Unique Regulatory Properties of Heterotetrameric Inositol 1,4,5-Trisphosphate Receptors Revealed by Studying Concatenated Receptor Constructs*♦

The ability of inositol 1,4,5-trisphosphate receptors (IP3R) to precisely initiate and generate a diverse variety of intracellular Ca2+ signals is in part mediated by the differential regulation of the three subtypes (R1, R2, and R3) by key functional modulators (IP3, Ca2+, and ATP). However, the contribution of IP3R heterotetramerization to Ca2+ signal diversity has largely been unexplored. In this report, we provide the first definitive biochemical evidence of endogenous heterotetramer formation. Additionally, we examine the contribution of individual subtypes within defined concatenated heterotetramers to the shaping of Ca2+ signals. Under conditions where key regulators of IP3R function are optimal for Ca2+ release, we demonstrate that individual monomers within heteromeric IP3Rs contributed equally toward generating a distinct 'blended' sensitivity to IP3 that is likely dictated by the unique IP3 binding affinity of the heteromers. However, under suboptimal conditions where [ATP] were varied, we found that one subtype dictated the ATP regulatory properties of heteromers. We show that R2 monomers within a heterotetramer were both necessary and sufficient to dictate the ATP regulatory properties. Finally, the ATP-binding site B in R2 critical for ATP regulation was mutated and rendered non-functional to address questions relating to the stoichiometry of IP3R regulation. Two intact R2 monomers were sufficient to maintain ATP regulation in R2 homotetramers. In summary, we demonstrate that heterotetrameric IP3R do not necessarily behave as the sum of the constituent subunits, and these properties likely extend the versatility of IP3-induced Ca2+ signaling in cells expressing multiple IP3R isoforms.

The importance of intracellular Ca 2ϩ as a second messenger is underscored by its distinctive ability to regulate a multitude of diverse cellular processes, including transcription, translation, secretion of fluids, muscle contraction, motility, fertilization, memory, apoptosis, and autophagy (1)(2)(3)(4)(5)(6)(7)(8). This remarkable capacity to precisely and often simultaneously regulate cellular events is thought to be due, at least in part, to the highly sophisticated spatial and temporal control of intracellular [Ca 2ϩ ] by a complement of specialized proteins, collectively termed the "Ca 2ϩ signaling toolkit" (7). Simplistically, this "toolkit" includes several Ca 2ϩ influx and release channels whose activation brings about a rise in basal intracellular Ca 2ϩ concentration. In addition, cytosolic Ca 2ϩ buffers, pumps, and transporters function to reduce the intracellular Ca 2ϩ concentration by extrusion into stores or the extracellular space (9). Essential components of this toolkit are the members of the endoplasmic reticulum-localized inositol 1,4,5-trisphosphate receptor (IP 3 R) 2 family.
Stimulation of cell surface receptors by growth factors, hormones, and neurotransmitters results in Ca 2ϩ mobilization as a result of the generation of IP 3 , which subsequently binds to and activates IP 3 Rs (10,11). There are three major subtypes of IP 3 Rs (R1, R2, and R3), encoded by a distinct gene (Itpr1, Itpr2, and Itpr3) (12,13). These ϳ300-kDa monomeric proteins co-translationally oligomerize into ϳ1200-kDa tetrameric Ca 2ϩ release channels (14). The three subtypes share ϳ60 -70% sequence homology and are conventionally divided into three functional domains. At the extreme N terminus lies the conserved ligandbinding domain, which is composed of a suppressor domain (SD) and a ligand-binding core (15). This is followed by a large, less conserved intermediary regulatory domain that contains several putative sites for regulation by different molecules, including Ca 2ϩ (serving as a co-agonist), adenosine triphosphate (ATP), protein binding partners, and post-translational modifications (16). Finally, at the C terminus lies a six-transmembrane domain that, in addition to being critical for receptor oligomerization (17) and ER localization (18,19), contains the ion-conducting pore between transmembrane helices 5 and 6 (20 -22).
The variation in primary amino acid sequence between the three subtypes results in each isoform exhibiting distinct IP 3 binding affinities and modulatory properties. For instance, IP 3 binding assays performed on the N-terminal 604 amino acids revealed that R2 has an ϳ3and ϳ12-fold greater IP 3 binding affinity than R1 and R3, respectively (23). Utilizing permeabilized cell Ca 2ϩ release assays, our group has shown that subtype sensitivity to ATP modulation follows a similar rank order, with R2 shown to have ϳ3and ϳ10-fold higher affinity for ATP than R1 and R3, respectively. Furthermore, although Ca 2ϩ release of R1 and R3 is potentiated at all [IP 3 ] by ATP, the activity of R2 is uniquely augmented only at submaximal [IP 3 ] (24,25). IP 3 Rs are also differentially regulated by kinases (PKA, PKG, and Akt) (26 -33), proteolytic enzymes (34), post-translational modifications (17), and a whole host of other functional modulators (12). Importantly, all known regulatory motifs are present in each monomer, although the stoichiometry necessary for modulation is unknown.
Further complexity potentially arises from the idea that IP 3 R subtypes assemble into heterotetramers. Although IP 3 R expression is ubiquitous, individual cell types express distinct complements of isoforms at varying proportions. Notably, most cell types express at least two subtypes (35)(36)(37). For instance, although the majority of neuronal IP 3 Rs are R1, most other peripheral cell types express R2 and R3 (38,39). Furthermore, numerous cross-linking (40,41), co-immunoprecipitation (39,42,43), and co-localization and immunostaining assays (44 -47) have supported the existence of heterotetrameric IP 3 R. Additionally, we demonstrated that homotetramers account for only a small proportion of IP 3 Rs in mouse pancreas (39).
Given that IP 3 Rs are regulated in a subtype-specific manner and that these subtypes likely oligomerize into heterotetramers, the obvious question relates to the contribution of each isoform within a heterotetramer to overall receptor function. Studies in cultured cells or isolated tissues fail to account for the exact proportions of each subtype within a tetramer. To address this, we recently described a strategy whereby concatenated IP 3 R cDNA constructs were engineered to result in expression of tetrameric channels with defined composition (39). As proof of principle, we showed that concatenated homo-or heterodimers encoding R1 or R2 were stably expressed in DT40 triple IP 3 R knock-out (3KO) cells, dimerized to form tetrameric channels localized to the ER membrane, were functionally responsive to G q -coupled, G-protein-coupled receptor stimulation, and exhibited identical single channel activity to those expressed from monomeric constructs (39).
