Molecular Basis of the Isoform-specific Ligand-binding Affinity of Inositol 1,4,5-Trisphosphate Receptors
2007; Elsevier BV; Volume: 282; Issue: 17 Linguagem: Inglês
10.1074/jbc.m609833200
ISSN1083-351X
AutoresMiwako Iwai, Takayuki Michikawa, Ivan Bosanac, Mitsuhiko Ikura, Katsuhiko Mikoshiba,
Tópico(s)Receptor Mechanisms and Signaling
ResumoThree isoforms of the inositol 1,4,5-trisphosphate (IP3) receptor (IP3R), IP3R1, IP3R2, and IP3R3, have different IP3-binding affinities and cooperativities. Here we report that the amino-terminal 604 residues of three mouse IP3R types exhibited Kd values of 49.5 ± 10.5, 14.0 ± 3.5, and 163.0 ± 44.4 nm, which are close to the intrinsic IP3-binding affinity previously estimated from the analysis of full-length IP3Rs. In contrast, residues 224–604 of IP3R1 and IP3R2 and residues 225–604 of IP3R3, which contain the IP3-binding core domain but not the suppressor domain, displayed an almost identical IP3-binding affinity with a Kd value of ∼2 nm. Addition of 100-fold excess of the suppressor domain did not alter the IP3-binding affinity of the IP3-binding core domain. Artificial chimeric proteins in which the suppressor domain was fused to the IP3-binding core domain from different isoforms exhibited IP3-binding affinity significantly different from those of the proteins composed of the native combination of the suppressor domain and the IP3-binding core domain. Systematic mutagenesis analyses showed that amino acid residues critical for type-3 receptor-specific IP3-binding affinity are involved in Glu-39, Ala-41, Asp-46, Met-127, Ala-154, Thr-155, Leu-162, Trp-168, Asn-173, Asn-176, and Val-179. These results indicate that the IP3-binding affinity of IP3Rs is specifically tuned through the intramolecular attenuation of IP3-binding affinity of the IP3-binding core domain by the amino-terminal suppressor domain. Moreover, the functional diversity in ligand sensitivity among IP3R isoforms originates from at least the structural difference identified on the suppressor domain. Three isoforms of the inositol 1,4,5-trisphosphate (IP3) receptor (IP3R), IP3R1, IP3R2, and IP3R3, have different IP3-binding affinities and cooperativities. Here we report that the amino-terminal 604 residues of three mouse IP3R types exhibited Kd values of 49.5 ± 10.5, 14.0 ± 3.5, and 163.0 ± 44.4 nm, which are close to the intrinsic IP3-binding affinity previously estimated from the analysis of full-length IP3Rs. In contrast, residues 224–604 of IP3R1 and IP3R2 and residues 225–604 of IP3R3, which contain the IP3-binding core domain but not the suppressor domain, displayed an almost identical IP3-binding affinity with a Kd value of ∼2 nm. Addition of 100-fold excess of the suppressor domain did not alter the IP3-binding affinity of the IP3-binding core domain. Artificial chimeric proteins in which the suppressor domain was fused to the IP3-binding core domain from different isoforms exhibited IP3-binding affinity significantly different from those of the proteins composed of the native combination of the suppressor domain and the IP3-binding core domain. Systematic mutagenesis analyses showed that amino acid residues critical for type-3 receptor-specific IP3-binding affinity are involved in Glu-39, Ala-41, Asp-46, Met-127, Ala-154, Thr-155, Leu-162, Trp-168, Asn-173, Asn-176, and Val-179. These results indicate that the IP3-binding affinity of IP3Rs is specifically tuned through the intramolecular attenuation of IP3-binding affinity of the IP3-binding core domain by the amino-terminal suppressor domain. Moreover, the functional diversity in ligand sensitivity among IP3R isoforms originates from at least the structural difference identified on the suppressor domain. The inositol 1,4,5-trisphosphate (IP3) 3The abbreviations used are: IP3, inositol 1,4,5-trisphosphate; IP3R, IP3 receptor. 3The abbreviations used are: IP3, inositol 1,4,5-trisphosphate; IP3R, IP3 receptor. receptors (IP3Rs) function as IP3-gated Ca2+ release channels located on intracellular Ca2+ stores, such as the endoplasmic reticulum (1Furuichi T. Kohda K. Miyawaki A. Mikoshiba K. Curr. Opin. Neurobiol. 