Mitochondria‐Homing Drug Mitochonic Acid 5 Improves Barth Syndrome Myopathy in a Human‐Induced Pluripotent Stem Cell Model and Barth Syndrome Drosophila Model

ABSTRACT

Barth syndrome (BTHS) is a rare disease caused by mutations in the tafazzin gene that affects the heart and muscles; however, to date, no clinically effective drugs are available. In BTHS, mitochondrial function is reduced owing to changes in cardiolipin metabolism. We developed mitochonic acid 5 (MA‐5), a small‐molecule compound that increases ATP levels, improves mitochondrial dynamics, and is effective in treating mitochondrial and muscle diseases. Therefore, this study examined the effectiveness of MA‐5 in treating BTHS. The mitochondrial functions of four isolated BTHS skin fibroblasts were examined. Human BTHS induced pluripotent stem cell (iPSC) were differentiated into myoblasts and cardiolipin metabolism and mitochondrial functions were analyzed. RNA‐seq was performed to clarify the metabolic changes. Using a Drosophila melanogaster model of BTHS, the effects of MA‐5 on motor performance and cardiac phenotype were examined. MA‐5 improved mitochondrial function and reduced cell death due to oxidative stress in skin fibroblasts of patients with BTHS. MA‐5 promoted ATP production and reduced oxidative stress in human BTHS iPS cell‐derived myoblasts. RNA‐seq analysis revealed that MA‐5 alleviated endoplasmic reticulum stress in BTHS cells. Administration of MA‐5 to BTHS Drosophila improved locomotor ability and tachycardia observed in patients with BTHS. Protein interaction analyses suggested colocalization of ATPase and the MA‐5‐binding protein mitofilin. These data suggested that MA‐5 improves BTHS dysfunction and may serve as a novel therapeutic agent for BTHS.

Barth syndrome (BTHS) is a rare disease caused by mutations in the tafazzin gene that affects the heart and muscles, but till date, no clinically effective drugs. Using Barth syndrome myopathy in human‐ iPS‐derived disease cells and Drosophila melanogaster model, a new mitochondria‐homing drug MA‐5, improves BTHS dysfunction and may serve as a new therapy for BTHS.

Abbreviations

BSO

L‐buthionine‐[S,R]‐sulfoximine

BTHS

Barth syndrome

CL

cardiolipin

FCCP

carbonyl cyanide‐p‐trifluoromethoxyphenylhydrazone

iPSC

inducible pluripotent cell

MA‐5

mitochonic acid‐5

MLCL

monolysocardiolipin

OCR

oxygen consumption rate

OXPHOS

oxidative phosphorylation

RNA‐Seq

RNA sequencing

ROS

reactive oxygen species

sIBM

sporadic inclusion body myositis

TAZ

tafazzin

1. Introduction

Barth syndrome (BTHS) is a rare X‐linked genetic disease characterized by cardiomyopathy, skeletal myopathy, neutropenia, and growth delay [1, 2, 3, 4, 5]. Disabling mutations or deletions in the tafazzin (TAZ) gene, located at Xq28, cause this disorder by reducing the remodeling of cardiolipin (CL), a principal phospholipid of the inner mitochondrial membrane, resulting in a high monolysocardiolipin (MLCL) to CL ratio (MLCL/CL ratio) [4, 6]. Alterations in CL composition lead to impaired energy metabolism in BTHS mitochondria, which is associated with increased production of reactive oxygen species (ROS), respiratory chain instability, and alterations in mitochondrial dynamics and biogenesis [7]. Traditional treatment approaches are largely symptomatic and focus predominantly on the management of cardiac symptoms through cardiac transplantation [8, 9], antibiotic prophylaxis, and granulocyte colony‐stimulating factor therapy [4].

Using BTHS model induced pluripotent stem cells (iPSC)‐derived cardiomyocytes (iPSC‐CMs), Wang et al. defined the metabolic, structural, and functional abnormalities associated with TAZ mutations and described the assembly of sparse and irregular sarcomeres [10]. Abnormalities in sarcomere assembly and myocardial contraction occurred due to abnormal cellular ATP levels. A metabolic switch from fatty acid oxidation to glycolysis was also observed in the hearts of BTHS knock‐in mice [11]. However, no detailed studies have been conducted on the myopathy caused by BTHS.

