Ca2+/calmodulin kinase II–dependent regulation of βIV-spectrin modulates cardiac fibroblast gene expression, proliferation, and contractility

Fibrosis is a pronounced feature of heart disease and the result of dysregulated activation of resident cardiac fibroblasts (CFs). Recent work identified stress-induced degradation of the cytoskeletal protein βIV-spectrin as an important step in CF activation and cardiac fibrosis. Furthermore, loss of βIV-spectrin was found to depend on Ca2+/calmodulin-dependent kinase II (CaMKII). Therefore, we sought to determine the mechanism for CaMKII-dependent regulation of βIV-spectrin and CF activity. Computational screening and MS revealed a critical serine residue (S2250 in mouse and S2254 in human) in βIV-spectrin phosphorylated by CaMKII. Disruption of βIV-spectrin/CaMKII interaction or alanine substitution of βIV-spectrin Ser2250 (βIV-S2254A) prevented CaMKII-induced degradation, whereas a phosphomimetic construct (βIV-spectrin with glutamic acid substitution at serine 2254 [βIV-S2254E]) showed accelerated degradation in the absence of CaMKII. To assess the physiological significance of this phosphorylation event, we expressed exogenous βIV-S2254A and βIV-S2254E constructs in βIV-spectrin-deficient CFs, which have increased proliferation and fibrotic gene expression compared with WT CFs. βIV-S2254A but not βIV-S2254E normalized CF proliferation, gene expression, and contractility. Pathophysiological targeting of βIV-spectrin phosphorylation and subsequent degradation was identified in CFs activated with the profibrotic ligand angiotensin II, resulting in increased proliferation and signal transducer and activation of transcription 3 nuclear accumulation. While therapeutic delivery of exogenous WT βIV-spectrin partially reversed these trends, βIV-S2254A completely negated increased CF proliferation and signal transducer and activation of transcription 3 translocation. Moreover, we observed βIV-spectrin phosphorylation and associated loss in total protein within human heart tissue following heart failure. Together, these data illustrate a considerable role for the βIV-spectrin/CaMKII interaction in activating profibrotic signaling.

Fibrosis is a pronounced feature of heart disease and the result of dysregulated activation of resident cardiac fibroblasts (CFs). Recent work identified stress-induced degradation of the cytoskeletal protein β IV -spectrin as an important step in CF activation and cardiac fibrosis. Furthermore, loss of β IV -spectrin was found to depend on Ca 2+ /calmodulin-dependent kinase II (CaMKII). Therefore, we sought to determine the mechanism for CaMKIIdependent regulation of β IV -spectrin and CF activity. Computational screening and MS revealed a critical serine residue (S2250 in mouse and S2254 in human) in β IVspectrin phosphorylated by CaMKII. Disruption of β IV -spectrin/CaMKII interaction or alanine substitution of β IV -spectrin Ser2250 (β IV -S2254A) prevented CaMKIIinduced degradation, whereas a phosphomimetic construct (β IV -spectrin with glutamic acid substitution at serine 2254 [β IV -S2254E]) showed accelerated degradation in the absence of CaMKII. To assess the physiological significance of this phosphorylation event, we expressed exogenous β IV -S2254A and β IV -S2254E constructs in β IV -spectrin-deficient CFs, which have increased proliferation and fibrotic gene expression compared with WT CFs. β IV -S2254A but not β IV -S2254E normalized CF proliferation, gene expression, and contractility. Pathophysiological targeting of β IV -spectrin phosphorylation and subsequent degradation was identified in CFs activated with the profibrotic ligand angiotensin II, resulting in increased proliferation and signal transducer and activation of transcription 3 nuclear accumulation. While therapeutic delivery of exogenous WT β IV -spectrin partially reversed these trends, β IV -S2254A completely negated increased CF proliferation and signal transducer and activation of transcription 3 translocation. Moreover, we observed β IV -spectrin phosphorylation and associated loss in total protein within human heart tissue following heart failure. Together, these data illustrate a considerable role for the β IV -spectrin/CaMKII interaction in activating profibrotic signaling.
Increased fibrosis, characterized by excessive accumulation of extracellular matrix (ECM) protein, is a critical component of the repair and remodeling response to chronic neurohumoral or biomechanical stress in tissues throughout the body, including heart, lung, kidney, liver, skeletal muscle, and skin (1). In heart, fibrosis is not only important for replacement of damaged/dead myocardium in response to ischemic injury but also promotes electrical and mechanical dysfunction in a wide range of cardiac diseases (2)(3)(4). Cardiac fibroblasts (CFs) are the underlying cell types responsible for fibrotic remodeling. Under baseline (nondiseased) conditions, CFs play an essential supportive role in maintaining ECM homeostasis. Following injury, however, CFs undergo a transition from their basal state to an activated phenotype, characterized by enhanced proliferation, migration, and ECM deposition leading to increased fibrosis (5). Importantly, a diverse array of signaling events, including neurohumoral, biomechanical, and paracrine factors, has been identified in the CF activation process (3), providing opportunities for therapeutic strategies aimed at convergent molecular targets shared among these numerous signaling cascades.
Spectrins are a family of proteins initially identified for their role in providing structural support to cellular membranes (6,7). The family includes two αand five β-spectrin gene products where dimers of αand β-spectrin interact with one another to form heterotetramers to facilitate interaction of membrane proteins with the actin cytoskeleton (8). Since their discovery, our understanding of spectrin function has grown to include critical spatiotemporal regulation of cell signaling events (9). Importantly, novel roles for the β IV -spectrin isoform have been discovered in organizing local signaling domains important for stress-induced cardiac remodeling, including arrhythmia and fibrosis (10)(11)(12)(13). Specifically, β IVspectrin targets a subpopulation of the multifunctional Ca 2+/ calmodulin-dependent protein kinase II (CaMKII) to the cardiac myocyte intercalated disc membrane for regulation of voltage-gated Na + channels. More recently, a broader role for β IV -spectrin/CaMKII has been identified in regulating gene programs in both cardiac myocytes and fibroblasts through interaction with the signal transducer and activation of transcription 3 (STAT3) (12,13). Importantly, β IV -spectrin sequesters STAT3 near the membrane and out of the nucleus under basal (unstimulated) conditions. CaMKII activation in response to long-term pacing in vitro or chronic pressure overload in vivo promotes loss of β IV -spectrin and redistribution of STAT3 with nuclear accumulation, ultimately leading to changes in gene expression.
Here, we tested the hypothesis that β IV -spectrin is a novel target for direct CaMKII phosphorylation with effects on stability/expression of β IV -spectrin. Moreover, given the identified role for CaMKII in regulating cardiac fibrosis, we sought to explore involvement of the β IV -spectrin/CaMKII/STAT3 regulatory nexus in modulating CF gene expression and phenotypes. Using MS guided by a computational screen for consensus motifs, a putative CaMKII phosphorylation site was identified in the β IV -spectrin C terminus (Ser2250 in mouse and Ser2254 in human). Heterologous expression of human β IV -spectrin constructs lacking the putative site (β IV -spectrin with alanine substitution at serine 2254 [β IV -S2254A]) or mimicking constitutive phosphorylation (β IV -spectrin with glutamic acid substitution at serine 2254 [β IV -S2254E]) demonstrated a significance for this site in modulating the rate of β IV -spectrin degradation in vitro. Adenoviral expression of WT and mutant β IV -spectrin constructs in isolated mouse CFs further illustrated the importance for β IV -spectrin phosphorylation in regulating spectrin stability, STAT3 localization, gene expression, cell proliferation, and contractility in genetic models of spectrin deficiency and in response to angiotensin II (AngII). Finally, using a novel custom phospho-specific β IVspectrin antibody, we identified increased β IV -spectrin phosphorylation at Ser2254 in human failing hearts, supporting an important role for this molecular pathway in human disease. These data advance our understanding of the dynamic range of β IV -spectrin function, including orchestration of its own degradation in response to chronic stress with coordinate changes in gene expression and cell function.

