoaffinity labeling of the IP3R isoforms could be caused by binding of 8-azido[α-32P]ATP to the latter site. For this purpose, [3H]IP3 binding was measured to the N-terminal 581 amino acids of IP3R1 and IP3R3 expressed in E. coli (Lbs-1 and Lbs-3, respectively) in the presence of various ATP concentrations, ranging from no addition to 10 mM (Fig. 3). ATP inhibited [3H]IP3 binding to Lbs-1 and Lbs-3 with an IC50 of 353 ± 45 µM and 4.0 ± 0.2 mM, respectively. Although there was a striking difference in the IC50 values for the interaction with Lbs-1 and Lbs-3, both IC50 values were much higher than those found for the inhibition of the 8-azido [α-32P]ATP labeling of IP3R1 and IP3R3 (1.6 µM and 177 µM, respectively). Moreover, IP3 (100 µM) was not able to inhibit the 8-azido[α-32P]ATP © Harcourt Publishers Ltd 2000

Fig. 3 IP3 binding to the IP3-binding domains Lbs-1 and Lbs-3 in the presence of various [ATP]. Expression of Lbs-1 and Lbs-3 in E. coli, preparation of the soluble fraction and partial purification of the proteins is described in Materials and Methods. IP3 binding to partially purified Lbs-1 (▲) and Lbs-3 (●) (6 µg each) was performed using the IP3-binding assay described in Materials and Methods. Increasing [ATP] displaced [3H]IP3 from Lbs-1 and Lbs-3 with IC50 values of 353 µM and 4 mM, respectively. Each result is the mean of three independent experiments, each performed in triplicate. The standard deviations were always smaller than 10%.

labeling of IP3R1 and IP3R3 (Table 1). We therefore conclude that IP3 and 8-azido[α-32P]ATP are acting on different binding sites on both IP3R1 and IP3R3. Nucleotide specificity of IP3R1 and IP3R3 In order to investigate the nucleotide specificity of IP3R1 and IP3R3, 8-azido[α-32P]ATP labeling of recombinant immunoprecipitated IP3R1 and IP3R3 was performed in the presence of different concentrations of unlabeled Cell Calcium (2000) 27(5), 257–267

262 K Maes, L Missiaen, P De Smet, S Vanlingen, G Callewaert, J B Parys, H De Smedt

A

C

B

D

Fig. 4 Effect of ATP and MgATP on IP3-induced Ca2+ release from A7r5 (A,B) and 16HBE14o- (C,D) cells. Non-mitochondrial Ca2+ stores, loaded to steady state with 45Ca2+, were incubated in efflux medium for 10 min, at which time 1 µM IP3 and 0.22 µM free Ca2+ with (● ●) or without (●) 0.5 mM ATP were added (A,C) for 2 min, as indicated by the horizontal bar. The same experiment was performed in the presence of 3 mM MgCl2, with (■ ■) or without (■) 0.5 mM ATP (B,D). Ca2+ release was plotted as fractional loss, i.e. the amount of Ca2+ leaving the stores in 2 min divided by the total store Ca2+ content at that time. Typical for three independent experiments each performed in duplicate.

Cell Calcium (2000) 27(5), 257–267

© Harcourt Publishers Ltd 2000

Differential modulation of IP3R by ATP 263

ATP, ADP, AMP and GTP (Table 2). For IP3R1, photoaffinity labeling was antagonized in the following order: ATP > ADP > AMP > GTP. In contrast, IP3R3 displayed a different nucleotide specificity: ATP ≈ GTP ≥ ADP ≥ AMP. In cell types that predominantly expressed IP3R1 (A7r5 cells) or IP3R3 (16HBE14o- cells) the dependence of IP3induced Ca2+ release on nucleotides shared qualitatively the same selectivity we observed for the purified receptors (Table 3). Particularly it became evident that the nucleotide-binding site of IP3R1 was more selective for ATP and ADP but much less for GTP, whereas for IP3R3, GTP was nearly as effective as ATP. We further examined whether caffeine, an adeninecontaining compound and a known inhibitor of IP3induced Ca2+ release, interacts with the ATP-binding sites on IP3R1 and IP3R3. Caffeine at a maximum concentration of 50 mM in the incubation buffer inhibited 8-azido[α-32P]ATP labeling by more than 60% in both IP3R isoforms (Table 1), confirming the hypothesis that caffeine and ATP are acting at the same site on the IP3R.

