Zonisamide attenuates pressure overload-induced myocardial hypertrophy in mice through proteasome inhibition

Abstract

Myocardial hypertrophy is a pathological thickening of the myocardium which ultimately results in heart failure. We previously reported that zonisamide, an antiepileptic drug, attenuated pressure overload-caused myocardial hypertrophy and diabetic cardiomyopathy in murine models. In addition, we have found that the inhibition of proteasome activates glycogen synthesis kinase 3 (GSK-3) thus alleviates myocardial hypertrophy, which is an important anti-hypertrophic strategy. In this study, we investigated whether zonisamide prevented pressure overload-caused myocardial hypertrophy through suppressing proteasome. Pressure overload-caused myocardial hypertrophy was induced in mice by trans-aortic constriction (TAC) surgery. Two days after the surgery, the mice were administered zonisamide (10, 20, 40 mg·kg⁻¹·d⁻¹, i.g.) for four weeks. We showed that zonisamide administration significantly mitigated impaired cardiac function. Furthermore, zonisamide administration significantly inhibited proteasome activity as well as the expression levels of proteasome subunit beta types (PSMB) of the 20 S proteasome (PSMB1, PSMB2 and PSMB5) and proteasome-regulated particles (RPT) of the 19 S proteasome (RPT1, RPT4) in heart tissues of TAC mice. In primary neonatal rat cardiomyocytes (NRCMs), zonisamide (0.3 μM) prevented myocardial hypertrophy triggered by angiotensin II (Ang II), and significantly inhibited proteasome activity, proteasome subunits and proteasome-regulated particles. In Ang II-treated NRCMs, we found that 18α-glycyrrhetinic acid (18α-GA, 2 mg/ml), a proteasome inducer, eliminated the protective effects of zonisamide against myocardial hypertrophy and proteasome. Moreover, zonisamide treatment activated GSK-3 through inhibiting the phosphorylated AKT (protein kinase B, PKB) and phosphorylated liver kinase B1/AMP-activated protein kinase (LKB1/AMPKα), the upstream of GSK-3. Zonisamide treatment also inhibited GSK-3’s downstream signaling proteins, including extracellular signal-regulated kinase (ERK) and GATA binding protein 4 (GATA4), both being the hypertrophic factors. Collectively, this study highlights the potential of zonisamide as a new therapeutic agent for myocardial hypertrophy, as it shows potent anti-hypertrophic potential through the suppression of proteasome.

Introduction

Pathological myocardial hypertrophy is a common compensatory reaction to prolonged hypertension caused by excessive collagen accumulation which induces cardiac stiffness. The shift from compensated myocardial hypertrophy to a decompensated state during stressful events increases heart-disease risks, including arrhythmias, myocardial infarction, or heart failure. As a predictor of heart failure, myocardial hypertrophy is an independent cardiac risk factor resulting in increased mortality and morbidity worldwide [1–5]. Current anti-hypertrophic drugs depend on the pharmacological suppression of key receptors related to neuroendocrine stimulation such as angiotensin II, catecholamines, and aldosterone. Despite the ability of pharmacological agents currently in clinical use to regulate activities of the neurohumoral axis, the incidence rate of heart failure continues to rise [6]. Based on these clinical observations, novel therapeutic strategies are urgently needed.

The ubiquitin-proteasome system (UPS) maintains intracellular protein homoeostasis. It is an adenosine triphosphate (ATP)-dependent proteolytic system resulting in the ubiquitination of target proteins through a set of enzymatic reactions and protein degradation by the 26 S proteasome, which is the central protease of the UPS [7]. The 26 S proteasome comprises a 20 S core protease and 19 S regulatory protein. The 19 S regulatory protein, with subunits (RPT) 1 and 4, recognises, deubiquitinates, and unfolds target proteins. The 20 S protease is the proteolytic centre of the proteasome and responsible for target protein degradation through chymotrypsin-, trypsin-, and caspase-like proteolytic activities, which links to the proteasome subunit beta type-5 (PSMB5), type-2 (PSMB2), and type-1 (PSMB1) [8]. The accumulation of polyubiquitinated proteins and proteasome dysfunction has been observed in many cardiac disorders [9–14]. A previous study has also shown that inhibition of the proteasome effectively alleviated myocardial hypertrophy [15]. Our previous study has shown that the proteasome inhibition mitigates myocardial hypertrophy, which was correlated to glycogen synthesis kinase 3 (GSK-3) activation in pressure overload rats subjected to abdominal aortic constriction (AAC) and hypertrophic primary neonatal rat cardiomyocytes triggered by angiotensin II (Ang II) [16]. GSK-3 is a key factor for regulating myocardial hypertrophy stimulated by pressure overload [17]. Myocardial hypertrophic stimuli activate the Akt (protein kinase B, PKB) pathway, leading to GSK-3 inactivation, resulting in reduced cardiac performance [18]. GSK-3 activity is also regulated by the liver kinase B1/AMP-activated protein kinase (LKB1/AMPKα) pathway. Inactivation of AMPKα attenuates pressure overload-induced myocardial hypertrophy by modulating GSK-3 activity [19]. The downstream myocardial hypertrophic factors of GSK-3 are phosphorylated extracellular signal-regulated kinase (ERK) [20] and nuclear translocation of GATA binding protein 4 (GATA4) [21].

