Regulation of mitochondrial cristae remodelling by acetylcholine alleviates palmitate-induced cardiomyocyte hypertrophy

Free Radical Biology and Medicine 145 (2019) 103–117

Contents lists available at ScienceDirect

Free Radical Biology and Medicine journal homepage: www.elsevier.com/locate/freeradbiomed

Original article

Regulation of mitochondrial cristae remodelling by acetylcholine alleviates palmitate-induced cardiomyocyte hypertrophy

T

Run-Qing Xuea, Ming Zhaoa, Qing Wua, Si Yanga, Yan-Ling Cuia, Xiao-Jiang Yua, Jiankang Liub, Wei-Jin Zanga,∗ a

Department of Pharmacology, School of Basic Medical Sciences, Xi'an Jiaotong University Health Science Center, Xi'an, PR China Frontier Institute of Science and Technol, and Key Laboratory of Biomedical Information Engineering of the Ministry of Education, School of Life Science and Technology, Xi'an Jiaotong University, Xi'an, PR China

b

ARTICLE INFO

ABSTRACT

Keywords: Palmitate Acetylcholine Cristae remodelling Mitofilin AMPK

Mitochondrial dysfunction is associated with obesity-induced cardiac remodelling. Recent research suggests that the cristae are the true bioenergetic components of cells. Acetylcholine (ACh), the major neurotransmitter of the vagus nerve, exerts cardio-protective effects against ischaemia. This study investigated the role of cristae remodelling in palmitate (PA)-induced neonatal rat cardiomyocyte hypertrophy and explored the beneficial effects of ACh. We found loose, fragmented and even lysed cristae in PA-treated neonatal cardiomyocytes along with declines in mitochondrial network and complex expression and overproduction of mitochondrial reactive oxygen species (ROS); these changes ultimately resulted in increased myocardial size. Overexpression of mitofilin by adenoviral infection partly improved cristae shape, mitochondrial network, and ATP content and attenuated cell hypertrophy. Interestingly, siRNA-mediated AMP-activated protein kinase (AMPK) silencing increased the number of cristae with a balloon-like morphology without disturbing mitofilin expression. Furthermore, AMPK knockdown abolished the effects of mitofilin overexpression on cristae remodelling and inhibited the interaction of mitofilin with sorting and assembly machinery 50 (Sam50) and coiled-coil helix coiled-coil helix domaincontaining protein 3 (CHCHD3), two core components of the mitochondrial contact site and cristae organizing system (MICOS) complex. Intriguingly, ACh upregulated mitofilin expression and AMPK phosphorylation via the muscarinic ACh receptor (MAChR). Moreover, ACh enhanced protein-protein interactions between mitofilin and other components of the MICOS complex, thereby preventing PA-induced mitochondrial dysfunction and cardiomyocyte hypertrophy; however, these effects were abolished by AMPK silencing. Taken together, our data suggest that ACh improves cristae remodelling to defend against PA-induced myocardial hypertrophy, presumably by increasing mitofilin expression and activating AMPK to form the MICOS complex through MAChR. These results suggest new and promising therapeutic approaches targeting mitochondria to prevent lipotoxic cardiomyopathy.

1. Introduction High levels of saturated free fatty acids, such as palmitate (PA), in the heart have been proposed to play roles in cardiac remodelling and heart failure in obesity and diabetes [1,2]. However, the nature of PAinduced cardiomyocyte hypertrophy has not been well characterized. Recent research has demonstrated mitochondrial dysfunction in the cardiovascular systems of patients and animals with metabolic syndrome [3,4]. Mitochondria are important hubs for cellular energy supply and signal communication; however, damaged mitochondria exhibit decreased levels of mitochondrial metabolic enzymes and ATP

and decreased membrane potential, which results in a burst of reactive oxygen species (ROS) production and cytochrome C release and leads to cell hypertrophy or apoptosis [5]. Therefore, the prevention and treatment of mitochondrial dysfunction are promising and reasonable strategies to maintain myocardial health in PA-induced cardiomyocyte hypertrophy. Consistent with the rule that “form follows function”, different forms of mitochondria are found in different cell types under different conditions [6]. Previous studies have often focused on mitochondrial division and fusion, but recent research suggests that mitochondrial cristae are dynamic biochemical compartments whose changes in shape

∗ Corresponding author. Department of Pharmacology, School of Basic Medical Sciences, Xi'an Jiaotong University Health Science Center, PO Box #77, No. 76 Yanta West Road, Xi'an City, 710061, Shaanxi Province, PR China. E-mail address: [email protected] (W.-J. Zang).

https://doi.org/10.1016/j.freeradbiomed.2019.09.025 Received 11 September 2019; Accepted 21 September 2019 Available online 22 September 2019 0891-5849/ © 2019 Elsevier Inc. All rights reserved.

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

modulate mitochondrial complex assembly and signal protein release in a process called cristae remodelling [7,8]. Mitochondrial cristae are mainly orchestrated by the mitochondrial contact site and cristae organizing system (MICOS) complex. Disturbance of MICOS complex formation is known to destroy cristae remodelling and disrupts mitochondrial and cellular function [9]. Nevertheless, the roles of mitochondrial cristae remodelling and its associated proteins in PA-induced cardiomyocyte hypertrophy remain poorly defined. Numerous studies have demonstrated the presence of autonomic imbalance in cardiovascular diseases induced by metabolic syndrome [10,11]. Enhancing vagal activity improves cardiac function by inhibiting mitochondrial dysfunction in obese rats [12]. Our previous experiments have also demonstrated the protective effect of pyridostigmine, an anticholinesterase, on mitochondrial cristae in obese mice [13]. In addition, acetylcholine (ACh), the major neurotransmitter of the vagus nerve, confers protective effects against ischaemia/reperfusion injury by improving mitochondrial biogenesis and function and inhibiting the mitochondrial unfolded protein response via muscarinic ACh receptor (MAChR)-AMP-activated protein kinase (AMPK) signalling [14,15]. However, the effects of ACh on cardiac hypertrophy and mitochondrial cristae remodelling in metabolic disorder models, such as those induced by PA, have not been elucidated. The present study aimed to explore whether mitochondrial cristae injury is involved in myocardial hypertrophy after PA treatment. We further examined the protective effects of ACh in cardiomyocytes against PA-induced damage to mitochondrial form and function, with a focus on the AMPK-associated pathway.

incubated with PA-BSA complex for 24 h; and the PA+5-aminoimidazole-4-carboxamide ribonucleotide (AICAR) group, in which cells were treated with 0.5 mmol/L AICAR for 2 h and then co-incubated with PABSA complex for 24 h. 2.4. Adenoviral infection An adenoviral vector encoding mitofilin was provided by GenePharma (GenePharma, China). NRVCs were infected with the adenoviral vector essentially as described previously [16]. Forty-eight hours after plating, the cells were infected with adenovirus for 2 h. Mitofilin protein overexpression was assessed by Western blot, and cell activity was measured by Cell Counting Kit-8 assay (Beyotime, China). 2.5. SiRNA transfection SiRNA oligonucleotides specific for AMPK and negative control (NC) siRNA oligonucleotides were provided by GenePharma. Cells were transfected with the siRNAs using Lipofectamine 2000 (Invitrogen, USA) for 8 h. The transfection medium was then replaced with DMEM containing 10% FBS, and the cells were grown for another 36 h in this medium. The efficiency of AMPK protein knockdown was assessed by Western blot. 2.6. ROS detection We assessed mitochondrial ROS using MitoSOX Red (Invitrogen). Briefly, cells were incubated with 5 μmol/L MitoSOX Red for 30 min at 37 °C in the dark and washed for 10 min with PBS. After fixation in 4% paraformaldehyde in PBS for 20 min at room temperature and washing for 10 min with PBS, the cells were incubated with DAPI (1:1000 dilution in PBS; Beyotime) for 15 min at 37 °C in the dark to label the nuclei and then washed again for 10 min with PBS. Fluorescent images were acquired using a confocal microscope (Nikon, Japan).

