Calmodulin Mediates Ca2+ Sensitivity of Sodium Channels
2004; Elsevier BV; Volume: 279; Issue: 43 Linguagem: Inglês
10.1074/jbc.m407286200
ISSN1083-351X
AutoresJames Kim, Smita Ghosh, Huajun Liu, Michihiro Tateyama, Robert S. Kass, Geoffrey S. Pitt,
Tópico(s)GABA and Rice Research
ResumoCa2+ has been proposed to regulate Na+ channels through the action of calmodulin (CaM) bound to an IQ motif or through direct binding to a paired EF hand motif in the Nav1 C terminus. Mutations within these sites cause cardiac arrhythmias or autism, but details about how Ca2+ confers sensitivity are poorly understood. Studies on the homologous Cav1.2 channel revealed non-canonical CaM interactions, providing a framework for exploring Na+ channels. In contrast to previous reports, we found that Ca2+ does not bind directly to Na+ channel C termini. Rather, Ca2+ sensitivity appears to be mediated by CaM bound to the C termini in a manner that differs significantly from CaM regulation of Cav1.2. In Nav1.2 or Nav1.5, CaM bound to a localized region containing the IQ motif and did not support the large Ca2+-dependent conformational change seen in the Cav1.2·CaM complex. Furthermore, CaM binding to Nav1 C termini lowered Ca2+ binding affinity and cooperativity among the CaM-binding sites compared with CaM alone. Nonetheless, we found suggestive evidence for Ca2+/CaM-dependent effects upon Nav1 channels. The R1902C autism mutation conferred a Ca2+-dependent conformational change in Nav1.2 C terminus·CaM complex that was absent in the wild-type complex. In Nav1.5, CaM modulates the Cterminal interaction with the III–IV linker, which has been suggested as necessary to stabilize the inactivation gate, to minimize sustained channel activity during depolarization, and to prevent cardiac arrhythmias that lead to sudden death. Together, these data offer new biochemical evidence for Ca2+/CaM modulation of Na+ channel function. Ca2+ has been proposed to regulate Na+ channels through the action of calmodulin (CaM) bound to an IQ motif or through direct binding to a paired EF hand motif in the Nav1 C terminus. Mutations within these sites cause cardiac arrhythmias or autism, but details about how Ca2+ confers sensitivity are poorly understood. Studies on the homologous Cav1.2 channel revealed non-canonical CaM interactions, providing a framework for exploring Na+ channels. In contrast to previous reports, we found that Ca2+ does not bind directly to Na+ channel C termini. Rather, Ca2+ sensitivity appears to be mediated by CaM bound to the C termini in a manner that differs significantly from CaM regulation of Cav1.2. In Nav1.2 or Nav1.5, CaM bound to a localized region containing the IQ motif and did not support the large Ca2+-dependent conformational change seen in the Cav1.2·CaM complex. Furthermore, CaM binding to Nav1 C termini lowered Ca2+ binding affinity and cooperativity among the CaM-binding sites compared with CaM alone. Nonetheless, we found suggestive evidence for Ca2+/CaM-dependent effects upon Nav1 channels. The R1902C autism mutation conferred a Ca2+-dependent conformational change in Nav1.2 C terminus·CaM complex that was absent in the wild-type complex. In Nav1.5, CaM modulates the Cterminal interaction with the III–IV linker, which has been suggested as necessary to stabilize the inactivation gate, to minimize sustained channel activity during depolarization, and to prevent cardiac arrhythmias that lead to sudden death. Together, these data offer new biochemical evidence for Ca2+/CaM modulation of Na+ channel function. Changes in cytoplasmic Ca2+ fluxes are the fundamental readouts of cellular electrical signals. It has become increasingly recognized that the