Zonisamide is a 1,2-benzisoxazole 3-methanesulfonamide originally synthesised in Japan in 1974. It has been approved for the treatment of seizures since 1989 [22, 23]. A series of experiments provided evidence that zonisamide acts through the blockade of voltage-gated sodium channels and inhibition of voltage-gated T-type calcium channels [24]. Our previous study using zonisamide showed that it attenuates myocardial hypertrophy induced by pressure overload in vitro and in vivo [25]. In addition, we also found that zonisamide prevents myocardial hypertrophy in high glucose-triggered primary neonatal rat cardiomyocytes and in type 2 diabetic mice [26]. Little is known about whether zonisamide acts on proteasome in the context of myocardial hypertrophy induced by pressure overload. Therefore, this study aims to explore the relationship between the cardioprotective effects of zonisamide and proteasome function.

In this study, we investigated whether the cardioprotective effects of zonisamide are related to the proteasome and GSK-3; we also evaluated the underlying mechanisms in a myocardial hypertrophic mouse model established using trans-aortic constriction (TAC) and Ang II-treated primary neonatal rat cardiomyocytes. Our data showed that zonisamide exerted a suppressive effect on proteasome activity but a stimulatory effect on GSK-3 activity, thus preventing myocardial hypertrophy.

Materials and methods

Reagents

Zonisamide (Z) and 18α-GA were obtained from Selleck Chemicals (Houston, Texas, USA). Ang II was obtained from Sigma-Aldrich (St. Louis, MO, USA). Synthetic fluorogenic proteasome substrates were obtained from R&D Systems (Minneapolis, MN, USA). The primary antibodies used are listed in Table 1. The horseradish peroxidase (HRP)-conjugated goat anti-rat and anti-rabbit immunoglobulin G (IgG) were obtained from Abcam (Cambridge, MA, USA).

Table 1.

Primary antibodies used in this study.

Antibody	Type	Dilution	Source	Cat. No.
α-MHC	pAb	1:1000 for WB	Proteintech	22281-1-AP
ANP	mAb	1:1000 for WB	Santa Cruz	sc-515701
AKT	pAb	1:2000 for WB	Cell Signalling Technology	4685
AKT-Ser473	mAb	1:1000 for WB	Cell Signalling Technology	4060
AMPKα	mAb	1:1000 for WB	Cell Signalling Technology	5831
AMPKα-Thr172	mAb	1:1000 for WB	Cell Signalling Technology	2535
β-MHC	pAb	1:1000 for WB	Proteintech	22280-1-AP
Collagen-1	pAb	1:1000 for WB	Proteintech	14695-1-AP
Collagen-3	pAb	1:1000 for WB	Proteintech	22734-1-AP
cTnT	mAb	1:150 for IF	Santa Cruz	sc-20025
ERK	mAb	1:1000 for WB	Cell Signalling Technology	4695
ERK1/2-Thr202/204	mAb	1:1000 for WB	Cell Signalling Technology	4370
GAPDH	pAb	1:1000 for WB	ZSGB-BIO	TA08
GATA4	mAb	1:50 for IF 1:1000 for WB	Abcam	ab307823
Lamin B1	pAb	1:3000 for WB	Bioworld	AP6001
LKB1	mAb	1:1000 for WB	Cell Signalling Technology	3047
LKB1-Ser428	mAb	1:1000 for WB	Cell Signalling Technology	3482
PSMB1	pAb	1:1000 for WB	Proteintech	11749-1-AP
PSMB2	mAb	1:3000 for WB	Abcam	ab249366
PSMB5	mAb	1:1000 for WB	Cell Signalling Technology	12919
RPT1	pAb	1:1000 for WB	Proteintech	14905-1-AP
RPT4	pAb	1:1000 for WB	Proteintech	15839-1-AP

pAb polyclonal antibody, mAb monoclonal antibody, WB Western blotting, IF immunofluorescence.

Animals

Newborn Sprague-Dawley rats (1-3 days old) and male C57BL/6 J mice (7-week-old) were purchased from the Hua Fukang Experimental Animal Centre (Beijing, China). All animal experiments in this study were approved by the Animal Experimentation Ethics Committee of Guangzhou Medical University (approval number: SYXK 2016-0168). The experimental mice were housed under a 12/12 h light/dark cycle. The temperature was kept at 23 °C ± 1 °C. The humidity was kept at 65% ± 5%. Food and water were provided ad libitum. Mice were randomly divided into five groups as follows: sham-operated group (Sham), trans-aortic constriction group (TAC), and one of three TAC+zonisamide (TAC + Z) groups (10 mg·kg⁻¹·d⁻¹, [T + Z 10]; 20 mg·kg⁻¹·d⁻¹, [T + Z 20]; and 40 mg·kg⁻¹·d⁻¹, [T + Z 40]). Two days after TAC surgery, mice in the TAC + Z groups were given the appropriate dose of zonisamide dissolved in ddH₂O via oral gavage daily for four weeks. Mice in the Sham group and the TAC group were administered the same volume of vehicle (ddH₂O) by gavage.

Trans-aortic constriction

Trans-aortic constriction (TAC) surgery was performed to induce pressure overload in C57BL/6 J mice according to a previously reported method [27]. Briefly, mice were anaesthetised with isoflurane and intubated. Then, a trans-sternal thoracotomy was performed. A 27-gauge needle served as a uniform model for constriction. The mice in the Sham group were subjected to the same surgery without aortic constriction.