1994; 4: 294-303Crossref PubMed Scopus (196) Google Scholar). Mammalian IP3R family consists of three isoforms (IP3R1, IP3R2, and IP3R3), and they form homotetrameric or heterotetrameric channels (2Monkawa T. Miyawaki A. Sugiyama T. Yoneshima H. Yamamoto-Hino M. Furuichi T. Saruta T. Hasegawa M. Mikoshiba K. J. Biol. Chem. 1995; 270: 14700-14704Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). There is evidence of a functional difference among the three isoforms of IP3R in terms of their IP3 sensitivity (3Newton C.L. Mignery G.A. Sudhof T.C. J. Biol. Chem. 1994; 269: 28613-28619Abstract Full Text PDF PubMed Google Scholar, 4Nerou E.P. Riley A.M. Potter B.V. Taylor C.W. Biochem. J. 2001; 355: 59-69Crossref PubMed Scopus (37) Google Scholar, 5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar) and cooperativity with respect to IP3 binding (5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). The intrinsic association constants of mouse IP3R1, IP3R2, and IP3R3 expressed in Sf9 cells are estimated to be 3.5 × 107, 1.7 × 108, and 3.4 × 106 (m-1), respectively (5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). IP3R2 exhibits both negative and positive cooperativity, whereas IP3R3 exhibits negative IP3 binding cooperativity (5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). This diversity of responsiveness to IP3 observed among the three IP3R isoforms may contribute to the generation of the different degree of IP3 sensitivity of the Ca2+ store. The molecular basis of the isoform-specific IP3-binding affinity, however, is not well understood. The IP3-binding domain of IP3R1 is composed of two functional domains, the amino-terminal suppressor domain and the carboxyl-terminal IP3-binding core domain (6Yoshikawa F. Morita M. Monkawa T. Michikawa T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1996; 271: 18277-18284Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). The IP3-binding core domain is the minimum region required for specific IP3 binding and is mapped within residues 226–578 of mouse IP3R1, a polypeptide of 2749 residues (6Yoshikawa F. Morita M. Monkawa T. Michikawa T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1996; 271: 18277-18284Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). The amino-terminal 225 amino acid residues of IP3R1 function as the suppressor for IP3 binding, and deletion of these residues results in significant enhancement of IP3 binding (6Yoshikawa F. Morita M. Monkawa T. Michikawa T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1996; 271: 18277-18284Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). The atomic resolution structures of both the suppressor domain (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar) and the IP3-binding core domain (8Bosanac I. Alattia J.R. Mal T.K. Chan J. Talarico S. Tong F.K. Tong K.I. Yoshikawa F. Furuichi T. Iwai M. Michikawa T. Mikoshiba K. Ikura M. Nature. 2002; 420: 696-700Crossref PubMed Scopus (272) Google Scholar) of mouse IP3R1 were solved by x-ray crystallography to 1.8- and 2.2-Å resolution, respectively. The IP3-binding core domain comprises two asymmetric domains, the β-domain and α-domain. A highly positive-charged pocket is created at the interface of these two domains to which an IP3 molecule binds. The 11 amino acid residues in the IP3-binding core domain of IP3R1 are responsible for the coordination of IP3 (8Bosanac I. Alattia J.R. Mal T.K. Chan J. Talarico S. Tong F.K. Tong K.I. Yoshikawa F. Furuichi T. Iwai M. Michikawa T. Mikoshiba K. Ikura M. Nature. 