Portal de Periódicos da CAPES

Fast Inactivation of Voltage-dependent Calcium Channels

2000; Elsevier BV; Volume: 275; Issue: 32 Linguagem: Inglês

10.1074/jbc.m000399200

1083-351X

Stephanie C. Stotz, Jawed Hamid, Renee Spaetgens, Scott E. Jarvis, Gerald W. Zamponi,

Nicotinic Acetylcholine Receptors Study

We recently described domains II and III as important determinants of fast, voltage-dependent inactivation of R-type calcium channels (Spaetgens, R. L., and Zamponi, G. W. (1999) J. Biol. Chem. 274, 22428–22438). Here we examine in greater detail the structural determinants of inactivation using a series of chimeras comprising various regions of wild type α1C and α1Ecalcium channels. Substitution of the II S6 and/or III S6 segments of α1E into the α1C backbone resulted in rapid inactivation rates that closely approximated those of wild type α1E channels. However, neither individual or combined substitution of the II S6 and III S6 segments could account for the 60 mV more negative half-inactivation potential seen with wild type α1E channels, indicating that the S6 regions contribute only partially to the voltage dependence of inactivation. Interestingly, the converse replacement of α1E S6 segments of domains II, III, or II+III with those of α1Cwas insufficient to significantly slow inactivation rates. Only when the I-II linker region and the domain II and III S6 regions of α1E were concomitantly replaced with α1Csequence could inactivation be abolished. Conversely, introduction of the α1E domain I-II linker sequence into α1C conferred α1E-like inactivation rates, indicating that the domain I-II linker is a key contributor to calcium channel inactivation. Overall, our data are consistent with a mechanism in which inactivation of voltage-dependent calcium channels may occur via docking of the I-II linker region to a site comprising, at least in part, the domain II and III S6 segments. We recently described domains II and III as important determinants of fast, voltage-dependent inactivation of R-type calcium channels (Spaetgens, R. L., and Zamponi, G. W. (1999) J. Biol. Chem. 274, 22428–22438). Here we examine in greater detail the structural determinants of inactivation using a series of chimeras comprising various regions of wild type α1C and α1Ecalcium channels. Substitution of the II S6 and/or III S6 segments of α1E into the α1C backbone resulted in rapid inactivation rates that closely approximated those of wild type α1E channels. However, neither individual or combined substitution of the II S6 and III S6 segments could account for the 60 mV more negative half-inactivation potential seen with wild type α1E channels, indicating that the S6 regions contribute only partially to the voltage dependence of inactivation. Interestingly, the converse replacement of α1E S6 segments of domains II, III, or II+III with those of α1Cwas insufficient to significantly slow inactivation rates. Only when the I-II linker region and the domain II and III S6 regions of α1E were concomitantly replaced with α1Csequence could inactivation be abolished. Conversely, introduction of the α1E domain I-II linker sequence into α1C conferred α1E-like inactivation rates, indicating that the domain I-II linker is a key contributor to calcium channel inactivation. Overall, our data are consistent with a mechanism in which inactivation of voltage-dependent calcium channels may occur via docking of the I-II linker region to a site comprising, at least in part, the domain II and III S6 segments. kilobase(s) base pair(s) cesium methanesulfonate tetraethylammonium Calcium entry through voltage-dependent calcium channels is important for a range of cellular processes, including neurotransmitter release and activation of Ca2+-dependent enzymes. Molecular cloning has identified the primary structures of at least 9 different neuronal Ca2+ channel α1 subunits (termed α1A through α1I (1Williams M.E. Feldman D.H. McCue A.F. Brenner R. Velicelebi G. Ellis S.B. Harpold M.M. Neuron. 1992; 8: 71-84Abstract Full Text PDF PubMed Scopus (439) Google Scholar, 2Tomlinson W.J. Stea A. Bourinet E. Charnet P. Nargeot J. Snutch T.P. Neuropharmacology. 1993; 32: 1117-1126Crossref PubMed Scopus (149) Google Scholar, 3Bech-Hansen N.T. Naylor M.J. Maybaum T.A. Pearce W.G. Koop B. Fishman G.A. Mets M. Musarella M.A. Boycott K.M. Nat. Genet. 1998; 19: 264-267Crossref PubMed Scopus (429) Google Scholar, 4Dubel S.J. Starr T.V. Hell J. Ahlijanian M.K. Enyeart J.J. Catterall W.A. Snutch T.P. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 5058-5062Crossref PubMed Scopus (282) Google Scholar, 5Williams M.E. Brust P.F. Feldman D.H. Patthi S. Simerson S. Maroufi A. McCue A.F. Velicelebi G. Ellis S.B. Harpold M.M. Science. 1992; 245: 389-395Crossref Scopus (416) Google Scholar, 6Fujita Y. Mynlieff M. Dirksen R.T. Kim M.S. Niidome T. Nakai J. Friedrich T. Iwabe N. Miyata T. Furuichi T. Furutama D. Mikoshiba K. Mori Y. Beam K.G. Neuron. 