Isolation of primary neonatal rat cardiomyocytes and cardiac fibroblasts

The primary neonatal rat cardiomyocytes (NRCMs) and neonatal rat cardiac fibroblasts (NRCFs) were isolated from the hearts of newborn Sprague-Dawley rats according to a previously reported method [28]. NRCMs and NRCFs were randomly divided into three groups: control (Con), Ang II 10 μM (Ang II), Ang II 10 μM + zonisamide 0.3 μM (Ang II + Z 0.3). Ang II and zonisamide were dissolved in ddH₂O. The cells were treated with Ang II and zonisamide for 24 h, and then were harvested for analysis.

Transthoracic echocardiography for cardiac evaluation

Noninvasive echocardiography for evaluation of cardiac performance in mice was performed using Visual Sonics (Vevo 3100; VisualSonics Inc., ON, Canada). Briefly, mice were anaesthetised with 1% isoflurane. M-mode echocardiography was performed to evaluate left ventricular (LV) ejection fraction (EF), fraction shortening (FS), and wall thickness including LV anterior and posterior walls at the diastolic and systolic stages (LVAW,d; LVAW,s; LVPW,d; and LVPW,s). Pulsed-wave Doppler was performed to obtain the isovolumetric relaxation time (IVRT) and ratio of early-to-late mitral inflow velocity (E/A) for the mitral valve (MV E/A ratio). The tissue Doppler imaging was performed to evaluate the peak early (E’) to late (A’) diastolic tissue velocity ratio (MV E’/A’). The heart rate was monitored and maintained within a consistent range (400-500 BPM).

Morphological analysis of cardiac tissue

The hearts of the mice were embedded in paraffin and cut into 3 μm sections. The areas of the cardiomyocytes were observed using hematoxylin and eosin (H&E) and fluorescein-conjugated wheat germ agglutinin (WGA) staining. To evaluate cardiac fibrosis, Masson’s trichrome and Sirius Red staining were performed. The sections were examined using light microscopy (Leica, Germany). The cardiomyocyte surface area and collagen deposition were determined using ImageJ software (NIH, Bethesda, USA).

Surface area measurement of primary neonatal rat cardiomyocytes

The surface area of primary neonatal rat cardiomyocytes (NRCMs) was measured using cardiac troponin T (cTnT) immunostaining [29]. Briefly, NRCMs were fixed, incubated with 1% Triton X-100, and blocked with 10% normal goat serum. Then, the cells were incubated with cTnT primary antibody overnight. After being washed with phosphate buffer saline (PBS), NRCMs were incubated with Alexa 647-conjugated goat anti-mouse IgG. The nuclei of NRCMs were incubated with 4′,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich). The images of NRCMs were obtained using a laser scanning confocal microscope (LSM880, Zeiss, Germany). The surface area of NRCMs was analysed using ImageJ software (NIH, Bethesda, USA).

Assay of cardiac proteasome activities

The cardiac proteasome activity was detected as previously reported [30]. Briefly, NRCMs or mouse heart tissues were homogenised in lysis buffer (Cell Signalling Technology, Danvers, MA, USA). The lysates of NRCMs or mouse heart tissue were incubated with the fluorogenic substrates. The intensity of fluorescence (345-nm and 445-nm emission) of the reaction products was calculated by the Spectra Max® M3 Microplate Reader (Molecular Devices LLC, San Jose, CA, USA). The proteasome activities of NRCMs and cardiac tissue were quantified according to the absorbance of the reaction products.

Western blotting

The cardiac tissue and cell lysates were homogenised in RIPA buffer. The proteins were extracted, and then separated onto 8%-15% sodium dodecyl sulfate-polyacrylamide gels. The proteins were transferred onto the polyvinylidene difluoride membranes. The polyvinylidene difluoride membranes were incubated with the primary antibodies. After being washed, the polyvinylidene difluoride membranes were incubated with appropriate secondary antibodies. The enhanced chemiluminescence was used for detecting the protein-antibody complexes. The internal controls used were the primary antibodies of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and Lamin B1.

Molecular docking analysis

Molecular docking analysis was used to explore the potential binding model and affinity of zonisamide for proteasome subunit beta types (PSMB) of the 20 S proteasome (PSMB1, PSMB2, and PSMB5), and proteasome-regulated particles (RPT) of the 19 S proteasome (RPT1, RPT4). Molecular docking was performed using the Molecular Operating Environment software (MOE, 2010.10, Chemical Computing Group Inc., Montreal, QC, Canada). The protein structure of the proteasome (PDB: RCSB PDB-8CVT: Human 19S-20S proteasome, state SD2) was used, and the docking site was identified using MOE Site Finder. The results of the docking analysis were generated using the Triangle Matcher method. London δG was used to evaluate the docking conformation scoring. The optimal geometric conformation of the best result was selected from the ligand interactions feature.

Statistical analysis

GraphPad Prism 7 (GraphPad, San Diego, CA, USA) was performed to process the statistical analyses following the manufacturer’s guidelines. The data were expressed as means ± standard error of the mean (SEM). The significance among groups was determined using one-way ANOVA. P < 0.05 was considered statistically significant.