2002; 420: 696-700Crossref PubMed Scopus (272) Google Scholar), and all of them except Gly-268 are conserved in other isoforms. The suppressor domain contains a β-trefoil fold and a helix-turn-helix structure inserted between two β-strands of the β-trefoil fold (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar). The conserved 7 amino acid residues, which are clustered on one side of the suppressor domain, were found to be critical for the suppression of IP3 binding (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar). These structural and functional analyses of the IP3-binding domain have been carried out mainly using the type-1 isoform, and the suppression ability of the amino-terminal regions of IP3R2 and IP3R3 has not been well characterized. In the present study, we analyzed the IP3-binding affinity of the IP3-binding domain of three isoforms and found that the intramolecular suppression of IP3 binding specifically tunes the IP3-binding affinity of IP3R isoforms. Systematic analyses of chimeric mutants and site-directed mutants created based on the three-dimensional structures of the suppressor domain provide us with a novel insight into the structural basis of the isoform-specific IP3-binding affinity of the three IP3R types. Gene Construction—Plasmids carrying the amino-terminal 604 amino acid residues of mouse IP3R1, IP3R2, and IP3R3 (T604m1, T604m2, and T604m3, respectively) were constructed as follows. The 1.8-kbp fragments of the three IP3R genes were amplified with the PCR method from pBact-STneoB-C1, pBluescriptII-C2, and pBact-STneoB-C3 (5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar), with three pairs of primers, P1/P4, P2/P5, and P3/P4 (supplemental Table S1), respectively. The fragments were digested with NdeI and EcoRI and were subcloned into the modified pRSET-A (Invitrogen), in which the hexa-histidine (His6) tag coding region was removed, to generate pRSET-T604m1, pRSET-T604m2, and pRSET-T604m3, respectively. To express amino acid residues 224–604 of IP3R2 and 225–604 of IP3R3 (supplemental Fig. S1), plasmids were constructed as follows. The 1.1-kbp fragments were amplified from pBluescriptII-C2 and pBact-STneoB-C3 with pairs of primers, P6/P5, and P7/P4 (supplemental Table S1), respectively. The fragments were digested with NdeI and EcoRI and were subcloned into the His6 tag-removed pRSET-A to generate pRSET-(224–604)m2 and pRSET-(225–604)m3, respectively. For the expression of residues 224–604 of mouse IP3R1, pET-D-(1–223)/T604 (9Yoshikawa F. Uchiyama T. Iwasaki H. Tomomori-Satoh C. Tanaka T. Furuichi T. Mikoshiba K. Biochem. Biophys. Res. Commun. 1999; 257: 792-797Crossref PubMed Scopus (42) Google Scholar) was used. For the expression of the His6-tagged suppressor domain (residues 2–223) of mouse IP3R1 (supplemental Fig. S1), pET23a-T7-IP3Rsup-6His (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar) was used. Plasmids carrying chimeric proteins, in which the suppressor domains were fused to the IP3-binding core domain from different isoforms (supplemental Fig. S1), were constructed using the technique of splicing by overlap extension with PCR (10Horton R.M. Hunt H.D. Ho S.N. Pullen J.K. Pease L.R. Gene (Amst.). 1989; 77: 61-68Crossref PubMed Scopus (2614) Google Scholar) and primers, P1, P3–5, P8, and P10–18 (supplemental Table S1). The combination of templates and primers used is summarized in supplemental Table S2. Mixtures of two DNA fragments separately produced in the first PCRs were used as templates for the second PCR (supplemental Table S2). The products of the second PCR were digested with NdeI and EcoRI and were subcloned into the His6 tag-removed pRSET-A to generate pRSET-(1–223)m2-(224–604)m1, pRSET-(1–224)m3-(224–604)m1, pRSET-(1–223)m1-(224–604)m2, pRSET-(1–224)m3-(224–604)m2, pRSET-(1–223)m1-(225–604)m3, and pRSET-(1–223)m2-(225–604)m3, respectively. Site-directed mutagenesis within the suppressor domain of T604m3 or (1–223)m1-(225–604)m3 (supplemental Fig. S2) to engineer systematic chimeric proteins as summarized in supplemental Fig. S2 was performed with a Quick Change site-directed mutagenesis kit (Stratagene) and primers containing the appropriate substitution (M1–M13, supplemental Table S3). Multiple mutants were generated by sequential mutagenesis. Only sense primers used for the site-directed mutagenesis