Results
Previous results revealed that chronic pressure overload induced prominent loss of β IV -spectrin in WT mice. However, mutant mice expressing truncated β IV -spectrin lacking an identified CaMKII interaction motif (mutant β IV -spectrin allele C-terminal region containing CaMKII interaction motif [qv 3J ] allele) resulted in maintained β IV -spectrin protein expression (12). Furthermore, in vitro studies showed that rapid pacing of WT myocytes under hyperphosphorylating conditions recapitulated loss of β IV -spectrin observed in vivo, which was prevented by the CaMKII inhibitor autocamtide-2-related inhibitory peptide as well as when using qv 3J myocytes (12). Based on these data, we hypothesized that CaMKII directly phosphorylates β IV -spectrin to induce degradation in response to stress. Initial bioinformatics screening of the fragment absent in the qv 3J β IV -spectrin allele revealed several putative  CaMKII phosphorylation motifs found in both mouse and  human, including serine residues 2250, 2268, 2301, 2435, and  2557 (2254, 2272, 2305, 2438, and 2560 in human) (Fig. 1A). Notably, Ser2250/2254 received the highest score as a predicted phosphorylation target (residue highlighted in red, Fig. 1A). We next performed MS analysis on immunoprecipitated β IV -spectrin derived from COS-7 cell lysates transfected with hemagglutinin (HA)-tagged versions of mouse or human β IV -spectrin encoding spectrin repeats 10 through the C terminus. These constructs were coexpressed with or without constitutively active CaMKII. Phosphorylated and nonphosphorylated tryptic peptide fragments were identified corresponding to β IV -spectrin residues 2247 to 2259 in mouse (mβ IV,2247-2259 ) and residues 2251 to 2263 in human (hβ IV,2251-2263 ) (associated tandem MS/MS [MS/MS] spectra are shown in Fig. 1B, Table S1). The percentage of phosphorylated mβ IV,2247-2259 (out of total detected mβ IV,2247-2259 ) increased from 0% in the absence of CaMKII to 36.2% when coexpressed with CaMKII. A similar effect was observed for the human sequence with an increase in phosphorylated hβ IV,2251-2263 from 15.9% to 48.3% in the absence and presence of CaMKII, respectively (Fig. 1D). Notably, human β IV -spectrin achieved higher percentages of phosphorylated peptide at both baseline and in response to constitutively active CaMKII coexpression, suggesting that differences in steady-state levels of phosphorylation may exist between mouse and human β IVspectrin. Importantly, the identified fragment and corresponding phosphorylation site (Ser2250 in the mouse and Ser2254 in human) are highly conserved across species (Fig. 1E), supporting the physiological significance. MS/MS spectra were unable to detect peptide fragments for the other remaining predicted CaMKII target motifs, preventing assessment of potential phosphorylation of additional sites.
Based on our MS results, we generated a rabbit affinitypurified polyclonal antibody for detection of phosphorylated β IV -spectrin Ser2250/2254. The antibody was tested by immunoblot on lysates generated from COS-7 cells transfected with human β IV -spectrin encoding repeats 10 through the C terminus (WT β IV -spectrin) alone, WT β IV -spectrin with constitutively active CaMKII, or a phosphoablated β IV -spectrin derived from the WT β IV -spectrin construct utilizing a serine to alanine mutation at residue 2254 (β IV -S2254A) and coexpressed with constitutively active CaMKII. Increased immunoreactive signal was observed with CaMKII coexpressed with WT β IV -spectrin but not β IV -S2254A (Fig. 2, A and B). In order to more thoroughly evaluate antibody specificity, preadsorption with the antigenic peptide used in antibody generation and affinity purification was performed. As expected, this resulted in the complete loss of signal when treated with the phosphorylated peptide (Fig. S1A). Further antibody characterization was performed by pretreatment of the membrane with alkaline phosphatase to eliminate immunoreactive signal corresponding to phosphorylation of β IV -spectrin. This resulted in signal loss for both WT β IV -spectrin alone (identified to have basal phosphorylation status in MS data; Fig. 1A) and WT β IV -spectrin with CaMKII but marginal loss for β IV -S2254A with CaMKII (Fig. S1, B and C). It is important to note that all validation experiments indicate some degree of crossreactivity with unphosphorylated epitopes. Treatment with the unmodified antigenic peptide also resulted in partial reduction of signal for all evaluated conditions, suggesting partial reactivity with unphosphorylated β IV -spectrin protein. Also, residual immunoreactive signal was apparent in cells expressing β IV -S2254A + CaMKII and in samples pretreated with alkaline phosphatase. Together, these data support the importance of including appropriate controls (i.e., normalization to total β IV -spectrin immunoreactive signal) when using the antibody. That said, these data support the ability of the antibody to detect phosphorylation of β IV -spectrin at S2254.
We next tested the hypothesis that CaMKII-dependent phosphorylation of β IV -spectrin modulates spectrin stability/ expression. We cotransfected COS-7 cells with constitutively active CaMKII and the WT and β IV -S2254A constructs. We also included a qv 3J construct lacking the C-terminal region containing CaMKII binding and phosphorylation site (starting at spectrin repeat 10 as the case for WT and β IV -S2254A). Cells were treated with cycloheximide (CHX) 48 h post-transfection to stop new protein synthesis, allowing for assessment of potential differences in protein degradation. Lysates were collected at baseline, as well as 1 and 2 h post-CHX treatment to evaluate the rate of β IV -spectrin loss. WT β IV -spectrin levels (expressed as percentage of baseline expression) were significantly lower 2 h post-CHX treatment compared with qv 3J or β IV -S2254A (Fig. 2, C and D), without dramatic changes at longer time points (up to 24 h, data not shown). Notably, there was no difference in relative β IV -S2254A and qv 3J expression at 2 h post-CHX. Given that MS results were unable to detect coverage of the other predicted phosphorylation sites (S2272, S2305, S2438, and S2560), a series of phosphoablated mutants were generated to similarly test their functional significance in regulating β IVspectrin loss. Coexpression of these additional mutants with constitutively active CaMKII led to no change in the rate of β IV -spectrin loss compared with WT ( Fig. S2), reinforcing a central role for Ser2254.To subsequently determine whether phosphorylation of Ser2254 was sufficient to induce spectrin degradation, COS-7 cells were transfected with phoshomimetic β IV -spectrin (β IV -S2254E) or WT constructs in the absence of constitutively active CaMKII (GFP cotransfected with β IVspectrin instead as the cotransfected control). Following treatment with CHX, we again assessed levels of β IV -spectrin expression, which identified a significant loss of spectrin with the β IV -S2254E but not WT construct, even in the absence of CaMKII coexpression (Fig. 3, A and B). Together, these data support an important role for Ser2254 phosphorylation in modulating stability/expression of β IV -spectrin.  CFs isolated from β IV -spectrin-deficient mice (mutant β IVspectrin allele lacking repeats 10 through C terminus [qv 4J ] mice expressing truncated β IV -spectrin lacking repeats 10 through the C terminus) were previously shown to have enhanced expression of profibrotic genes, increased proliferation, increased contractility, and associated in vivo fibrosis compared with WT animals (13). Therefore, as a first step in assessing the potential physiological significance of β IV -spectrin phosphorylation at S2254, we evaluated CF phenotype in qv 4J CFs, as a model for β IV -spectrin deficiency. qv 4J CFs were subjected to adenoviral expression of human phosphoablated (Ad.β IV -S2254A) or phosphomimetic (Ad.β IV -S2254E) β IVspectrin or control (Ad.GFP) (Fig. 4A). Mouse and human β IVspectrin share 96% sequence homology, supporting use of human constructs in mouse CFs. Importantly, qv 4J CFs expressing Ad.β IV -S2254A showed a significant reduction in proliferation at both 48 and 72 h post-transduction compared with Ad.β IV -S2254E or control (Fig. 4B). Furthermore, delivery of Ad.β IV -S2254A significantly reduced CF contractility in relation to Ad.β IV -S2254E or Ad.GFP treated cells, as assessed by collagen gel volume compaction (Fig. 4, D and E) as another functional readout for profibrotic activity. At the same time, a significant reduction in select profibrotic genes was observed in Ad.β IV -S2254A expressing CFs compared with Ad.β IV -S2254E or control (Fig. 4C). Interestingly, while CFs expressing Ad.β IV -S2254E showed a similar proliferation rate compared with control, they had significant enhancement of profibrotic gene expression. This change in fibrotic gene expression was specific to the recognized β IV -spectrin/STAT3 pathway, as evaluation of additional fibrosis-associated genes (pdgfra, postn, and vim) previously shown to be unaffected by β IV -spectrin/STAT3 (13) remained unregulated between groups (Fig. S3).
To ensure that observed phenotypic changes with expression of β IV -spectrin constructs were not an artifact of the qv 4J model (expression of truncated β IV -spectrin), we repeated a subset of experiments using our fibroblast-specific β IV -spectrin model utilizing a periostin MerCreMer crossed with β IV -spectrinfloxed mice (inducible fibroblast-specific β IV -spectrin knockout mouse [β IV ifKO]) ( Fig. S4A) (13). Periostin expression is highly specific to activated CFs and can be induced with AngII. Therefore, β IV ifKO and control (β IV -floxed, Cre−) mice    Figure 2. CaMKII-dependent phosphorylation of β IV -spectrin at S2254 promotes degradation. A, representative immunoblots and B, densitometric measurements for phospho-β IV -spectrin (S2254) and total β IV -spectrin from COS-7 cells transfected for 48 h with WT β IV -spectrin ± CaMKII T287D or phosphoablated β IV -spectrin (S2254A) with CaMKII T287D. *p < 0.05 and **p < 0.01 by one-way ANOVA followed by Tukey HSD test (F = 42.67; p < 0.0001); n = 7. C, representative immunoblots (bands at expected size for expressed constructs comprising repeat 10 through the spectrin C terminus) and D, densitometric measurements from COS-7 cells transfected with WT, qv 3J , or S2254A β IV -spectrin coexpressed with constitutively active CaMKII, T287D. Cells were cultured for 48 h after which degradation assays were performed by treating COS-7 cells with CHX (20 μM) to stop new protein synthesis for 1 and 2 h to measure the rate of β IV-spectrin loss. β IV -spectrin expression was normalized against cotransfected CaMKII to account for transfection control. were treated for 7 days with tamoxifen/AngII for activation of the MerCreMer periostin promoter to induce β IV -spectrin KO in CFs (Fig. S4, A and B), as previously described (13). CFs isolated from β IV ifKO mice were then transduced with Ad.β IV -S2254A, Ad.β IV -S2254E, or Ad.GFP. Consistent with the phenotype observed in the qv 4J -derived CFs, Ad.β IV -S2254A treatment reduced contractility (as assessed by collagen gel compaction) compared with Ad.β IV -S2254E and Ad.GFP treated cells. Collectively, these data support a functional role for the phosphorylation of β IV -spectrin at S2254 in its ability to regulate CF gene expression and profibrotic activity.
β IV -spectrin alters CF gene expression by controlling subcellular localization of the transcription factor STAT3 through physical interaction and sequestration at the membrane (12,13). Notably, conditions of β IV -spectrin loss through genetic or acquired means untethers STAT3, resulting in greater nuclear expression and transcriptional activity. To determine whether phosphorylation of β IV -spectrin at S2254, and subsequent changes in β IV -spectrin stability, altered STAT3 localization, immunostaining was performed on qv 4J CFs expressing Ad.β IV -S2254A, Ad.β IV -S2254E, or Ad.GFP. Consistent with previous observations, control (Ad.GFP) qv 4J CFs displayed   CaMKII phosphorylates β IV -spectrin in heart prominent nuclear localization of STAT3 because of the absence of STAT3/β IV -spectrin binding in the qv 4J allele (Fig. 4, F and G). Expression of Ad.β IV -S2254A but not Ad.β IV -S2254E in qv 4J CFs significantly reduced the relative amount of STAT3 in the nucleus compared with control, consistent with proliferation, gene expression, and compaction assay results. Interestingly, Ad.β IV -S2254A was colocalized with STAT3 in a diffuse and perinuclear pattern, whereas Ad.β IV -S2254E showed a distinct punctate pattern (Fig. 4F).
Beyond genetic models of β IV -spectrin deficiency (qv 4J and β IV ifKO), we sought to test whether β IV -spectrin phosphorylation at Ser2250/2254 influenced CF phenotype under pathophysiological stress conditions involving AngII treatment, which alters CF phenotype (including increased proliferation) in part through activation of CaMKII (14)(15)(16)(17)(18). CFs treated with AngII (1 μM) showed an increase in phosphorylated β IVspectrin (expressed as fraction of total) with an overall decrease in total β IV -spectrin expression (Fig. 5, A-C)   determine whether phosphorylation of β IV -spectrin regulated the CF functional response to AngII, WT CFs were transduced with Ad.β IV -WT, Ad.β IV -S2254A, or Ad.GFP prior to treatment with AngII. Consistent with immunoblot data, AngII treatment led to a loss in β IV -spectrin signal in Ad.GFPexpressing CFs (Fig. 5D), together with a significant increase in STAT3 nuclear signal and increased proliferation compared with Ad.GFP-expressing CFs (Fig. 5, D-G). Expression of Ad.β IV -WT significantly reduced AngII-induced translocation of STAT3 into the nucleus and reduced the proliferation rate compared with Ad.GFP-expressing CFs. Interestingly, Ad.β IV -S2254A-expressing CFs were resistant to AngII-induced STAT3 nuclear accumulation with levels comparable to untreated Ad.GFP-expressing CFs (Fig. 5, D and F). Similarly, Ad.β IV -S2254A-expressing CFs treated with AngII showed a similar proliferation rate to untreated Ad.GFP CFs. Together, these studies establish a link between β IV -spectrin phosphorylation, STAT3 localization, and CF profibrotic phenotype. Finally, as a first step in determining whether our findings in mouse translate to human, levels of phosphorylated β IVspectrin were evaluated in ventricular lysates from normal (nonfailing) and failing human hearts. A significant increase in phosphorylated β IV -spectrin (as a fraction of total β IV -spectrin) was observed together with a decrease in overall β IV -spectrin levels in heart failure-compared control (Fig. 6, A-C), suggesting a potential role for this regulatory site in human disease.

