Ankyrin-B Regulates Ca_v2.1 and Ca_v2.2 Channel Expression and Targeting

Background: Calcium channels control membrane excitability. The mechanisms underlying Ca_v2.1/Ca_v2.2 targeting are not well understood.

Results: Ankyrin-B associates with Ca_v2.1/Ca_v2.2. Loss of ankyrin-B results in reduced expression of Ca_v2.1/Ca_v2.2 in select brain regions.

Conclusion: Ankyrin-B plays a role in the expression of Ca_v2.1/Ca_v2.2.

Significance: Results identify pathway for membrane targeting of calcium channels and regulation of membrane excitability.

Abstract

N-type and P/Q-type calcium channels are documented players in the regulation of synaptic function; however, the mechanisms underlying their expression and cellular targeting are poorly understood. Ankyrin polypeptides are essential for normal integral membrane protein expression in a number of cell types, including neurons, cardiomyocytes, epithelia, secretory cells, and erythrocytes. Ankyrin dysfunction has been linked to defects in integral protein expression, abnormal cellular function, and disease. Here, we demonstrate that ankyrin-B associates with Ca_v2.1 and Ca_v2.2 in cortex, cerebellum, and brain stem. Additionally, using in vitro and in vivo techniques, we demonstrate that ankyrin-B, via its membrane-binding domain, associates with a highly conserved motif in the DII/III loop domain of Ca_v2.1 and Ca_v2.2. Further, we demonstrate that this domain is necessary for proper targeting of Ca_v2.1 and Ca_v2.2 in a heterologous system. Finally, we demonstrate that mutation of a single conserved tyrosine residue in the ankyrin-binding motif of both Ca_v2.1 (Y797E) and Ca_v2.2 (Y788E) results in loss of association with ankyrin-B in vitro and in vivo. Collectively, our findings identify an interaction between ankyrin-B and both Ca_v2.1 and Ca_v2.2 at the amino acid level that is necessary for proper Ca_v2.1 and Ca_v2.2 targeting in vivo.

Introduction

Neurons are highly polarized cells, where structural, functional, and molecular differences within the neuronal physiology underlie the ability to receive, process, and transmit information. The specific expression profile of ion channels, receptors, and cytoskeletal proteins present within select neuronal membrane domains plays a primary role in defining local function of the entire cell. Recently, studies focused on the role of cytoskeletal proteins in the genesis and maintenance of local membrane domains in excitable cells have revealed a new and exciting avenue of study for these proteins in physiology and disease (1, 2).

Calcium ions are critical second messengers involved in the regulation of multiple neuronal cellular processes, but primarily synaptic transmission (3–7). Ca²⁺ entry across the plasma membrane occurs via voltage-dependent calcium channels (VDCCs).² To maintain the tight spatio-temporal distribution of intracellular Ca²⁺ necessary to mediate a diverse array of functions, VDCCs have evolved with respect to their biophysical properties, regulation, and subcellular localization (8). The physiological importance of VDCCs is further exemplified by the variety of mutations in VDCC channel components linked to neurological and motor diseases (9). VDCCs have been classified into five groups, identified as T, L, N, P/Q, and R, based on electrophysiological and pharmacological properties (9–12). N- and P/Q-type currents are observed primarily in neurons and co-localize with a subset of docked vesicles at the synapse, where they control exocytosis (13–15) and initiate synaptic neurotransmission (3, 4, 16). VDCCs are heteromultimers composed of the pore-forming α1 subunit and associated auxiliary subunits, Ca_vβ and α2δ (for review, see Ref. 4). Although auxiliary subunits have been demonstrated to enhance α1B expression (17, 18), the mechanisms underlying Ca_v channel membrane expression and localization are not clearly defined.

Ankyrins are a class of membrane adaptor proteins necessary for the targeting and regulation of select membrane and cytoplasmic proteins (19). In the brain, ankyrin-G is a necessary component for the functional organization of initial segments and nodes of Ranvier. Targeted disruption of cerebellar ankyrin-G results in severe ataxia, loss of ability to fire action potentials, and loss of Na_v1.6 from the initial segment (20, 21). Additionally, there is loss of nodal β_IV spectrin and KCNQ2/3 channels with indiscriminate localization of the cell adhesion molecule neurofascin (21, 22). Loss of neurofascin expression at the initial segment of Purkinje neurons results in disruption of interneuron synapses connecting Purkinje neurons in the cerebellum (23). Likewise, ankyrin-B has been demonstrated to play a critical role in the establishment and maintenance of excitable domains in brain (24). Loss of ankyrin-B in mice results in significant nervous system defects, including hypoplasia of the corpus callosum and pyramidal tracts, dilation of the lateral ventricles, and deterioration of long axon tracts (25). Notably, recent work outside of the brain has defined new roles for ankyrin-B for targeting select voltage-gated Ca²⁺ channels in the pacemaking cells of the cardiac sinoatrial node (26, 27). We therefore tested the role of ankyrin-B for targeting select calcium channels in the brain.