1993; 10: 465-598Abstract Full Text PDF PubMed Scopus (205) Google Scholar, 7Bourinet E. Soong T.W. Sutton K. Slaymaker S. Mathews E. Monteil A. Zamponi G.W. Nargeot J. Snutch T.P. Nat. Neurosci. 1999; 2: 407-415Crossref PubMed Scopus (364) Google Scholar, 8Mori Y. Friedrich T. Kim M.S. Mikami A. Nakai J. Ruth P. Bosse E. Hofmann F. Flockerzi V. Furuichi T. Mikoshiba K. Imoto K. Tanabe T. Numa S. Nature. 1991; 350: 398-402Crossref PubMed Scopus (706) Google Scholar, 9Sather W.A. Tanabe T. Zhang J.F. Mori Y. Adams M.E. Tsien R.W. Neuron. 1993; 11: 291-303Abstract Full Text PDF PubMed Scopus (382) Google Scholar, 10Stea A. Tomlinson W.J. Soong T.W. Bourinet E. Dubel S.J. Vincent S.R. Snutch T.P. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 10576-10580Crossref PubMed Scopus (307) Google Scholar, 11Cribbs L.L. Lee J.H. Yang J. Satin J. Zhang Y. Daud A. Barclay J. Williamson M.P. Fox M. Rees M. Perez-Reyes E. Circ. Res. 1998; 83: 103-109Crossref PubMed Scopus (518) Google Scholar, 12Perez-Reyes E. Cribbs L.L. Daud A. Lacerda A.E. Barclay J. Williamson M.P. Fox M. Rees M. Lee J.H. Nature. 1998; 391: 896-900Crossref PubMed Scopus (639) Google Scholar, 13Soong T.W. Stea A. Hodson C.D. Dubel S.J. Vincent S.R. Snutch T.P. Science. 1993; 260: 1133-1136Crossref PubMed Scopus (438) Google Scholar, 14Williams M.E. Marubio L.M. Deal C.R. Hans M. Brust P.F. Philipson L.H. Miller R.J. Johnson E.C. Harpold M.M. Ellis S.B. J. Biol. Chem. 1994; 269: 22347Abstract Full Text PDF PubMed Google Scholar, 15Tottene A. Moretti A. Pietrobon D. J. Neurosci. 1997; 16: 6342-6363Google Scholar)) that encode the previously identified native L-, P-, N, -Q-, T-, and R-types (for review, see Refs. 16McCleskey E.W. Curr. Opin. Neurobiol. 1994; 4: 304-312Crossref PubMed Scopus (107) Google Scholar and 17Stea A. Soong T.W. Snutch T.P. Neuron. 1995; 15: 929-940Abstract Full Text PDF PubMed Scopus (194) Google Scholar). Calcium channels, like many other voltage-dependent ion channels, undergo a series of conformational changes in response to voltage, resulting in their opening, closing, and inactivation. Voltage-dependent inactivation of calcium channels is an important intrinsic process that prevents the breakdown of the calcium gradient as well as excessive calcium entry that is toxic to most cells (18Choi D.W. Trends Neurosci. 1988; 11: 465-469Abstract Full Text PDF PubMed Scopus (1609) Google Scholar, 19Orrenius S. McConkey D.J. Bellomo G. Nicotera P. Trends Pharmacol. Sci. 1989; 10: 281-285Abstract Full Text PDF PubMed Scopus (790) Google Scholar, 20Orrenius S. Nicotera P. J. Neural Transm. Suppl. 1994; 43: 1-11PubMed Google Scholar). In addition, many pharmacological agents interact predominantly with inactivated channels (21Hawthorn M.H. Ferrante J.N. Kwon Y.W. Rutledge A. Luchowski E. Bangalore R. Triggle D.J. Eur. J. Pharmacol. 1992; 229: 143-148Crossref PubMed Scopus (15) Google Scholar, 22Hering S. Aczel S. Kraus R.L. Berjukow S. Striessnig J. Timin E.N. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 13323-13328Crossref PubMed Scopus (62) Google Scholar). Unlike sodium (23Vassilev P. Scheuer T. Catterall W.A. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 8147Crossref PubMed Scopus (172) Google Scholar, 24Eaholtz G. Scheuer T. Catterall W.A. Neuron. 1994; 12: 1041-1048Abstract Full Text PDF PubMed Scopus (136) Google Scholar) and potassium (25Zagotta W.N. Hoshi T. Aldrich R.W. Science. 1990; 250: 506-507Crossref PubMed Scopus (609) Google Scholar, 26Hoshi T. Zagotta W.N. Aldrich R.W. Neuron. 1991; 7: 436-454Abstract Full Text PDF Scopus (569) Google Scholar, 27Isacoff E.Y. Jan Y.N. Jan L.Y. Nature. 1991; 342: 86-90Crossref Scopus (276) Google Scholar) channels, the mechanisms that govern calcium channel voltage-dependent inactivation are not fully understood. Although a number of structural moieties of the calcium channel α1 subunit have been implicated in being important in fast calcium channel inactivation (7Bourinet E. Soong T.W. Sutton K. Slaymaker S. Mathews E. Monteil A. Zamponi G.W. Nargeot J. Snutch T.P. Nat. Neurosci. 1999; 2: 407-415Crossref PubMed Scopus (364) Google Scholar,22Hering S. Aczel S. Kraus R.L. Berjukow S. Striessnig J. Timin E.N. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 13323-13328Crossref PubMed Scopus (62) Google Scholar, 28Zhang J.F. Ellinor P.T. Aldrich R.W. Tsien R.W. Nature. 1994; 372: 97-100Crossref PubMed Scopus (177) Google Scholar, 29Zamponi G.W. Soong T.W. Bourinet E. Snutch T.P. J. Neurosci. 1996; 16: 2430-2443Crossref PubMed Google Scholar, 30Herlitze S. Hockerman G.H. Scheuer T. Catterall W.A. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 1402-1406Crossref PubMed Scopus (169) Google Scholar, 32Hering S. Aczel S. Grabner M. Doring F. Berjukow S. Mitterdorfer J. Sinnegger M.J. Striessnig J. Degtiar V.E. Wang Z. Glossmann H. J. Biol. Chem. 1996; 271: 24471-24475Abstract Full Text Full Text PDF PubMed Scopus (75) Google Scholar, 33Hering S. Berjukow S. Aczel S. Timin E.N. Trends Pharmacol. Sci. 1998; 19: 439-443Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar), the detailed mechanism underlying the inactivation process remain unknown, and there have been few systematic attempts to resolve this issue. By creating a series of chimeras between non-inactivating (L-type) α1C and rapidly inactivating (R-type) α1E rat brain calcium channels, we recently demonstrated that multiple structural domains determine the voltage dependence and rates of calcium channel inactivation (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Here, we present novel evidence implicating the domain II and III S6 segments and the domain I-II linker region as key elements in setting the rate of calcium channel inactivation. Using a number of additional chimeras derived from α1C and α1E channels, we demonstrate that insertion of either the domain II S6, III S6, or I-II linker regions of α1E into α1C is sufficient to confer α1E-like inactivation kinetics. Consistent with these data, removal of inactivation from α1E required the concomitant substitution of all three regions with α1C sequence. Based on this evidence, we propose a model in which the I-II linker forms a hinged lid that may dock at the domain II and III S6 regions of the channel. We previously introduced convenient silent restriction enzyme sites (obtained from Life Technologies, Inc. and from New England Biolabs) into the cDNAs encoding for wild-type rat brain α1E (rbE-II, GenBankTM accession number L15453) and α1C (rbC-II, GenBankTMaccession number M67515). AvrII was inserted at the beginning of the I-II linker, and SalI was inserted at the beginning of the II-III linker (see Ref. 34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar) to facilitate the creation of a series of chimeras encompassing various combinations of transmembrane domains of the two parent channels. To permit exchange of the domain II and III S6 segments, an additional pair of restriction sites was introduced into several of these chimeras at exactly complimentary positions; an AgeI restriction site was introduced 20 amino acids 5′ to the beginning of the II S6 segment, and an AatII site was created 5 amino acids 5′ to the beginning of the III S6 segment. The residues prior to the beginning of II S6 and III S6 are identical in both channels, and hence, the resultant chimeras only differ in their S6 regions. To permit exchange of the domain I-II linker region, a silent NarI site was introduced into the ECCC sequence at the beginning of domain II. The EECC and CCEE chimeras (in the pMT2 expression vector) were used as the template for mutagenesis to introduce unique AgeI sites near the beginning of the II S6 segments. Both constructs were first cut with SalI (II-III linker, 3′-polycloning site) and recircularized to reduce their length by about 5 kb1 before proceeding with mutagenesis. Using the QuikChange kit (Stratagene), we created silent mutations at bp 2109 of CCEE and bp 1815 of EECC. Restriction digests confirmed successful addition of the sites, and the coding region was sequenced to confirm the absence of errors. To construct the chimeras, CCEE (+AgeI) and EECC (+AgeI) were cut with KpnI and AgeI, and the resulting 2-kb fragments from each construct were exchanged via ligation. Finally, the excised 5-kb SalI fragment was reintroduced into both constructs to yield two full-length clones: α1E (IIS6C)/pmT2 and α1C(IIS6E)/pmT2. The EEEC and CCCE chimeras (in the pMT2 expression vector) were used as the templates for mutagenesis to introduce unique AatII sites near the beginning of the III S6 segments. Using the QuikChange kit, a silent mutation at bp 4002 of EEEC and a non-silent mutation at bp 3384 of CCCE were introduced. However, because the non-silent mutation of CCCE involved a substitution to the corresponding residue in the EEEC sequence (serine to valine), the substitution became inconsequential in the completed chimera. Successful addition of the sites was confirmed by restriction digests, and the coding region was sequenced to confirm the absence of errors. To construct the chimeras, CCCE + AatII and EEEC +AatII were cut with NotI and AatII restriction enzymes, and the resulting 4-kb fragments from each construct were exchanged and religated to yield two full-length clones: α1E (IIIS6C)/pmT2 and α1C(IIIS6E)/pmT2. To create double chimeras, α1E (IIS6C) and α1E (IIIS6C) were cut withSalI (II-III linker, 3′ polylinker, ∼5 kb). Subsequently, the 5-kb SalI fragment derived from α1E(IIIS6C) was ligated into α1E (IIS6C) to produce α1E (II/IIIS6C). An analogous approach was used to create α1C (II/IIIS6E) from α1C (IIS6E) α1C (IIIS6E). In each case, the correct orientation was determined using restriction digest patterns. To create CeCCC, a unique non-silentSplI site was generated by site-directed mutagenesis (QuikChange) at the very end of the domain I-II linker regions of CECC and CCCC at exactly complimentary positions in a stretch of residues that is completely conserved between both parent channels (VFYW). AKpnI-SplI fragment (∼1.5 kb) was excised from CECC and ligated into likewise digested CCCC to produce CeCCC but still