Discussion
Here, we sought to identify the molecular mechanism for CaMKII-dependent regulation of β IV -spectrin expression and CF activity. Using a computational screen, MS analysis, and in vitro functional assays, we identified a putative CaMKII phosphorylation site in the β IV -spectrin C terminus conserved across species, including mouse and human (S2250/S2254). We observed that ablation of the CaMKII phosphorylation site in β IV -spectrin (β IV -S2254A) conferred resistance to CaMKIIinduced degradation in vitro, whereas a phosphomimetic construct (β IV -S2254E) demonstrated accelerated loss even in the absence of CaMKII. Using a custom antibody, we observed that AngII treatment of isolated mouse CFs induced phosphorylation of β IV -spectrin together with increased nuclear localization of STAT3, increased proliferation, and contractility. Adenoviral expression of WT β IV -spectrin was able to partially prevent AngII-induced changes in STAT3 localization and proliferation in CFs. Moreover, CFs expressing Ad.β IV -S2254A treated with AngII were almost indistinguishable from untreated control with respect to STAT3 localization and CF phenotype. Finally, we reported an increase in phosphorylated β IV -spectrin in samples from human failing hearts compared with nonfailing controls, suggesting a significant role for β IV -spectrin phosphorylation and subsequent protein loss in pathologic cardiac remodeling. Our findings expand the dynamic nature of β IV -spectrin to include coordination of its own disassembly in the presence of chronic stress. Furthermore, the system is imbued with a mechanism for reporting its fate back to the cell through STAT3dependent changes in gene expression.
Recent studies have identified a role for β IV -spectrin in mediating the fibrotic response in heart. Both cardiomyocyteand CF-specific deletion of β IV -spectrin led to a prominent increase in fibrosis, arrhythmia, and contractile dysfunction (12,13). Conversely, disruption of β IV -spectrin/CaMKII interaction (qv 3J allele) preserved β IV -spectrin expression, abrogated fibrosis, and improved cardiac function in response to pressure overload. Together, these studies intimately tie β IVspectrin expression with the onset of pathologic remodeling through both genetic and acquired disease pathways. Our new findings reveal the mechanistic relationship between β IVspectrin stability and CaMKII via direct phosphorylation of a conserved residue in the β IV -spectrin C terminus. Although the phosphorylation site identified here is distinct from and proximal to (44 amino acids upstream) the previously identified CaMKII-binding site in β IV -spectrin (give residues), these studies cannot rule out the possibility that CaMKII phosphorylation alters CaMKII binding to β IV -spectrin itself with potential disruptions in associated signaling (10). While studies here concentrate on implications of β IV -spectrin phosphorylation/expression related to STAT3 signaling, it will be interesting in the future to assess consequences for CaMKII signaling.
β IV -spectrin expression is closely linked to that of α IIspectrin, which associate together in heterotetrameric  Figure 6. Increased β IV -spectrin phosphorylation in human failing hearts. A, representative immunoblots for phospho-β IV -spectrin (S2254) and total β IVspectrin in human left ventricular nonfailing (NF) and heart failure (HF) tissue samples. B, associated summary data for phospho-β IV -spectrin normalized to total β IV -spectrin and C, total β IV -spectrin normalized to GAPDH. HF samples comprise both ischemic and nonischemic samples (groups combined). *p < 0.05 and **p < 0.01 by Student's two-tailed t test; NF: n = 7, HF: n = 10. Summary data are presented as mean ± SEM.
complexes. Indeed, genetic deletion of α II -spectrin was found to promote marked reductions in β IV -spectrin (19), suggesting a destabilizing effect by the loss of interacting partners. Interestingly, α II -spectrin also experiences protein loss in heart failure, whereas mouse models of cardiac-specific deletion of α II -spectrin increases fibrosis at baseline and disease states (20), reflecting states of remodeling consistent with β IV -spectrin loss. Despite this, it is not clear what role α II -spectrin might play in the CaMKII-dependent degradation of β IVspectrin and subsequent fibrotic remodeling. Is α II -spectrin stability simply dependent on signaling events targeting β IV -spectrin, or might there exist independent regulation of α II -spectrin that could dictate outcomes for β IV -spectrin expression and fibrosis? Overall, this codependent regulation may even suggest that multiple disease pathways are able to converge on this spectrin complex to facilitate activation of common underlying remodeling events. The identification of β IV -spectrin as a novel target for CaMKII adds to a growing list of CaMKII substrates involved in disease remodeling (21)(22)(23)(24). Importantly, CaMKII has long been distinguished as a critical factor in mediating pathologic electrical remodeling, hypertrophic signaling, and fibrosis (22,25,26). Genetic models of CaMKII overexpression result in spontaneous hypertrophy, decreased cardiac function, and increased mortality (27)(28)(29). In contrast, genetic and pharmacologic inhibition of CaMKII provides protection from chronic stress, limiting the development of cardiomyopathy and maintaining cardiac output in both ischemic and nonischemic disease models (30)(31)(32)(33). Functional effects of CaMKII activity are mediated through post-translational modifications to a range of ion-handling proteins such as phospholamban (34), ryanodine receptor (35), and the voltage-gated channels Ca v 1.2 (36) and Na v 1.5 (10,37), as well as more direct hypertrophic drivers like the class 2 histone deacetylases 4 and 5 (38). Notably, our previous investigations found that the β IVspectrin qv 3J mouse with disruption of CaMKII/β IV -spectrin interaction led to reduced fibrosis but maintained a hypertrophic response to chronic pressure overload (12), delineating the multiple etiologies of CaMKII-directed remodeling and the pivotal importance of β IV -spectrin stability with regard to fibrosis. Indeed, direct inhibition of CaMKII in CFs has been shown to reduce CF activation and fibrosis (17,18), further reinforcing the role for a CaMKII/β IV -spectrin axis of regulation.
CaMKII resides at the hub of a vast signaling network and is well equipped to integrate a host of pathophysiological signaling inputs. Neurohumoral signaling, ischemic stress, inflammation, reactive oxygen species, and diabetes are all associated with activation of CaMKII resulting in cardiomyopathy and fibrosis (30,39,40), whereas specific profibrotic cytokines and neurohumoral factors such as AngII (41), interleukin-6 (42), tumor necrosis factor beta (17), and bone morphogenetic protein 4 (43) also provide CaMKII activation. Based on our findings that CaMKII-dependent phosphorylation of β IV -spectrin modulates gene expression and activity of CFs at baseline and in response to AngII, it is interesting to consider the possibility that the CaMKII/β IV -spectrin axis exists as a central common pathway for fibrotic remodeling response to stress.
Our data support that the CaMKII/β IV -spectrin signaling node resides upstream of the transcription factor STAT3, whereas loss of β IV -spectrin has been shown to drive fibrotic remodeling by promoting redistribution of STAT3 into the nucleus and subsequent changes in profibrotic genes (12,13). These data add to the emerging status of STAT3 as a master controller of the fibrotic response (44). Pharmacologic inhibition of STAT3 directly (12,45,46) or genetic ablation of upstream STAT3 activators, such as interleukin-6 (42), have successfully attenuated cardiac fibrosis in both the atria and/or ventricles following ischemia and pressure overload, reducing adverse electrical remodeling and improving heart function. Moreover, STAT3 activation and nuclear translocation has been shown to be upstream of CaMKII activity and can be impaired by direct CaMKII inhibition (42). Importantly, the investigation here is not only supportive of these data but identifies β IV -spectrin as the bridge between these signaling events, showing that stable expression of phosphoablated β IVspectrin is able to reduce nuclear STAT3 and CF proliferation even in the presence of CaMKII activation.
Characteristically, CaMKII activity often leads to altered function for its target substrate; however, we describe here the unexpected consequence of induced protein degradation. While there is precedence for phosphorylation-dependent degradation of target substrates for other protein kinases (47)(48)(49), such a mechanistic role for CaMKII is less established. Up to now, just one such target has been identified in the protein Liprinα1, a regulator for neuronal dendrite maturation (50). Interestingly, CaMKII phosphorylation of Liprinα1 activates a PEST sequence in the target protein, characterized by the enrichment of proline, glutamate, serine, and threonine residues. Such sequences facilitate basal regulation of protein half-lives as well as induced degradation following posttranslational modification, recruiting both calpain and proteosome-dependent degradation pathways. It will be interesting to further analyze the β IV -spectrin C terminus for putative PEST sequences to determine whether a similar mechanism may be involved in the behavior observed here. Importantly, other clues to the mechanisms responsible for CaMKII-dependent spectrin loss may be found in the confocal imaging studies of β IV -spectrin localization. Specifically, phosphomimetic and WT β IV -spectrin stimulated with AngII showed significant punctate patterns of localization (Figs. 4F and 5D), consistent with those for both proteosomal complexes and/or lysosomal bodies (51,52). Notably, these puncta were absent when expressing β IV -spectrin under nondegrading conditions such as phosphoablated β IV -spectrin with AngII (Fig. 5D), highlighting a clear association with Ser2254 phosphorylation status, β IV -spectrin stability, and nuclear STAT3 accumulation.
Overall, this study identifies β IV -spectrin as a novel target of CaMKII and reveals CaMKII/β IV -spectrin signaling to play an important role in modulating the response of CFs to neurohumoral stress. Given the paramount role that CaMKII activation plays in multiple cardiac disease pathways and the established significance for β IV -spectrin in regulating cardiac fibrosis, the β IV -spectrin/CaMKII axis represents a comprehensive event and attractive therapeutic candidate in fibrotic remodeling.