Results

Zonisamide reduced pressure overload-induced cardiac dysfunction in mice

As shown in Fig. 1a, zonisamide was orally administered for four consecutive weeks after TAC surgery. A representative sample of echocardiography performed for cardiac evaluation is shown in Fig. 1b. The echocardiographic analysis demonstrated deterioration of cardiac performance in the TAC group with decreased LV EF and LV FS (Fig. 1c), along with LV wall thickening (Fig. 1d, e) and increased LV mass (Fig. 1f) and IVRT (Fig. 1g). The changes were accompanied by decreased MV E/A (Fig. 1h) and MV E’/A’ (Fig. 1i). The depressed LV EF and LV FS are powerful predictors of deteriorated LV systolic function [31, 32]. The prolonged IVRT is an early indicator of LV diastolic dysfunction [33]. The lower MV E/A and MV E’/A’, two of the studied parameters of LV diastolic filling [34, 35], indicated LV diastolic dysfunction. These aberrant echocardiographic parameters were restored after zonisamide intervention. During the assessment, heart rate remained stable (Fig. 1j). These findings suggested that treatment with zonisamide markedly improved pressure overload-induced cardiac dysfunction in mice.

Fig. 1. Zonisamide improved cardiac function in mice with pressure overload.

a Experimental design. The male C57BL/6 J mice (7-week-old) were subjected to either sham or trans-aortic constriction (TAC) surgery. Two days after surgery, mice in the zonisamide groups (T + Z 10, T + Z 20, and T + Z 40) were treated with zonisamide (10, 20, or 40 mg·kg⁻¹·d⁻¹) by intragastric (i.g.) administration for four consecutive weeks. The Sham and TAC groups received vehicle for the same duration. b The representative M-mode echocardiograms were captured (left panel). The pulsed-wave Doppler imaging was performed to evaluate the early peak transmitral flow velocity (E) to late peak transmitral flow velocity (A) ratio (MV E/A) (middle panel). The tissue Doppler imaging was performed to evaluate the peak early (E’) to late (A’) diastolic tissue velocity ratio (MV E’/A’) (right panel). c Evaluation of left ventricular (LV) ejection fraction (EF) and fraction shortening (FS). d Evaluation of LV end-systolic and end-diastolic anterior wall thickness (LVAW,s; LVAW,d). e Evaluation of LV end-systolic and end-diastolic posterior wall thickness (LVPW,s; LVPW,d). f Evaluation of LV mass. g Evaluation of LV isovolumetric relaxation time (IVRT). h The peak E to peak A ratio for mitral valve (MV E/A). i The peak E’ to peak A’ ratio for mitral valve (MV E’/A’). j The heart rate (HR). Data were analysed using one-way ANOVA and presented as mean ± SEM (n = 10). **P < 0.01 vs. Sham; ^#P < 0.05, ^##P < 0.01 vs. TAC.

Zonisamide ameliorated pressure overload-induced myocardial hypertrophy and myocardial fibrosis in mice

Four weeks after TAC surgery, pressure overload stimulation triggered pronounced myocardial hypertrophy in mice, as evidenced by significantly enlarged cardiac size (Fig. 2a), increased heart weight (HW) to tibia length (TL) ratio (HW/TL) (Fig. 2b), increased heart weight (HW) to body weight (BW) ratio (HW/BW) (Fig. 2c) and increased cardiomyocyte surface (Fig. 2d-f) compared to those in sham operation mice. Zonisamide treatment decreased the HW/TL and HW/BW ratios and reversed myocardial hypertrophy. Masson’s trichrome and Sirius Red staining revealed collagen deposition in the hearts in the TAC group, indicating myocardial fibrosis (Fig. 2d, g, h). Treatment with zonisamide dramatically decreased interstitial collagen contents compared to that in the TAC group.

Fig. 2. Zonisamide alleviated myocardial hypertrophy and myocardial fibrosis in the hearts of mice with pressure overload induced by TAC.

a Images of mice hearts from the Con, TAC, and zonisamide (T + Z 10, T + Z 20, and T + Z 40) groups. b Quantitative measurement of the heart weight (HW) to tibia length (TL) ratio (HW/TL) (n = 5). c Quantitative measurement of the heart weight (HW) to the body weight (BW) ratio (HW/BW) (n = 5). d–h Cardiomyocytes surface area was assessed using haematoxylin and eosin (H&E) and wheat germ agglutinin (WGA) staining; collagen volume was assessed using Masson’s trichrome and Sirius Red staining (scale bar: 500 μm and 20 μm for H&E staining; 20 μm for WGA staining, Masson’s trichrome staining and Sirius Red staining, n = 4). Data were analysed using one-way ANOVA and presented as mean ± SEM. *P < 0.05, **P < 0.01 vs. Sham; ^#P < 0.05, ^##P < 0.01 vs. TAC.

Zonisamide ameliorated Ang II-induced myocardial hypertrophy and myocardial fibrosis in vitro

To evaluate the effects of zonisamide in vitro, primary NRCMs and NRCFs were treated with Ang II to induce myocardial hypertrophy and fibrosis, respectively (Fig. 3a). The concentration of zonisamide (0.3 μm) for the in vitro assays have been demonstrated to be effective in Ang II-triggered primary NRCMs and NRCFs in our previous study [25]. For determining zonisamide’s effect on myocardial hypertrophy, the alteration in the surface area of NRCMs were explored using immunofluorescence analysis of cardiac troponin T (cTnT). The NRCMs triggered by Ang II had an obvious increase in the surface area, which was attenuated by zonisamide (Fig. 3b, c). Western blotting showed that the NRCMs incubated with Ang II had a lower alpha myosin heavy chain (α-MHC) level, but higher beta myosin heavy chain (β-MHC) and atrial natriuretic peptide (ANP) levels than those in control cells, which are pathological markers of myocardial hypertrophy. It was also shown that Ang II induced the upregulation of myocardial fibrotic markers collagen type I (Collagen-1) and collagen type III (Collagen-3) in the NRCFs. Zonisamide treatment significantly reversed myocardial hypertrophy and fibrosis in vitro (Fig. 3d-f). Taken together, these findings indicated that the efficiency of zonisamide in mitigating myocardial hypertrophy and myocardial fibrosis in vitro.