In this report, we examine the contribution of individual isoforms within an IP 3 R heterotetramer toward shaping the Ca 2ϩ signals that are generated. Under conditions where key regulators of IP 3 R function are optimal for Ca 2ϩ release (5 mM ATP, 200 nM free Ca 2ϩ ) (24,25,48,49), individual monomers within a heteromeric IP 3 R were found to contribute equally toward dictating the channels' "blended" sensitivity to IP 3 . IP 3 binding assays further revealed that this distinct IP 3 sensitivity is likely dictated by the unique apparent IP 3 binding affinity of heteromeric IP 3 Rs. In contrast, under suboptimal conditions for Ca 2ϩ release through the IP 3 Rs, where [ATP] was varied, a single isoform was found to dictate the ATP regulatory properties of heteromeric IP 3 Rs. Finally, using concatenated receptors with mutations in defined numbers of subunits, we address for the first time fundamental questions relating to the stoichiometry of IP 3 R regulation by ATP.

Materials and Methods
Reagents Used-All restriction enzymes and DNA T4 ligase were obtained from New England Biolabs. Fetal bovine serum, RPMI 1640 media, chicken serum, penicillin/streptomycin, and ␤-mercaptoethanol were obtained from Gibco/Life Technologies, Inc. G418 sulfate (Geneticin) was obtained from Invitrogen. Fura-2 AM and MgFluo4 AM were obtained from Molec-ular Probes by Life Technologies, Inc. Enhanced chemiluminescent substrate and DyLight TM 800CW secondary antibodies were from Thermo Scientific. The D c protein assay kit, Tris base, glycine, horseradish peroxidase-conjugated secondary antibodies, and all reagents used for SDS-PAGE were from Bio-Rad. All materials for native-PAGE, including native-PAGE 3-12% BisTris gels, were obtained from Novex by Life Technologies, Inc. CHAPS was obtained from G Biosciences. D-Myoinositol 1,4,5-trisphosphate hexapotassium salt and protein A/G PLUS-agarose beads were obtained from Santa Cruz Biotechnology. [ 3 H]IP 3 was obtained from PerkinElmer Life Sciences. KCl and NaCl were obtained from Amresco. Glucose was obtained from Calbiochem EMD Millipore. Mouse anti-Chicken IgM was obtained from Southern Biotech. All other materials were obtained from Sigma.
Generation of Concatenated IP 3 R Constructs-R1 and R2 dimeric constructs were generated by designating IP 3 R cDNA encoding the corresponding isoforms as "head" and "tail" and modifying them appropriately using a QuikChange mutagenesis strategy as described before (39). R3 head subunits were generated by silently mutating NcoI sites in the rat R3 coding sequence (primers 1-6). After the start codon, an alanine codon was added to enable insertion of a new NcoI site, in addition to a Kozak sequence (primers 7 and 8). The head subunit was further modified by deleting a stop codon and introducing a nucleotide sequence encoding the first half of the linker followed by an AgeI site (primers 9 and 10), immediately after the R3 coding sequence. The R3 tail subunit was generated by introducing an AgeI site followed by a nucleotide sequence encoding the second half of the linker immediately before the start codon (primers 11 and 12). Additionally, a blunt end restriction site (HpaI) was inserted immediately after the stop codon (primers 13 and 14). The coding sequence for all constructs was confirmed by sequencing. To generate IP 3 R dimers encoding homo-and heterodimeric receptors, the appropriate fragments were directly ligated between the two arms of pJAZZ Mamm linear vector to generate a construct encoding two IP 3 R subunits connected with a 14-amino acid linker within a single open reading frame. The tetrameric R1R1R1R2 construct was engineered by ligating, C to N terminus, subunits 1-3 of the R1R1R1R1 homotetramer and the R2 tail subunit, generated as described previously (39).
Generation of ATPB-binding Site Mutants-A QuikChange mutagenesis strategy was employed to modify the Walker A-like motif ATPB site in cDNAs encoding the mouse R2 monomer, the R2 dimer head and tail subunits. Briefly, forward and reverse mutagenic primers (primers 15 and 16) were generated by Integrated DNA Technologies to encode mutations at G5940, G5946, and G5955, in addition to silently introducing a PspOMI restriction site. These mutations code for amino acid substitutions G1969A, G1971A, and G1974A, respectively. The coding sequence for all constructs was confirmed by sequencing. The R2R2 ⌬ and R2 ⌬ R2 ⌬ dimers were generated using the same approach used to generate the wild type R2 homodimer, as described above.
Cell Culture and Generation of DT40 Cells Stably Expressing IP 3 Rs-DT40-3KO cells and DT40-3KO cells stably expressing IP 3 R constructs were maintained at 39°C and 5% CO 2 in RPMI 1640 media supplemented with 10% FBS, 1% chicken serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 10 M ␤-mercaptoethanol. Cells were maintained by subculturing every 3 days. Selection was carried out using G418 sulfate (geneticin). DT40-3KO cell transfection and stable line generation were performed as described previously (34). Cells stably expressing monomeric or concatenated IP 3 R constructs were lysed using Triton X-100 lysis buffer as described before.
Western Blotting-DT40-3KO cells or DT40 cells stably expressing monomeric or concatenated IP 3 R constructs were lysed using Triton X-100 lysis buffer (50 mM Tris base, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, pH 8), supplemented with protease inhibitors. All lysates were incubated on ice for 30 min, interspersed with occasional vortexing and sonication to ensure adequate disruption of membranes. Lysates were cleared by centrifugation (16,000 ϫ g, 10 min, 4°C) and resolved on SDS-PAGE. The proteins were transferred to a nitrocellulose membrane (Bio-Rad) and immunoblotted using isoform-specific primary antibodies and corresponding secondary antibodies. Membranes were visualized using an Odyssey infrared imaging system (LICOR Biosciences).
Native-PAGE-DT40-3KO cells expressing various constructs were harvested by centrifugation and lysed in CHAPS lysis buffer (40 mM NaCl, 25 mM HEPES, 10 mM CHAPS, 1 mM EDTA, pH 7.4) supplemented with protease inhibitors. After 30 min on ice at 4°C, lysates were cleared by centrifugation at 16,000 ϫ g for 10 min at 4°C. Cleared lysates were mixed with of 4ϫ sample buffer, 5% G-250 sample additive, and fractionated 3-12% native-PAGE TM Novex gels. Separated proteins were transferred to polyvinylidene difluoride (PVDF) membranes and probed using the indicated primary antibodies and the appropriate horseradish peroxidase-conjugated secondary antibodies. Protein bands were detected using enhanced chemiluminescent substrate.