are shown in supplemental Table S3. Substitution of amino acid residues 61–122 of T604m3 with amino acid residues 62–121 of T604m1 was carried out using the technique of splicing by overlap extension with PCR (supplemental Table S2) and primers P3, P19, P20, and P21 (supplemental Table S1). The product of the second PCR was digested with NdeI, and the NdeI fragment (0.4 kbp) was replaced with the NdeI fragment (0.4 kbp) of pRSET-T604m3. The nucleotide sequences of all of PCR products and site-directed mutants used in this study were confirmed by DNA sequencing with a 3130 Genetic analyzer (Applied Biosystems). Expression and Purification of Recombinant Proteins—Recombinant proteins were expressed in Escherichia coli BL21-codonplus (Stratagene) as described previously (9Yoshikawa F. Uchiyama T. Iwasaki H. Tomomori-Satoh C. Tanaka T. Furuichi T. Mikoshiba K. Biochem. Biophys. Res. Commun. 1999; 257: 792-797Crossref PubMed Scopus (42) Google Scholar). Protein purification was performed with a HiTrap heparin HP column (GE Healthcare) according to the method described previously (5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). The His6-tagged suppressor domain of mouse IP3R1 (amino acid residues 2–223) was purified with a ProBond resin (Invitrogen) as described previously (5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). Protein concentrations were determined with a protein assay kit (Bio-Rad) and bovine serum albumin as a standard. IP3 Binding Assay—An equilibrium IP3 binding analysis of purified soluble proteins was performed as described previously (9Yoshikawa F. Uchiyama T. Iwasaki H. Tomomori-Satoh C. Tanaka T. Furuichi T. Mikoshiba K. Biochem. Biophys. Res. Commun. 1999; 257: 792-797Crossref PubMed Scopus (42) Google Scholar), except for the reaction condition with a cytosol-like medium (110 mm KCl, 10 mm NaCl, 5 mm KH2PO4, and 50 mm Hepes-KOH, pH 7.4, at 4 °C). Purified protein (0.02–0.8 μg) was incubated with 0.14–8.68 nm [3H]IP3 (New England Nuclear/DuPont) and various concentrations of unlabeled IP3 (Dojindo) in a binding buffer (cytosol-like medium containing 1 mm dithiothreitol and 0.5 mm EGTA). To avoid tracer depletion, the amount of the protein and the concentration of [3H]IP3 were adjusted for each experiment. Nonspecific binding was measured in the binding buffer without adding the protein. Nonlinear regression of IP3 binding data with the Hill-Langmuir equation, F=[IP3]Kd+[IP3](Eq. 1) where F is the fraction of the recombinant protein that binds IP3, [IP3] is the concentration of IP3, and Kd is the apparent dissociation constant, was performed with Igor Pro (version 4.04, Wavematrics) software. Modeling of the Suppressor Domain in IP3R3—The structure of the IP3R1 suppressor domain (residues 7–223) (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar) was used for homology modeling of the suppressor domain structure of IP3R3, residues 6–224. The sequence alignment between IP3R1 and IP3R3 was generated with ClustalW (11Thompson J.D. Higgins D.G. Gibson T.J. Nucleic Acids Res. 1994; 22: 4673-4680Crossref PubMed Scopus (54899) Google Scholar) and used for modeling with Modeller (12Sali A. Blundell T.L. J. Mol. Biol. 1993; 234: 779-815Crossref PubMed Scopus (10292) Google Scholar). Because the loop linking the two helices in the Arm sub-domain of IP3R1 (residues 76–81) was not defined in the original structure, this portion of the suppressor domain was also modeled in the type-3 isoform. Thirty models were calculated and superimposed with Molmol (13Koradi R. Billeter M. Wuthrich K. J. Mol. Graph. 