ion channels controlling this electrical activity are subject to feedback modulation by these very Ca2+ fluxes. Among the best characterized responses are Ca2+ activation of small conductance K+ (SK) 1The abbreviations used are: SK, small conductance K+; CaM, calmodulin; FPLC, fast protein liquid chromatography; MOPS, 4-morpholinepropanesulfonic acid; GST, glutathione S-transferase; CT, C terminus; CTs, C termini; WT, wild-type; LQTS3, long QT syndrome subtype 3; BrS, Brugada syndrome; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl. channels to generate slow after-hyperpolarizations (1Bond C.T. Maylie J. Adelman J.P. Ann. N. Y. Acad. Sci. 1999; 868: 370-378Crossref PubMed Scopus (139) Google Scholar) and Ca2+-dependent inactivation of L-type Ca2+ channels (Cav1.2) to limit Ca2+ entry during action potentials (2Budde T. Meuth S. Pape H.C. Nat. Rev. Neurosci. 2002; 3: 873-883Crossref PubMed Scopus (169) Google Scholar). For both cases, the Ca2+-binding protein calmodulin (CaM) is an obligate channel subunit that serves as the Ca2+ sensor (3Xia X.M. Fakler B. Rivard A. Wayman G. Johnson-Pais T. Keen J.E. Ishii T. Hirschberg B. Bond C.T. Lutsenko S. Maylie J. Adelman J.P. Nature. 1998; 395: 503-507Crossref PubMed Scopus (740) Google Scholar, 4Peterson B.Z. DeMaria C.D. Adelman J.P. Yue D.T. Neuron. 1999; 22: 549-558Abstract Full Text Full Text PDF PubMed Scopus (722) Google Scholar, 5Zühlke R.D. Pitt G.S. Deisseroth K. Tsien R.W. Reuter H. Nature. 1999; 399: 159-162Crossref PubMed Scopus (744) Google Scholar). Identification of CaM in these roles has helped foster a new understanding of this ubiquitous Ca2+-binding protein. Previously, the model for CaM action had been enzymatic activation through disinhibition; binding of Ca2+/CaM to an autoinhibitory peptide, usually an amphipathic α-helix, exposes the active site of an enzyme (6Hoeflich K.P. Ikura M. Cell. 2002; 108: 739-742Abstract Full Text Full Text PDF PubMed Scopus (602) Google Scholar). In contrast, biochemical, functional, and structural analyses of CaM regulation of SK and Cav1.2 channels revealed that these channels contain large CaM-binding pockets in their cytoplasmic C termini composed of multiple noncontiguous sequences (3Xia X.M. Fakler B. Rivard A. Wayman G. Johnson-Pais T. Keen J.E. Ishii T. Hirschberg B. Bond C.T. Lutsenko S. Maylie J. Adelman J.P. Nature. 1998; 395: 503-507Crossref PubMed Scopus (740) Google Scholar, 7Pitt G.S. Zuhlke R.D. Hudmon A. Schulman H. Reuter H. Tsien R.W. J. Biol. Chem. 2001; 276: 30794-30802Abstract Full Text Full Text PDF PubMed Scopus (250) Google Scholar, 8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar, 9Schumacher M.A. Rivard A.F. Bachinger H.P. Adelman J.P. Nature. 2001; 410: 1120-1124Crossref PubMed Scopus (501) Google Scholar, 10Keen J.E. Khawaled R. Farrens D.L. Neelands T. Rivard A. Bond C.T. Janowsky A. Fakler B. Adelman J.P. Maylie J. J. Neurosci. 1999; 19: 8830-8838Crossref PubMed Google Scholar). Furthermore, although the details of CaM binding and Ca2+/CaM action differ between the SK and Cav1.2 channels, both constitutively bind apoCaM, which sits poised as a resident Ca2+ sensor. The recent structural determination of Ca2+/CaM in complex with the edema factor of Bacillus anthracis also revealed a large non-canonical CaM-binding site (11Drum C.L. Yan S.Z. Bard J. Shen Y.Q. Lu D. Soelaiman S. Grabarek Z. Bohm A. Tang W.J. Nature. 2002; 415: 396-402Crossref PubMed Scopus (357) Google Scholar) and demonstrated a new mode of CaM action by promoting active-site remodeling (6Hoeflich K.P. Ikura M. Cell. 2002; 108: 739-742Abstract Full Text Full Text PDF PubMed Scopus (602) Google Scholar). These and other recent studies have greatly expanded the repertoire of CaM interaction motifs and modes of CaM function. Recognition of an "IQ" motif within the Cav1.2 channel CaM-binding pocket (5Zühlke R.D. Pitt G.S. Deisseroth K. Tsien R.W. Reuter H. Nature. 