carrying the non-silent SplI site. Finally, another round of site-directed mutagenesis was used to remove the non-silentSplI site, thereby restoring the original amino acid sequence (VFYW). To create CEEE(II+IIIS6C), anAvrII fragment cut from CCCC (900 bp before 5′ polylinker, I-II linker region, ∼2 kb) and ligated into likewise-digested α1E(II/IIIS6C). To create CECC(IIS6C), CEEE(II/IIIS6C) and EECC were cut with SalI, and the fragment from the latter chimera (corresponding to domains III and IV of α1C) was ligated into the former construct. The correct orientation of the constructs was confirmed via restriction digests. To create CcEEE(II/IIIS6C), we first introduced a silent NarI site into the EEEE(II/IIIS6C) sequence ∼20 amino acid residues before the end of the domain I-II linker region. ECCC contained an endogenousNarI site at an exactly complementary position. A 1900-bpNarI fragment (1000 bp before the 5′ polylinker, end of I-II linker) was excised from ECCC substituted in the EEEE(II/IIIS6C +NarI) construct to give rise to EcEEE(II/IIIS6C), with the lowercase letter indicating the origin of the I-II linker. This construct did not express functionally in HEK cells but was used to create the chimera CcEEE(II/IIIS6C) by substituting domain I of EcEEE(II/IIIS6C) with that of CCCC using AvrII. Correct orientations of the inserts were confirmed via restriction enzyme digests. We previously provided a detailed description of the procedures for transient expression of the wild type and chimeric calcium channels in human embryonic kidney tsa-201 cells and their electrophysiological analysis via whole cell patch clamp (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Unless stated otherwise, the external and internal recording solutions contained, respectively, 20 mm BaCl2, 1 mm MgCl2, 10 mm HEPES, 40 mm tetraethylammonium chloride, 10 mm glucose, 65 mm CsCl (pH 7.2 with tetraethylammonium hydroxide) and 108 mm cesium methanesulfonate (CsMS), 4 mmMgCl2, 9 mm EGTA, 9 mm HEPES (pH 7.2 with tetraethylammonium hydroxide), thus minimizing the possibility of contamination from calcium-dependent inactivation processes. Pipette resistances were typically on the order of 3 to 4 MΩ, and series resistance was compensated by 85% to minimize voltage errors. Currents were typically elicited from holding potentials of −100 mV (or −130 mV for α1E and other chimeras which activated more negatively) to various test potentials using Clampex software (Axon Instruments). However, to obtain steady state inactivation curves, a 5-s conditioning pulse preceded a test depolarization to +10 mV. The rate of inactivation was assessed by considering both the percentage of current that had inactivated over a time course of 125 ms and by mono-exponential fits to the time course of inactivation. Data were analyzed using Clampfit (Axon Instruments) and Sigmaplot 4.0 (Jandel Scientific). Steady state inactivation curves and macroscopic current voltage relations were analyzed using the Boltzmann equation (see Ref. 34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). All error bars are standard errors, numbers in parentheses displayed in the figures reflect numbers of experiments, and p values were determined by Student'st tests. We previously reported that wild type α1C and α1E channels exhibited pronounced differences in their inactivation profiles (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Since completion of our original study, we switched our internal recording solution (105 mm CsCl, 25 mm tetraethylammonium chloride, 11 mm EGTA, and 10 mm HEPES, pH 7.2) to a solution composed of 108 mm CsMS, 4 mm MgCl2, 9 mm EGTA, 9 mm HEPES (pH 7.2), which we found to yield more stable recordings. Hence, it was necessary to reassess the inactivation properties of the two wild type channels under our present experimental conditions. Fig. 1 compares the inactivation profiles of wild type α1C and α1E channels, coexpressed with α2-δ and β1b subunits. As shown in the figure, the two wild type channels exhibit diametrically different inactivation properties such that the half-inactivation potential of α1E is 60 mV more negative than that of α1C. Furthermore, the rate of inactivation, expressed either as the time constant for current decay, τ, or the percentage of current that has inactivated over a time course of 125 ms, is 2–4-fold greater for α1E than α1C, depending on the test potential. These differences in the inactivation profiles of the two channels are qualitatively consistent with our previous recordings (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar); however, a closer comparison with our previous work reveals three quantitative differences. First, the current densities obtained in the CsMS internal solution were on average twice as large as those observed in CsCl (not shown). Second, the half-inactivation potentials of α1Eand α1C were shifted, respectively, by 10 to 20 mV in the hyperpolarizing direction, whereas the half-activation potential was not significantly affected (p > 0.05). Finally, the inactivation rates in internal CsMS were considerably accelerated. Neither the substitution of negative counter ion nor the presence or absence of internal magnesium was found to account for the effects (not shown). However, internal TEA ions significantly affected the inactivation properties of the channel such that the presence of 25 mm internal TEA resulted in a significant slowing of the inactivation rate (Fig. 1, C and D) and an ∼10-mV rightward shift in the midpoint of the steady state inactivation. Thus, TEA ions, commonly thought to be inert for voltage-dependent calcium channels, exert a pronounced effect on calcium channel gating. Our previous work showed that multiple transmembrane domains were involved in the inactivation process of α1E channels, with domain II and III contributing to the greatest extent (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). We theorized that calcium channel inactivation may occur via a mechanism reminiscent of that underlying C-type inactivation common to many types of voltage-dependent potassium channels, a process that is believed to involve pore constriction mediated by the S6 segments of the channel (35Choi K.L. Aldrich R.W. Yellen G. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 5092-5095Crossref PubMed Scopus (394) Google Scholar, 36Kukuljan M. Lebarca P. Latorre R. Am. J. Physiol. 1995; 268: C425-C436Crossref PubMed Google Scholar). To assess a putative role of the S6 segments in fast calcium channel inactivation, we created chimeras in which the S6 regions of domains II/III of α1C were replaced by the corresponding regions of α1E. As seen in Fig. 2 A, replacement of the II S6 or III S6 segments of α1C with that of α1E dramatically increased the inactivation rate of α1C to levels observed with wild type α1E channels. These data suggest that the individual S6 segments in domains II and III are important determinants of calcium channel inactivation. To test whether the effects of domains II and III were additive, we also examined a double chimera α1C(II/IIIS6E) in which the S6 segments of domains II and III of α1E were inserted into α1C concomitantly (Fig. 2). The double replacement did not result in further speeding of the inactivation kinetics, nor could it enhance the slower, α1C-like rates that persisted at relatively hyperpolarized test potentials in the two single S6 chimeras. Thus, although the presence of a single "inactivating" S6 segment is sufficient to confer many aspects of the more rapid inactivation of the wild type α1E channels, even their combination cannot account for all of the voltage dependence associated with the inactivation rates. If the presence of a single S6 segment in domain II or III of α1E is sufficient to confer rapid inactivation kinetics onto α1C, one might expect that replacement of only one of those two S6 segments in α1Eshould be ineffective in removing inactivation. To test this hypothesis, we examined two additional chimeras, α1E(IIS6C) and α1E (IIIS6C). As seen in Fig.3, A and B, the two chimeras exhibited inactivation rates that did not differ significantly from those of the wild type α1E channel, except at relatively hyperpolarized test potentials. However, even simultaneous substitution of the II S6 and III S6 regions of α1E with those of α1C (i.e. α1E(II/IIIS6C)) did not significantly slow inactivation (Fig.3 C), suggesting the presence of an additional region in the α1E sequence that is independently capable of maintaining inactivation. To identify the putative region sustaining rapid inactivation of α1E, we first inserted the α1C domain I into EEEE(II/IIIS6C), creating the CEEE(II/IIIS6C) chimera, and still, α1E-like inactivation persisted (Fig. 3 D). However, upon substitution of most of the α1C domain I-II linker region into this construct, inactivation was virtually abolished (Fig. 3 E), suggesting that it was the domain I-II linker region of α1E that maintained α1E-like inactivation rates in the α1E (II/IIIS6) construct. To unequivocally show the importance of the domain I-II linker region, we created two additional chimeras (EcEEE and EcEEE (II/IIIS6C)); however, neither construct expressed functionally in tsa-201 cells. Nonetheless, if our hypothesis is correct, then an α1C channel containing the domain I-II linker of α1E should exhibit rapid inactivation kinetics. Data obtained with such a chimera (CeCCC) are shown in Fig. 3 F. As evident from the figure, the CeCCC construct exhibited inactivation kinetics that were significantly faster than those seen with the wild type α1C channels, consistent with our hypothesis. We also examined an additional construct that contained the domain I-II linker plus the first five transmembrane segments of domain II of α1E (thus retaining the "non-inactivating" domain II S6 and III S6 regions of α1C), and similar to that of the CeCCC chimera, the CECC(IIS6C) construct exhibited inactivation kinetics that were significantly more rapid than those of the wild type α1Cchannel (n = 12, not