Introduction
Individuals with diabetes are at an increased risk of developing cardiovascular disease, a risk that is 2-3 times higher than that of the general population (1). Furthermore, 19-26% of diabetes patients develop heart failure (HF), as blood glucose levels are closely associated with the risk of HF (2). Diabetic cardiomyopathy (DCM) is the prevailing cardiovascular complication of diabetes and has emerged as a significant cause of death among diabetic patients (3). At present, DCM is defined as a cardiac disorder based on diabetes, without any other cardiac disease, and ultimately leads to HF (4).
The molecular mechanisms underlying DCM are multi-faceted and the incorporation of ferroptosis has propelled research to a new stage (5). The concept of ferroptosis was initially proposed by Dixon et al (6) and was characterized as an iron-dependent regulated cell death caused by lethal lipid peroxidation. The lipophilic radical-trapping antioxidant, ferrostatin-1 (Fer-1), is an inhibitor of ferroptosis (7) and glutathione peroxidase 4 (GPX4) is an endogenous lipid peroxide (LPO) scavenger that serves a pivotal role in the defense against ferroptosis (8). Our previous research revealed the involvement of ferroptosis in DCM, which was concomitant with the suppression of GPX4 and dysfunction of mitochondria (9). Due to the distinctive characteristics of myocardial cells, the abundance of mitochondria, mitochondrial dysfunction (10) and mitochondrial oxidative stress (11) are particularly important in the pathogenesis of DCM. The presence of shrunken mitochondria, increased mitochondrial membrane density and the disappearance of mitochondrial cristae in cells undergoing ferroptosis indicates the involvement of the mitochondria in ferroptosis and suggests the mitochondria as the core site (12). It is noteworthy that GPX4 exists as isoforms in the cytosol (cytoGPX4), mitochondria (mitoGPX4) and nucleus (nuclGPX4), exerting its anti-ferroptosis efficacy (13). Tadokoro et al (14) confirmed that mitoGPX4-mediated mitochondria-dependent ferroptosis has a pivotal role in the progression of doxorubicin-induced cardiomyopathy (DIC). However, the investigation of mitochondria-dependent ferroptosis in DCM remains limited.
SS-31, also known as MTP-131, elamipretide and Bendavia, is a tetrapeptide and a novel mitochondria-targeting antioxidant (15). Current research indicates that SS-31 exhibits exceptional biocompatibility and safety (16) and has demonstrated notable therapeutic potential for different types of cardiomyopathy, such as hypertensive cardiomyopathy (17), dilated cardiomyopathy (18) and DIC (19). Nevertheless, to the best of the authors' knowledge, the effect of SS-31 on DCM remains unexplored. In addition to the direct effect on mitochondria, SS-31 also mitigates oxidative stress and ferroptosis by regulating the related signaling pathways. The relationship between SS-31 and ferroptosis has also been validated in neurodegenerative diseases (20,21), and this evidence points SS-31 towards GPX4, which is a key regulatory factor in ferroptosis.
Therefore, the present study investigated the association between mitochondria-dependent ferroptosis and the pathogenesis of DCM and explored a potential therapeutic strategy for DCM by alleviating mitochondria-dependent ferroptosis using SS-31.
Materials and methods
Cells and treatment
H9C2 rat cardiomyocyte cells were obtained from the Chinese National Infrastructure of Cell Line Resource (cat. no. 1101RAT-PUMC000219). The H9C2 cells were cultured in DMEM containing 5.5 mmol/l glucose (cat. no. 10567022; Gibco; Thermo Fisher Scientific, Inc.), supplemented with 10% fetal bovine serum (cat. no. 10100147C; Gibco; Thermo Fisher Scientific, Inc.), 100 U/ml penicillin and 100 μg/ml streptomycin (cat. no. B540732; Sangon Biotech Co., Ltd.). The cell incubator (370; Thermo Fisher Scientific, Inc.) was set at 37°C with a 5% CO2 environment. An in vitro DCM model was established by treating H9C2 cells with 35 mmol/l glucose for 24 h, according to the protocol reported by previous studies (9,10).
As shown in Fig. 1, the H9C2 cells were divided into four groups: i) Control group, cultured in 5.5 mmol/l glucose for 24 h; ii) high glucose (HG) group, cultured in 35 mmol/l glucose for 24 h; iii) HG + Fer-1 group, cultured in 35 mmol/l glucose and 10 μM Fer-1 for 24 h; and iv) HG + SS-31 group, cultured in 35 mmol/l glucose and 10, 20 or 50 μM SS-31 for 24 h.