Our findings define a role for ankyrin-B in neuronal cell biology by demonstrating ankyrin-B-dependent interactions that are required for proper Ca_v2.1 and Ca_v2.2 expression in brain. We observed a significant decrease in Ca_v2.1 and Ca_v2.2 expression in cortex, cerebellum, and brain stem tissues of ankyrin-B-deficient mice. In defining the structural requirement for this interaction, we demonstrate that the ankyrin-B membrane-binding domain associates directly with Ca_v2.1 and Ca_v2.2. Specifically, this interaction occurs in the DII/III loop region of both VDCCs. Finally, a single amino acid mutation at residue Tyr-797 in Ca_v2.1 and Tyr-788 in Ca_v2.2 is sufficient to disrupt association of ankyrin-B with Ca_v2.1 or Ca_v2.2. Collectively, our data demonstrate a necessary role for ankyrin-B in the expression of Ca_v2.1 and Ca_v2.2 in brain.

EXPERIMENTAL PROCEDURES

Animals

Mice were age-matched wild-type, ankyrin-B^+/−, and ankyrin-G cerebellar-specific knock-out littermates. Animals were backcrossed at least 20 generations (99.8% pure) into the C57BL/6 background (Jackson Laboratories). Ankyrin-G cerebellar-specific knock-out mice were generated as described (25). Animals were handled according to approved protocols and animal welfare regulations of the Institutional Animal Care and Use Committee at Ohio State University. All mice were housed and handled identically.

Antibodies

The following primary antibodies were used for immunoblotting and/or immunostaining: anti-ankyrin-B (polyclonal), anti-ankyrin-G (polyclonal), anti-Ca_v2.2 (polyclonal; Thermo Scientific and Alomone Labs), anti-Ca_v2.1 (polyclonal; Alomone Labs); anti-GAPDH (monoclonal; Fitzgerald Industries International); anti-GFP (NeuroMabs). Secondary antibodies used for immunoblotting were HRP-conjugated donkey anti-rabbit (Jackson Laboratories) and HRP-conjugated donkey anti-mouse (Jackson Laboratories). Secondary antibody used for immunostaining were Alexa Fluor 568 goat anti-rabbit Ig (Invitrogen). Ca_v2.1 and Ca_v2.2 antibodies were validated using immunoblots from HEK293 cells transfected with either GFP-Ca_v2.1 or GFP-Ca_v2.2. Ca_v2.1 Ig specifically recognized GFP-Ca_v2.1, but not GFP-Ca_v2.2. Conversely, Ca_v2.2 Ig specifically recognized GFP-Ca_v2.2, but not GFP-Ca_v2.1.

Ca_v2.1 and Ca_v2.2 Constructs

Primers were designed to clone amino acids 708–1194 (Ca_v2.1 DII/DIII loop), 788–1194, and 808–1194 from rat full-length Ca_v2.1 cDNA based on the sequence obtained from GenBank (NM_012918.3). To facilitate subcloning into pcDNA3.1⁺, primers were designed to incorporate 5′ EcoRI and 3′ XhoI restriction sites. To facilitate subcloning into pEGFP-C3, primers were designed to incorporate 5′ EcoRI and 3′ BamHI restriction sites. Full-length Ca_v2.1 constructs were generated from rat Ca_v2.1. To facilitate subcloning into pAcGFP1-N2, primers were designed to incorporate 5′ HindIII and 3′ SalI restriction sites. Primers were designed to clone amino acids 710–1139 (Ca_v2.2 DII/III loop), 779–1139, and 799–1139 from a murine brain cDNA library based on the sequence obtained from GenBank (NM_001042528). To facilitate subcloning into pcDNA3.1⁺ (Invitrogen), primers were designed to incorporate 5′ EcoRI and 3′ XhoI restriction sites. To facilitate subcloning into pEGFP-C3 (Clontech), primers were designed to incorporate 5′ EcoRI and 3′ BamHI restriction sites. Full-length Ca_v2.2 constructs were generated from rat Ca_v2.2. To facilitate subcloning into pAcGFP1-N2, primers were designed to incorporate 5′ SalI and 3′ KpnI restriction sites. All sequences were verified prior to experiments.

Site-directed Mutagenesis

Primers designed to insert the Y788E mutation into the Ca_v2.2 DII/III loop in pcDNA3.1⁺ and pEGFP-C3 and the full-length Ca_v2.2 in pAcGFP1-N2 or the Y797E in Ca_v2.1 DII/III loop in pcDNA3.1⁺ and pEGFPC3 and the full-length Ca_v2.1 in pAcGFP1-N2 were used in concert with the Stratagene QuikChange Site-directed Mutagenesis kit and manufacturer's instructions. All sequences were verified prior to experiments.

In Vitro Translation/Binding

Ca_v2.1 and Ca_v2.2 constructs were in vitro translated using rabbit reticulocyte lysate, [³⁵S]methionine (20 μCi of Redivue l-[³⁵S]methionine; GE Healthcare), T7 polymerase, and 0.63 ng of plasmid DNA (TNT Coupled Rabbit Reticulocyte Lysate System; Promega). For binding experiments, 20 μg of purified GST, ankyrin-B membrane-binding domain (MBD)-GST, ankyrin-B spectrin-binding domain (SBD)-GST, or ankyrin-B C-terminal domain (CTD)-GST was coupled to glutathione-Sepharose for 2 h at 4 °C in IV Binding Buffer (50 mm Tris-HCl (pH 7.35), 1 mm EDTA, 1 mm EGTA, 150 mm NaCl, 0.1% Triton X-100). Following extensive washing in IV Wash Buffer (50 mm Tris-HCl (pH 7.35), 1 mm EDTA, 1 mm EGTA, 350 mm NaCl, 0.1% Triton X-100), conjugated beads were incubated with Ca_v2.1 or Ca_v2.2 in vitro translated products overnight at 4 °C in IV Wash Buffer. The beads were washed five times in IV Wash Buffer, eluted, and separated by SDS-PAGE. Gels were stained with Coomassie Blue to show the presence of GST proteins before drying the gel (Bio-Rad Laboratories). Radiolabeled proteins were detected by phosphorimaging or standard autoradiography.