2. Materials and methods 2.1. Primary cardiomyocyte isolation and culture All experiments were performed according to the Guidelines for the Care and Use of Laboratory Animals and with the permission of the Ethics Committee of Xi'an Jiaotong University. Primary neonatal rat ventricular cardiomyocytes (NRVCs) were isolated from 1- to 3-day old neonatal Sprague-Dawley rats. The hearts were removed, washed in PBS, cut into small pieces and then digested with 0.1% collagenase for 20 min. The mixture was centrifuged at 1000 rpm for 5 min, and the precipitate was resuspended and seeded in culture dishes with lowglucose Dulbecco's modified Eagle medium (DMEM) containing 10% (v/v) foetal bovine serum (FBS), 100 U/mL penicillin and 100 μg/mL streptomycin. The medium was renewed every 24 h. The cells were cultured at 37 °C in a humidified atmosphere (with 5% CO2). The cardiomyocytes were separated from fibroblasts based on their differential adhesion.

2.7. Mitochondrial membrane potential measurement and mitochondrial morphological analysis The mitochondrial membrane potential was determined using JC-1 staining (Beyotime). Briefly, cells were loaded with a JC-1 staining solution (1:1000 dilution in PBS) for 30 min at 37 °C in the dark and rinsed for 10 min with PBS. JC-1 fluorescence was detected with a confocal microscope. Red fluorescence indicated potential-dependent aggregation of JC-1 in the mitochondria, while green fluorescence indicated the presence of the monomeric form of JC-1 in the cytosol after depolarization of the mitochondrial membrane. Differences in the ratios of the red and green fluorescence intensities reflected changes in mitochondrial membrane potential. In addition, mitochondrial morphology was observed by red fluorescence microscopy (Nikon, Japan).

2.2. PA-bovine serum albumin (BSA) complex preparation Sodium PA (Sigma, USA) was dissolved in PBS at 70 °C and then conjugated with fatty acid-free BSA (Sigma) at a 6:1 M ratio.

2.8. Immunoprecipitation and Western blot analysis

2.3. Drug treatments

Immunoprecipitation and Western blot analysis were performed as previously described [13,17]. The following primary antibodies and dilutions were used: anti-atrial natriuretic peptide (ANP, 1:1000 dilution; Invitrogen), anti-brain natriuretic peptide (BNP, 1:1000 dilution; AbSci, USA), anti-β-myosin heavy chain (β-MHC, 1:1000 dilution; AbSci), anti-cytochrome C (1:500 dilution; Cell Signalling Technology, USA), anti-coiled-coil helix coiled-coil helix domain-containing protein 3 (CHCHD3, 1:1000 dilution; Invitrogen), anti-mitofilin (1:500 dilution; Invitrogen; 1:1000 dilution; Bioworld, USA), anti-sorting and assembly machinery 50 (Sam50, 1:500 dilution; Proteintech, China), anticomplex I (1:2000 dilution; Invitrogen), anti-complex II (1:2000 dilution; Invitrogen), anti-complex III (1:2000 dilution; Invitrogen), anticomplex IV (1:2000 dilution; Invitrogen), anti-complex V (1:2000 dilution; Invitrogen), anti-AMPKα (1:500 dilution, Cell Signalling

The NRVCs were serum-starved in FBS-free DMEM for 12 h and then randomly divided into the following experimental groups: the BSA group, in which cells were treated with BSA alone; the PA group, in which cells were treated with 200 μmol/L PA-BAS complex for 24 h; the ACh group, in which cells were treated with 10−6 mol/L ACh for 2 h and then co-incubated with BSA for 24 h; the PA + ACh group, in which cells were treated with 10−6 mol/L ACh for 2 h and then coincubated with PA-BSA complex for 24 h; the PA + ACh + atropine (Atro) group, in which cells were treated with 10−5 mol/L Atro and 10−6 mol/L ACh for 2 h and then co-incubated with PA-BSA complex for 24 h; the PA + ACh + compound C (CC) group, in which cells were treated with 10 μmol/L CC and 10−6 mol/L ACh for 2 h and then co104

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Technology), anti-phospho-AMPKα (Thr172; 1:500 dilution; Cell Signalling Technology) and anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 1:5000 dilution, Sinopept, China). The relative intensities of the bands were measured with ImageJ software.

3.2. Mitochondrial cristae remodelling was involved in PA-induced cardiomyocyte hypertrophy Mitochondria adapt their morphology to support their cellular functions [18]. To determine the morphological changes in the mitochondrial network during hypertrophic NRVC growth, cultured NRVCs were treated with PA (200 μmol/L, 0–48 h) and stained with JC1. After 24 h, PA-treated cardiomyocytes exhibited a less connected mitochondrial network than untreated cardiomyocytes (Fig. 2A). These morphological observations were further confirmed by the percentage of NRVCs displaying a fragmented mitochondrial morphology. As shown in Fig. 2B, PA significantly increased the percentage of NRVCs with fragmented mitochondria from 8.54 ± 0.78 (0 h) to 42.91 ± 2.55 and 68.34 ± 2.52 after 24 and 48 h, respectively. We further explored changes to the mitochondrial cristae during hypertrophic cardiomyocyte growth. The results are shown in Fig. 2C and E. Based on electron micrographs, five characteristic mitochondrial morphologies were identified, as shown in Fig. 2D [19,20], in which the severity of mitochondrial cristae damage progresses from left to right. Among the control cells (0 h), 65.99% of mitochondria contained thin and long lamellar-shaped cristae; however, after 12 h, the mitochondrial cristae mainly presented thin and long lamellae (36.21%) and open and long lamellae (42.90%). Because the changes in mitochondrial function and hypertrophic markers between 0 and 12 h were not significant (Fig. 1), it is reasonable to believe that the mitochondrial cristae in NRVCs treated with PA for 12 h were relatively healthy. However, the proportion of aberrant mitochondria with open and short lamellar and balloon-like or non-lamellar cristae dramatically increased from 16.87% at 0 h to 72.98% and 82.22% at 24 and 48 h, respectively. Notably, the proportion of mitochondria that displayed balloon-like and non-lamellar cristae increased from 34.78% at 24 h to 64.75% at 48 h, which suggested that mitochondrial cristae damage continued to worsen during PA treatment. These microscopic observations indicate that perturbations to mitochondrial shape, especially the shape of mitochondrial cristae, may occur in PA-induced cardiomyocyte hypertrophy. For all subsequent experiments, PA (200 μmol/L) treatment was carried out for 24 h.