1996; 14: 51-55Crossref PubMed Scopus (6454) Google Scholar) (root mean square deviation = 1.2 Å). Only the mean model is shown. Characterization of IP3 Binding to the Amino-terminal 604 Residues of Three IP3R Isoforms—Nontagged amino-terminal 604 amino acid residues of the three isoforms of mouse IP3Rs, T604m1, T604m2, and T604m3 (supplemental Fig. S1), were expressed in E. coli and were purified on a HiTrap heparin HP column. Recombinant T604 proteins showed an apparent molecular mass of ∼65 kDa (Fig. 1A). We measured the IP3-binding activity of these purified proteins using 3H-labeled IP3. Fig. 1B shows the relationship between the normalized amount of IP3 bound to the purified proteins and the IP3 concentration applied. Because the data of the three isoforms were well fitted with the Hill-Langmuir equation (see “Experimental Procedures”) and each protein possesses a single IP3-binding site, IP3 binding to the purified amino-terminal 604 amino acid residues seemed to occur independently. The apparent dissociation constants of T604m1, T604m2, and T604m3 were estimated to be 49.5 ± 10.5 nm (n = 3), 14.0 ± 3.5 nm (n = 3), and 163.0 ± 44.4 nm (n = 9) (mean ± S.D.), respectively. These values are well consistent with the intrinsic dissociation constants estimated from the analyses of homotetrameric IP3R channels expressed in Sf9 cells under the same experimental condition (28.6 nm for IP3R1, 5.9 nm for IP3R2, and 294.0 nm for IP3R3) (5Iwai M. Tateishi Y. Hattori M. Mizutani A. Nakamura T. Futatsugi A. Inoue T. Furuichi T. Michikawa T. Mikoshiba K. J. Biol. Chem. 2005; 280: 10305-10317Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar) (Fig. 1C). These results suggest that the type-specific IP3-binding affinity originates from the amino-terminal 604 amino acid residues of each IP3R type. Comparison of the IP3-binding Affinity of the IP3-binding Domain of Three Isoforms without the Suppressor Domain—The IP3-binding domain of IP3R1 is functionally divided into two parts: the amino-terminal suppressor domain and the carboxyl-terminal IP3-binding core domain (6Yoshikawa F. Morita M. Monkawa T. Michikawa T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1996; 271: 18277-18284Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar, 14Yoshikawa F. Iwasaki H. Michikawa T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1999; 274: 328-334Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). The IP3-binding core domain has been defined as a minimum essential region for specific IP3 binding, which resides within amino acid residues 226–578 of IP3R1 (6Yoshikawa F. Morita M. Monkawa T. Michikawa T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1996; 271: 18277-18284Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). Because the amino acid residues 224–604 of mouse IP3R1, designated as (224–604)m1 (supplemental Fig. S1B), are expressed well in E. coli (9Yoshikawa F. Uchiyama T. Iwasaki H. Tomomori-Satoh C. Tanaka T. Furuichi T. Mikoshiba K. Biochem. Biophys. Res. Commun. 1999; 257: 792-797Crossref PubMed Scopus (42) Google Scholar), we used it for the measurement of the IP3-binding affinity of the IP3-binding domain without the suppressor domain. Supplemental Fig. S1A shows portions of the amino acid sequence alignments among the three mouse IP3R isoforms. To compare the unsuppressed IP3-binding affinity among the three isoforms, we purified the bacterially expressed (224–604)m1, residues 224–604 of mouse IP3R2 ((224–604)m2), and the residues 225–604 of mouse IP3R3 ((225–604)m3) (Fig. 2A). We found that all of the IP3 binding data of these proteins were well fitted with the Hill-Langmuir equation (Equation 1) (Fig. 2, B–D) with statistically indistinguishable apparent dissociation constants, 1.78 ± 0.63 nm (n = 3) for (224–604)m1, 1.51 ± 0.01 nm (n = 3) for (224–604)m2, and 2.04 ± 0.65 nm (n = 3) for (225–604)m3 (mean ± S.D.) (Fig. 2E). These results indicate that the presence of the amino-terminal 223, 223, and 224 amino acid residues of mouse IP3R1, IP3R2, and IP3R3 results in the suppression of IP3 binding by 27.8-fold (Fig. 2B), 9.3-fold (Fig. 2C), and 81.1-fold (Fig. 2D), respectively. Moreover, the isoform-specific IP3-binding affinity of the native IP3Rs reflects the different degree of the suppression of IP3 binding by the suppressor domain of each isoform. Mechanism of IP3 Binding Suppression—To assess the mechanism for the suppression, we investigated