Size Exclusion Chromatography-A small section of frozen bovine salivary gland (Pel-Freez Biologicals Inc., Rogers, AR) was excised, washed in ice-cold homogenization buffer (10 mM Tris, 150 mM sucrose, 150 mM KCl, 1.5 mM MgCl 2 , pH 7.5) containing protease inhibitors, and flash-frozen in liquid N 2 . In the presence of liquid N 2 , the salivary section was ground to a fine powder and lysed on ice for 1 h using CHAPS lysis buffer (10 mM CHAPS, 25 mM HEPES, 120 mM NaCl, 1 mM EDTA, 1 mM DTT, pH 8) containing protease inhibitors. During the hour, the lysate was subject to repetitive vortexing, needle homogenization (26-gauge 5/8-inch), and mechanical homogenization using an RZR 2021 homogenizer (Heidolph Instruments, Germany) with 50 strokes at 1200 rpm. The lysates were then centrifuged twice at 1000 ϫ g for 15 min at 4°C. The supernatants were transferred to polyallomer Beckman centrifuge tubes and centrifuged at 130,000 ϫ g twice for 30 min at 4°C. A small sample of the lysate was mixed with SDS loading buffer and saved as the input control. 1100 l of the clarified extract was injected onto the Sephacryl-400 column. 36 ϫ 2-ml fractions were collected post-void at a 1 ml/ml flow rate. Equivalent amounts of each fraction were fractionated on 5% SDS-PAGE, transferred to nitrocellulose, and probed with the indicated antibodies. Co-immunoprecipitation was performed on pooled fractions corresponding to ϳ1.1 MDa by incubation with protein A/G plus-agarose beads and the described isoform-specific antibodies overnight at 4°C, tumbled end over end. Beads were washed with Triton X-100 lysis buffer, resuspended in SDS loading buffer, and boiled at 75°C for 5 min. The samples were resolved on 5% SDS-PAGE, transferred to nitrocellulose membranes, and probed using the indicated antibodies.
Fluorescence Imaging-DT40 cells expressing defined IP 3 R constructs were loaded with 2 M Fura-2 AM on a glass coverslip mounted onto a Warner chamber at room temperature for 20 -30 min. Loaded cells were perfused with HEPES imaging buffer (137 mM NaCl, 4.7 mM KCl, 1.26 mM CaCl 2 , 1 mM Na 2 HPO 4 , 0.56 mM MgCl 2 , 10 mM HEPES, 5.5 mM glucose, pH 7.4) and stimulated with the desired agonist. Ca 2ϩ imaging was performed using an inverted epifluorescence Nikon microscope with a ϫ40 oil immersion objective. Cells were alternately excited at 340 and 380 nm, and emission was monitored at 505 nm. Images were captured every second with an exposure of 10 ms and 4 ϫ 4 binning using a digital camera driven by TILL Photonics software.

96-Well Permeabilized Cell IP 3 -induced Ca 2ϩ Release (IICR)
Assays-This protocol was modified from one described previously (50). DT40-3KO cells stably expressing defined IP 3 R constructs were grown up to near confluency, harvested, and washed twice with HEPES imaging buffer. Cells were incubated in darkness with 20 M MgFluo4 AM for 1 h at room temperature on a rocker. After loading, cells were washed, resuspended in Ca 2ϩ -free media (140 mM KCl, 20 mM NaCl, 20 mM PIPES, 1 mM EGTA, and 2 mM MgCl 2 , pH 7.0), and permeabilized using 10 g/ml saponin. Permeabilization was confirmed by visualizing trypan blue accumulation. Permeabilized cells were subsequently washed, resuspended in Mg 2ϩ -free media (140 mM KCl, 20 mM NaCl, 20 mM PIPES, 1 mM EGTA, 375 M CaCl 2 , pH 7.0), and dispensed into a black-walled flat bottom 96-well plate (ϳ500,000 cells/well). The plate was spun at 200 ϫ g for 2 min to plate cells to the bottom of each well. The plate was allowed to rest for 30 min prior to commencing the assay. Fluorescence imaging was carried out using FlexStation 3 from Molecular Devices (excitation 490 nm and emission 525 nm) and analyzed by using SoftMax Pro Microplate Data Acquisition and Analysis software. Stores were loaded by adding 1.5 mM Mg-ATP to activate SERCA. Upon loading, SERCA was disabled using cyclopiazonic acid, and carbonyl cyanide p-trifluoromethoxyphenylhydrazone was added to uncouple mitochondria. IP 3 Rs were then activated by the addition of varying [IP 3 ] in the presence of 5 mM ATP. The initial rates of Ca 2ϩ release were determined by fitting the curves to a single exponential function (OriginPro 6.1). Dose-response curves were plotted using the determined rates.
Single Cell Permeabilized DT40 Cell IICR Assays-DT40-3KO cells stably expressing defined IP 3 R constructs were loaded with 20 M MgFura-2 AM for 50 -60 min on a glass coverslip mounted onto a Warner chamber at room temperature in darkness. Cells were subsequently permeabilized by superfusion of 40 M ␤-escin in intracellular medium (ICM) (125 mM KCl, 19 mM NaCl, 10 mM HEPES, 1 mM EGTA, pH 7.3). The duration of permeabilization was dependent on the flow rate, and care was taken to prevent excessive (internal membrane) permeabilization by careful monitoring of the fluorescence. Permeabilized cells were then washed in ICM without ␤-escin for 15 min to allow for removal of cytosolic dye. The internal stores of permeabilized cells were loaded by activating SERCA through superfusion with ICM containing 1.4 mM MgCl 2 , 3 mM Na-ATP, and 0.65 mM CaCl 2 (free [Ca 2ϩ ] of 200 nM (MaxChelator freeware)). Upon stabilization of fluorescence, MgCl 2 was removed from the superfused solution to disable SERCA and allowed to flow through for 1 min. Varying [IP 3 ] in the absence or presence of varying [ATP] were applied through superfusion to induce unidirectional Ca 2ϩ release from internal stores. After washing out IP 3 , stores were repeatedly refilled and released to allow for repeated stimulations. Each experiment contained between 30 and 60 cells in a field of view and was performed a minimum of three times. The initial rates of Ca 2ϩ release were determined by fitting the release curves of each individual cell to a single exponential function (OriginPro 6.1). Dose-response curves were plotted using the determined rates. All imaging was performed on an inverted epifluorescence Nikon microscope with a ϫ40 oil immersion objective. Cells were alternately excited at 340 and 380 nm, and emission was monitored at 505 nm. Images were captured every 5 s during permeabilization, loading, and disabling and every second during release. This was done with an exposure of 10 ms and 4 ϫ 4 binning using a digital camera driven by TILL Photonics software.