Treatments for BTHS are limited to heart transplantation for cardiac symptoms or symptomatic therapy to alleviate the burden [4]; however, there have been no reports on the treatment of myopathy. Recently, we reported that a newly synthesized indole derivative, mitochonic acid 5 (MA‐5), increased cellular ATP levels and improved the survival of fibroblasts from patients with mitochondrial disease [12]. MA‐5 modulates mitochondrial ATP synthesis independent of oxidative phosphorylation and the electron transport chain. MA‐5 targets the mitochondrial protein mitofilin/IMMT/Mic60 at the crista junction of the inner membrane and improves mitochondrial shape. In addition, MA‐5 accelerates ATP synthase dimerization and supercomplex formation [13]. MA‐5 also promotes ATP production by changing the mitochondrial cristae structure; therefore, even if a genetic abnormality or change occurs in any part of complex I–V involved in ATP synthesis, it will enhance ATP production to some extent and improve survival [13]. In addition, MA‐5 increases ATP levels and improved cell survival and mitochondrial dynamics in myopathies such as sporadic inclusion body myositis (sIBM) [14]. sIBM is the most common idiopathic inflammatory myopathy and several reports have suggested the involvement of mitochondrial abnormalities in its etiology. Furthermore, MA‐5 increases cellular ATP levels, reduces mitochondrial ROS levels, and protects cells against sIBM myoblast death. MA‐5 also improved the survival of sIBM skin fibroblasts as well as mitochondrial morphology and dynamics in these cells. These data suggest that MA‐5 can be used as an alternative therapeutic strategy for muscle diseases [14].

In this study, we differentiated iPSC carrying the TAZ mutation, which causes BTHS, into myoblasts and examined the effects of MA‐5 in these cells. Furthermore, we examined the effects of MA‐5 on the locomotor function and heart rate of TAZ^−/− Drosophila melanogaster exhibiting BTHS.

2. Materials and Methods

2.1. Materials

MA‐5 was synthesized and characterized in our laboratory. LC/MS‐grade methanol, acetonitrile, and 2‐propanol were obtained from Kanto Chemical Company (Tokyo, Japan), while LC/MS‐grade ammonium acetate, acetic acid, and formic acid were purchased from Fujifilm Wako Chemicals (Osaka, Japan). BSO (L‐buthionine‐[S,R]‐sulfoximine) was purchased from Sigma‐Aldrich (St. Louis, MO).

2.2. Human Skin Fibroblasts

Fibroblasts obtained by skin biopsy from 4 BTHS patients were collected in Chiba Children's Hospital under the approval of the Ethics Committee of Tohoku University (UMIN000014874), and written informed consent was obtained from all subjects. Fibroblasts were cultured in 1.0 g/L low‐glucose DMEM with 10% FBS (Gibco, Waltham, MA) at 37°C in 5% CO₂.

2.3. iPSC and Differentiation Into Myoblasts

Control (PGP1) and BTHS iPSC (BTHH) [10] were kindly provided by Dr. William T. Pu (Harvard Stem Cell Institute). The iPSCs were maintained in Stemflex (Gibco, Waltham, MA) on Geltrex (Thermo Fisher Scientific, Waltham, MA) coated dishes. The differentiation was done with a Skeletal muscle differentiation kit (Myocea, San Diego, CA) according to a previous report [15].

2.4. Measurement of ATP Production

Fibroblasts from BTHS patients or iPS‐myoblasts derived from PGP1 and BTHH were cultured in 96‐well plates at a density of 3.0 × 10³ cells per well (n = 4–6). The ATP level was measured 6 h after DMSO or MA‐5 treatment (10 μM) by an ATP measurement kit (Toyo Ink, Tokyo, Japan).

2.5. Cell Viability Assay and LDH Level Assay

Cell viability assay was performed as previously reported [12, 13, 14]. Barth syndrome fibroblasts or iPS‐myoblast were cultured in normal growth medium until semi‐confluent and then plated at 3 × 10³ cells/0.2 mL assay medium per each well of 96 well flat bottom culture plate in assay medium (n = 4). After 24 h incubation, BSO at 100 μM or dimethyl sulfoxide (DMSO) for control at 0.1% were added and cultured for another 24 h. After each compound was applied at the indicated final concentrations, the fibroblasts were cultured for 48 h, and cell viability was measured by Cell count Reagent SF (Nacalai Tesque, Kyoto, Japan). LDH assay was also performed (n = 4) with the medium of this plate as previously reported [13, 14], using an LDH cytotoxicity detection kit (Takara, Shiga, Japan).