shown). Taken together, this supports the idea that the presence of either one of three regions in α1E, the domain I-II linker or II S6 or III S6 regions, is sufficient to preserve rapid inactivation, whereas the concomitant replacement of these three regions with the corresponding elements of α1C is required to confer slow inactivation kinetics. We have previously shown that domains II and III could account for the majority of differences in half-inactivation potential between the two wild type channels, whereas domains I and IV contributed to a lesser extent (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). To test whether the effects of domains II and III could be attributed to the S6 regions, we compared the half-inactivation potentials of the wild type and chimeric calcium channels (Fig. 4). Replacement of the II S6 region of α1E did not affect the half-inactivation potential, and the reverse substitution in α1C resulted in only a small ∼10-mV hyperpolarizing shift in steady state inactivation kinetics, which was paralleled by a comparable change in half-activation potential (Fig. 4). Thus, it seems unlikely that the domain II S6 region contributes in a meaningful manner to the determination of steady state inactivation kinetics despite its pronounced effects on inactivation rate. Replacement of the domain III S6 region resulted in more substantial (∼15 mV) hyperpolarizing shifts in half-inactivation potential that were not mirrored by activation potential shifts. In view of the 60-mV spread between the wild type channels, the contributions from the S6 regions were relatively minor, suggesting that other regions in domains II and III may determine the voltage dependence of inactivation. Indeed, both the CeCCC and CECC (IIS6C) constructs inactivated 20 mV more negatively than the wild type α1C channel (Fig. 4) despite sharing a common half-activation potential with α1C, thereby implicating the domain I-II linker. Thus, although insertion of a single inactivating structure of α1E is sufficient to confer the rapid inactivation profile in an essentially all or none fashion, multiple substitutions, including some that do not affect inactivation rates (i.e.domain I in Fig. 4D of Ref. 34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar), are required to account for the differences in voltage dependence of inactivation of the wild type channels. We previously presented evidence that the calcium channel domains II and III are critical determinants of both the voltage dependence and the rate of inactivation (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Each of those two domains contributed to about half of the observed differences in half-inactivation potential between α1C and α1E, and insertion of either domains II or III of α1E into the α1Csequence conferred all of the rapid inactivation kinetics of α1E (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Here, we have more narrowly identified the regions involved in determining the inactivation rate. As seen from Table I, with one exception (ECEE), any construct containing the α1E sequence in the domain IIS6, III S6, or I-II linker regions exhibited α1E-like inactivation kinetics. In contrast, only constructs carrying α1C sequence in each of those regions inactivated slowly, consistent with the idea that the above regions are the central structural elements involved in the control of inactivation rates.Table IDependence of the inactivation rate on the presence of the domain I-II linker, and domain II S6 and III S6 regionsConstructII S6III S6I-IIE (IIS6 or IIIS6 or I-II)RateCCCCCCCSlowEEEEEEE√FastCEECEEE√FastCCECCEC√FastEECCECE√FastCECCECE√FastCEEEEEE√FastCCEECEC√FastECCCCCCSlowCCCECCCSlowECEE*CEC√Slow*CCCC (II S6E)ECC√FastCCCC (III S6E)CEC√FastCCCC (II/III S6E)EEC√FastEEEE (II S6C)CEE√FastEEEE (III S6C)ECC√FastEEEE (II/III S6C)CCE√FastCEEE (II/III S6C)CCE√FastCcEEE (II/III S6C)CCCSlowCeCCCCCE√FastCECC (II S6C)CCE√FastThe data for the inter-domain chimeras (third through tenth lines) were taken from Spaetgens and Zamponi (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). The terminology "fast" refers to α1E-like inactivation. The check marks indicate the presence of an α1E sequence in domain II S6 or III S6 or in the domain I-II linker. The asterisk denotes a chimera which we found to be an outlier in our original paper (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Open table in a new tab The data for the inter-domain chimeras (third through tenth lines) were taken from Spaetgens and Zamponi (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). The terminology "fast" refers to α1E-like inactivation. The check marks indicate the presence of an α1E sequence in domain II S6 or III S6 or in the domain I-II linker. The asterisk denotes a chimera which we found to be an outlier in our original paper (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). The involvement of the I-II linker would be consistent with the observation that two separate point mutations in this region can slow the inactivation of α1A calcium channels (7Bourinet E. Soong T.W. Sutton K. Slaymaker S. Mathews E. Monteil A. Zamponi G.W. Nargeot J. Snutch T.P. Nat. Neurosci. 