METTL3‐m6A‐SLC25A11 Axis Promotes Chronic Hypoxia–Induced Cardiomyocyte Ferroptosis

Abstract

Background

Chronic hypoxia is commonly associated with various cardiovascular diseases, with cardiomyocyte death being the most frequently observed alteration. Mitochondrial dysfunction is another consequence seen in the hypoxic heart. However, the mechanistic linkage between mitochondrial dysfunction and cardiomyocyte death in the hypoxic heart remains unclear. Solute carrier family 25 member 11 (SLC25A11) is essential for mitochondrial function via transporting glutathione to mitochondria and is possibly involved in ferroptosis. However, the role of SLC25A11 in chronic hypoxia–induced cardiomyocyte ferroptosis remains unknown.

Methods

SLC25A11 overexpression and short hairpin RNA vector were constructed and introduced into the AC16 cardiomyocytes and human induced pluripotent stem cell–derived cardiomyocytes exposed to chronic hypoxia, and the cell viability, mitochondrial function, and ferroptosis were evaluated. Myocardial injury and cardiac function were also assessed in hypoxic mouse models.

Results

Our findings reveal that chronic hypoxia induced ferroptosis and mitochondrial dysfunction and decreased cell viability and SLC25A11 expression. SLC25A11 overexpression inhibited chronic hypoxia–induced ferroptosis and mitochondrial dysfunction. SLC25A11 silencing‐induced ferroptosis is reversed by iron chelator deferoxamine. Chronic hypoxia–induced increased the N6‐methyladenosine (m6A) level of SLC25A11 3′UTR and decreased expression of SLC25A11 were reversed by methyltransferase‐like 3 inhibitor STM2457. m6A‐binding protein YTH domain family 2 binds to the SLC25A11 3′UTR. STM2457 reversed chronic hypoxia–induced cardiomyocyte ferroptosis and mitochondrial dysfunction, while SLC25A11 knockdown abolished the effects of STM2457. The ameliorative effect of STM2457 in cardiomyocyte injury was proved in the hypoxia mouse model.

Conclusions

This study is the first to demonstrate the protective effect of methyltransferase‐like 3 inhibition, via SLC25A11 m6A modification, against chronic hypoxia–induced cardiomyocyte ferroptosis and reveals the possibility that inhibiting activation of methyltransferase‐like 3–m6A–SLC25A11 may provide cardioprotective therapy for chronic hypoxia–induced cardiovascular diseases.

Nonstandard Abbreviations and Acronyms

GPX4

glutathione peroxidase 4

GSSG

glutathione/oxidized glutathione

hiPSC‐CMs

human induced pluripotent stem cell–derived cardiomyocyte

m6A

N6‐methyladenosine

SLC25A11

solute carrier family 25 member 11

METTL3

methyltransferase‐like 3

OCR

oxygen consumption rate

PTGS2

prostaglandin‐endoperoxide synthase 2

RIP

RNA immunoprecipitation

ROS

reactive oxygen species

YTHDF2

YTH domain family 2

Clinical Perspective.

What Is New?

• Our observation reveals that methyltransferase‐like 3–mediated N6‐methyladenosine modification of solute carrier family 25 member 11) regulates ferroptosis in chronic hypoxia–induced cardiomyocytes.

What Are the Clinical Implications?

• By elucidating the potential mechanisms underlying methyltransferase‐like 3–N6‐methyladenosine– solute carrier family 25 member 11 ferroptosis, our results offer opportunities for future precision research, to mitigate the burden of chronic hypoxia–induced cardiovascular diseases.

The heart consumes a large amount of ATP and pumps oxygen‐rich blood to all other embryonic organs. To achieve this, the heart needs a large supply of oxygen to support its pumping function. However, lung diseases, high‐altitude environments, and anemia can lead to a chronic lack of oxygen in cardiomyocytes. ¹ Chronic hypoxia is closely related to the occurrence of various cardiovascular diseases and leads to cardiac fibrosis, heart failure, and even cardiomyocyte death. ² , ³ An in‐depth understanding of the pathological changes in cardiomyocyte death following chronic hypoxia will provide clues for better drug development and therapeutic strategies in the future.

Ferroptosis is an iron‐dependent necroptosis characterized by glutathione depletion and lipid peroxidation, which occurs as a result of hyperoxygenation of polyunsaturated fatty acid residues of phospholipids by nonheme iron‐containing lipoxygenase and insufficient ability of glutathione peroxidase 4 (GPX4) to remove oxidized phospholipids. ⁴ Ferroptosis is considered to be an essential event for the occurrence and development of diseases, so emerging strategies against ferroptosis have been developed for the treatment of organ and tissue damage. ⁵ , ⁶ Studies have found that ferroptosis is an important mechanism of hypoxic myocardial cell loss. ⁷ , ⁸ However, the potential involvement of ferroptosis in chronic hypoxia–induced cardiovascular diseases remains unclear.

The solute carrier family 25 member 11 (SLC25A11), also known as the 2‐oxoglutarate carrier, is an anion carrier that regulates the electrically neutral exchange between 2‐ketoglutarate and some dicarboxylic acids and serves as a key factor for glutathione transport from cytoplasm to mitochondria. ⁹ Due to the importance of the glutathione–GPX4 antioxidant system for mitochondrial function and ferroptosis, the regulation of the mitochondrial glutathione pool is essential for ferroptosis. ¹⁰ One study found that SLC25A11 was involved in regulating mitochondrial glutathione levels and cell ferroptosis. ⁹ However, the role of SLC25A11 in ferroptosis in chronic hypoxia‐induced cardiomyocytes remains unclear.