Tissue Preparation and Homogenization

For immunoblot and co-immunoprecipitation (co-IP) analysis, brain tissues (cortex, cerebellum, and brain stem) were flash-frozen in liquid nitrogen and ground into a fine powder. The powder was resuspended in 2 volumes of ice-cold homogenization buffer (50 mm Tris-HCl (pH 7.35), 10 mm NaCl, 0.32 m sucrose, 5 mm EDTA, 2.5 mm EGTA, 1 mm PMSF, 1 mm AEBSF, 10 μg/ml leupeptin, and 10 μg/ml pepstatin) and homogenized using a hand-held homogenizer (27, 28). The homogenate was centrifuged at 1000 × g at 4 °C to remove nuclei. Triton X-100 and deoxycholate were added to the postnuclear supernatant for final concentrations of 0.75% Triton X-100 and 1% deoxycholate. The lysate was pelleted at high speed for 15 min at 4 °C. The resulting supernatant was quantitated by bicinchoninic acid assay prior to analysis.

Immunoblots

Immunoblots from anti-ankyrin-B, anti-Ca_v2.1, anti-Ca_v2.2, and anti-ankyrin-G blots were evaluated by densitometry and expression normalized to anti-GAPDH blots (29, 30). Histograms represent expression as a percentage of wild-type expression (wild-type expression normalized to 100%).

Tissue Preparation for Immunostaining

Freshly extracted tissues from wild-type and ankyrin-B^+/− mice were fixed in 4% paraformaldehyde for 12 h and the tissues transferred to 10, 20, and 30% sucrose solutions for 12 h each. Tissues were cryosectioned to 10-μm thickness. Cryosections were rehydrated in PBS prior to preblocking in 3 mg/ml BSA in PBS. Primary antibodies were made in a vehicle of 3 mg/ml BSA with 0.1% Triton X-100 in PBS and incubated on sections overnight at 4 °C. The slides were washed three times in vehicle and incubated with Alexa Fluor donkey anti-rabbit 568 secondary antibodies for 3 h at 4 °C. Following three washes in vehicle, the slides were mounted with VectaShield (Vector Laboratories) and no. 1 coverslips. Images were collected on a Zeiss 510 Meta confocal microscopy using Carl Zeiss software.

Co-IP from Brain Lysates

Protein A-conjugated agarose beads (AffiGel; Bio-Rad) were incubated with either control Ig or anti-Ca_v2.2 Ig, anti-Ca_v2.1 Ig, or anti-ankyrin-B Ig in co-IP binding buffer (PBS with 0.1% Triton X-100 and protease inhibitor mixture (Sigma)) for 12 h at 4 °C. Beads were centrifuged and washed three times in ice-cold PBS. Wild-type cortex, cerebellum, or brain stem tissue lysate were added to the washed beads, along with protease inhibitor mixture and co-IP binding buffer, and incubated for 12 h at 4 °C. The reactions were washed three times in ice-cold co-IP buffer. The samples were eluted and the proteins separated by SDS-PAGE prior to immunoblots with ankyrin-B, Ca_v2.1, or Ca_v2.2 Ig. Experiments were performed multiple times with similar results. Due to the low copy number of Ca_v2.1 and Ca_v2.2 in brain regions, lysate inputs were scaled up. For experiments where Ca_v2.1 or Ca_v2.2 Igs were immobilized on beads, 1 mg of cortex and cerebellum lysate was used, whereas 2 mg of brain stem lysate was used. Input lanes represent 10% of total lysate used. For experiments where ankyrin-B Ig was immobilized on beads, 1 mg of lysate for each brain region was utilized. Input lanes represent 5% of total lysate for cortex and cerebellum and 10% of total lysate for brain stem.

Co-IP from Transfected Cells

Protein A-conjugated agarose beads were incubated with either control IgG or affinity-purified anti-ankyrin-B Ig for 12 h at 4 °C. Beads were centrifuged and washed three times in ice-cold PBS. 100 μg of transfected HEK lysate was added to the washed beads, along with protease inhibitor mixture and IP buffer, and incubated for 12 h at 4 °C. The reactions were washed three times in ice-cold IP buffer. The samples were eluted and the proteins separated by SDS-PAGE prior to immunoblots with anti-GFP Ig. For co-IP using full-length calcium channel-transfected cells, input lanes represent 20% of total lysate input (400 μg). For GFP-Ca_v channel loop transfections, input lanes represent 10% of total lysate input (200 μg).

Pulldowns

One hundred micrograms of wild-type cortex, cerebellum, or brain stem lysate was added to 10 μg of GST, GST-ankyrin-B MBD, GST-ankyrin-B SBD, or GST-ankyrin-B CTD-conjugated beads, along with 500 μl of Binding Buffer, protease inhibitor mixture (Sigma), and PMSF. The reactions incubated for 12 h at 4 °C. The beads were centrifuged and washed three times in binding buffer. The proteins were eluted, separated by SDS-PAGE, and immunoblotted using anti-Ca_v2.1 Ig and anti-Ca_v2.2 Ig.