Cell viability assay
The cell viability was assessed by Cell Counting Kit-8 (CCK-8; cat no. E606335; Sangon Biotech Co., Ltd.). The H9C2 cells were seeded into 96-well plates at a density of 2×103 cells/well and allowed to adhere. Following a 6-h serum starvation period for cell cycle synchronization, the cells were treated as aforementioned. After 24 h of treatment, the 100 μl medium in each well was replaced with 10% CCK-8 reagent and the cells were incubated for another 2 h. The absorbance at 450 nm was then measured using a microplate reader (Multiskan GO; Thermo Fisher Scientific, Inc.). The cell viability was determined as follows: (Aexperiment-Ablank)/(Acontrol-Ablank) x100%.
Animal model
A total of 40 male C57BL/6J mice (aged 6-8 weeks, weighing 18-20 g) were purchased from Shanghai Slack Laboratory Animal Co., Ltd. [production license no. SCXK (Shanghai) 2017-0005]. The mice were housed in the Zhejiang Chinese Medical University Laboratory Animal Research Center (Hangzhou, China) under a specific pathogen-free environment (room temperature, 22±2°C; humidity, 55±5%; 12-h light/dark cycle), with unrestricted access to food and water. All experimental procedures were approved by the Animal Ethical and Welfare Committee of ZCMU (approval no. IACUC-20220307-10).
Prior to the experiment, the mice were acclimated for 1 week. As shown in Fig. 1, the diabetes model was established by overnight fasting followed by a single intraperitoneal injection of 150 mg/kg streptozotocin (cat. no. V900890; Sigma-Aldrich; Merck KGaA). Tail snip was used to obtain tail-tip blood, and blood-glucose was detected by a glucose meter (Accu-Chek® Performa test strips and an Accu-Chek® Performa Combo; both Roche Diabetes Care GmbH) 3, 5 and 7 days after injection. A random blood-glucose value of ≥16.7 mmol/l was classified as having diabetes, which was maintained for 4 weeks to induce DCM (9). Subsequently, the mice were divided into 4 groups (n=8): i) Control group, healthy mice were intraperitoneally injected with saline for 4 weeks; ii) DCM group, DCM model mice were intraperitoneally injected with saline for 4 weeks; iii) DCM + Fer-1 group, DCM model mice were intraperitoneally injected with 1 mg/kg/day Fer-1 (cat. no. HY-100579; MedChemExpress) for 4 weeks; and iv): DCM + SS-31 group, DCM model mice were intraperitoneally injected with 2.5 mg/kg/day SS-31 (cat. no. B27916; Shanghai Yuanye Biotechnology, Co., Ltd.) for 4 weeks.
The humane end points, which involve weight loss, loss of appetite, weakness (inability to eat or drink) and cardiovascular system, were established for the present study. However, none of the mice in the present study reached these endpoints.
Echocardiograms
The mice were anesthetized with 1% isoflurane in 100% oxygen and placed on a heating pad to keep warm (22). Then, transthoracic echocardiography was performed using an ultrasound system (VisualSonics Vevo 2100; FUJIFILM VisualSonics, Inc.) with a 30 MHz probe (MX400; FUJIFILM VisualSonics, Inc.). The left ventricle internal diameter in diastole (LVIDd), left ventricle internal diameter in systole (LVIDs), interventricular septum in diastole (IVSd), interventricular septum in systole (IVSs), left ventricular posterior wall in diastole (LVPWd) and left ventricular posterior wall in systole (LVPWs) were measured in M-mode. Subsequently, the left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), fractional shortening (FS) and ejection fraction (EF) were calculated using the Teichholtz formula as follows (22): LVEDV=[7/(2.4 + LVIDd)] x LVIDd3; LVESV=[7/(2.4 + LVIDs)] x LVIDs3; FS=[(LVIDd - LVIDs)/LVIDd] x100%; and EF=[(LVEDV - LVESV)/LVEDV] x100%.
