Cardiac proteasome activity in muscle ring finger-1 null mice at rest and following synthetic glucocorticoid treatment

Abstract

Muscle ring finger-1 (MuRF1) is a muscle-specific E3 ubiquitin ligase that has been implicated in the regulation of cardiac mass through its control of the ubiquitin proteasome system. While it has been suggested that MuRF1 is required for cardiac atrophy, a resting cardiac phenotype has not been reported in mice with a null deletion [knockout (KO)] of MuRF1. Here, we report that MuRF1 KO mice have significantly larger hearts than age-matched wild-type (WT) littermates at ≥6 mo of age and that loss of cardiac mass can occur in the absence of MuRF1. The objective of this study was to determine whether changes in proteasome activity were responsible for the cardiac phenotypes observed in MuRF1 KO mice. Cardiac function, architecture, and proteasome activity were analyzed at rest and following 28 days of dexamethasone (Dex) treatment in 6-mo-old WT and MuRF1 KO mice. Echocardiography demonstrated normal cardiac function in the enlarged hearts in MURF1 KO mice. At rest, heart mass and cardiomyocyte diameter were significantly greater in MuRF1 KO than in WT mice. The increase in cardiac size in MuRF1 KO mice was related to a decrease in proteasome activity and an increase in Akt signaling relative to WT mice. Dex treatment induced a significant loss of cardiac mass in MuRF1 KO, but not WT, mice. Furthermore, Dex treatment resulted in an increase in proteasome activity in KO, but a decrease in WT, mice. In contrast, Akt/mammalian target of rapamycin signaling decreased in MuRF1 KO mice and increased in WT mice in response to Dex treatment. These findings demonstrate that MuRF1 plays an important role in regulating cardiac size through alterations in protein turnover and that MuRF1 is not required to induce cardiac atrophy.

left ventricular hypertrophy is an independent risk factor for a number of cardiac pathologies, including dyssynchrony, myocardial infarction, and heart failure (5, 23, 27, 43). The regulation of cardiac mass is paramount to attenuating the onset of progressive dysfunction and the onset of heart failure. Thus, identification of the molecular mechanisms that control cardiac growth is important in developing therapeutic targets that regulate heart size, possibly leading to improvements in the outcome of patients with pathological hypertrophy.

Muscle-specific ring finger proteins (MuRFs) are part of an important subfamily of ring finger proteins that are specifically expressed in striated muscle and possess E3 ubiquitin ligase activity (37). The isoform MuRF1 has been shown to interact with titin (8, 32, 37), cardiac troponin I (19, 28), β-myosin heavy chain (19), cardiac troponin C, myosin light chain 2, and troponin T (53). Although its role in striated muscle continues to be evaluated, MuRF1 appears to facilitate polyubiquitin chain formation via lysine 48 linkages on targeted substrates, leading to recognition and substrate degradation by the 26S proteasome complex (20). In this capacity, MuRF1 is believed to be involved in muscle atrophy (19).

The manipulation of MuRF1 expression in mouse models has revealed novel findings related to its importance in the regulation of striated muscle mass. Skeletal muscle mass is spared in mice lacking MuRF1 following certain atrophy-inducing conditions (4, 34). With respect to the heart, the genetic deletion of MuRF1 alone has not been shown to lead to any significant alterations in cardiac mass or function at rest but has been shown to induce greater hypertrophy in response to transaortic banding (51). In contrast, deletions of MuRF1 and either MuRF2 or MuRF3 lead to significant cardiac hypertrophy (19, 52). In the case of double MuRF1/MuRF2 deletions, there is no cardiac dysfunction in the mice that survive, although 75% of the mice die as infants due to grossly enlarged hearts (52), while double MuRF1/MuRF3 deletions lead to enlarged hearts and early-onset congestive heart failure (19).

MuRF1 has been hypothesized to regulate skeletal and cardiac size through its regulation of the ubiquitin proteasome system; however, cardiac proteasome activity has not been directly measured in mice with the null deletion of MuRF1 under any condition. Thus the main objective of the present study was to measure cardiac mass, proteasome activity, and functional capacity in wild-type (WT) and MuRF1 knockout (KO) mice at rest and following treatment with the synthetic glucocorticoid dexamethasone (Dex). Our results reveal that the null deletion of MuRF1 leads to physiological hypertrophy of the heart through what appears to be significant decreases in 20S and 26S proteasome activities coupled with an increase in Akt activation and protein synthesis. In response to Dex, a significant loss of cardiac mass was observed in 6-mo-old MuRF1 KO, but not WT, mice as the result of an increase in proteasome activity and a decrease in protein synthesis. These results suggest that MuRF1 may be controlling the activity of the proteasome directly or indirectly and suggest that MuRF1 plays an important role in controlling protein turnover in cardiac muscle.

MATERIALS AND METHODS

Animals.

The generation of mice with a null deletion of MuRF1 is described elsewhere (4). To disrupt the MuRF1 gene, MuRF1 genomic DNA spanning coding exons 1–4 and most of exon 5 was replaced with a LacZ/neomycin cassette. Homozygous KO and WT mice were generated by intercrossing heterozygous MuRF1 mice. Thirty-two 6-mo-old and sixteen 3- to 4-mo-old female WT and MuRF1 KO mice were kept under a 12:12-h light-dark cycle and fed standard diets. All animal procedures were approved by the Institutional Animal Care and Use Committee at the University of California, Davis.

Dex treatment.

