Protein phosphatase 2A (PP2A) is a multifunctional protein phosphatase with critical roles in excitable cell signaling. In the heart, PP2A function is linked with modulation of β-adrenergic signaling and has been suggested to regulate key ion channels and transporters including Na/Ca exchanger, ryanodine receptor, inositol 1,4,5-trisphosphate receptor, and Na/K ATPase. Although many of the functional roles and molecular targets for PP2A in heart are known, little is established regarding the cellular pathways that localize specific PP2A isoform activities to subcellular sites. We report that the PP2A regulatory subunit B56α is an in vivo binding partner for ankyrin-B, an adapter protein required for normal subcellular localization of the Na/Ca exchanger, Na/K ATPase, and inositol 1,4,5-trisphosphate receptor. Ankyrin-B and B56α are colocalized and coimmunoprecipitate in primary cardiomyocytes. Using multiple strategies, we identified the structural requirements on B56α for ankyrin-B association as a 13 residue motif in the B56α COOH terminus not present in other B56 family polypeptides. Finally, we report that reduced ankyrin-B expression in primary ankyrin-B+/− cardiomyocytes results in disorganized distribution of B56α that can be rescued by exogenous expression of ankyrin-B. These new data implicate ankyrin-B as a critical targeting component for PP2A in heart and identify a new class of signaling proteins targeted by ankyrin polypeptides.
efficient excitable cell function requires precisely orchestrated signaling pathways to modulate the function of diverse membrane, cytosolic, and nuclear proteins. Specifically, the collaboration between finely tuned protein phosphorylation/dephosphorylation pathways is critical for neuronal communication and cardiac excitation-contraction coupling. Findings over the past decade clearly demonstrate that a local organization of kinase and phosphatase proteins with specific effector protein pathways facilitates efficient signaling. This local organization may occur by direct interaction of kinases and phosphatases with anchoring or scaffolding proteins, cytoskeleton, or even by direct interaction with target ion channels, transporters, or receptors (21, 36, 42, 44, 67). However, the cellular pathways underlying the biogenesis/targeting of kinases or phosphatases to specialized membrane protein complexes are not clearly resolved.
Protein phosphatase 2A (PP2A) is a multifunctional serine/threonine phosphatase with critical roles in ion channel/transporter regulation (3, 34, 41, 88), Wnt signaling (39, 63, 72, 84), transformation (2, 6, 27, 75), cell polarization (23), circadian rhythm (5, 68, 85), and cell survival and apoptosis (37, 40, 64, 73, 76, 81). In heart, PP2A function is necessary for normal excitation-contraction coupling, and PP2A levels are altered in heart failure (1, 57, 58). Numerous studies have suggested that PP2A activity is associated with critical cardiac receptors, ion channels, and transporters including β-adrenergic receptor (10, 12, 22), L-type Ca2+-channels (11, 22), ryanodine receptor (43), Na/Ca exchanger (70), inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] receptor (13), and Na/K ATPase (35).
PP2A is a heterotrimer made up of a 65-kDa structural A subunit, a highly conserved 36-kDa catalytic C subunit, and a variable regulatory B subunit. The PP2A core enzyme is a dimer consisting of the A and C subunits. The A subunit acts as a scaffold that mediates the interaction between the catalytic and regulatory/targeting B subunits. Finally, the B subunits are responsible for substrate specificity, subcellular localization, and enzymatic activity of the entire PP2A holoenzyme (26, 45, 55); however, the molecular pathways underlying the specificity for these cellular functions are currently unknown.
Ankyrin polypeptides are required for proper membrane localization of ion channels, transporters, and cell adhesion molecules. Ankyrin activity is critical for normal function of neurons, epithelia, and erythrocytes (17, 29, 30, 33, 59, 87). Over the past five years, ankyrin dysfunction has been linked to severe cardiac arrhythmia in mice and humans (9). Mice heterozygous for a null mutation in ankyrin-B display a number of cardiac phenotypes including bradycardia, conduction defects, polymorphic ventricular tachycardia, and sudden cardiac death in response to catecholaminergic stimuli (51).
