Dysfunction of the β₂-spectrin-based pathway in human heart failure

β₂-Spectrin is altered in human disease and in animal cardiovascular disease models. Dysregulation of β₂-spectrin is associated with heart failure and arrhythmia.

Abstract

β₂-Spectrin is critical for integrating membrane and cytoskeletal domains in excitable and nonexcitable cells. The role of β₂-spectrin for vertebrate function is illustrated by dysfunction of β₂-spectrin-based pathways in disease. Recently, defects in β₂-spectrin association with protein partner ankyrin-B were identified in congenital forms of human arrhythmia. However, the role of β₂-spectrin in common forms of acquired heart failure and arrhythmia is unknown. We report that β₂-spectrin protein levels are significantly altered in human cardiovascular disease as well as in large and small animal cardiovascular disease models. Specifically, β₂-spectrin levels were decreased in atrial samples of patients with atrial fibrillation compared with tissue from patients in sinus rhythm. Furthermore, compared with left ventricular samples from nonfailing hearts, β₂-spectrin levels were significantly decreased in left ventricle of ischemic- and nonischemic heart failure patients. Left ventricle samples of canine and murine heart failure models confirm reduced β₂-spectrin protein levels. Mechanistically, we identify that β₂-spectrin levels are tightly regulated by posttranslational mechanisms, namely Ca²⁺- and calpain-dependent proteases. Furthermore, consistent with this data, we observed Ca²⁺- and calpain-dependent loss of β₂-spectrin downstream effector proteins, including ankyrin-B in heart. In summary, our findings illustrate that β₂-spectrin and downstream molecules are regulated in multiple forms of cardiovascular disease via Ca²⁺- and calpain-dependent proteolysis.

NEW & NOTEWORTHY

β₂-Spectrin is altered in human disease and in animal cardiovascular disease models. Dysregulation of β₂-spectrin is associated with heart failure and arrhythmia.

the spectrin-based cytoskeleton is critical for the biogenesis, maintenance, and regulation of excitable and nonexcitable cells (14). In vertebrates, the spectrin tetrameric lattice is typically comprised of α/β-spectrin heterotetramers (2 α-spectrin genes, 5 β-spectrin genes) linked with actin. Furthermore, through direct interactions with ankyrin family proteins, spectrins integrate the cytoskeleton with a host of membrane ion channels, transporters, and cell adhesion molecules (5, 13, 14). In humans, spectrin gene variants are linked with hereditary spherocytosis (2) and spinocerebellar ataxia (32). In mice, spectrin deficiency is associated with progressive hindlimb paralysis, deafness, and tremor (54). Finally, spectrin-associated ankyrin molecules are linked with a host of phenotypes in humans and animals, including spherocytosis (2), ataxia (70), bipolar disorder, diabetes (24, 36), skeletal muscle disease (4), and cardiovascular disease (12, 62).

While spectrins have been extensively studied in congenital forms of disease, we know little regarding their roles or regulation in more common forms of acquired disease, particularly in heart failure and arrhythmia. In heart, there are two primary β-spectrin gene products (β₂-spectrin and β_IV spectrin). While β_IV spectrin function has been extensively examined over the past 5 yr (19, 29–31, 38, 44), we still know little regarding the potential functions of β₂-spectrin in cardiac function, and essentially nothing is known regarding β₂-spectrin function in common forms of cardiovascular disease.