N6‐methyladenosine (m6A) methylation is the most common reversible internal posttranscriptional modification of mammalian mRNA, which is dynamically regulated by methyltransferases (writer), demethylases (erasers), and m6A‐binding proteins (reader). ¹¹ The important role of m6A modification in the regulation of gene expression has attracted attention and has been recognized in the process of growth, development, and disease progression. ¹² However, little understanding has been presented of the role of m6A modification in hypoxic cardiomyocyte injury. As a study reported, m6A glycosylation levels were significantly increased, and inhibition of the methyltransferase‐like 3 (METTL3) completely abolished the ability of cardiomyocytes to undergo hypertrophy in response to growth stimulation. ¹³ In vitro cell culture studies showed that METTL3 was shown to affect the survival of hypoxic cardiomyocytes by regulating the m6A modification of the gene. ¹⁴ , ¹⁵ , ¹⁶ These studies have shown that METTL3‐mediated m6A modification is essential for the survival of hypoxic cardiomyocytes; however, the target genes of METTL3‐m6A modification have not been elucidated.

Primary cardiomyocytes and immortalized cardiac cell lines are widely used alternatives for studying cellular mechanisms and mimicking conditions in vivo. Isolated primary cardiomyocytes are the best option, being disadvantaged mainly by the time‐dependent isolation process, short culture life span, need for a large number of animals, and limited genetic manipulation possibilities. ¹⁷ In contrast, secondary cell lines are easier to handle, but they possess lower similarity to real cardiac myocytes. An increasing number of studies have been using the immortalized human cardiomyocyte cell line AC16 generated from adult human ventricular cardiomyocytes, ¹⁸ , ¹⁹ originally established by Davidson et al in 2005. ²⁰ In addition, human induced pluripotent stem cell–derived cardiomyocytes (hiPSC‐CMs) are a powerful in vitro model for studying cardiac biology, disease mechanisms, and drug responses. ²¹ They combine the genetic relevance of human cells with the scalability of stem cell technology, making them increasingly popular in research and drug discovery. Here, we therefore focused on the role of METTL3‐m6A modification in regulating SLC25A11 levels in chronic hypoxia–induced AC16 cardiomyocytes and hiPSC‐CMs. Our results suggest that chronic hypoxia promotes ferroptosis of cardiomyocytes by inhibiting SLC25A11 expression by METTL3‐m6A.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Cell Culture and Treatment

Human AC16 cardiomyocytes were cultured and passaged in DMEM supplemented with fetal bovine serum (10%) and streptomycin/penicillin (1%). Human induced pluripotent stem cell–derived cardiomyocytes were obtained from Nuwacell Biotechnologies Co., Ltd. (Anhui, China) and maintained in Matrigel‐coated 6‐well plates using mTESR1 medium. hiPSC‐CMs were achieved using a differentiation protocol as previously described. ²¹ Normoxic culture conditions were 5% CO₂ in a humidified atmosphere at 37 °C. To simulate hypoxic injury, AC16 cells and hiPSC‐CMs were grown in an incubator containing 94% N₂, 1% O₂, and 5% CO₂. To explore the involvement of ferroptosis in chronic hypoxia–induced injury, AC16 cells were treated with 10 μM of ferroptosis inducer erastin (Selleck Chemicals, Houston, TX), 100 μM of ferroptosis inhibitor iron chelator deferoxamine (Selleck Chemicals), or vehicle in the absence or presence of hypoxia stimuli for 48 hours. STM2457 is a highly potent, specific, and bioavailable inhibitor of METTL3, suitable for in vivo investigations. ²² To explore the involvement of METTL3‐mediated m6A modification, METTL3 inhibitor STM2457 (5 μM; Selleck Chemicals) ²³ was used to treat AC16 cells and hiPSC‐CMs in normoxia or hypoxia stimuli for 48 hours.

Plasmids and Cell Transfection

RNA interference mediated by short hairpin RNAs (shRNAs) against SLC25A11 (shSLC25A11–1/2/3, respectively) and scramble shRNA as the negative control were provided by RiboBio (Guangzhou, China) and incorporated into pLKO.1 plasmid. Cotransfection of recombinant pLKO.1‐shSLC25A11 plasmids, psPAX2, and pMD2G packaging vectors was conducted in 293T cells through Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Two days after transfection, the secreted virus particles were harvested, concentrated through ultracentrifugation, and transduced into AC16 cells.

RNA interference mediated by small interfering RNAs (siRNAs) against YTH domain family 2 (YTHDF2) (siYTHDF2‐1/2, respectively) or scramble siRNA as the negative control was constructed by RiboBio (Guangzhou, China). The SLC25A11 overexpression vector was constructed via the pcDNA3.1 vector recombined SLC25A11 gene with the blank pcDNA3.1 vector as the negative control. AC16 cells and hiPSC‐CMs were cultured in a 6‐well plate until 70% cell confluence was reached. The siRNA and recombinant overexpression plasmids transfecting AC16 cells and hiPSC‐CMs were performed by Lipofectamine 2000. Cells were collected after 48 hours of transfection for further analysis.

Cell Counting Kit‐8 Assay

AC16 cells that underwent 24 hours of culture in a 96‐well plate and 24 hours of treatment in normoxic or hypoxic conditions were collected for cell viability analysis via Cell Counting Kit‐8 assay (Dojindo, Japan). Cells in each well were incubated with Cell Counting Kit‐8 solution (10 μL) for 4 hours. Samples in the 96‐well plate were placed in a ThermoMax microplate reader (Olympus, Center Valley, PA) to monitor the absorbance of each well at 450 nm.