1999; 399: 159-162Crossref PubMed Scopus (744) Google Scholar, 12Zühlke R.D. Reuter H. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 3287-3294Crossref PubMed Scopus (162) Google Scholar) fostered the identification of similar motifs in homologous regions of other channels and subsequent attempts to identify whether these channels are likewise Ca2+/CaM-regulated. First described as the binding site in myosins for CaM-like essential light chains (13Cheney R.E. Mooseker M.S. Curr. Opin. Cell Biol. 1992; 4: 27-35Crossref PubMed Scopus (336) Google Scholar), IQ motifs were subsequently found in apoCaM-binding proteins such as neuromodulin (14Alexander K.A. Wakim B.T. Doyle G.S. Walsh K.A. Storm D.R. J. Biol. Chem. 1988; 263: 7544-7549Abstract Full Text PDF PubMed Google Scholar). Many proteins containing CaM-binding IQ motifs have now been identified (15Rhoads A.R. Friedberg F. FASEB J. 1997; 11: 331-340Crossref PubMed Scopus (749) Google Scholar), providing a loose consensus sequence of IQXXXRXXXXR for identifying additional CaM-binding proteins. The voltage-gated Na+ channels form one family of ion channels in which an IQ motif has been identified. After identifying CaM as binding partner for the Nav1.2 C terminus in a yeast two-hybrid screen, Mori et al. (16Mori M. Konno T. Ozawa T. Murata M. Imoto K. Nagayama K. Biochemistry. 2000; 39: 1316-1323Crossref PubMed Scopus (105) Google Scholar) recognized an IQ motif in their bait construct at a position homologous to the IQ motif in Cav1.2. This led to a series of studies looking for Ca2+/CaM regulation of Na+ channel function (17Tan H.L. Kupershmidt S. Zhang R. Stepanovic S. Roden D.M. Wilde A.A. Anderson M.E. Balser J.R. Nature. 2002; 415: 442-447Crossref PubMed Scopus (196) Google Scholar, 18Deschenes I. Neyroud N. DiSilvestre D. Marban E. Yue D.T. Tomaselli G.F. Circ. Res. 2002; 90: E49-E57Crossref PubMed Google Scholar, 19Herzog R.I. Liu C. Waxman S.G. Cummins T.R. J. Neurosci. 2003; 23: 8261-8270Crossref PubMed Google Scholar). Because Na+ channels initiate the rapid upstroke of the cardiac and neuronal action potentials, Ca2+/CaM regulation of Na+ channel activity offers an intriguing way for the fine-tuning of membrane excitability. The clinical implications of such regulation are underscored by findings that mutations within or near the IQ motif are pathogenic. Mutations in the major cardiac sodium channel (Nav1.5) are arrhythmogenic, placing patients at risk for sudden cardiac death (20Makita N. Horie M. Nakamura T. Ai T. Sasaki K. Yokoi H. Sakurai M. Sakuma I. Otani H. Sawa H. Kitabatake A. Circulation. 2002; 106: 1269-1274Crossref PubMed Scopus (155) Google Scholar, 21Rook M.B. Bezzina Alshinawi C. Groenewegen W.A. van Gelder I.C. van Ginneken A.C. Jongsma H.J. Mannens M.M. Wilde A.A. Cardiovasc. Res. 1999; 44: 507-517Crossref PubMed Scopus (171) Google Scholar), and a mutation in a neuronal sodium channel (Nav1.2) is associated with a familial form of autism (22Weiss L.A. Escayg A. Kearney J.A. Trudeau M. MacDonald B.T. Mori M. Reichert J. Buxbaum J.D. Meisler M.H. Mol. Psychiatry. 2003; 8: 186-194Crossref PubMed Scopus (241) Google Scholar). Because of the homology between Na+ and Ca2+ channels with respect to their IQ motifs and the location of these motifs within the channels, the search for a defined role for Ca2+/CaM in Na+ channel function followed the model originally developed for Cav1.2, in which interactions of Ca2+/CaM with the IQ motif in Cav1.2 accelerate channel inactivation (2Budde T. Meuth S. Pape H.C. Nat. Rev. Neurosci. 