2.9. Transmission electron microscopy (TEM) NRVCs were harvested, fixed with 2.5% glutaraldehyde and 4% paraformaldehyde in 0.1 mol/L phosphate buffer for 2 h at 4 °C, postfixed with 1% osmium tetroxide in 0.1 mol/L phosphate buffer, dehydrated, infiltrated with propylene oxide and embedded in epoxy resin. Ultrathin sections were doubled-stained with uranyl acetate and lead citrate. Images of the myocardial cell ultrastructure were captured with a transmission electron microscope (Hitachi, Japan). 2.10. Assessment of ATP content Intracellular ATP content was determined using an enhanced ATP assay kit (Beyotime) according to the manufacturer's instructions. The ATP content was assessed using a multimode microplate reader with a luminometer (FLUOstar Omega, Germany). 2.11. Immunofluorescence staining After treatment, NRVCs were washed with PBS and then fixed with 4% paraformaldehyde for 30 min at room temperature. After rinsing with PBS, the cells were permeabilized in 0.1% Triton X-100 for 30 min at room temperature. The cells were washed twice with PBS and incubated in 1% BSA for 30 min at room temperature. After washing with PBS, the NRVCs were incubated with antibodies against mitofilin (1:200 dilution; Bioworld) and α-actinin (1:100 dilution; Invitrogen) in blocking solution overnight at 4 °C. After washing with PBS, the cells were further incubated with secondary Alexa 488-conjugated antibodies (1:200 dilution in PBS; Thermo Fisher Scientific) for 30 min at 37 °C. The cells were washed three times with PBS and then incubated with DAPI (1:1000 dilution in PBS; Beyotime) for 15 min at 37 °C. The samples were then examined under a confocal fluorescence microscope (Nikon, Tokyo, Japan), and fluorescence images were analysed with ImageJ.

3.3. Mitofilin overexpression partly restored mitochondrial morphology, reversed mitochondrial dysfunction and decreased PA-induced cardiomyocyte hypertrophy

2.12. Statistical analysis

Recent studies have shown that mitofilin plays a key role in mitochondrial cristae remodelling. In our experiments, mitofilin expression gradually decreased during PA-induced hypertrophic NRVC growth (Supplementary Fig. 3); therefore, we hypothesized that mitofilin overexpression would ameliorate PA-induced cell hypertrophy. Mitofilin was overexpressed in PA-stimulated NRVCs with an adenoviral vector (Ad-mitofilin) (Supplementary Fig. 4). A greater proportion of mitochondria displayed thin/open and long lamellar-shaped cristae in the PA + Ad-mitofilin group than in the PA + NC adenovirus (AdNC) group (48.37% and 26.12%, respectively), and the proportion of mitochondria with balloon-like and non-lamellar cristae decreased from 37.98% in the PA + Ad-NC group to 20.77% in the PA + Ad-mitofilin group (Fig. 3A and B). In addition, infection of cardiomyocytes with Admitofilin increased mitochondrial network, as determined by the parameters described previously (Fig. 3C and D). Furthermore, infection with Ad-mitofilin prevented the PA-induced decrease in ATP content (Fig. 3E) and markedly inhibited the PA-induced increase in cell surface area (Fig. 3F and G). Taken together, these results suggest that improving mitochondrial cristae remodelling by overexpressing mitofilin inhibits PA-induced mitochondrial dysfunction and myocardial hypertrophy.

The data are shown as the means±SEMs. Statistical analysis was conducted using one-way ANOVA followed by Student's t-test or Tukey's multiple comparisons test. For all analyses, values of P < 0.05 indicated statistical significance. 3. Results 3.1. Mitochondrial dysfunction was observed during PA-induced cardiomyocyte hypertrophy To explore the processes underlying obesity-induced cardiac hypertrophy, we first determined changes in the levels of hypertrophic markers in PA-treated NRVCs. NRVCs were subjected to PA (200 μmol/ L) treatment for 6, 12, 24, 36, and 48 h. As shown in Fig. 1A–D, the expression of cardiomyocyte hypertrophy markers increased in response to PA, peaking at 24 h. A robust increase in mitochondrial ROS content was also observed at 24 h (Fig. 1E–G), and the protein levels of mitochondrial complexes I, II, III, and V were lower in NRVCs treated with PA for 24 h than in control NRVCs (Fig. 1H-M). These data suggest that mitochondrial damage was involved in the PA-induced hypertrophy of NRVCs. 105

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 1. PA-induced cardiomyocyte hypertrophy was accompanied by mitochondrial damage. (A–D) PA treatment for different lengths of time increased the expression of ANP, BNP and β-MHC in NRVCs. n = 6 independent experiments. (E–G) Mitochondrial ROS production in NRVCs was measured following PA treatment for different lengths of time. Scale bar, 50 μm n = 4 independent experiments; more than 100 cells were measured in each experiment. (H–M) Lysates from NRVCs incubated with PA for different lengths of time were immunoblotted using antibodies against mitochondrial complexes I–V, and GAPDH served as the internal control. n = 4 independent experiments. The data are represented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. *P < 0.05, **P < 0.01, and ***P < 0.001 vs the 0 h group.

106

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

(caption on next page)

107

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 2. Mitochondrial cristae remodelling was observed in PA-induced cardiomyocyte hypertrophy. (A) Representative confocal images showing the morphologies of mitochondria treated with PA for 0, 12, 24, and 48 h. Scale bar, 25 μm. (B) The percentage of cells displaying fragmented mitochondria in each group, as quantified in three or four randomly chosen fields of view (n > 200 cells) per experiment. n = 4 independent experiments. (C) NRVCs were treated with PA and analysed by TEM (upper panel). The magnified images in the yellow squares are shown in the lower panel. Scale bar, 100 nm. (D) Representative cristae morphology observed by TEM. Scale bar, 100 nm. (E) Quantification of the proportion of mitochondria exhibiting each cristae morphology shown in Fig. 2C. The numbers indicate the numbers of mitochondria that were counted. The data are presented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. ***P < 0.001 vs the 0 h group. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.4. ACh increased mitofilin expression and protected mitochondrial cristae via MAChR

mitofilin, which was abolished by Atro incubation. Specifically, according to the TEM results shown in Fig. 4D and E, the proportion of mitochondria with open and short lamellar-like cristae was 18.91% lower in the PA + ACh group than in the PA group; the proportion of mitochondria with balloon-like and non-lamellar cristae was also 22.81% lower in the PA + ACh group than in the PA group. However, pretreatment with Atro inhibited the effect of ACh on PA-treated

To elucidate whether ACh can protect mitochondrial cristae, we first examined the effect of ACh on mitofilin expression using an immunofluorescence assay. As shown in Fig. 4A–C, pretreatment of PA-treated NRVCs with ACh significantly upregulated the protein expression of

Fig. 3. Overexpression of mitofilin partly improved mitochondrial cristae remodelling and function and inhibited PA-induced cardiomyocyte hypertrophy. (A) Mitochondrial cristae morphology in PAtreated NRVCs following Ad-NC or Ad-mitofilin infection was observed by TEM. Scale bar, 100 nm. (B) Quantification of cristae with the morphologies shown in Fig. 3A. The numbers indicate the numbers of mitochondria that were counted. (C) Representative confocal images showing the morphologies of mitochondria in Ad–NC– or Ad-mitofilin-infected NRVCs treated with PA. Scale bar, 25 μm. (D) Analysis of cells with fragmented mitochondria, as quantified in three or four randomly chosen fields of view (n > 200 cells) per experiment. n = 4 independent experiments. The data are presented as the mean ± SEM. Comparisons between two groups were conducted with unpaired Student's t-tests. (E) ATP content in BSA- or PA-treated NRVCs infected with Ad-NC or Ad-mitofilin. n = 6 independent experiments. (F) Representative immunofluorescence staining of NRVCs with α-actinin antibodies (red); the nuclei were stained with DAPI (blue). Scale bar, 100 μm. (G) Statistical analysis of cardiomyocyte surface area. n = 6 independent experiments; more than 100 cells were measured in each experiment. The data are presented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. ***P < 0.001 vs the BSA + Ad-NC group; #P < 0.05 and ###P < 0.001 vs the PA + Ad-NC group. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

108

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

(caption on next page)