We recently identified β₂-spectrin as a nodal protein required for the biogenesis and maintenance of the cardiomyocyte submembrane cytoskeleton (62). Mice deficient for cardiac β₂-spectrin display defects in membrane and sarcoplasmic reticulum (SR) ion channel and transporter targeting as well as cytoskeletal organization (62). Furthermore, mice selectively lacking β₂-spectrin in heart [“β₂-spectrin conditional knockout (cKO) mice”] show severe myocyte electrical phenotypes and arrhythmias (62). Finally, human gene variants that alter β₂-spectrin/ankyrin-B interactions are associated with arrhythmia (62). As a first step in defining the regulation of β₂-spectrin in common forms of cardiovascular disease, we analyzed β₂-spectrin regulation in acquired human cardiovascular disease as well as animal heart failure and arrhythmia models. β₂-Spectrin levels were reduced in disparate forms of human cardiovascular disease, including atrial fibrillation and ischemic and nonischemic heart failure. Furthermore, β₂-spectrin levels were reduced in large and small animal models of cardiovascular disease, including canine nonischemic heart failure and murine pressure overload-induced heart failure. Mechanistically, in line with data in other organs, we determined that β₂-spectrin levels are tightly regulated by posttranslational mechanisms, namely Ca²⁺-dependent calpain proteases. In fact, we discovered that Ca²⁺- and calpain-dependent regulation of β₂-spectrin alters downstream binding partners, including ankyrin-B. In summary, our findings illustrate that β₂-spectrin and downstream molecules are highly regulated in multiple forms of cardiovascular disease through mechanisms that include Ca²⁺- and calpain-dependent proteolysis.

MATERIALS AND METHODS

Animals.

All experiments were performed in age-matched (2–4 mo) littermates (56). All animal procedures were approved and were in accordance with institutional guidelines (Institutional Animal Care and Use Committee; Ohio State University).

Statistics.

P values were determined with the unpaired Student's t-test (2-tailed) for single comparisons. Multiple comparisons were analyzed by use of ANOVA. The Bonferroni test was used for post hoc testing. Data were all normally distributed by Shapiro-Wilk test and are presented as means ± SE. If the data distribution failed normality tests with the Shapiro-Wilk test, rank-based ANOVA and the Dunn multiple-comparisons test were performed.

Canine heart failure.

All canine procedures were approved by the Ohio State University Institutional Animal Care and Use Committee (primary investigator Carnes) and conformed to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Before studies, dogs were verified to have normal cardiac function by routine electrocardiograms and echocardiographic examinations during butorphanol tartrate (0.5 mg/kg im) sedation. Dogs had a right ventricle (RV) pacemaker lead implanted in the RV apex, and heart failure was induced by tachypacing for 4 mo as previously described (63). To assess time dependence during the progression of heart failure, echocardiograms and electrocardiogram were measured at baseline, after 1 mo, and 4 mo of pacing in the 4-mo heart failure group. A second group of dogs was RV tachypaced for 1 mo at 180 beats/min, and echocardiograms were measured at baseline and at the end of the pacing protocol as previously reported (52, 53). An age-matched group of healthy dogs was used as controls and studied in parallel. Functional parameters of canines are noted (43). Following experiments, tissue samples were snap-frozen and stored at −80°C.

Human tissue.

For atrial tissue, experimental protocols were approved by the ethics committee of the Medical Faculty Mannheim, University of Heidelberg (no. 2011-216N-MA). Each patient provided written informed consent. During routine cannulation procedures in patients undergoing open-heart surgery for cardiac bypass grafting and valve replacement, the tip of the right atrial appendage was removed and immediately snap-frozen in liquid nitrogen. Appendages were obtained from 12 sinus rhythm and 10 paroxysmal atrial fibrillation patients. The atrial fibrillation group included patients in sinus rhythm at surgery with a history of at least one episode of self-terminating atrial fibrillation lasting <7 days (paroxysmal atrial fibrillation).

For human heart failure samples, left ventricular (LV) tissue was obtained from explanted hearts of patients undergoing heart transplantation at The Ohio State University. Approval for use of human subjects was obtained from the Institutional Review Board of Ohio State University. LV tissue from healthy donor hearts not suitable for transplantation (subclinical atherosclerosis, age, no matching recipients) was obtained through Lifeline of Ohio. The investigation conforms to the principles outlined in the Declaration of Helsinki. As noted (9), samples included the following: nonischemic heart failure: n = 7 (4 male/3 female), age = 52 ± 13 yr, and LV ejection fraction of 14.5 ± 5.2%; nonfailing controls: n = 5 (2 male/3 female), age = 47 ± 12 yr. Ischemic heart failures samples were defined as the presence of any epicardial coronary vessels with ≥75% stenosis or any history of myocardial infarction or coronary revascularization (either percutaneous transluminal coronary angioplasty or coronary artery bypass grafting) (16).