Collection of blood and tissue samples
At the end of the experiment, the cardiac puncture was performed before sacrifice. In brief, after being anesthetized by sodium pentobarbital (50 mg/kg), a needle was inserted at the top of the sternum to collected 200 μl blood slowly. After collecting blood, a lethal dose of sodium pentobarbital (150 mg/kg) was immediately injected, followed by dissection and collection of cardiac tissue. The serum was then separated by centrifugation (1,500 x g, 10 min, 4°C) and stored at -80°C. A small piece of the left ventricle tissues was fixed in 4% paraformaldehyde (cat. no. E672002; Sangon Biotech Co., Ltd.) at 4°C overnight, or 2.5% glutaraldehyde (cat. no. A17876; Alfa Aesar; Thermo Fisher Scientific, Inc.) at 4°C overnight, for histopathological analysis and transmission electron microscopy (TEM). The remaining samples were rapidly frozen by liquid nitrogen and stored at -80°C.
Biochemical analysis
The serum samples were examined using an automated biochemical analyzer (3100; Hitachi High-Technologies Corporation) for the quantification of blood glucose, lactate dehydrogenase (LDH) and creatine kinase isoenzymes (CK-MB) levels.
Histopathology
The left ventricular tissues fixed in 4% paraformaldehyde (4°C overnight) were subjected to histopathological examination using H&E (hematoxylin for 10 min, eosin for 2 min, room temperature) and Sirius red (30 min, room temperature) staining to evaluate the pathological changes. Images were captured under a fluorescence upright microscope (Axioscope A1; Zeiss GmbH).
TEM
The left ventricular tissues fixed with 2.5% glutaraldehyde (4°C overnight) underwent post-fixation (1% OsO4+1.5%K3[Fe(CN)3] for 1 h at 4°C, TCH for 60 min, 1% OsO4 post-fixation for 1 h at 4°C, 2% uranyl acetate solution for overnight at 4°C), dehydration (50-100% gradient ethanol), resin penetration and embedding (acetone: EMBed 812=1:1 for 1 h, acetone: EMBed 812=1:3 for overnight, two changes of pure EMBed 812 for 8 h), polymerization (60°C for 48 h) and sectioning (60-80 nm, then onto the copper screen). The myofilament and mitochondria of the myocardial cells were imaged by a transmission electron microscope (H-7500; Hitachi High-Technologies Corporation).
Mitochondria isolation
The mitochondria were isolated using a Tissue Mitochondria Isolation kit (cat. no. C3606; Beyotime Institute of Biotechnology) and a Mitochondria Isolation kit for Cultured Cells (cat. no. 89874; Thermo Fisher Scientific, Inc.). Subsequently, RIPA lysis buffer (cat. no. C500005; Sangon Biotech Co., Ltd.) containing protease inhibitor (cat. no. C600380; Sangon Biotech Co., Ltd.) was used to lyse the mitochondria and extract the mitochondrial proteins.
Redox status determination
In total, 2×105 H9C2 cells/well were seeded into 6-well plates. The malondialdehyde (MDA), 4-Hydroxynonenal (4-HNE), ferrous iron and glutathione (GSH) content of the cell and tissue samples were determined using a Malondialdehyde Content Assay kit (cat. no. D799762; Sangon Biotech Co., Ltd.), 4-Hydroxynonenal ELISA kit (cat. no. D751041; Sangon Biotech Co., Ltd.), Ferrous Iron Colorimetric Assay kit (cat. no. E-BC-K773-M; Wuhan Elabscience Biotechnology Co., Ltd.) and GSH Content Assay kit (cat. no. D799614; Sangon Biotech Co., Ltd.), respectively. The protein concentrations were quantified using a BCA Protein Assay kit (cat. no. C503051; Sangon Biotech Co., Ltd.). The results were detected using a microplate reader and expressed in terms of protein concentration.
Total LPO assay
The total LPO assay was conducted using BODIPY 581/591 C11 (cat. no. D3861; Thermo Fisher Scientific, Inc.), which shifts fluorescence properties from red to green signals upon oxidation. Following the protocol described by Martinez et al (23), H9C2 cells were incubated with 2.5 μM BODIPY 581/591 C11 at 37°C for 30 min. After washing with PBS, fluorescence images were captured using a fluorescence inverted microscope (Axio Observer D1; Zeiss GmbH) in the FITC and Rhodamine channel. ImageJ software (version 1.53t; National Institutes of Health) was used for fluorescence intensity analysis.
Adenosine triphosphate (ATP) detection
The ATP content and protein concentration in the cell and tissue samples were measured using the ATP Assay kit (cat. no. S0026; Beyotime Institute of Biotechnology) and the BCA Protein Assay kit, respectively. Data were obtained using a Luminometer (GloMax 20/20; Promega Corporation) or a microplate reader, with the final results expressed in terms of protein concentration.