Mouse models
Adult (2-4 months) C57BL/6J male and female WT and truncated β IV -spectrin (qv 4J ) littermate mice were used. The qv 4J mice were obtained from Jackson Laboratory and express a Sptnb4 allele with a spontaneous insertion point mutation at C4234T (Q1358 → Stop) resulting in a premature stop codon proximal to the STAT3-binding region in β IV -spectrin (53,54). Periostin MerCreMer mice (55) were crossbred with β IV -spectrinfloxed mice (12) to obtain tamoxifen-inducible β IV -ifKO mice. Adult male and female β IV -ifKO or control (β IV -spectrin floxed, Cre−) mice were treated with tamoxifen (Milli-poreSigma) dissolved in corn oil (75 mg/kg i.p. injection daily) and AngII (2.16 mg/kg i.p. injection daily) for 1 week to induce MerCreCre expression under regulation of the periostin promoter (56,57). Animal studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH) following protocols that were reviewed and approved by the Institutional Animal Care and Use Committee at The Ohio State University.
Bioinformatics screening of putative CaMKII phosphorylation target sites in β IV -spectrin A C-terminal fragment of the β IV -spectrin protein was evaluated for potential CaMKII target phosphorylation motifs using the Group-Based Prediction System (GPS 3.0) (58). Five potential sites of interest were identified that were conserved between mouse and human β IV -spectrin sequences (residues refer to position in human sequence): Ser2254, Ser2272, Ser2305, Ser2438, and Ser2560.

