The PDZ Motif of the α1C Subunit Is Not Required for Surface Trafficking and Adrenergic Modulation of CaV1.2 Channel in the Heart*

Background: The mechanisms responsible for CaV1.2 regulation by the α1C C terminus are unknown. Results: Trafficking, basal function, and adrenergic modulation of CaV1.2 were not altered in cardiomyocytes of transgenic mice expressing PDZ-deleted α1C. Conclusion: PDZ-mediated interactions are not required for CaV1.2 trafficking and function in the heart. Significance: The regulation of CaV1.2 by auxiliary proteins does not depend on the PDZ ligand motif in the heart. Voltage-gated Ca2+ channels play a key role in initiating muscle excitation-contraction coupling, neurotransmitter release, gene expression, and hormone secretion. The association of CaV1.2 with a supramolecular complex impacts trafficking, localization, turnover, and, most importantly, multifaceted regulation of its function in the heart. Several studies hint at an important role for the C terminus of the α1C subunit as a hub for multidimensional regulation of CaV1.2 channel trafficking and function. Recent studies have demonstrated an important role for the four-residue PDZ binding motif at the C terminus of α1C in interacting with scaffold proteins containing PDZ domains, in the subcellular localization of CaV1.2 in neurons, and in the efficient signaling to cAMP-response element-binding protein in neurons. However, the role of the α1C PDZ ligand domain in the heart is not known. To determine whether the α1C PDZ motif is critical for CaV1.2 trafficking and function in cardiomyocytes, we generated transgenic mice with inducible expression of an N-terminal FLAG epitope-tagged dihydropyridine-resistant α1C with the PDZ motif deleted (ΔPDZ). These mice were crossed with α-myosin heavy chain reverse transcriptional transactivator transgenic mice, and the double-transgenic mice were fed doxycycline. The ΔPDZ channels expressed, trafficked to the membrane, and supported robust excitation-contraction coupling in the presence of nisoldipine, a dihydropyridine Ca2+ channel blocker, providing functional evidence that they appropriately target to dyads. The ΔPDZ Ca2+ channels were appropriately regulated by isoproterenol and forskolin. These data indicate that the α1C PDZ motif is not required for surface trafficking, localization to the dyad, or adrenergic stimulation of CaV1.2 in adult cardiomyocytes.

PDZ domains are protein interaction motifs that bind to specific C-terminal sequences of their interacting proteins. PDZs are relatively promiscuous interaction domains that may have specificity for more than one target protein. The pore-forming subunits of both Ca V 1.2 and Ca V 1.3, two subtypes of L-type Ca 2ϩ channels, contain evolutionarily conserved class 1 PDZ domain-binding C-terminal motifs (Fig. 1). To efficiently activate cAMP-response element-binding protein and gene expression in hippocampal neurons, interactions between the ␣ 1C subunit and proteins with a PDZ domain have been shown to be required (21), although the PDZ motif is not required for correct subcellular distribution or membrane expression of Ca V 1.2 in dendrites (22). The role of the PDZ ligand motif of Ca V 1.2, therefore, differs from that of Ca V 1.3, which depends on its PDZ motif for association with Shank and insertion into the postsynaptic membrane (23), and Ca V 2.2, which depends on its PDZ motif for insertion into the presynaptic space (24). The PDZ of ␣ 1C has also been shown to interact with proteins containing PDZ domains, such as enigma homolog 1 (25), which is a protein kinase D1-scaffolding protein, Cypher/ZASP (26), which is an protein kinase A anchoring protein, membrane-associated guanylate kinase-inverted proteins, Na ϩ /H ϩ exchanger regulatory factor 1/2, and neuronal nitric oxide synthase (nNOS) (27). In the heart, the PDZ-dependent association of ␣ 1C with Cypher/ZASP has been proposed to facilitate ␤-adrenergic-mediated phosphorylation of Ca V 1.2, and deletion of the ␣ 1C C-terminal PDZ motif significantly impaired PKA phosphorylation of Ser 1928 of heterologously expressed Ca V 1.2 (26).