Lipid and Mitochondrial Reactive Oxygen Species

Lipid reactive oxygen species (ROS) levels in AC16 cells and hiPSC‐CMs were measured by the fluorescence method using the BODIPY 581/591 C11 kit (Thermo Fisher Scientific, Waltham, MA; D3861). In brief, cells were washed with PBS and were collected by 800g of centrifugation. Cell precipitates were stained with C11‐BODIPY (581/591) (10 μM) for 30 minutes at 37°C under dark conditions. Lipid ROS levels were evaluated by measuring the fluorescence intensity at 581/591 nm of excitation/emission on a CytoFLEX flow cytometer (Beckman Coulter, Brea, CA). To detect mitochondrial ROS production, the AC16 cells were stained with 5 nM MitoSOX for 15 minutes, and the fluorescence signals were analyzed using CytoFLEX flow cytometry.

Extracellular Flux Analysis

Mitochondrial energy metabolism was analyzed by real‐time seahorse analysis of oxygen consumption rates (OCRs). In brief, cells at a density of 1×10 ⁴ /well were planted in XF‐24 culture plates and incubated for 24 hours at 37 °C and 5% CO₂. Before the experiment, cells were transferred to a CO₂‐free incubator, and the culture medium was replaced with XF Base Medium. The Seahorse XF Cell Mito Stress Test Kit (103016–100; Agilent Technologies, Santa Clara, CA) was used to add specific inhibitors.

Detection of Ferroptosis‐Related Markers

The glutathione/oxidized glutathione (GSSG) ratio was determined via a glutathione/GSSG kit (E‐BC‐K097‐M; Elabscience, Houston, TX). Malondialdehyde levels in the cell lysates were assessed with a malondialdehyde detection kit (A003‐1; Nanjing Jiancheng Bio, China). Intracellular Fe²⁺ was visualized via an Iron Assay Kit (Abcam, Cambridge, MA; ab83366) and measured by a colorimetric microplate reader or via the fluorescence probe FerroOrange (Dojindo, Mashiki, Japan; F374) and a confocal microscope. Experimental procedures were performed on the basis of the protocols provided by the manufacturer.

Quantitative Real‐Time Polymerase Chain Reaction

The total RNA sample was prepared from AC16 cells homogenate or myocardial tissue homogenate using the TRIzol kit (Thermo Fisher Scientific). The RNA sample was the template for the synthesis of the cDNA with the Hifair II First Strand cDNA Synthesis SuperMix. The cDNA products were used as the templates for real‐time polymerase chain reaction (PCR) using the Hieff qPCR SYBR Green Master Mix. PCR amplification procedures were conducted on an ABI 7300 real‐time PCR system (Applied Biosystems, Carlsbad, CA). Reagents used for quantitative PCR were purchased from Yeasen Biotechnology Co., Ltd. (Shanghai, China). The primers were as follows: SLC25A11‐F: 5′‐GTCCGTCAAGTTCCTGTTTGG‐3′; SLC25A11‐R: 5′‐AGCCGACAGCCCAGTGTA‐3′; YTHDF2‐F: 5′‐CAGGCAAGGCCCAATAATGC‐3′; YTHDF2‐R: 5′‐AAGTAGGGCATGGCTGTGTC‐3′; ACTB‐F: 5′‐GTCACCAACTGGGACGACAT‐3′; ACTB‐R: 5′‐TAGCAACGTACATGGCTGGG‐3′.

Western Blot

Total protein was extracted from cardiomyocytes or myocardial tissues with RIPA lysis buffer (Beyotime, Jiangsu, China) in the presence of protease and phosphatase inhibitor and phenylmethylsulfonyl fluoride (1%; Beyotime) at 4°C. The lysate was centrifuged at 12000g for 20 minutes, and the supernatant was collected, which was the total protein. An equal volume of protein (25 μg) was used for protein expression analysis. Protein samples were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and then transferred to a polyvinylidene fluoride membrane. Proteins in the polyvinylidene fluoride membrane were sequentially subjected to blocking, primary antibody incubation, and secondary antibodies. The primary antibodies used are as follows: SLC25A11 (Proteintech, Rosemont, IL; 12 253‐1‐AP), prostaglandin‐endoperoxide synthase 2 (PTGS2; Abcam; ab179800), GPX4 (Abcam; ab125066), YTHDF2 (Abcam; ab220163), and β‐actin (Proteintech; 66 009‐1‐Ig). Protein bands were visualized using an enhanced chemiluminescence kit, and their gray values were quantified by ImageJ software (National Institutes of Health, Bethesda, MD). The level of β‐actin was recognized as an internal control for quantifying target proteins.

m6A Methylated RNA Immunoprecipitation Quantitative PCR

The experimental m6A methylated RNA immunoprecipitation quantitative PCR procedures were conducted as in a previous study. ²⁴ Briefly, total RNA was extracted from AC16 cells and quantified by spectrophotometry. A total of 100 μg RNA was ultrasonically fragmented into fragments in immunoprecipitation buffer in the presence of RNase inhibitor. For immunoprecipitation, RNA fragments were absorbed by the protein A/G magnetic beads conjugated antibody (anti‐m6A antibody or anti–immunoglobulin G) at 4 °C. RNA fragments adsorbed by beads were separated with N6‐methyladenosine 5‐monophosphate sodium salt (Sigma‐Aldrich, St. Louis, MO) and used as samples for detecting m6A‐modified SLC25A11 by the quantitative real‐time PCR method.

