C-terminal Fragments of the a 1C (Ca V 1.2) Subunit Associate with and Regulate L-type Calcium Channels Containing C-terminal-truncated a 1C Subunits*

, L-type Ca 2 1 channels in native tissues have been found to contain a pore-forming a 1 subunit that is often truncated at the C terminus. However, the C terminus contains many important domains that regulate channel function. To test the hypothesis that C-terminal fragments may associate with and regulate C-terminal-truncated a 1C (Ca V 1.2) subunits, we performed elect- rophysiological and biochemical experiments. In tsA201 cells expressing either wild type or C-terminal-trunca-ted a 1C subunits in combination with a b 2a subunit, tru- ncation of the a 1C subunit by as little as 147 amino acids led to a 10–15-fold increase in currents compared with those obtained from control, full-length a 1C subunits. Dialysis of cells expressing the truncated a 1C subunits with C-terminal fragments applied through the patch pipette reconstituted the inhibition of the channels seen with full-length a 1C subunits. In addition, C-terminal deletion mutants containing a tethered C terminus also exhibited the C-terminal-induced inhibition. Immunoprecipitation assays demonstrated the association of the C-terminal fragments with truncated a 1C subunits. In addition, glutathione S -transferase pull-down assays demonstrated that the C-terminal inhibitory fragment could associate with at least two domains within the C terminus. The results support the hypothesis the C-terminal fragments of the a 1C subunit can associate with C-terminal-truncated a M M m M Ba of the size of peak of the initially recorded whole cell Ba potential and peak inward current plotted on the ordinate versus time starting from when the cell was patched and intracellular access was first obtained on the ordinate. Peak currents were analyzed using the ISO-2 (MFK, Frankfurt/Main, Germany) anal- ysis software and normalized to the whole cell capacitance. The data for each condition were pooled and expressed as the mean 6 S.E. All experiments were repeated at least four times.

The voltage-activated L-type Ca 2ϩ channels are heteromeric proteins minimally composed of a pore-forming ␣ 1 subunit and accessory ␣ 2 ␦ and ␤ subunits (1,2). Each ␣ 1 subunit contains four repeated domains containing a total of 24 membranespanning domains as well as a long hydrophilic C terminus which contains important regulatory domains that contribute to channel regulation. For example, the C terminus of the ␣ 1C subunit constitutes ϳ30% of the total mass of the ␣ 1C subunit (3) and is critical for membrane targeting of the channels (4), the regulation of the channels by protein phosphorylation (5), and the binding of Ca 2ϩ -binding proteins such as calmodulin and sorcin (6 -8). In addition, the C terminus of ␣ 1C appears to contain inhibitory domains because deletion of up to ϳ70% of the C-terminal 665 amino acids leads to increased currents (9).
A puzzling observation that has been made in several laboratories is that the C terminus of several L-type Ca 2ϩ channels appears to be truncated in many native tissues. For example, when the ␣ 1C subunit was isolated from cardiac myocytes, only 10 -15% of the total protein was a full-length 240-kDa ␣ 1C subunit, whereas the majority migrated on SDS gels as a ϳ190-kDa protein that was lacking the distal ϳ50 kDa of the C terminus (10). Similar observations have been made for the ␣ 1C subunit expressed in brain (11) and the ␣ 1S subunit isolated from skeletal muscle (12,13). If the C termini of these proteins were truly absent, this would have major implications for channel regulation. Although many protease inhibitors have been used to try to prevent this truncation, none has altered the proportion of full-length to truncated protein (14). In marked contrast to what has been observed in native systems, fulllength ␣ 1C subunits have been observed in heterologous expression systems (5,15,16). Despite the fact that ϳ90% of the ␣ 1C subunits appeared to be truncated when isolated from cardiac myocytes, immunofluorescence imaging suggested that the C terminus of the ␣ 1C subunit was present in a 1:1 ratio with the ␣ 1C subunit and co-localized with channel subunits in cardiac myocytes (10). This finding led us to consider the possibility that the processing of the ␣ 1C subunit might have physiological importance and that the C-terminal fragments might remain functionally associated with the channels (17). Such a scenario would allow for maintenance of the important regulatory functions ascribed to the cleaved C terminus. As a first test of this hypothesis, we demonstrated that an exogenous protease, chymotrypsin, cleaved the full-length 240-kDa expressed ␣ 1C subunit in vitro into a "body" of 190 kDa, similar to what is observed in native systems (10), and C-terminal fragments of 30 -50 kDa that remained associated with the membrane (17).
Here we report studies in which we have tested the ability of C-terminal fragments to associate with and regulate the conductance of C-terminally cleaved ␣ 1C subunits expressed in intact cells.

EXPERIMENTAL PROCEDURES
Materials-All reagents were obtained from general sources unless otherwise stated. Antibodies Card I, Card C (15), and CT1 1,2 as well as * This work was supported by National Institutes of Health Grant HL23306 (to M. M. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Antibody Preparation-To generate an additional antibody that would recognize C-terminal fragments of the ␣ 1C subunit, a fusion protein encoding amino acid residues 1907-2171 (termed CT4, see Fig.  1B) in the C terminus of the ␣ 1C subunit was produced. The sequence in the CT4 region was subcloned into an expression vector pQE32 (Qiagen), resulting in an in-frame fusion of the 6ϫ-His tag to the CT4 residues. The 6ϫ-His-CT4 was produced and purified under native conditions from bacteria (15). Purified CT4 fusion proteins were injected into a rabbit, and polyclonal antibodies were prepared at Bethyl Laboratories (Montgomery, TX). The specificity of this antibody was tested using the wild type ␣ 1C subunit, the ␣ 1C ⌬1905 subunit that lacks the CT4 fragment (Fig. 1), and CT4 expressed in tsA201 cells (see Fig. 1C).