HEK293 Cell Culture and Transfection of Ca_v2.1 and Ca_v2.2 Fragments

HEK293 cells were maintained in Dulbecco's modified essential medium (Invitrogen) supplemented with 10% fetal bovine serum (FBS; HyClone) and 0.1% penicillin/streptomycin. Cells were cultured at 37 °C in 5% CO₂. Cultured cells were split 24 h prior to transfection and plated to obtain 30% confluence at time of transfection. Effectene reagent (Qiagen) was used to transfect cells with 0.2 μg Ca_v2.2 DII/III pEGFP-C3, Ca_v2.2 DII/III Y788E pEGFP-C3, or pEGFP-C3 alone. Transfection was carried out for 7 h at 37 °C and the cells recovered overnight. After 48 h, the cells were fixed with 2% paraformaldehyde, stained with ToPro (to discriminate nuclear material), and evaluated by confocal microscopy. Likewise, transfections using Ca_v2.1 DII/III pEGFP-C3, Ca_v2.1 DII/III Y797E pEGFP-C3, and pEGFP-C3 alone were also performed using the same protocol.

HEK293 Cell Culture and Transfection for Confocal Microscopy

HEK293 cells were maintained in Dulbecco's modified essential medium supplemented with 10% FBS and 0.1% penicillin/streptomycin. Cells were cultured at 37 °C in 5% CO₂. Cultured cells were split 24 h prior to transfection and plated on MatTek plates to obtain 90% confluence at time of transfection. Lipofectamine (Invitrogen) was used to transfect cells with 1 μg of Ca_v2.2-GFP, 1 μg of Ca_v2.2 Y788E-GFP, 1 μg of Ca_v2.1-GFP, 1 μg of Ca_v2.1 Y797E-GFP, or pAcGFP1-N2 alone. Transfection was carried out for 5 h at 37 °C and the cells recovered overnight. After 24 h, the cells were fixed with 2% paraformaldehyde, mounted with VectaShield (with DAPI), and evaluated by confocal microscopy.

HEK293 Cell Culture and Transfection for Co-immunoprecipitations

HEK293 cells were maintained in Dulbecco's modified essential medium supplemented with 10% FBS and 0.1% penicillin/streptomycin. Cells were cultured at 37 °C in 5% CO₂. Cultured cells were split 24 h prior to transfection and plated on 100-mm² plates to obtain 90% confluence at time of transfection. Lipofectamine (Invitrogen) was used to transfect cells with 1 μg of Ca_v2.2-GFP, 1 μg of Ca_v2.2 Y788E-GFP, 1 μg of Ca_v2.1-GFP, 1 μg of Ca_v2.1 Y797E-GFP, or pAcGFP1-N2 alone. Untransfected cells were used as a negative control. Transfection was carried out for 5 h at 37 °C and the cells recovered overnight. After 24 h, the cells were lysed in IP buffer (25 mm Tris-HCl (pH 7.4), 150 mm NaCl, 1 mm EDTA, 1% Nonidet P-40, 5% glycerol, protease inhibitor mixture (Sigma)) and centrifuged to remove cellular debris, and the protein was quantitated by bicinchoninic acid assay (Pierce).

Statistics

Data are presented as a mean ± S.D., and statistical significance was assessed using Student's t test. The null hypothesis was rejected for p < 0.05.

RESULTS

Ca_v2.1 and Ca_v2.2 Expression Is Reduced in Ankyrin-B^+/− Brain

Ankyrins are key players for targeting ion channels in excitable cells. Recent work defined new roles for ankyrin-B for targeting select voltage-gated Ca²⁺ channels in the pacemaking cells of the cardiac sinoatrial node (26, 27). We therefore tested the role of ankyrin-B for targeting select calcium channels in the nervous system. Specifically, we investigated Ca_v2.1 and Ca_v2.2 subunit expression in ankyrin-B^+/− cortex, cerebellum, and brain stem. As expected, ankyrin-B expression was significantly reduced in ankyrin-B^+/− cortex, cerebellum, and brain stem lysates (Fig. 1A; decreased 53, 52, and 51%, respectively; n = 4, p < 0.0001), whereas the expression levels of the related ankyrin gene product, ankyrin-G (Fig. 1B), and GAPDH (loading control; Fig. 1E) were unchanged. We observed decreased Ca_v2.1 expression in cortex (41%), cerebellum (20%), and brain stem (35%) in ankyrin-B^+/−-derived tissues compared with wild-type lysates (Fig. 1C; n = 4, p < 0.05). Ca_v2.2 expression was also reduced in ankyrin-B^+/− cortex (49%), cerebellar (40%), and brain stem tissue (50%) compared with lysates from wild-type littermates (Fig. 1D; n = 4, p < 0.05).

FIGURE 1.

Ca_v2.1 and Ca_v2.2 expression levels are decreased in ankyrin-B^+/− brain lysates. Ankyrin-B, Ca_v2.1, and Ca_v2.2 expression levels are significantly decreased in cortex, cerebellum, and brain stem in ankyrin-B^+/− mice compared with wild-type mice (A, C, and D, respectively). There are no significant differences in expression between wild-type and ankyrin-B^+/− lysates of ankyrin-G (B) or GAPDH (E, loading control). For each panel, n = 4 for each genotype (p < 0.05).