2002; 3: 873-883Crossref PubMed Scopus (169) Google Scholar). Studies with similar approaches on three different Na+ channels yielded inconsistent results, however (17Tan H.L. Kupershmidt S. Zhang R. Stepanovic S. Roden D.M. Wilde A.A. Anderson M.E. Balser J.R. Nature. 2002; 415: 442-447Crossref PubMed Scopus (196) Google Scholar, 18Deschenes I. Neyroud N. DiSilvestre D. Marban E. Yue D.T. Tomaselli G.F. Circ. Res. 2002; 90: E49-E57Crossref PubMed Google Scholar, 19Herzog R.I. Liu C. Waxman S.G. Cummins T.R. J. Neurosci. 2003; 23: 8261-8270Crossref PubMed Google Scholar). Although it has been proposed that Ca2+/CaM may regulate Na+ channels in an isoform-specific manner (19Herzog R.I. Liu C. Waxman S.G. Cummins T.R. J. Neurosci. 2003; 23: 8261-8270Crossref PubMed Google Scholar), this cannot explain the conflicting results with common isoforms among these studies. Direct binding of Ca2+ to a paired EF hand motif in Nav1.5 channels has recently been proposed as an alternative mode of Ca2+ regulation (23Wingo T.L. Shah V.N. Anderson M.E. Lybrand T.P. Chazin W.J. Balser J.R. Nat. Struct. Mol. Biol. 2004; 11: 219-225Crossref PubMed Scopus (118) Google Scholar). Structural modeling of the proximal portion of the Nav1.5 C terminus had predicted overall homology to the N-terminal lobe of CaM, but unlikely binding capacity for Ca2+, since many key acidic residues used for specific Ca2+ binding in CaM are not conserved in the Nav1.5 C terminus (24Cormier J.W. Rivolta I. Tateyama M. Yang A.-S. Kass R.S. J. Biol. Chem. 2002; 277: 9233-9241Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar). Nevertheless, the new model presented by Wingo et al. (23Wingo T.L. Shah V.N. Anderson M.E. Lybrand T.P. Chazin W.J. Balser J.R. Nat. Struct. Mol. Biol. 2004; 11: 219-225Crossref PubMed Scopus (118) Google Scholar) and their functional analysis of a pro-arrhythmic mutation within this region, if correct, could offer an explanation for the differences reported among the studies focused on CaM. New information on CaM interaction with and regulation of Ca2+ channels (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar, 25Erickson M.G. Liang H. Mori M.X. Yue D.T. Neuron. 2003; 39: 97-107Abstract Full Text Full Text PDF PubMed Scopus (199) Google Scholar) has offered a mirror with which to reexamine CaM interaction with Na+ channels. The specific rationale for this approach derives from the appreciation that CaM interaction domains cannot be predicted from the primary amino acid sequence, as demonstrated by the crystal structures of CaM in complex with SK channels and B. anthracis edema factor (9Schumacher M.A. Rivard A.F. Bachinger H.P. Adelman J.P. Nature. 2001; 410: 1120-1124Crossref PubMed Scopus (501) Google Scholar, 10Keen J.E. Khawaled R. Farrens D.L. Neelands T. Rivard A. Bond C.T. Janowsky A. Fakler B. Adelman J.P. Maylie J. J. Neurosci. 1999; 19: 8830-8838Crossref PubMed Google Scholar) and highlighted by the recent work on CaM regulation of Cav1.2 (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar). If sequences outside the IQ motif are essential for CaM binding and regulation of Nav1 channels, as they are with Cav1.2, then a more systematic approach than the original focus on the IQ motifs might yield more definitive results. Another motivation for this approach came from the specific biochemical tools developed for understanding Ca2+/CaM regulation of Cav1.2 (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar), which offered new opportunities for exploring Ca2+/CaM interaction with Nav1 channels. With this framework, we explored the biochemistry of