Expression of Channel Subunits and C-terminal Fragments of the ␣ 1C Subunit in Mammalian Cells-Various C-terminal fragments of the ␣ 1C subunit (see Fig. 1) were expressed in tsA201 cells (HEK293 cells transformed with large T-antigen (18)) using the following strategies. The C-terminal domain (CT) containing amino acids 1623-2171 of ␣ 1C was excised as a BglII/BamHI fragment. The CT4 (containing amino acids 1907-2171) and CT23 (containing amino acids 1623-1905) fragments were derived from pCR3␣ 1C and pCR3␣ 1C ⌬1905 constructs as BamHI/BamHI and BglII/BamHI fragments, respectively. The CT7 fragment containing amino acids 2026 -2171 was derived from pCR3␣ 1C ⌬2024 construct as a BamHI/BamHI fragment. A fusion protein expression vector, pCR3His/Myc, was derived from pCR3 (Invitrogen) by inserting a 6ϫ-His tag and six copies of the Myc epitope into the multiclonal sites of the original vector to allow expression of a protein with both 6ϫ-His-and Myc tag fused to the N terminus. To create expression constructs encoding different regions of the C terminus of the ␣ 1C subunit, the cDNA fragments of CT, CT4, CT7, and CT23 were subcloned into the BamHI-digested pCR3His/Myc vector, resulting in in-frame fusions of the C-terminal fragments to the 6ϫ-His and six copies of Myc tags on the vector. TsA201 cells were maintained in Dulbecco's modified Eagle's medium (Life Technologies, Inc. containing 10% fetal bovine serum (Life Technologies, Inc. and 1% penicillin/ streptomycin at 37°C in 5% CO 2 . Transient expression of the Myctagged C-terminal fusion proteins in tsA201 cells along with wild type or mutant rabbit ␣ 1C (Ca V 1.2) subunits (3) and the rat ␤ 2 a subunit (19) was carried out using a total of 30 -40 g of plasmid DNA/100-mm plate and either the calcium phosphate precipitation method or the transfection reagent Effectene from Qiagen as described (15,20). For electrophysiological experiments, tsA201 cells were plated onto 6-cm culture dishes to achieve ϳ40 -60% confluency. The indicated ␣ 1C subunit construct cDNAs were co-transfected with ␤ 2a subunit cDNA at 1.5 g each along with 0.2 g of CD8 cDNA (as an expression indicator (21)). On the day of the experiment transfected cells were washed with phosphate-buffered saline, dissociated using trypsin-EDTA (Life Technologies, Inc.) and transferred onto either 35-mm culture dishes or 12-mm glass coverslips, each previously coated with rat tail collagen type VII (1 g/ml) to achieve ϳ40% confluency. Cells were allowed to settle on the plate for 2 h before electrophysiological recordings.
Expression of C-terminal Fragments in Bacteria and GST Fusion Protein Pull-down Assays-To express C-terminal fragments of ␣ 1C in bacteria as GST fusion proteins, the cDNAs encoding CT, CT4, CT7, CT8, CT12, CT14, or CT23 (see Fig. 1) were subcloned into the BamHI site of the GST fusion protein expression vector pGEX-5X-2 (Amersham Pharmacia Biotech). The GST-tagged C-terminal fusion proteins were expressed in Escherichia coli BL21 and purified using glutathioneagarose after standard procedures (Amersham Pharmacia Biotech). CT7 and CT8 also were expressed and purified as 6ϫ-His-tagged fusion proteins using pQE32-CT7 and pQE30-CT8 and standard procedures. The fusion proteins were used in electrophysiological assays and in GST pull-down experiments. The GST pull-down experiments were performed using various GST-C-terminal fusion proteins containing CT, CT4, CT7, CT8, CT12, CT14, CT23 (see Fig. 1B) as well as NT (Nterminal amino acids 1-154), L1 (loop between conserved domains I and II, amino acids 437-554), and L2 (loop between conserved domains I and II, amino acids 785-930). In each reaction, purified 6ϫ-His-CT7 was diluted into 0.5 ml of binding buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.4, plus protease inhibitors (15)) and added to 50 l of glutathione-Sepharose beads precoupled to either GST alone (control) or to the various GST-C-terminal, N-terminal, or loop constructs. Incubations were carried out for 4 -6 h at 4°C with agitation. The GST beads were washed 4 times with 1 ml of washing buffer (300 mM NaCl, 50 mM Tris-HCl, pH 7.4), and after the last wash, 1/5 volume of SDS-polyacrylamide gel electrophoresis Laemmli buffer (22) was added to each reaction. The proteins were separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose, and the bound 6ϫ-His-CT7 was detected by immunoblotting with the CT4 antibody.