Ankyrin-B contains an NH2-terminal “membrane-binding” domain, a “spectrin-binding” domain (SBD), and a “regulatory” domain. Recent findings suggest a critical role for the SBD in ankyrin-B function. Specifically, of all ANK2 (encodes ankyrin-B) human arrhythmia variants reported to date, the A4274G variant (leading to ankyrin-B E1425G) is associated with the most severe human cardiac phenotypes and is located in the COOH terminus of the SBD (51, 52). Based on the identification of this variant and the lack of information regarding additional functional roles for the SBD, we sought to identify binding partners for the COOH-terminal region of the SBD.
Using the SBD as bait, we screened a human heart library to identify novel ankyrin-associated proteins. Our screen identified B56α, a targeting/regulatory subunit of the serine/threonine phosphatase PP2A (45, 79), as a candidate ankyrin-binding partner. Moreover, we identified a unique 13 amino acid domain on B56α required for ankyrin-B association. A series of complementary studies reveals that ankyrin-B and B56α, but not other B56 family members, are physiological binding partners in cardiac myocytes. Furthermore, we report the unexpected finding that ankyrin-B is required for proper B56α targeting in cardiomyocytes. Thus ankyrin-B not only organizes ion channels and transporters but is also a critical targeting component for PP2A. These results provide exciting new insights into the cellular pathways responsible for the biogenesis of local signaling domains at specific subcellular domains in excitable cells.
MATERIALS AND METHODS
Mice used in these studies were neonatal wild-type C57BL/6 mice and ankyrin-B+/− and ankyrin-B−/− littermates (C57BL/6), 1 to 2 days old. Sprague-Dawley rats (250–275 g) and adult wild-type C57BL/6 mice were used as source of cardiac tissue and isolated adult ventricular myocytes. Animals were handled according to approved protocols and animal welfare regulations of the Institutional Animal Care and Use Committee.
Yeast two-hybrid assays.
The COOH-terminal region of the human 220-kDa ankyrin-B SBD (nucleotides 3282–4329) was cloned into pAS2-1 (Clontech) to create GAL4-ankyrin-B SBD Δ (SBDΔ; residues 1094–1443). With the use of a lithium acetate-based transformation protocol, the bait plasmid was expressed into AH109 yeast (ADE2, HIS3, lacZ selection) (50). We did not observe autoactivation of the bait plasmid on plates lacking adenine-leucine-tryptophan (ALT) or adenine-histidine-leucine-tryptophan (AHLT). AH109 yeast expressing GAL4-SBD were mated with pretransformed yeast harboring the human heart Matchmaker cDNA library in pACT2 (Clontech). Yeast were grown in nutrient-rich medium, and mated yeast were then grown on medium lacking AHLT (−AHLT). Clones displaying significant growth on −AHLT medium were replated on −AHLT plates to confirm positive selection. Interaction bait/prey controls for the screen included TD1–1/pLAM5 (negative, Clontech) and pVA3/TD1–1 (positive). Approximately 3.2 × 106 independent clones were screened.
Mouse neonatal cardiomyocytes.
To generate primary cardiomyocyte cultures, hearts were dissected from P1 or P2 mice and placed in 2 ml of Ham's F-10 medium (Mediatech). Atrial tissue was removed, and the ventricular chambers were rinsed to remove any remaining blood. Hearts were transferred into 1.5 ml of 0.05% trypsin and 200 μM EDTA in Ham's F-10 medium. Hearts were minced into ∼40 small pieces using forceps and small scissors and incubated in trypsin-EDTA at 37°C. After 15 min, the heart pieces were gently triturated and incubated for an additional 15 min. A mixture of 200 μl of soybean trypsin inhibitor (2 mg/ml; Worthington) and 200 μl of collagenase (0.2 mg/ml; 1,980 U/mg; Sigma) was incubated with the cells for 50 min at 37°C. The cell suspension was pelleted, resuspended in complete medium (40% DMEM, 40% Ham's F-10 medium, and 20% FBS), and plated on plastic dishes. After 5 h, the nonadherent cells (cardiomyocytes) were aspirated from the plate, pelleted, resuspended in complete medium, and plated on fibronectin-coated (Roche) coverslips or glass tissue culture plates (Mattek). After 24 h, cardiomyocytes were washed with Ham's F-10 medium, and the complete medium was replaced with defined medium to prevent an overgrowth of fibroblasts. Defined medium (100×) consists of 100 μg/ml insulin, 500 μg/ml transferrin, 100 nM LiCl, 100 nM NaSeO4, and 10 nM thyroxine. Medium was replaced every 24 h. For rescue experiments, cardiomyocytes were transfected using Effectene (Qiagen) after exactly 84 h in culture. For each 14-mm coverslip, the transfection reaction was performed in 2 ml of defined medium with 0.4 μg endotoxin-free plasmid DNA and 10 μl of Effectene transfection reagent. Cardiomyocytes were transfected for 9 h, washed extensively in Ham's F-10 medium, and then transferred to complete medium. After 12 h, cardiomyocytes were washed and incubated with defined medium. At 6 to 7 days postisolation, cardiomyocytes were fixed and processed for immunofluorescence. Transfection efficiency in the presented experiments was <10% of cardiomyocytes.