The roles of the ␣ 1C PDZ ligand motif in modulating trafficking, E-C coupling, and ␤-adrenergic modulation of Ca 2ϩ current has not been directly tested in the heart. We determined the role of the PDZ motif in native cardiomyocytes by creating a transgenic mouse expressing an ␣ 1C subunit harboring a deletion of the C-terminal PDZ ligand motif.

EXPERIMENTAL PROCEDURES
Reagents-Nisoldipine (Santa Cruz Biotechnology) was dissolved weakly at a concentration of 30 mM in ethanol, protected from light, and diluted with ethanol on the day of the experiment to 3 mM. The final dilution of nisoldipine to 300 nM was in the extracellular recording solution. All other chemicals were acquired from Sigma.
Animals-The pseudo-WT (pWT) ␣ 1C and the ⌬PDZ (C-terminal four amino acid residues deleted) constructs were generated by fusing the rabbit Cacna1c cDNA (accession no. X15539) to the clone 26 vector containing the modified murine ␣-myosin heavy chain (MHC), tetracycline-inducible promoter ("responder" line) vector (a gift from Drs. Jeffrey Robbins and Jeffrey Molkentin) (28,29). The ␣ 1C subunit was engineered to be dihydropyridine (DHP)-insensitive with the substitutions T1066Y and Q1070M (30,31), and a 3ϫ FLAG epitope was ligated in-frame to the N terminus of ␣ 1C . Transgenic founder mice were identified with genomic DNA utilizing polymerase chain reactions using the following PCR primers: forward within clone 26 vector, CTT CCA GCC CTC TCT TTC TC; reverse ␣ 1C , CAG CTG CGT TGG CAT TCA TGT TG.
Transgenic-positive mice were bred with cardiac-specific (␣MHC), doxycycline-regulated, codon-optimized reverse transcriptional transactivator (rtTA) mice (obtained via Mutant Mouse Regional Resource Centers (MMRRC)) (32) to generate double-transgenic mice. In addition to the set of PCR primers above, mice carrying both transgenes were selected using the following rtTA PCR primers: forward rtTA, GTG ATT AAC AGC GCA CTG GAG; reverse rtTA, CAA ACA GTT CGA TAG CTT GCC G. Additionally, founder lines were selected on the basis of their lack of transgenic ␣ 1C expression in the absence of doxycycline. Mice were fed food impregnated with 0.2 g/kg doxycycline to induce expression (Bio Serv, catalog no. S3888) for 1-2 days. The Institutional Animal Care and Use Committee at Columbia University approved all animal experiments.
Immunoblots and Immunofluorescence-Cardiomyocytes were isolated (33) from 8-to 12-week-old non-transgenic and doxycycline-fed transgenic mice. Cardiomyocytes were homogenized in a 1% Triton X-100 buffer containing 50 mM Tris-HCl (pH7.4) 150 mM NaCl, 10 mM EDTA, 10 mM EGTA, and protease inhibitors. The lysates were incubated on ice for 30 min, centrifuged at 14,000 rpm at 4°C for 10 min, and supernatants were collected. Proteins were size-fractionated on SDS-PAGE, transferred to nitrocellulose membranes, and probed with HRP-conjugated anti-FLAG (Sigma) antibody and anti-␣ 1C and HRP-conjugated goat anti-rabbit antibodies. Detection was performed with a charge-coupled device camera (Carestream Imaging). Loading normalization was performed with anti-tubulin antibody (Santa Cruz Biotechnology). For immunofluorescence, isolated cardiomyocytes were fixed for 15 min in 4% paraformaldehyde. Indirect immunofluorescence was performed using a 1:200 rabbit anti-FLAG antibody (Sigma) and 1:200 FITC-labeled goat anti-rabbit antibody (Sigma). Images were acquired using a confocal microscope.