Female mice were maintained on a standard chow diet with unlimited water during 28 days of treatment with the synthetic glucocorticoid Dex. Dex (water-soluble; catalog no. D2915, Sigma) was provided in the drinking water at a concentration of 0.03 mg/ml, which should deliver a dose of 3 mg/kg per mouse on the basis of pilot dosing studies in C57Bl/6 mice where daily water consumption was measured. During the course of the treatment, fresh Dex-treated water was provided every other day, at which time water consumption was measured, confirming that the mice were indeed receiving a dose of 3 mg/kg. The effectiveness of the dosing was verified at the end of the experiment by measurement of spleen mass, which decreases in response to glucocorticoids because of the apoptosis of splenocytes.

Echocardiography.

Echocardiography was performed in the week before and on day 29 immediately after Dex treatment. Systolic function was assessed by M-mode echocardiography and two-dimensional analysis using a Siemans Sequoia C512 ultrasound system with an Acuson 15L8 probe. The measurements represented the average of six selected cardiac cycles from at least two separate scans performed in random-blind fashion, with papillary muscles used as a point of reference for consistency in the level of the scan. The measurement definitions are described elsewhere (36). End diastole was defined as the maximal left ventricular diastolic dimension and end systole as the peak of posterior wall motion. Fractional shortening (FS) was calculated from left ventricular dimensions using the following equation: FS = [(EDD − ESD)/EDD] × 100%, where EDD and ESD represent end-diastolic and end-systolic dimension, respectively.

Histology.

Hearts were excised, weighed, placed in relaxation medium (100 mM KCl, 10 mM imidazole, 2 mM EDTA, 5 mM MgCl₂, and 4 mM ATP), fixed in 4% paraformaldehyde for 24 h, and embedded in paraffin. Heart sections were stained with hematoxylin and eosin for evaluation of general histology. Immunohistochemistry was performed with laminin (1:1,000 dilution; Sigma); hematoxylin was used as the counterstain to locate centrally located nuclei. Stained laminin slides were further used for the measurement of myocyte cross-sectional area (μm²). Digital images were obtained under ×400 total magnification and analyzed by Axiovision software (Zeiss). For each muscle, six nonoverlapping regions of the left ventricle were analyzed (∼500 cardiomyocytes/muscle).

ELISA-based measurements of total ubiquitinated and polyubiquitinated proteins.

Heart homogenate samples (1 μg) were incubated overnight at 4°C to optimize binding to the bottom of 96-well ELISA plates (Santa Cruz Biotechnology). On the next day, samples were incubated in blocking buffer (1% BSA and 1× PBS + Tween 20), rinsed three times in 1× PBS, and incubated with anti-ubiquitin (1:3,000 dilution; Santa Cruz Biotechnology) or anti-polyubiquitin (1:2,000 dilution; catalog no. FK1, Biomol). The anti-ubiquitin antibody detects all polyubiquitinated, as well as monoubiquitinated and free ubiquitin, proteins in the sample, while the polyubiquitin antibody detects only polyubiquitinated proteins (see Fig. 3A). After three rinses in 1× PBS + Tween 20, secondary antibody conjugated to horseradish peroxidase was added. Lastly, the TMB substrate was added to initiate a color change reaction proportional to horseradish peroxidase activity. Sulfuric acid (2.5 M) was added to stop the enzyme-substrate reaction. Total ubiquitinated and polyubiquitinated proteins were measured spectrophotometrically at a wavelength of 450 nm. Absorbance values for wells containing 1% BSA were used as background controls. Purified ubiquitin and a pentaubiquitinated chain (Biomol) were used to validate the specificity of the polyubiquitin antibody.

Fig. 3.

A: 20 and 60 ng of purified pentaubiquitin (Penta-Ub, solid bars) were incubated with the polyubiquitinated (FK1) antibody. PBS served as a blank, and BSA (open bars) and ubiquitin (Ub, gray bars) were used as negative controls. FK1 interacted with pentaubiquitin but and not monoubiquitin. B and C: results from ELISAs of heart samples measuring ubiquitinated (Total-Ub) and polyubiquitinated (Poly-Ub) proteins from WT and MuRF1 KO mice following no treatment (control, solid bars) or 28 days of Dex treatment (open bars). Values are means ± SE of 3–4 mice. *Significantly different from control (P < 0.05).

RNA isolation and quantitative PCR.

Total RNA was isolated from mechanically homogenized cardiac tissue in 1 ml of TRIzol reagent. All centrifugation was performed at 12,000 rpm at 4°C. Homogenized samples were centrifuged for 15 min, and the resulting top aqueous layer was transferred to a microcentrifuge tube containing 200 μl of chloroform. After a 15-min incubation period, the samples were centrifuged for 15 min. The top aqueous layer was added to 0.5 ml of isopropyl alcohol, mixed well, and centrifuged for 10 min. The RNA pellet was washed with 75% ethanol, air-dried, and resuspended in diethylpyrocarbonate water for RT-PCR analysis. cDNA was made using a Quantitect RT kit (Qiagen, Valencia, CA). The resulting cDNA was analyzed by quantitative PCR with unlabeled primers in SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) for 40 cycles at an annealing temperature of 59°C. Each sample was run in triplicate. The primer sequences were as follows, forward and reverse, respectively: 5′-GCTGGTGGAAAACATCATTGACAT-3′ and 5′-CATCGGGTGGCTGCCTTT-3′ for MuRF1, 5′-GACTGGACTTCTCGACTGCC-3′ and 5′-TCAGGGATGTGAGCTGTGAC-3′ for muscle atrophy F-box (MAFbx), 5′-ATAGACCATGTAGAAGCCTAGCCTTT-3′ and 5′GGCTTTTATTGTCAGTTACATGCTTTATAG-3′ for metallothionein 2, 5′CAAAATCGCCCTATTCCTCA-3′ and 5′- AGACCCAGCTTCGTTCTCCT-3′ for ribosomal protein L39 (Rpl39). A study identifying housekeeping genes under proliferative conditions identified Rpl39 as a suitable candidate (39). Furthermore, a microarray analysis of Dex-treated WT and MuRF1 KO skeletal muscle, performed in our laboratory, found Rpl39 to be unchanged (unpublished results). Relative expression levels were based on standard curves of plasmid copy numbers for each gene and normalized to Rpl39 expression.