Deletion constructs of human B56α were generated by PCR. B56α constructs correspond to residues 1–486 (nucleotides 1–1458), 1–189 (nucleotides 1–567), 190–486 (nucleotides 570–1458), 271–486 (nucleotides 813–1458), 313–486 (nucleotides 939–1458), 398–486 (nucleotides 1194–1458), 447–486 (nucleotides 1341–1458), 473–486 (nucleotides 1419–1458), 447–472 (nucleotides 1341–1416), and 1–472 (nucleotides 1–1416). Constructs were ligated into pACT2 and sequenced to ensure that no nucleotide variants or deletions were introduced by the PCR protocol.
Solubilization of cardiac proteins.
Adult heart immunoprecipitations were performed as described (51). Briefly, the adult mouse heart was dissected and rinsed in PBS plus 0.32 M sucrose and 2 mM Na-EDTA. Tissue was flash frozen in liquid nitrogen and ground into a fine powder. The powder was resuspended in four volumes of 50 mM Tris·HCl (pH 7.35), 10 mM NaCl, 0.32 M sucrose, 5 mM Na EDTA, 2.5 mM Na EGTA, 1 mM PMSF, 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride hydrochloride, 10 μg/ml leupeptin, and 10 μg/ml pepstatin using a Dounce homogenizer. The homogenate was centrifuged at 1,000 g to remove nuclei. Triton X-100 and deoxycholate were added to the postnuclear supernatant for final concentrations of 1.5% Triton X-100 and 0.75% deoxycholate. The lysate was pelleted at 100,000 g for 1 h at 4°C, and the supernatant was recleared at 100,000 g for 1 h to remove residual large membranes or vesicles. The resulting supernatant was used for immunoprecipitation as described (48) or for binding experiments.
Antibodies include affinity-purified antibodies to green fluorescent protein (GFP) (48) and ankyrin-B (48). The antibody against B56α was purchased from BD Biosciences (San Jose, CA). PP2A/c antibody was purchased from Transduction Laboratories (Lexington, KY) and PP2A/A antibody from Oxford Biomedical Research (Oxford, MI). Mouse monoclonal anti-ryanodine receptor 2 was purchased from Affinity Bioreagents (Golden, CO). Monoclonal anti-α-actinin was obtained from Sigma (St. Louis, MO).
Neonatal cardiomyocytes were washed with PBS (pH 7.4) and fixed in 2% paraformaldehyde. Cells were blocked/permeabilized in PBS containing 0.075% Triton X-100 and 2 mg/ml BSA and incubated in primary antibody overnight at 4°C. After PBS washes, cells were incubated in secondary antibody (Alexa 488, 568; Molecular Probes) for 2 h at room temperature and mounted using Vectashield (Vector) and No. 1 coverslips. Images were collected on Zeiss 510 Meta confocal microscope [63 power-oil 1.40 numerical aperature (Zeiss), pinhole equals 1.0 Airy Disc] using Carl Zeiss Imaging software. Images were imported into Adobe Photoshop for cropping and linear contrast adjustment.
Preparation and staining of adult ventricular cells.
For adult rat and mouse ventricular myocyte cultures, dissociated cells were prepared from adult Sprague-Dawley rats (250–275 g) or adult C57Bl/6 mice as previously described (15, 16). Following dissociation, ventricular cardiomyocytes were resuspended in NaHCO3-buffered medium 199 supplemented with 10−7 M insulin, 5 mM creatine, 2 mM l-carnitine, 0.2% BSA, 5 mM taurine, 1% penicillin-streptomycin, and 3 mM N-(2-mercaptopropionyl)-glycine. Myocytes were then used for biochemistry or immunostaining following fixation in ice-cold 100% ethanol (47, 51).