Role of Mg2+ in stimulation of IP3-induced Ca2+ release by ATP We investigated a possible difference between free ATP and MgATP in stimulating the IP3R, by photoaffinity labeling of recombinant IP3R1 and IP3R3 in the presence and in the absence of 10 mM MgCl2 (Table 1). Total ATP concentrations of 10 µM and 100 µM were used for IP3R1 and IP3R3, respectively. In the presence of Mg2+, free [ATP] was calculated to be 0.112 µM and 1.13 µM, respectively. Photoaffinity labeling of IP3R1 and IP3R3 remained the same in the presence or absence of Mg2+. This finding was further explored by a functional analysis of IP3induced Ca2+ release. Permeabilized A7r5 cells (Fig. 4, A and B) or 16HBE14o- cells (Fig. 4, C and D) slowly lost their accumulated 45Ca2+ during incubation in the efflux medium. A short exposure to 1 µM IP3 increased the Ca2+ release from the stores. The Ca2+ release was enhanced if 0.5 mM ATP was added at the time of IP3 addition. The same experiment was performed in the presence of 3 mM MgCl2. In this condition, [MgATP] was close to 0.5 mM, and the free [Mg2+] and the free [ATP] were calculated to be 2.5 and 0.012 mM, respectively. IP3-induced Ca2+ release from permeabilized A7r5 cells and 16HBE14ocells was partially inhibited in the presence of 3 mM MgCl2. This phenomenon was described earlier [34–37]. For this reason it is difficult to compare the effect of ATP in a quantitative way. It was however, evident that the fractional loss was increased by the addition of 0.5 mM ATP, both in the presence and in the absence of Mg2+. This was the case for both cell types expressing different patterns of IP3Rs. Taken together, these findings indicate © Harcourt Publishers Ltd 2000

that MgATP as well as free ATP stimulate IP3-induced Ca2+ release by binding to the same site on the IP3R.

DISCUSSION IP3R1 and IP3R3 have a different ATP affinity and a different nucleotide specificity The different IP3R isoforms (IP3R1, IP3R2 and IP3R3) are expressed in a tissue-specific manner [30,38–40]. Recently, it has been shown that cell types expressing a particular IP3R isoform differ significantly in their response to ATP [12,13]. IP3-induced Ca2+ release was strongly enhanced by ATP in IP3R1-expressing cells whereas a less significant effect of ATP was observed in IP3R3-expressing cells [13]. A similar difference in ATP sensitivity was observed for the IP3-induced Ca2+ release from A7r5 cells, expressing mainly IP3R1, and from 16HBE14o- cells, expressing predominantly IP3R3 [12]. In the present study, we demonstrated that this difference in ATP sensitivity was due to the molecular properties of the different IP3R isoforms expressed in those cell types. We labeled recombinant IP3R1 and IP3R3, expressed in Sf9 insect cells, with the photoaffinity label 8-azido [α-32P]ATP and found that IP3R1 had a 100-fold higher affinity for ATP than IP3R3, with IC50 values of 1.6 µM and 177 µM, respectively. The IC50 of ATP binding to recombinant IP3R1 was in the micromolar range but somewhat lower than that for purified cerebellar IP3R1 (5–17 µM) [5,8]. ATP stimulated IP3-induced Ca2+ release from 16HBE14o- cells, which express predominantly IP3R3, with an EC50 of 341 µM [12], which is in good agreement with our data concerning the ATP affinity of recombinant IP3R3. For A7r5 cells, ATP stimulated IP3-induced Ca2+ release with an EC50 of 32 µM [12]. A7r5 cells from rat embryonic aorta express 75% IP3R1 and 25% IP3R3 [40]. Preliminary data indicate extensive heterotetramer formation in this cell type (Parys, unpublished data). Since Miyakawa et al. [13] showed that the ATP sensitivity of IP3R3 became dominant over that of IP3R1 in cells expressing both IP3R1 and IP3R3, this can explain the apparently lower ATP sensitivity of A7r5 cells. The difference in ATP affinity between IP3R1 and IP3R3 can be explained by the difference in the primary sequence of IP3R1 and IP3R3. No typical consensus sequence for adenine nucleotide binding is evident from the primary sequences. The IP3R1 however, contains two glycine-rich ‘GXGXXG’ sequences (ATPA, residues 1773–1780 and ATPB, residues 2016–2021) which were described earlier as ‘putative’ ATP-binding domains [2,14–16,41]. In a previous study [17], we have expressed the cDNA domains of IP3R1 which contain these glycinerich motifs as GST-fusion proteins in bacteria. We demonstrated that both sequences were able to bind ATP with Cell Calcium (2000) 27(5), 257–267