Proteasome activity.

20S and 26S proteasome activity was assayed as previously described (21). All assays were carried out in a total volume of 100 μl in 96-well opaque plates. The final composition of the 20S assay buffer was 250 mM HEPES, 5 mM EDTA, and 0.03% SDS (pH 7.5). The final composition of the 26S assay buffer was 50 mM Tris, 1 mM EDTA, 150 mM NaCl, 5 mM MgCl₂, 50 μM ATP, and 0.5 mM DTT (pH 7.5). Cardiac muscle was homogenized by a Dounce tissue grinder in 26S buffer. Samples were centrifuged for 30 min at 12,000 g. The resulting supernatant was used to assess proteasome activity. The individual caspase-like (β₁-subunit), trypsin-like (β₂-subunit), and chymotrypsin-like (β₅-subunit) activity of the 20S and 26S proteasome was measured by calculating the difference between fluorescence units recorded with and without the specific inhibitors in the reaction medium. The β₁-subunit was initiated by addition of 10 μl of 1 mM Z-Leu-Leu-Glu-7-amido-4-methylcoumarin (Peptides International). The β₂-subunit was initiated by addition of 10 μl of 1 mM Boc-Leu-Ser-Thr-Arg-7-amido-4-methylcoumarin (Bachem). The β₅-subunit assays were initiated by addition of 10 μl of 1 mM succinyl-Leu-Leu-Val-Tyr-7-amido-4-methylcoumarin (Bachem). The β₁-, β₂-, and β₅-subunit assays were conducted in the absence and presence of its respective proteasomal inhibitors: 40 nM Z-Pro-Nle-Asp-al (Biomol), 40 μM epoxomicin, and 10 μM epoxomicin (Peptides International). These substrates were cleaved by the proteasome subunits, releasing free 7-amido-4-methylcoumarin (AMC), which was then detected fluorometrically. Each well contained 20 μg of homogenized heart sample. Released AMC was measured using a Fluoroskan Ascent fluorometer (Thermo Electron) at an excitation wavelength of 390 nm and an emission wavelength of 460 nm. Fluorescence was measured at 15-min intervals for 2 h. All assays were linear in this range, and each sample was assayed in quadruplicate. Proteasome activity is expressed as mean ± SE.

Cathepsin L assay.

The cathepsin L assay was carried out in a total volume of 100 μl in 96-well opaque plates. The final composition of the cathepsin buffer was 200 mM sodium acetate, 2 mM EDTA, and 2 mM DTT (pH 5.5). Cathepsin activity was measured by calculating the difference between fluorescence units recorded with and without its specific inhibitor (cathepsin L inhibitor I, Calbiochem) in the reaction medium. Activity was initiated by addition of 10 μM Z-Phe-Arg-7-amido-4-methylcoumarin (Peptides International). The substrates were cleaved by cathepsin, releasing free AMC, which was detected fluorometrically. Each well contained 34 μg of homogenized heart sample. Released AMC was measured using a Fluoroskan Ascent fluorometer at an excitation wavelength of 390 nm and an emission wavelength of 460 nm. Fluorescence was measured at 15-min intervals for 2 h. All assays were linear in this range, and each sample was assayed in quadruplicate. Cathepsin L activity is expressed as mean ± SE.

Western blots.

Mouse heart samples (25 μg) were prepared in denaturing sample loading buffer, separated by 10% SDS-PAGE, and transferred to a polyvinylidene difluoride membrane. The membranes were incubated overnight at 4°C in 5% milk and 1× Tris-buffered saline-Tween with the following antibodies: polyclonal rabbit Rpt6 (1:1,000 dilution; commercially made and affinity purified by 21st Century Biochemicals), Rpt1 (1:1,250 dilution; Biomol), proteasome activator 28α (PA28α, 1:1,000 dilution; Biomol), β₅-subunit (1:1,250 dilution; Biomol), β-tubulin (1:1,000 dilution; Santa Cruz Biotechnology), phosphorylated and total Akt (1:1,000 dilution; Cell Signaling), phosphorylated and total mammalian target of rapamycin (mTOR, 1:1,000 dilution; Cell Signaling), and eukaryotic translation elongation factor 2 (1:1,000 dilution; Cell Signaling). After three rinses in 1× Tris-buffered saline-Tween, membranes were incubated with corresponding secondary antibodies (Vector). Membranes were visualized with chemiluminiscent substrate (Millipore). Protein expression was normalized to β-tubulin or eukaryotic translation elongation factor 2, which were shown not to change based on Ponceau staining.

Protein synthesis measurements.

Four-month-old female WT and MuRF1 KO mice were treated with Dex for 14 days (n = 3–4/group). On the final day of treatment, an initial blood sample was collected from the tail of each animal at 8 AM. Each animal was then given an intraperitoneal injection of ²H₂O (Sigma) based on the following calculation: (body weight × 0.75) × 0.03. After a 4-h equilibration period, during which food was removed from the cage, animals were anesthetized with 2–4% isoflurane, and a second blood sample was collected via cardiac puncture. The heart and hindlimb muscles were excised, weighed, and frozen in liquid nitrogen. The tissues, along with the two blood samples collected from each animal, were sent to the Mouse Metabolic Phenotyping Center at Case Western Reserve University for determination of the protein fractional synthesis rate (1). These mice were part of a different study examining skeletal muscle atrophy (2a).

