HDAC6 contributes to pathological responses of heart and skeletal muscle to chronic angiotensin-II signaling

Abstract

Little is known about the function of the cytoplasmic histone deacetylase HDAC6 in striated muscle. Here, we addressed the role of HDAC6 in cardiac and skeletal muscle remodeling induced by the peptide hormone angiotensin II (ANG II), which plays a central role in blood pressure control, heart failure, and associated skeletal muscle wasting. Comparable with wild-type (WT) mice, HDAC6 null mice developed cardiac hypertrophy and fibrosis in response to ANG II. However, whereas WT mice developed systolic dysfunction upon treatment with ANG II, cardiac function was maintained in HDAC6 null mice treated with ANG II for up to 8 wk. The cardioprotective effect of HDAC6 deletion was mimicked in WT mice treated with the small molecule HDAC6 inhibitor tubastatin A. HDAC6 null mice also exhibited improved left ventricular function in the setting of pressure overload mediated by transverse aortic constriction. HDAC6 inhibition appeared to preserve systolic function, in part, by enhancing cooperativity of myofibrillar force generation. Finally, we show that HDAC6 null mice are resistant to skeletal muscle wasting mediated by chronic ANG-II signaling. These findings define novel roles for HDAC6 in striated muscle and suggest potential for HDAC6-selective inhibitors for the treatment of cardiac dysfunction and muscle wasting in patients with heart failure.

a frequent comorbidity in patients with heart failure is exercise intolerance, which is due to cardiac insufficiency as well as skeletal muscle atrophy (cachexia) and dysfunction (15, 16). Muscle atrophy and impaired exercise capacity are associated with reduced quality of life and poor prognosis, and thus there is intense interest in understanding the molecular basis of muscle wasting in the context of chronic heart failure.

A common pathway governing cardiac and skeletal muscle remodeling involves the renin-angiotensin system. Angiotensin II (ANG II) is a secreted peptide that promotes hypertension and has direct effects on the myocardium, inducing pathological cardiac hypertrophy and fibrosis. Drugs that reduce ANG-II synthesis (angiotensin-converting enzyme inhibitors) and block ANG-II receptor signaling (angiotensin receptor blockers) are first-line therapy for heart failure (18). In addition to promoting cardiac dysfunction, ANG II also has been shown to cause muscle atrophy by reducing protein synthesis and stimulating ubiquitin-proteasome-mediated turnover of protein in skeletal muscle (3, 19). The clinical significance of these findings was highlighted by studies showing that angiotensin-converting enzyme inhibitors increase muscle mass and function in elderly individuals (5, 14).

Histone deacetylases (HDACs) catalyze removal of acetyl groups from lysine residues in a variety of proteins and are grouped into four classes (I, II, III, and IV) (8). Class-II HDACs are further divided into two subclasses, IIa (HDACs 4, 5, 7, 9) and IIb (HDACs 6 and 10). Class-IIa HDACs clearly regulate muscle remodeling by functioning as endogenous repressors of cardiac hypertrophy (12), and by promoting skeletal muscle atrophy by stimulating expresssion of E3 ubiquitin ligases such as muscle RING-finger protein-1 (MuRF1) and atrogin-1 (4, 13, 20). In contrast to class IIa HDACs, little is known about the role(s) of class IIb HDACs in muscle. Here, we demonstrate that HDAC6, which is a cytoplasmic HDAC that deacetylates cytoskeletal proteins, functions downstream of ANG II to control adverse remodeling of cardiac and skeletal muscle. HDAC6 null mice have improved systolic cardiac function due in part to increased myofibrillar force generation, and HDAC6 deletion prevents ANG II-mediated skeletal muscle wasting. These findings define novel roles for HDAC6 in striated muscle and suggest potential utility for HDAC6-selective inhibitors for the treatment of adverse muscle remodeling.

MATERIALS AND METHODS

Experimental animals.