2.6. Measurement of Mitochondrial Function by a Flux Analyzer

Measurement of the mitochondrial function was performed as we reported previously [14]. The oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) of fibroblasts from BTHS patients and normal volunteers (n = 3) were measured using XFe24 (Agilent Technologies, Santa Clara, CA). For iPS‐myoblasts, OCR and ECAR (n = 14–16) were measured using XFe96 (Agilent Technologies). Briefly, cells were cultured in assay medium (70 mM sucrose, 220 mM mannitol, 10 mM KH₂PO₄, 5 mM MgCl₂, 2 mM HEPES, 1.0 mM EGTA and 0.2% [w/v] fatty acid‐free BSA, pH 7.2) without CO₂ for 60 min. Then, after equilibration, the three respiration rates were measured in fibroblasts following injections of four inhibitors (1 μM oligomycin, 1 μM carbonyl cyanide‐p‐trifluoromethoxyphenylhydrazone (FCCP), 1 μM rotenone and antimycin) of mitochondrial oxidative phosphorylation (OXPHOS).

2.7. Immunostaining

The cells were fixed with 4% PFA/PBS at 4°C for 15 min after a brief wash with PBS. Fixed cells were washed with PBS and then blocked with PBST (PBS/0.4% Triton X‐100)/5% normal donkey serum (Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 h at room temperature. Then, primary antibodies, anti‐Pax3/Pax7 (CAT. #MAB2457, R&D systems, Minneapolis, MN), and anti‐MYOD (CAT. #554130, BD Biosciences, Franklin Lakes, NJ) were diluted in blocking solution and incubated with the fixed cells for 3 h at room temperature. After washing with PBST, secondary antibodies were incubated for 30 min at room temperature. Fluorescent images of the iPS‐myoblasts were captured by BZ‐X800 (Keyence, Osaka, Japan).

2.8. Mitochondrial Superoxide Analysis

To determine the mitochondrial ROS production of iPS‐myoblasts, we used MitoSOX Red mitochondrial superoxide indicator (Invitrogen, Waltham, MA) as described [12]. Briefly, cells were treated with 5 μM MitoSOX and the nuclear staining dye Hoechst‐33 342 (1:1000, Sigma‐Aldrich, St, Louis, MO). Fluorescence was recorded for MitoSOX (n = 6) (excitation (Ex) 510 nm and emission (Em) 580 nm) and normalized to Hoechst 33342 (Ex 350 nm and Em 461 nm) in a SpectraMax M2e (Molecular Devices, San Jose, CA).

2.9. RNA Sequencing (RNA‐Seq)

iPS‐myoblasts (n = 4) were processed with DMSO or MA‐5 (10 μM) for 24 h. RNA was extracted with RNeasy plus (Qiagen, Hilden, Germany). RNA sequence library preparation was performed using a QIAseq Stranded RNA Library Kit (Qiagen), according to the manufacturer's instructions. Illumina libraries were converted into circular single‐stranded DNA libraries and RNA sequencing was performed by DNAFORM (Kanagawa, Japan). Sequence reads in Fastq format were assessed for quality using FastQC. Raw reads were filtered to remove low‐quality reads using Trimmomatic (v0.39). The obtained reads were mapped to human GRCh38 genome using STAR (version 2.7.10a). Reads on annotated genes were counted using StringTie software (v2.2.0). Gene expression levels were measured using the Bioconductor package, edgeR (version 3.15). For gene set enrichment analysis, genes commonly differentially expressed across all group comparisons were subjected to functional annotation using Enrichr. The identified genes were uploaded to the Enrichr online platform for comprehensive enrichment analysis. We conducted Gene Ontology (GO) enrichment analysis using the 2023 version of GO databases, including biological process, molecular function, and cellular component categories. Additionally, MGI phenotype enrichment analyses were performed using the 2024 MGI mammalian phenotype database. Statistical significance of enrichment was determined using Fisher's exact test, with adjusted p‐values used to evaluate enrichment significance. Pathways and functional categories with adjusted p‐values less than 0.05 were considered statistically significant in our analysis.