Statistical analysis.

Two-way ANOVA or Student's t-test was performed using SigmaStat 3.1 (Systat Software, San Jose, CA) and basic statistics on Microsoft Excel 2007 (Microsoft, Seattle, WA). A paired t-test was performed to measure body mass loss. Results are expressed as means ± SE, with significance set as P < 0.05.

RESULTS

Untreated MuRF1 KO hearts are significantly larger than WT hearts but show no impairment in function.

Initial analysis of 6-mo-old female mice revealed that cardiac mass as a function of wet weight and heart-to-body mass ratio was significantly greater in MuRF1 KO than WT mice (Fig. 1). Examination of older (up to 18-mo-old) mice revealed that cardiac mass was enlarged in the MuRF1 KO mice relative to WT at all ages. In agreement with the increased mass, mean cross-sectional area was significantly greater in cardiomyocytes of MuRF1 KO than of WT mice (Fig. 2). To determine whether cardiac enlargement affected function, echocardiography was performed, and no difference was measured for fractional shortening or ejection fraction in the hearts of WT and MuRF1 KO mice (Table 1). These data suggest that deletion of MuRF1 resulted in physiological hypertrophy of the heart.

Fig. 1.

Cardiac mass was significantly greater in female muscle-specific ring finger protein 1 (MuRF1) knockout (KO) than age-matched wild-type (WT) mice. A and B: heart wet weight and heart-to-body weight ratio (HW:BW) in WT (solid bars) and MuRF1 KO (open bars) mice. Values are means ± SE of 3–16 mice per group. †Significantly different from WT (P < 0.05).

Fig. 2.

Cross-sectional area of cardiomyocytes increased significantly following 28 days of dexamethasone (Dex) treatment in WT but not MuRF1 KO mice. A: representative laminin-stained cross sections from hearts of WT and MuRF1 KO mice following no treatment [control (Con)] or 28 days of Dex treatment. Images were analyzed at ×400 total magnification. B: cardiomyocyte area in WT and MuRF1 KO mice following no treatment (control, solid bars) or Dex treatment (open bars). Values are means ± SE of 4 mice per group. *Significantly different from control (P < 0.05). †Significantly different from WT (P < 0.05).

Table 1.

Left ventricular measurements by echocardiography M-mode analysis

	Control		Dex Treatment
	WT	KO	WT	KO
LVEDD, mm	3.05 ± 0.09	3.23 ± 0.11	2.53 ± 0.08*	2.98 ± 0.11
LVESD, mm	1.18 ± 0.24	1.3 ± 0.147	0.93 ± 0.08	1.3 ± 0.07
PWd, mm	0.88 ± 0.11	0.83 ± 0.09	0.68 ± 0.06	0.85 ± 0.06
PWs, mm	1.35 ± 0.10	1.4 ± 0.19	1.3 ± 0.08	1.3 ± 0.07
FS, %	61.18 ± 6.17	60.78 ± 4.00	64.38 ± 3.25	56.28 ± 1.67
EF, %	92.28 ± 3.38	92.95 ± 1.86	94.85 ± 1.13	91.00 ± 0.97

Values are means ± SE of 4 mice per group. Dex, dexamethasone; WT, wild-type mice; KO, muscle-specific ring finger-1 protein knockout mice; LVEDD and LVESD, left ventricular end-diastolic and end-systolic dimension; PW_d and PW_s, posterior wall thickness in diastole and systole; FS, fractional shortening; EF, ejection fraction.

^*Significantly different from control.

Proteasome activity is significantly lower in untreated MuRF1 KO mice.

As an E3 ubiquitin ligase, MuRF1 is responsible for the modification of selected substrates through the addition of mono-, di-, or polyubiquitin chains, leading to altered protein function or degradation by the 26S proteasome. Thus, using an ELISA-based assay, we measured the amount of total ubiquitination and polyubiquitination in the hearts of WT and MuRF1 KO mice. The specificity of the polyubiquitin antibody to exclusively recognize polyubiquitinated proteins was confirmed by its specific activity to purified pentaubiquitin (Fig. 3A). The amount of total ubiquitination and polyubiquitination was similar in WT and MuRF1 KO hearts under resting conditions (Fig. 3, B and C).

Since protein degradation is mediated, to a large degree, through the activation of the ubiquitin proteasome pathway in striated muscle, 26S (ATP-dependent) and 20S (ATP-independent) proteasome activity was measured. The 26S proteasome is composed of an inner 20S core flanked by two 19S subunits on each end. Degradation is thought to occur at the catalytic β₁-, β₂-, and β₅-subunits of the inner 20S core, which possess caspase-, trypsin-, and chymotrypsin-like cleavage, respectively. Measurement of 20S and 26S catalytic activities revealed that, with the exception of 20S β₅-subunit activity, the activity of all other 20S and 26S catalytic proteolytic subunits was significantly lower in untreated MuRF1 KO than WT hearts (Fig. 4).

Fig. 4.

Catalytic subunit activity levels of 20S (A) and 26S (B) proteasomes in hearts of WT and MuRF1 KO mice in response to no treatment (control, solid bars) or 28 days of Dex treatment (open bars). Activities of the 3 catalytic subunits, β₁ (caspase-like), β₂ (trypsin-like), and β₅ (chymotrypsin-like), are expressed relative to untreated WT control. Values are means ± SE of 4 mice per group. *Significantly different from control (P < 0.05). †Significantly different from WT (P < 0.05).

The larger heart mass in MuRF1 KO hearts is correlated with increased Akt phosphorylation.