Immunoblots.

Tissue was harvested and immediately placed in ice-cold homogenization buffer (in mM: 25 Tris·HCl, 150 NaCl, 1 EDTA; supplemented with 1:1,000 protease inhibitor cocktail and 1% Nonidet P-40) (51). Following bicinchoninic acid assay (BCA) quantitation, tissue lysates were electrophoresed using the Mini-PROTEAN tetra cell (Bio-Rad) and a 4–15% precast TGX gel (Bio-Rad). Gels were transferred to a nitrocellulose membrane using the Mini-PROTEAN tetra cell (Bio-Rad) and blocked for 1 h at room temperature. Primary antibody incubation was carried out overnight at 4°C. Densitometric analyses were performed using ImageLab software (Bio-Rad). For all experiments, protein values were normalized against an internal loading control validated against the specific pathology. With the exception of human atrial fibrillation samples that use calsequestrin as the standard loading control (11, 12, 66), GAPDH was used. Dominant high-molecular-mass bands (>220 kDa) were quantified in each sample to avoid potential bias toward quantification of low-molecular-mass degradation products.

Antibodies.

Antibodies used were mouse monoclonal anti-Na⁺/Ca²⁺ exchanger-1 (R3F1; Swant), ankyrin-B (36), Ca²⁺/calmodulin-dependent protein kinase (CaMK) IIδ pT287 (Badrilla), anti-voltage-gated Ca²⁺ channel 1.2 (Affinity Bioreagents), α-actinin (Sigma), β₂-spectrin (Abcam), GAPDH (Fitzgerald), ryanodine receptor 2 (Abcam), and sarcoendoplasmic reticulum 2 (Affinity Bioreagents).

Transverse aortic constriction.

Transverse aortic constriction (TAC) was performed as described (23). Briefly, mice were anesthetized (100 mg/kg ip ketamine + 5 mg/kg xylazine), intubated, and placed on a respirator (120 breaths/min, 0.1 ml tidal volume). Anesthesia was monitored by repeated hindlimb response to pinch. Aorta was exposed via a midline sternotomy. A 6.0 Prolene suture was placed around the aorta distal to the brachiocephalic artery. The suture was tightened around a blunted 27-gauge needle placed next to the aorta. The needle was removed, and the chest was closed. Echocardiography was performed at regular intervals for 2 mo to assess cardiac function. With our laboratory protocol, this protocol produces decreased ejection fraction and increased heart-to-body weight ratio compared with sham mice (23). Age-matched littermates were used as sham-operated controls. Hearts were harvested from dead mice (2% Avertin, 20 μl/g ip) via rapid thoracotomy. Adequate anesthesia was monitored via pain response to hindlimb toe pinch.

Echocardiography.

Eight-week-old male C57/BL6 mice weighing >20 g were used for these experiments. Digital images were obtained at a frame rate of 180 images/s. Transthoracic echocardiogram was performed using the Vevo 2100 (Visualsonics). The mice were anesthetized using 2.0% isoflurane in 95% O₂-5% CO₂ at a rate of ∼0.8 l/min. Anesthesia was maintained by administration of oxygen and ∼1% isoflurane. Electrode gel was placed on the ECG sensors of the heated platform, and the mouse was placed supine on the platform to monitor electrical activity of the heart. A temperature probe was inserted in the rectum of the mouse to monitor core temperature of ∼37°C. The MS-400 transducer was used to collect the contractile parameters of the heart in the short-axis M-mode. LV wall thickness [interventricular septum (IVS) and posterior wall (PW) thickness] and internal dimensions at diastole and systole (LVIDd and LVIDs, respectively) were measured. LV fractional shortening [(LVIDd − LVIDs)/LVIDd], relative wall thickness [(IVS thickness + PW thickness)/LVIDd], and LV mass [1.05(IVS thickness + LVIDd + PW thickness)3 − LVIDd3] were calculated from the M-mode measurements.