A cell suspension of isolated adult rat or mouse cardiomyocytes was pelleted, washed once with PBS, repelleted, and resuspended in 6 ml of lysis buffer containing 50 mM Tris·HCl, 10 mM NaCl, 2 mM DTT, protease inhibitor cocktail, and 1% Nonidet P-40 (pH 7.4). The suspension was homogenized and centrifuged at 100,000 g to remove large membranes. The resulting supernatants were used for pull-down experiments. Glutathione S-transferase (GST), GST-B56α, and GST-B56γ were purified using standard techniques. Biotinylated peptides were generated against the COOH-terminal residues of mouse B56α (SGSGAYNMHSILSNTSAE) and B56γ (SGSGAKANPQVLKKRIT; Biosyn). An S-G-S-G linker was inserted between the biotin and the B56 sequence. Peptides were purified and bound (10 μg/each) to streptavidin-sepharose beads. GST, GST-B56α, and GST-B56γ beads were incubated with 0.5 ml of cell extract (1.2 mg protein/ml) overnight with gentle agitation at 4°C. Unbound supernatants were removed for later analysis. After being extensively washed, bound protein was eluted using sample buffer and analyzed by SDS-PAGE. Immunoblots were performed using antibodies to cognate binding partners of B56α or B56γ, the PP2A structural subunit, PP2A/A, and the PP2A catalytic subunit PP2A/c. Additionally, blots were probed with affinity-purified Ig raised against the ankyrin-B regulatory domain (49).
PP2A-targeting subunit interacts with the ankyrin-B in yeast.
We performed a yeast two-hybrid screen of human heart library to identify candidate binding partners for the ankyrin-B SBD. The bait construct was designed to encode the COOH-terminal region of the SBD for two reasons. First, previous ankyrin yeast two-hybrid SBD screens identify β2-spectrin as a high percentage of positive interacting clones. β2-Spectrin binding activity is located to the NH2-terminal region of this 62 kDa domain (53). Therefore, we designed our bait construct COOH terminal to this site (see Fig. 1A) to increase our efficiency for identifying nonspectrin clones. Second, the ankyrin-B E1425G arrhythmia variant is located in the SBD COOH terminus (51). This specific variant is associated with the most severe human cardiac phenotypes including polymorphic ventricular tachycardia and sudden cardiac death (51, 52, 69). Therefore, this specific region of the ankyrin-B SBD is likely critical for normal ankyrin-B function in vivo.
We screened 3.2 × 106 independent clones and identified 63 in-frame candidate-interacting proteins. Two of the 63 clones encoded COOH-terminal cDNA sequence of B56α, a PP2A targeting/regulatory subunit preferentially expressed in heart (31, 45, 79). The longest clone (clone 163) represented human nucleotides 162–1458 (Fig. 1). The second positive clone represented residues 915–1458 (Fig. 1). Clone 163 did not interact with a bait plasmid containing the full ankyrin-G SBD (not shown). These results demonstrate that the ankyrin-B SBD specifically interacts with residues in the COOH-terminal region (residues 305–486; nucleotides 915–1458) of B56α in yeast.
B56α colocalizes with ankyrin-B in primary cardiomyocytes.
A prerequisite for an ankyrin-B/B56α in vivo interaction is their colocalization in cardiac myocytes. Ankyrin-B expression in both neonatal and adult cardiomyocytes has been extensively studied and is known to be localized in isolated adult ventricular myocytes at both M- and Z-lines and in primary neonatal cardiomyocytes at M-lines (Z-disc/transverse-tubule domains not yet developed) (47–53, 78). We examined potential colocalization of ankyrin-B with endogenous B56α using cultured neonatal mouse ventricular myocytes. As previously demonstrated, ankyrin-B is primarily localized over the M-line in these immature cells [Fig. 2 and supplemental Fig. 1 (note: the supplemental figure may be found with the online version of this article); alternates with α-actinin labeling (46, 49)]. Labeling of endogenous B56α revealed a strong overlap with ankyrin-B immunofluorescence over the M-line (Fig. 2). We also observed minor populations of B56α-positive staining (<5%) at sites distinct from ankyrin-B (Fig. 2; note arrows in Fig. 2B). Therefore, ankyrin-B and B56α are coexpressed at specific cellular sites in neonatal primary cardiomyocytes that are consistent with a potential ankyrin-B/B56α interaction.