Homologous Competitive IP 3 Binding Assay-DT40-3KO cells stably expressing IP 3 R constructs were lysed using Triton X-100 lysis buffer supplemented with protease inhibitors. Lysates were incubated on ice for 30 min, interspersed with occasional vortexing, and subsequently cleared by centrifugation twice at 16,000 ϫ g for 20 min at 4°C. An overnight immunoprecipitation was performed on the cleared lysate by adding protein A/G PLUS-agarose beads and antibody against the specific isoform at 4°C. Post-immunoprecipitation, the beads were washed with binding buffer (50 mM Tris base, 1 mM EDTA, pH 8, plus 1 mM ␤-mercaptoethanol added freshly upon use) and equally aliquoted into Eppendorf tubes. The binding reaction was performed in a 100-l volume containing the immunoprecipitated protein, 2.5 nM [ 3 H]IP 3 , and varying concentrations of unlabeled IP 3 for 1 h at 4°C, vortexed every 10 min. After incubation, beads were pelleted by centrifugation (16,000 ϫ g for 2 min), and the supernatants were carefully aspirated. The beads were incubated with 500 l of 1% SDS overnight at room temperature. The next day, the tube contents were transferred to scintillation vials, mixed with scintillation liquid, and bound radioactivity measured using a liquid scintillation counter. Nonspecific binding was defined as the radioactivity measured in the presence of 50 M cold IP 3 . Specific binding was determined by subtracting the counts/min values obtained in the presence of 50 M cold from the counts/min values obtained with other conditions. Total specific binding was determined as the binding observed in the absence of cold IP 3 . All values were normalized to total specific binding. Normalized specific binding from three experiments was averaged, and curves were fit using a logistic dose-response equation using the OriginPro 6.1 software.
Patch Clamp Experiments-Single IP 3 R channel potassium currents (I K ) were measured in the on-nucleus voltage clamp configuration of the patch clamp technique using PClamp9 and an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA), as described previously (49). Pipette solutions contained 140 mM KCl, 10 mM HEPES, 1,2-bis(2-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid, 200 nM free Ca 2ϩ , and the indicated concentrations of IP 3 and ATP. Free [Ca 2ϩ ] was calculated using Max Chelator freeware and verified fluorometrically. Traces were consecutive 3-s sweeps recorded at Ϫ100 mV, sampled at 20 kHz, and filtered at 5 kHz. A minimum of 15 s of recordings was considered for data analyses. Current/voltage relationships were generated by obtaining multiple sweeps at the indicated holding potential. Pipette resistances were typically 20 megohms and seal resistances were Ͼ5 gigaohms.
Data Analysis-Single channel openings were detected by half-threshold crossing criteria using the event detection protocol in Clampfit9. We assumed that the number of channels in any particular nucleus was represented by the maximum number of discrete stacked events observed during the experiment. Even at low P o , stacking events were evident (data not shown). Only patches with one apparent channel were considered for analyses of mean open and closed times and open probability. The slope conductances were determined from the linear fits of the current-voltage relationships where g ϭ I K /(V Ϫ V K ). Equation parameters were estimated using a non-linear least squares algorithm.

IP 3 R Isoforms Oligomerize Forming Heterotetramers in
Native Tissues-The ubiquitous expression of IP 3 Rs is well established, with numerous studies unequivocally demonstrating that individual cell types express at least two if not all three isoforms (38,39). Co-immunoprecipitation experiments performed on lysates isolated from individual cell types or tissues have shown that these isoforms do associate (39,42,43,51). However, they fail to exclude the possibility that these observed interactions are the result of larger intermolecular complexes of homotetrameric channels. To address this issue, lysates from bovine salivary glands, which express all three isoforms (Fig.  1A), were fractionated by gel filtration. A total of 36 ϫ 2-ml fractions were collected across the entire elution spectrum, and alternate fractions were resolved on SDS-PAGE and probed using antisera against the C terminus of R1 or R2 (Fig. 1B). Although some immunoreactivity was detected at higher molecular weights, this may be attributed to the association of native tetrameric IP 3 Rs with various accessory proteins or the formation of macromolecular complexes of homo-and heterotetrameric IP 3 Rs. Fractions corresponding to the approximate molecular mass of the native IP 3 R (ϳ1000 -1200 kDa) were pooled and subject to co-immunoprecipitation. A band was detected at ϳ250 kDa when immune complexes from the pooled fractions were captured with ␣-R1 and probed with ␣-R2CT (Fig. 1C). This indicates that R1 and R2 are clearly capable of associating with each other to form a heterotetrameric IP 3 R complex, thereby providing the first definitive evidence of endogenous heterotetramer formation in a native tissue.
Establishing Stable Cell Lines Expressing Homo-or Heterodimer IP 3 R Constructs-Having confirmed that different isoforms oligomerize to form heterotetrameric IP 3 Rs natively, an obvious question is as follows. What is the contribution of each isoform within a heterotetramer to overall receptor function? To address this, cDNA constructs encoding concatenated IP 3 R homo-and heterodimers were engineered and stably expressed in DT40-3KO cells, an IP 3 R null background cell line (39). Fig. 2A demonstrates that concatenated homo-or heterodimers encoding R2 or R3 can be stably expressed in DT40-3KO cells. Immunoblots probed with ␣-R2NT show that monomeric R2 migrated at ϳ250 kDa, whereas dimers containing R2 migrated at ϳ500 kDa, as expected. No bands were observed for 3KO or R3 monomers or homodimers (R3R3). Similarly, immunoblots probed with ␣-R3 demonstrated expression of monomeric (ϳ250 kDa) and dimeric (ϳ500 kDa) R3 migrating as expected, with no immunoreactivity seen in lanes containing 3KO or R2 monomers or homodimer (R2R2) lysates. Additionally, no lower order degradation products were detected in cells expressing dimers, firmly establishing that the composition of concatenated IP 3 Rs being studied was defined. These dimeric proteins are also capable of oligomerizing to form tetramers (Fig. 2B). Under non-denaturing blue native-PAGE conditions, IP 3 R monomers or dimers migrated at ϳ1100 -1200 kDa when probed with either ␣-R2NT or ␣-R3, corresponding to the established molecular mass of assembled tetrameric IP 3 Rs.