Mitochondrial membrane potential (MMP) assay
The MMP assay was conducted using JC-1 (cat. no. C2006; Beyotime Institute of Biotechnology). In a higher MMP, JC-1 aggregates in the matrix of the mitochondria, forming J-aggregates and emitting red fluorescence; conversely, once MMP is decreased, JC-1 remains as a monomer and does not aggregate in the matrix of the mitochondria and instead emits green fluorescence. Briefly, H9C2 cells were incubated with 10 μg/ml JC-1 at 37°C for 30 min. After washing with buffer, fluorescence images were captured using a fluorescence inverted microscope (Axio Observer D1; Zeiss GmbH) in the FITC and Rhodamine channel. ImageJ software (version 1.53t; National Institutes of Health) was used for fluorescence intensity analysis.
Mitochondrial LPO (mitoLPO) assay
The fluorescent dye, MitoPeDPP (cat. no. M466; Dojindo Laboratories, Inc.), localizes to the inner mitochondrial membrane and emits strong fluorescence upon oxidation. Briefly, H9C2 cells were incubated with 0.5 μM MitoPeDPP at 37°C for 15 min. After washing with PBS, fluorescence images were captured using a fluorescence inverted microscope (Axio Observer D1; Zeiss GmbH) in the FITC channel. ImageJ software (version 1.53t; National Institutes of Health) was used for fluorescence intensity analysis.
Mitochondrial ferrous iron assay
The fluorescent dye, Mito-FerroGreen (cat. no. M489; Dojindo Laboratories, Inc.), can detect ferrous ions specifically in the mitochondria, with the fluorescence intensity exhibiting a positive correlation with the concentration of ferrous ions in the mitochondria. Briefly, H9C2 cells were incubated with 5 μM Mito-FerroGreen at 37°C for 30 min. After washing with serum-free medium, fluorescence images were captured using a fluorescence inverted microscope in the FITC channel. ImageJ software (version 1.53t; National Institutes of Health) was used for fluorescence intensity analysis.
Reverse transcription-quantitative (RT-q) PCR
Total RNA from cells on 6-well plate or 50 mg tissue was extracted by RNAiso Plus reagent (cat. no. 9109; Takara Bio, Inc.) and quantified using a micro-spectrophotometer (Quickdrop; Molecular Devices, LLC). Subsequently, cDNA was synthesized using PrimeScript RT Master Mix (cat. no. RR036; Takara Bio, Inc.). The qPCR assay was next performed using TB Green® Premix Ex Taq II (cat. no. RR820; Takara Bio, Inc.) and a Real-Time PCR system (ABI 7500; Applied Biosystems; Thermo Fisher Scientific, Inc.) with the following thermocycling conditions: 95°C for 30 sec, 1 cycle; 95°C for 5 sec and 60°C for 34 sec, 40 cycles. All experiments were performed according to the manufacturers' instructions and included at least three biological replicates. β-actin served as the reference gene and the relative expression levels were calculated using the 2−ΔΔCq method (24). The qPCR primer sequences are listed in Table I. The primers for totalGPX4 were designed to target all isoforms of GPX4, while the primers for mitoGPX4 were specifically designed to recognize the unique sequence of mitoGPX4.
Western blotting
Western blotting analysis was performed using total and mitochondrial protein. The total proteins were extracted using RIPA lysis buffer containing protease inhibitor, and a tissue grinder (Scientz-48; Ningbo Scientz Biotechnology, Co., Ltd.) for the tissue samples. After the protein concentration was quantified by a BCA protein assay kit, 30 μg protein for each sample was separated using a 10% SDS-PAGE gel, then transferred to a PVDF membrane (cat. no. F619534; Sangon Biotech Co., Ltd.). After blocking with 5% Block Buffer (cat. no. A600669; Sangon Biotech Co., Ltd.) for 1 h at room temperature, the membrane was incubated with anti-GPX4 (1:1,000; cat. no. ab125066; Abcam), anti-β-actin (1:1,000; cat. no. ab8227; Abcam) and anti-voltage dependent anion channel 1 (1:1,000; cat. no. ab15895; Abcam) primary antibodies at 4°C overnight, then with a goat anti-rabbit (1:2,000; cat. no. ab6721; Abcam) secondary antibody (HRP) at room temperature for 1 h. Finally, the bands was visualized by ECL luminescence reagent (cat. no. C500044; Sangon Biotech Co., Ltd.) and a chemiluminescent imaging system (5200multi; Tanon Science and Technology Co., Ltd.). ImageJ software (version 1.53t; National Institutes of Health) was used for grayscale analysis.