We performed immunostaining experiments using adult mouse cardiomyocytes to confirm neonatal cardiomyocyte results. As previously demonstrated (47, 51), ankyrin-B is concentrated at the M-line of adult mouse cardiomyocytes (alternates with ryanodine receptor 2 Z-line staining; Fig. 3, A and C) but also displays a minor population of immunostaining over the Z-line (47, 51). Similar to ankyrin-B staining, and consistent with our results from neonatal cardiomyocytes (Fig. 2), we observed significant colocalization of endogenous B56α with ankyrin-B over the M-line. No specific signal was observed in myocytes treated with secondary antibody alone. Moreover, similar to ankyrin-B, we observed a second minor population of B56α localized to the cardiomyocyte Z-line. Therefore, in both neonatal and adult mouse cardiomyocytes, the localization of B56α is consistent with a potential interaction with ankyrin-B.
B56α associates with ankyrin-B in heart.
We evaluated the potential ankyrin-B/B56α interaction in cardiac tissue using GST pull-down assays. GST-B56α was expressed in bacteria and purified using glutathione-sepharose beads. Detergent-soluble lysate was generated from ventricular cardiomyocytes isolated from adult rat heart and incubated with glutathione-sepharose coupled to GST (control) or GST-B56α. Following extensive washes, bound protein was eluted from the beads and analyzed by SDS-PAGE and immunoblot analysis. We first evaluated the ability of GST-B56α to associate with the catalytic (PP2A/c) and structural/scaffolding (PP2A/A) subunits of PP2A, known binding partners of B56α in heart and other tissues (74, 79). Immunoblot analysis of pull-down experiments with antibodies against PP2A/c and PP2A/A revealed that GST-B56α but not GST alone associated with a significant fraction of detergent-soluble PP2A/c and PP2A/A in heart extracts (Fig. 4A). An analysis of the binding experiments using affinity-purified ankyrin-B IgG revealed that GST-B56α but not GST alone associated with 220-kDa ankyrin-B, the predominant ankyrin-B isoform in heart [Fig. 4A (51, 78)]. These pull-down data confirm our findings in yeast and demonstrate that 220-kDa ankyrin-B and B56α interact in cardiac lysates.
We further evaluated the ankyrin-B/B56α interaction using coimmunoprecipitation assays from detergent-soluble lysates generated from adult mouse heart. Lysates were prepared as described in materials and methods, and B56α was immunoprecipitated using B56α-specific antibody and protein A sepharose. Following extensive washes, bound protein was eluted and analyzed by immunoblot. As shown in Fig. 4B, B56α Ig was able to coimmunoprecipitate 220-kDa ankyrin-B from cardiac lysate. In contrast, we did not observe ankyrin-B in control coimmunoprecipitation experiments (control Ig). Therefore, consistent with our findings in yeast and pull-down experiments, coimmunoprecipitation experiments support an in vivo interaction between ankyrin-B and B56α.
Defining the ankyrin-B-binding domain on B56α.
We used the two-hybrid system to identify the structural requirements on B56α for ankyrin-B-binding activity. We generated 10 B56α baits corresponding to combinations of seven domains in pACT-2 using standard molecular techniques. Based on the original yeast two-hybrid clone (Fig. 1B), all clones, with the exception of two, contained the COOH-terminal region of B56α (Fig. 5). Bait clones were transformed into AH109 yeast containing the COOH-terminal region of the ankyrin-B SBD. As expected, clone 163 as well as full-length B56α interacted with ankyrin-B SBD in yeast (Fig. 5). Two-hybrid assays revealed that the COOH terminus of B56α is required for interaction with ankyrin-B in yeast (Fig. 5). Specifically, the COOH-terminal domain bait alone was sufficient for ankyrin-B binding activity (Fig. 5). Moreover, a COOH-terminal B56α truncation lacking the final 13 amino acids did not interact with ankyrin-B, whereas a bait of only the final 13 amino acids displayed ankyrin-B binding activity in yeast (Fig. 5).