Co-immunoprecipitation of C-terminal Fragments with the Channel Complexes-TsA201 cells were transiently co-transfected with a combination of the channel subunits and different C-terminal fusion proteins. Approximately 40 h post-transfection, whole cell lysates were prepared in lysis buffer (20 mM Na 2 HPO 4 , pH 7.4, 150 mM NaCl, 1% Triton X-100 plus protease inhibitors (15)). The whole cell lysates were diluted 1:3 with lysis buffer without Triton X-100 so that the final concentration of Triton X-100 was 0.3%. The diluted cell lysates were incubated with either one of the ␣ 1C subunit-specific antibodies, Card I or CI2, coupled to Ultra-link protein G (Pierce) at 4°C overnight. The immunoprecipitates were washed with wash buffer (20 mM Na 2 HPO 4 , pH 7.4, 150 mM NaCl, and 0.1% Triton X-100, 3ϫ1 ml) and analyzed using SDS-polyacrylamide gel electrophoresis and immunoblotting. The detection of the ␣ 1C subunits was with the Card I or CI2 antibodies, as specified in the figure legends. Detection of co-immunoprecipitated fusion proteins was with the CT4 or anti-Myc antibodies.
In experiments where cross-linking was performed before immunoprecipitation, cells were lysed in lysis buffer, and cross-linking was performed in the presence of a 20 mM phosphate and 0.1 mM of the nickel (Ni(II)) complex of the tripeptide NH 2 -Gly-Gly-His-COOH (GGH-Ni(II)) (23). The nickel-peptide complex was formed by mixing a 1:1 molar ratio of nickel acetate and GGH in water. After a 5-min equilibration, the solution was added to cell lysates to a final volume of 0.4 ml. The cross-linking reactions were initiated by the addition of magnesium monoperoxyphthalic acid hexahydrate (0.1 mM) and incubated at room temperature for 10 min. The reactions were quenched by the addition of 1 l of 0.02 M thiourea. The cell lysates were then diluted and subjected to immunoprecipitation with the Card I antibody as described above.
Electrophysiological Assays-Ba 2ϩ currents were obtained at room temperature (ϳ22 Ϯ 1°C) from the transfected human embryonic kidney cells using the whole cell patch voltage clamp technique. The currents were recorded utilizing an EPC-7 patch clamp amplifier (List) whose analog output signal was low pass-filtered at 3 kHz and then digitally sampled at 5 kHz using an ISO-2 data acquisition system (MFK, Frankfurt/Main, Germany). The input series resistance was ϳ70% electronically compensated by adjusting the circuit to a point just below that which would begin to cause ringing. Patch pipette electrodes having resistances of 2-5 megaohms when filled with internal solution were manufactured from borosilicate glass capillary tubing (Warner Instrument Corp, #GC150F-10) using a P-97 micropipette puller (Sutter Instrument Co.); the pipette tips were polished using a Narashige MF-83 pipette polisher. The internal solution used to fill the pipette electrodes contained 60 mM CsCl, 5 mM HEPES, 1 mM MgCl 2 , 5 mM MgATP, 0.1 mM MgGTP, 10 mM ethylene glycol-bis(␤-aminoethyl ether) N,N,NЈ,NЈ tetraacetic acid (EGTA), 78 mM methane sulfonic acid, and 78 mM n-methyl-D-glucamine (corrected to pH 7.3 with CsOH; if necessary, adjusted to 330 mosmol with methanesulfonate N-methyl-D-glucamine). To perform electrophysiological experiments, the cells were placed in a bath chamber that was perfused constantly with a solution containing 30 mM NaCl, 10 mM BaCl, 5 mM HEPES, 20 mM CsCl, 1 mM MgCl 2 , 78 mM methanesulfonic acid, and 78 mM N-methyl-D-glucamine. The pH was corrected to 7.4 with NaOH, and osmolarity was adjusted to 330 mosmol.
Immediately before attempting to obtain whole cell patches, anti-CD8 antibody-coated Dynabeads (Dynal, Oslo, Norway) were added to the bath chamber to identify which cells had been successfully transfected cells (21). For some experiments, the internal (pipette) solutions also contained one of several indicated peptides at a concentration of 1 g of protein/ml, and whole cell patch Ba 2ϩ current was recorded either in the presence or absence of the particular peptide. Data depicting the effect of each of the several peptides to modify the size of the peak of the initially recorded whole cell Ba 2ϩ current evoked in response to 50-ms depolarizing pulses to ϩ10 mV at 0.1 Hz from a holding potential of Ϫ90 mV were recorded and are presented as the peak inward current plotted on the ordinate versus time starting from when the cell was patched and intracellular access was first obtained on the ordinate. Peak currents were analyzed using the ISO-2 (MFK, Frankfurt/Main, Germany) analysis software and normalized to the whole cell capacitance. The data for each condition were pooled and expressed as the mean Ϯ S.E. All experiments were repeated at least four times.