Proteolysis experiments.

Ca²⁺ and calpain proteolysis experiments were performed on mouse heart and whole brain tissues as adapted from Ref. 59. Wild-type mouse heart and brain homogenates were prepared from freshly dissected tissues. Samples were homogenized in homogenization buffer (10 mM Tris·HCl and 0.32 M sucrose). Tissues were centrifuged at 4°C at 600 revolutions/min for 10 min, and the supernatant was quantitated by BCA (ThermoScientific). Supernatants were diluted with homogenization buffer to a concentration of 6.0 mg/ml. Lysate (45 μl) was added to 45 μl cleavage buffer (40 mM Tris·HCl, 50 mM NaCl, 2 mM DTT; final concentration in reaction: 20 mM Tris·HCl, 50 mM NaCl, 1 mM DTT) with varying concentrations of CaCl₂ (0, 1, or 10 mM CaCl₂; final CaCl₂ concentrations 0, 0.5, and 5 mM). Samples that used MDL-28170 were preincubated with 100 μM MDL-28170 (made in 100% DMSO; Sigma) for 2 min at room temperature before addition of cleavage buffer. Samples were incubated at 37°C for 30 min, and the reactions were stopped with 100 μl Laemmli + β-mercaptoethanol (19:1 dilution). Sample (30 μl) was separated on 4–15% Tris glycine gels, and proteins were transferred to nitrocellulose. Blots were incubated with anti-ankyrin-B, anti-β₂-spectrin (Abcam), and anti-GAPDH (Fitzgerald) in 5% nonfat dry milk/Tween 20-Tris-buffered saline (TTBS) overnight at 4°C, followed by three washes in TTBS. Following incubation in secondary antibody (donkey antimouse, donkey antirabbit; Jackson Laboratories) for 2 h at 4°C, blots were washed in TTBS and developed using a Bio-Rad Clarity ECL kit.

RESULTS

β₂-Spectrin is expressed in human and mouse heart.

As a first step to investigate the role of β₂-spectrin in human cardiovascular function, we evaluated the chamber-specific expression of β₂-spectrin protein in nonfailing human heart. We observed β₂-spectrin in all four chambers of nonfailing human heart, although levels were significantly reduced in atria compared with ventricle (n = 4, P < 0.05; Fig. 1, A and B). In mice, β₂-spectrin was expressed in all four cardiac chambers. β₂-Spectrin protein levels were similar between right and left atria but significantly reduced in RV of mouse heart compared with LV (n = 4, P < 0.05; Fig. 1, C and D). In summary, β₂-spectrin is expressed in all four chambers of nonfailing human and mouse heart, with species-specific differences in relative chamber protein expression.

Fig. 1.

β₂-Spectrin protein expression in human and mouse cardiac chambers. A and B: β₂-spectrin protein levels in heart chambers from nonfailing human heart (n = 4 samples/chamber, *P < 0.05). C and D: β₂-spectrin protein levels in heart chambers from wild-type mouse (n = 4 samples/chamber, *P < 0.05). LV, left ventricle; RV, right ventricle; LA, left atrium; RA, right atrium.

β₂-Spectrin levels are reduced in human atrial fibrillation.

We recently linked dysfunction in the β₂-spectrin-ankyrin-B pathway with congenital human arrhythmia (62). Furthermore, ankyrin-B loss-of-function mutations cause human atrial fibrillation (12), and ankyrin-B protein levels are reduced in atrial samples of patients with atrial fibrillation (12). Because ankyrin-B requires β₂-spectrin for cellular targeting and stability (62), we investigated whether loss of β₂-spectrin may potentially occur upstream of reduced ankyrin-B levels in atrial fibrillation. We examined the expression of β₂-spectrin protein in right atrial tissue from patients with normal sinus rhythm vs. patients with paroxysmal atrial fibrillation. Tissue of patients with documented paroxysmal atrial fibrillation displayed significant reduction in β₂-spectrin protein levels compared with samples from individuals in normal sinus rhythm (n = 12 patients in sinus rhythm; n = 10 patients with paroxysmal atrial fibrillation; P < 0.05; Fig. 2). We observed no change in the expression of calsequestrin 2, a standard loading control not altered in atrial fibrillation (10, 12, 66). In summary, β₂-spectrin protein levels are significantly reduced in right atrial samples of patients with acquired forms of human atrial fibrillation. This loss provides a potential rationale for reduced ankyrin-B expression in human atrial fibrillation (12).