Experiments were approved by and performed in accordance with the Institutional Animal Care and Use Committee at the University of Colorado Denver. Ten week-old, male, wild-type (WT) and HDAC6 null mice (sv/129 strain) (7) were infused with 1.0 μg·kg⁻¹·min⁻¹ ANG II (Bachem) for 8 wk using osmotic minipumps (Alzet). Ten-week-old male mice were also employed for transverse aortic constriction (TAC) and sham surgeries. Tubastatin A (50 mg/kg) or vehicle control (50:50 DMSO:PEG-300) was delivered to mice every day by intraperitoneal injection starting the day of minipump implantation. Body composition was measured by dual-energy X-ray absorptiometry using a mouse densitometer (PIXImus2, Lunar).

Hemodynamics.

Echocardiography was performed using a Vevo770 System equipped with a 30-mHz frequency mechanical transducer (VisualSonics), as previously described (21). GraphPad Prism software was used to generate graphs and analyze data. One-way ANOVA with Newman-Keuls post hoc test (P < 0.05) was used to determine statistical differences.

Myofibril mechanical studies.

Frozen left ventricular (LV) biopsies were skinned using Triton X-100 and homogenized. Myofibrils, mounted on a force recording apparatus, were Ca²⁺-activated (pCa 4.5–5.8) and fully relaxed (pCa 9) by fast solution switching at 15°C at an initial sarcomere length of 2.2 μm, as previously described (17). Customized LabView software was used to acquire and generate mechanical data.

Flow cytometery.

Flow cytometry was performed with single cell suspensions of both ventricles using a BD Cantos II Flow Cytometer (5,000 forward and side scattered-gated cells were captured per sample), as previously described (21). Fluorescently conjugated cell surface antibodies were used at a 1:100 dilution: anti-CD45-V500 (BD Biosciences, 561487), anti-CD3-APC-Cy7 (eBioscience, 47-0032), anti-CD11b-APC (eBioscience, 17-0112-81), anti-CD34-PE (BD Biosciences, 551387), anti-CD335 (NKp46)-PB (eBioscience, 48-3351). Data were collected and visualized using FACSDiva software (BD Biosciences).

Fibrosis analysis.

LVs were fixed in paraformaldehyde and stained using picrosirus red dye, as previously described (21). Quantification of picrosirius red staining was completed by determining the average stained pixels² per total pixels² in images of the LV (18 images per animal).

Immunoblotting and quantitative PCR.

Cardiac and skeletal muscle homogenates were prepared as previously described (11). The following antibodies were used for immunoblotting: HDAC6 (Assay Biotek, C0226), HDAC3 (Cell Signaling Technology, cs-3949), acetyl-tubulin (Cell Signal Technology, 3971), total tubulin (Santa Cruz Biotechnology, sc-23948), GAPDH (Santa Cruz Biotechnology, sc-20357), LC-3 (Novus Biologicals, NB100-2220) and calnexin (Santa Cruz Biotechnology, sc-11397).

Quantitative PCR (qPCR) was performed using Absolute qPCR SYBR Green ROX mix (Thermo Scientific) on a StepOne qPCR instrument (Applied Biosystems). The following PCR primers were used: MuRF1 (forward), 5′-GCTATGGAGAACCTGGAGAAGCA-3′; MuRF1 (reverse), 5′-CGGAAACGACCTCCAGACAT-3′; atrogin-1 (forward), 5-GAACATCATGCAGAGGCTGA; and atrogin1 (reverse), 5′-TAGCCGGTCTTCACTGAGC-3′.

RESULTS AND DISCUSSION

To determine whether HDAC6 is involved in cardiac remodeling mediated by chronic ANG-II signaling, 10-wk-old, male, HDAC6 null mice and (WT) littermates were infused with ANG II for 8 wk using osmotic minipumps. WT mice treated with ANG II developed LV systolic dysfunction, as revealed by echocardiographic assessment of ejection fraction (Fig. 1, A and B). In contrast, cardiac function was completely preserved in HDAC6 null mice.

Fig. 1.