Luciferase Reporter Gene Assay

SLC25A11‐3′UTR fragments containing m6A motifs were cloned into the downstream of firefly luciferase in the pmirGLO plasmid (Promega, Fitchburg, WI). AC16 cells were cultured in 24‐well plates and cultured in conditions of 5% CO₂ and 37 °C until 60% cell confluence. Cells were then transfected with either the pmirGLO‐SLC25A11 3′UTR luciferase reporter plasmid or the internal reference plasmid pRL‐TK vector and received hypoxia stimulation in the presence/absence of STM2457 treatment. Finally, a dual‐luciferase assay was conducted.

RNA Stability

To evaluate RNA stability of SLC25A11 in AC16 cells under conditions of chronic hypoxia and STM2457, SLC25A11 mRNA was determined using quantitative PCR in cells treated with actinomycin D (5 μg/mL; Sigma) for the indicated time (0, 3, and 6 hours).

RNA Immunoprecipitation Assay

YTHDF2 binding with SLC25A11 mRNA was examined by RNA immunoprecipitation (RIP) assay using EZ‐Magna RIP RNA‐Binding Protein Immunoprecipitation Kit (Millipore, Burlington, MA). Briefly, the total RNA sample was extracted from AC16 cells via RIP lysis buffer in the presence of RNase inhibitor (Millipore) and protease inhibitor. The RNA sample was mixed with beads conjugated with antibody (anti‐YTHDF2; Abcam; ab220163) in RIP buffer. RNA that was absorbed by bead conjugated with the anti–immunoglobulin G antibody (Cell Signaling Technology, Danvers, MA) was the negative control. RNA in YTHDF2 immunoprecipitation was purified to determine SLC25A11 expression using PCR.

Animal Model

The male C57BL6/J mice (aged 8 weeks) were obtained from Vital River Laboratory Animal Technology Co., Ltd. (Zhejiang, China). For animal model construction and treatment, mice were randomly divided into 4 groups (n=9 per group): a normoxia+vehicle group, a hypoxia+vehicle group, a hypoxia+30 mg/kg STM2457 and a hypoxia+60 mg/kg STM2457 group. The hypoxic injury was induced via mice living in a ventilated hypoxic cabin for 21 days with a system of a mixture of 10% oxygen and 90% nitrogen (Yuyan Science, China). STM2457 treatment was achieved by mice receiving 7 days of intravenous injections of STM2457 (30 or 60 mg/kg of mouse weight once daily) ²³ and another 21 days of the same dose injection under hypoxic conditions. At 21 days after hypoxia injury, myocardial function was measured by echocardiography, and the mice were then euthanized for myocardial tissue harvesting for subsequent histological and molecular analysis. Isolation and culture of murine cardiomyocytes were performed as previously described. ²⁵ Briefly, the mice in normoxia+vehicle, hypoxia+vehicle, and hypoxia+60 mg/kg STM2457 groups were euthanized, the heart was surgically removed and connected to a Langendorff perfusion system, followed by perfusion with a collagenase solution for up to 50 minutes. The heart was transferred into Petri dishes, the atria were removed, and undigested parts of the heart muscle were eliminated using a filter. For functional analysis, Ca²⁺ was stepwise increased to a final concentration of 1 mM. Isolated cardiomyocytes were cultured in laminin‐coated dishes using minimal essential medium (Gibco, Waltham, MA) supplemented with 2.5% FBS and 50 U/mL penicillin and 50 μg/mL streptomycin. All animal experiments were approved by the Animal Care and Use Committee of Luodian Hospital, Baoshan District, and done in compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Hematoxylin and Eosin Staining

Hematoxylin and eosin staining was performed to evaluate the myocardial histomorphological changes. Myocardial tissues were washed with normal saline and prepared into 5‐μm tissue sections. Sections received hematoxylin and eosin staining sequentially. Histomorphological features were observed by a Leica DM3000 microscope (Leica, Wetzlar, Germany) and photographed.

Echocardiography

The animal was anesthetized using 2% isoflurane via inhalation and fixed in the supine position. Echocardiography was conducted by 2 observers in a blinded manner to record images along the parasternal short axis. Echocardiographic images were stored for offline analysis by Vevo 2100 software (VisualSonics, Bothell, WA). Cardiac function was assessed using 4 indicators, including left ventricular (LV) end‐systolic diameter, LV end‐diastolic diameter, LV ejection fraction, and LV fractional shortening by the echocardiography software.

Statistical Analysis

Data processing and statistics were processed by Prism version 8.4.2 (GraphPad Software, La Jolla, CA). Significance among groups was assessed by unpaired 2‐tailed Student’s t test or 1‐way ANOVA followed by Dunnett’s post hoc test. Data are represented as means±SDs. Statistical significance was defined as P<0.05.

Results

Chronic Hypoxia Induces Ferroptosis and Inhibits SLC25A11 Expression in Cardiomyocytes

Chronic hypoxia induced a decrease in AC16 cell viability, and this inhibition was enhanced when ferroptosis was induced by erastin and weakened when ferroptosis was inhibited by deferoxamine (Figure 1A), suggesting increased ferroptosis mediated the chronic hypoxia–induced cardiomyocyte injury. Ferroptosis is an iron‐dependent nonapoptotic form of cell death characterized by the iron‐induced accumulation of lipid ROS. ²⁶ Therefore, the lipid ROS and intracellular Fe²⁺ levels in AC16 cells were also assessed via flow cytometry and biochemical analysis, respectively. Contrary to the trend change of cell activity, Fe²⁺ content and lipid ROS level were increased in the chronic hypoxic condition, and these increases were enhanced by erastin and weakened by deferoxamine (Figure 1B and 1C), which further supports that chronic hypoxia induces ferroptosis in cardiomyocytes. However, deferoxamine decreased Fe²⁺ content but had no effect on the AC16 cell viability under normoxic conditions (Figure S1). Besides, we found that the SLC25A11 expression level decreased with the time course of chronic hypoxia in AC16 cells (Figure 1D and 1E).