Given the greater mass and absence of cardiac dysfunction in the MuRF1 KO mice, we examined the Akt/mTOR pathway. Phosphorylation of Akt was significantly higher in the hearts of control KO than control WT mice (Fig. 5). No difference in mTOR phosphorylation (Fig. 5) or glycogen synthase kinase 3β phosphorylation (data not shown) was found between untreated WT and KO mice.

Fig. 5.

A: representative Western blots from WT and MuRF1 KO hearts probed with antibodies against phosphorylated Akt (P-Akt), total Akt, phosphorylated mammalian target of rapamycin (P-mTOR), total mTOR, and eukaryotic translation elongation factor 2 (EEF2). WT and MuRF1 KO samples were run on the same gel. B and C: phosphorylated-to-total Akt ratio, phosphorylated and total Akt protein expression, phosphorylated-to-total mTOR ratio, and phosphorylated and total protein expression in hearts of WT and MuRF1 KO mice in response to no treatment (control, solid bars) or Dex treatment (open bars). Protein expression (optical density) relative to WT control is shown. Phosphorylated and total forms of Akt and mTOR were normalized to EEF2 expression. Values are means ± SE of 3–4 mice per group. *Significantly different from control (P < 0.05). †Significantly different from WT (P < 0.05).

Dex treatment induces cardiac atrophy in 6-mo-old MuRF1 KO mice.

After characterizing the differences between 6-mo-old WT and MuRF1 KO hearts at rest, we evaluated the response of WT and MuRF1 KO hearts to extended Dex treatment. Measurement of spleen mass demonstrated that WT and MuRF1 KO mice received equivalent and effective doses of Dex over the 28-day treatment period (Table 2). Dex treatment induced a significant loss of body weight in WT (13%) and MuRF1 KO (8%) mice (Table 2). A significant loss of skeletal muscle mass also was observed in WT and MuRF1 KO mice after 28 days of Dex treatment (Table 2).

Table 2.

Body and organ masses of Dex-treated 6-mo-old WT and MuRF1 KO mice

	Control		Dex Treatment
	WT	KO	WT	KO
Body mass, g	23.33 ± 1.27	26.80 ± 0.94	21.79 ± 0.89‡	25.66 ± 1.15‡
Heart mass, mg	115 ± 4.74	159 ± 7.64†	114 ± 3.49	139 ± 4.72*
Heart-to-body mass ratio, mg/g	4.97 ± 0.27	5.92 ± 1.44†	5.22 ± 0.28	5.47 ± 0.22
Gastrocnemius mass, mg	139 ± 5.2	164 ± 2.86†	105 ± 4.04*	125 ± 6.29*
Spleen mass, mg	82.7 ± 6.6	81.7 ± 1.6	39.9 ± 1.1*	40.9 ± 1.41*

Values are means ± SE of 6–8 mice per group.

^*Significantly different from control within a genotype (P < 0.05).

^†Significantly different from control WT (P < 0.05).

^‡Significant effect of Dex treatment on body weight (P < 0.05, by paired t-test).

Dex treatment for 28 days induced a 12.5% loss in cardiac mass in MuRF1 KO mice but had no effect on cardiac mass in WT mice (Table 2). Under resting conditions, the heart-to-body mass ratio was significantly higher in MuRF1 KO than WT mice. After Dex treatment, the heart-to-body mass ratio trended upward in WT mice due to a significant loss of body weight and no change in heart mass. A greater percent loss of heart mass (12.5%) than body mass (4%) resulted in a small decrease in the heart-to-body mass ratio in MuRF1 KO mice. The effects of Dex on cardiac structure were further evaluated by measurement of cardiomyocyte cross-sectional area; mean cardiomyocyte cross-sectional area was significantly increased by Dex treatment in WT hearts but was not changed in MuRF1 KO hearts (Fig. 2).

Past studies showed that exogenous glucocorticoid use is associated with negative effects on cardiac wall thickness and end-diastolic dimensions (51), cardiac output (44), and systolic function (3). After 28 days of Dex treatment, all animals underwent echocardiography to assess changes in wall thickness and systolic function. Dex treatment caused no major physiological dysfunction in the hearts of WT or MuRF1 KO mice. The only significant change was a decrease in left ventricular end-diastolic dimensions in the hearts of WT mice (Table 2).

MuRF1 and MAFbx expression do not increase in the heart following chronic Dex treatment.

Analysis of mRNA expression by quantitative PCR revealed no significant change in MuRF1 expression in WT hearts after 28 days of Dex treatment. Furthermore, MAFbx expression was unchanged in WT and KO hearts following Dex treatment (Fig. 6). Expression of metallothionein 2, a Dex-responsive gene (35), significantly decreased in WT and MuRF1 KO hearts following Dex treatment, demonstrating that the Dex dose used in this study was sufficient to induce a response at the mRNA level of selective genes.

Fig. 6.

mRNA expression of muscle atrophy F-box (MAFbx), MuRF1, and metallothionein 2 (MT2) in hearts of WT and MuRF1 KO mice in response to no treatment (control) or 28 days of Dex treatment. Data are expressed relative to WT control. *Significant difference between control and Dex-treated mice (P < 0.05).

Proteasome activity is altered in WT and MuRF1 KO mice in response to Dex treatment.

After 28 days of Dex treatment, total ubiquitination was significantly lower in the MuRF1 KO, but not WT, hearts (Fig. 3). In contrast, polyubiquitination was unaltered in WT and KO mice following Dex treatment (Fig. 3). Because Dex treatment significantly reduced MuRF1 KO cardiac mass, alterations in proteasome activity were evaluated. Dex treatment significantly decreased β₁- and β₅-subunit activity in the 20S and 26S proteasomes of WT mice (Fig. 4). Conversely, in MuRF1 KO hearts, Dex treatment significantly increased 20S β₅-subunit and 26S β₂-subunit activity. All other proteasomal proteolytic activities were unchanged following Dex treatment.