Histone deacetylase 6 (HDAC6) deletion improves cardiac function in mice treated chronically with angiotensin II (ANG II). Wild-type (WT) and HDAC6 knockout (KO) mice were treated with ANG II for 8 wk via osmotic minipumps. Control animals were given sham pumps lacking ANG II. M-mode echocardiographic images (A) were used to quantify ejection fraction (EF; B). C: left ventricular (LV) weight-to-tibia length (TL) ratios were determined at necropsy. D and E: fibrosis in LV sections was quantified using picrosirius red dye. F: flow cytometry was used to quantify the presence of the indicated cell populations in ventricles from ANG II-treated mice. Mac, macrophages; Neut, neutrophils; NK, NK cells; Non-Gran, nongranulocytes. G: homogenates of LVs were assessed by immunoblotting with the indicated antibodies. H: autophagic activity in LVs was measured by immunoblotting to detect conversion of LC3-I to LC3-II. LC3-II levels were quantified by densitometry. I: representative LC3 immunoblot. J: WT mice were treated with ANG II for 2 wk and were injected daily with the HDAC6 inhibitor tubastatin A (Tub A) or vehicle control. Tub A rescued ANG II-mediated impairment of LV EF. For B–D, H, and J, values represent means ± SE. *P < 0.05 vs. sham controls.

HDAC6 deletion did not blunt ANG II-mediated cardiac hypertrophy, as evidenced by echocardiography (data not shown) and quantification of LV mass upon animal death (Fig. 1C). Likewise, picrosirius red staining of LV sections revealed that both WT and HDAC6 null mice developed interstitial fibrosis in response to chronic ANG-II treatment, with a trend toward enhanced fibrosis in the absence of HDAC6 (Fig. 1, D and E).

We recently showed that class I HDACs regulate differentiation of fibrocytes, which are bone marrow-derived cells that have features of both monocytes and fibroblasts and are able to adopt a mesenchymal phenotype and contribute to fibrosis in response to pathological stress (21). Fibrocytes are defined by coexpression of CD34 (stem cell marker), CD45 (hematopoietic cell marker), monocyte markers (e.g., CD11), and either collagen or α-smooth muscle actin (mesenchymal markers) (9). To address whether HDAC6 regulates accumulation of fibrocytes or inflammatory mediators in the heart, flow cytometry was performed with single-cell suspensions of ventricles from WT and HDAC6 null mice treated with ANG II for 8 wk. As shown in Fig. 1F, the only CD45⁺ cell population that was altered in the hearts of HDAC6 null mice compared with WT mice was fibrocytes, which were more prevalent in the absence of HDAC6.

Biochemical analyses of LV homogenates confirmed the absence of HDAC6 protein in null mice and a dramatic increase in acetylation of the HDAC6 substrate tubulin in these mice (Fig. 1G). Prior studies in non-muscle cells have defined a role for HDAC6 in the control of authophagy (10). Consistent with this, conversion of LC3-I to LC3-II, which is a hallmark of autophagy, was reduced in LVs of HDAC6 null mice treated with ANG II (Fig. 1, H and I).

To address whether pharmacological inhibition of HDAC6 recapitulates the protective effect of HDAC6 deletion, mice were infused with ANG II in the absence or presence of tubastatin A, a highly selective small molecule inhibitor of HDAC6 (2). Mice treated with tubastatin A and ANG II had significantly improved ejection fraction compared with vehicle-treated controls (Fig. 1J), supporting the notion that HDAC6 catalytic activity impairs systolic function of the heart in the setting of chronic ANG-II signaling.

Mice were subjected to LV pressure overload by TAC to determine whether HDAC6 serves a generalizable role in the control of systolic cardiac function. WT mice developed mild but significant LV dysfunction following 4 wk of TAC (Fig. 2, A and B), as well as marked LV hypertrophy (Fig. 2, C–E). In agreement with findings made with ANG II, HDAC6 null mice were resistant to TAC-mediated LV dysfunction but still developed LV hypertrophy in response to pressure overload. The trend toward elevated LV interstitial fibrosis in response to TAC was similar in WT and HDAC6 null mice (Fig. 2F).

Fig. 2.