Progressive C-terminal Deletions of the Full-length ␣ 1C Ltype Ca 2ϩ Channel Results in Potentiated Channel Currents-
Previous studies demonstrate that truncation of 307-472 amino acids from the C terminus of ␣ 1C (in mutants ␣ 1C ⌬1856, ␣ 1C ⌬1733, and ␣ 1C ⌬1700) produced increased channel currents in Xenopus oocytes (9), suggesting that the C terminus contains inhibitory elements. To test the hypothesis that C-terminal fragments of the ␣ 1C subunit might associate with and regulate the conductance of C-terminal-truncated ␣ 1C subunits, we determined if expressed C-terminal fragments could reconstitute the inhibition of currents when co-expressed with C-terminaltruncated mutants of ␣ 1C . To do so, we first prepared and analyzed currents from several different mutants. Using 10 mM Ba 2ϩ as a charge carrier, whole cell current density was compared from channels containing the full-length ␣ 1C subunit or from the C-terminal deletion mutants ␣ 1C ⌬2024, ␣ 1C ⌬1905, and ␣ 1C ⌬1733 (see Fig. 1). Each construct was transiently expressed in tsA-201 cells, and all channel constructs were co-expressed with the rat ␤ 2a subunit (19). The current-voltage relationships of channels containing full-length ␣ 1C , ␣ 1C ⌬2024, ␣ 1C ⌬1905, or ␣ 1C ⌬1733 were determined (Fig. 2). The fulllength channel displayed a characteristic L-type current-voltage profile with the maximal peak I Ba at 0 mV and a mean current density of 4.97 Ϯ 1.3 (mean Ϯ S.E., n ϭ 4) pA/pF. In marked contrast, all three truncation mutants displayed significantly larger currents at all voltages between Ϫ20 and ϩ40 mV. The current-voltage (I-V) relationship obtained from cells expressing the ␣ 1C ⌬2024 subunit displayed maximal current at ϩ10 mV with an average peak I Ba of 60.1 Ϯ 19.2 pA/pF (n ϭ 5). Similarly, currents from cells expressing the ␣ 1C ⌬1905 or ␣ 1C ⌬1733 subunits also exhibited peak currents between 0 to ϩ10 mV with current densities of 59.3 Ϯ 12.9 pA/pF (n ϭ 4) and 44.7 Ϯ 13.4 pA/pF (n ϭ 6), respectively. These results demonstrated that expression of any of the three truncation mutants gave rise to markedly larger currents in mammalian cells in a manner similar to that previously described for the truncation mutants ␣ 1C ⌬1856, ␣ 1C ⌬1733, and ␣ 1C ⌬1700 that were expressed in Xenopus oocytes. More importantly, the results demonstrated that a C-terminal truncation of the ␣ 1C subunit by as little as 147 amino acids (␣ 1C ⌬2024) displayed markedly enhanced currents when compared with the intact full-length channel. This suggested that the inhibitory motif might be contained between amino acids 2025 and 2171.
Cytosolic Application of ␣ 1C C-terminal Fragments to Cells Expressing Truncated ␣ 1C Channels Resulted in Time-dependent Reduction of I Ba -We next tested if application of C-terminal fragments would inhibit the C-terminal-truncated ␣ 1C subunits. We first tested if the CT fragment, corresponding to amino acids 1622-2171 ( Fig. 1), would result in inhibition of I Ba from ␣ 1C ⌬2024 and ␣ 1C ⌬1905. GST-CT was applied through the patch clamp pipette at a peptide concentration of 1 g/ml pipette solution. During recordings, cells were maintained at Ϫ90 mV holding potential and depolarized to ϩ10 mV for 50 ms at 10-s intervals. In the control cells, currents in both the ␣ 1C ⌬1905 (Fig. 3)-and ␣ 1C ⌬2024 (data not shown)-expressing cells exhibited a time-dependent increase in currents that began immediately upon patching the cells. Peak currents were achieved within ϳ3 min after access to the cell was achieved and were maintained with minimal decrease (Ͻ10%) during the 15-min recording period (Fig. 3). In cells exposed to CT, peptide currents also showed an initial increase in current density (Fig. 4). However, a reduced peak, compared with control, was attained within 2 min of open access to the cytosol (Fig. 4). Current continued to decline such that I Ba was markedly reduced compared with control after a 6 -10-min exposure of either ␣ 1C ⌬2024 or ␣ 1C ⌬1905 to CT (Fig. 4, A and B, and Table I). The inhibitory effect of CT was essentially abolished when the CT peptide was boiled for 5 min before addition to pipette solution (Table I). These data demonstrated that the inhibition of channels could be reconstituted by applying the CT peptide to cells expressing the C-terminal-truncated mutants, ␣ 1C ⌬2024 and ␣ 1C ⌬1905. Furthermore the structural integrity of the peptide was necessary for the peptide to exert its inhibitory effect. These data provided initial support for the hypothesis that C-terminal fragments of ␣ 1C can associate with and regulate channel activity in intact cells.
The Distal Portion of the C-terminal Tail Contains a Channel subunit. The C terminus of each construct is depicted with the starting and ending sites labeled. The beginning of the C terminus is amino acid residue 1507, and the last amino acid is 2171. WT, wild type. C, Western blots depicting the C-terminal fusion proteins used in the electrophysiological assays. The left side of the blot depicts the specificity of the CT4 antibody. TsA201 cells were transfected with either the wild type ␣ 1C and ␤ 2 subunits (lanes 1 and 4) or the ␣ 1C ⌬1905 and ␤ 2 subunits (lanes 2 and 5) or with 6ϫ-His-CT4 only. Lysates from cells containing channel subunits were immunoprecipitated with the Card I antibody (lanes 1, 2, 4, and 5). Lysates from cells expressing CT4 were enriched for CT4 by concentration on nickel resin (lanes 3 and 6). Western blotting was performed with the CT4 antiserum (lanes 1-3) or with pre-immune sera (lanes 4 -6). Inhibitory Domain-We next attempted to define smaller regions of the C terminus that might be responsible for channel inhibition. We focused initially on fragments CT4 and CT7, which were approximately equivalent to the fragments deleted from ␣ 1C ⌬1905 and ␣ 1C ⌬2024, respectively (see Fig. 1). Application of the GST-CT4 peptide to cells expressing either ␣ 1C ⌬2024 or ␣ 1C ⌬1905 caused a marked inhibition of currents in a manner similar to the CT peptide ( Fig. 5 and Table I). An early and reduced current peak was observed within 4 min of application of CT4; current density was markedly reduced compared with control in ␣ 1C ⌬2024and ␣ 1C ⌬1905-expressing cells from 4 to 10 min after the start of the perfusion (Fig. 5, A and  B, Table I). Boiling the CT4 peptide again abolished the inhibitory effect (Table I).