Figure 1. Chronic hypoxia induces ferroptosis and inhibits SLC25A11 expression in AC16 cells.

A, Cell viability; (B) Fe²⁺; (C) lipid ROS; and (D, E) SLC25A11 expression in AC16 cells treated with 10 μM of erastin (an inducer for ferroptosis), 100 μM of deferoxamine (an inhibitor for ferroptosis), or vehicle in the presence of hypoxia stimuli for 48 hours (n=3). One‐way ANOVA followed by Dunnett’s post hoc test was used. SLC25A11 indicates solute carrier family 25 member 11.

SLC25A11 Overexpression Inhibits Chronic Hypoxia–Induced Ferroptosis in Cardiomyocytes

Given the significant downregulation of SLC25A11 expression in chronic hypoxia–induced AC16 cells, we overexpressed SLC25A11 in AC16 cells (Figure 2A and 2B) to observe its effect on ferroptosis. Cell viability analysis results showed that SLC25A11 overexpression reversed the chronic hypoxia–induced decrease in cell viability (Figure 2C). Moreover, the mitochondrial glutathione/GSSG ratio, ROS, and OCR levels were also measured to examine the role of SLC25A11, as a mitochondrial glutathione transporter, in chronic hypoxia–induced AC16 cells. As shown in Figure 2D through 2F, chronic hypoxia decreased glutathione/GSSG ratio and OCR level and increased mitochondrial ROS level, which were reversed by SLC25A11 overexpression. The results of ferroptosis analysis showed that SLC25A11 overexpression reversed chronic hypoxia–induced changes in ferroptosis‐related markers, including the inhibition of malondialdehyde, Fe²⁺, and lipid ROS levels (Figure 2G through 2I). However, SLC25A11 overexpression increased glutathione/GSSG ratio and decreased malondialdehyde and Fe²⁺ contents but had no effect on the AC16 cells’ viability under normoxic conditions (Figure S2A–D). In addition, SLC25A11 overexpression reversed chronic hypoxia–induced changes in GPX4 and PTGS2 expression, including promotion of GPX4 expression and inhibition of PTGS2 expression, in chronic hypoxia–induced AC16 cells (Figure 2J). Similarly, SLC25A11 overexpression also reversed chronic hypoxia‐induced changes in cell viability, glutathione/GSSG ratio, OCR, and ferroptosis‐related markers in hiPSC‐CMs (Figure S3A–F).

Figure 2. SLC25A11 overexpression inhibits chronic hypoxia–induced ferroptosis in AC16 cells.

SLC25A11 expression in AC16 cells transfected with the SLC25A11 expression vector was examined by (A) qRT‐PCR and (B) western blot. C, Cell viability; (D) glutathione/GSSG ratio; (E) oxygen consumption rate; (F) mitochondrial ROS; (G) malondialdehyde; (H) Fe²⁺; (I) lipid ROS; and (J) expressions of SLC25A11, GPX4, and PTGS2 in AC16 cells transfected with SLC25A11 expression vector or blank vector in the presence of hypoxia stimuli for 48 hours (n=3). One‐way ANOVA followed by Dunnett’s post hoc test was used. GPX4 indicates glutathione peroxidase 4; GSSG, glutathione/oxidized glutathione; PTGS2, prostaglandin‐endoperoxide synthase 2; qRT‐PCR, quantitative real‐time polymerase chain reaction; ROS, reactive oxygen species; and SLC25A11, solute carrier family 25 member 11.

SLC25A11 Silencing Inhibits Cell Viability and Induces Ferroptosis in Cardiomyocytes

SLC25A11 was knocked down in AC16 cells by shRNA vector transduction (Figure 3A and 3B) to observe its effect on cell viability and ferroptosis. Cell viability analysis results showed that SLC25A11 knockdown reduced AC16 cell viability, and the reduction was reversed by deferoxamine, an inhibitor for ferroptosis (Figure 3C). The results of the ferroptosis analysis showed that SLC25A11 knockdown increased lipid ROS production (Figure 3D) and Fe²⁺ content (Figure 3E), inhibited GPX4 expression and increased PTGS2 expression, while these changes were reversed when ferroptosis was inhibited by deferoxamine in AC16 cells (Figure 3F).

Figure 3. SLC25A11 silencing inhibits cell viability and induces ferroptosis in AC16 cells.

SLC25A11 expression in AC16 cells transduced with the SLC25A11 shRNA vector was measured by (A) qRT‐PCR and (B) western blot. C, Cell viability; (D) lipid ROS; (E) Fe²⁺; and (F) expressions of GPX4 and PTGS2 in AC16 cells transduced with SLC25A11 shRNA or shNC vector with or without deferoxamine or vehicle in the normoxic condition for 48 hours (n=3). One‐way ANOVA followed by Dunnett’s post hoc test was used. GPX4 indicates glutathione peroxidase 4; qRT‐PCR, quantitative real‐time polymerase chain reaction; PTGS2, prostaglandin‐endoperoxide synthase 2; ROS, reactive oxygen species; SLC25A11, solute carrier family 25 member 11; shNC, scramble short hairpin RNA as the negative control; and shRNA, short hairpin RNA.