264 K Maes, L Missiaen, P De Smet, S Vanlingen, G Callewaert, J B Parys, H De Smedt

high affinity. The ATP affinity found in this study for the full-size IP3R1 (IC50 = 1.6 µM) corresponds very well with the IC50 values (≈ 1 µM) previously found for ATPA and ATPB, the ‘GXGXXG’ sequences present in IP3R1. However, we observed a much lower affinity for the fullsize IP3R3 (IC50 = 177 µM) in comparison to the affinity found for ATPB (IC50 ≈ 1 µM), the only ‘GXGXXG’ sequence which is also conserved in IP3R3. It is conceivable that the accessibility of the ATP-binding site (ATPB) expressed as a small fusion protein is higher than for the intact, properly folded protein, resulting in a higher affinity. It is also possible that the ATP-binding site expressed as a GST-fusion protein only represents the core of the ATP-binding site in the intact receptor [42]. As a consequence, ATPB could have a diverging affinity for ATP than the intact IP3R3. In this context, it should be noted that ATP activated gating of the Xenopus IP3R with an affinity of 270 µM [19], which is close to the ATP affinity found for IP3R3. The primary sequence of the Xenopus IP3R is very similar to that of IP3R1 [43], but the ‘ATPA’ sequence is not fully conserved. Only the glycine-rich sequence ‘ATPB’ is conserved in Xenopus IP3R which is similar to the situation for IP3R3 and could explain the low affinity found for the Xenopus isoform. It is known that ATP concentrations in the millimolar range inhibit IP3-induced Ca2+ release by competing with IP3 for the IP3-binding site [6–11]. Since GTP and pyrophosphates are as effective as ATP in inhibiting IP3induced Ca2+ release, this effect seems to be mediated by the pyrophosphate regions of the nucleotides [8]. We showed that photoaffinity labeling with 8-azido[α32 P]ATP of IP3R1 and IP3R3 occurs at the stimulatory ATP-binding sites and not at the IP3-binding sites. IP3 (100 µM) was unable to inhibit 8-azido[α-32P]ATP labeling of IP3R1 and IP3R3, indicating that IP3 and ATP clearly act at different sites on the IP3R. Moreover, ATP inhibited IP3 binding to the IP3-binding domains of IP3R1 and IP3R3 with IC50 values that exceeded those found for the 8-azido[α-32P]ATP labeling (353 µM and 4.0 mM, respectively). Interestingly, there was a strong difference in the ATP sensitivity of the IP3-binding domains of the two isoforms. As a result, both isoforms have a bellshaped dependence on ATP, but IP3R3 has a much lower affinity for the inhibitory as well as for the stimulatory phase. In this study, we also investigated the nucleotide specificities of IP3R1 and IP3R3. For IP3R1, the 8-azido[α32 P]ATP labeling was antagonized in the following order: ATP > ADP > AMP > GTP. The same order of potency was found for the displacement of [α-32P]ATP from purified cerebellar IP3R1 [8] and for the stimulatory effects of nucleotides on IP3-induced Ca2+ release in various systems [4–7, 18]. In our previous study [17], we observed the same nucleotide specificity for GST-fusion proteins Cell Calcium (2000) 27(5), 257–267