Fig. 3. Zonisamide alleviated angiotensin II (Ang II)-triggered myocardial hypertrophy and myocardial fibrosis in vitro.

a The primary neonatal rat cardiomyocytes (NRCMs) and neonatal rat cardiac fibroblasts (NRCFs) were isolated from the hearts of newborn Sprague-Dawley rats. Zonisamide and Ang II were incubated with NRCMs or NRCMs for 24 h. b Representative micrographs of cardiac troponin T (cTnT) (red) and 4′,6-diamidino-2-phenylindole (DAPI) (blue) immunofluorescence in the NRCMs (scale bar: 10 μm). c Quantitative measurement of surface area of the NRCMs (n = 4). d–f Western blot results and relative protein levels of alpha myosin heavy chain (α-MHC), beta myosin heavy chain (β-MHC), and atrial natriuretic peptide (ANP) in NRCMs, collagen type I (Collagen-1) and collagen type III (Collagen-3) in NRCFs (n = 5). Data were analysed using one-way ANOVA. Values are presented as mean ± SEM. *P < 0.05 vs. Con; ^#P < 0.05, ^##P < 0.01 vs. Ang II.

Zonisamide suppressed proteasome activities and subunits in TAC mice and hypertrophic NRCMs

The chymotrypsin-, trypsin- and caspase-like proteasome activities increased in mice cardiac tissue in the TAC group, which was consistent with that in the hypertrophic NRCMs in the Ang II-treated group. Zonisamide administration significantly inhibited proteasome activities in TAC mice (Fig. 4a), which is consistent with the findings observed in NRCMs (Fig. 4b). To investigate other mechanisms of zonisamide-induced inhibition of proteasome, we further investigated the effects of zonisamide on proteasome subunits using Western blotting. Figure 4c, d show that protein levels of proteasome subunit beta types (PSMB) of the 20 S proteasome (PSMB1, PSMB2 and PSMB5) and proteasome-regulated particles (RPT) of the 19 S proteasome (RPT1, RPT4) were obviously upregulated in mice hearts subjected to pressure overload compared to those in the control, but were reversed to normal levels after treatment with zonisamide. These changes were consistent with those observed in NRCMs (Fig. 4e, f). The nuclear PSMB5 level, which responds to the chymotrypsin-like activity of the proteasome, was abnormally elevated in the Ang II group, which was prevented by zonisamide administration (Fig. 4g). These results indicated that zonisamide significantly decreased proteasome activities, which was correlated with the suppression of 20 S and 19 S proteasome subunits.

Fig. 4. Zonisamide inhibited the proteasome activities and subunits in TAC mice and hypertrophic NRCMs.

Proteasome activities in cardiac tissue (a) (n = 6) and NRCMs (b) (n = 4) were analysed using fluorescent peptide substrates. c, d Representative Western blotting images and relative protein levels of the 20 S proteasome subunits (PSMB1, PSMB2, and PSMB5) and 19 S proteasome subunits (RPT1, RPT4) in cardiac tissue (n = 6). e, f Representative Western blotting images and relative protein levels of the 20 S proteasome subunits (PSMB1, PSMB2, and PSMB5) and 19 S proteasome subunits (RPT1, RPT4) in the NRCMs (n = 5). g Representative Western blotting image and relative protein level of PSMB5 in the nuclei of NRCMs (n = 4). Data were analysed using one-way ANOVA. Values are presented as the mean ± SEM. For (a and d), **P < 0.01 vs. Sham; ^#P < 0.05, ^##P < 0.01 vs. TAC. For (b, f, and g), *P < 0.05, **P < 0.01 vs. Con; ^#P < 0.05, ^##P < 0.01 vs. Ang II.

Zonisamide alleviated myocardial hypertrophy via suppression of activated proteasome

To further verify whether zonisamide alleviates myocardial hypertrophy by directly restraining proteasome function. NRCMs were incubated with 18α-GA (a widely used proteasome agonist [36]). The proteasome activities and expression of proteasome subunits were then measured. We found that 18α-GA significantly reversed the suppressive effects of zonisamide on Ang II-activated proteasome activities (Fig. 5a) and subunits (Fig. 5b, c). In addition, 18α-GA abolished the suppressive effects of zonisamide on Ang II-triggered myocardial hypertrophy, as evidenced by decreased α-MHC, increased β-MHC and ANP expression (Fig. 5b, c), and enlarged surface area of NRCMs (Fig. 5d, e). The above findings demonstrated that zonisamide’s inhibitory effect on myocardial hypertrophy was related to weakened proteasome activities and proteasome subunits.

Fig. 5. Zonisamide alleviated myocardial hypertrophy by suppressing activated proteasome.