109

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 4. ACh improved mitochondrial cristae morphology through the MAChR-mitofilin pathway. (A–C) The protein levels of mitofilin (red) in NRVCs were analysed by immunostaining using the indicated antibodies, and the nuclei were stained with DAPI (blue) and quantified. Scale bar, 50 μm n = 4 independent experiments; more than 100 cells were measured in each experiment. (D) The effect of ACh on cardiac mitochondrial cristae morphology was assessed by TEM. Scale bar, 100 nm. (E) Quantification of cristae with the morphologies shown in Fig. 4D. The numbers indicate the numbers of mitochondria that were counted. The data are presented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. ***P < 0.001 vs the BSA group; ###P < 0.001 vs the PA group; &&&P < 0.001 vs the PA + ACh group. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

74.14% (in the PA + Ad-mitofilin + AMPK siRNA group). These results indicate that AMPK regulates mitofilin to strengthen mitochondrial cristae. Recent studies have confirmed the interactions between mitofilin and other proteins located on the mitochondrial membrane that form the MICOS complex to repair mitochondrial cristae. Based on these data, we hypothesized that AMPK regulates the participation of mitofilin in the formation of the MICOS complex. As shown in Fig. 7C and D, compared with the control, both AMPK siRNA and PA treatment decreased the interaction of mitofilin with Sam50 and CHCHD3 (two representative core proteins of the MICOS complex). Compared with that in PA + Ad-NC + NC siRNA group cells, the interaction between mitofilin and Sam50/CHCHD3 was significantly greater in PA + Ad-mitofilin group cardiomyocytes, and this effect of mitofilin overexpression was reversed by AMPK knockdown. Taken together, these results support the conclusions that PA treatment downregulates the participation of mitofilin in MICOS complex formation and that AMPK mediates the interaction of mitofilin with other proteins in the MICOS complex.

mitochondria, decreasing the percentage of mitochondria with thin/ open and long lamellae from 66.47% (in the PA + ACh group) to 30.13% (in the PA + ACh + Atro group). Treatment of control cells with ACh alone had no significant effect on mitofilin expression or cristae remodelling. These findings suggest that ACh can upregulate mitofilin expression to protect against mitochondrial cristae remodelling in PA-treated cardiomyocytes through MAChR. In addition, ACh improved mitochondrial function in PA-treated NRVCs. As shown in Fig. 5A–C, ACh inhibited PA-induced mitochondrial ROS production in NRVCs. In addition, ACh relieved changes in mitochondrial membrane potential in NRVCs exposed to PA, as indicated by a significant increase in the red/green fluorescence ratio in the PA + ACh group compared to the PA group (Fig. 5D and E). Moreover, ACh increased the expression of mitochondrial complexes I, II, III and V in PA-treated NRVCs (Fig. 5F–K). 3.5. AMPK played a key role in mitochondrial cristae and was activated by ACh through MAChR

3.7. Knockdown of AMPK abolished the effects of ACh on MICOS complex formation and cardiomyocyte hypertrophy in PA-treated NRVCs

AMPK is involved in the regulation of mitochondrial morphology and function. As shown in Fig. 6A, AMPK phosphorylation was markedly decreased in the PA group compared with the BSA group; this decrease in AMPK phosphorylation was reversed by ACh treatment, the effect of which was abolished by Atro treatment. Furthermore, the pAMPK level was downregulated by CC in ACh- and PA-treated NRVCs and upregulated by AICAR in PA-treated NRVCs (Fig. 6B). Moreover, as shown in Fig. 6C and D, ACh treatment restored ATP content and decreased cytochrome C release from mitochondria to the cytoplasm in the presence of PA. This effect was abolished by CC. Furthermore, AICAR had an effect similar to that of ACh. These results suggest that AMPK plays a crucial role in the mito-protective effect of ACh on PAtreated NRVCs. To elucidate the role of AMPK in mitochondrial cristae remodelling, AMPK expression in NRVCs was inhibited by transfection with siRNA (Supplementary Fig. 5). As shown in Fig. 6E and F, unlike the mitochondrial cristae in the BSA + NC siRNA group, those in the BSA + AMPK siRNA group were disordered. These morphological observations were further confirmed by quantification of the percentages of mitochondria with thin/open and long, lamellar-shaped cristae, which dramatically decreased from 81.54% in the BSA + NC siRNA group to 41.59% in BSA + AMPK siRNA group, although there was no significant difference in mitofilin expression between these two groups (Supplementary Fig. 6). In summary, these results indicate that AMPK is an important regulator of mitochondrial cristae remodelling that also mediates the effects of ACh on mitochondria in NRVCs treated with PA.

We further used AMPK siRNA transfection to examine the effect of ACh on MICOS complex formation in PA-treated NRVCs. As shown in Fig. 8A and B, treatment with ACh reversed the decreases in mitofilinSam50 and mitofilin-CHCHD3 interactions. Transfection with AMPK siRNA inhibited the protective effects of ACh. Additionally, ACh reversed the effects of PA on mitochondrial ROS production and cardiomyocyte hypertrophy, but these effects were lost in NRVCs following AMPK knockdown (Fig. 8C–F), confirming that ACh activates AMPK in PA-treated NRVCs. 4. Discussion This study yielded several findings. (1) PA induced the formation of aberrant mitochondria with balloon-like and non-lamellar cristae accompanied by mitochondrial fragmentation and dysfunction (e.g., decreased mitochondrial complex expression and increased mitochondrial ROS production), which eventually led to cellular hypertrophy in NRVCs. (2) Overexpression of mitofilin partly mitigated mitochondrial cristae remodelling and functional damage; furthermore, mitofilin overexpression in PA-treated NRVCs decreased cardiomyocyte size, and ACh upregulated mitofilin via MAChR. (3) AMPK silencing by siRNA transfection physiologically disrupted mitochondrial cristae morphology, and AMPK was activated through MAChR by ACh in NRVCs treated with PA. (4) AMPK knockdown inhibited the positive effect of mitofilin overexpression on PA-treated cardiomyocytes and decreased the interactions of mitofilin with Sam50 and CHCHD3 to form the MICOS complex. (5) AMPK siRNA transfection attenuated the ACh-induced increases in MICOS complex formation, mitochondrial ROS production, and cardiomyocyte hypertrophy. Overall, these findings show that AMPK promotes the involvement of mitofilin in MICOS complex formation, which may be an essential intermediate step regulating mitochondrial cristae remodelling, and that ACh improves mitochondrial damage during PA-induced myocardial hypertrophy through the MAChR-mitofilin and MAChR -AMPK-MICOS pathway (Fig. 9). A growing body of evidence indicates that mitochondrial

3.6. AMPK promoted the interaction of mitofilin with other proteins associated with mitochondrial cristae remodelling To investigate the relationship between the AMPK and mitofilin proteins in the pathological regulation of mitochondrial cristae, both mitofilin overexpression and AMPK knockdown were carried out in PAtreated NRVCs, the results of which are shown in Fig. 7A and B. Transfection of NRVCs with AMPK siRNA blocked the effects of mitofilin overexpression on mitochondrial cristae following PA treatment and increased the proportion of NRVCs containing mitochondria with open and short lamellar cristae, balloon-like cristae and non-lamellar cristae from 50.87% (in the PA + Ad-mitofilin + NC siRNA group) to 110