Statistical analysis
The results were presented as the mean ± standard deviation. Statistical analysis was performed using SPSS software (version 19.0; IBM Corp.). Statistical comparisons were conducted by one-way ANOVA follow by the Bonferroni post hoc test. P<0.05 was considered to indicate a statistically significant difference.
Results
SS-31 restores mitochondrial dysfunction induced by high glucose in H9C2 cells
First, in vitro experiments were conducted to reveal the restorative effect of SS-31 on high glucose-induced mitochondrial dysfunction in H9C2 cells. Briefly, 35 mM glucose was employed to induce high glucose damage, then the cells were treated with 10, 20 or 50 μM SS-31. Meanwhile, a positive control group was treated with 10 μM Fer-1, a ferroptosis inhibitor. As shown in Fig. 2A, the cell viability assay results indicated that 35 mM glucose significantly suppressed H9C2 cell viability (P<0.01), while treatment with SS-31 or Fer-1 effectively restored this impaired cell viability (P<0.01). The MMP was assessed by employing JC-1 fluorescence staining (Fig. 2B and C). It was observed that high glucose resulted in a decrease in MMP in H9C2 cells (P<0.01), while both Fer-1 and SS-31 enhanced the MMP in high glucose-treated H9C2 cells (P<0.01) with a stronger restorative effect observed in the HG + SS-31 group. Furthermore, the results of the ATP content analysis demonstrated that the HG group exhibited a significant decrease in ATP (P<0.01), which was effectively restored by Fer-1 (P<0.01) or SS-31 (P<0.01), with SS-31 exhibiting a superior efficacy in restoring ATP content (Fig. 2D). These results suggested that SS-31 could effectively restore the mitochondrial dysfunction induced by high glucose in H9C2 cells.
SS-31 alleviates LPO damage induced by high glucose in H9C2 cells
In the present study, MDA and 4-HNE were detected as reliable markers of LPO (Fig. 3A and B). Compared with the control group, the HG group exhibited a significant elevation in MDA (P<0.01) and 4-HNE (P<0.01) levels. Furthermore, the treatments with Fer-1 and SS-31 both significantly suppressed the production of MDA and 4-HNE (P<0.01). The visualization of totalLPO and mitochondrial (mito)LPO in H9C2 cells was achieved using BODIPY 581/591 C11 and mitoPEDPP, respectively (Fig. 3C and D). The results demonstrated a significant increase in both totalLPO (P<0.01) and mitoLPO (P<0.01) in high glucose-treated H9C2 cells, with the mitoLPO exhibiting higher an accumulation compared with totalLPO. Furthermore, the treatments with Fer-1 and SS-31 showed effective reductions in the accumulation of totalLPO (P<0.01) and mitoLPO (P<0.01). Notably, SS-31 had a stronger capacity to counteract the increase in mitoLPO compared with Fer-1. These results suggested that both Fer-1 and SS-31 were effective in alleviating LPO accumulation, with SS-31 exhibiting an improved efficacy in alleviating mitoLPO.
SS-31 promotes mitoGPX4 to alleviate mitochondria-dependent ferroptosis
Accumulation of labile iron is a pivotal contributor to ferroptosis; therefore, Mito-FerroGreen was employed for the visualization and quantification of mitochondrial ferrous ions (Fig. 4A and B) and the Ferrous Iron Colorimetric Assay kit was employed for the quantification of the total ferrous ions (Fig. 4C) in H9C2 cells. The results showed a significant accumulation of total (P<0.01) and mitochondrial (P<0.01) ferrous ions in H9C2 cells treated with high glucose. Although both Fer-1 (P<0.01) and SS-31 (P<0.05) exhibited the capacity to alleviate the accumulation of total and mitochondrial ferrous ions, Fer-1 exerted a more notable effect.