HDAC6 deletion improves cardiac function in mice subjected to LV pressure overload. Mice underwent sham surgery or transverse aortic constriction (TAC) and were analyzed after 4 wk. A–D: M-mode echocardiographic images were used to quantify EF, interventricular septum (IVS) thickness and LV posterior wall (LVPW) thickness (d, diastole). E: LV weight-to-TL ratios were determined at necropsy. F: fibrosis in LV sections was quantified using picrosirius red dye. Values represent means ± SE. *P < 0.05 vs. sham controls.

We hypothesized that preservation of LV function in HDAC6 null mice is partly due to alterations in myofibrillar mechanical properties. To address this possibility, myofibril force generation and relaxation were quantified using small bundles (2 to 3 strands) of LV myofibrils connected to a force transducer and a position controller (Fig. 3A). As shown in Fig. 3B, myofibrils from HDAC6-deficient mice treated with ANG II exhibited a dramatic increase in calcium-activated force generation compared with WT controls, even though kinetics of myofibril activation and relaxation were unchanged between WT and HDAC6 null animals (Fig. 3, C–F). Additionally, resting tension of myofibrils in the inactive state was higher in HDAC6 null animals (Fig. 3G), suggesting that either the structural scaffolding of the myofibril is more rigid or actomyosin cross bridges are engaged during the resting state despite the removal of calcium from the solution.

Fig. 3.

HDAC6 (HD6) deletion augments myofibril inotropy. A: representative image of a small bundle of LV myofibrils mounted on a force recording apparatus. B–G: mechanical characteristics of LV myofibrils from WT and HDAC6 null mice treated with ANG II for 8 wk. H: to determine myofibril calcium sensitivity (pCa50) and the Hill coefficient, the ratio of force generated at submaximal (P) and maximal calcium (P0) activation was plotted; multiple calcium concentrations were used (pCa: 9, 5.8, 5.7, 5.6, and 4.5; and pCa = −log [Ca²⁺]). I: immunoblotting was performed to compare the amount of HDAC6 and HDAC3 in total LV homogenates vs. purified myofibrils. GAPDH (cytoplasmic marker) and troponin I (sarcomeric protein) served as controls to confirm myofibril enrichment. *P < 0.05 vs. WT + ANG II.

To further define the mechanistic basis for elevated force observed with myofibrils from HDAC6-deficient mice, pCa50 and cooperativity (Hill Coefficient) were determined. pCa50 is a measure of the avidity with which calcium binds and releases from the troponin complex to promote force generation. In contrast, cooperativity defines the efficiency with which contractile and regulatory proteins of the sarcomere function together to activate actomyosin cross bridges and thereby promote sarcomere contraction and relaxation. Whereas pCa50 was similar between WT and HDAC6 null animals, the Hill coefficient was significantly lower in the HDAC6 null animals (Fig. 3H), indicating a dramatic decrease in cooperativity. These results suggest the possibility that HDAC6 controls how sarcomeric proteins function together to promote actomyosin cross-bridge formation and force generation, akin to how oxygen binding by hemoglobin is related to parameters such as pH and oxygen tension.

Fractionation studies were performed to begin to address whether HDAC6 directly affects sarcomeric proteins. As shown in Fig. 3I, HDAC6 was readily detected in whole LV homogenates from WT mice and, remarkably, HDAC6 also copurified with myofibrils. In contrast, the class I HDAC, HDAC3, was not detected in myofibril preparations. These results suggest the possibility that HDAC6 controls cardiac muscle function by deacetylating a sarcomeric protein(s).

ANG II is known to cause skeletal muscle wasting in humans and in rodents (19). During the course of our studies, it became apparent that HDAC6 null mice treated with ANG II were significantly larger than WT controls (Fig. 4A). WT mice treated with ANG II failed to gain weight during the 8 wk of treatment (Fig. 4, B and C). Strikingly, although HDAC6 null mice initially lost weight upon exposure to ANG II, the final body weight of these animals neared that of vehicle-treated controls at study completion.

Fig. 4.