Cellular Electrophysiology-The isolated cardiomyocytes were superfused with 140 mM tetraethylammonium-Cl, 1.8 mM CaCl 2 , 1 mM MgCl 2 , 10 mM glucose, and 10 mM HEPES, adjusted to pH 7.4 with CsOH. All experiments were performed at room temperature, ϳ22°C. Membrane currents were measured by whole-cell patch clamp method using a MultiClamp 700B amplifier (Axon Instruments). The pipette solution contained (in mM) 135 mM CsCI, 10 mM EGTA, 1 mM MgCl 2 , 2 mM magnesium-ATP, 2.0 mM CaCl 2 , and 10 mM HEPES, adjusted to pH 7.2 with CsOH. Pipette series resistances were usually Ͻ1 M⍀ after 60% compensation. Leak currents and capacitance transients were subtracted by a P/4 protocol. To measure Ca 2ϩ peak currents, the cell membrane potential was held at Ϫ70 mV and stepped to ϩ10 mV for 350 ms every 10 s. To evaluate the current-voltage relationship for Ca 2ϩ currents, the same protocol was repeated with steps between Ϫ50 mV to ϩ50 mV in 10-mV increments.

JOURNAL OF BIOLOGICAL CHEMISTRY 2167
Statistical Analysis-Results are presented as mean Ϯ S.E. For comparisons between two groups, unpaired Student's t test was used. Statistical analyses were performed using Prism 6 (GraphPad Software). For multiple group comparisons, a oneway analysis of variance followed by Sidek post hoc test was performed using Prism 6. Differences were considered statistically significant at values of p Ͻ 0.05.

RESULTS
Generation of Inducible, Cardiac-specific ⌬PDZ a 1C Transgenic Mice-A class I PDZ domain-binding motif VSXL is present in the cardiac/neuronal ␣ 1C subunit, conforming to the consensus sequence X[T/S]X COOH , where X is any residue and is a hydrophobe (34). The PDZ domain-binding motif is conserved across many species, including human, rabbit, rodents, and zebrafish and in the Caenorhabditis elegans L-type Ca 2ϩ channel subunit (egl-19) (23) (Fig. 1). The role for the PDZ ligand domain in the heart is unknown. Cultured neurons transfected, using a Ca 2ϩ phosphate technique, with ␣ 1C lacking the PDZ ligand domain exhibited attenuated cAMP-response element-binding protein responses, indicating an important role for the PDZ ligand motif in the subcellular localization of ␣ 1C in neurons (21). Although adenoviruses have been used to express Ca V 1.2 subunits in cardiomyocytes (35,36), creation of adenoviruses encoding ␣ 1C is difficult because of the ␣ 1C insert size and because viral infection requires that cardiomyocytes be cultured for extended periods. Although useful for investigating biophysical properties, many relevant aspects of Ca V 1.2 channel targeting and functional modulation in heart cannot be studied in heterologous expression systems, which lack the complex cytoarchitecture and intracellular environment of adult cardiomyocytes.
To circumvent these problems, we generated transgenic mice featuring inducible, cardiac-specific expression of DHPresistant, FLAG epitope-tagged ␣ 1C (Fig. 2A). This approach preserves the hormonal regulation of Ca V 1.2 by limiting ␣ 1C overexpression, is relatively rapid, and enables functional screening of ␣ 1C mutants in cardiomyocytes freshly isolated from adult mice. We previously generated transgenic mice with inducible expression of DHP-resistant (T1066Y/Q1070M), (37). These transgenic mice were crossed with ␣-MHC rtTA transgenic mice. To determine the importance of the PDZ domain-binding site in ␣ 1C in cardiomyocytes, we generated a transgenic mouse line expressing ␣ 1C with a deletion of the four C-terminal amino acid residues in the background of an N-terminal 3ϫ FLAG epitope tag and DHP resistance (⌬PDZ). These transgenic mice were crossed with ␣-MHC-rtTA, and doubletransgenic mice were identified by PCR (Fig. 2B). Several ⌬PDZ-␣ 1C founder transgenic lines were originally generated, and all lines demonstrated doxycycline-induced ␣ 1C expression after crossing with the ␣MHC-rtTA mice. The results were consistent across all founder lines and gender, and therefore were pooled.