Ankyrin-B specifically interacts with the α-isoform of the B56 family.
Translation of the ankyrin-binding motif on human B56α revealed 13 residues that are highly conserved between human, mouse, and rat (Fig. 6A). B56α contains a core region domain that is highly conserved among the other B56 isoforms [β, δ, γ1, γ2, γ3, ε (45, 55, 77)]. Based on our original yeast clone, we could not rule out the possibility that other B56 family members could also interact with ankyrin-B. We performed sequence analyses of the B56 family of polypeptides to determine whether other B56 isoforms might display sequence homology to the COOH terminus of B56α and therefore might also interact with ankyrin-B. Whereas the COOH-terminal sequences of B56γ isoforms [γ1, γ2, and γ3, derived from alternative splicing products (56)] display high homology, the COOH-terminal amino acids of other B56 isoforms are strikingly divergent (Fig. 6B). In fact, the final 13 residues of B56α do not display sequence homology to any coding sequence present in other B56 gene products. These results strongly suggest that ankyrin-B interacts with only the α-isoform of the B56 family of PP2A targeting proteins.
To test whether ankyrin-B preferentially associated with the B56α subunit, we performed pull-down experiments using purified GST-B56α and GST-B56γ fusion proteins. Consistent with our previous experiments (Fig. 4), we observed the interaction of GST-B56α with ankyrin-B from detergent-soluble lysates of adult heart (Fig. 7A). In contrast, we were unable to observe the interaction of GST-B56γ1 using an identical binding protocol (Fig. 7A). Control experiments showed that the levels of GST-B56α and GST-B56γ1 bound to the beads were equivalent (see Coomassie-stained gel in Fig. 7A, bottom). Furthermore, the GST-B56γ1 chimera was still biologically active since this affinity matrix bound PP2A/c and PP2A/A from heart extracts as expected (data not shown). These data support sequence predictions that ankyrin-B specifically interacts with only the B56α isoform of the B56 family.
B56α COOH terminus is sufficient to interact with ankyrin-B.
Yeast two-hybrid data define the B56α COOH-terminal residues as critical for interaction with ankyrin-B (Fig. 5). To determine whether these residues are sufficient for ankyrin-B association, we performed pull-down experiments using biotinylated peptides of the COOH-terminal residues of B56α and B56γ. NH2-terminal biotinylated peptides were coupled to streptavidin-agarose and incubated with detergent-soluble lysates of adult mouse heart. Following the binding reaction, the agarose beads were extensively washed, and bound protein was eluted and analyzed by immunoblot using ankyrin-B Ig. As predicted from our experiments in yeast, the COOH-terminal B56α peptide was sufficient to purify ankyrin-B from cardiac lysate (Fig. 7B). Moreover, we were unable to observe binding of the divergent B56γ1 peptide to ankyrin-B using identical experimental conditions (Fig. 7B). Thus the COOH terminus of B56α is required and sufficient for interaction with cardiac ankyrin-B.
Ankyrin-B is required for the subcellular targeting of B56α in primary cardiomyocytes.