Based on our immunoblot findings, we performed confocal analysis for the localization of Ca_v2.1 and Ca_v2.2 in wild-type and ankyrin-B^+/− mouse cerebellum. In the cerebellum, Ca_v2.1 and Ca_v2.2 are expressed in Purkinje neurons and localized in the cell body (31, 32). Ankyrin-B is also expressed in the cerebellar Purkinje neuron cell body, as well as throughout the molecular layer (33). As shown in Fig. 2, whereas ankyrin-B, Ca_v2.1, and Ca_v2.2 are robustly expressed in Purkinje neurons from wild-type cerebellum, we observed a significant decrease of both Ca_v2.1 and Ca_v2.2 expression in response to the reduction in ankyrin-B expression (Fig. 2).

FIGURE 2.

Ankyrin-B^+/− mice display a significant decrease in ankyrin-B, Ca_v2. 1, and Ca_v2.2 expression in cerebellar Purkinje neurons. A and B, ankyrin-B expression is significantly reduced in ankyrin-B^+/− Purkinje neurons. Note that ankyrin-B is expressed in both Purkinje cell layer and molecular layer. C–F, Ca_v2.1 (C and D) and Ca_v2.2 (E and F) expression is significantly reduced in the Purkinje neurons of ankyrin-B^+/− mice. Scale bar denotes 20 μm. G and H, Ca_v2.1 and Ca_v2.2 antibodies are validated. Untransfected HEK293 cells and HEK293 cells transfected with either full-length GFP-Ca_v2.1 or GFP-Ca_v2.2 were analyzed by co-immunoprecipitation or immunoblotting (IB). In G, Ca_v2.1 Ig recognizes GFP-Ca_v2.1 but not GFP-Ca_v2.2. In H, Ca_v2.2 Ig recognizes GFP-Ca_v2.2 but not GFP-Ca_v2.1.

Ankyrin-G Is Not Required for Ca_v2.1 or Ca_v2.2 Expression

Although encoded from different genes, ankyrin-B and ankyrin-G display similar domain organization, and both target ion channels and transporters in diverse excitable cell types (2). As both ankyrin-B and ankyrin-G are co-expressed in cell types harboring Ca_v2.1 and Ca_v2.2, we evaluated whether ankyrin-G is required for Ca_v2.1 and Ca_v2.2 expression as shown for ankyrin-B (Figs. 1 and 2). In contrast to findings for ankyrin-B-deficient tissue, we observed no difference in Ca_v2.1 or Ca_v2.2 expression in cerebellar lysates generated from a well characterized mouse model homozygous for a null mutation in cerebellar ankyrin-G (Fig. 3 and Ref. 20). These data demonstrate that ankyrin-B, but not ankyrin-G, is required for regulating Ca_v2.1 and Ca_v2.2 expression in the cerebellum. Moreover, these data suggest that ankyrins-G and -B play nonredundant roles in calcium channel targeting in brain.

FIGURE 3.

Ankyrin-G is not required for Ca_v2.1 or Ca_v2.2 expression. Although ankyrin-G cerebellar-specific knock-out mice display a significant reduction in the expression of ankyrin-G (A), there are no significant differences in the expression of Ca_v2.1 or Ca_v2.2 between wild-type and ankyrin-G cerebellar-specific knock-out cerebellar lysates (A and B, n = 4; N.S.). Ankyrin-G levels by immunoblotting in ankyrin-G cerebellar-specific knock-out mice were significantly reduced and similar to background.

Ankyrin-B Associates with Ca_v2.1 and Ca_v2.2 in Brain

Considering that Ca_v2.1 and Ca_v2.2 expression levels were significantly reduced in ankyrin-B^+/− brain tissue (Figs. 1 and 2), we evaluated the ability of ankyrin-B to associate with Ca_v2.1 and Ca_v2.2 in brain. Co-IP analysis using detergent-soluble lysates from adult mouse cortex, cerebellum, and brain stem demonstrated that affinity-purified anti-Ca_v2.1 Ig co-immunoprecipitated ankyrin-B from cortex (Fig. 4A), cerebellum (Fig. 4B), and brain stem (Fig. 4C). Likewise, anti-Ca_v2.2 Ig co-immunoprecipitated ankyrin-B from cortex, cerebellum, and brain stem lysates (Fig. 4, D–F). We observed no interaction between ankyrin-B and control Ig in any tissue lysate (Fig. 4, A–F). To confirm this interaction, we performed reciprocal co-immunoprecipitations utilizing affinity-purified ankyrin-B Ig. Ankyrin-B Ig co-immunoprecipitated Ca_v2.1 and Ca_v2.2 from detergent-soluble lysates from cortex, cerebellum, and brain stem, whereas we observed no interaction with a control Ig (Fig. 4, G–L). Collectively, these data support an in vivo interaction between ankyrin-B and Ca_v2.1, as well as an in vivo interaction between ankyrin-B and Ca_v2.2, in mouse cortex, cerebellum, and brain stem.

FIGURE 4.