In cardiomyocytes isolated from non-transgenic mice (C57Bl/6), native ␣ 1C was detected as a full-length, ϳ240 kDa band and a cleaved, ϳ210 kDa band using an anti-␣ 1C antibody created against an internal epitope within the intracellular loop of domains II and III. Native ␣ 1C in non-transgenic mice cannot be detected using an anti-FLAG antibody (Fig. 2C). Both the pWT ␣ 1C transgenic channels and the transgenic channels were detected using anti-FLAG antibody. The ratios of cleaved to full-length pWT and ⌬PDZ transgenic ␣ 1C were not significantly different, implying that the PDZ motif is not required for cleavage of the ␣ 1C distal C terminus.
Confirming the expression of the transgene, immunofluorescence staining of fixed cardiomyocytes from pWT and ⌬PDZ mutant transgenic mice with an anti-FLAG antibody showed a membrane distribution of expressed ␣ 1C subunits consistent with t-tubular localization (Fig. 2D). No staining was detected in cardiomyocytes when the anti-FLAG antibody was omitted.
The PDZ Motif Is Not Required for Trafficking to the Dyad or Initiating E-C Coupling-Cardiomyocyte contraction requires Ca 2ϩ influx via Ca V 1.2, which triggers sarcoplasmic reticulum Ca 2ϩ release. We assessed the localization of the transgenic channels in the dyad by determining whether E-C coupling could be maintained in the presence of nisoldipine. In control non-transgenic cardiomyocytes, 300 nM nisoldipine eliminated contraction to electric field stimulation at 1 Hz (Fig. 3, A and D). By contrast, in cardiomyocytes from both the pWT-␣ 1C and the ⌬PDZ-␣ 1C transgenic mice, the effect of nisoldipine was greatly diminished, and E-C coupling was preserved (Fig. 3, B-D). Taken together, these results demonstrate that the PDZ ligand motif is not required for the surface expression and subcellular localization of Ca V 1.2 to the dyad. Furthermore, the PDZ ligand motif is not required for the initiation of E-C coupling in mice.
Functional, Inducible Expression of ⌬PDZ-␣ 1C in Cardiomyocytes-To measure transgenic channel selectivity, we chose a concentration of 300 nM nisoldipine as optimal because nisoldipine (300 nM) blocked Ͼ98% of heterologously expressed WT Ca V 1.2 current in tsA-201 cells but only blocked 34.6 ϩ 2.5% of DHP-insensitive ␣ 1C (37). We measured Ca V 1.2 currents in adult cardiomyocytes from non-transgenic and transgenic mice. Nisoldipine (300 nM) inhibited 92.4% Ϯ 1.6% of endogenous peak Ca 2ϩ current in cardiomyocytes isolated from nontransgenic mice (n ϭ 12) (Fig. 4, A and F) but 63.4 Ϯ 4.7% of peak current in cardiomyocytes isolated from doxycycline-fed pWT-␣ 1C transgenic mice (n ϭ 30) (Fig. 4, B, D, and F) and 64.4% Ϯ 2.8% of peak current in cardiomyocytes isolated from doxycycline-fed ⌬PDZ mutant transgenic mice (n ϭ 25 cardiomyocytes, p Յ 0.001) (Fig. 4, C, E, and F). The voltage dependence of Ca V 1.2 activation for ⌬PDZ ␣ 1C was not different from the endogenous channels because the current-voltage curves in the absence and presence of nisoldipine were identical. These findings imply that, at least under basal conditions, the modulation of the transgenic, PDZ-deleted Ca V 1.2 channels by accessory proteins was similar to endogenous Ca V 1.2 channels.