Fig. 2.

β₂-Spectrin levels are reduced in human atrial fibrillation. B: atrial samples of patients with documented paroxysmal atrial fibrillation (pAF, n = 10) display reduced β₂-spectrin protein expression compared with individuals in normal sinus rhythm (SR, n = 12; *P < 0.05). A: calsequestrin was used as a loading control.

β₂-Spectrin protein levels are altered in human heart failure.

To evaluate a broader role for β₂-spectrin in common forms of cardiovascular disease arrhythmia, we next evaluated β₂-spectrin protein levels in LV lysates of patients with nonischemic heart failure compared with lysates from nonfailing hearts. We observed a significant reduction in β₂-spectrin protein levels in LV samples from patients with documented nonischemic heart failure vs. nonfailing heart samples (P < 0.05; Fig. 3B). Activated CaMKIIδ was used as a control for the heart failure phenotype (35, 38, 46) (n = 5 nonfailing; n = 5 nonischemic heart failure; P < 0.05; Fig. 3D). Consistent with loss of β₂-spectrin in human nonischemic heart failure samples, we observed a significant reduction in ankyrin-B protein expression compared with nonfailing human hearts (n = 5 nonfailing; n = 5 nonischemic heart failure; P < 0.05; Fig. 3J). Notably, while we observed a significant decrease in β₂-spectrin protein levels in samples from human nonischemic heart failure (Fig. 3B), β₂-spectrin transcript levels were not significantly different between nonfailing and nonischemic heart failure samples [n = 3/group; P = not significant (NS)]. Thus, β₂-spectrin protein expression is likely reduced in nonischemic heart failure samples by posttranscriptional mechanisms.

Fig. 3.

β₂-Spectrin protein levels are reduced in human heart failure samples. A and B: β₂-spectrin protein expression was significantly decreased in LV samples from human nonischemic heart failure (NIHF) compared with nonfailing (NF) human LV heart samples (n = 5 NF, n = 5 NIHF; *P < 0.05). C–J: cardiac protein expression in human NIHF. In line with prior studies, compared with LV samples from NF hearts, LV samples from patients with documented NIHF displayed a significant reduction in ankyrin-B protein expression while showing a significant increase in Ca²⁺/calmodulin-dependent protein kinase (CaMK) IIδ pT287 expression (*P < 0.05). K and L: β₂-spectrin protein levels are reduced in human ischemic heart failure (IHF) samples. β₂-Spectrin protein expression was significantly decreased in LV samples from patients with ischemic heart failure compared with NF human LV heart samples (n = 5 NF, n = 5 IHF; *P < 0.05).

We next evaluated β₂-spectrin protein levels in LV lysates of patients with ischemic heart failure compared with lysates from nonfailing hearts. Consistent with findings from nonischemic heart failure samples, we observed a significant decrease in β₂-spectrin protein levels in all samples (n = 5 nonfailing; n = 5 failing; P < 0.05; Fig. 3M).

β₂-Spectrin protein levels are reduced in small and large animal models of heart failure.

β₂-Spectrin levels are reduced in human heart failure. However, molecular mechanisms underlying heart disease may differ between human and animal disease models, as well as between large and small animal models. We therefore examined β₂-spectrin protein expression in large and small animal models of heart failure. We first performed analysis of β₂-spectrin levels in LV samples isolated from mice following 6 wk of transaortic constriction [TAC; ejection fraction reduced ∼50% (23)]. This well-characterized disease model provides the ability to assess phenotypes in a nonischemic model of heart failure, including alterations in signaling pathways activated by mechanical stretch (45, 58). Consistent with our findings in human nonischemic heart failure, LV samples from TAC mice displayed a significant decrease in β₂-spectrin protein expression (n = 4/group; P < 0.05; Fig. 4, A and B) compared with sham control animals. Together, β₂-spectrin protein levels are consistently decreased in human heart failure as well as large and small animal models of heart failure.