Similar experiments were performed with GST-CT7. Surprisingly, GST-CT7 was without effect on currents from cells expressing either ␣ 1C ⌬2024 or ␣ 1C ⌬1905. However, since truncation of the CT7 fragment in the ␣ 1C ⌬2024 mutant resulted in a loss of inhibition (Fig. 2), we next asked if a different type of CT7 fusion protein, 6ϫ-His-CT7, could inhibit the C-terminaltruncated mutants. Conceivably, the GST construct may have hampered the presentation of CT7 to the channels. Application of 6ϫ-His-CT7 to cells expressing ␣ 1C ⌬1905 caused a marked reduction from control within 6 -8 min after access to the cytosol (Fig. 6A, Table I). The 6ϫ-His-CT7 also caused inhibition of currents from ␣ 1C ⌬2024 (Fig. 6B, Table I). The effects of CT7 were similar to those of CT and CT4, although the inhibition appeared to be less robust than that caused by CT4. In particular, the inhibition of currents from ␣ 1C ⌬2024 by CT7 appeared to develop more slowly and to a smaller extent that that caused by CT4 (compare Figs. 5 and 6, Table I). Nevertheless, these results demonstrated that application of CT7 could reconstitute inhibition of currents from either ␣ 1C ⌬1905 or ␣ 1C ⌬2024. The effects of CT7 were resistant to boiling. It is well known that the activity of many small proteins, such as calmodulin, can be resistant to denaturation (24), and thus, the lack of inhibition of the Ba 2ϩ current during intracellular dialysis with the small, boiled 6ϫ-His-CT7 was not surprising.
We also tested the effects of other peptides derived from the C terminus of ␣ 1C . As a control, we tested GST alone and found it had no effects on the currents (Fig. 7, A and B, Table I). Peptide GST-CT23, corresponding to amino acids 1622-1905, also had no effect on channel currents obtained from either ␣ 1C ⌬2024 and ␣ 1C ⌬1905 (Fig. 7, A and B, Table I). Since CT4 appeared to be more efficacious than CT7 (the C-terminal half of CT4), we tested whether CT8, which corresponded to the N-terminal half of CT4 (amino acids 1905-2024, Fig. 1), had any inhibitory activity. Conceivably CT4 might be a better inhibitor because an additional inhibitory domain might be contained within the CT8 fragment. However, neither GST-CT8 (data not shown) nor 6ϫ-His-CT8 (Fig. 7, A and B, Table  I) had any effect on channel currents. Taken together, these results indicated that the inhibition of channel currents by the C terminus could be reconstituted by applying fragments as small as CT7 (corresponding to the most distal 144 amino acids of the C terminus) to cells expressing truncated ␣ 1C subunits. The more effective inhibition of currents by CT4 compared with CT7 may have been due to a more effective association of CT4 with the truncated channels (see below).
Channels Containing the ␣ 1C ⌬1733 Subunit Were Insensitive   FIG. 2. Enhanced I Ba current in ␣ 1C truncation mutants. Wild type or C-terminal-truncated ␣ 1C subunits (refer to Fig. 1) were expressed in tsA201 cells along with the ␤ 2a subunit. Barium currents were measured as described under "Experimental Procedures." Cells were maintained at a Ϫ90-mV holding potential and depolarized for 50 ms at indicated potentials (pulse applied at 10-s intervals). Currents were normalized to cell capacitance. The results shown are the means Ϯ S.E. The numbers in parentheses indicate the number of experiments performed for each construct; data were obtained from a minimum of three individual transfections. to Inhibitory Peptides-The CT7 and CT4 peptides induced functional inhibition of currents generated through ␣ 1C ⌬2024 and ␣ 1C ⌬1905. We asked whether the inhibition could be produced with an ␣ 1C subunit that was truncated farther upstream, such as the ␣ 1C ⌬1733 truncation mutant. Interestingly, application of CT4 to cells expressing the ␣ 1C ⌬1733 truncation mutant caused a modest inhibition, but this was not significantly different from control (Fig. 8, Table I). Similarly, CT7, both at the concentration (1 g/ml, data not shown) that caused inhibition of ␣ 1C ⌬2024 and ␣ 1C ⌬1905 and at a 4-fold higher concentration (Fig. 8, Table I) caused little or no inhibition of currents from ␣ 1C ⌬1733. These results suggested that the section of the C terminus between amino acids 1733 and 1905 was necessary for the inhibition caused by pipette application of either CT4 or CT7. Conceivably amino acids 1733-1905 might be the "receptor" for CT7 that allows for channel inhibition. Alternatively, CT7 might interact with another domain to cause inhibition, and amino acids 1733-1905 might play a role in helping to present CT7 to the inhibitory receptor.   As such, amino acids 1733-1905 might act to anchor or stabilize the inhibitory domain within CT7.