A recent cryo-EM structure indicates that the C and N termini of adjacent IP 3 R monomers are situated in close proximity to the assembled tetrameric channel (52). Nevertheless, a major concern when designing and utilizing the concatenated constructs was whether linking the N terminus, containing the IP 3 binding domain, to the C terminus of the adjacent subunit, containing the pore domain, would constrain the tetrameric channel and thereby affect the conformational change required for receptor function. Therefore, homologous competitive IP 3 binding assays were carried out to ensure that the monomers and concatenated dimers bound IP 3 with similar affinities. Monomers of the three isoforms exhibited similar apparent affinities to those previously reported (23), with R2 having the highest apparent affinity (14 Ϯ 4 nM) followed by R1 (50 Ϯ 5 nM) and finally R3 (171 Ϯ 19 nM) (Fig. 2C). Furthermore, the data clearly show that the apparent IP 3 binding affinities of the monomers and homodimers are indistinguishable, with R2R2 having the highest affinity at 18 Ϯ 4 nM, followed by R1R1 (44 nM Ϯ 7 nM), and finally R3R3 (186 Ϯ 17 nM). Analysis of the slopes also revealed the binding for R1 and R1R1 is non-cooperative, consistent with previous work by Iwai et al. (53). However, contrary to their findings, we show R3 and R3R3 also exhibit non-cooperative binding, whereas R2 and R2R2 display negative cooperative binding (Hill coefficients of 0.7 and 0.8, respectively).
To ensure that the monomers and homodimers were functionally indistinguishable, a direct analysis of IP 3 R activity was FIGURE 1. IP 3 R subtypes oligomerize to form heterotetramers in native tissues. Cleared lysates from bovine salivary gland were prepared using CHAPS lysis buffer. A, lysates were resolved on 5% SDS-PAGE and immunoblotted with ␣-R1, ␣-R2CT, or ␣R3. Lysates from DT40 3KO cells stably expressing R1, R2, or R3 were used as controls. B, bovine salivary gland lysates cleared through ultracentrifugation were subject to size exclusion chromatography. Alternate fractions were resolved on 5% SDS-PAGE and probed using ␣-R1 or ␣-R2CT. IP 3 R immunoreactivity was quantified using ImageJ and normalized to peak immunoreactivity. C, fractions 9 -11 corresponding to native IP 3 R (ϳ1000 -1200 kDa) were pooled at subject to co-immunoprecipitation using ␣-R1 and protein A/G beads. Immunoprecipitates (IP) were probed using ␣-R1 or ␣-R2CT. Lysates from DT40-3KO cells stably expressing R1, R2, or R3 were used as controls. Representative immunoblots are shown.
carried out using a unidirectional permeabilized cell IICR assay on a Flexstation 96-well plate reader. Under optimal conditions for IP 3 R activity (200 nM free Ca 2ϩ , 5 mM ATP) (24,25,48,49), the three isoforms follow the previously established rank order of IP 3 sensitivity, with R2 exhibiting the lowest EC 50 (392 Ϯ 74 nM) followed by R1 (1.84 Ϯ 0.32 M) and R3 (6.01 Ϯ 0.61 M) (Fig. 2D). Furthermore, these data clearly show that the homodimers are indistinguishable from their parent monomers; R2R2 exhibiting an IP 3 sensitivity of 497 Ϯ 5 nM, followed by R1R1 with 2.14 Ϯ 0.23 M and R3R3 with 6.58 Ϯ 0.47 M. Interestingly, an analysis of the slopes revealed that although R1, R1R1, R2, and R2R2 were found to display non-cooperative to marginally cooperative release (Hill coefficients of 1.0, 1.2, 1.2, and 1.5, respectively), R3 and R3R3 exhibited positive cooperative release (2.48 and 2.09, respectively). Although the reasons for these differences are unclear, they could be explained by isoform-specific differences in functional regulation by various accessory proteins (12), post-translational modifications (54), or reactive oxygen species (55). Finally, although the observed cooperativity of IP 3 -mediated Ca 2ϩ release is not as high as reported previously (56 -58), Taylor and Laude (59) have suggested that positive feedback by Ca 2ϩ could amplify the steepness of IP 3 Fig. 3 shows that R2-containing heterodimers exhibited an intermediate IP 3 sensitivity between that of the two constituent isoforms (Fig. 3A). For example, channels assembled from R2R3 and R3R2 heterodimers exhibited comparable EC 50 of 2.6 Ϯ 0.14 and 2.3 Ϯ 0.15 M, respectively, both intermediate to the IP 3 sensitivities of R2R2 and R3R3 (Fig. 3A). Similarly, the R1R2 heterodimer exhibited an EC 50 of 1.2 Ϯ 0.17 M, an EC 50 intermediate to those exhibited by R1R1 and R2R2 (Fig. 3B). This suggests that both isoforms within heterodimers contribute toward determining its unique IP 3 sensitivity at optimal [ATP] and [Ca 2ϩ ]. Both heterodimers were also shown to have marginally cooperative release.
Blended IP 3 Sensitivity Dictated by Apparent IP 3 Binding Affinity-The three major essential regulators of IP 3 R function are IP 3 , Ca 2ϩ , and ATP (12). As [ATP] and [Ca 2ϩ ] were tightly clamped in the permeabilized cell assays to achieve maximal release, we speculated that the apparent IP 3 binding affinity was potentially the major determinant of the intermediate IP 3 sensitivity observed with the heterodimers under these conditions. Using homologous competitive binding IP 3 assays, we focused on the R3R2 heterodimer as R2 and R3 have the most distinct binding affinities. We hypothesized that binding reflecting the two independent binding affinities of the constituent isoforms would be observed and that the binding curve generated could be fit by a bimodal logistic equation. However, the R3R2 heterodimer exhibited a unique unimodal apparent IP 3 binding affinity of 94 Ϯ 6 nM (Fig. 4) approximately intermediate to the affinities of R2R2 and R3R3 (Fig. 2C). Notably, unlike the neg-

FIGURE 2. IP 3 R concatenated dimers are stably expressed in DT40 3KO cells, form tetramers, and are functionally indistinguishable from IP 3 R monomers.
A, lysates from DT40-3KO or DT40-3KO cells stably expressing IP 3 R constructs were resolved on 4% SDS-PAGE and probed with ␣-R2NT or ␣-R3. B, lysates from DT40-3KO cells stably expressing IP 3 R constructs were resolved on 3-12% native-PAGE TM Novex gels. Immunoblots were processed using ␣-R2NT or ␣-R3. C, [ 3 H]IP 3 binding curves of R1, R2, and R3 monomers, compared with R1R1, R2R2, and R3R3 homodimers, generated using homologous competitive binding assays. D, [IP 3 ]-response relationships of R1, R2, and R3 monomers, compared with R1R1, R2R,2 and R3R3 homodimers. High throughput permeabilized cell IICR assays were carried out using FlexStation 3. IP 3 -mediated Ca 2ϩ release was induced through addition of varying [IP 3 ] in the presence of 5 mM ATP and 200 nM free Ca 2ϩ . All values were normalized to the maximal release rate. Each point is the mean Ϯ S.E. of eight wells from at least three experiments. atively cooperative binding exhibited by R2R2 (Fig. 2C), binding in R3R2 was non-cooperative resembling R3 and R3R3.