Role of PDZ Motif in the Adrenergic Modulation of Ca V 1.2-
The PDZ-dependent association of ␣ 1C with Cypher/ZASP, functioning as an protein kinase A anchoring protein (AKAP) has been proposed recently to facilitate ␤-adrenergic-mediated phosphorylation of Ca V 1.2. Deletion of the ␣ 1C C-terminal PDZ motif, preventing association of Ca V 1.2 and Cypher/

The Role of the Ca V 1.2 PDZ Motif in the Heart
ZASP, significantly impaired isoproterenol-stimulated PKA phosphorylation of Ser 1928 of heterologously expressed Ca V 1.2 (26). The role of the PDZ ligand-binding motif in regulating the functional modulation of Ca V 1.2 by the adrenergic system was not assessed.
Freshly isolated cardiomyocytes were isolated from doxycycline-treated ⌬PDZ transgenic mice. In the presence of nisoldipine, isoproterenol increased peak Ca V 1.2 current by a mean of 1.8 Ϯ 0.1-fold, identical to the isoproterenol-induced augmentation of current in pWT ␣ 1C transgenic cardiomyocytes (p ϭ not significant, pWT␣ 1C versus ⌬PDZ) (Fig. 5, A-C and F).
In the presence of nisoldipine, forskolin increased peak Ca V 1.2 current by a mean of 1.7 Ϯ 0.04-fold increase in cardiomyocytes isolated from the ⌬PDZ mice, nearly identical to the 1.8 Ϯ 0.1-fold increase in pWT ␣ 1C cardiomyocytes (Fig. 5, D-F). Taken together, these results demonstrate that the Ca V 1.2 PDZ ligand motif is not required for surface expression, dyadic localization, E-C coupling, or adrenergic modulation of Ca 2ϩ currents in cardiomyocytes.

DISCUSSION
The current understanding regarding mechanisms underlying Ca V 1.2 trafficking and modulation derives from studies of recombinant channels reconstituted in heterologous cells. An important limitation is that heterologous cells lack the complex cytoarchitecture and intracellular milieu of adult cardiomyocytes. To address this, we developed an approach that utilizes transgenic mice expressing doxycycline-inducible, cardiacspecific, DHP-resistant ␣ 1C . Importantly, the transgenic wildtype channels are transported appropriately to the dyad, can initiate E-C coupling, demonstrate normal activation and inactivation properties, and are fully and appropriately regulated by ␤-adrenergic stimulation. Compared with knockin mouse models, this approach is both cost-effective and rapid and, perhaps more importantly, enables the induced brief expression of mutant channels in adults, permitting the comparison of WT and mutant ␣ 1C structure-function mechanisms in the absence of developmental abnormalities and heart failure. This is not possible in knockin mouse models, which display embryonic lethality (such as ␣ 1C 1904 and ␣ 1C 1796 knockin mice) (17,18). Importantly, this approach avoids the need to culture cardiomyocytes, which may cause dedifferentiation, and uses an inducible expression system so that we can carefully titrate the level of expression.
The most important potential limitation of this approach is that the transgenic Ca 2ϩ channels may compete with endogenous channels for limited binding partners, therefore altering the stoichiometry of the channels. Our experiments were designed to limit overexpression by exposing the animals to doxycycline for only 1-2 days and selecting for transgenic founder lines with relatively low expression. The findings that the trafficking and function of the pWT ␣ 1C transgenic channels were normal suggest that limited overexpression does not adversely affect the normal characteristics of the channels. The trafficking and function of the ⌬PDZ mutant channels were similar to both endogenous and pWT ␣ 1C channels, demonstrating that the PDZ motif of ␣ 1C is not required in adult cardiomyocytes.