NRCMs were treated with Ang II in the presence or absence of zonisamide for 24 h. Before being harvested for analysis, NRCMs were incubated with 18α-glycyrrhetinic acid (18α-GA) (2 mg/ml) for 3 h. a Proteasome activities of the NRCMs were analysed using fluorescent peptide substrates (n = 5). b, c Representative Western blot images and analysis of the 20 S proteasome subunits (PSMB1, PSMB2, and PSMB5), 19 S proteasome subunits (RPT1, RPT4), α-MHC, β-MHC, and ANP (n = 5). d Representative macrographic images of immunofluorescence of cTnT (red) and DAPI (blue) (scale bar: 10 μm). e Quantitative evaluation of surface area of NRCMs using ImageJ software (n = 5). Data were analysed using one-way ANOVA. Values are presented as mean ± SEM. *P < 0.05, **P < 0.01 vs. Con; ^#P < 0.05, ^##P < 0.01 vs. Ang II; ^$$P < 0.01 vs. Ang II + Z 0.3.

The effects of zonisamide on the activity of GSK-3 and myocardial hypertrophic factors

In our previous study, it was revealed that proteasome inhibition resulted in elevated GSK-3 activity (decreased GSK-3 phosphorylation) through inactivation of the Akt pathway and LKB1/AMPKα pathway which led to the downregulation of myocardial hypertrophic factors, including phosphorylation of ERK and nuclear aggregation of GATA4 [16]. To determine whether GSK-3 and its upstream/downstream were involved in zonisamide-induced proteasome inhibition, p-GSK-3α and p-GSK-3β (inactivation of GSK-3), p-AKT (activation of the Akt), p-LKB1 (activation of LKB1), p-AMPKα (activation of AMPKα), p-ERK (activation of ERK), and nuclear GATA4 in vivo were detected using Western blotting. The protein levels of p-GSK-3α, p-GSK-3β, p-AKT, p-LKB1, p-AMPKα, p-ERK, and nuclear GATA4 were increased in the TAC group as compared to the Sham group but downregulated in the zonisamide-treated groups (Fig. 6a-d). The changes in the NRCMs were consistent with the in vivo results (Fig. 6e-h). For evaluating the nuclear translocation of GATA4, immunofluorescence was performed. The nuclear aggregation of GATA4 was observably elevated in the Ang II group but decreased with zonisamide incubation (Fig. 6i, j). These results suggested that zonisamide activated GSK-3 but inactivated AKT and LKB1/AMPKα, the upstream regulatory factors of GSK-3. Zonisamide also decreased the protein levels of p-ERK and nuclear aggregation of GATA4, two hypertrophic factors which are downstream of GSK-3.

Fig. 6. The effects of zonisamide on glycogen synthesis kinase 3 (GSK-3), AKT (protein kinase B, PKB), liver kinase B1(LKB1), AMP-activated protein kinase (AMPKα), extracellular signal-regulated kinase (ERK), and GATA binding protein 4 (GATA4).

a–d Representative Western blot images and analysis of phosphorylated (p-) GSK-3α, p-GSK-3β, p-AKT, p-LKB1, p-AMPKα, p-ERK, and nuclear GATA4 in vivo (n = 4). e–h Representative Western blot images and analysis of phosphorylated (p-) GSK-3α, p-GSK-3β, p-AKT, p-LKB1, p-AMPKα, p-ERK, and nuclear GATA4 in vitro (n = 4). i Representative micrographs of immunofluorescence of GATA4 (green), cTnT (red) and DAPI (blue) in NRCMs (scale bar: 5 μm). j Quantitative measurement of GATA4 distribution in nucleus using ImageJ software (n = 5). Data were analysed using one-way ANOVA. Values are presented as mean ± SEM. For (b and d), *P < 0.05, **P < 0.01 vs. Sham; ^#P < 0.05, ^##P < 0.01 vs. TAC. For (f, h, and j), *P < 0.05, **P < 0.01 vs. Con; ^##P < 0.01 vs. Ang II.

The molecular docking analysis of the interaction between zonisamide and proteasome

In this study, zonisamide inhibited protein levels and activities of proteasome subunit beta types (PSMB) of the 20 S proteasome (PSMB1, PSMB2, and PSMB5), and proteasome-regulated particles (RPT) of the 19 S proteasome (RPT1, RPT4). To the best of our knowledge, the interaction between zonisamide and proteasome has not been previously reported. Therefore, we used molecular docking analysis to determine the interaction between zonisamide and proteasome. The chemical structure of zonisamide is shown in Fig. 7a. For PSMB1, the amino acids of the S-chain (HIS36 and SER34) formed three hydrogen bonds with zonisamide, the HIS36 also formed π - π stacking interaction with zonisamide. In addition, GLU31, ILE35, TYR132, TYR124, ALA138, ALA143, and TYR144 participated in the formation of van der Waals forces (Fig. 7b). For PSMB2, the CYS63 and ARG88 on the J chain formed six hydrogen bonds with zonisamide. In addition, the pockets composed of ILE84, THR77, ARG81, THR58, ARG60 on the J chain and TYR156 and ASN155 on the I chain formed strong van der Waals bonds with zonisamide (Fig. 7c). For PSMB5, The R-THR2 and R-THR22 on the R chain formed three hydrogen bonds with zonisamide, while the R-VAL32, R-ALA50, R-ALA47, and R-GLY48 on the R chain formed van der Waals interactions with zonisamide (Fig. 7d). For RPT1, the three amino acids GLY219, LYS222, and THR223 on the A-chain formed five hydrogen bonds with zonisamide. In addition, THR220, GLY221, LEU224, and ARG227 formed van der Waals interactions with zonisamide (Fig. 7e). For RPT4, PRO176 on the E-chain formed a hydrogen bond with zonisamide. In addition, PRO175, THR178, LEU303, and ASN339 formed strong van der Waals bonds with zonisamide (Fig. 7f). Therefore, we speculated that zonisamide may represent a novel inhibitor of proteasome (Fig. 8).