The data presented in Fig. 4 suggest that the four individual IP 3 ligand-binding domains within a tetramer, rather than functioning as four independent sites, interact and contribute to determine the overall binding affinity of the assembled channel. Previous studies add support to the notion that adjacent subunits may interact to influence function. For example, Boehning and Joseph (40) demonstrate through a series of biochemical cross-linking and mutagenesis assays that the C terminus of one monomer interacts with the N terminus of another monomer, regardless of subtype, to influence tetramer functionality. This idea is also reinforced by a recent cryo-EM structure, which suggests that the N-terminal ligand-binding domain interacts with multiple contact sites in the C terminus of adjacent subunits to initiate the conformational changes required for channel gating (52). In total, our observations raise the possibility that IP 3 R heterotetramers with different subunit combinations could exhibit unique IP 3 binding affinities, thus extending the repertoire of IP 3 regulation of IP 3 R and in turn the sensitivity of Ca 2ϩ release.
R2 Determines the ATP Regulatory Properties of R3R2 Heterodimers-The previous data suggest that the IP 3 binding affinity of heterodimers is dictated by a contribution from both constituent isoforms. We next investigated whether a similar situation occurs when heterotetramers are regulated by ATP. Fig. 5, A-C, illustrates by using single cell IICR assays that the presence of saturating ATP (5 mM) markedly potentiated Ca 2ϩ release from cells expressing R2R2 only at submaximal [IP 3 ], consistent with previously reported findings using R2 monomers (24,49). Moreover, the presence of ATP appeared to induce a shift in IP 3 sensitivity, reducing the EC 50 from 2.45 Ϯ 0.93 M (0 ATP) to 0.62 Ϯ 0.18 M (5 mM ATP). In contrast, R3R3 requires ATP to become maximally activated with 5 mM ATP enhancing the rate of release even at 100 M IP 3 (from 0.16 Ϯ 0.02 s Ϫ1 to 0.29 Ϯ 0.02 s Ϫ1 ), again consistent with earlier reports (Fig. 5, D-F) (24). Furthermore, unlike R2R2, ATP does not appear to markedly alter IP 3 sensitivity (0 ATP, 6.69 Ϯ 1.06 M, versus 5 mM ATP, 7.26 Ϯ 1.16 M). It should also be noted that the EC 50 values observed using the permeabilized single cell IICR assays are comparable with those observed using the Flexstation 3-based IICR assays. Next, we investigated the sensitivity of Ca 2ϩ release to ATP in cells expressing these constructs (Fig. 6, A and B). Single cell IICR assays were performed where ATP levels were varied between 0 and 5 mM ATP at fixed [IP 3 ] and [Ca 2ϩ ]. Release was induced at 1 M for R2R2 and 10 M for R3R3-expressing cells because the differences in raw Ca 2ϩ release rates at 0 and 5 mM ATP were high at these [IP 3 ]. Fig. 6 shows that R2R2 clearly exhibited a considerably higher sensitivity to ATP (34 Ϯ 15 M) than R3R3 (447 Ϯ 175 M). These values are again comparable with those observed for R2 and R3 monomer constructs (ϳ41 M versus ϳ400 M, respectively). Having defined the properties of the homomeric R2R2 and R3R3 constructs, we then investigated the characteristics of Ca 2ϩ release in cells expressing R3R2 heterodimers. Strikingly, the presence of 5 mM ATP potentiated Ca 2ϩ release only at submaximal [IP 3 ] (Fig. 5, G-I) while reducing the EC 50 from 4.17 Ϯ 0.84 M to 2.07 Ϯ 0.31 M. Additionally, R3R2 heterodimers exhibited an ATP sensitivity of 47 Ϯ 19 M (Fig. 6C). Remarkably, these data demonstrate that both the mode of regulation of heterotetrameric R2R3 channels and the ATP sensitivity are identical to that observed for R2 channels. These data strongly suggest that R2 dictates the mode of ATP regulation in heteromers containing equal proportions of R2 and R3.
Previous investigations into IP 3 R regulation by ATP in native tissues or cultured cells corroborate our findings by demonstrating that the regulatory properties appear to be dictated by a specific isoform. For instance, Ca 2ϩ release assays performed on mouse pancreatic acini revealed that native IP 3 Rs exhibited an ATP sensitivity of 38 M and were only modulated at submaximal [IP 3 ], comparable with the properties of R2 (60). This is despite the fact that mouse pancreatic acini have roughly equal expression of R2 and R3, with the vast majority of receptors likely heterotetrameric in nature (39). Interestingly, pancreatic acini from R2 knock-out mice exhibited an ATP sensitivity of 450 M and were regulated by ATP at all [IP 3 ], identical to R3. Similar findings were observed in AR42J (R2 predominant) and RinM5F (R3 predominant) cells, which exhibited ATP sensitivities of 10 and 430 M, respectively. Interestingly, when RinM5F cells were transiently transfected with cDNA encoding R2, a 10-fold increase in ATP sensitivity was observed. Furthermore, Ca 2ϩ release was no longer potentiated at all [IP 3 ], rather only at submaximal [IP 3 ] (60). More recently,   using the on-nucleus patch clamp technique, our group showed that the open probability (P o ) of R1R2 heterodimers was only enhanced by ATP at low [IP 3 ], comparable with the properties exhibited by R2R2 (39). Taken together, these data suggest that a general property of R2 is to dictate the mode of ATP regulation of heteromers containing equal proportions of R2, irrespective of the additional subtypes in the tetramer. These data are in stark contrast to IP 3 binding, where a "blending" of the characteristics of the two constituent isoforms was observed. Indeed, the high "R2-like" sensitivity to ATP exhibited by R3R2 may suggest that the presence of two R2 subunits prevents ATP from binding to and regulating R3 altogether.