1999; 2: 407-415Crossref PubMed Scopus (364) Google Scholar, 30Herlitze S. Hockerman G.H. Scheuer T. Catterall W.A. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 1402-1406Crossref PubMed Scopus (169) Google Scholar). In addition, overexpression of the I-II linker regions of α1A was found to speed inactivation of the α1A channel (31Cens T. Restituito S. Galas S. Charnet P. J. Biol. Chem. 1999; 274: 5483-5490Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar). Also consistent with our data, amino acid substitutions in the domain III S6 region have been reported to affect inactivation kinetics (32Hering S. Aczel S. Grabner M. Doring F. Berjukow S. Mitterdorfer J. Sinnegger M.J. Striessnig J. Degtiar V.E. Wang Z. Glossmann H. J. Biol. Chem. 1996; 271: 24471-24475Abstract Full Text Full Text PDF PubMed Scopus (75) Google Scholar, 37Kraus R.L. Sinnegger M.J. Glossmann H. Hering S. Striessnig J. J. Biol. Chem. 1998; 273: 5586-5590Abstract Full Text Full Text PDF PubMed Scopus (180) Google Scholar). Zhang et al. (28Zhang J.F. Ellinor P.T. Aldrich R.W. Tsien R.W. Nature. 1994; 372: 97-100Crossref PubMed Scopus (177) Google Scholar) implicate exclusively the domain I S6 region in the fast inactivation process by utilizing a series of chimeras between rabbit α1A and marine ray α1E calcium channels, which contrasts with our observations that rapid inactivation kinetics do not require the presence of α1E domain I nor is replacement of this region with α1C sequence sufficient to abolish inactivation. Both α1E and α1Ashare identical domain II and III S6 regions and differ in their domain I S6 regions in only one position (methionine in α1A versus valine in α1E (38Stea A. Soong T.W. Snutch T.P. North R.A. Handbook of Receptors and Channels: Ligand- and Voltage-gated Ion Channels. CRC Press Inc., Boca Raton, FL1995: 113-141Google Scholar)). Interestingly, the α1C sequence also contains a valine residue in this position, thus perhaps masking any subtle effects of domain I in our experiments. Alternatively, it is possible that an amino acid substitution in the domain I S6 regions secondarily affects inactivation by altering the conformation of the associated I-II linker region. We previously presented evidence that domains II and III accounted for much of the difference in half inactivation potentials seen with the wild type channels (34Spaetgens R.L. Zamponi G.W. J. Biol. Chem. 1999; 274: 22428-22438Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Within domain III, we can attribute a significant effect to the S6 segment, indicating that the III S6 region is involved in controlling both the rate and some of the voltage dependence of inactivation. A similar argument can be made for the domain I-II linker region. In contrast, the II S6 region had little effect on the voltage dependence of inactivation despite being an important feature for determining the inactivation rate. Conversely, domain I does not affect inactivation rate but does contribute to voltage dependence (compare α1E (II/IIS6C) and CEEE (II/IIIS6C)). Thus, despite some overlap, the structural determinants governing the rates and voltage dependence of inactivation appear to be distinct. Whereas the critical determinants of inactivation rate are fairly localized, the mechanism controlling the voltage dependence of inactivation appears to be more globally distributed across the calcium channel α1 subunit. Our original intent was to gather additional evidence in support of our hypothesis that fast calcium channel inactivation might occur via a mechanism reminiscent of the C-type inactivation process, which is thought to involve a pore collapse mediated by the four S6 segments (i.e. Refs. 35Choi K.L. Aldrich R.W. Yellen G. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 5092-5095Crossref PubMed Scopus (394) Google Scholar and 36Kukuljan M. Lebarca P. Latorre R. Am. J. Physiol. 1995; 268: C425-C436Crossref PubMed Google Scholar). Our data implicating the domain II and III S6 regions fit with such a model. However, the critical involvement of the cytoplasmic domain I-II linker region argues against simple pore collapse. As a result, our data are best described by a model in which the I-II linker region forms a cytoplasmic gating particle (39Armstrong C.M. Bezanilla F. J. Gen. Physiol. 1977; 70: 447-590Crossref Scopus (764) Google Scholar) similar to that proposed for the domain III-IV linker of voltage-dependent sodium channels (23Vassilev P. Scheuer T. Catterall W.A. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 8147Crossref PubMed Scopus (172) Google Scholar, 24Eaholtz G. Scheuer T. Catterall W.A. Neuron. 1994; 12: 1041-1048Abstract Full Text PDF PubMed Scopus (136) Google Scholar) (see Fig.5). If so, then the S6 regions might perhaps serve as the docking site for the inactivation gate. The current belief that that S6 segments line the inner vestibule of the pore would be consistent with such a mechanism (40Doyle D.A. Morais-Cabral J. Pfuetzner R.A. Kuo A. Gulbis J.M. Cohen S.L. Chait B.T. MacKinnon R. Science. 