In addition to proteasome activity, the expression of several proteins associated with the proteasome was measured in the hearts of WT and MuRF1 KO mice (Fig. 7). Expression of Rpt1 and Rpt6, proteasomal ATPases that are components of the 19S proteasome, was used to assess the total amount of 19S proteasome in the heart. Rpt1 and Rpt6 expression was unchanged in response to Dex in WT and MuRF1 KO mice. β₅-Subunit expression was used to assess the total amount of 20S proteasome in the heart. Expression of the β₅-subunit was significantly lower in WT mice after Dex treatment, which correlated with the observed decrease in β₅-subunit activity. In contrast, in MuRF1 KO mice, β₅-subunit protein expression was significantly higher following treatment and correlated with the observed increase in the β₅-subunit. PA28α, an activator of the 20S proteasome (41), was significantly decreased only in WT hearts following Dex treatment. Overall, these findings suggest that deletion of MuRF1 enhanced proteasome activity in the heart in response to synthetic glucocorticoids.

Fig. 7.

A: Western blots from WT and MuRF1 KO hearts probed with antibodies against Rpt1, Rpt6, PA28α, β₅-subunit, and β-tubulin. Four control and treated samples from WT and MuRF1 KO mice were run on separate gels. Representative blots show 3 samples each from control and treated WT and KO mice. White lines indicate where lanes have been spliced together in formatting of the representative gel. B–E: Rpt1, Rpt6, PA28α, and β₅-subunit protein expression (optical density) relative to untreated control groups, in hearts of WT and MuRF1 KO mice in response to no treatment (control, solid bars) or Dex treatment (open bars). Expression of Rpt1, Rpt6, and β₅-subunit was normalized to β-tubulin expression. Values are means ± SE of 4 mice per group. *Significantly different from control (P < 0.05).

Cathepsin L activity is unchanged in response to Dex.

Cathepsin L is a lysosomal endopeptidase that has been shown to be upregulated at the mRNA level in response to Dex treatment (15, 31). In this study, cathepsin L activity did not change following Dex treatment and was similar between WT and MuRF1 KO mice (data not shown).

Markers of protein synthesis decrease in Dex-treated MuRF1 KO hearts.

Dex treatment induced a significant increase in Akt phosphorylation in WT hearts but caused a significant decrease in Akt phosphorylation in KO hearts (Fig. 5). The ratio of phosphorylated to total mTOR did not change in WT hearts in response to Dex. However, the amount of phosphorylated and total mTOR significantly decreased in the hearts of MuRF1 KO mice following Dex treatment (Fig. 5). These data suggest a decrease in protein synthesis in the hearts of MuRF1 KO mice in response to Dex treatment.

Cardiac atrophy is not observed in 4-mo-old MuRF1 KO mice following 14 days of Dex treatment.

The response of 6-mo-old WT and MuRF1 KO hearts to 28 days of Dex treatment was opposite to that previously reported (51). Therefore, we examined the response of WT and MuRF1 KO hearts from 4-mo-old mice to 14 days of Dex treatment. The hearts were taken from mice used in another study that was designed to examine the effects of short-term Dex treatment on skeletal muscle atrophy (2a). Similar to previous results (51), we found that, following 14 days of Dex treatment, WT hearts underwent significant atrophy, whereas no loss of mass was observed in the hearts of MuRF1 KO mice (Fig. 8).

Fig. 8.

Effect of 14 days of Dex treatment on cardiac wet weight (A), fractional synthesis rate (FSR, B), activity of β₅-catalytic subunit in 20S and 26S proteasomes (C), and Akt and mTOR phosphorylation (D) in 4-mo-old WT and MuRF1 KO mice in response to no treatment (control, solid bars) or 14 days of Dex treatment (open bars). Western blots were obtained from WT and MuRF1 KO hearts probed with antibodies against total and phosphorylated Akt and mTOR. Activity of β₅-catalytic subunit in 20S and 26S proteasomes is expressed relative to WT control. Values are means ± SE of ≥3–4 mice per group. *Significant (P < 0.05) difference between control and Dex-treated mice within a genotype. †Significantly different from control WT.

Selected analyses on hearts from 4-mo-old WT and MuRF1 KO mice were used for comparison with the 6-mo-old mice. It has been reported that the β₅-catalytic subunit has the highest proteolytic capacity among the proteasomal subunits (26), and since these activities changed in response to 28 days of Dex treatment in the 6-mo-old mice, we chose to measure only 20S and 26S β₅-subunit proteasome activities following 14 days of Dex treatment. A significant decrease in 20S β₅-subunit activity and no change in 26S β₅-subunit activity were observed in Dex-treated WT hearts, while no significant changes were observed in the MuRF1 KO hearts (Fig. 8).

At 4 mo, untreated MuRF1 KO hearts were also significantly larger than WT hearts (Fig. 8). Comparable to hearts from 6-mo-old mice, Akt phosphorylation was significantly higher in untreated MuRF1 KO than WT hearts at 4 mo of age. No significant change was measured for Akt or mTOR phosphorylation after 14 days of Dex treatment in WT or MuRF1 KO mice.

Prior to tissue collection, 4-mo-old animals were treated with ²H₂O for measurement of fractional rates of protein synthesis in cardiac and skeletal muscles. Fractional rates of protein synthesis significantly decreased in MuRF1 KO mice following 14 days of Dex treatment (Fig. 8).