R2 Dictates the Temporal Pattern of Ca 2ϩ Release of R3R2 Heterodimers-As a consequence of the distinctive binding and regulatory properties of each IP 3 R isoform, the submaximal stimulation of B cell receptors in DT40 cells expressing a single isoform results in each subtype generating distinctive intracellular Ca 2ϩ release "signatures." This temporal pattern may be considered as the integrated response of the particular IP 3 R subtype to the multiple regulatory inputs experienced after B cell receptor stimulation. For instance, our group and others have shown that R2-expressing cells produce robust long lasting Ca 2ϩ oscillations, in stark contrast to the monophasic Ca 2ϩ spikes produced by R1 (24,39,61). A previous study from our laboratory has reported that modulation of R2 activity by ATP is important for maintaining Ca 2ϩ oscillations (24). ␣-IgM stimulation of cells expressing R3 monomers and homodimers reproducibly produced robust yet short lasting Ca 2ϩ oscillations (Fig. 7B), qualitatively and quantitatively distinctive from the oscillatory pattern generated by R2 expressing cells (Fig.  7A). Indeed, quantification of the number of transients elicited by cells (Fig. 7C) shows a larger percentage of responding cells expressing R3 monomers, or homodimers produced few oscillations (1-3 transients), compared with R2R2. Remarkably, R2R3 and R3R2 heterodimers (Fig. 7, D and E) both generated robust long lasting oscillatory Ca 2ϩ signals, similar to those generated by R2 monomers and homodimers. This is quantitatively represented in Fig. 7F, which shows that a smaller percentage of cells produced 1-3 transients, comparable with R2R2. We had also previously reported that R1R2 heterodimers produce long lasting Ca 2ϩ oscillations, similar to R2 (39). Together, these data further support the notion that two R2 subunits within a heterotetramer are sufficient to dictate the overall modulation of Ca 2ϩ release, including regulation by ATP binding, necessary to initiate an oscillatory profile of Ca 2ϩ signals generated by heteromeric IP 3 Rs.
Stoichiometry of ATP Regulation-By virtue of being able to define the exact composition of tetramers being expressed, this approach using concatenated IP 3 R can also be used to answer questions related to the stoichiometry required for channel function and regulation. The primary sequence of the three subtypes contains several putative ATP-binding Walker A-type motifs (GXGXXG) (24,25). However, mutagenic analysis has revealed that only a single such motif, the ATPB site in R2, is critical for ATP regulation of Ca 2ϩ release (16). This has led to the suggestion that sites other than the Walker A-type motifs or alternate accessory proteins are responsible for regulation by ATP in R1 and R3. Briefly, triple glycine-alanine substitutions at G1969A, G1971A, and G1974A in R2 monomers (R2 ⌬ ) were shown to result in 5 mM ATP being completely unable to potentiate Ca 2ϩ release at any [IP 3 ]. Additionally, R2 ⌬ lost the ability to produce the ␣-IgM-stimulated robust Ca 2ϩ oscillations characteristic of R2 (24).
We next examined the ATP regulatory properties of ⌬ATPB constructs by conducting permeabilized single cell IICR assays. Fig. 8, E-G, shows that Ca 2ϩ release was not potentiated by 5 mM ATP in R2 ⌬ R2 ⌬ , exhibiting an IP 3 sensitivity of 2.57 Ϯ 0.72 M. This EC 50 was comparable with that observed in Fig. 5A. Interestingly, the ability of 5 mM ATP to potentiate Ca 2ϩ release at submaximal [IP 3 ] was retained by R2R2 ⌬ . The presence of ATP visibly induced a shift in IP 3 sensitivity, reducing the EC 50 from 2.37 Ϯ 0.97 M (0 ATP) to 0.71 Ϯ 0.16 M (5 mM ATP). These values are also comparable with those seen in Fig. 5A. These observations clearly suggest that two functional ATPB motifs within an R2 homotetramer are sufficient to both facilitate ATP regulation and induce Ca 2ϩ oscillations. The inability of R2 ⌬ R2 ⌬ to maintain the characteristic Ca 2ϩ oscillations may be explained by the reduced sensitivity to IP 3 observed in Fig. 8D. We also sought to examine the contribution of the R2 ATPB motif to ATP regulation in cells expressing heterodimers. R2 dictates the mode of ATP regulation in R3R2 heterodimers. Characteristically, ATP potentiates Ca 2ϩ release only at submaximal [IP 3 ] (Fig. 5, G-I). Although the sites critical for regulation in R3 are unknown, a ⌬ATPB mutation in the R2 subunit of the heterodimer would still be predicted to leave the ATP regulatory sites in R3 intact and capable of binding ATP. Accordingly, we hypothesized that incorporation of an R2 ⌬ATPB mutation into R3R2 (R3R2 ⌬ ) would disable R2's ability to dictate ATP regulation and induce a shift to an R3 mode of regulation. Remarkably, however, permeabilized IICR assays (Fig. 9, A and B) demonstrated that R3R2 ⌬ is not regulated by ATP. For example, 5 mM ATP fails to augment Ca 2ϩ release at any [IP 3 ]. Similar observations were made with R1R2 ⌬ (Fig. 9, C  and D). This lack of sensitivity to ATP suggests that the two R2 ⌬ subunits within the heterotetramer were still likely determining the mode of ATP regulation, despite the R3/R1 subunits being unmodified and fully functional. It also suggests that the higher "R2-like" sensitivity to ATP exhibited by R3R2 (Fig. 6C) is unlikely to be the only cause of the dominant R2 activity with regard to regulation of Ca 2ϩ release. One possible explanation is that the presence of two R2 subunits within the heterotetramer is sufficient to prevent ATP binding to R3 or R1, thereby abrogating modulation by ATP.
Two R2 Subunits Are Necessary to Dictate ATP Regulation of Heterotetramers-Two R2 subunits within a heterotetramer are sufficient for R2 to dictate both the ATP regulatory proper-ties and the pattern of ␣-IgM-stimulated intracellular Ca 2ϩ signals. We next asked the question whether one R2 subunit was sufficient to dictate the function of heterotetramers. A concatenated tetramer containing three R1 subunits and one R2 subunit was successfully generated (R1R1R1R2) and stably expressed in DT40-3KO cells (Fig. 10A). Immunoblots probed with ␣-R1 show that monomeric R1 migrate at ϳ250 kDa followed by the R1R1 dimer, R1R1R1 trimer, and R1R1R1R2 heterotetramer at increasing molecular masses. The R1R1R1R2 concatemer was also functionally responsive to stimulation with submaximal [␣-IgM] (Fig. 10B), producing largely single Ca 2ϩ spikes that are qualitatively similar to the patterns exhibited by R1R1. Furthermore, quantitative analysis of the Ca 2ϩ signals revealed that over ϳ50% of responding cells produced 1-2 transients, generating a histogram profile similar to that of the R1R1 (ϳ70% producing 1-2 transients) and distinct from the R2R2 (ϳ5%) (Fig. 10C). These data, together with the previous report showing that R1R2 heterodimers produce long lasting Ca 2ϩ oscillations characteristic of R2 (39), suggest that one R2 subunit within a heterotetramer is not sufficient to dictate the patterns of Ca 2ϩ signals generated.