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 5. ACh diminished PA-induced mitochondrial dysfunction. (A–C) ACh treatment attenuated the PA-induced increases in mitochondrial ROS levels. Scale bar, 50 μm n = 4 independent experiments; more than 100 cells were measured in each experiment. (D–E) The mitochondrial membrane potential in NRVCs was measured by JC-1 staining and qualified by changes in the red/green fluorescence ratio. Scale bar, 50 μm n = 4 independent experiments; more than 100 cells were measured in each experiment. (F–K) The effect of ACh on the expression of mitochondrial complexes I–V in PA-treated NRVCs. n = 4 independent experiments. The data are presented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. *P < 0.05, **P < 0.01, and ***P < 0.001 vs the BSA group; #P < 0.05, ##P < 0.01, and ###P < 0.001 vs the PA group. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

dysfunction plays an important role in cardiomyopathy induced by metabolic disorders. Generally, excessive levels of free fatty acids lead to mitochondrial uncoupling and ROS overproduction and therefore disturb mitochondrial function. Furthermore, mitochondrial damage

decreases ATP production, and damaged mitochondria produce amounts of ROS that trigger detrimental reactions, resulting in cardiac dysfunction [21,22]. Recent research has shown that mitochondrial cristae are the true bioenergetic components in cells [23] and that 111

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 6. AMPK regulated mitochondrial cristae remodelling and was activated by ACh through MAChR. (A) ACh increased the phosphorylation of AMPK after PA induction, but this effect was blocked by Atro. n = 4 independent experiments. (B) The effects of ACh and AICAR (an AMPK agonist) on AMPK phosphorylation in PAtreated NRVCs and the effect of CC (an AMPK inhibitor) on AMPK phosphorylation in PA-treated NRVCs treated with ACh. n = 4 independent experiments. (C) The effects of ACh and AICAR on ATP production in PA-treated NRVCs and the effect of CC on ATP production in PA-treated NRVCs treated with ACh. n = 6 independent experiments. (D) The effects of ACh and AICAR on cytochrome C release in PA-treated NRVCs and the effect of CC on cytochrome C release in PA-treated NRVCs treated with ACh. n = 3 independent experiments. (E) AMPK knockdown in NRVCs inhibited mitochondrial cristae remodelling. Scale bar, 100 nm. (F) Quantification of cristae with the morphologies shown in Fig. 6E. The numbers indicate the numbers of mitochondria that were counted. The data are presented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. **P < 0.01 vs the BSA group; ##P < 0.01 and ###P < 0.001 vs the PA group; & P < 0.05, &&P < 0.01, and &&&P < 0.001 vs the PA + ACh group.

respiratory chain supercomplex assembly and respiratory efficiency depend on cristae shape [7]. However, cristae remodelling in lipotoxic cardiomyopathy has been poorly explored. Here, we report observing not only increased mitochondrial fragmentation, which has been the focus of most related research [24], but also, and more interestingly,

enormous numbers of loose, small and even lysed cristae in PA-treated NRVCs in parallel with increased mitochondrial production of ROS and decreased expression of mitochondrial respiratory chain complexes I, II, III, and V. These findings are similar to those of studies on T cells in which mitochondrial fusion has been found to result in compact cristae, 112

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 7. Knockdown of AMPK abolished the effect of mitofilin overexpression on mitochondrial cristae and decreased the participation of mitofilin in MICOS complex assembly in PA-treated NRVCs. (A) Knockdown of AMPK abolished the effect of mitofilin overexpression on mitochondrial cristae in PA-treated NRVCs. Scale bar, 100 nm. (B) Quantification of cristae with the morphology shown in Fig. 7A. The numbers indicate the numbers of mitochondria that were counted. (C–D) Immunoprecipitation and Western blot analysis results showing the interactions of mitofilin with CHCHD3 and Sam50 in NRVCs transfected with AMPK siRNA. n = 4 independent experiments. (E) Schematic diagram of the effect of AMPK on mitochondrial cristae remodelling. AMPK knockdown led to abnormal cristae and decreased MICOS complex formation. The data are presented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. **P < 0.01 and ***P < 0.001 vs the BSA + Ad-NC + NC siRNA group; #P < 0.05 and ##P < 0.01 vs the PA + Ad-NC + NC siRNA group; &&P < 0.01 vs the PA + Ad-mitofilin + NC siRNA group.

113

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 8. AMPK siRNA diminished the effects of ACh on mitochondrial cristae and cardiomyocyte hypertrophy in PA-treated NRVCs. (A–B) Changes in the interactions between mitofilin and CHCHD3 and between mitofilin and Sam50 after transfection of PA-treated NRVCs treated with or without ACh with AMPK/NC siRNA. n = 6 independent experiments. (C–D) The effect of AMPK/NC siRNA transfection on mitochondrial ROS production in PA-treated NRVCs treated with or without ACh. Scale bar, 25 μm n = 4 independent experiments; more than 100 cells were measured in each experiment. (E–F) Changes in cardiomyocyte surface area after AMPK/NC siRNA transfection in PA-treated NRVCs with or without ACh treatment. Scale bar, 100 μm n = 4 independent experiments; more than 100 cells were measured in each experiment. The data are presented as the mean ± SEM. One-way ANOVA and Tukey's multiple comparisons test. **P < 0.01 and ***P < 0.001 vs the PA + NC siRNA group; ##P < 0.01 and ###P < 0.001 vs the PA + ACh + NC siRNA group.

prolific oxidative phosphorylation and divided mitochondria, resulting in disordered cristae and redox imbalance [25]. Taken together, these findings suggest cristae remodelling as a treatment target for metabolic disorder-associated cardiac remodelling. Mitofilin is a structural protein of the inner mitochondrial membrane that plays an important role in maintaining cristae shape [26]. Baseler et al. showed that mitofilin is diminished in the hearts of patients with type 1 diabetes and that transgenic overexpression of mitofilin attenuates diabetes-induced changes in cardiac and mitochondrial function [27,28]. In contrast, Gutiérrez et al. found that the

expression of mitofilin in obese rats is not significantly downregulated compared to that in control rats [29]. Interestingly, we found that in vitro, myocardial mitofilin expression decreased continuously with increasing PA incubation time. Moreover, overexpression of mitofilin not only improved cristae remodelling but also promoted mitochondrial network formation and function and decreased PA-induced myocardial hypertrophy. Notably, Zhang et al. revealed that mitofilin is upregulated in pathological cardiac hypertrophy and that mitofilin overexpression promotes cardiac hypertrophy under pressure overload stress [30], indicating that mitofilin may play various roles; further 114

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

Fig. 9. ACh improved mitochondrial cristae morphology and function in cardiomyocytes with PAinduced hypertrophy, possibly through upregulation of mitofilin and activation of AMPK to promote MICOS complex formation via MAChR. ACh: acetylcholine; AMPK: AMP-activated protein kinase; ANP: atrial natriuretic peptide; ATP: adenosine triphosphate; BNP: brain natriuretic peptide; β-MHC: β-myosin heavy chain; CHCHD3: coiled-coil helix coiled-coil helix domain-containing protein 3; MAChR: muscarinic acetylcholine receptor; MICOS: mitochondrial contact site and cristae organizing system; Sam50: sorting and assembly machinery 50; PA: palmitate; ROS: reactive oxygen species; Ψm: mitochondrial membrane potential.