expressing the putative nucleotide-binding domains of IP3R1. Little is known about the nucleotide specificity of IP3R3. In this study, we measured a different nucleotide specificity for IP3R3 in comparison to IP3R1 in photoaffinity labeling experiments. While the nucleotide-binding site of IP3R1 was specific for ATP and ADP, the nucleotide-binding site on IP3R3 showed a different rank order of specificity and preferentially recognized ATP and GTP. These nucleotide specificities for IP3R1 and IP3R3 were confirmed in 45Ca2+ flux experiments with A7r5 cells, expressing mainly IP3R1 and 16HBE14o- cells, expressing predominantly IP3R3. Although in these cell lines heterotetramers may still be present, the nucleotide-binding properties apparently reflect those of the predominant isoform expressed. We conclude that IP3R1 and IP3R3 are differentially regulated by ATP, which is reflected by a different ATP sensitivity and a different nucleotide specificity. Physiological relevance In normal physiological conditions, the total [ATP] in the cell is estimated to be in the millimolar range. This implicates that, according to the ATP affinities we determined (1.6 µM for IP3R1 and 177 µM for IP3R3), the nucleotidebinding sites on IP3R1 and IP3R3 should always be occupied. The free [ATP] in the cell is calculated to vary from 420 µM to 540 µM [44–46]. In a recent study, it was suggested that only free ATP and not MgATP enhanced IP3induced Ca2+ release [19]. The ATP affinity of the IP3R and particularly this of IP3R3, would then be close to the free [ATP] in the cell, and would have direct physiological significance. However, in our study as well as in other publications [4,5,7,8], MgATP was as effective as free ATP in stimulating IP3-induced Ca2+ release or in binding to the ATP-binding site of the IP3R. The endoplasmic reticulum (ER) and the mitochondria are in close physical proximity, particularly at sites of Ca2+ release [47,48]. Because of the rapid uptake of the released Ca2+ by the mitochondria [49–51], the [Ca2+] in the mitochondrial matrix can change, affecting the mitochondrial membrane potential [52] and the activities of the mitochondrial enzymes necessary for ATP synthesis [53,54]. As a result, these processes can cause local changes in intracellular [ATP], which may approach the submillimolar range in which the IP3R3 is activated by ATP. Depletion of intracellular ATP can also occur during ischemia. After several minutes of ischemia, ATP levels in the brain can fall below 0.1 mM [55]. It becomes evident from these examples that ATP concentrations can reach the submillimolar range and consequently, the Ca2+ release mediated by IP3R3 could be affected whereas the ATP-binding site(s) on IP3R1 would still be saturated. All © Harcourt Publishers Ltd 2000

Differential modulation of IP3R by ATP 265

cell types however co-express different IP3R isoforms [30,40,56], which are mostly organized in heterotetrameric structures [57–60]. Miyakawa et al. [13] showed that the property of IP3R3 in terms of ATP sensitivity became dominant over that of IP3R1 in cells expressing both IP3R1 and IP3R3. IP3R2 was found to be insensitive to ATP and also this property became dominant in heterotetramers. As a result, most functional IP3R tetramers could have a much lower ATP affinity than that found for IP3R1 homotetramers.

USA). We thank Dr D. C. Gruenert (University of California, San Francisco, CA, USA) for the supply of 16HBE14o- cells. We thank Dr C. W. Taylor (University of Cambridge, UK) for stimulating discussions. J B P is Research Associate and P D S is Research Assistant of the Foundation for Scientific Research-Flanders (FWO). This work was supported by grant 99/08 of the Concerted Actions, by grant P4/23 of the Interuniversity Poles of Attraction Program of the Belgian State and by grants 3.0207.99 and G.0322.97 of the FWO.