GSH serves as the crucial substrate for GPX4 to effectively exert its anti-LPO function. Similar to GPX4, GSH is also distributed in the cytosol and mitochondria. As shown in Fig. 4D, totalGSH (P<0.01) and mitoGSH (P<0.01) were significantly decreased in high glucose-treated H9C2 cells, and Fer-1 failed to restore the suppression in totalGSH (P>0.05) and mitoGSH (P<0.05) levels. By contrast, SS-31 not only reinstated the totalGSH levels (P<0.05) but also restored the mitoGSH levels (P<0.01). Detection of GPX4 expression in high glucose-treated H9C2 cells demonstrated that totalGPX4 (P<0.01) and mitoGPX4 (P<0.01) mRNA expression was significantly reduced (Fig. 4E); however, only the protein expression of totalGPX4 was significantly inhibited (P<0.01; Fig. 4F and G). Additionally, treatment with Fer-1 did not affect the expression of totalGPX4 (P>0.05) and mitoGPX4 (P>0.05), whereas SS-31 significantly upregulated both the mRNA and protein expression of totalGPX4 (P<0.01) and mitoGPX4 (P<0.01), with a higher increase in mitoGPX4 than totalGPX4 protein expression (Fig. 4E-G). Therefore, it was hypothesized that, although Fer-1 and SS-31 exhibited efficacy in alleviating ferroptosis induced by high glucose in H9C2 cells, their underlying mechanisms were different. Specifically, SS-31 demonstrated the capacity to activate the GSH/GPX4 pathway, while more notably activating the mitoGSH/mitoGPX4 pathway in mitochondria.
SS-31 alleviates myocardial injury in DCM mice
To further investigate the therapeutic effects of SS-31 on DCM in vivo, a DCM mouse model was established and administered intraperitoneal injections of 2.5 mg/kg/day SS-31 or 1 mg/kg/day Fer-1 for 4 weeks. As shown in Table II and Fig. 5A, at the end of the experiment, the DCM mice exhibited significant diabetic symptoms with hyperglycemia (P<0.01) and a low body weight (P<0.01). However, treatment with Fer-1 or SS-31 did not ameliorate hyperglycemia in the DCM mice (P>0.05), but SS-31 did improve body weight (P<0.05). In terms of the cardiac parameters, there was a significant decrease in cardiac weight (P<0.01), FS (P<0.01) and EF (P<0.01) and a significant increase in LDH (P<0.01) and CK-MB (P<0.01) in the DCM mice, indicating cardiac damage. Conversely, treatment with Fer-1 or SS-31 effectively restored the cardiac weight (P<0.01), FS (DCM + Fer-1, P<0.05; DCM + SS-31, P<0.01) and EF (DCM + Fer-1, P<0.05; DCM + SS-31, P<0.01) and reduced LDH (P<0.01) and CK-MB (P<0.01). Notably, the DCM + SS31 group displayed lower CK-MB levels compared with the DCM + Fer-1 group, highlighting the improved cardioprotective effect of SS-31 over Fer-1 on DCM mice. The detailed echocardiographic results are available in Table SI.
The histopathological staining of cardiac tissues is shown in Fig. 5B and C. The H&E staining showed that the myocardial fibers were intact and aligned in the control group, but disorganized and even fractured in the DCM group. In the DCM + SS-31 group, the disorder of the myocardial fibers was improved. Sirius red staining revealed collagen fiber deposition in the myocardial tissues of the DCM group, whereas treatment with Fer-1 or SS-31 alleviated this pathology. These results suggested that both Fer-1 and SS-31 exhibited cardioprotective effects in DCM mice.
SS-31 alleviates mitochondria-dependent ferroptosis in DCM mice
The ultrastructure of the cardiac tissue from DCM mice was visualized by TEM (Fig. 6A). There were significant alterations in the ultrastructure of the cardiac tissue in DCM mice, including sarcomere disruption, mitochondrial swelling, disarray and outer membrane rupture, as well as disappearance or fragmentation of the mitochondrial cristae. Treatment with SS-31 effectively corrected the mitochondrial damage; it restored swelling, preserved the integrity of the inner and outer membranes and enhanced the abundance of cristae. Conversely, Fer-1 intervention exhibited a relatively modest effect on the ultrastructure of mitochondria. The results of the ATP content examination (Fig. 6B) demonstrated that the DCM cardiac tissue exhibited a significant decrease in ATP content (P<0.01). However, treatment with Fer-1 (P<0.05) or SS-31 (P<0.01) restored the ATP content, with SS-31 exhibiting a more significant efficacy. These results suggested that SS-31 exhibited an improved efficacy compared with Fer-1 in alleviating the mitochondrial dysfunction in the cardiac tissues of DCM mice.