the effect of the addition of the bacterially expressed His6-tagged amino acid residues 2–223 of mouse IP3R1 (H(2–223)m1) (supplemental Fig. S1) on the IP3-binding activity of (224–604)m1. As shown in supplemental Fig. S3 the addition of 100-fold excess (molar ratio) purified H(2–223)m1 did not significantly alter the IP3-binding affinity of (224–604)m1. These results indicate that the suppression of IP3 binding requires a covalent bond between the suppressor domain and the IP3-binding core domain. Functional Difference among the Suppressor Domains of the Three IP3R Isoforms—We created six chimeric proteins composed of a suppressor domain fused with an IP3-binding core domain from different isoforms (supplemental Fig. S1B) to analyze the mechanism underlying the generation of isoform-specific IP3-binding affinity. All of the equilibrium IP3 binding data obtained for these chimeric proteins showed good fits with the Hill-Langmuir equation (Equation 1, data not shown) with apparent dissociation constants of 44.2 ± 13.4 nm (n = 3) for (1–223)m2-(224–604)m1, 73.0 ± 17.3 nm (n = 3) for (1–224)m3-(224–604)m1, 15.8 ± 3.6 nm (n = 3) for (1–223)m1-(224–604)m2, 40.7 ± 2.6 nm (n = 3) for (1–224)m3-(224–604)m2, 65.9 ± 7.4 nm (n = 9) for (1–223)m1-(225–604)m3, and 53.7 ± 5.2 nm (n = 3) for (1–223)m2-(225–604)m3 (mean ± S.D.). The results of the statistical analyses of these data are shown in Fig. 3. All of the chimeric proteins showed >10-fold lower IP3-binding affinity (Fig. 3, A–C) compared with the proteins that do not possess the suppressor domain (Kd ≈ 2 nm) (Fig. 2). The artificial chimeric proteins, however, revealed apparent dissociation constants that significantly differed from those of the proteins composed of the native combination of the suppressor domain and the IP3-binding core domain (Fig. 3, A–C). This suggests that the suppressor domain alone is not a prime determinant of the affinity for IP3. When we compare the same data in respect of the IP3-binding core domain (Fig. 3, D–F), the different character of the IP3-binding core domain of the three isoforms became obvious. The IP3-binding core domain of the type-1 isoform exhibited an almost identical IP3-binding affinity despite the type of the suppressor domain connected (Fig. 3D). The type-2 IP3-binding core domain showed indistinguishable affinity for IP3 when it was fused with the type-1 and type-2 suppressor domains, but the chimeric protein with the type-3 suppressor domain showed an affinity ∼3-fold lower compared with the native combination of the type-2 isoform (Fig. 3E). The IP3-binding core domain of the type-3 isoform strictly required the type-3 suppressor domain to generate its native IP3-binding affinity (Fig. 3F). These results indicate that: 1) the isoform-specific IP3-binding affinity is predominantly determined by the IP3-binding core domain rather than the suppressor domain; 2) the suppressor domains of the type-1 isoform and type-2 isoform are mutually interchangeable for both the IP3-binding core domains of IP3R1 and IP3R2 to generate their native IP3-binding affinity; and 3) the type-3-specific site(s) in the suppressor domain may be critical for the proper suppression of the IP3-binding core domain of IP3R3. Critical Loop Regions in the Suppressor Domain for the Suppression of IP3R3—To elucidate the structural basis for the suppression of the type-3 isoform, we created a series of mutated proteins based on the amino acid sequence difference between IP3R1 and IP3R3 and on the three-dimensional structure of the suppressor domain of IP3R1 (supplemental Fig. S2). The suppressor domain forms a hammer-like structure with the head subdomain and the arm subdomain (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar). The head subdomain consists of two single-turn α-helices and 12 β-strands, which form a β-trefoil fold (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar). The arm subdomain is made of a helix-turn-helix structure, which is inserted between the fourth