CaM interaction with the C termini of both Nav1.2 and Nav1.5. These are two potentially informative isoforms because both have pathogenic mutations in or near the IQ motif that offer the possibility to further define a role for CaM regulation of channel function. Computational Analysis—Sequence alignment of the C terminus of Cav1.2 (GenBank™/EBI Data Bank accession number X15539), Nav1.2 (accession number M94055), and Nav1.5 (accession number M77235) was performed using Proteinsolve (Stratagene) based on a modified version of the Boyko weight matrix (26Wishart D.S. Boyko R.F. Willard L. Richards F.M. Sykes B.D. Comput. Appl. Biosci. 1994; 10: 121-132PubMed Google Scholar). The secondary structure prediction of the Cav1.2 C terminus was determined using Proteinsolve based on the consensus prediction from multiple methods (homology, hydrophobic moment, GOR (40Garnier J. Osguthorpe D.J. Robson B. J. Mol. Biol. 1978; 120: 97-120Crossref PubMed Scopus (3423) Google Scholar), Chou-Fasman (41Chou P.Y. Fasman G.D. Biochemistry. 1974; 13: 211-222Crossref PubMed Scopus (1833) Google Scholar, 42Chou P.Y. Fasman G.D. Biochemistry. 1974; 13: 222-245Crossref PubMed Scopus (2547) Google Scholar), and motifbased), and prediction of Nav1.2 and Nav1.5 C termini was based on previous work by Cormier et al. (24Cormier J.W. Rivolta I. Tateyama M. Yang A.-S. Kass R.S. J. Biol. Chem. 2002; 277: 9233-9241Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar). Construction of cDNA Plasmids—A cDNA encoding the C terminus of Nav1.2 was purchased from Open Biosystems after identifying an appropriate expressed sequence tag. DNA sequences corresponding to amino acids 1777–1937 of Nav1.2 and amino acids 1773–1940 of Nav1.5 were amplified by PCR with primers containing endonuclease restriction sites. Products were digested with the appropriate enzymes and ligated into pET28a+ (Novagen) to produce cDNAs encoding His6-tagged proteins. The CaM expression plasmid has been described (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar). Mutations were generated using QuikChange (Stratagene). Protein Expression and Purification—BL21 cells were transformed by electroporation with the appropriate plasmid(s). A 10-ml starter culture was grown with the appropriate antibiotic(s) for 2 h at 37 °Cand then transferred to 1-liter flasks, where the culture was grown until A600 = 0.3. The culture flasks were cooled in an ice-water bath, and protein expression was induced with 1.0 mm isopropyl β-d-thiogalactopyranoside for 72 h at 16 °C. Cells were harvested and resuspended in 500 mm NaCl, 20 mm Tris, 5 mm imidazole, and 25% glycerol (pH 7.5) supplemented with EDTA-free Complete protease inhibitor mixture (Roche Applied Science), and bacterial cell lysates were prepared by passage through a French pressure cell. The lysates were centrifuged at 100,000 × g for 90 min, and the supernatants were then applied to Talon metal affinity resin (Clontech). The proteins were eluted with 250 mm imidazole, aliquoted, and stored at -20 °C in 25% glycerol for further use. Gel Filtration Analysis—Gel filtration was performed over a Superdex 200 HR 10/30 column on an AKTA FPLC (Amersham Biosciences) in 500 mm NaCl and 20 mm Tris (pH 7.5) supplemented with CaCl2 (10 μm) or EGTA (5 mm). The following protein standards (Amersham Biosciences) were used to calibrate the elutions: aldolase (158 kDa), albumin (67 kDa), ovalbumin (43 kDa), and chymotrypsinogen (26 kDa). Protein Concentration Determination—The concentration of CaM was determined by UV absorbance at 277 nm using a molar extinction coefficient of 3300 m-1 cm-1. The concentration of CaM in complex with fusion proteins of Nav1 C termini was