Ankyrin-B co-immunoprecipitates with Ca_v2.1 and Ca_v2.2. A–F, Ca_v2.1 Ig and Ca_v2.2 Ig co-immunoprecipitate ankyrin-B from detergent-soluble lysates of cortex (A and D), cerebellum (B and E), and brain stem (C and F), respectively. No significant interaction was observed with control (Ctrl) Ig. G–L, ankyrin-G Ig co-immunoprecipitates Ca_v2.1 and Ca_v2.2 from detergent-soluble lysates of cortex (G and J), cerebellum (H and K), and brain stem (I and L), respectively. No significant interaction was observed with control Ig. IB, immunoblotting.

We further evaluated the neuronal ankyrin-B/Ca_v2.1 and ankyrin-B/Ca_v2.2 interactions using pulldown experiments from mouse detergent-soluble cortex, cerebellum, and brain stem lysates. Ankyrin-B comprises three distinct structural domains: MBD, SBD, and CTD (Fig. 5A). Of note, GST-ankyrin-B MBD, but neither GST-ankyrin-B SBD or GST-ankyrin-B CTD interacted with Ca_v2.1 nor Ca_v2.2 from cortex (Fig. 5B), cerebellum (Fig. 5C), and brain stem (Fig. 5D; fusion proteins are shown in Fig. 5E). Collectively, these data demonstrate an interaction between Ca_v2.1 and Ca_v2.2 and ankyrin-B within multiple regions of the brain. These data further identify the structural requirements on ankyrin-B for Ca_v channel binding as the ANK repeats of the membrane-binding domain.

FIGURE 5.

Ankyrin-B associates with Ca_v2.1 and Ca_v2.2. A, ankyrin-B comprises a membrane-binding domain, spectrin-binding domain, and a C-terminal regulatory domain. B–D, GST-ankyrin-B MBD associates with both Ca_v2.1 and Ca_v2.2 from detergent-soluble lysates of cortex (B), cerebellum (C), and brain stem (D). No interaction was observed between lysates and GST-ankyrin-B SBD, GST-ankyrin-B CTD, or GST alone. E, GST fusion proteins were utilized for pull-down experiments. Fusion proteins were purified from bacteria and analyzed by Coomassie Blue staining. Whereas GST migrates at 29 kDa, GST-ankyrin-B MBD migrates at 94 kDa. GST-ankyrin-B SBD migrates at 62 kDa, and GST-ankyrin-B CTD migrates at 79 kDa.

Ankyrin-B Interacts with the Ca_v2.1 and Ca_v2.2 DII/III Loops

Next, we evaluated the requirements on Ca_v2.1 and Ca_v2.2 for neuronal ankyrin-B binding. Previous studies have suggested that a region in the DII/III loop region of Ca_v2.2 contains a region postulated to play a role in the targeting of calcium channels (34) (Fig. 6, A and B). This DII/III region contains a 20-amino acid motif that is highly conserved among species (Fig. 6, C and D). Three constructs based on Ca_v2.1 and Ca_v2.2 sequences were designed to evaluate the role of this motif in ankyrin-B/Ca_v2.1 and ankyrin-B/Ca_v2.2 association (Fig. 6, E and F). The first construct encompassed the entire DII/III loop (Ca_v2.1 amino acids 708–1194; Ca_v2.2 amino acids 710–1139); the second initiated proximal to the putative ankyrin-binding motif (ABM; Ca_v2.1 amino acids 788–1194; Ca_v2.2 amino acids 779–1139); and the third began distal to the ABM (Ca_v2.1 amino acids 808–1194; Ca_v2.2 amino acids 799–1139). In vitro binding with immobilized GST-ankyrin-B MBD revealed that the full-length Ca_v2.1 DII/III loop and the Ca_v2.1 construct containing residues 788–1194 displayed binding activity for GST-ankyrin-B MBD (Fig. 6G). In contrast, the Ca_v2.1 construct lacking residues 788–808 (construct 808–1194) had binding activity similar to GST alone (Fig. 6G). Similarly, full-length Ca_v2.2 DII/III loop and the Ca_v2.2 construct containing residues 779–799 (construct 779–1139) displayed binding activity for GST-ankyrin-B MBD (Fig. 6H). The Ca_v2.2 construct lacking residues 779–799 (construct 799–1139) had binding activity similar to GST alone (Fig. 6H). These data identify a 20-amino acid motif in both the Ca_v2.1 and Ca_v2.2 DII/III loops that is necessary for interaction with ankyrin-B.

FIGURE 6.

Ca_v2.1 and Ca_v2.2 DII/DIII motif associates with ankyrin-B. A and B, location of putative ABM in the DII/DIII cytoplasmic loop of Ca_v2.1 and Ca_v2.2 is shown. C and D, the putative ABM in Ca_v2.1 and Ca_v2.2 is conserved across species and harbors a central conserved tyrosine (bolded). E and F, three constructs were designed to evaluate the role of this amino acid sequence in ankyrin-B binding. These constructs include the entire DII/III loop, a portion of the loop that begins just prior to the putative ABM, and a portion of the loop that begins just after the putative ABM. G and H, in vitro binding assays demonstrate that, whereas the full-length DII/III loops of Ca_v2.1 and Ca_v2.2 and the smaller loop portion that contains the putative ABM are able to associate with GST-ankyrin-B MBD, the smaller loop portion that lacks this domain demonstrates no binding with GST-ankyrin-B MBD. I and J, Coomassie Blue staining demonstrating that equal loading of GST constructs was performed prior to imaging (denoted by asterisks).