is involved in regulating cristae remodelling. In fact, Toyama et al. found that AMPK interacts with mitochondrial fission factor (MFF) to participate in mitochondrial fission in response to energy stress [35]. Wang et al. further showed that AMPK activation suppresses Drp1mediated mitochondrial fission in a mouse model of diabetes-accelerated atherosclerosis [36]. These results indicate that AMPK plays an indispensable role in mitochondrial fission. In contrast, in our study, transfection with AMPK siRNA led to misarranged and lysed cristae, which further confirmed the necessity of AMPK in the maintenance of mitochondrial cristae morphology. In summary, the protective effects of ACh on PA-treated mitochondrial cristae are mediated in part by AMPK activation. The morphology of the inner mitochondrial membrane has been found to be altered by the recently confirmed MICOS complex. The MICOS complex suppsorts the interactions of a wide variety of different proteins in different mitochondrial locations [9]. Mitofilin was the first MICOS complex core component shown to interact with multiple proteins (Sam50, located at the outer mitochondrial membrane, and CHCHD3, located at the inner mitochondrial membrane) [37,38]. Generally, MICOS complex dysfunction alters cristae shape, and vice versa [39], which is consistent with our observation that the interactions between mitofilin and Sam50/CHCHD3 were reduced following PA-induced cristae injury and that both mitofilin overexpression and ACh treatment increased protein-protein interactions to improve cristae structure. Although a stable MICOS complex is required to maintain mitochondrial cristae structure and function, the mechanisms that underlie MICOS complex regulation remain poorly defined. Previous research has shown that OMA1, an inner mitochondrial membrane protease, maintains mitochondrial ultrastructure through specific interactions with mitofilin to influence the recruitment of MICOS complex components without affecting mitochondrial fission- and fusion-associated proteins [40]. Inconsistent with these results, OPA1, a well-known mitochondrial fusion protein, has been found to function upstream of the MICOS complex to regulate cristae number and stability [41]. Surprisingly, our studies showed that AMPK knockdown blocked the mitofilin-Sam50 and mitofilin-CHCHD3 interactions and inhibited the positive effects of mitofilin overexpression and ACh on cristae remodelling and mitochondrial function, indicating that AMPK is also an important regulator of MICOS complex formation. However, further studies exploring the potential mechanism by which AMPK regulates the MICOS complex are needed. ACh has a dramatic ability to maintain mitochondrial homeostasis. Compared to the mechanisms of other drugs targeting mitochondria,

study on mitofilin is needed to explain its roles in multiple cardiac hypertrophy models. Taken together, our findings suggest a link between decreased mitofilin expression, cristae damage, and mitochondrial fragmentation and dysfunction in cardiomyocytes with PA-induced hypertrophy. Therefore, mitofilin regulation may be a potential strategy to protect against mitochondria-associated myocardial injury. Increasing evidence has shown a correlation between sympathetic overactivity/vagus nerve withdrawal and metabolic syndrome-related cardiovascular disease, and vagus nerve stimulation has been demonstrated to prevent cardiac dysfunction by improving mitochondrial function in obese insulin-resistant rats [12]. Our previous studies revealed that pyridostigmine, a cholinesterase inhibitor, improves mitochondrial and cristae morphology and function to alleviate obesityinduced cardiac remodelling [13]. Additionally, ACh benefits cardiomyocytes and endothelial cells by mediating mitophagy and the mitochondrial unfolded protein response after hypoxia/reoxygenation [15,31]. However, little information on the effects of ACh on mitochondrial and cristae shape in metabolic disorder-related cardiovascular disease is available. Importantly, this study indicated that ACh upregulates mitofilin via MAChR and therefore attenuates changes in mitochondrial cristae structure to defend against PA-induced mitochondrial and myocardial damage. However, since the percentage of relatively healthy mitochondria (mitochondria displaying thin/open and long lamellar-shaped cristae) was significantly lower in PA-treated NRVCs overexpressing mitofilin (48.37%) than in PA-treated NRVCs with ACh treatment (66.47%), it is reasonable to speculate that other mechanisms participate in the regulation of mitochondrial cristae remodelling by ACh. We further explored the signalling pathway by which ACh treatment improved mitochondrial health. Previous studies have demonstrated that vagal activation prevents cardiovascular damage by affecting AMPK-related pathways in the context of many pathological conditions, including ischaemia, pressure overload and long-term high-fat diet consumption [17,32,33]. We also found that ACh treatment increased AMPK phosphorylation during PA treatment and that Atro abrogated the effects of ACh, indicating that ACh treatment increases AMPK phosphorylation, at least in part, through MAChR. AMPK is a crucial regulator of energy metabolism and mitochondrial bioenergetics [34]. Consistent with this role, we found that ACh and AICAR, an AMPK activator, promoted ATP production and that CC, an AMPK inhibitor, reversed the protective effect of ACh. Furthermore, AMPK activation restricted cytochrome C release from mitochondria, and because 85% of the total cytochrome C is stored in cristae, we hypothesized that AMPK 115

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al.

the enhancement of vagal activity by ACh represents strengthening of a built-in physiological protective mechanism. Moreover, ACh and improved vagal activity exert mito-protective effects in multiple ways, such as through redox state modulation, Ca2+ regulation and mitochondrial biogenesis [42]. The present study shows that ACh alleviates PA-induced cristae damage by both upregulating mitofilin expression and activating AMPK to promote MICOS complex formation. Notably, our previous research confirmed that pyridostigmine activates vagal activity and improves cardiac mitochondrial cristae shape in a mouse model of metabolic syndrome [13]; however, there is insufficient evidence to prove the link between vagal activation and cristae modulation. Therefore, further research on the effects of ACh treatment and vagal nerve stimulation on mitochondrial cristae remodelling in animal models of lipotoxic cardiomyopathy is needed. Additionally, the concentration of ACh used in our research is much higher than the concentration that would be present in vivo; indeed, the concentration of ACh produced in response to vagus nerve stimulation in the rabbit heart is in the nanomolar range [43]. However, in the synaptic space, ACh concentrations can reach millimolar levels due to quantal release [44,45]. Furthermore, the therapeutic concentration of ACh is usually higher than that encountered under physiological conditions [46,47]. In the present study, 10−6 mol/L ACh had prominent effects on PA-induced mitochondrial cristae injury and myocardial hypertrophy; thus, 10−6 mol/L ACh was used throughout our study, which is consistent with the ACh concentrations used in previous reports [31,48]. Of note, our current study does have some limitations. Cardiac remodelling caused by obesity is a consequence of multiple factors, and several mechanisms have been proposed, such as cardiac metabolism alterations, inflammation, insulin signalling impairment, hyperglycaemia, pressure/volume overload, extracellular matrix changes, fibrosis, mitochondrial dysfunction and oxidative stress; furthermore, these mechanisms are not independent but rather are mutually reinforcing [22,49,50]. The present study was conducted purely in vitro and showed excess supply of saturated fatty acids to mitochondrial cristae in cardiomyocytes. Although lipid accumulation-induced mitochondrial dysfunction plays a fatal role in intracellular metabolic disorder [13], the present research does not prove the involvement of mitochondrial cristae remodelling in PA-induced cardiomyocyte metabolic damage (for example, insulin insensitivity). Additionally, although there were multiple positive effects of ACh on PA-induced mitochondrial remodelling, ACh cannot be developed into a drug for clinical treatment because of its ease of hydrolysis and multiple effects throughout the body. However, ACh is the major neurotransmitter in the vagus nerve, and in addition to vagal stimulation, some novel pharmacological approaches that work by increasing ACh concentrations, such as choline (a cholinergic drug) and pyridostigmine (a reversible acetylcholinesterase inhibitor), might augment vagal nerve function [51]. Therefore, the in vitro mito-protective effects of ACh as a pharmacological tool provide a theoretical basis for the use of mitochondria-targeting drugs that activate the vagus nerve. In summary, the salient findings of this study are that ACh can improve mitochondrial cristae morphology and function in cardiomyocytes with PA-induced hypertrophy by upregulating mitofilin and activating AMPK to promote MICOS complex formation through MAChR. These findings provide important insights into the molecular mechanisms underlying the regulation of mitochondrial cristae remodelling and the cardio-protective effects of ACh, which may aid in the development of future therapeutic approaches and pharmacological treatments to prevent or alleviate lipotoxic cardiomyopathy.