METTL3 Induces SLC25A11 m6A Modification in a YTHDF2‐Dependent Manner

m6A sites of SLC25A11 3′UTR were found by SRAMP prediction (Figure 4A). We determined m6A levels of SLC25A11 3′UTR in AC16 cells (Figure 4B). Compared with normoxia, chronic hypoxia increased m6A levels, while this increase was reversed by STM2457 (Figure 4C), a highly potent, specific, and bioavailable inhibitor of METTL3, which is suitable for in vitro and in vivo investigations. ²² Results of luciferase activity showed that luciferase activity of SLC25A11 3′UTR decreased by chronic hypoxia was reversed by STM2457 (Figure 4D). Accordingly, the inhibition of mRNA and protein levels of SLC25A11 induced by chronic hypoxia was reversed by STM2457 (Figure 4E and 4F). In addition, we found that the half‐life of SLC25A11 reduced by chronic hypoxia was abrogated by STM2457 (Figure 4G). YTHDF2, the reader of m6A modification regulators, induces the decay of m6A‐modified mRNA. ²⁷ We knocked down the YTHDF2 in AC16 cells by siRNA transfection (Figure 4H and 4I) to examine its effect on SLC25A11 expression. YTHDF2 knockdown restored SLC25A11 mRNA and protein levels in the condition of chronic hypoxia (Figure 4J and 4K). Moreover, RIP‐PCR results showed that YTHDF2 was enriched in SLC25A11 3′UTR, suggesting YTHDF2 bound to SLC25A11 3′UTR (Figure 4L). These data indicate that METTL3‐mediated m6A modification and YTHDF2‐mediated SLC25A11 mRNA stability synergistically regulate SLC25A11 expression.

Figure 4. METTL3 induces SLC25A11 m6A modification in a YTHDF2‐dependent manner.

A, m6A sites of SLC25A11 3′UTR is predicted by SRAMP. B, m6A levels of SLC25A11 3′UTR in AC16 cells (n=3). C, m6A levels and (D) luciferase activity of SLC25A11 3′UTR in AC16 cells treated with 5 μM METTL3 inhibitor STM2457 or vehicle in the presence of hypoxia stimuli for 48 hours (n=3). E, mRNA and (F) protein levels of SLC25A11 in AC16 cells treated with 5 μM STM2457 or vehicle in the presence of hypoxia stimuli for 48 hours (n = 3). (G) The half‐life of SLC25A11 mRNA in AC16 cells treated with 5 μM STM2457 or vehicle in the presence of hypoxia stimuli for 48 hours (n=3). YTHDF2 expression in AC16 cells transfected with YTHDF2 siRNAs was measured by (H) qRT‐PCR and (I) western blot (n=3). Levels of (J) mRNA and (K) protein of SLC25A11 in AC16 cells transfected with YTHDF2 siRNA or siNC under hypoxia stimuli for 48 hours (n=3). L, The binding between YTHDF2 and SLC25A11 3′UTR (n = 3). C–K, One‐way ANOVA followed by Dunnett’s post hoc test or (B, L) unpaired 2‐tailed Student t test was used. m6A indicates N6‐methyladenosine; METTL3, methyltransferase‐like 3; qRT‐PCR, quantitative real‐time polymerase chain reaction; siNC, scramble small interfering RNA as the negative control; siRNA, small interfering RNA; SLC25A11, solute carrier family 25 member 11; and YTHDF2, YTH domain family 2.

METTL3 Promotes Cardiomyocyte Ferroptosis by Inducing SLC25A11 m6A Modification

The inhibition of AC16 cell viability induced by chronic hypoxia was reversed by STM2457, but the effect of STM2457 was weakened by SLC25A11 knockdown (Figure 5A). Moreover, the increased glutathione/GSSG ratio and OCR level and decreased mitochondrial ROS level induced by STM2457 were reversed by SLC25A11 knockdown in AC16 cells under the chronic hypoxic condition (Figure 5B through 5D). The results of ferroptosis analysis showed that decreased malondialdehyde, Fe²⁺, lipid ROS levels, and PTGS2 expression and increased GPX4 expression induced by STM2457 were reversed by SLC25A11 knockdown in AC16 cells under the chronic hypoxic condition (Figure 5E through 5H). The effects of METTL3‐mediated SLC25A11 m6A modification on the cell viability, glutathione/GSSG ratio, malondialdehyde, and Fe²⁺ contents in AC16 cells under the normoxic condition were shown in Figure S4A through S4D. Similarly, SLC25A11 knockdown also reversed STM2457‐induced changes in cell viability, glutathione/GSSG ratio, OCR, and ferroptosis‐related markers in hiPSC‐CMs under the chronic hypoxic condition (Figure S5A through S5G).

Figure 5. METTL3 promotes ferroptosis in chronic hypoxia–induced AC16 cells by inducing SLC25A11 m6A modification.

A, Cell Counting Kit‐8; (B) glutathione/GSSG ratio; (C) mitochondrial ROS; (D) oxygen consumption rate, (E) malondialdehyde; (F) lipid ROS; (G) Fe²⁺; and (H) expressions of SLC25A11, GPX4, and PTGS2 in AC16 cells treated with 5 μM STM2457 or vehicle and transduced with SLC25A11 shRNA or shNC vector in the presence of hypoxia stimuli for 48 hours (n=3). One‐way ANOVA followed by Dunnett’s post hoc test was used. GPX4 indicates glutathione peroxidase 4; GSSG, glutathione/oxidized glutathione; m6A, N6‐methyladenosine; METTL3, methyltransferase‐like 3; PTGS2, prostaglandin‐endoperoxide synthase 2; ROS, reactive oxygen species; shNC, scramble short hairpin RNA as the negative control; shRNA, short hairpin RNA; SLC25A11, solute carrier family 25 member 11; and YTHDF2, YTH domain family 2.