2.10. Reverse Transcription and Quantitative PCR

cDNA (n = 4) was synthesized using a ReverTra Ace qPCR RT Kit (TOYOBO, Osaka, Japan). Quantitative real‐time PCR was performed using SYBR Green PCR reagents according to the manufacturer's instructions on a StepOnePlus Real Time PCR System (Thermo Fisher Scientific, Waltham, MA). The primer sequences for ATF4, CHOP, GRP94, EDEM, XBP1, and BiP/GPR78 are listed [16] (Table S1). The cycle threshold (Ct) was calculated using the Ct method. The relative mRNA expression levels were normalized to that of GAPDH.

2.11. SDS‐PAGE and Western Blotting

Equal amounts of protein (n = 4; 15 μg) were separated by SDS‐PAGE on a 12% gel. The proteins were transferred onto a polyvinylidene difluoride membrane (BIO‐RAD, #1704273, Hercules, CA). The membrane was incubated overnight with primary antibodies against IRE1 (phospho‐S724, CAT. #ab124945, Abcam), Phospho‐IRE1 (phospho‐S724, CAT. # ab14945, Abcam), PERK (CAT. #12185, Cell Signaling Technology, Danvers, MA), Phospho‐PERK (Thr980, CAT. #3179, Cell Signaling Technology, Danvers, MA), β‐actin (CAT. #sc47778, Santa Cruz Biotechnology), ATP assay cocktail (Total OXPHOS Rodent WB antibody Cocktail (CAT. # ab110413, Abcam), mitofilin (CAT. # 10179‐1‐AP, Proteintech, CAT.# ZRB2003, 1G19 ZooMAb rabbit monoclonal) and incubated for 1 h with HRP‐conjugated secondary antibodies (CAT. #32430 and CAT. #32460, Thermo Fisher Scientific, Waltham, MA)). The protein bands were detected using the enhanced chemiluminescent plus system. The band intensity was measured with ImageJ (NIH, Bethesda, MD). The quantification of the Western blot was repeated across three independent experiments.

2.12. Electron Microscopy

Electron microscopic analysis was performed as previously reported [17]. Briefly, iPS‐derived myoblasts were perfused with 4% paraformaldehyde for 15 min, and the skeletal muscle of Drosophila was dissected and fixed in 2% formaldehyde and 2.5% glutaraldehyde in PBS for 2 h. The tissues were then washed with PBS and post‐fixed in 2% osmium tetroxide for 2 h. All samples were washed twice in PBS, dehydrated in ascending grades of ethanol, and embedded in TAAB Epon 812 (TAAB, Reading, UK). Serial sections (70 nm in thickness) were cut on an Ultracut UCT ultramicrotome (Leica Microsystems, Wetzlar, Germany). After being stained with uranyl acetate and lead citrate, they were examined on a JEM 1400 transmission electron microscope (JEOL, Tokyo, Japan) operated at 80 kV. Digital images were collected by a JEOL charge‐coupled device (CCD) camera. The morphological parameters were compared in iPS‐myoblasts (n = 6–11; number of mitochondria), and the number of abnormal mitochondria with vesicles and altered morphology was counted in Drosophila (n = 10–34; number of mitochondria) by blinded analysis. Area, perimeter, and circularity were quantified with Image J.

2.13. Drosophila Genetics and Husbandry

w ¹¹¹⁸ flies were used as wild type. Homozygous TAZ ^−/− fly strain was a kind gift from Dr. Michael Schlame (New York University School of Medicine) [18]. TAZ ^−/− flies were used as BTHS model. All Drosophila melanogaster strains were maintained at 25°C and 60% humidity. Our study examined male flies because male animals exhibited less variability in phenotype. Flies developed on cornmeal food containing deionized water, 6 g/L agar (Nacalai Tesque), 37.6 g/L EBIOS, 84.7g/L cornmeal 94.1 g/L glucose (Fujifilm Wako Chemicals, Osaka, Japan), 4.2 g/L ethyl p‐hydroxybenzoate (Nacalai Tesque) and 17.6 mL/L ethanol. 100 mM MA‐5 diluted by DMSO (Fujifilm Wako Chemicals) was stored at 4°C. For each treatment, MA‐5 was diluted by 5% sucrose (Nacalai Tesque) into 100 nM, absorbed by filter paper, and administered to flies. For the administration of MA‐5, the flies were transferred to vials with filter paper (ADVANTEC, Tokyo, Japan) soaked with 500 μL of MA‐5 (100 nM) solution. The flies were transferred to normal cornmeal‐food vials and rested for 2 days after the MA‐5 treatment. The flies were reared in dark conditions with avoidance of light to maintain the effect of MA‐5 as MA‐5 decomposes when exposed to light.