DISCUSSION

The MuRF1 gene has been shown to associate with sarcomeric contractile and regulatory machinery (8, 19, 28, 32, 37, 53) and appears to serve an important role in mediating striated muscle atrophy (4, 7, 10, 33, 45). Although many of its roles continue to be explored, as an E3 ubiquitin ligase, MuRF1 is thought to influence muscle size primarily through the regulation of the ubiquitin proteasome system. Thus one prediction is that proteasome activity would be suppressed in mice with a null deletion of MuRF1 under stressful conditions and possibly at rest; however, proteasome activity has not been reported in hearts of MuRF1 KO mice under any condition. Furthermore, no significant resting cardiac phenotype has been reported in mice with the deletion of only MuRF1. This could be related to the finding that MuRF1 expression becomes significant only postnatally (38) and, thus, has a limited role in cardiac development or that the other MuRF1 family members, MuRF2 and MuRF3, can compensate for the loss of MuRF1 in the adult. Alternatively, it could be that most studies examined mice at relatively young ages and did not evaluate a late phenotype. In studying the effects of MuRF1 deletion on skeletal muscle atrophy, we noticed that the hearts of MuRF1 KO mice were generally larger than the hearts of their WT littermates. Therefore, the primary objective of this study was to determine whether alterations in proteasome activity could explain the adult cardiac phenotype observed in the MuRF1 KO mice. Additionally, we investigated whether changes in proteasome activity could explain the recent observation that MuRF1 KO mice were resistant to cardiac atrophy in response to the synthetic glucocorticoid Dex (51).

The main findings of this study are that 1) deletion of MuRF1 results in physiological hypertrophy of the heart that is evident at ∼6 mo of age and persists with age and 2) cardiac atrophy can occur in the absence of MuRF1. A novel finding is that the variable response of hearts in WT and MuRF1 KO mice appears to be related to differential changes in pathways controlling the ubiquitin proteasome system as well as protein synthesis.

The regulation of muscle size is related to the balance between protein degradation and synthesis. It is estimated that the entire array of proteins in the heart is replaced every 30 days (9), and as much as 80% of intracellular protein turnover is performed by the proteasome (29). Thus very small changes in proteasome activity can have drastic effects on the overall rate of myocardial protein turnover, potentially leading to altered heart function (22). Under resting conditions, significant differences were found in cardiac mass between the WT and MuRF1 KO mice at 6 mo of age; thus we examined whether proteasome activity was altered in mice with a null deletion of MuRF1. The activity of five of the six catalytic proteasome subunits was significantly lower in MuRF1 KO than WT hearts, which could partly explain the significantly greater heart mass in MuRF1 KO mice. Interestingly, proteasome activity is not suppressed in the skeletal muscles of MuRF1 KO mice at rest (unpublished observations).

Next, we measured proteasome activity in the hearts of WT and MuRF1 KO mice following chronic glucocorticoid treatment. We found that, following 28 days of Dex treatment in 6-mo-old animals, only the hearts from MuRF1 KO mice exhibited significant loss of cardiac mass, which is contrary to a previous report (51). After 28 days of Dex treatment, activities of the β₁- and β₅-subunits in the 20S (42% and 30% decrease, respectively) and 26S (18% and 20% decrease, respectively) proteasomes were significantly decreased in WT mice. Conversely, significant increases in activity of the β₅-subunit in the 20S proteasome (10%) and the β₂-subunit in the 26S proteasome (15%) were found in Dex-treated MuRF1 KO mice. The changes in β₅-subunit activities were associated with changes in protein expression of the β₅-subunit. Furthermore, in Dex-treated WT mice, there was a decrease in the amount of PA28α, a regulator of 20S proteasome activity (41), which could contribute to the decrease in proteasome activity. Since the β₅-subunit possesses the highest proteolytic capacity among the proteasomal subunits (26), significant changes in β₅-subunit activity alone could account for the maintained heart mass in the WT mice and the significant decrease in heart mass in the MuRF1 KO mice.

Given that MuRF1 is associated with muscle atrophy and increases in ubiquitin-proteasome system (UPS)-mediated degradation in skeletal muscle, deletion of MuRF1 might be predicted to lead to a decrease in basal ubiquitination or a decrease in 26S proteasome activity. Under resting conditions, we found reduced proteasome activity in MuRF1 KO mice; however, contrary to our expectations, the activity of several proteasome subunits increased in the MuRF1 KO mice but decreased in the WT mice in response to Dex treatment. The response of the UPS to synthetic glucocorticoids is mainly derived from literature evaluating rat skeletal muscle and is generally thought to increase, although few studies have actually measured proteasome subunit activity (2, 11). Furthermore, there are limited data for the response of cardiac muscle to long-term glucocorticoid treatment. Proteasome activity has previously been shown to be depressed in mice transitioning to heart failure (50) and in aging hearts (6) but increased in the hypertrophic response to aortic band-induced pressure overload (14). MuRF1 may thus play a role in regulating cardiac mass and contribute significantly to the protein quality control in the heart. In a recent review (46), it was suggested that reactivation of UPS activity could be an appropriate therapy, especially in cases of heart failure. The current study demonstrates that lack of MuRF1 does not prevent changes in proteasome activity and, in fact, can lead to increases in proteasome activity under certain conditions.