Fig. 7. Crystal structure and bindings of zonisamide with proteasome subunit beta types (PSMB) of the 20 S proteasome (PSMB1, PSMB2, and PSMB5), and proteasome-regulated particles (RPT) of the 19 S proteasome (RPT1, RPT4).

a The chemical structure of zonisamide. b–f The overview of PSMB (PSMB1, PSMB2, and PSMB5) and RPT (RPT1, RPT4) bound to zonisamide (zonisamide shown as blue sticks). The three-dimensional ligand-interaction map of PSMB (PSMB1, PSMB2, and PSMB5) and RPT (RPT1, RPT4) bound to zonisamide. Individual residues are labelled according to their numbering in the PSMB (PSMB1, PSMB2, and PSMB5) and RPT (RPT1, RPT4), hydrogen bonds interactions are depicted in yellow dashed lines. Two-dimensional ligand-interaction map of PSMB (PSMB1, PSMB2, and PSMB5) and RPT (RPT1, RPT4) bound to zonisamide (bottom).

Fig. 8. Schematic diagram of zonisamide on pressure overload-induced myocardial hypertrophy.

Zonisamide inhibited the proteasome (including proteasome activities and subunits). The inhibited proteasome triggered GSK-3α/β activation, and the AKT and/or the LKB1/AMPKα pathways may be involved. The activation of GSK-3 (GSK-3α, GSK-3β) prevented myocardial hypertrophy via the inhibition of myocardial hypertrophic factors, including ERK and GATA4. Alternatively, the LKB1/AMPKα pathway may regulate myocardial hypertrophy directly, thereby improving cardiac performance.

Discussion

Myocardial hypertrophy caused by pressure overload is considered an adaptive response that eliminates stress on the ventricular wall during the initial phase of cardiac remodelling [37]. However, prolonged hypertrophy is likely to exert adverse effects, as evidenced by massive amounts of patient data demonstrating strong associations between hypertrophic growth and the incidence of heart failure [38]. Nevertheless, dihydrochlorothiazide combined with triamterene or Ang II receptor antagonists, which were used to protect cardiac function in patients with pressure overload-induced myocardial hypertrophy, slightly increased median survival time [39]. Although several treatments have been proposed, no specific drugs have been developed. Consequently, the development of new drugs is urgently needed. An emerging body of research shows that elevated proteasome activity is associated with several diseases. To date, a consensus has been reached that increased proteasome activity and overexpression of proteasome subunits accelerate the deterioration of myocardial hypertrophy [40], and there is growing evidence that partial inhibition of proteasome activity exerts an anti-hypertrophic effect [15]. Some of the proteasome inhibitors have been developed for multiple diseases, including neurodegenerative diseases and certain types of cancer. The proteasome inhibitors such as bortezomib, carfilzomib, and ixazomib, for instance, have been approved for malignant tumours treatment by the U.S. Food and Drug Administration (FDA) [41–43]. Zonisamide was originally developed in Japan and approved by the FDA in 2000 as adjuvant therapy for partial seizures [44–46]. Our previous studies were the first to demonstrate that zonisamide exerts therapeutic efficacy in the prevention of myocardial hypertrophy in a rat model of pressure overload, and prevents diabetic cardiomyopathy in a mouse model of type 2 diabetes mellitus [25, 26]. In the present study, we showed that zonisamide ameliorates pressure overload-induced myocardial hypertrophy in mice. Haemodynamically, zonisamide increased EF, FS, E/A ratios, and E’/A’ ratios but decreased LV ventricular thickness, LV mass, and IVRT, suggesting that zonisamide relieves myocardial hypertrophy and ameliorates the decline in cardiac function. Morphologically, zonisamide displayed a suppressive effect on hypertrophic histology in mouse heart tissue, observed using H&E and WGA staining, with decreased cardiomyocyte surface area. This anti-hypertrophic effect was characterised by decreased HW/TL, HW/BW, and heart size. Hypertrophic cardiomyopathy is accompanied by collagen deposition which is a hallmark of the condition [47]. In this study, we found that mice with pressure overload developed myocardial fibrosis with abundant collagen deposition, which was mitigated by zonisamide treatment. In this study, an in vitro myocardial hypertrophy model was built by treating NRCMs with Ang II to reveal the underlying molecular mechanism of zonisamide. A previous study demonstrated that pathological myocardial hypertrophy in adult hearts is marked by a decrease in α-MHC level but an increase in β-MHC [48] and ANF [49] levels. Our data showed that zonisamide attenuated Ang II-induced myocardial hypertrophy in vitro, as indicated by the diminished surface area, reduced expression of ANP and β-MHC, and increased expression of α-MHC in the NRCMs. Furthermore, in vitro myocardial fibrosis model was established by treating NRCFs with Ang II. The protein levels of Collagen-1 and Collagen-3, two markers for myocardial fibrosis [50], were remarkably increased in the NRCFs treated with Ang II, and this increase was attenuated by zonisamide administration, consistent with the results of the in vivo histological evaluation of myocardial fibrosis.

