Melatonin activates Parkin translocation and rescues the impaired mitophagy activity of diabetic cardiomyopathy through Mst1 inhibition

Abstract Mitophagy eliminates dysfunctional mitochondria and thus plays a cardinal role in diabetic cardiomyopathy (DCM). We observed the favourable effects of melatonin on cardiomyocyte mitophagy in mice with DCM and elucidated their underlying mechanisms. Electron microscopy and flow cytometric analysis revealed that melatonin reduced the number of impaired mitochondria in the diabetic heart. Other than decreasing mitochondrial biogenesis, melatonin increased the clearance of dysfunctional mitochondria in mice with DCM. Melatonin increased LC3 II expression as well as the colocalization of mitochondria and lysosomes in HG‐treated cardiomyocytes and the number of typical autophagosomes engulfing mitochondria in the DCM heart. These results indicated that melatonin promoted mitophagy. When probing the mechanism, increased Parkin translocation to the mitochondria may be responsible for the up‐regulated mitophagy exerted by melatonin. Parkin knockout counteracted the beneficial effects of melatonin on the cardiac mitochondrial morphology and bioenergetic disorders, thus abolishing the substantial effects of melatonin on cardiac remodelling with DCM. Furthermore, melatonin inhibited Mammalian sterile 20‐like kinase 1 (Mst1) phosphorylation, thus enhancing Parkin‐mediated mitophagy, which contributed to mitochondrial quality control. In summary, this study confirms that melatonin rescues the impaired mitophagy activity of DCM. The underlying mechanism may be attributed to activation of Parkin translocation via inhibition of Mst1.


| INTRODUCTION
The global prevalence of diabetes is estimated to affect 439 million adults (aged 20-79 years) by 2030. There will be a 69% increase in the number of diabetic patients in developing countries between 2010 and 2030. 1 Diabetic cardiomyopathy accounts for a significant increase of mortality in diabetic patients. 2 However, current treatment strategies for cardiomyopathy in diabetic patients are the same as in non-diabetic patients and do not address the underlying causes of cardiac dysfunction. 3,4 Mitochondria are organelles that produce the energy that is required to drive the endergonic reactions of cell life and are part of the signalling process. 5 The heart is rich in mitochondria to maintain normal contractile function due to the high energy demands of this organ as well as other tissues. [6][7][8] Cardiac mitochondria play a critical role and act as a power plant for cardiomyocytes through oxidative phosphorylation. [9][10][11] Increasing evidence suggests that the cardiovascular complications of diabetes converge on mitochondria as an epicentre for cardiomyocyte damage. 12 Cardiomyocytes are particularly susceptible to diabetic insults caused by dysfunctional mitochondria. 13,14 Mitochondria that are damaged during diabetic stress can produce reactive oxygen species (ROS) and release death-inducing factors, thus augmenting cardiomyocyte injury. [14][15][16] Effective clearance of these damaged mitochondria may prevent cardiomyocyte damage induced by diabetes and limit cardiac dysfunction. 13,17 Autophagy is the primary mechanism for mitochondrial quality control and removes whole mitochondria. In contrast to nonselective autophagy, mitophagy is a more specific mechanism that maintains a healthy mitochondrial population. [18][19][20] During mitophagy, damaged mitochondria are selectively sequestered by phagophores, encapsulated and subsequently fused with lysosomes to recycle their essential components. 20 Mitophagy ensures that cardiomyocytes maintain a functional network of mitochondria and provides strong protection against diabetic insults. Mitophagy can be regulated by several pathways during mitochondrial quality control. 13,20,21 Amongst these, the PINK1/PARKIN pathway is well-studied and is involved in the labelling of damaged mitochondria for mitophagy degradation. 22 Thus, a better understanding of whether PINK1/PAR-KIN-mediated mitophagy participates in the pathogenesis of diabetic cardiomyopathy is still needed as well as whether they can be targeted therapeutically.
Melatonin (N-acetyl-5-methoxytryptamine) is predicted to have evolved an estimated 3.0-2.5 billion years ago. 23,24 The chemical structure of melatonin has remained very stable for billions of years, and its structure is identical from cyanobacteria to human beings. [25][26][27] In recent years, a vast number of studies have documented the involvement of melatonin in cardiac protection. [28][29][30][31][32][33] Melatonin presumably enters mitochondria through oligopeptide transporters. 34 Measurement of the subcellular distribution of melatonin showed that the concentration of melatonin in mitochondria greatly exceeds that in blood. 35,36 This evidence suggests that melatonin may specifically target mitochondria. Onphachanh and colleagues demonstrated that melatonin stimulates PINK1 expression via the MT2/Akt/NF-κB pathway, which is important for the prevention of neuronal cell injury under high glucose conditions. 37 Our previous study also demonstrated that melatonin alleviates cardiac remodelling and dysfunction in DCM. 38