Effect of caffeine

REFERENCES

Caffeine, an adenine-containing compound, inhibits the IP3R [20–26]. We previously reported that the inhibition of IP3-induced Ca2+ release by caffeine can be prevented by ATP [24,26]. The relative potencies of various nucleotides for counteracting the inhibition by caffeine [26] and for activating the stimulatory ATP-binding sites [4–8,18] were the same. Caffeine also inhibited [α32 P]ATP binding to the putative ATP-binding domains of IP3R1 [17]. In the present study, caffeine (50 mM) inhibited 8-azido[α-32P]ATP labeling of IP3R1 and IP3R3 by more than 60%. These data suggest that caffeine acted at the level of the ATP-binding sites of both IP3R isoforms. It is conceivable that endogenous caffeine-like compounds may act as negative regulators of the IP3R by preventing ATP to bind to its site and to stimulate IP3-induced Ca2+ release.

1. Berridge MJ. Inositol trisphosphate and calcium signalling. Nature 1993; 361: 315–325. 2. Mignery GA, Südhof TC. The ligand-binding site and transduction mechanism in the inositol 1,4,5-trisphosphate receptor. EMBO J 1990; 9: 3893–3898. 3. Patel S, Joseph SK, Thomas AP. Molecular properties of inositol 1,4,5-trisphosphate receptors. Cell Calcium 1999; 25: 247–264. 4. Missiaen L, Parys JB, De Smedt H et al. Effect of adenine nucleotides on myo-inositol 1,4,5-trisphosphate-induced calcium release. Biochem J 1997; 325: 661–666. 5. Ferris CD, Huganir RL, Snyder SH. Calcium flux mediated by purified inositol 1,4,5-trisphosphate receptor in reconstituted lipid vesicles is allosterically regulated by adenine nucleotides. Proc Natl Acad Sci USA 1990; 87: 2147–2151. 6. Iino M. Effects of adenine nucleotides on inositol 1,4,5-trisphosphate-induced calcium release in vascular smooth muscle cells. J Gen Physiol 1991; 98: 681–698. 7. Bezprozvanny I, Ehrlich BE. ATP modulates the function of inositol 1,4,5-trisphosphate-gated channels at two sites. Neuron 1993; 10: 1175–1184. 8. Maeda N, Kawasaki T, Nakade S et al. Structural and functional characterization of inositol 1,4,5-trisphosphate receptor channel from mouse cerebellum. J Biol Chem 1991; 266: 1109–1116. 9. Willcocks AL, Cooke AM, Potter BV, Nahorski SR. Stereospecific recognition sites for [3H]inositol 1,4,5-trisphosphate in particulate preparations of rat cerebellum. Biochem Biophys Res Commun 1987; 146: 1071–1078. 10. Guillemette G, Balla T, Baukal AJ, Spät A, Catt KJ. Intracellular receptors for inositol 1,4,5-trisphosphate in angiotensin II target tissues. J Biol Chem 1987; 262: 1010–1015. 11. Nunn DL, Taylor CW. Liver inositol 1,4,5-trisphosphate-binding sites are the Ca2+-mobilizing receptors. Biochem J 1990; 270: 227–232. 12. Missiaen L, Parys JB, Sienaert I et al. Functional properties of the type 3 InsP3 receptor in 16HBE14o- bronchial mucosal cells. J Biol Chem 1998; 273: 8983–8986. 13. Miyakawa T, Maeda A, Yamazawa T, Hirose K, Kurosaki T, Iino M. Encoding of Ca2+ signals by differential expression of IP3 receptor subtypes. EMBO J 1999; 18: 1303–1308. 14. Furuichi T, Yoshikawa S, Miyawaki A, Wada K, Maeda N, Mikoshiba K. Primary structure and functional expression of the inositol 1,4,5-trisphosphate-binding protein P400. Nature 1989; 342: 32–38. 15. Mignery GA, Newton CL, Archer BTIII, Südhof TC. Structure and expression of the rat inositol 1,4,5-trisphosphate receptor. J Biol Chem 1990; 265: 12679–12685. 16. Ferris CD, Snyder SH. Inositol 1,4,5-trisphosphate-activated calcium channels. Annu Rev Physiol 1992; 54: 469–488.