and fifth β-strands of the head subdomain (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar). We hypothesized that the critical regions for the suppression of the type-3 isoform are located within the 11 loop segments that connect β-strands (Figs. 4A and S2), because in the case of β-trefoil fold architecture most of the surface properties are dictated by the residues from the loop segment, rather than the residues comprising the barrel and the triangular array of the core structure (15Murzin A.G. Lesk A.M. Chothia C. J. Mol. Biol. 1992; 223: 531-543Crossref PubMed Scopus (301) Google Scholar). To evaluate this hypothesis, we first constructed a mutant protein in which all of the non-conserved amino acid residues within the β-strand regions of T604m3 were substituted with residues appearing in IP3R1 (supplemental Fig. S2). Non-conserved residues are located within the first, fourth, eighth, ninth, and twelfth β-strand regions (β1, β4, β8, β9, and β12, respectively, supplemental Fig. S2) (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar). The mutant protein, (β1,β4,β8,β9,β12)m1/T604m3, exhibited an apparent dissociation constant of 254.0 ± 48.5 nm (n = 3), which was indistinguishable from that of T604m3 (163.0 ± 44.4 nm)(n = 9) (Fig. 4B). These results indicate that the critical sites for type-3-specific suppression are located within the loop regions, but not within the β-strand regions, of the suppressor domain. Type-3-specific amino acid sequences are located within the first, third, fourth, fifth, seventh, eighth, and tenth loop regions (L1, L3, L4, L5, L7, L8, and L10, respectively, supplemental Fig. S2) (7Bosanac I. Yamazaki H. Matsu-ura T. Michikawa T. Mikoshiba K. Ikura M. Mol. Cell. 2005; 17: 193-203Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar). To identify critical sites within the suppressor domain for the suppression of IP3 binding to the type-3 IP3-binding core domain, we created seven mutant proteins in which the amino acid sequences of individual type-3-specific loop regions, L1, L3, L4, L5, L7, L8, and L10, were substituted with residues appearing in IP3R1 (supplemental Fig. S2). We found that both L5m1/T604m3 (97.4 ± 26.2 nm, n = 7) and L7m1/T604m3 (105.0 ± 7.8 nm, n = 6) exhibited an apparent dissociation constant, which was significantly different from that of T604m3 (Fig. 5A). These results showed that both L5 and L7 are required for the suppression of IP3 binding to the IP3-binding core domain of IP3R3. We then constructed a mutant protein in which type-3-specific amino acid residues within both L5 and L7 in T604m3 were simultaneously changed to the residues that appear in IP3R1 (supplemental Fig. S2). This mutant protein, (L5,L7)m1/T604m3, exhibited an apparent dissociation constant (80.6 ± 9.9 nm, n = 6) that was lower than those of L5m1/T604m3 and L7m1/T604m3 (Fig. 5B), but the effect of the simultaneous replacement was not statistically significant when compared with the effects of the single loop replacements. The apparent dissociation constant of (L5,L7)m1/T604m3 is significantly different from that of (1–223)m1-(225–604)m3 (Fig. 5B). We then created five mutant proteins, (L1,L5,L7)m1/T604m3, (L3,L5,L7)m1/T604m3, (L4,L5,L7)m1/T604m3, (L5,L7,L8)m1/T604m3, and (L5,L7,L10)m1/T604m3, in which the type-3-specific amino acid residues in three loop regions, including L5 and L7, were substituted with residues appearing in IP3R1 (supplemental Fig. S2). Among these mutants, (L3,L5,L7)m1/T604m3 and (L5,L7,L8)m1/T604m3 exhibited an apparent dissociation constant (69.2 ± 10.6 nm (n = 5) and 77.4 ± 12.7 nm (n = 5), respectively) that was not different significantly from that of (1–223)m1-(225–604)m3 (p = 0.525 and 0.066, respectively, Fig. 5C). To confirm these results, we created mutant proteins in which all of the type-1-specific amino acid residues within L3, L5, and L7 in (1–223)m1-(225–604)m3 or within L5, L7, and L8 in (1–223)m1-(225–604)m3 were substituted with residues appearing in IP3R3
Referência(s)