Acknowledgements This research was supported by grants from the National Natural Science Foundation of China (Nos. 81770293, 81970221, and 91649106). Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.freeradbiomed.2019.09.025. References [1] I. Zlobine, K. Gopal, J.R. Ussher, Lipotoxicity in obesity and diabetes-related cardiac dysfunction, Biochim. Biophys. Acta Mol. Cell Biol. Lipids 1861 (2016) 1555–1568. [2] B. Chaurasia, S.A. Summers, Ceramides – lipotoxic inducers of metabolic disorders, Trends Endocrinol. Metab. 26 (2015) 538–550. [3] J.L. Fetterman, M. Holbrook, D.G. Westbrook, J.A. Brown, K.P. Feeley, R. BretónRomero, E.A. Linder, B.D. Berk, R.M. Weisbrod, M.E. Widlansky, N. Gokce, S.W. Ballinger, N.M. Hamburg, Mitochondrial DNA damage and vascular function in patients with diabetes mellitus and atherosclerotic cardiovascular disease, Cardiovasc. Diabetol. 15 (2016). [4] M.A. Aon, C.G. Tocchetti, N. Bhatt, N. Paolocci, S. Cortassa, Protective mechanisms of mitochondria and heart function in diabetes, Antioxid. Redox Sign. 22 (2015) 1563–1586. [5] J.D. Schilling, The mitochondria in diabetic heart failure: from pathogenesis to therapeutic promise, Antioxid. Redox Sign. 22 (2015) 1515–1526. [6] L. Pernas, L. Scorrano, Mito-morphosis: mitochondrial fusion, fission, and cristae remodeling as key mediators of cellular function, Annu. Rev. Physiol. 78 (2016) 505–531. [7] S. Cogliati, C. Frezza, M.E. Soriano, T. Varanita, R. Quintana-Cabrera, M. Corrado, S. Cipolat, V. Costa, A. Casarin, L.C. Gomes, E. Perales-Clemente, L. Salviati, P. Fernandez-Silva, J.A. Enriquez, L. Scorrano, Mitochondrial cristae shape determines respiratory chain supercomplexes assembly and respiratory efficiency, Cell 155 (2013) 160–171. [8] S. Cogliati, J.A. Enriquez, L. Scorrano, Mitochondrial cristae: where beauty meets functionality, Trends Biochem. Sci. 41 (2016) 261–273. [9] M. van der Laan, S.E. Horvath, N. Pfanner, Mitochondrial contact site and cristae organizing system, Curr. Opin. Cell Biol. 41 (2016) 33–42. [10] L.R. Wulsin, P.S. Horn, J.L. Perry, J.M. Massaro, R.B. D'Agostino, Autonomic imbalance as a predictor of metabolic risks, cardiovascular disease, diabetes, and mortality, J. Clin. Endocrinol. Metab. 100 (2015) 2443–2448. [11] R. Pop-Busui, Cardiac autonomic neuropathy in diabetes: a clinical perspective, Diabetes Care 33 (2010) 434–441. [12] B. Samniang, K. Shinlapawittayatorn, T. Chunchai, W. Pongkan, S. Kumfu, S.C. Chattipakorn, B.H. KenKnight, N. Chattipakorn, Vagus nerve stimulation improves cardiac function by preventing mitochondrial dysfunction in obese-insulin resistant rats, Sci. Rep. 6 (2016). [13] R. Xue, X. Yu, M. Zhao, M. Xu, Q. Wu, Y. Cui, S. Yang, D. Li, W. Zang, Pyridostigmine alleviates cardiac dysfunction via improving mitochondrial cristae shape in a mouse model of metabolic syndrome, Free Radical Biomed. 134 (2019) 119–132. [14] L. Sun, M. Zhao, X. Yu, H. Wang, X. He, J. Liu, W. Zang, Cardioprotection by acetylcholine: a novel mechanism via mitochondrial biogenesis and function involving the PGC-1α pathway, J. Cell. Physiol. 228 (2013) 1238–1248. [15] M. Xu, X. Bi, X. He, X. Yu, M. Zhao, W. Zang, Inhibition of the mitochondrial unfolded protein response by acetylcholine alleviated hypoxia/reoxygenation-induced apoptosis of endothelial cells, Cell Cycle 15 (2016) 1331–1343. [16] M.K. Gupta, F.G. Tahrir, T. Knezevic, M.K. White, J. Gordon, J.Y. Cheung, K. Khalili, A.M. Feldman, GRP78 interacting partner Bag5 responds to ER stress and protects cardiomyocytes from ER stress-induced apoptosis, J. Cell. Biochem. 117 (2016) 1813–1821. [17] M. Xu, R. Xue, Y. Lu, S. Yong, Q. Wu, Y. Cui, X. Zuo, X. Yu, M. Zhao, W. Zang, Choline ameliorates cardiac hypertrophy by regulating metabolic remodelling and UPRmt through SIRT3-AMPK pathway, Cardiovasc. Res. 115 (2019) 530–545. [18] V. Eisner, M. Picard, G. Hajnoczky, Mitochondrial dynamics in adaptive and maladaptive cellular stress responses, Nat. Cell Biol. 20 (2018) 755–765. [19] M.G. Sun, J. Williams, C. Munoz-Pinedo, G.A. Perkins, J.M. Brown, M.H. Ellisman, D.R. Green, T.G. Frey, Correlated three-dimensional light and electron microscopy reveals transformation of mitochondria during apoptosis, Nat. Cell Biol. 9 (2007) 1057–1065. [20] H. Otera, N. Miyata, O. Kuge, K. Mihara, Drp1-dependent mitochondrial fission via MiD49/51 is essential for apoptotic cristae remodeling, J. Cell Biol. 212 (2016) 531–544. [21] V. Parra, H.E. Verdejo, M. Iglewski, A. Del Campo, R. Troncoso, D. Jones, Y. Zhu, J. Kuzmicic, C. Pennanen, C. Lopez Crisosto, F. Jaña, J. Ferreira, E. Noguera, M. Chiong, D.A. Bernlohr, A. Klip, J.A. Hill, B.A. Rothermel, E.D. Abel, A. Zorzano, S. Lavandero, Insulin stimulates mitochondrial fusion and function in cardiomyocytes via the Akt-mTOR-NFκB-Opa-1 signaling pathway, Diabetes 63 (2013) 75–88. [22] H. Bugger, E.D. Abel, Mitochondria in the diabetic heart, Cardiovasc. Res. 88 (2010) 229–240. [23] L.E. Formosa, M.T. Ryan, Mitochondrial OXPHOS complex assembly lines, Nat. Cell