We next carried out an examination of the tetramer's regulation by ATP. We have previously reported that R1R2 heterodimers, forming tetramers containing two R1 and two R2 subunits, exhibited regulatory properties and single channel characteristics similar to R2R2 dimers (39). Specifically, on-nucleus patch clamp electrophysiology showed that the presence of 5 mM ATP in the patch pipette only enhanced channel activity at low (1 M) [IP 3 ] and not at maximal (10 M) [IP 3 ]. This was also confirmed using permeabilized single cell IICR assays (histogram, Fig. 10D). Furthermore, although ATP regulates the P o of both subtypes by increasing their total activity, it does so in biophysically distinct manners (49). At the single channel level, IP 3 Rs enter a "bursting" state in the presence of IP 3 regardless of subtype, with each burst characterized by relatively constant open and closed times. Accordingly, any rise in P o is the result of an increase in the duration spent bursting. Uniquely, the presence of saturating [ATP] augmented the P o of R1 and R1R1 homodimers by increasing the duration of bursts. In contrast, the P o of R2 and R2R2 homodimers was augmented by an increase in the number of bursts over a given period. An examination of R1R2 single channel activity revealed that ATP enhanced channel P o by increasing the number of bursts, indicating that R2 also dictated the single channel kinetics of heterotetramers containing equal numbers of R1 and R2 subunits (49). Strikingly, however, single channel analysis of R1R1R1R2 activity revealed that 5 mM ATP augmented the P o at both submaximal (1 M) and maximal (10 M) [IP 3 ] (Fig. 10, E and F), following a distinctly R1 mode of ATP regulation. Furthermore, it did so in a biophysical manner characteristic of R1: by increasing duration spent bursting rather than the number of bursts over a given period, a characteristic typical of R1 (49). Permeabilized single cell IICR assays confirm these data (Fig.  10, H-J), showing that ATP potentiates Ca 2ϩ release at both submaximal and maximal [IP 3 ], similar to the pattern displayed by R1R1 homodimers (Fig. 10G). These data are consistent with the requirement of at least two R2 subunits being necessary to dictate the mode of regulation of Ca 2ϩ release by ATP. Conclusion-Heterotetrameric IP 3 Rs likely play major roles in the generation of diverse intracellular Ca 2ϩ concentration signals in cells expressing multiple IP 3 R subtypes, which is a situation that is the norm in the majority of cells. However, elucidating the specific regulatory properties of heterotetrameric IP 3 R has been challenging to date, largely because of the difficulty in defining the complement of constituent isoforms expressed in a given situation. This problem is compounded by the difficulty in predicting how subtype-specific regulation of individual monomers, within the tetramer, might contribute to the channel's overall activity. To circumvent this issue, we utilized our previously described concatemer approach to produce pre-assembled IP 3 Rs with defined subunit composition, and we investigated fundamental characteristics of Ca 2ϩ release through heterotetrameric IP 3 Rs (39). This experimental paradigm has provided unique and non-intuitive insight into how important subtype-specific regulatory events contribute to channel activity, together with the stoichiometry of input required. Specifically, we show that IP 3 binding to individual ligand-binding domain monomers, although not markedly cooperative, is nevertheless influenced by binding to neighboring subunits to establish the overall binding affinity (Fig. 11A). This unique IP 3 binding affinity likely plays a role in dictating the intermediate IP 3 sensitivity exhibited by heterotetramers containing equal proportions of R2 and either R1 or R3 (Fig. 11B). The recent IP 3 R cryo-EM structure, which shows the N terminus of one subunit interacting with multiple contact points in the C terminus of adjacent subunits, supports this notion of inter-subunit cross-talk within a heterotetramer. Furthermore, it raises the tantalizing prospect that cells expressing different IP 3 R heterotetramer populations may exhibit distinct IP 3 binding affinities and sensitivities that suit the physiological needs of the cell.
Our data have also shown that, in terms of regulation of Ca 2ϩ release by ATP, the presence of two R2 subunits within a tetramer are necessary and sufficient for R2 to dictate the regulatory characteristics and ATP sensitivity of the heterote- tramer, regardless of the other subtypes present in the tetramer (Fig. 11, C and D). Previous work performed in pancreatic acini, which express relatively equal proportions of R2 and R3, also confirms that R2 dictates ATP regulatory properties of native IP 3 Rs in these cells. However, the physiological relevance of this dominance is unclear. One possibility is that the mode of ATP regulation and higher ATP sensitivity exhibited by R2 containing heterotetramers may play critical roles in determining metabolic fate at ER mitochondrial microdomains, especially during periods of metabolic stress when [ATP] are reduced. Indeed, our group previously showed that during periods of reduced ATP production, WT pancreatic acini generated larger Ca 2ϩ signals than in cells where R2 was knocked out (60). In such instances, the higher ATP sensitivity exhibited by R2-containing heterodimers contributes to the maintenance of metabolic homeostasis by potentiating ER mitochondrial Ca 2ϩ flux and stimulating ATP production via Ca 2ϩ -sensitive dehydrogenases (62,63).
In all, these distinct properties likely add further diversity to the spatial and temporal properties of Ca 2ϩ signals that can be evoked through an individual heterotetrameric IP 3 R. Furthermore, we have extended these studies to begin to understand the stoichiometry of modulatory input of IP 3 R by mutagenesis of defined numbers of regulatory motifs within the tetramer, specifically demonstrating that two ATPB sites within R2 homotetramers are required for ATP to potentiate Ca 2ϩ release. We envision that this powerful experimental platform can be utilized to define the stoichiometry of other regulatory inputs, as well as illuminate the consequences of incorporating subunits encoding splice variants or mutations associated with human diseases within tetrameric IP 3 Rs on Ca 2ϩ release.
Author Contributions-R. C. designed and stably expressed constructs, collected and analyzed the data, drafted the manuscript, and prepared the figures. K. J. A. was responsible for designing and establishing the concatemer strategy. L. E. W. collected and analyzed data obtained through single channel electrophysiology. D. I. Y. was responsible for the conception and design of concatenated strategy, as well as data analysis, generation of figures, and editing of the manuscript. All authors approved the final version. FIGURE 11. Novel insights into the function and regulation of IP 3 R heterotetramers. Individual subtypes are represented as distinct colors as follows: R1, red; R2, blue; and R3, green. A, R3R2 heterodimers, which generate heterotetramers containing an equal proportion of R2 and R3, exhibit unique IP 3 binding properties (affinity, cooperativity) that are distinct to that of the two constituent subtypes. B, at optimal conditions for Ca 2ϩ release (200 nM Ca 2ϩ , 5 mM ATP), heterodimers containing R2 and either R1 or R3 exhibit an IP 3 sensitivity intermediate to that of the two constituent subtypes that is potentially reflective of the heterodimers' unique IP 3 binding affinity. C, when ATP levels are reduced to submaximal concentrations, two R2 subunits are sufficient to dictate the regulatory properties of heterotetramers containing equal proportions of R2 and either R1 or R3. D, one R2 monomer is not sufficient to maintain R2 regulatory dominance in a heterotetramer containing three R1 subunits.