MS
A fragment of β IV -spectrin from spectrin repeat 10 through the C terminus (representing a natively expressed truncated version of β IV -spectrin) was cloned from mouse and human complementary DNA (cDNA) libraries into pcDNA3.1 with an HA tag for heterologous expression, as previously described (12). HA-tagged β IV -spectrin was expressed in COS-7 cells (American Type Culture Collection catalog no. CRL-1651, RRID: CVCL 0224) and purified using Anti-HA antibody (3724; Cell Signaling) loaded onto TrueBlot agarose beads (00-8800-25; Rockland). A subset of cells were coexpressed with constitutively active CaMKII (phosphomimetic mutation of Thr287 to Asp[T287D]), generated as previously described (12). About 300 ng of β IV -spectrin plasmid was transfected alone or codelivered with 300 ng CaMKII T287D using 2 μl Lipofectamine 2000 (11668030; Thermo Fisher Scientific), according to the manufacturer's instructions. MS and subsequent analysis was performed by the Proteomics and Metabolomics Laboratory, Lerner Research Institute, The Cleveland Clinic Foundation, Cleveland, OH, USA. The protein samples were fractionated on an SDS-PAGE and submitted for analysis, and a small area around the band was cut from the gel. The gel pieces were washed with water and dehydrated in acetonitrile. The bands were then reduced with DTT and alkylated with iodoacetamide prior to the in-gel digestion. All bands were digested in gel by adding 5 μl of 10 ng/μl trypsin in 50 mM ammonium bicarbonate and incubating overnight at room temperature. The peptides that were formed were extracted from the polyacrylamide in two aliquots of 30 μl 50% acetonitrile with 5% formic acid. These extracts were combined and evaporated to <10 μl in Speedvac and then resuspended in 1% acetic acid to make up a final volume of 30 μl for LC-MS analysis. The LC-MS system was a Dionex Ultimate 3000 nano-flow HPLC interfacing with a ThermoScientific Fusion Lumos mass spectrometer system. The HPLC system used an Acclaim PepMap 100 precolum (75 μm × 2 cm, C18, 3 μm, 100 A) followed by an Acclaim PepMap RSLC analytical column (75 μm × 15 cm, C18, 2 μm, 100 A). About 5 μl volumes of the extract were injected, and the peptides eluted from the column by an acetonitrile/0.1% formic acid gradient at a flow rate of 0.3 μl/min were introduced into the source of the mass spectrometer online. The microelectrospray ion source is operated at 2.5 kV. The digest was analyzed using the datadependent multitask capability of the instrument acquiring full scan mass spectra to determine peptide molecular weights and product ion spectra to determine amino acid sequence in successive instrument scans. The data were analyzed by using all collision-induced dissociation spectra collected in the experiment to search the Chlorocebus sabaeus reference sequence protein database and more specifically against the sequence of mouse and human β IV -spectrin as expressed in the pcDNA3.1 construct. In order to more accurately measure the relative abundance of the phosphopeptides across these samples, a second LC-MS/MS experiment was performed.
Tandem mass spectra were extracted using the Spectrum Selector node bundled into Proteome Discoverer 2.4 (Ther-moFisher). Charge state deconvolution and deisotoping were not performed. The data were analyzed using Proteome Discoverer, version 2.4, with the search engine Sequest. The databases used to search the MS/MS spectra included the Reference Sequence C. sabaeus database containing 61,803 entries along with the sequences for mouse Spectrin beta chain (Q62261) and human Spectrin beta chain (Q9H254). An automatically generated decoy database (reversed sequences) was generated for false discovery rate (FDR) calculation. The search was performed looking for fully tryptic peptides with a maximum of three missed cleavages. Oxidation of methionine, acetylation of protein N terminus, and phosphorylation at S, T, and Y were set as dynamic modifications. Carbamidomethylated cysteine was set as a static modification. The precursor mass tolerance for these searches was set to 10 ppm, and the fragment ion mass tolerance was set to 0.6 Da. Peptide and protein identifications were accepted if they could be established at greater than 5.0% probability to achieve an FDR less than 0.1%. Protein identifications were accepted if they could be established at greater than 5.0% probability to achieve an FDR less than 1.0% and contained at least three identified peptides. The positively identified phosphorylated peptides were further subjected to manual inspection with the requirements of a mass measurement less than 5 ppm, and a majority of the ions present were consistent with the identified phosphopeptide. The targeted experiments involve the analysis of specific β IV -spectrin peptides, including the phosphorylated and unmodified forms. The chromatograms for these peptides were plotted based on known fragmentation patterns, and the peak areas of these chromatograms were used to determine the extent of phosphorylation.