HDAC6 deletion blocks ANG II-mediated skeletal muscle wasting. A: representative image of WT and HDAC6 knockout mice after 8 wk of treatment with ANG II. B: change in animal body weight over the course of the 8-wk study. C: total mouse body weight at the beginning and end of the study. D–F: dual-energy X-ray absorptiometry (DEXA) scan analysis of mice after 8 wk of treatment with ANG II or sham osmotic minipumps. G: gastrocnemius-to-TL ratios were determined at the time of necropsy. H and I: quantitative PCR analysis of atrogin-1 and muscle RING-finger protein-1 (MURF1) ubiquitin ligase mRNA expression in gastrocnemius homogenates from WT and HDAC6 knockout mice treated with ANG II for 2 or 8 wk. J: autophagic activity in gastrocnemius muscle was measured by immunoblotting to detect conversion of LC3-I to LC3-II. LC3-II levels were quantified by densitometry. K: representative LC3 immunoblot. Values for all graphs represent means ± SE. *P < 0.05 vs. WT + ANG II (H–J).

To further define the differences in body weight between WT and HDAC6 null mice, animals were assessed by dual-energy X-ray absorptiometry (DEXA) scan. As shown in Fig. 4D, whole body lean mass was significantly reduced in WT mice treated with ANG II for 8 wk. However, consistent with body weight measurements, ANG II failed to reduce lean mass in mice lacking HDAC6. Fat mass and bone mineral content were not significantly altered by ANG II or the absence of HDAC6 (Fig. 4, E and F). Increased skeletal muscle mass in HDAC6 null mice was confirmed upon weighing gastrocnemii at the time of necropsy (Fig. 4G). Elevated muscle mass in HDAC6 null mice was associated with reduced expression of atrogin-1 and MuRF1, two ubiquitin ligases that serve a central role in the control of muscle wasting (Fig. 4, H and I). Furthermore, consistent with findings in the heart, autophagy was diminished in skeletal muscle of HDAC6 null mice, as revealed by reduced conversion of LC3-I to LC3-II (Fig. 4, J and K).

There is a paucity of information regarding the function of class-IIb HDACs in striated muscle. We previously demonstrated that HDAC6 catalytic activity is elevated in the heart during pathological hypertrophy (11), and Brundel and colleagues (22) defined a role for HDAC6 in the control of atrial fibrillation. With regard to skeletal muscle, treatment of cultured myoblasts with a pharmacological inhibitor of HDAC6 was shown to impair differentiation of the cells into myotubes (6). Alamdari et al. (1) also showed that levels of HDAC6 protein in skeletal muscle are modestly reduced in the setting of sepsis, and HDAC6 mRNA expression is transiently decreased in skeletal muscle following dexamethasone infusion. Here, we demonstrate that HDAC6 contributes to cardiac dysfunction and skeletal muscle wasting in response to ANG-II signaling. HDAC6 deletion in mice does not block cardiac hypertrophy or fibrosis but dramatically improves myofibril force generation. We show that HDAC6 copurifies with cardiac myofibrils, suggesting a possible role for this HDAC in the control of sarcomere protein acetylation and function.

The molecular basis for the observed protection of skeletal muscle in HDAC6 null mice exposed to ANG II remains unknown, and it should be noted that we cannot rule out the possibility that deletion of this HDAC alters animal feeding behavior. Nonetheless, the current findings justify the need for continued efforts to define the function and regulation of HDAC6 in striated muscle. We acknowledge that a limitation of the current study is the reliance on mice in which the gene encoding HDAC6 was deleted globally. Thus we cannot exclude the possibility that observed phenotypes reflect functions of HDAC6 in muscle and non-muscle cells. Future studies with myocyte-restricted HDAC6 null mice in various models of pathological cardiac and skeletal muscle remodeling should be particularly informative and may suggest novel therapeutic approaches for the treatment of heart failure and muscle wasting.

GRANTS

B. S. Ferguson was supported by American Heart Association Postdoctoral Fellowship 12POST10680000. B. S. Ferguson. and K. B. Schuetze received funding from National Institutes of Health (NIH) T32 Training Grant 5T32-HL-007822-12. S. M. Williams. and J. I. Spiltoir were funded by NIH Predoctoral Fellowship TL5TL1-RR-025778-04. M. Y. Jeong was supported by the Sarnoff Foundation. T. A. McKinsey was supported by NIH Grants HL-116848 and AG-043822.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

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