A Conserved Tyrosine Residue Regulates Association with Ankyrin-B

Although inhibition of calcium channels is voltage-dependent, it is mediated by direct interaction of G protein βγ subunits with the α1 pore-forming subunit of the channel (35, 36). Additionally, phosphorylation by tyrosine kinases has been shown to inhibit calcium channels (37) with G protein-dependent inhibition of calcium current due in part to a decrease in the open probability of the channel, resulting in reduced current density (38–40). We had previously defined a 20-amino acid motif in the DII/III loops of Ca_v2.1 and Ca_v2.2 that associates with ankyrin-B (Fig. 6). Notably, this motif contains a conserved tyrosine residue. To evaluate the role of this tyrosine in the regulation of Ca_v2.1 and Ca_v2.2 association with ankyrin-B, we incorporated a Y797E mutation in the Ca_v2.1 DII/III loop and evaluated the effect on in vitro binding (Fig. 7A). Mutagenesis of the Tyr-797 residue to a glutamic acid resulted in loss of association with GST-ankyrin-B MBD by in vitro binding analysis (Fig. 7B). Mutation of the analogous residue, Tyr-788, in the Ca_v2.2 DII/III to a glutamic acid (Fig. 7D) demonstrated a similar loss of in vitro binding with GST-ankyrin-B MBD (Fig. 7E).

FIGURE 7.

A conserved tyrosine residue in the Ca_v2. 1 and Ca_v2.2 DII/III loop is necessary for ankyrin-B interaction in vitro. A and D, Ca_v2.1 and Ca_v2.2 DII/III loop domains contain a conserved tyrosine residue that was experimentally mutated to a glutamic acid to assess ankyrin-B binding. B, GST-ankyrin-B MBD was able to associate with the Ca_v2.1 DII/III loop domain. In contrast, there was a loss of interaction with the domain containing the Y797E mutation. There was no appreciable binding with GST alone. E, Ca_v2.2 DII/III loop domain associated with GST-ankyrin-B MBD is shown. However, there was no appreciable association between GST-ankyrin-B MBD and the Ca_v2.2 DII/III Y788E. There was no detectable binding to GST alone. C and F, Coomassie Blue stains demonstrate that equal loading of GST constructs was performed prior to imaging.

DII/DIII Tyrosine Residues Are Critical for Ca_v2.1/Ca_v2.2 Targeting

Ankyrin proteins are critical for targeting ion channels and transporters in diverse cell types including neurons. In fact, ion channel or transporter ankyrin-binding motifs alone are sufficient to target GFP and other markers in heterologous cells and primary cultures (41). We therefore tested (i) whether the DII/DIII loops of Ca_v2.1 and Ca_v2.2 (that harbor ankyrin-binding motifs) were sufficient to behave as a dominant target signal in HEK293 cells, and (ii) whether the single tyrosine mutation in each channel altered this trafficking pattern. DII/DIII loops from Ca_v2.1 and Ca_v2.2 were subcloned in-frame into the pEGFP-C3 expression vector and transfected into HEK293 cells (express endogenous ankyrin-B). Confocal analysis revealed that unlike GFP that was exclusively found in the HEK293 cell nucleus (Fig. 8, A and D), GFP-Ca_v2.1 DII/III loop and GFP-Ca_v2.2 DII/III were in fact targeted outside of the nucleus (versus GFP) and had a diffuse pattern of expression throughout the cell (Fig. 8, B and E). Consistent with these findings, both GFP-Ca_v2.1 DII/III loop and GFP-Ca_v2.2 DII/III loop interacted with ankyrin-B in parallel co-IP experiments from transfected cells (Fig. 8, G–I). In contrast, GFP-Ca_v2.1 DII/III Y797E and GFP-Ca_v2.2 DII/III Y788E lacked ankyrin-B-binding activity (Fig. 8, G–I) and were restricted in localization to a tight and consistent perinuclear distribution (Fig. 8, C and F). These data demonstrate that the Ca_v2.1 and Ca_v2.2 DII/DIII loops harbor dominant targeting motifs that divert them from the nucleus. Further, these data implicate the Tyr-797 in the Ca_v2.1 and the Tyr-788 residue in the Ca_v2.2 as critical components in the regulation of calcium channel targeting.

FIGURE 8.

Ca_v2.1/Ca_v2.2 tyrosine residues in ABM are critical for cellular targeting. HEK293 cells transfected with pEGFP-C3 demonstrated a primarily nuclear localization pattern (A and D), whereas HEK293 cells transfected with Ca_v2.1 DII/III-GFP (B) or Ca_v2.2 DII/III-GFP (E) displayed a diffuse cellular pattern of expression unique from the nucleus. C and F, unlike WT constructs, both Ca_v2.1 Y797E-GFP and Ca_v2.2 Y788E-GFP were distributed in a tight perinuclear pattern. Scale bars denote 8 μm. G–I, a conserved tyrosine residue in the Ca_v2.1 and Ca_v2.2 DII/III loop is necessary for ankyrin-B interaction in vitro. G, in control experiments, ankyrin-B Ig did not interact with GFP alone expressed in HEK293 cells. H and I, ankyrin-B Ig co-immunoprecipitates GFP fusion protein containing the Ca_v2.1 or Ca_v2.2 DII/DII loop from transfected HEK293 cells. In contrast, ankyrin-B Ig does not co-immunoprecipitate GFP fusion proteins of Ca_v2.1 DII/DIII Y797E or Ca_v2.2 DII/III Y788E in parallel experiments. IB, immunoblotting.