Heterologous expression and degradation assays
In order to assess the role of CaMKII activity and β IVspectrin phosphorylation on the stability of β IV -spectrin, COS-7 cells (American Type Culture Collection CRL-1651) were transfected with 300 ng of the HA-β IV -spectrin pcDNA3.1 construct and 100 ng of constitutively active CaMKII (T287D) or enhanced GFP as previously described (12). Transfections were performed using 2 μl Lipofectamine 2000, according to the manufacturer's protocol. HA-β IV -spectrin pcDNA3.1 was also mutated at serine residue 2254 (human βIV-spectrin) as well as additional serine residues 2272, 2305, 2438, and 2560 using the Agilent site-directed mutagenesis kit with primers listed in the following table Underlining indicates mutant bases. Cells were incubated 36 h after transfection, at which time a baseline sample was collected. Remaining wells were then treated with 20 μg/ml CHX in incomplete Dulbecco's modified Eagle's medium (DMEM) to inhibit new protein synthesis. The drug was kept on the cells for 1 and 2 h, at which time cells were collected and lysates prepared for the evaluation of β IV -spectrin expression. About 20 μg of protein was loaded into a 10% polyacrylamide gel. An HA-antibody (catalog no. 3724; Cell Signaling) was used to assess levels of β IV -spectrin and normalized to the cotransfected CaMKII or GFP to account for variations in transfection efficiency.