determined based on a standard of known CaM concentrations, as described above, on a Coomassie Blue-stained gel. Fluorescence Spectroscopy—Before obtaining fluorescence spectra, all solutions and proteins were treated with Chelex (Bio-Rad) to remove any Ca2+ present. Fluorescence spectra were obtained on a Photon Technology International QuantaMaster spectrofluorometer in a 2-ml quartz cuvette (Hellma) in buffer containing 20 mm Tris (pH 7.4) and 100 mm NaCl. Intrinsic tyrosine fluorescence spectra were obtained at λex = 280 nm and monitored for fluorescent emission between 290 and 390 nm. Spectra for 5 μm CaM alone and in complex with the fusion proteins of Nav1 C termini were obtained upon sequential addition of 100 μm EGTA, 140 μm CaCl2, and 400 μm EGTA. Ca2+ titration experiments were carried out with Ca2+ buffers containing 20 mm MOPS (pH 7.2), 100 mm KCl, 1 mm MgCl2, and various concentrations of CaCl2 and EGTA. Free Ca2+ concentrations were precalibrated in buffers obtained from Molecular Probes, Inc. or calculated using WEBMAXC STANDARD (www.stanford.edu/~cpatton/maxc.html) to obtain buffers with a range of free Ca2+ concentrations not well represented by the available buffer set. Glutathione S-Transferase (GST) Pull-down Assays—The GST-Nav1.5 III–IV linker fusion protein and the GST control were bound to glutathione-Sepharose 4B (Amersham Biosciences). The Nav1 C terminus·CaM fusion protein complex was incubated with bound GST or GST-Nav1.5 III–IV linker for 60 min at 25 °C in 150 mm NaCl, 50 mm Tris, and 0.1% Triton X-100 (pH 7.4) supplemented with 1 mm Ca2+ or 1 mm EGTA. The bound complexes were then washed extensively with the appropriate buffer, eluted in SDS sample buffer, separated by SDS-PAGE, and visualized by Coomassie Blue staining. Expression of Recombinant Na+ Channels and Electrophysiology—Na+ channels were expressed in human embryonic kidney 293 cells at 22 °C as described previously (27Abriel H. Cabo C. Wehrens X.H. Rivolta I. Motoike H.K. Memmi M. Napolitano C. Priori S.G. Kass R.S. Circ. Res. 2001; 88: 740-745Crossref PubMed Scopus (108) Google Scholar). CD8-positive cells were identified using Dynabeads (M-450, Dynal, Inc.) and were patch-clamped 48 h after transfection. Membrane currents were measured using whole cell patch-clamp procedures with Axopatch 200B amplifiers (Axon Instruments, Inc., Foster City, CA). Protocols and solutions for measurement of Na+ channel current tetrodotoxin-sensitive Na+ channel currents and sustained caret (Isus) are described in detail in a previous publication (28Clancy C.E. Tateyama M. Liu H. Wehrens X.H. Kass R.S. Circulation. 2003; 107: 2233-2237Crossref PubMed Scopus (125) Google Scholar). As a first step in using insights from the recent work on Cav1.2·CaM (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar) to analyze CaM binding to Nav1 channels, we aligned the primary sequences of Nav1.2 and Nav1.5 C termini with the Cav1.2 C terminus and looked for similarities within the regions found important for CaM interaction with the Cav1.2 C terminus (Fig. 1A). We especially focused on homology outside the IQ motif, as identification of additional regions that contribute to CaM interaction in the Nav1 channels could provide important insight into CaM interaction and function. The Nav1.2 and Nav1.5 sequences are very similar to each other (76% identical) in this proximal portion of the C terminus. Compared with the Cav1.2 sequence, they show significant similarity (18 and 19% identical, respectively) that extends throughout the entire region. Secondary structure predictions of the Na+ and Ca2+ channel C termini also showed remarkable correlations (Fig. 1A). These results led us