To test the above possibilities, we determined whether constructs containing a "tethered" CT4 or CT7 domain might exhibit the inhibited currents. Two additional constructs were made that contained internal deletions within the C terminus but intact CT4 or CT7 segments: These constructs, ␣ 1C ⌬1733-1905 and ␣ 1C ⌬1733-2024, were essentially the ␣ 1C ⌬1733 construct containing a tethered CT4 or CT7, respectively (see Fig.  1). Currents from these constructs were recorded and compared with those from ␣ 1C ⌬1733. The current-voltage relationships demonstrated that ␣ 1C ⌬1733-1905 and ␣ 1C ⌬1733-2024 each had current profiles that were similar to ␣ 1C ⌬1733, but the peak currents were drastically reduced compared with those from the ␣ 1C ⌬1733 subunit (Fig. 9, Table II). In contrast, both ␣ 1C ⌬1733-1905 and ␣ 1C ⌬1733-2024 produced peak currents that were comparable with those of the "fully inhibited" wild type ␣ 1C subunit (compare Fig. 9 with Fig. 2). These results were consistent with the idea that a tethered CT4 or CT7 could inhibit channels lacking the fragment corresponding to amino acids 1733-1905 or 1733-2024. To further test the concept that the tethered CT7 was responsible for the inhibition of the channels, we created an additional construct, ␣ 1C ⌬1733-1905⌬2024, which was similar to ␣ 1C ⌬1733-1905 but in addition contained a deletion of CT7. Thus, this construct was predicted to produce the large currents seen with other constructs lacking CT7. Indeed, the currents from ␣ 1C ⌬1733-1905⌬2024 were large and comparable to those obtained from ␣ 1C ⌬1733 and much greater that those from the parent construct ␣ 1C ⌬1733-1905 (Fig. 9, Table II). These results confirmed that the tethered CT7 in ␣ 1C ⌬1733-1905 was responsible for causing channel inhibition, as the removal of CT7 in ␣ 1C ⌬1733-1905⌬2024 gave rise to large, "uninhibited" currents. That the tethered CT4 or CT7 could cause inhibition in FIG. 7. Effect of other C-terminal peptides on channel current. Cells expressing the ␣ 1C ⌬1905 (A, B, and C) or ␣ 1C ⌬2024 (D, E, and F) subunits were presented with GST alone (A and D), GST-CT23 (B and E), or 6ϫ-His-CT8 (C and F) (refer to Fig. 1). All fusion proteins were applied at 1 g peptide/ml pipette solution, and results shown are the means Ϯ S.E. with the number of experiments shown in parentheses.

TABLE II
Peak currents from mutants containing or lacking tethered C-terminal fragments Data are peak currents from the experiments depicted in Fig. 9.

Construct
Peak currents (x Ϯ S.E.) n 52.5 Ϯ 13.0 4 constructs lacking amino acids 1733-1905 suggested that CT7 interacted with something other than these amino acids to cause inhibition of the channels. However, since CT7 could not inhibit ␣ 1C ⌬1733, which lacks these amino acids, the data are consistent with the idea that amino acids 1733-1905 play an important role in presenting CT7 to the inhibitory receptor. This concept suggests that CT7 might have multiple binding sites.
Co-immunoprecipitation of the C-terminal Fusion Proteins with the Channel Subunits-We performed biochemical studies to determine that CT4 and CT7 could associate with ␣ 1C subunits in tsA cells. First, CT4 and CT7 in the pCR3His/Myc vector were transiently transfected into tsA201 cells, and expression was ascertained by SDS-polyacrylamide gel electrophoresis and immunoblotting. Staining with the anti-Myc or CT4 antibodies revealed expression of CT4 and CT7 (see Figs. 10 and 11). Next, we tested whether the C-terminal fragments of the ␣ 1C subunit could directly associate with the channel subunits. C-terminal fragments were co-expressed with fulllength wild type ␣ 1C or ␣ 1C ⌬2024, ␣ 1C ⌬1905, ␣ 1C ⌬1733 subunits and the ␤ 2a subunits. Whole cell lysates were prepared from the transfected cells and immunoprecipitated with the CI2 antibody. When the channel subunits were co-expressed with CT4 and immunoprecipitated with the CI2 antibody, which is directed against the II-III loop of ␣ 1C and does not recognize CT4, CT4 was co-immunoprecipitated with the ␣ 1C subunits (Fig. 10). The ␣ 1C subunits in the immunoprecipitates were detected on the blot using the Card I antibody (Fig. 10,  upper panel), and the co-immunoprecipitated CT4 was detected by the anti-myc antibody (Fig. 10, lower panel). As a negative control, the CI2 antibody did not immunoprecipitate the CT4 fusion protein in the absence of the ␣ 1C subunits (Fig. 10). Surprisingly, CT4 co-immunoprecipitated not only with ␣ 1C ⌬2024 and ␣ 1C ⌬1905 but also with ␣ 1C ⌬1733 (Fig. 10), which was not inhibited by CT4 (Fig. 8). However, more CT4 was present in the immunoprecipitates with wild type ␣ 1C and ␣ 1C ⌬2024 or ␣ 1C ⌬1905 compared with those containing ␣ 1C ⌬1733. These results suggested that there might be multiple interaction sites for CT4 and that ␣ 1C ⌬1733 lacks an interaction site that is necessary for channel inhibition. The results are consistent with the concept that amino acids 1733-1905 may be critical for allowing inhibition by CT4. The observation that CT4 could bind to full-length ␣ 1C suggested that the interaction sites for CT4 were available in this protein.