Isolation of primary mouse ventricular CFs
Mouse CFs were isolated from left and right ventricles under sterile conditions, as described (59). Briefly, mouse hearts were minced in 2 mg/ml collagenase II (Worthington) dissolved in 1× Ham's F-10 buffer (Corning). After digestion, the extract was filtered and centrifuged. The supernatant was discarded, and cells were resuspended in DMEM; 1×, supplemented with 10% fetal bovine serum (FBS), 1% L-glutamine, and 1% penicillin/streptomycin. Cells were allowed to adhere to culture plates for approximately 4 to 5 h prior to media removal containing nonadherent cells (e.g., endothelial, myocytes). Fresh feeding media were replenished, and cells were grown for 5 to 7 days to 80% to 100% confluency at 37 C in 5% CO 2 . All cell experiments were conducted at passage 1.

Adenoviral generation
Adenoviral β IV -spectrin constructs were generated by cloning the HA-tagged β IV -spectrin from the pcDNA3.1 construct into the adenovirus construct Ad.CIG (cytomegalovirus promoter combined with internal ribosome entry site expression of enhanced GFP in the second position and β IVspectrin in the first). The resulting adenoviral plasmid was transfected along with the psi5 vector into CRE8 cells for the production and amplification of packaged viral constructs as described here (60).

CF proliferation assay
CFs were seeded into 12-well culture-treated plates, as described (13). Briefly, cells were adhered for 24 h with serum starvation. At the same time, cells were transduced with adenoviruses (Ad.) Ad.GFP (control), Ad.WT-β IV -spectrin, Ad.β IV -S2254A, or Ad.β IV -S2254E between WT and qv 4J isolated CFs. The next day medium was replaced with 10% FBS/ DMEM/penicillin/streptomycin for qv 4J CFs or 2% FBS/ DMEM/penicillin/streptomycin for WT cells. For qv 4J CFs, cells were trypsinized at 24, 48, and 72 h postplating. WT CFs were treated 100 nM AngII (1158; R&D Systems) and trypsinized at 24 and 72 h post-AngII stimulation. Cell pellets were resuspended in a fixed volume and manually counted using a hemacytometer to calculate total cell numbers. Accuracy and reproducibility of manual counting was previously confirmed through BrdU proliferation assay (Cell Signaling) (13).

Collagen gel formation and macroscopic gel contraction measurements
Type I rat-collagen gels (1 mg/ml) were prepared by mixing 10× PBS, sterile water, acidic rat tail collagen, and 1 M NaOH. Cells were added (100,000 cells/ml) and mixed before gelation. Cell-collagen mixtures were cast into 24-well culture plates and incubated at 37 C in 5% CO 2 for 45 min. After incubation, 1 ml of culture feeding media was added, and the gels were released from wells. Collagen gels were photographed immediately after release and again the next day (16 h). Photographs were analyzed using NIH ImageJ software (61). Specifically, the diameter of each gel was measured in perpendicular directions and then averaged. As previously described, isotropic compaction was assumed to measure the volume ratio of gels before and after compaction (62).