to hypothesize that CaM interaction with Na+ channels may also involve multiple noncontiguous sequences in addition to the IQ motifs. We next tested whether we could isolate a stable complex between CaM and the C terminus of Nav1.2 or Nav1.5. Generation of a complex between CaM and the C terminus of Cav1.2 was a particularly helpful tool in identifying the Cav1.2 CaM-binding pocket. We first tested whether coexpression of CaM is necessary to maintain solubility of the Nav1.2 and Nav1.5 C termini in a bacterial expression system since this requirement formed a linchpin in our model of CaM interaction with Cav1.2 (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar). When we expressed amino acids 1777–1937 of Nav1.2 or amino acids 1773–1940 of Nav1.5 (referred to as Nav1.2 CT or Nav1.5 CT, respectively, to designate the His6-tagged fusion proteins), corresponding to the Cav1.2 construct previously tested (amino acids 1507–1669), little of the material in the cell extract was soluble in either case (Fig. 1B), similar to when the construct of the Cav1.2 C terminus was expressed alone. With CaM coexpression, however, the Nav1 C termini (CTs) remained soluble. CaM co-purified with the His6-tagged Nav1.2 or Nav1.5 CT on a metal affinity column with an apparent 1:1 stoichiometry (Fig. 1B). Each complex migrated on a gel filtration column as a single species, heterodimers with an apparent molecular mass of ∼40 kDa (Fig. 1C). The ability of CaM to bind Ca2+ was not necessary for generation of these soluble complexes since we also obtained abundant soluble material when either of the Nav1 CTs was coexpressed with CaM1234, a mutant in which Ca2+ binding to all four EF hands is lost (data not shown); this complex migrated on the gel filtration column identically to one with wild-type (WT) CaM in either Ca2+ or EGTA. These results show that the Nav1.2 and Nav1.5 CTs are able to form stable complexes with Ca2+/CaM and apoCaM, suggesting that CaM interaction with these Na+ channels, as with Cav1.2, is constitutive and Ca2+-independent. We next tested whether the Nav1.2 CT·CaM or Nav1.5 CT·CaM complex displayed an altered mobility on a gel filtration column. Since the Ca2+-induced mobility shift of the Cav1.2·CaM complex was interpreted as the conformational change that underlies Ca2+-dependent gating and provided an important part of the model for CaM regulation of Cav1.2 (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar), a similar Ca2+-dependent conformational change in Nav1 CT·CaM complexes would be informative. For the Nav1 CT·CaM complexes, we did not detect any difference in mobility in either EGTA- or 10 μm Ca2+-containing buffer (Fig. 1C). This result raised the possibility that the Nav1 CT·CaM complexes are Ca2+-insensitive. Alternatively, Ca2+ interaction with the Nav1 CT·CaM complexes might induce a change not detectable by gel filtration chromatography. We therefore looked for Ca2+ enhancement of intrinsic Tyr fluorescence as a more sensitive test of Ca2+ binding to CaM. This well characterized property reports changes in the hydrophobicity of the environment near Tyr99 and Tyr138 in the CaM C-terminal lobe that occur as a result of a local conformational change induced by Ca2+ binding (29Wallace R.W. Tallant E.A. Dockter M.E. Cheung W.Y. J. Biol. Chem. 1982; 257: 1845-1854Abstract Full Text PDF PubMed Google Scholar), as demonstrated in Fig. 2C. For the Ca2+-insensitive CaM1234 mutant, there was no Ca2+ enhancement of Tyr fluorescence (Fig. 2F). By this measure, Ca2+ clearly bound to the C-terminal lobe of CaM when in complex with either the Nav1.2 or Nav1.5 CT (Fig. 2, A and B). Ca2+ enhanced the intrinsic Tyr fluorescence (black