Fig. 4.

β₂-Spectrin protein expression in canine and mouse models of heart failure. A and B: LV samples isolated from mice following 6 wk of transaortic constriction (TAC) displayed a significant decrease in β₂-spectrin protein expression compared with samples from sham control (wild-type) mice (n = 4 sham, n = 4 TAC; *P < 0.05). C and D: β₂-spectrin protein levels are significantly reduced in LV samples of a chronic canine heart failure model at 1 and 4 mo compared with control heart samples (n = 4 control, n = 4 heart failure at 1 or 4 mo; *P < 0.05).

While small animal rodent models have been widely used to study the molecular mechanisms underlying cardiac disease pathogenesis, there are inherent differences in cardiac physiology between mice and humans (e.g., heart rate, metabolism, and heart size). We therefore used a large animal preclinical model of nonischemic heart failure. This specific canine disease model is valuable as the chronic tachypacing-induced cardiomyopathy for several reasons: 1) it is nonischemic in nature, 2) it does not require significant structural or exogenous hormonal stresses on the heart, and 3) it displays a remodeling process that takes place over a more extended time frame (1–4 mo) compared with most animal disease models (3, 43, 64). Thus, this model provides the ability to assess β₂-spectrin regulation in response to long-term nonischemic stress. Similar to human data, we observed a significant decrease in β₂-spectrin protein expression in LV samples from this canine model of heart failure (3, 43) compared with nonfailing control hearts at both 1 and 4 mo (n = 4/group; P < 0.05; Fig. 4, C and D). In summary, both small and large animal models of heart failure showed significant alterations in β₂-spectrin levels, suggesting a potential common upstream pathway for protein regulation.

β₂-Spectrin protein levels are regulated by Ca²⁺ and calpain.

In other organs, degradation of the spectrin-based cytoskeleton is linked with cellular remodeling (8, 49, 69). Furthermore, this remodeling is tightly linked with cleavage of the spectrin-based cytoskeleton by Ca²⁺-dependent calpain proteases (18, 42, 61). Heart failure is characterized by alterations in cardiac Ca²⁺ handling and elevated reactive oxygen species, both known to activate cardiac calpain proteases (48, 67). We therefore examined the impact of altered Ca²⁺-mediated proteolysis on β₂-spectrin protein expression. Tissue preparations from adult mouse hearts (or brain as positive control) were analyzed after 30 min of exposure to pathological Ca²⁺ concentrations to activate calpain proteases. In response to elevations in Ca²⁺, β₂-spectrin protein levels in both brain and heart lysates were significantly reduced compared with control samples (P < 0.05; Fig. 5) (35). The selectivity of Ca²⁺-dependent activation of calpain in β₂-spectrin degradation was evaluated in proteolysis experiments ± calpain inhibitor MDL-28170 (35). In both heart and brain lysates the Ca²⁺-dependent degradation of β₂-spectrin was blocked by MDL-28170 (Fig. 5). We observed similar results for the β₂-spectrin downstream partner ankyrin-B in both heart and brain (Fig. 5). Thus, β₂-spectrin levels are tightly regulated by local cellular Ca²⁺ concentrations through activation of calpain. Furthermore, these data provide a logical explanation for reduced β₂-spectrin and ankyrin-B protein levels in pathophysiological conditions where Ca²⁺ levels are increased (12, 35, 68).

Fig. 5.

β₂-Spectrin expression is regulated by Ca²⁺- and calpain-dependent proteolysis. A and B: cardiac β₂-spectrin and ankyrin-B levels are reduced in response to elevated Ca²⁺ concentrations (0.5 or 5.0 mM CaCl₂; n = 3/group; *P < 0.05). Moreover, this loss was prevented in the presence of the calpain inhibitor MDL-28170 [100 μM; P = not significant (NS) between control and MDL-28170-treated samples]. C and D: β₂-spectrin and ankyrin-B levels are reduced in brain lysate response to elevated Ca²⁺ concentrations (0.5 or 5.0 mM CaCl₂; n = 3/group; *P < 0.05). This loss was prevented in the presence of the calpain inhibitor MDL-28170 (100 μM; P = NS between control and MDL-28170-treated samples).