The distal C terminus of the ␣ 1C subunit regulates Ca V 1.2 trafficking and function in the heart, but the mechanisms and determinants are unknown. The deficits in Ca V 1.2 trafficking and ␤-adrenergic regulation seen in ␣ 1C 1904 (17) and ␣ 1C 1796 knockin mice could be due to multiple factors, including deficits in macromolecular complex formation (17,18). In vitro binding assays demonstrated that the ␣ 1C PDZ ligand motif was required for interactions between Ca V 1.2 and several PDZ domain proteins expressed in cardiomyocytes, such as NHERF1, MAGI-3, and nNOS (27). This sequence, or a nearly identical sequence, is present in human, monkey, mouse, rat, marmoset, rabbit, and zebrafish ␣ 1C . On the basis of the postulated role of the PDZ motif of ␣ 1C in regulating cAMP-response element-binding protein signaling in neurons, potentially via its effects on subcellular localization (21) (35) and in an ␣ 1C S1928A knockin mouse (46). The role for Ca 2ϩ channels in mediating cAMP-response element-binding protein signaling in the heart has not been delineated. PDZ-domain proteins are important modulators of K ϩ (47, 48) and Na ϩ channel (49 -51) function and ␤-adrenergic signaling (52) in the heart. PDZ domain interactions are regulated and ⌬PDZ (E) Ca V 1.2 acquired before (black trace) and 3 min after superfusion of 300 nM nisoldipine. Insets, series of whole-cell Ca V 1.2 currents recorded from a series of pulses between Ϫ40 mV and ϩ 60 mV from a holding potential of Ϫ70 mV in the absence of nisoldipine (black traces) and 3 min after 300 nM nisoldipine (red traces). pF, picofarad. F, fraction of DHP-resistant current density. Data are mean ϩ S.E. ***, p Ͻ 0.001 by Student's t test, analysis of variance, and Sidak post hoc test.

The Role of the Ca V 1.2 PDZ Motif in the Heart
by phosphorylation and facilitate intracomplex phosphorylation. The adaptor SAP97 facilitates the phosphorylation of K V 4.2 and K V 4.3 by Ca 2ϩ -calmodulin kinase II (47) and of the ␤1-adrenergic receptor by PKA (53). In the heart, both Na V 1.5 and Kir2.1 interact independently with PDZ domain proteins, regulating the surface expression and function of the channels (54 -56). The expression of Na V 1.5 also affects the turnover and regulation of Kir2.1 channels in the heart via PDZ-dependent interactions (55). Mice with deletion of the Na V 1.5 PDZ motif displayed reduced Na V 1.5 expression and current, specifically at the lateral myocyte membrane, whereas expression and function at the intercalated disks were not altered (51). Ca V 1.3, which is expressed in neurons and in cardiac sinoatrial and atrioventricular nodes and atria, but not in ventricles (57), also contains a C-terminal PDZ ligand motif and binds erbin (58) and shank, both neuronal PDZ domain proteins (23,59).
The most direct approach to determine the functional role of the PDZ domain proteins in modulating Ca V 1.2 is to express a PDZ-deleted ␣ 1C in cardiomyocytes. This approach avoids potential nonspecific effects of deleting the PDZ-binding domain proteins instead because they also interact with other ion channels and ␤-adrenergic receptors. Using this approach, we found that deletion of the ␣ 1C PDZ ligand motif does not alter Ca V 1.2 trafficking, basal function, or modulation by the sympathetic nervous system in adult cardiomyocytes. We cannot exclude the possibility that the ␣ 1C PDZ ligand motif may be responsible for functions during cardiac development or may have a role in other tissues. An advantage of our approach is that the short-term induction of expression minimizes the compensatory effects that can be observed in more traditional transgenic expression or knockin mice.
Taken together, we show that the ␣ 1C PDZ ligand motif is not required for trafficking to the surface or dyads, basal function, and adrenergic modulation in the heart. The mechanisms responsible for how the distal C terminus regulates Ca V 1.2 trafficking and function in cardiomyocytes have yet to be discovered.