CONCLUSION In summary, we have demonstrated that the different ATP sensitivity of IP3-induced Ca2+ release in different cell types can be attributed to the differences in ATPbinding properties of IP3R1 and IP3R3. Both isoforms have regulatory ATP-binding sites which are different from their IP3-binding site. IP3R3 had a 100-fold lower sensitivity for ATP than IP3R1, and a much broader substrate specificity. For both isoforms, MgATP binding was as effective as ATP binding and both could be antagonized by caffeine. ACKNOWLEDGEMENTS We thank Lea Bauwens, Jerry Renders, Luce Heremans, Anja Florizoone, Marina Crabbé, Hilde Van Weijenbergh, Irène Willems, Yves Parijs and Raphael Verbist for their skilful technical assistance. We acknowledge the generous gifts of the p400C1 plasmid containing the IP3R1 cDNA by Drs K. Mikoshiba and A. Miyawaki (University of Tokyo, Japan) and the pCB6+IP3R3 plasmid containing the IP3R3 cDNA by Dr G. I. Bell (Howard Hughes Medical Institute, University of Chicago, IL, © Harcourt Publishers Ltd 2000

Cell Calcium (2000) 27(5), 257–267

266 K Maes, L Missiaen, P De Smet, S Vanlingen, G Callewaert, J B Parys, H De Smedt

17. Maes K, Missiaen L, Parys JB et al. Adenine-nucleotide binding sites on the inositol 1,4,5-trisphosphate receptor bind caffeine, but not adenophostin A or cyclic ADP-ribose. Cell Calcium 1999; 25: 143–152. 18. Kaplin AI, Snyder SH, Linden DJ. Reduced nicotinamide adenine dinucleotide-selective stimulation of inositol 1,4,5-trisphosphate receptors mediates hypoxic mobilization of calcium. J Neurosci 1996; 16: 2002–2011. 19. Mak DD, McBride S, Foskett JK. ATP regulation of type 1 inositol 1,4,5-trisphosphate receptor channel gating by allosteric tuning of Ca2+ activation. J Biol Chem 1999; 274: 22231–22237. 20. Parker I, Ivorra I. Caffeine inhibits inositol trisphosphatemediated liberation of intracellular calcium in Xenopus oocytes. J Physiol Lond 1991; 433: 229–240. 21. Missiaen L, Taylor CW, Berridge MJ. Luminal Ca2+ promoting spontaneous Ca2+ release from inositol trisphosphate-sensitive stores in rat hepatocytes. J Physiol Lond 1992; 455: 623–640. 22. Brown GR, Sayers LG, Kirk CJ, Michell RH, Michelangeli F. The opening of the inositol 1,4,5-trisphosphate-sensitive Ca2+ channel in rat cerebellum is inhibited by caffeine. Biochem J 1992; 282: 309–312. 23. Hirose K, Iino M, Endo M. Caffeine inhibits Ca2+-mediated potentiation of inositol 1,4,5-trisphosphate-induced Ca2+ release in permeabilized vascular smooth muscle cells. Biochem Biophys Res Commun 1993; 194: 726–732. 24. Missiaen L, Parys JB, De Smedt H, Himpens B, Casteels R. Inhibition of inositol trisphosphate-induced calcium release by caffeine is prevented by ATP. Biochem J 1994; 300: 81–84. 25. Bezprozvanny I, Bezprozvannaya S, Ehrlich BE. Caffeineinduced inhibition of inositol 1,4,5-trisphosphate-gated calcium channels from cerebellum. Mol Biol Cell 1994; 5: 97–103. 26. Missiaen L, Parys JB, De Smedt H et al. Inhibition of inositol trisphosphate-induced calcium release by cyclic ADP-ribose in A7r5 smooth-muscle cells and in 16HBE14o- bronchial mucosal cells. Biochem J 1998; 329: 489–495. 27. Parys JB, De Smedt H, Missiaen L, Bootman MD, Sienaert I, Casteels R. Rat basophilic leukemia cells as model system for inositol 1,4,5-trisphosphate receptor IV, a receptor of the type II family: functional comparison and immunological detection. Cell Calcium 1995; 17: 239–249. 28. Sienaert I, Huyghe S, Parys JB et al. ATP-induced Ca2+ signals in bronchial epithelial cells. Pflügers Arch 1998; 436: 40–48. 29. Yoneshima H, Miyawaki A, Michikawa T, Furuichi T, Mikoshiba K. Ca2+ differentially regulates the ligand-affinity states of type 1 and type 3 inositol 1,4,5-trisphosphate receptors. Biochem J 1997; 322: 591–596. 30. De Smedt H, Missiaen L, Parys JB et al. Isoform diversity of the inositol trisphosphate receptor in cell types of mouse origin. Biochem J 1997; 322: 575–583. 31. Vanlingen S, Parys JB, Missiaen L, De Smedt H, Wuytack F, Casteels R. Distribution of inositol 1,4,5-trisphosphate receptor isoforms, SERCA isoforms and Ca2+ binding proteins in RBL-2H3 rat basophilic leukemia cells. Cell Calcium 1997; 22: 475–486. 32. Sipma H, De Smet P, Sienaert I et al. Modulation of inositol 1,4,5-trisphosphate binding to the recombinant ligand-binding site of the type 1 inositol 1,4,5-trisphosphate receptor by Ca2+ and calmodulin. J Biol Chem 1999; 274: 12157–12162. 33. Blondel O, Bell GI, Moody M, Miller RJ, Gibbons SJ. Creation of an inositol 1,4,5-trisphosphate-sensitive Ca2+ store in secretory granules of insulin-producing cells. J Biol Chem 1994; 269: 27167–27170.