Immunoblotting and immunostaining
Primary ventricular CF lysates were prepared by washing cells in PBS and scraping in PhosphoSafe Extraction Buffer (71296; Millipore) supplemented with Protease Inhibitor Cocktail and analyzed using SDS-PAGE and immunoblotting, as described (10,12,54). Briefly, equal protein loading was achieved using standard bicinchoninic acid protein assay protocols and verified by Ponceau staining of immunoblots. The following antibodies were used for immunoblotting: HA antibody (for heterologous β IV -spectrin) (1:1000 dilution; 3724; Cell Signaling), CaMKII (1:2000 dilution; #A010-56AP; Badrilla), phospho-β IV -spectrin (custom antibody generated by Gen-Script using expression of the peptide sequence RQE{pSRT} VDQPEETARRC for generation in rabbit and affinity purification; 1:1000 dilution), total β IV -spectrin (1:1000; MABN1727; Sigma), GFP (1:1000; 2555; Cell Signaling), and GAPDH (1:5000; 10R-G109A; Fitzgerald). Specificity of custom phospho-β IV -spectrin antibody was tested using the phosphopeptide used in antibody generation and affinity purification as well as the unmodified antigenic peptide sequence as a peptide blocker. Blocking peptides were used at a concentration of 1.1875 μg/ml by incubating with antibody solution overnight at 4 C before application on blots. Partial loss of signal with the unmodified antigenic peptide (Fig. S1) and persistence of signal in our S2254A β IV -spectrin construct ( Fig. 2A) suggest some level of reactivity with epitopes adjacent to the phosphorylation site, as expected of polyclonal antibodies. Further evaluation was conducted by pretreatment of the immunoblots with alkaline phosphatase enzyme (Antarctic Phosphatase; NEB #M0289) at 150 units/ml in reaction buffer 10 mM Tris-HCl (pH 8.0 at 37 C), 5 mM MgCl 2 , 100 mM KCl, 0.02% Triton X-100, and 0.1 mg/ml bovine serum albumin. ImageJ software (61) was used in quantitation of immunoreaction signal.
CFs were seeded for immunostaining in 12-well plates with laminin-coated glass coverslips. Cells were initially seeded in DMEM incomplete media. Experiments using qv 4J -derived CFs were treated with adenovirus at the time of plating, and the media changed the next day to 10% FBS/DMEM/penicillin/streptomycin and left another 48 h (72 h culture total). WT-derived CFs were treated with adenovirus at the time of plating and after 24 h were then treated with control or AngII (100 nM) in 2% FBS/DMEM/penicillin/streptomycin for another 24 h. Cells were then fixed in 4% paraformaldehyde and permeabilized with Triton X-100 for 10 min. Nonspecific binding was blocked using blocking buffer comprising 1.5% bovine serum albumin, 3% goat serum in 1× PBS (Invitrogen) overnight at 4 C. Primary antibodies (β IV -spectrin [1:100; Millipore; N393/76], total STAT3 [1:200; Cell Signaling; 4904]) were prepared in blocking buffer and added for overnight incubation at 4 C. After washing, secondary Alexa Flour 568 antimouse (1:500; Invitrogen; A11004), 488 anti-rabbit (1:500; Invitrogen; A11008), and 633 phalloidin (1:200; Invitrogen; A22284) antibodies were prepared in blocking buffer and incubated for 2 h at room temperature. After washing, 4 0 ,6diamidino-2-phenylindole (DAPI) (Vecta Laboratories) was applied, and dishes were stored at 4 C in the dark until they were imaged using confocal microscopy. Images were acquired on a Nikon Ti2 Microscope in a blinded fashion. Multiple random fields were selected from preparation for analysis. STAT3 localization analysis was conducted by identifying nuclear area through DAPI staining and cell body areas through phalloidin staining. Background noise from acquired images was removed using a 3 × 3 median filter in MATLAB (MathWorks). Threshold values were established from these measurements, and binary images were generated. ImageJ software was then used to identify percent signal within DAPIdefined boundaries, relative to total signal within phalloidindefined boundaries.

Quantitative real-time PCR
Total RNA from mouse CFs was extracted with TRIzol Reagent following manufacturer's instructions. Multiscribe reverse transcriptase (Invitrogen) was used for the first-strand cDNA synthesis (20 ng/μl) using random primers. Quantitative real-time PCRs (qPCRs) were performed in triplicate on cDNA samples in 96-well optical plates using PowerUp SYBR Green Universal PCR Master Mix (Invitrogen) and a Quant-Studio 3 Real-Time PCR System (Applied Biosystems). qPCR data were analyzed using relative standard curve method, and two delta Ct was used to calculate fold changes in relative gene expression. qPCR products were confirmed by melt-curve analysis, amplicon length, and DNA sequencing. Rpl-7 levels were used as a normalization control. Experiments were conducted in technical triplicates. Primers for gene targets are listed herewith.

Human tissue
Left ventricular failing heart tissue was obtained from explanted hearts from patients undergoing heart transplantation at The Ohio State University by the Cardiac Transplant Team. Left ventricular tissue from nonfailing donor hearts (without a history of cardiac dysfunction) not suitable for transplantation was obtained through the Lifeline of Ohio Organ Procurement Organization. All samples were flushed with ice-cold cardioplegic solution and flash frozen within 45 min of removal and kept frozen at −80 C. Frozen samples were ground before treatment with PhosphoSafe Extraction Reagent (71296; Millipore) supplemented with protease inhibitor (P8340; Sigma) and followed up by sonication to facilitate homogenization. The local institutional review board approved the use of human subject tissue. This investigation conforms to the principles outlined in the Declaration of Helsinki. Human hearts used in this study were deidentified and labeled with six digit random codes for reference to clinical descriptions.

Statistics
SigmaPlot 14.0 (SYSTAT) was used for statistical analysis. A 2-tailed t test was used to determine p values for single comparisons. For multiple comparisons, a one-way ANOVA with Tukey honest significant difference post hoc test was used (data presented as mean ± SEM). The null hypothesis was rejected for p value <0.05.

Data availability
The MS proteomics data have been deposited to the Pro-teomeXchange Consortium via the PRIDE (63) partner repository with the dataset identifier PXD025776 and 10.6019/ PXD025776. All remaining data are contained within the article.
Supporting information-This article contains supporting information.