In other experiments CT7 was co-expressed with either wild type ␣ 1C ␤ 2a or with the various C-terminal deletion mutants of ␣ 1C in combination with ␤ 2a subunits in tsA cells. In contrast to CT4, the CT7 fusion protein did not co-immunoprecipitate with either the wild type or the mutant ␣ 1C subunits (data not shown). However, because CT7 was effective in inhibiting channel activity but appeared less robust in causing inhibition than CT4 (Figs. 4 and 5; Table I), we reasoned that the association of CT7 might have been weaker and disrupted during the detergent solubilization and repeated washings of the immunoprecipitates. Thus, we asked if inclusion of a cross-linking agent during the immunoprecipitation might allow for detection of the association of CT7 with the channels. For these studies we used the cross-linking agent Ni(II) complex of the tripeptide NH 2 -Gly-Gly-His-COOH, which has been shown to be highly specific in that only proteins that specifically associated could be cross-linked (23). The lysates from cells expressing CT7 and mutant or wild type ␣ 1C subunits were incubated with the cross-linking agent for 10 min at room temperature, and the reactions were quenched with thiourea. When the channels were immunoprecipitated with the CI2 (data not shown) or Card I antibodies (Fig. 11), both of which are directed against the internal II-III linker of ␣ 1C and do not recognize CT7, CT7 was co-immunoprecipitated (Fig. 11). Taken together, these results demonstrated that the CT7 could directly associate with the channels in intact cells. In addition, the results supported the observations that, although the inhibi- FIG. 10. Co-immunoprecipitation of CT4 with the channel subunits. TsA201 cells were co-transfected with CT4, the rat ␤ 2a subunit, and wild type or mutant ␣ 1C subunits as indicated. Whole cell lysates were prepared from the transfected cells, and the channel subunits were immunoprecipitated with the CI2 antibody. For cells expressing CT4 alone, CT4 was concentrated by absorption onto a nickel resin. The immunoprecipitates (lanes [1][2][3][4] or concentrated lysates (lane 5) were electrophoresed on a 5-15% gradient (acrylamide) SDS gel and transferred to a filter for immunoblotting. The filter was cut in half, and the immunoprecipitated ␣ 1C subunits (wild-type or deletion mutants) were detected on the top portion of the immunoblot with the Card I antibody, whereas the co-immunoprecipitated CT4 fusion proteins were detected on the bottom portion using the anti-Myc antibody. The lane marked CT4 alone represents the concentrated CT4 from cells transfected with only this vector.
FIG. 11. Co-immunoprecipitation of CT7 with the channel subunits. TsA201 cells were co-transfected with CT7, the rat ␤ 2a subunit, and wild type or mutant ␣ 1C subunits as indicated. Whole cell lysates were prepared from the transfected cells and subjected to cross-linking with the Ni(II)-GGH complex as described under "Experimental Procedures." The lysates were diluted, and the channel subunits were immunoprecipitated with the Card I antibody. For cells expressing CT7 alone, CT7 was concentrated by absorption onto a nickel resin. Other conditions were as in Fig. 10. The Western blot was cut in half, and the immunoprecipitated ␣ 1C subunits (wild type or deletion mutants) were detected on the top portion of the immunoblot with the CI2 antibody, whereas the co-immunoprecipitated CT7 fusion proteins were detected on the bottom portion using the CT4 antibody. The lane marked CT7 alone represents the concentrated CT7 from cells transfected with only this vector. tory domain appeared to be contained within CT7, CT4 appeared to associate with the channels more effectively.
Interactions of CT7 with the C Terminus of ␣ 1C in GST Pull-down Assays-Since CT7 appeared to contain the inhibitory domain, it was of interest to identify the binding sites for CT7. GST pull-down assays were performed with bacterially expressed 6ϫ-His-tagged CT7 and GST constructs derived from the intracellular domain of ␣ 1C . CT7 bound to GST constructs of CT, CT4, CT12, CT23, and CT14 but not to CT8, NT, L1, L2, or GST alone (Fig. 12). A common site shared by CT, CT4, CT12, and CT23 is contained within amino acids 1733-1905. These data along with those obtained from the electrophysiological and co-immunoprecipitation data suggest that one interaction site for CT7 is amino acids 1733-1905. However, CT7 also bound to CT14 (amino acids 1623-1733), which is upstream of 1733-1905. This is consistent with the idea that there are multiple interaction sites for CT7. That the CT14 site is upstream of amino acid 1733 is consistent with the result that CT7 associates with ␣ 1C ⌬1733 but more poorly than to other ␣ 1C constructs that contain both the CT14 site and amino acids 1733-1905. In addition, since CT7 can bind to ␣ 1C ⌬1733 but cannot inhibit channel activity, it suggests that binding to CT14 is not sufficient for channel inhibition. It may be that binding of CT7 to amino acids 1733-1905 is important for presenting CT7 to sites that are necessary to cause inhibition, whereas in the constructs such as ␣ 1C ⌬1733-1905 and ␣ 1C ⌬1733-2024, the tethering of CT7 is sufficient to allow it to interact with the inhibitory receptor.