DISCUSSION

Dilated cardiomyopathies, heart failure, and arrhythmias are a significant health burden (20, 50). Despite improving medical therapies, cardiac resynchronization therapies, and LV assist devices (1), heart failure remains a global epidemic. The lifetime risk of developing heart failure is 20% (41) with a 5-yr age-adjusted mortality at 59 and 45% for men and women, respectively (39). Furthermore, the global burden of the most common arrhythmia, atrial fibrillation, continues to increase and have untoward effects on patients (17, 68). Thus, the identification of pathways underlying the development and progression of heart failure and arrhythmias is essential for the creation of new diagnostics and treatments.

Emerging data have emphasized the role of cytoskeletal protein modulation in the progression of heart failure (25–28), but as a whole the role of the myocyte cytoskeleton has remained relatively understudied compared with its membrane protein partners (e.g., ion channels, receptors). Our findings provide data regarding the regulation of β₂-spectrin protein expression in both structural and electrical heart disease, specifically demonstrating significant reductions in β₂-spectrin expression in ischemic and nonischemic heart failure and atrial fibrillation. Mechanistically, alterations in cellular Ca²⁺ induce loss of β₂-spectrin levels in heart via proteolytic cleavage by calpains. Our findings illustrate significant remodeling in animal and human disease that underlies several classes of cytoskeletal and membrane protein reorganization in heart failure. Our data demonstrate that β₂-spectrin is a key component of this cytoskeletal network, and lack of spectrin in heart leads to adverse remodeling in mice, culminating in a heart failure phenotype susceptible to arrhythmias and eventual death (62). Notably, humans with gene variants that block interactions between β₂-spectrin and ankyrin-B display arrhythmia phenotypes (62).

An important finding of this study is the link of β₂-spectrin with the development of heart failure. Our laboratory previously demonstrated that, following acute insult or chronic injury, ankyrin-B is posttranslationally degraded by oxidative stress and/or increased Ca²⁺. Furthermore, these phenotypes are blocked by the calpain inhibitor MDL-28170 (35). Our new findings and the work of other groups in noncardiac systems support upstream degradation of β₂-spectrin by Ca²⁺-dependent calpain proteases as the source of this degradation and subsequent cytoskeletal/membrane remodeling. The selective and rapid degradation of spectrins via distinct intracellular pathways was first demonstrated using chicken erythrocytes (69). Later it was demonstrated that an influx of Ca²⁺ leads to rapid cleavage of β₂-spectrin, and in brain prominent calpain cleavage products appear at 166 and 120 kDa (18). In our studies, while we observed parallel changes in ankyrin-B with β₂-spectrin in heart, we did not observe altered expression levels of Na⁺/Ca²⁺ exchanger. This observation may be due to the known presence of ankyrin-dependent and ankyrin-independent populations of Na⁺/Ca²⁺ exchanger (47) or due to secondary remodeling of Na⁺/Ca²⁺ exchanger in response to acute or chronic stress. In fact, work from Goldhaber and Philipson has elegantly demonstrated striking and coordinated remodeling of L-type Ca²⁺ current and inward Na⁺/Ca²⁺ exchanger current pathways in heart (21, 55).

Finally, based on known protein partners, it is exciting to speculate on new nonstructural roles for β₂-spectrin in cardiovascular physiology. β₂-Spectrin has key roles in cardiogenesis, and global loss of β₂-spectrin results in defects in heart development linked with defects in TGF-β signaling (40) (involved in translocation of Smad3 and Smad4) (65). Furthermore, the β₂-spectrin partner α-spectrin is known to have multiple roles in organizing Mena/vasodilator-stimulated phosphoprotein targeting and thus regulation of actin assembly/dynamics (6). Furthermore, α-spectrin displays its own E2/E3 ubiquitin-conjugating/-ligating enzymatic activity (22, 34). Thus, it will be important for future experiments to integrate both signaling and structural analysis to define the diversity of β₂-spectrin functions at both baseline and in disease.