circles; λem = 308 nm) of both the Nav1.2 CT·CaM and Nav1.5 CT·CaM complexes, which was completely reversed by the addition of EGTA. Sequential addition of Ca2+ and EGTA to the same sample ensured that the increase in intensity was a property of the protein and not a result of changes in protein concentration due to pipetting errors. To rule out that the Tyr fluorescence enhancement was confounded by Ca2+-induced changes in the fluorescent properties of the five additional Tyr residues in Nav1.2 or the three in Nav1.5, we repeated these studies with Nav1 CT·CaM1234 complexes and found no Ca2+-enhanced Tyr fluorescence (Fig. 2, D–F). For CaM in complex with either the Nav1.2 or Nav1.5 CT, the apparent K0.5 for Ca2+ binding to the C-terminal lobes showed an ∼6-fold increase (K0.5 = 12.6 ± 2.8 μm for Nav1.2 CT·CaM, 12.6 ± 3.9 μm for Nav1.5 CT·CaM, and 1.9 ± 0.2 μm for CaM alone; n = 3) and a loss of the cooperativity between the Ca2+-binding sites (Fig. 2I). We also tested for Ca2+ binding to a putative EF hand motif within the Nav1 proteins (23Wingo T.L. Shah V.N. Anderson M.E. Lybrand T.P. Chazin W.J. Balser J.R. Nat. Struct. Mol. Biol. 2004; 11: 219-225Crossref PubMed Scopus (118) Google Scholar) by measuring the intrinsic fluorescence of a single conserved Trp residue (Trp1802 in Nav1.2 and Trp1798 in Nav1.5) located in the Nav1 EF hand motifs. Since CaM does not contain Trp, changes in the fluorescence could be attributed to Ca2+ binding to the Nav1 EF hand motifs. With selective Trp excitation at λex = 295 nm, we observed peak emission at 325 nm that was insensitive to Ca2+ (Fig. 2, G and H). Results with the Nav1.2 construct similarly showed no Ca2+-sensitive changes. In addition, we did not detect any Ca2+-dependent increase in intrinsic Trp fluorescence for either the Nav1.2 or Nav1.5 CT in complex with CaM1234 with excitation at λex = 280 nm (Fig. 2, D–F). Together with the Tyr fluorescence studies, these data demonstrate that Ca2+ binds to CaM in complex with the Nav1 CTs, but not to the Nav1 CTs themselves. The subsequent conformational change is markedly restricted compared with Cav1.2 (8Kim J. Ghosh S. Nunziato D.A. Pitt G.S. Neuron. 2004; 41: 745-754Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar), however, as indicated by the lack of an apparent shift in migration on the gel filtration column. These dissimilarities led us to examine more closely the contributions of sequences outside the IQ motifs to CaM interaction with the Nav1 CTs. Although the ability of CaM to promote solubilization of the Nav1 CTs (Fig. 1) supported our hypothesis that these proteins, like Cav1.2, contain a CaM-binding pocket consisting of multiple noncontiguous sequences, the lack of a conformational change on the gel filtration column suggested that the details of CaM interaction with the Nav1 CTs likely differed from those of CaM interaction with Cav1.2. Our specific approach was to examine whether mutations outside the Nav1 IQ motifs would disrupt CaM interaction, similar to what we observed for Cav1.2. Long QT syndrome subtype 3 (LQTS3) and Brugada syndrome (BrS) mutations in the C terminus were chosen as candidates for Nav1.5 because these mutations, which place patients at risk for arrhythmogenic sudden cardiac death (30Keating M.T. Sanguinetti M.C. Cell. 2001; 104: 569-580Abstract Full Text Full Text PDF PubMed Scopus (862) Google Scholar), are within the Nav1.5 CT that we tested and alter channel gating behavior; if the mutations also alter CaM interaction, this would provide an important correlation between CaM interaction and channel function and would therefore parallel the changes in channel gating resulting from disruptio
Referência(s)