Conflicts of interest The authors declare no conflicts of interest. 116

Free Radical Biology and Medicine 145 (2019) 103–117

R.-Q. Xue, et al. Biol. 20 (2018) 511–513. [24] C.A. Galloway, Y. Yoon, Mitochondrial dynamics in diabetic cardiomyopathy, Antioxid. Redox Sign. 22 (2015) 1545–1562. [25] M.D. Buck, D. O Sullivan, R.I. Klein Geltink, J.D. Curtis, C. Chang, D.E. Sanin, J. Qiu, O. Kretz, D. Braas, G.J.W. van der Windt, Q. Chen, S.C. Huang, C.M. O Neill, B.T. Edelson, E.J. Pearce, H. Sesaki, T.B. Huber, A.S. Rambold, E.L. Pearce, Mitochondrial dynamics controls T cell fate through metabolic programming, Cell 166 (2016) 63–76. [26] V.S. Van Laar, P.A. Otero, T.G. Hastings, S.B. Berman, Potential role of mic60/ mitofilin in Parkinson's disease, Front. Neurosci. 12 (2019). [27] W.A. Baseler, E.R. Dabkowski, C.L. Williamson, T.L. Croston, D. Thapa, M.J. Powell, T.T. Razunguzwa, J.M. Hollander, Proteomic alterations of distinct mitochondrial subpopulations in the type 1 diabetic heart: contribution of protein import dysfunction, Am. J. Physiol. Regul. Integr. Comp. Physiol. 300 (2011) R186–R200. [28] D. Thapa, C.E. Nichols, S.E. Lewis, D.L. Shepherd, R. Jagannathan, T.L. Croston, K.J. Tveter, A.A. Holden, W.A. Baseler, J.M. Hollander, Transgenic overexpression of mitofilin attenuates diabetes mellitus-associated cardiac and mitochondria dysfunction, J. Mol. Cell. Cardiol. 79 (2015) 212–223. [29] G. Gutiérrez-Salmeán, P. Ortiz-Vilchis, C.M. Vacaseydel, L. Garduño-Siciliano, G. Chamorro-Cevallos, E. Meaney, S. Villafaña, F. Villarreal, G. Ceballos, I. RamírezSánchez, Effects of (−)-epicatechin on a diet-induced rat model of cardiometabolic risk factors, Eur. J. Pharmacol. 728 (2014) 24–30. [30] Y. Zhang, J. Xu, Y. Luo, X. An, R. Zhang, G. Liu, H. Li, H. Chen, D. Liu, Overexpression of mitofilin in the mouse heart promotes cardiac hypertrophy in response to hypertrophic stimuli, Antioxid. Redox Sign. 21 (2014) 1693–1707. [31] L. Sun, M. Zhao, Y. Yang, R. Xue, X. Yu, J. Liu, W. Zang, Acetylcholine attenuates hypoxia/reoxygenation injury by inducing mitophagy through PINK1/parkin signal pathway in H9c2 cells, J. Cell. Physiol. 231 (2016) 1171–1181. [32] S.S. Kong, J.J. Liu, X.J. Yu, Y. Lu, W.J. Zang, Protection against ischemia-induced oxidative stress conferred by vagal stimulation in the rat heart: involvement of the AMPK-PKC pathway, Int. J. Mol. Sci. 13 (2012) 14311–14325. [33] Y. Lu, Q. Wu, L. Liu, X. Yu, J. Liu, M. Li, W. Zang, Pyridostigmine protects against cardiomyopathy associated with adipose tissue browning and improvement of vagal activity in high-fat diet rats, Biochim. Biophys. Acta (BBA) - Mol. Basis Dis. 1864 (2018) 1037–1050. [34] S. Herzig, R.J. Shaw, AMPK: guardian of metabolism and mitochondrial homeostasis, Nat. Rev. Mol. Cell Biol. 19 (2017) 121–135. [35] E.Q. Toyama, S. Herzig, J. Courchet, L.T. Jr, O.C. Losón, K. Hellberg, N.P. Young, H. Chen, F. Polleux, D.C. Chan, Metabolism. AMP-activated protein kinase mediates mitochondrial fission in response to energy stress, Science 351 (2016) 275–281. [36] Q. Wang, M. Zhang, G. Torres, S. Wu, C. Ouyang, Z. Xie, M.H. Zou, Metformin suppresses diabetes-accelerated atherosclerosis via the inhibition of Drp1-mediated mitochondrial fission, Diabetes 66 (2017) 193–205. [37] M. Darshi, S.S. Taylor, Mitochondrial ChChD3 acts as a scaffold for mitofilin, Sam50 and PKA, FASEB J. 22 (2008) 620–645.

[38] D. Tarasenko, M. Barbot, D.C. Jans, B. Kroppen, B. Sadowski, G. Heim, W. Möbius, S. Jakobs, M. Meinecke, The MICOS component Mic60 displays a conserved membrane-bending activity that is necessary for normal cristae morphology, J. Cell Biol. 216 (2017) 889–899. [39] A.K. Alkhaja, D.C. Jans, M. Nikolov, M. Vukotic, O. Lytovchenko, F. Ludewig, W. Schliebs, D. Riedel, H. Urlaub, S. Jakobs, M. Deckers, MINOS1 is a conserved component of mitofilin complexes and required for mitochondrial function and cristae organization, Mol. Biol. Cell 23 (2012) 247–257. [40] R.M. Levytskyy, M.P. Viana, O. Khalimonchuk, Protease OMA1 modulates mitochondrial metabolism and cristae structure through interaction with MICOS complex, FASEB J. 32 (2018) 543–546. [41] C. Glytsou, E. Calvo, S. Cogliati, A. Mehrotra, I. Anastasia, G. Rigoni, A. Raimondi, N. Shintani, M. Loureiro, J. Vazquez, L. Pellegrini, J.A. Enriquez, L. Scorrano, M.E. Soriano, Optic atrophy 1 is epistatic to the core MICOS component MIC60 in mitochondrial cristae shape control, Cell Rep. 17 (2016) 3024–3034. [42] X. He, M. Zhao, X. Bi, L. Sun, X. Yu, M. Zhao, W. Zang, Novel strategies and underlying protective mechanisms of modulation of vagal activity in cardiovascular diseases, Br. J. Pharmacol. 172 (2015) 5489–5500. [43] T. Kawada, T. Akiyama, S. Shimizu, A. Kamiya, K. Uemura, M. Li, M. Shirai, M. Sugimachi, Detection of endogenous acetylcholine release during brief ischemia in the rabbit ventricle: a possible trigger for ischemic preconditioning, Life Sci. 85 (2009) 597–601. [44] L.V. Borovikova, S. Ivanova, M. Zhang, H. Yang, G.I. Botchkina, L.R. Watkins, H. Wang, N. Abumrad, J.W. Eaton, K.J. Tracey, Vagus nerve stimulation attenuates the systemic inflammatory response to endotoxin, Nature 405 (2000) 458–462. [45] A.G. Engel, X. Shen, D. Selcen, S. Sine, New horizons for congenital myasthenic syndromes, Ann. Acad. Sci. 1275 (2012) 54–62. [46] M.H. Kim, M.O. Kim, J.S. Heo, J.S. Kim, H.J. Han, Acetylcholine inhibits long-term hypoxia-induced apoptosis by suppressing the oxidative stress-mediated MAPKs activation as well as regulation of Bcl-2, c-IAPs, and caspase-3 in mouse embryonic stem cells, Apoptosis 13 (2008) 295–304. [47] Q. Qin, J.M. Downey, M.V. Cohen, Acetylcholine but not adenosine triggers preconditioning through PI3-kinase and a tyrosine kinase, Am. J. Physiol. Heart Circ. Physiol. 284 (2003) H727–H734. [48] D. Li, J. Liu, B. Liu, H. Hu, L. Sun, Y. Miao, H. Xu, X. Yu, X. Ma, J. Ren, W. Zang, Acetylcholine inhibits hypoxia-induced tumor necrosis factor-α production via regulation of MAPKs phosphorylation in cardiomyocytes, J. Cell. Physiol. 226 (2011) 1052–1059. [49] E.D. Abel, S.E. Litwin, G. Sweeney, Cardiac remodeling in obesity, Physiol. Rev. 88 (2008) 389–419. [50] M.A. Aon, C.G. Tocchetti, N. Bhatt, N. Paolocci, S. Cortassa, Protective mechanisms of mitochondria and heart function in diabetes, Antioxid. Redox Sign. 22 (2015) 1563–1586. [51] L. Liu, M. Zhao, X. Yu, W. Zang, Pharmacological modulation of vagal nerve activity in cardiovascular diseases, Neurosci. Bull. 35 (2019) 156–166.

117