Based on the established role of ankyrin-B in regulating the localization of diverse ion channels, transporters, and cell adhesion molecules in a host of tissues, we hypothesized that the role of the ankyrin-B/B56α interaction was to target B56α to specific cardiomyocyte domains. To test whether ankyrin-B activity was required for B56α targeting in cardiomyocytes, we evaluated the localization of B56α in neonatal wild-type cardiomyocytes and cardiomyocytes either heterozygous (ankyrin-B+/−) or homozygous (ankyrin-B−/−) for a null mutation in ankyrin-B (71, 78). As previously reported (47, 51), wild-type cultured neonatal cardiomyocytes display a striated expression of ankyrin-B over the M-line (Fig. 8A). As expected (Fig. 2), endogenous B56α labeling shows significant overlap with ankyrin-B (Fig. 8A). As shown in Fig. 8B, reduced expression of ankyrin-B immunostaining is observed in ankyrin-B+/− cardiomyocytes [220-kDa ankyrin-B expression reduced ∼50% (51)] that are accompanied by a striking loss of B56α immunofluorescence at sites where ankyrin-B expression is reduced (Fig. 8B, asterisks). Interestingly, there was no difference in the levels of B56α expression at local cardiomyocyte domains where ankyrin-B expression appeared unaffected (Fig. 8B; note colocalization over remaining M-line striations). Finally, as shown in Fig. 8C, ankyrin-B−/− (null) cardiomyocytes are void of ankyrin-B immunoreactivity. Likewise, consistent with a role for ankyrin-B in B56α targeting, we observe nearly a complete loss of B56α localization to the M-line of ankyrin-B−/− cardiomyocytes. We did, however, observe small regions of B56α staining above background in both ankyrin-B+/− and ankyrin-B−/− cardiomyocytes that were unaffected by the reduction in ankyrin-B expression (see white arrows in Fig. 8, B and C). As shown in supplemental Fig. 1 (neonatal cardiomyocytes double-labeled with ankyrin-B and α-actinin) and as previously reported (49, 50, 52), reduced expression of ankyrin-B levels in the neonatal cardiomyocyte does not affect the expression or localization of cardiomyocyte proteins unrelated to ankyrin-B function. Taken together, these data strongly support a role for ankyrin-B in the targeting of the primary pool of B56α to striated regions in neonatal cardiac myocytes.
We performed “rescue” assays to determine whether the loss of B56α localization in cardiomyocytes with reduced ankyrin-B expression was due specifically to monogenic loss of ankyrin-B. Ankyrin-B+/− cardiomyocytes that display decreased levels and abnormal localization of B56α were transfected with cDNA encoding GFP-220-kDa ankyrin-B. After 56 h, cardiomyocytes were fixed and stained with antibodies against GFP and B56α. As expected, nontransfected cardiomyocytes displayed decreased expression and abnormal localization of B56α (Fig. 9A). In contrast, cardiomyocytes expressing GFP-ankyrin-B displayed B56α distribution similar to those observed in wild-type cardiomyocytes (Fig. 9B). Together, these data demonstrate a crucial role for ankyrin-B in the subcellular localization of the PP2A targeting subunit B56α in primary cardiomyocytes.
This study identifies an ankyrin-dependent pathway required for the targeting of the B56α subunit of the PP2A complex in primary cardiomyocytes. Ankyrin-B and B56α are colocalized in primary cardiomyocytes and associate in yeast as well as in lysates from isolated adult cardiomyocytes. This study identifies the structural requirements on B56α for ankyrin-B association as a 13 amino acid motif at the B56α COOH terminus that is not found in other B56 family proteins. Finally, using wild-type cardiomyocytes and cardiomyocytes with reduced ankyrin-B expression, we demonstrate that ankyrin-B is required for the subcellular targeting of B56α in primary cardiomyocytes. Together, our surprising new findings demonstrate a novel role for ankyrin-B in the subcellular organization of a critical cardiac phosphatase, PP2A.
Total cellular PP2A activity is increased during heart failure (57, 58), and several studies implicate this elevated phosphatase signaling with cardiac myocyte dysfunction. Whereas acute increases in PP2A activity markedly lower the gain in excitation-contraction coupling (15), other studies show that chronic overexpression of the catalytic subunit of PP2A (PP2A/c) results in a dilated cardiomyopathy accompanied by reduced contractility in transgenic mice (19). Yet it is likely that B subunit-mediated targeting of PP2A rather than global changes in phosphatase activity is crucial for cardiac function. For example, the overexpression of a mutant PP2A scaffolding subunit (PP2A/A), which cannot bind B subunits, resulted in a transgenic mouse line with mislocalized cardiac PP2A that was accompanied with a dilated cardiomyopathy (4). Although new information on the subcellular organelles that sequester B56γ1 and B56α in cardiac myocytes is emerging (20), there is little information on the binding partner targets for any B subunits in heart cells. Thus the new results presented here identify, for the first time, a binding partner for B56α in heart that is critical for its subcellular distribution within sarcomeric structures.