Suppression of METTL3 Activity Inhibits Ferroptosis and Improves Cardiac Function of Chronic Hypoxia–Induced Mice

The effectiveness of METTL3 inhibitor STM2457 on ferroptosis was verified in myocardial tissues of chronic hypoxia–induced mice. Hematoxylin and eosin staining found normal morphology of cardiac tissues and uniform staining in the control mice; remarkable necrosis, cellular edema, and inflammatory infiltration in the chronically hypoxic mice; and considerably reduced myocardial injury in the 30 and 60 mg/kg STM2457 group (Figure 6A). Consistent with the changes of the ferroptosis index in cells induced by chronic hypoxia, chronic hypoxia increased malondialdehyde content (Figure 6B), Fe²⁺ content (Figure 6C), inhibited GPX4 and SLC25A11 expression, and increased PTGS2 expression in myocardial tissues (Figure 6D and 6E). The changes in these indexes were reversed by STM2457 (Figure 6B through 6E). Furthermore, the effect of METTL3 inhibitor STM2457 on ferroptosis was also verified in primary murine cardiomyocytes isolated from the mice treated with chronic hypoxia with or without 60 mg/kg STM2457. Chronic hypoxia decreased cell viability (Figure 6F), glutathione/GSSG ratio (Figure 6G), OCR level (Figure 6H), increased malondialdehyde (Figure 6I), Fe²⁺ levels (Figure 6J and 6K), inhibited GPX4 and SLC25A11 expression and increased PTGS2 expression (Figure 6L), which were reversed by METTL3 inhibitor STM2457.

Figure 6. Suppression of METTL3 activity inhibits ferroptosis in chronic hypoxia‐induced mice.

A–E, Mice were treated with chronic hypoxia with or without 30 or 60 mg/kg STM2457. A, Representative images of hematoxylin and eosin staining (scale bar, 100 μm). B, malondialdehyde; (C) Fe²⁺; and (D, E) expressions of SLC25A11, GPX4, and PTGS2 in myocardial tissues (n = 3 or 6). F–L, Primary murine cardiomyocytes were isolated from the mice treated with chronic hypoxia with or without 60 mg/kg STM2457. F, Cell viability, (G) glutathione/GSSG ratio, (H) oxygen consumption rate, and (I) malondialdehyde level in primary murine cardiomyocytes (n=3). J, K, Representative images and fluorescence intensity of FerroOrange staining primary murine cardiomyocytes (scale bar, 50 μm) (n = 3). L, Expressions of SLC25A11, GPX4, and PTGS2 in primary murine cardiomyocytes (n=3). One‐way ANOVA followed by Dunnett’s post hoc test was used. GPX4 indicates glutathione peroxidase 4; METTL3, methyltransferase‐like 3; GSSG, glutathione/oxidized glutathione; PTGS2, prostaglandin‐endoperoxide synthase 2; and SLC25A11, solute carrier family 25 member 11.

The cardiac function of mice in all groups was evaluated by echocardiography, and representative images are shown in Figure 7A. Compared with normoxic conditions, LV ejection fraction and LV fractional shortening were significantly reduced after chronic hypoxia treatment (Figure 7B and 7C), and LV end‐diastolic diameter and LV end‐systolic diameter were significantly increased (Figure 7D and 7E). These changes in cardiac function indexes were restored by the METTL3 inhibitor STM2457 (Figure 7A through 7E).

Figure 7. Suppression of METTL3 activity elevates the cardiac function of chronic hypoxia‐induced mice.

A, Echocardiography images. B–E, The image‐based cardiac function indexes included LVEF (B), LVFS (C), LVDd (D), and LVDs (E) (n=9). One‐way ANOVA followed by Dunnett’s post hoc test was used. LVDd, left ventricular end‐diastolic diameter; LVDs, left ventricular end‐systolic diameter; LVEF, left ventricular ejection fraction; LVFS, left ventricular fractional shortening; and METTL3, methyltransferase‐like 3.

Discussion

Chronic hypoxia is a cytopathological state that affects a variety of cell functions and survival by regulating metabolic and cell death pathways (including ferroptosis). ²⁸ , ²⁹ Previous evidence has suggested that ferroptosis is considered to be an important pathological mechanism of hypoxic myocardial cell loss, but the specific mechanism is still unclear. ³⁰ In this study, for the first time, we identified the SLC25A11 downregulation as an important target in chronic hypoxia–induced cardiomyocytes. Mechanistically, we found that SLC25A11 expression was decreased in cardiomyocytes under the chronic hypoxic condition, and SLC25A11 overexpression reversed chronic hypoxia–induced ferroptosis and mitochondrial dysfunction. Moreover, SLC25A11 expression regulation was dependent on chronic hypoxia–induced METTL3‐dependent m6A methylation.