1998; 280: 69-77Crossref PubMed Scopus (5746) Google Scholar, 41MacKinnon R. Cohen S.L. Kuo A. Lee A. Chait B.T. Science. 1998; 280: 106-109Crossref PubMed Scopus (370) Google Scholar, 42Lopez G.A. Nung Y. Jan L.Y. Nature. 1994; 367: 179-182Crossref PubMed Scopus (148) Google Scholar), but it may well be possible that other regions of the channel could be part of the docking interaction. A putative role of the domain I-II linker as the inactivation gate would fit the previously reported effects of point mutations in the α1A calcium channel I-II linker (7Bourinet E. Soong T.W. Sutton K. Slaymaker S. Mathews E. Monteil A. Zamponi G.W. Nargeot J. Snutch T.P. Nat. Neurosci. 1999; 2: 407-415Crossref PubMed Scopus (364) Google Scholar, 30Herlitze S. Hockerman G.H. Scheuer T. Catterall W.A. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 1402-1406Crossref PubMed Scopus (169) Google Scholar) and the ability of overexpressed α1A I-II linker to accelerate the inactivation rate of α1A calcium channels. This model could also account for the effects of the calcium channel β subunits, which are known to physically bind to the I-II linker region, on inactivation rate (10Stea A. Tomlinson W.J. Soong T.W. Bourinet E. Dubel S.J. Vincent S.R. Snutch T.P. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 10576-10580Crossref PubMed Scopus (307) Google Scholar, 43Pragnell M. DeWaard M. Mori Y. Tanabe T. Snutch T.P. Campbell K.P. Nature. 1994; 368: 67-70Crossref PubMed Scopus (552) Google Scholar, 44Olcese R. Neely A. Qin N. Wei X. Birnbaumer L. Stefani E. J. Physiol. (Lond.). 1996; 497: 675-686Crossref Scopus (45) Google Scholar). The antagonistic effects of the β2a subunit on inactivation (44Olcese R. Neely A. Qin N. Wei X. Birnbaumer L. Stefani E. J. Physiol. (Lond.). 1996; 497: 675-686Crossref Scopus (45) Google Scholar) could perhaps arise from a restricted mobility of the domain I-II linker region as a consequence of anchoring the palmitoylated N terminus of this subunit to the plasma membrane (45Qin N. Platano D. Olcese R. Costantin J.L. Stefani E. Birnbaumer L. Proc. Natl. Acad. Sci. U. S. A. 1998; 14: 4690-4695Crossref Scopus (156) Google Scholar). A hinged-lid model could also accommodate our observations that intracellular TEA slows inactivation rates. If a TEA molecule acting as a low affinity blocker were to compete with the I-II linker for its docking site, one would expect to observe a slowing of the macroscopic time course of inactivation. Such a mechanism would not be without precedent, as TEA prevents inactivation gate closure of shaker B potassium channels (35Choi K.L. Aldrich R.W. Yellen G. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 5092-5095Crossref PubMed Scopus (394) Google Scholar), and open channel block of batrachotoxin-activated cardiac sodium channels by local anesthetics and related compounds prevents fast inactivation (46Zamponi G.W. Sui X. Codding P.W. French R.J. Biophys. J. 1993; 65: 2324-2334Abstract Full Text PDF PubMed Scopus (18) Google Scholar, 47Zamponi G.W. French R.J. Biophys J. 1993; 65: 2335-2347Abstract Full Text PDF PubMed Scopus (57) Google Scholar). How can a hinged-lid mechanism account for the observation that the presence of either the domain IIS6, III S6, or the domain I-II linker region of α1E was generally sufficient to mediate rapid inactivation? Within the framework of our model, the domain I-II linker of α1C would have the ability to dock to either the domain II S6 or III S6 regions of α1E. Conversely, the domain I-II linker of α1E would have to be capable of interacting with either one of the domain II S6 or III S6 regions of α1C. In contrast, the relative lack of inactivation of L-type channels would require an inability of the α1CI-II linker region to dock effectively to the domain II S6 and III S6 regions when of α1C origin. Biochemical evidence, however, will ultimately be required to prove a putative existence of a physical binding interaction between the domain I-II linker and the S6 segments. In summary, a hinged-lid model of inactivation can nicely account for our data as well as the key observations reported in the literature. The redundancy of the structural elements that are sufficient to maintain rapid inactivation underlines the fundamental importance of this process for the precise control of calcium entry and, thus, prevention of accumulation of toxic levels of intracellular calcium (18Choi D.W. Trends Neurosci. 1988; 11: 465-469Abstract Full Text PDF PubMed Scopus (1609) Google Scholar, 19Orrenius S. McConkey D.J. Bellomo G. Nicotera P. Trends Pharmacol. Sci. 1989; 10: 281-285Abstract Full Text PDF PubMed Scopus (790) Google Scholar, 20Orrenius S. Nicotera P. J. Neural Transm. Suppl. 1994; 43: 1-11PubMed Google Scholar). We thank Dr. T. P. Snutch for the wild type calcium channel cDNA constructs.