An important result in this study was the precise mapping of ankyrin-B binding domain to a small COOH-terminal sequence on B56α. An emerging theme is that the biology of PP2A is specified in part by an intimate association with its substrates. For example, PP2A binds to the β2-adrenergic receptor (60), CaMKIV (80), and to CaMKIIδ (86). There is little information on the particular domains of the PP2A holoenzyme trimeric complex that define such subcellular targeting. Although yeast two-hybrid studies have shown that PP2A/c, and not B, subunits interact directly with the α1-subunit of the L-type Ca2+ channel (11, 22), a more common targeting motif is seen in ryanodine receptor/PP2A interactions where PP2A is tethered through a B-targeting subunit, PR130 (43, 65). Yeast two-hybrid studies have defined distinct binding activities for B56 subunits although the sites have not been defined with precision. For example, the amino-terminal half of B56δ contains a binding site for the transcription factor HAND1 (18), whereas two-thirds of the COOH-terminal region of B56α contains a binding site for the double-stranded RNA-dependent kinase (PKR), which plays a role in translational control (83). The identification of a short, unique 13 amino acid sequence that comprises the ankyrin-B-binding activity on B56α not only defines B56 targeting activity with new resolution but also reveals that ankyrin-B binds only this B56 subunit at its site. Recent analyses of the crystal structure of the PP2A holoenzyme reveal that this COOH-terminal domain projects from the ternary complex, a structure consistent with an ankyrin-B targeting function (82). Taken together, these results provide new insight into the precision of the molecular mechanisms of PP2A targeting in cardiac cells.
Ankyrin-B is critical for normal cardiac function. Decreased ankyrin-B expression in mice or ankyrin-B loss-of-function variants in humans result in cardiac phenotypes including bradycardia, atrial fibrillation, conduction defects, polymorphic ventricular arrhythmia, and sudden cardiac death (51, 52, 69). Abnormal ankyrin-B activity in cardiomyocytes results in abnormal cardiomyocyte Ca2+ homeostasis leading to extrasystoles (51). These cellular phenotypes are accompanied by aberrant targeting of three ankyrin-B-associated proteins: Na/Ca exchanger, Na/K ATPase, and Ins(1,4,5)P3 receptor (8, 47, 51). The relative contribution of the loss of Na/Ca exchanger, Na/K ATPase, and Ins(1,4,5)P3 receptor in the generation of ankyrin-B-based arrhythmia is unknown. In fact, based on mouse models lacking each membrane protein, it is unlikely that the loss of a single component above could generate the phenotypes associated with ankyrin-B dysfunction (24, 25, 54, 61, 66). Whereas the loss of multiple membrane components is likely to exacerbate the phenotype, an alternative explanation for the arrhythmia is the loss of additional regulatory proteins from the ankyrin-B, Na/K ATPase, Na/Ca exchanger, and Ins(1,4,5)P3 receptor protein complex. Ankyrins are phosphoproteins (49), and the activities of Na/K ATPase, Na/Ca exchanger, and Ins(1,4,5)P3 receptor have been previously shown to be regulated by phosphorylation (14, 28, 32). In fact, PP2A specifically has been suggested to regulate the activities of Na/Ca exchanger and Ins(1,4,5)P3 receptor in excitable cells (13, 70). Recently, it has been shown that PP2A associates with the Na/K ATPase at the apical junctional complex in epithelial cells (62), and in alveolar epithelial cells, dephosphorylation of the Na/K ATPase by PP2A allows its recruitment to the plasma membrane (35). Future experiments designed to investigate the functional role for ankyrin-B-dependent recruitment of PP2A will be critical for determining the role of this interaction for normal cardiac electrical activity. Furthermore, future studies should explore how identified human ANK2 SBD arrhythmia variants (e.g., E1425G) affect B56α binding and activity. Finally, we consistently observed a minor population of B56a (<5%) that appeared to be localized in the cardiomyocyte by an ankyrin-B-independent pathway (Figs. 2 and 8). The role of this pool is currently unknown and provides an obvious target for future investigation.
In summary, our findings demonstrate a new functional role for ankyrin-B in the recruitment of a key signaling protein to specialized subcellular domains in cardiomyocytes. Specifically, our new results identify a new role for the ankyrin-B SBD in mediating interaction with PP2A. Our new results demonstrate that peripheral components of ion channel/transporter macromolecular complexes play key roles in the local regulation of membrane signaling and cellular excitability.
This work was supported by National Heart, Lung, and Blood Institute Grants R01-HL-084583 and R01-HL-083422 (to P. J. Mohler) and P01-HL-07079 (to T. B. Rogers).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by the American Physiological Society