Overwhelming evidence supports that suppression of the proteasome may be beneficial to the hypertrophic heart [15]. We previously showed that the suppressed proteasome led to GSK-3 activation, resulting in amelioration of myocardial hypertrophy [16]. However, whether the cardioprotective effects of zonisamide are related to proteasome function has not been investigated yet. In the present study, we provided novel evidence that (i) zonisamide inhibits the proteasome activities, including chymotrypsin-, trypsin-, and caspase-like proteolytic activities, and (ii) zonisamide inhibits the proteasome subunits and nuclear protein level of PSMB5, indicating that zonisamide suppresses proteasome through different mechanisms. We treated the NRCMs with 18α-GA, a specific proteasome activator, to confirm whether the anti-hypertrophic effect of zonisamide is dependent on proteasome inhibition. The data showed that 18α-GA treatment increased proteasome activities and subunits. Meanwhile, the anti-hypertrophic effect of zonisamide was attenuated by cotreatment with 18α-GA, suggesting that the anti-hypertrophic effect of zonisamide is proteasome-dependent. Our findings are consistent with those obtained using the proteasome inhibitor epoxomicin in the prevention of myocardial hypertrophy [12].

GSK-3 is reported to play a critical role in cell growth, metabolism, gene expression, and cytoskeletal integrity [51]. GSK-3 manifests commonly as 2 congeners, GSK-3α and GSK-3β. Increased phosphorylation at Ser21 of GSK-3α and Ser9 of GSK-3β inhibits GSK-3 activity [17]. The activation of GSK-3 also slows the advancement of myocardial hypertrophy [52, 53]. One of our previous studies of the brain indicated that lactacystin, a specific proteasome inhibitor, could reduce the phosphorylation of GSK-3β [30]. Another of our previous studies of the heart showed that quercetin relieved myocardial hypertrophy via proteasome inhibition and subsequent GSK-3α and GSK-3β activation. The proteasome inhibition occurred upstream of GSK-3 [16]. In the current study, we also investigated the phosphorylation at Ser21 of GSK-3α and Ser9 of GSK-3β. We found that the activity of GSK-3 was suppressed, which was consistent with the previous findings for GSK-3α [52] and GSK-3β [21, 53]. Zonisamide treatment markedly increased the activity of GSK-3 as evidenced by lower phosphorylation of GSK-3α and GSK-3β. AKT upstream of GSK-3β is triggered by myocardial hypertrophy stimuli [54]. The AMP/ATP ratio is increased in a state of pressure overload. AMPKα (a serine/threonine protein kinase) functions as an energy transducer and is activated and involved in the vital molecular biological responses in eukaryotic cells [55]. LKB1 is a ubiquitin ligase whose expression corresponds to the proteasome activity [56]. In this study, the activities of AKT, AMPKα and LKB1 were distinctly increased in Ang II-treated NRCMs and TAC mice but decreased in the zonisamide-treated groups as evidenced by the low-expression of p-AKT, p-AMPKα, and p-LKB1. In addition, we detected myocardial hypertrophic factors, including p-ERK and GATA4, which are downstream targets of GSK-3. Compared to the control group, p-ERK increased, and GATA4 aggregated to the nucleus in the hypertrophic NRCMs in the Ang II treated group, which is consistent with the previous findings [57, 58]. In the zonisamide-treated group, the changes were reversed to the normal levels, which have not been reported yet. Therefore, this study confirmed inhibitory effect of zonisamide on myocardial hypertrophic factors for the first time. In previous studies, it was found that inhibition of proteasome could restrain myocardial hypertrophy through modulation of the ERK activity [58]. MG132, a proteasome inhibitor, eliminated the elevation of GATA4 level in embryonic stem cells [59]. These studies indicated that myocardial hypertrophic factors, such as ERK and GATA4, could be regulated by the proteasome. Therefore, based on our findings, it can be concluded that zonisamide may regulate the myocardial hypertrophic factors through proteasome inhibition.

Several proteasome inhibitors (bortezomib, carfilzomib, and ixazomib) have been proven to be clinically useful in curing certain types of malignancies [41–43]. Considering that the process of developing new drugs to treat any disease is often labourious, expensive, and prone to failure [60], repurposing old drugs for new indications is a promising tool to speed up the drug discovery process [61]. This research demonstrated that zonisamide, an anti-epileptic drug [62], shows tremendous promise in myocardial hypertrophy. The mechanism underlying the cardioprotective effect of zonisamide still needs further investigation. It has been reported that phenotypic measurements may significantly affected by the diversity of sex [63, 64]. We did not investigate the therapeutic effect of zonisamide in female mice. Studies including female mice are needed.

Conclusions

In summary, our study showed that zonisamide is effective in reducing features associated with myocardial hypertrophy. The anti-hypertrophic effect of zonisamide, reported here for the first time, is a result of proteasome inhibition, which is correlated with GSK-3 activation. Therefore, our results suggest that zonisamide functions as a novel proteasome inhibitor that supports repurposing for the treatment of myocardial hypertrophy.

Author contributions

QW designed the study; QW carried out the study and wrote the paper; QW performed the experiments; WJL, XYM, JSC, and XYZ participated in the animal experiment; XYY and YHL conceived, and supervised the study and helped in discussing the data.

Competing interests

The authors declare no competing interests.