We confirmed these findings in the context of full-length Ca_v2.1 and Ca_v2.2. We observed both GFP-Ca_v2.1 and GFP-Ca_v2.2 with diffuse expression across transfected HEK293 cells, whereas GFP-Ca_v2.1 Y797E and full-length GFP-Ca_v2.2 Y788E displayed abnormal targeting with a clustered perinuclear distribution (Fig. 9, A–E). Consistent with these findings, whereas GFP-Ca_v2.1 and GFP-Ca_v2.2 associated with ankyrin-B in co-IP experiments, ankyrin-B Ig was unable to co-immunoprecipitate either full-length GFP-Ca_v2.1 Y797E or full-length GFP-Ca_v2.2 Y788E from HEK293 cell lysates (Fig. 9, F–H). These results further support a critical role for ankyrin-B-mediated Ca_v2.1 and Ca_v2.2 expression and targeting.

FIGURE 9.

Ca_v2. 1/Ca_v2.2 tyrosine residues in ABM are critical for cellular targeting and ankyrin interaction. A, HEK293 cells transfected with pEGFP-C3 demonstrated a primarily nuclear localization pattern. B and D, in contrast, HEK293 cells transfected with full-length GFP-Ca_v2.1 (B) or GFP-Ca_v2.2 (D) displayed a diffuse cellular pattern of expression unique from the nucleus or perinuclear region. C and E, unlike WT Ca_v2.1 and Ca_v2.2 constructs, both full-length GFP-Ca_v2.1 Y797E and full-length GFP-Ca_v2.2 Y788E-GFP were distributed in a tight perinuclear pattern (white arrowheads). Scale bars denote 8 μm. F, control experiments demonstrate that ankyrin-B Ig does not associate with GFP alone. G, ankyrin-B co-immunoprecipitates full-length GFP-Ca_v2.1, but not full-length GFP-Ca_v2.1 Y797E from transfected HEK293 lysates. H, ankyrin-B co-immunoprecipitates full-length GFP-Ca_v2.2, but not full-length GFP-Ca_v2.2 Y788E from transfected HEK293 lysates.

DISCUSSION

Our findings demonstrate a role of ankyrin-B for Ca_v2.1 and Ca_v2.2 channel expression and targeting in brain. Loss of ankyrin-B results in decreased expression of Ca_v2.1 and Ca_v2.2 in cortex, cerebellum, and brain stem. Moreover, mutation of a single residue in the Ca_v2.1 and Ca_v2.2 DII/III loop results in loss of association with ankyrin-B and abnormal cellular targeting. Ankyrin-B has been demonstrated to play a critical role in the establishment and maintenance of excitable domains in brain. Loss of ankyrin-B in mice results in significant nervous system defects, including hypoplasia of the corpus callosum and pyramidal tracts, dilation of the lateral ventricles, and deterioration of long axon tracts (25). Specifically, these phenotypes are associated with loss of L1CAM proteins throughout the brain. Interestingly, mutations in the human L1CAM gene cause mental retardation as well as a collection of neurological symptoms referred to as CRASH syndrome (corpus callosum hypoplasia, retardation, aphasia, spastic paraplegia, and hydrocephalus) (42). The ankyrin-B^−/− mouse model, although lethal at post-natal day 1, exhibits loss of L1CAM from axons and exhibits many of the CRASH symptoms, although with greater severity (25). No humans with null mutations in the ankyrin-B gene (ANK2) have been identified; however, given that heterozygote individuals with loss-of-function mutations in the ANK2 gene are present in the population (43) and are found in patients with cardiac arrhythmia phenotypes (26, 44–47), it is possible that rare homozygous or mixed heterozygotes may exhibit CRASH symptoms. S1224L and Y1229H mutations in the human L1CAM gene are associated with CRASH syndrome (48). Moreover, these mutations are located within the highly conserved ankyrin-binding site and abolish association of L1 with ankyrin-B (48).

Ca_v2.1 and Ca_v2.2 channel activity in brain is necessary for proper initiation of synaptic transmission. Ca_v2.2^−/− animals demonstrate normal motor coordination; however, these mice exhibit reduced responses to pain stimuli (49) and elevated aggression (50). Ca_v2.1^−/− animals demonstrate dystonia at postnatal day 10–12 and rarely survive past weaning (51). Likewise, Ca_v2.1^−/− mice exhibit reduced responses to pain stimuli (52). As noted above, ankyrin-B^−/− die at birth, therefore we currently lack data on the in vivo phenotypes of ankyrin-B deficiency in an adult mouse. Future studies utilizing inducible ankyrin-B knock-out strategies in the postnatal mouse will be important for understanding the role of this cytoskeletal protein in synaptic function.

Finally, although our data link ankyrin-B with regulation of targeting via the DII/DIII loop of both Ca_v2.1 and Ca_v2.2 via specific tyrosine residues, it is possible that other protein interactions in this loop may also play critical roles in ion channel processing and membrane regulation. For example, the Ca_v2.2 DII/DIII loop maintains a SNARE binding or synprint region that has been linked with Ca_v channel targeting (34). It will be important in future work to understand the potential dual regulation of Ca_v channel targeting by ankyrin and other proteins.