In the cardiac tissues of DCM mice, there was a significant increase in MDA (P<0.01) and 4-HNE (P<0.01) levels and an accumulation in total ferrous ions (P<0.01), indicating severe LPO damage (Fig. 6C-E). However, treatment with Fer-1 or SS-31 effectively suppressed the generation of MDA and 4-HNE as well as the accumulation of total ferrous ions, with SS-31 demonstrating a stronger inhibition of 4-HNE production (P<0.05) and Fer-1 demonstrating a stronger alleviation of ferrous ion accumulation (P<0.05).
Similar to the in vitro experiments, the GSH/GPX4 and mitoGSH/mitoGPX4 pathways in the cardiac tissues of DCM mice were also analyzed. In the DCM group, both the totalGSH (P<0.01) and mitoGSH (P<0.01) levels were significantly reduced, which were only effectively restored by SS-31 treatment (P<0.01; Fig. 6F). Consistent with the aforementioned in vitro results, both the totalGPX4 and mitoGPX4 mRNA (Fig. 6G) and protein (Fig. 6H and I) expression levels were significantly suppressed in the cardiac tissues of DCM mice (P<0.01), and Fer-1 failed to reverse this inhibition (P>0.05); however, SS-31 did counteract this inhibition (P<0.01). Furthermore, it is noteworthy that the decrease in mitoGPX4 relative to totalGPX4 was comparatively more pronounced in the DCM mice during the in vivo experiments, indicating that mitochondria-dependent ferroptosis may play a more critical role in the pathogenesis of DCM. The primary mechanism of SS-31 in alleviating ferroptosis was through activation of the mitoGSH/mitoGPX4 pathway.
Discussion
SS-31, a mitochondria-targeting antioxidant peptide, exhibits favorable water solubility and cell permeability, with a receptor and transporter-independent cellular uptake mechanism (25). SS-31 selectively binds to cardiolipin in the inner mitochondrial membrane via both electrostatic and hydrophobic interactions. This selective binding demonstrates robust affinity towards the mitochondria while preventing cardiolipin oxidation, thereby stabilizing the cytochrome c and respiratory chain complex (26,27). Consequently, SS-31 can directly alleviate mitochondrial oxidative stress as well as protect mitochondrial function and ATP synthesis (15). Previously, research on mitochondria-dependent ferroptosis presented a novel avenue and trajectory for the prevention and treatment of cardiomyopathy (14). Mitochondria, as highly dynamic double-membrane organelles, function as the metabolic center for carbohydrates, lipids and proteins and have a pivotal role in energy metabolism, signal transduction and cell death regulation. Under physiological conditions, mitochondria account for 30-40% of the volume of cardiomyocytes, and almost all (>95%) of the ATP generated in the heart is derived from mitochondrial oxidative phosphorylation. Diabetes, as the underlying cause of DCM, leads to cellular metabolic dysfunction, such as iron metabolism and mitochondrial dysfunction. The subsequent excessive labile iron triggers LPO damage through the Fenton reaction, while impaired mitochondrial function disrupts redox homeostasis (28). Therefore, the heart is susceptible to mitochondria-dependent ferroptosis. Previously, we established the involvement of ferroptosis in the pathogenesis of DCM (9). In the present study, it was hypothesized that mitochondria-dependent ferroptosis was a crucial determinant in the pathogenesis of DCM and proposed the administration of SS-31 as a targeted therapeutic approach for DCM.
In the present study, DCM mouse and high glucose-treated H9C2 cell models were established to investigate the potential mechanisms of SS-31 treatment. Meanwhile, Fer-1 (a ferroptosis inhibitor) was employed as a positive control, which alleviates ferroptosis by eliminating the labile iron pool (8). The serum cardiac injury biomarkers, echocardiograms and histopathological results confirmed the establishment of the in vivo model, and assessment of the H9C2 cell viability indicated the establishment of the in vitro model, consistent with our previous report (9). In the in vivo experiments, both Fer-1 and SS-31 did not improve hyperglycemia in the DCM mice, while SS-31 exhibited stronger cardioprotective effects than Fer-1, indicating that the cardioprotective effect of SS-31 was not related to glycemic control. The mitochondrial dysfunction of the DCM group was verified by the decrease in ATP content, the attenuation of the MMP and the disruption in the mitochondrial ultrastructure. Treatment with SS-31 exhibited a stronger restorative effect on ATP and the MMP compared with Fer-1, while effectively restoring the mitochondrial ultrastructure. Additionally, the DCM group exhibited severe LPO damage, with more significant mitochondrial LPO damage than total LPO damage and treatment with SS-31 effectively alleviated the mitochondrial LPO damage. Considering the contribution of the labile iron pool to ferroptosis, intracellular and mitochondrial iron ions were also assessed in the in vitro experiment. In the model group, the ferrous ions accumulated both intracellularly and in the mitochondria; however, SS-31 treatment mitigated the accumulation of ferrous ions, albeit with a weaker efficacy compared with Fer-1. These findings suggested that the involvement of mitochondria-dependent ferroptosis was critical in DCM pathogenesis and provided evidence for SS-31 alleviating mitochondria-dependent ferroptosis.