| Animals and pharmaceutical intervention
The in vivo study was implemented in accordance with the NIH guidelines on the use of laboratory animals. All animal protocols were approved by the Institutional Animal Care of the Fourth Military Medical University.
Mst1-null (Mst1 −/− ) and Mst1 transgenic (Mst1 Tg) mice (C57BL/6 background) were generated by K&D Gene Technology (Wuhan, China). After multiplying, gene-modified mice were identified by western blot and real-time PCR analyses. The littermates of wild-type mice were used as controls. As previously described, 39 eight-week-old mice (20-25 g, male) were given an intraperitoneal injection of streptozotocin (50 mg/kg for 5 consecutive days) to induce the diabetes model. Blood glucose was measured 7 days after the last dose of STZ, only those with random blood glucose levels ≥16.6 mmol/L can be labelled with diabetes. After dissolving with ethanol, melatonin was diluted in distilled water at a dose of 20 mg/kg/d and then administered by oral gavage for 4 weeks. 39 Several groups were set as follows: (a) wild type (WT, n = 32);

| Primary neonatal cardiomyocyte culture, transfection and treatment
The ventricles of the neonatal heart were isolated from 1-day-old wild-type C57BL/6 mice and digested enzymatically as previously described. 40,41 Cardiomyocytes were adhered to dishes and grown

| Fluorescence image analysis
Colocalization of mitochondria with LC3, lysosome or Parkin was used to visually assess mitophagy. 43 Adenoviruses with GFP-LC3 were incubated with primary cardiomyocytes for 24 hours to visualize LC3 punctae. Cells were dyed with MitoTracker Red CMXRos (MTR, 50 nmol/L) for 25 minutes at 37°C to visualize mitochondria and stained with LysoTracker Green DND-26 (LTG, 100 nmol/L) for 60 minutes to detect lysosomes. Following staining by MTR and fixation with 4% paraformaldehyde, anti-Parkin and FITC-conjugated secondary antibodies were used to visualize the Parkin distribution according to previously described protocols. 39 Images were visualized under an Olympus FV1000 laser confocal microscope and analysed using ImageJ software. Data were performed in three independent experiments.

| Adult cardiomyocyte isolation
Mice were anaesthetized with 4% chloral hydrate and fastened in the supine position. As previously reported, 44 the heart was removed and perfused with Ca 2+ free Tyrode's Solution (Sigma, L6402) by Langendorff-based methods to clear the blood. Next, continuous perfusion with Tyrode's Solution containing 0.1% collagenase II was applied to pretreat cardiac tissues that were minced into small chunks. After efficient mixing and digesting, cells were incubated in DMEM containing 10% foetal bovine serum and then filtered through a 200-μm strainer. Following centrifugation, isolated adult cardiomyocytes were reserved for analysis.

| Flow cytometry
Flow cytometric analysis of cellular mitochondria was performed.
Cardiomyocytes were incubated with Trypsin-EDTA for 5 minutes at 37°C and then suspended in DMEM containing 10% foetal bovine serum. Adherent cells were separated to monocells by soft pipetting.

| Assessment of cardiomyocyte apoptosis
A TUNEL assay kit was used to evaluate the apoptosis ratio of the cardiomyocytes as described before. 39 Data were representative of three independent experiments; random 20 fields with ×200 were assessed per group.

| Cardiac function assessment
Four months after STZ injection, echocardiography was employed to evaluate cardiac function. First, the mice were anaesthetized through inhaling 1.5%-2% isoflurane for examination. As previously described, 45 The left ventricular end-systolic diameter (LVESD) and left ventricular end-systolic diameter (LVESD) were quantified through a M-mode echocardiography system with a 15-MHz linear transducer (VisualSonics Vevo 2100, Toronto, ON, Canada). Then, LVEF and LVFS were calculated using computer algorithms. Those diameters were obtained from three consecutive cardiac cycles and averaged. Data were representative of three biological repeats.

| Transmission electron microscopy
Transmission electron microscopy (TEM) was used to observe the mitochondrial ultrastructure and typical autophagosomes engulfing a mitochondrion. 39,46 The detailed protocols were described in a previous report. 45 The observation of TEM was repeated three times, and 12 fields with ×9900 or cells were observed per group.