Cell Calcium (2000) 27(5), 257–267

34. Bootman MD, Missiaen L, Parys JB, De Smedt H, Casteels R. Control of inositol 1,4,5-trisphosphate-induced Ca2+ release by cytosolic Ca2+. Biochem J 1995; 306: 445–451. 35. Volpe P, Alderson-Lang BH, Nickols GA. Regulation of inositol 1,4,5-trisphosphate-induced Ca2+ release. I. Effect of Mg2+. Am J Physiol 1990; 258: C1077–1085. 36. White AM, Varney MA, Watson SP et al. Influence of Mg2+ and pH on NMR spectra and radioligand binding of inositol 1,4,5-trisphosphate. Biochem J 1991; 278: 759–764. 37. Van Delden C, Foti M, Lew DP, Krause KH. Ca2+ and Mg2+ regulation of inositol 1,4,5-triphosphate binding in myeloid cells. J Biol Chem 1993; 268: 12443–12448. 38. Newton CL, Mignery GA, Südhof TC. Co-expression in vertebrate tissues and cell lines of multiple inositol 1,4,5trisphosphate (InsP3) receptors with distinct affinities for InsP3. J Biol Chem 1994; 269: 28613–28619. 39. Wojcikiewicz RJ. Type I, II, and III inositol 1,4,5-trisphosphate receptors are unequally susceptible to down-regulation and are expressed in markedly different proportions in different cell types. J Biol Chem 1995; 270: 11678–11683. 40. De Smedt H, Missiaen L, Parys JB et al. Determination of relative amounts of inositol trisphosphate receptor mRNA isoforms by ratio polymerase chain reaction. J Biol Chem 1994; 269: 21691–21698. 41. Wierenga RK, Hol WG. Predicted nucleotide-binding properties of p21 protein and its cancer-associated variant. Nature 1983; 302: 842–844. 42. Wierenga RK, Terpstra P, Hol WG. Prediction of the occurrence of the ADP-binding βαβ-fold in proteins, using an amino acid sequence fingerprint. J Mol Biol 1986; 187: 101–107. 43. Kume S, Muto A, Aruga J et al. The Xenopus IP3 receptor: structure, function, and localization in oocytes and eggs. Cell 1993; 73: 555–570. 44. Dunne MJ, West-Jordan JA, Abraham RJ, Edwards RH, Petersen OH. The gating of nucleotide-sensitive K+ channels in insulinsecreting cells can be modulated by changes in the ratio ATP4–/ADP3– and by nonhydrolyzable derivatives of both ATP and ADP. J Membr Biol 1998; 104: 165–177. 45. Kargacin ME, Kargacin GJ. Predicted changes in concentrations of free and bound ATP and ADP during intracellular Ca2+ signaling. Am J Physiol 1997; 273: C1416–1426. 46. Maechler P, Wang H, Wollheim CB. Continuous monitoring of ATP levels in living insulin secreting cells expressing cytosolic firefly luciferase. Febs Lett 1998; 422: 328–332. 47. Rizzuto R, Pinton P, Carrington W et al. Close contacts with the endoplasmic reticulum as determinants of mitochondrial Ca2+ responses. Science 1998; 280: 1763–1766. 48. Simpson PB, Mehotra S, Lange GD, Russell JT. High density distribution of endoplasmic reticulum proteins and mitochondria at specialized Ca2+ release sites in oligodendrocyte processes. J Biol Chem 1997; 272: 22654–22661. 49. Rizzuto R, Brini M, Murgia M, Pozzan T. Microdomains with high Ca2+ close to IP3-sensitive channels that are sensed by neighboring mitochondria. Science 1993; 262: 744–747. 50. Csordás G, Thomas AP, Hajnóczky G. Quasi-synaptic calcium signal transmission between endoplasmic reticulum and mitochondria. EMBO J 1999; 18: 96–108. 51. Babcock DF, Hille B. Mitochondrial oversight of cellular Ca2+ signaling. Curr Opin Neurobiol 1998; 8: 398–404. 52. Loew LM, Carrington W, Tuft RA, Fay FS. Physiological cytosolic Ca2+ transients evoke concurrent mitochondrial depolarizations. Proc Natl Acad Sci USA 1994; 91: 12579–12583.