SLC25A11 is considered to be an essential regulatory protein of the mitochondrial antioxidant system because it acts as the key oxoglutarate carrier for the transport of cytoplasmic glutathione to mitochondria. ³¹ Regulation of the mitochondrial glutathione pool can affect metabolic disorders caused by mitochondrial dysfunction in various kinds of cells. ³² , ³³ , ³⁴ Studies have shown that SLC25A11 interacts with the mitochondrial outer membrane protein FUNDC2 (FUN14 domain‐containing 2) to regulate mitochondrial glutathione and ferroptosis. ⁹ Although the mechanism of the mitochondria‐mediated ferroptosis pathway is still unclear, several studies suggest that mitochondria may participate in ferroptosis, in which reduced glutathione participates in ferroptosis by maintaining iron homeostasis and assisting GPX4 to reduce phospholipid peroxidation. ³² Studies have also demonstrated that SLC25A11 promotes the survival of cancer cells in the hypoxic microenvironment by promoting oxidation reduction and energy balance. ³⁵ Our results for the first time demonstrated that SLC25A11 overexpression suppressed common indexes of ferroptosis in chronic hypoxia–induced cardiomyocytes, such as increased glutathione/GSSG ratio and GPX4 expression, and decreased malondialdehyde level and PTGS2 expression. In addition to the GPX4–glutathione antioxidant system, malondialdehyde is the final decomposition product of membrane lipid oxidative metabolism, and its increase represents lipid peroxidation damage, which is positively correlated with ferroptosis. ³⁶ PTGS2 (prostaglandin endoperoxide synthase 2; known as cyclooxygenase‐2) is a critical enzyme for prostaglandin production. Although the exact role of PTGS2 in cardiovascular diseases remains unclear, some studies suggest that it plays a role in cardiac function and pathology, such as promoting inflammation, the mitochondrial respiratory chain of cardiomyocytes and cardiomyocyte hypertrophy, ³⁷ , ³⁸ , ³⁹ , ⁴⁰ which is considered to be an effective marker of ferroptosis. ⁴¹ Overall, we concluded that SLC25A11 is a key regulatory site of ferroptosis injury in chronic hypoxia–induced cardiomyocytes and has the potential to be used as a therapeutic target.

Hypoxia triggers m6A methylation changes in a variety of cells, which affect cell survival. ⁴² , ⁴³ m6A methylation modification is involved in hypoxic adaptive changes in cardiovascular cells, including fibroblasts and cardiomyocytes. ¹⁶ , ²⁴ , ⁴⁴ Although there is little evidence that SLC25A11 expression is regulated by m6A methylation, we identified methylation sites in SLC25A11 3′UTR on the basis of SRAMP prediction and found that the m6A methylation level of SLC25A11 was significantly increased in cardiomyocytes treated with chronic hypoxia. This elevated methylation was reversed by the METTL3 inhibitor STM2457. These results suggest that chronic hypoxia may suppress SLC25A11 expression by inducing m6A methylation. In addition to methyltransferases, which are responsible for initiating the m6A methylation modification process, reading proteins activate downstream regulatory pathways by recognizing m6A methylated bases. ⁴⁵ Our data revealed that the inhibitory effect of chronic hypoxia on SLC25A11 was partially relieved after YTHDF2 knockdown, indicating that YTHDF2 and METTL3 coordinate to promote m6A postmethylation degradation of SLC25A11 mRNA and decrease its expression under chronic hypoxic conditions. METTL3 is considered to be an important methyltransferase that regulates m6A modification, which plays a significant role in modulating tumor cell growth, nervous system diseases, cardiovascular diseases, and other pathological processes. ¹³ , ⁴⁶ , ⁴⁷ METTL3‐m6A has been proven to be involved in hypoxic cardiomyocyte injury, ¹⁵ and METTL3‐m6A can promote myocardial fibrosis and apoptosis after myocardial infarction. ¹⁶ METTL3‐m6A also mediates myocardial ferroptosis induced by sepsis. ⁴⁸ This study is the first to demonstrate that the METTL3‐m6A‐YTHDF2‐SLC25A11 pathway regulates ferroptosis in hypoxic cardiomyocytes.

Neonatal mammalian hearts are capable of regenerating by inducing cardiomyocyte proliferation after injury, and the regenerative capacity of the neonatal mammalian heart gradually diminishes with development and maturation. ⁴⁹ Furthermore, hypoxia exposure can improve cardiac structure and function in mice with acute myocardial infarction by regulating mitochondrial function. ⁵⁰ Mice placed in a hypoxia chamber also demonstrated an increase in the proliferation of right ventricular myocytes, with no effect on LV myocytes. ⁵¹ However, our results showed that chronic hypoxia also decreased cell viability and induced ferroptosis of the primary murine cardiomyocytes isolated from adult mice. Therefore, additional studies are needed to investigate the association between cardiomyocyte proliferation, ferroptosis, and regenerative capacity under the chronic hypoxic condition. In line with our findings, a previous study demonstrated that cardiomyocyte apoptosis was observed after the exposure of the mice to chronic hypoxia. ⁵² Additionally, GPX4 as phospholipid hydroperoxide glutathione peroxidase inhibits cell death, including apoptosis and ferroptosis, by reducing lipid peroxidation, ⁵³ and thus further research is warranted to investigate whether SLC25A11 can play a significant role in cardiovascular diseases by influencing lipid peroxidation‐mediated cardiomyocyte apoptosis under the chronic hypoxic condition.

Conclusions

In conclusion, our study, for the first time, identified the downregulated SLC25A11 as a critical regulatory point of chronic hypoxia–induced cardiomyocyte ferroptosis. Moreover, chronic hypoxia–induced downregulated SLC25A11 is a process of METTL3‐mediated m6A methylation modification. Our results reveal a perspective into the underlying mechanisms of chronic hypoxia–induced cardiomyocyte injury, where increased expression of SLC25A11 protects cardiomyocytes from oxidation disorders and ferroptosis.

Sources of Funding

This research was supported by grants from the Shanghai Baoshan District Medical Health Project (2023‐E‐36).

Disclosures

The authors declare no conflict of interest.

Supporting information

Data S1