| Mitochondria and cytosol isolation
Mitochondria and cytosol were isolated using a commercial mitochondria isolation kit according to the manufacturer's instructions.
First, fresh, minced heart tissues or primary collected cardiomyocytes were homogenized in Mitochondria Isolation Solution containing PMSF on ice for 15 minutes. After grinding with a glass homogenizer, centrifugation at 800 g was performed for 10 minutes at 4°C.
The supernatant was collected in another tube and centrifuged again at 11 000 g for 10 minutes. The pellet was resuspended with Mitochondrial Lysate Solution to obtain mitochondrial proteins. The supernatant fraction was centrifuged at 12 000 g for 20 minute to obtain cytosol proteins.

| Mitochondria functional analysis
Citrate synthase (CS) activity, ATP contents and reactive oxygen species (ROS) production were measured to estimate the mitochondrial status. CS activity was measured by a commercial assay kit (Sigma) following the manufacturer's instructions. 41 An ATP bioluminescent assay kit (S0026; Beyotime, Shanghai, China) was used to detect the ATP level of heart tissues. 38 After homogenizing and centrifuging at 12 000 g for 15 minutes at 4°C, the supernatant was mixed with the working reagent and then assessed by a microplate luminescence luminometer. As before, the contents of ROS were assessed by electron paramagnetic resonance spectroscopy according to Mellin's methods. 41 Data were representative of three biological repeats.

| Western blot
The detailed protocol of the western blot assay has been described previously. 38,39 Cardiac tissues or cardiomyocytes were harvested and homogenized with RIPA buffer on ice. After centrifuging at 13 000 g for 15 minutes and quantifying with a Bradford protein assay, protein samples were separated by SDS-PAGE, transferred, incubated with antibodies, and scanned by a chemiluminescence system. Finally, the bands were quantified using the Image Pro Plus software (Media Cybernetics, Rockville, MD, USA). Data were representative of three independent experiments.

| Statistical analyses
Numerical data are expressed as the mean ± standard deviation (SD) of at least three independent experiments. Significant differences between treatments were compared using an unpaired Student's t test, one-way ANOVA followed by Fisher's post hoc comparison test or two-way ANOVA with multiple post hoc comparisons. Two-sided tests were used throughout the study, and a P value <0.05 was considered statistically significant. SPSS software package version 14.0 (SPSS, Chicago, IL, USA) was used for data analysis, and graphical representations were prepared by Prism software (GraphPad).

| Melatonin enhances impaired mitochondria elimination in diabetic hearts
In agreement with a previous report, 39   Melatonin consistently increased the number of typical autophagosomes containing damaged mitochondria in high-glucosetreated cardiomyocytes. These effects were obviously reversed by Parkin knockdown, as observed by electron microscopy ( Figure 4F).
Next, the mitophagy levels were quantified by the colocalization of GFP-LC3 and MitoTracker Red via confocal imaging. The colocalization of GFP-LC3 and MitoTracker Red also indicated that melatonin enhanced mitophagy to purge damaged mitochondria mostly through a Parkin-dependent pathway ( Figure 4G-I, Figure S1).

| Parkin deletion reverses the protective effects of melatonin on the DCM phenotype
Clearance of damaged mitochondria in cardiomyocytes is crucial for mitochondrial quality control and cardiac function. We also investigated whether melatonin enhanced Parkin-mediated mitophagy through Mst1 inhibition in mice with DCM. Mst1 knockout abolished the rise in expression of Parkin exerted by melatonin in diabetic mice hearts ( Figure 6E-G). Additionally, the increased mitochondrial Parkin and phosphorylated Parkin (Ser65) expression induced by melatonin was also abolished by Mst1 knockout (Figure 6H and I).
Taken together, these findings indicate that melatonin inhibits Mst1 phosphorylation, thus enhancing Parkin-mediated mitophagy to eliminate dysfunctional mitochondria which contributes to mitochondrial quality control. In the past, it was believed that mitochondria degradation through autophagy was of low specificity. In recent years, it has become clear that mitophagy is capable of selectively removing dysfunctional mitochondria to maintain mitochondrial quality and homoeostasis. 7,20 However, the involvement of mitophagy in the pathogenesis of diabetic cardiomyopathy is far from clear. 13 Here, we investigated the effects of melatonin on mitophagy in mice subjected to a diabetic insult. Typical autophagosomes engulfing mitochondria were found in non-diabetic mouse hearts by using transmission electron microscopy, 46

| Limitations
Neonatal cardiomyocytes rather than adult cardiomyocytes were used to investigate possible mechanisms. Further studies should be performed to explore how melatonin inhibits Mst1 expression and phosphorylation. The protective effects of melatonin on mice with type 2 diabetes should also be elucidated in further studies.