In addition to the direct effects on the mitochondria, recent research has reported that SS-31 can also activate the GSH/GPX pathway to alleviate ferroptosis; however, the research is still restricted and limited to neurological injury disorders. First, Zhang et al (20) demonstrated that SS-31 effectively alleviates hippocampal ferroptosis and ameliorates cognitive dysfunction in sevoflurane-induced neonatal mice. Second, Liu et al (21) discovered that SS-31 activates the GSH/GPX4 pathway to suppress hippocampal ferroptosis in epileptic rats. It is worth noting that before the concept of ferroptosis was proposed, Dai et al (17) demonstrated that SS-31 improves hypertensive cardiomyopathy by inhibiting NADPH oxidase 4 (NOX4). Subsequently, NOX4 was shown to be a facilitator of ferroptosis (29). The GSH/GPX4 pathway has a central role in limiting LPO and ferroptosis; intracellular GSH is synthesized from cystine and serves as the substrate for GPX4 (30). GPX4 eliminates the cellular toxicity of LPO by oxidizing GSH to glutathione disulfide (8). GPX4 is a highly conserved gene that encodes three distinct GPX4 proteins, which are respectively localized within the mitochondria, cytosol and nucleus. Previous research has demonstrated the presence of the GSH/GPX4 pathway not only in the cytosol but also within mitochondria (mitoGSH/mitoGPX4), where it inhibits mitochondria-dependent ferroptosis (14). In the present study, total and mitoGSH were quantified and the expression levels of total and mitoGPX4 were assessed. In the in vivo DCM group, the depletion of mitoGSH was more significant than the totalGSH depletion. Although no significant decrease in mitoGPX4 protein expression was observed in the HG group during the in vitro experiment, a reduction in mRNA was detected. It was hypothesized that this phenomenon may be attributed to the metabolic rate of mRNA being faster than proteins; therefore, the GPX4 protein expressed before the 24-h high glucose exposure was possibly incompletely degraded, resulting in a discordance between mRNA and protein expression. In addition, the in vivo experimental results demonstrated a synchronicity between mRNA and protein expression. These findings of the present study further supported the hypothesis that the inactivation of mitoGSH/mitoGPX4, leading to mitochondria-dependent ferroptosis, has a pivotal role in the pathogenesis of DCM. Furthermore, only SS-31 exhibited the capacity to activate the mitoGSH/mitoGPX4 pathway.
The complete GPX4 genomic DNA contains 8 exons, with exons 3-8 present in all three isoforms of GPX4 protein. Alternative splicing does not participate in the production of these three GPX4 isoforms, which are instead determined by different transcription start sites. Exon 1 of GPX4 genomic DNA contains two translational start codons, the first for mitoGPX4 and the second for cytoGPX4; exon 2 contains the third translational start codon for nuclGPX4. Each translational start codon has its own distinct transcription start site (31). This may explain why the expression of mitoGPX4 was unsynchronized with totalGPX4.
The present study still has certain limitations. Since the difference between mitoGPX4 and cytoGPX4 only exists in the first exon, which governs mitochondrial localization, precise and effective RNA interference targeting mitoGPX4 becomes unattainable. There is also currently an absence of a precise inhibitor for mitoGPX4; therefore, further comprehensive investigations were impeded in the present study. The employment of Fer-1 in the present study had no effect on the mitoGSH/mitoGPX4 pathway, despite alleviating LPO damage. This observation implied that activation of the mitoGSH/mitoGPX4 pathway by SS-31 was not due to the alleviation of LPO damage.
In conclusion, the present study demonstrated that mitochondria-dependent ferroptosis serves as a pathogenic mechanism underlying DCM and can be alleviated by SS-31. The specific mechanism involved activation of the mitoGSH/mitoGPX4 pathway and mitigating the accumulation of ferrous ions in mitochondria. Considering the safety profile and multi-organ protective effects of SS-31, it is a promising drug for the prevention and treatment of DCM.