Significance.

Our findings support significant remodeling of cellular cytoskeleton in various disease states. The spectrin pathway is emerging as a new target for therapeutic interventions, and there are early data that suggest that spectrin breakdown products could serve as biomarkers for disease severity and prognosis. Recent data suggest that calpain-2 activation degrades β₂-spectrin into four major fragments (110, 108, 85, and 80 kDa) in rat brain after injury (37). In fact, based on this disease pathway, spectrin degradation products have been proposed as early markers for tissue and organ damage (33, 37, 57, 60). Based on our data and these findings in brain (33, 60), it is exciting to speculate that β₂-spectrin breakdown products may serve as an early diagnostic marker for cardiovascular phenotypes, including myocardial infarction, acute heart failure, or ventricular arrhythmia.

Study limitations.

While a number of findings in our study are descriptive, the findings provide potential new methods to assess both acute and chronic cardiovascular damage, as well as complement small rodent studies performed using β₂-spectrin cKO mice (62). However, it is important to note that the downstream pathways observed in β₂-spectrin cKO mice (62) and β₂-spectrin global null mice [both with complete loss of β₂-spectrin (40)] likely differ from those observed in acquired forms of cardiovascular disease (reduced only partially across disease pathologies). In fact, an important finding from this study is the consistent loss of β₂-spectrin expression across diverse etiologies. These findings may represent compensatory remodeling in these pathologies. Alternatively, these findings may support multiple populations of β₂-spectrin (calpain sensitive vs. calpain resistant) in the myocyte. While our findings are consistent with data in other organ systems, an important future study may be to perform similar experiments in the presence of in vivo calpain inhibitors to analyze the potential preservation of spectrin levels following acute stress. Furthermore, as it appears that loss of β₂-spectrin has a role in congenital and acquired heart disease, it will be important to further define both up- and downstream molecular pathways. Finally, like ankyrins, spectrins are highly regulated in vivo, both by complex alternative splicing events and posttranslational modifications (7). Thus, while our findings support alterations in the dominant high-molecular-mass form(s) of β₂-spectrin, it will be critical for future studies to define the likely large spectrum of β₂-spectrin splice product in heart, as well as how each product is regulated at the organ, cell, and subcellular level.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants HL-114893 (T. J. Hund), HL-084583, HL-083422, and HL-114383 (P. J. Mohler); the James S. McDonell Foundation (T. J. Hund); the American Heart Association (P. J. Mohler); a Robert Wood Johnson Harold Amos Faculty Development Grant and the National Center for Advancing Translational Sciences (TL1TR001069, S. A. Smith).

DISCLOSURES

The authors declare that they have no conflicts of interest with the contents of this article.

AUTHOR CONTRIBUTIONS

S.S., M.M., P.W., N.V., P.F.B., P.M.J., A. Kilic, C.A.C., D.D., M.R., and T.J.H. conception and design of research; S.S., L.H., C.F.K., A. Kempton, L.D., T.W., P.W., N.V., P.M.J., C.A.C., D.D., M.R., and T.J.H. performed experiments; S.S., L.H., C.F.K., A. Kempton, L.D., J.C., T.W., P.W., N.V., A. Kilic, C.A.C., D.D., and T.J.H. analyzed data; S.S., L.H., C.F.K., A. Kempton, L.D., J.C., N.V., A. Kilic, and D.D. interpreted results of experiments; S.S., L.H., C.F.K., L.D., J.C., N.V., and D.D. prepared figures; S.S. and C.F.K. drafted manuscript; S.S., C.F.K., L.D., N.V., P.F.B., P.M.J., C.A.C., M.R., T.J.H., and P.J.M. edited and revised manuscript; S.S., D.D., T.J.H., and P.J.M. approved final version of manuscript.