In summary, the present study has demonstrated that Cterminal fragments derived from the ␣ 1C subunit can associate with and regulate the activity of L-type Ca 2ϩ channels containing a truncated ␣ 1C subunit. Truncation of the ␣ 1C subunit by 147 amino acids, as in the ␣ 1C ⌬2024 subunit, was sufficient to relieve the inhibition caused by the presence of a full-length C terminus. In addition, application of the fragment corresponding to the deleted amino acids, CT7, to cells expressing channels containing an ␣ 1C subunit truncated at either position 2024 or 1905 reconstituted channel inhibition. Amino acids 1733-1905 appeared to be important to allow for this reconstituted inhibition by the CT7 or CT4 fragments, as neither CT7 nor CT4 were able to effectively inhibit channels truncated at position 1733. However, amino acids 1733-1905 did not appear to be the only receptor for CT7, as the tethered CT7 in mutants ␣ 1C ⌬1733-1905 and ␣ 1C ⌬1733-2024 appeared to be capable of inhibiting currents in the absence of amino acids 1733-1905. It is likely that these amino acids help to position CT7 to a receptor that allows for the inhibitory effects to be expressed.
Many studies demonstrate that the ␣ 1C subunit appears to be truncated at the C terminus to a protein of ϳ190 kDa in native tissues including heart and brain (10,11,25,26). In contrast, in heterologous expression systems, the ␣ 1C subunit has been found to be a full-length protein (e.g. Refs. 5, 15, and 16). It was not obvious if the truncation observed in the native systems was an artifact that occurred upon channel isolation or the result of a physiological processing event. However, earlier findings demonstrated that the C-terminal domain could be visualized in intact cardiac myocytes using an immunocytochemical approach (10). In addition, we found that exogenous chymotrypsin can cleave full-length expressed ␣ 1C subunits into a 190-kDa body that is very similar to the ϳ190-kDa fragment observed in native tissues and C-terminal fragments of 35-50 kDa (17). Interestingly, the C-terminal fragments remained associated with the membrane after cleavage (17). A proline-rich domain was identified between amino acids 1974 and 2000 and found to be important for the tethering of the C-terminal fragments to the membrane (17). Here we have presented complimentary findings that demonstrated that the C-terminal fragments associated with C-terminal-truncated channels and regulated channel activity. The domain termed CT7 was found to contain an inhibitory domain; however CT4, which contained the CT7 sequence, appeared to interact with the ␣ 1C subunit more effectively. Conceivably the difference in abilities of CT7 and CT4 to associate with the channels was due to the proline-rich domain. CT7 lacks the proline-rich domain (17), whereas CT4 contains this motif.
It appears that the C terminus of the ␣ 1C subunit is a complex and important component of this protein. The present results together with those of an earlier study (17) support the FIG. 13. Schematic depicting how distal C terminus (DCT) of the L-type Ca 2؉ channel regulates channel conductance. The C terminus of the full-length ␣ 1C subunit contains several regulatory domains within the C terminus that are identified and located schematically. Our previous work indicates that the C terminus of the majority of the full-length channel is cleaved proximal to the 1900amino acid region into both proximal and distal portions. The present data (obtained using the CT4 and CT7 peptides) suggest that the portion of the distal C terminus between amino acid 2024 and 2171 (corresponding to CT7) serves as an inhibitory domain. The efficacy for inhibition is enhanced by the peptide region 1909 -2024 (included in CT4 but absent from CT7). Inhibition mediated by the cleaved peptide inhibitory domain requires the peptide region between 1733 and 1905 on the proximal portion of the C terminus, and therefore, we suggest that this region contains a distal C terminus binding region. In this way the distal C terminus functions to inhibit and thus regulate channel conductance through and in response to an as yet undefined mechanism. Also, shown for completeness are several other known regulatory domains including the EF-hand, calmodulin binding domain (CBD) (7), the proline-rich domain (PRD) (17), and the Ser-1928 residue (the only site capable of in vivo phosphorylation) (5). hypothesis that the ␣ 1C subunit may undergo a physiologically important processing event in native systems and that the C-terminal fragments may remain associated with the channels to allow for channel regulation. A simplified version of our view of how the C terminus of the channel functions to regulate channel conductance is depicted in the schematic presented in Fig. 13. An extension of these conclusions is that if the endogenous C-terminal fragments do remain associated with the channels, one might predict that injection of the exogenous C-terminal peptides into cardiac myocytes would have little or no effect. Indeed, in preliminary experiments, we tested the effects of CT4 on L-type currents from adult rabbit cardiac myocytes and found that the exogenously applied CT4 did not inhibit their whole cell L-type Ca 2ϩ current (data not shown). Further experiments will be necessary to define the point of cleavage and the nature of the C-terminal fragments in native systems.
Other important proteins involved in signaling are known to undergo regulated proteolysis (27). Examples include the sterol regulatory element-binding proteins, the amyloid precursor protein that is linked to Alzheimer's disease, and Notch, a protein that is important in developmental signaling (27). Future studies will be required to unravel the events associated with the processing of the L-type Ca 2ϩ channels and to further understand how the processing regulates channel activity.