© Harcourt Publishers Ltd 2000

Differential modulation of IP3R by ATP 267

53. Rutter GA, Burnett P, Rizzuto R et al. Subcellular imaging of intramitochondrial Ca2+ with recombinant targeted aequorin: significance for the regulation of pyruvate dehydrogenase activity. Proc Natl Acad Sci USA 1996; 93: 5489–5494. 54. Hajnóczky G, Robb-Gaspers LD, Seitz MB, Thomas AP. Decoding of cytosolic calcium oscillations in the mitochondria. Cell 1995; 82: 415–424. 55. Abe K, Kogure K, Yamamoto H, Imazawa M, Miyamoto K. Mechanism of arachidonic acid liberation during ischemia in gerbil cerebral cortex. J Neurochem 1987; 48: 503–509. 56. Ross CA, Danoff SK, Schell MJ, Snyder SH, Ullrich A. Three additional inositol 1,4,5-trisphosphate receptors: molecular cloning and differential localization in brain and peripheral tissues. Proc Natl Acad Sci USA 1992; 89: 4265–4269.

© Harcourt Publishers Ltd 2000

57. Wojcikiewicz RJ, He Y. Type I, II and III inositol 1,4,5trisphosphate receptor co-immunoprecipitation as evidence for the existence of heterotetrameric receptor complexes. Biochem Biophys Res Commun 1995; 213: 334–341. 58. Monkawa T, Miyawaki A, Sugiyama T et al. Heterotetrameric complex formation of inositol 1,4,5-trisphosphate receptor subunits. J Biol Chem 1995; 270: 14700–14704. 59. Nucifora FC Jr., Sharp AH, Milgram SL, Ross CA. Inositol 1,4,5-trisphosphate receptors in endocrine cells: localization and association in hetero- and homotetramers. Mol Biol Cell 1996; 7: 949–960. 60. Joseph SK, Lin C, Pierson S, Thomas AP, Maranto AR. Heteroligomers of type I and type III inositol trisphosphate receptors in WB rat liver epithelial cells. J Biol Chem 1995; 270: 23310–23316.

Cell Calcium (2000) 27(5), 257–267