The Dex dose given in our study was sufficient to induce loss of body weight and splenocyte apoptosis, with no impact on food or water consumption (data not shown). Thus, 3 mg/kg is an effective dose of Dex in mice, even though proteasome activity was not increased in WT mice. After 28 days of Dex treatment, loss of skeletal muscle mass was not significantly different between WT and MuRF1 KO mice. This finding contrasts with our recent finding of skeletal muscle sparing in 4-mo-old MuRF1 KO mice after 14 days of Dex treatment (3 mg/kg; Ref. 2a). Interestingly, proteasome activity was not increased in WT or KO skeletal muscle in response to Dex treatment. Thus glucocorticoids do not always affect proteasome activity and have more often been shown to adversely affect protein synthesis (25, 54). Furthermore, it appears that skeletal and cardiac muscle may respond differently to synthetic glucucorticoids.

Recently, Willis et al. (51) reported cardiac atrophy in 3- to 4-mo-old C57Bl/6 WT mice in response to 14 days of Dex treatment. In contrast, they reported no cardiac atrophy in MuRF1 KO mice in response to the same glucocorticoid treatment and concluded that expression of the MuRF1 gene was critical for inducing atrophy of cardiac muscle. The contradictory findings between our study and the study of Willis et al. could be related to a number of experimental differences, including sex (females in this study vs. 50:50 male-female mix) and the dose (3 mg/kg vs. 5 mg/kg) and route (oral vs. subcutaneous injection) of Dex administration. While these differences must be considered, we believe that the most likely factors are the difference in the age of the animals and the duration of the treatment. In our initial study, 5- to 6-mo-old mice were used, and our results in WT mice are consistent with results reported in the literature for more mature adult mice (31). Our studies using 3- to 4-mo-old mice and a shorter duration of Dex treatment produced results similar to those reported by Willis et al. Thus the data suggest that the response of the heart to Dex may be age-dependent. However, the additional analysis of protein synthesis and the proteasome suggests that the length of exposure to Dex may also be a variable.

We attempted to identify the mechanisms responsible for the loss of cardiac mass in 3- to 4-mo-old WT mice after Dex treatment. In response to Dex treatment, fractional protein synthesis rates trended upward in WT hearts and significantly decreased in MuRF1 KO hearts. Similar to the older WT hearts, 20S β₅-subunit activity significantly decreased in the young WT hearts in response to Dex treatment. However, 20S β₅-subunit activity in the MuRF1 KO mice was unchanged after 14 days of Dex treatment, which contrasts with the significant increase observed after 28 days. These responses to 14 days of Dex treatment are consistent with the insignificant change in WT heart mass and a decrease in mass in the KO hearts following 28 days of Dex treatment. However, the wet weight data showed the exact opposite. We cannot explain the discrepancy; however, it is possible that the early loss of mass in the WT hearts is due to a decrease in blood volume, and with extended time on Dex treatment, the cardiac mass would be normal.

Since we observed physiological hypertrophy of MuRF1 KO hearts, we looked for changes in Akt-mediated signaling, which is often associated with physiological, but not pathological, cardiac hypertrophy (9, 12, 13, 17, 24). Protein synthesis rates and selected proteins involved in protein translation have been shown to be elevated in mice with the null deletion of MuRF1 and MuRF2 (52). Here, we show that activation of Akt is significantly higher in the resting hearts of MuRF1 KO than WT mice. In examining other components of the pathway (mTOR, S6 protein kinase 1, and glycogen synthase kinase 3β), we did not detect significant increases in activation under resting conditions (data not shown). The increased Akt activation could also be affecting other pathways, such as the FOXO transcription factors or angiogenesis, leading to changes in cardiac size (42, 47, 48).

Mixed results have been reported for the effect of exogenous glucocorticoid treatment on Akt/mTOR signaling in the heart, i.e., decreases (40) and no change (18, 30) in Akt phosphorylation. In this study, Akt/mTOR signaling differed in WT and MuRF1 KO mice in response to Dex treatment. Akt activation significantly increased in WT hearts and significantly decreased in MuRF1 KO hearts after Dex treatment. Although no change in mTOR activation was observed in either group, the absolute amount of phosphorylated and total mTOR was significantly lower in Dex-treated MuRF1 KO hearts. The change in Akt activation in WT hearts following Dex treatment supports the increased cross-sectional area observed in these hearts. The decreases in Akt activation and mTOR protein levels in MuRF1 KO hearts following Dex treatment suggest a decrease in protein synthesis, which likely contributed to the loss of cardiac mass. The lack of a change in cross-sectional area in the MuRF1 KO hearts is difficult to explain; however, there are previous reports of changes in cardiac mass with no change in myocyte cross-sectional area (16, 49, 55). Alternatively, changes in mass could be reflected in alterations in the length of the cardiomyoctes or loss of myocytes, which were not measured in this study.

In summary, this study reports, for the first time, that deletion of MuRF1 can lead to physiological hypertrophy of the heart in unchallenged mice, likely due to changes in proteasome activity and protein synthesis. The present data also demonstrate that cardiac atrophy can be induced in mice with a null deletion of MuRF1. This study provides strong support for the hypothesis that alterations in proteasome activity are important in regulating cardiac mass under physiological and stressful conditions. Interestingly, these data reveal that proteasome activity can increase in the absence of MuRF1, suggesting that MuRF1 may be controlling the activity of the proteasome directly or indirectly. These data highlight the need to identify the in vivo substrates of MuRF1 and to further examine how MuRF1 might be controlling protein turnover through regulation of synthesis and degradation.

GRANTS

This work was supported by National Institutes of Health Grants DK-75801 (S. C. Bodine) and HL-096819 (A. V. Gomes). Partial support for D. T. Hwee was provided by training fellowships from the Howard Hughes Medical Institute-Integrating Medicine into Basic Science (56006769) and the National Heart, Lung, and Blood Institute (T32 HL-086350). Portions of this work were supported by the Mouse Metabolic Phenotyping Center at Case Western Reserve University.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.