2.14. Motor Function Experiments (Climbing Assay)

The flies were collected after hatching in 4 days and divided into 10 groups of 15–20 flies for each. The first climbing assay was conducted from the next day after collection. The climbing ability assay was conducted every ten days [19]. The behavior of the flies was recorded by a GoPro HERO8 camera (GoPro, San Mateo, CA). Before the climbing assay, the flies were transferred into empty vials. The flies were tapped to the bottom of the vial and given 20 s to climb. Afterwards, the ratio of flies that could climb to 5 cm was calculated, the assay for each group was repeated three times, and the average scores were calculated (n = 10).

2.15. Drosophila Cardiac Analysis

The flies were raised at 25°C on standard cornmeal‐agar medium (control food) or those supplemented with MA‐5 of final concentration 10 μM (MA‐5 food). The heart rate was measured as described previously [20] with partial modification. Briefly, white prepupae were collected and placed on a glass‐bottom dish with a drop of sterilized water and viewed through an inverted microscope with a 10× objective. Images of beating hearts were taken through Axiocam 506 mono (ZEISS, Jena, Germany) with 5 ms exposure and 3 × 3 binning. Imaging was done at a controlled room temperature (25°C ± 1°C). After the image acquisition, the constant heartbeat was confirmed with the waveform of intensity change acquired by the edge‐tracing of the heart, then the heart rate was counted manually. Each heart was traced 30 s for 3 times and the average was used as data for an individual. For each experimental group, more than 5 flies were used for the measurement (n = 5).

2.16. Metabolite Extraction

Frozen flies (14.4 ~ 19.1 mg) were immediately plunged into methanol (20 mL/mg) containing an internal standard (500 nM reserpine), and homogenized at 1500 rpm for 5 min to inactivate enzymes. Then, 200 μL of the solution was transferred to another glass tube, 100 μL of chloroform and 20 μL of deionized water were added and mixed thoroughly. The solution was centrifuged at 2400× g for 20 min at room temperature, and 100 μL of the supernatant was transferred to a glass insert vial and used as a sample for liquid chromatography mass spectrometry (LC–MS).

2.17. Cardiolipin (CL) and Monolysocardiolipin (MLCL) Measurement

The number of the samples was n = 3 for iPS‐myoblasts and n = 6 pools of 5–6 individuals for flies. The LC system was the Agilent 1290 Infinity HPLC (Agilent Technologies, Palo Alto, CA). This system was equipped with an automatic degasser, binary pump, thermostated column compartment, and autosampler. Chromatographic separation was performed using an Acquity UPLC HSS T3 C18 column (2.1 i.d. ×50 mm, 1.8 μm; Waters, Milford, MA), and the column temperature was maintained at 45°C. The mobile phase consisted of acetonitrile‐methanol–water (3:1:1), with 5 mM ammonium formate as the eluent A and isopropanol with 5 mM ammonium formate as the eluent B. The initial mobile phase was 100% eluent A at a flow rate of 0.3 mL/min. The gradient profiles of eluent B were 40%, 64%, 64%, 82.5%, 85%, and 95% at 5, 7.5, 12, 12.5, 19, and 20 min, respectively. The injection volume was 5 μL. MS data were acquired from a 6530 Accurate‐Mass Q‐TOF mass spectrometer using the dual ESI of G3251A (Agilent). Samples were analyzed by positive ion electrospray MS. The MS condition was as follows: drying gas 10 L/min at 350°C, nebulizer 55 psig, capillary voltage 3500 V, fragmentor 150 V, skimmer 90 V, OCT1 RF Vpp 500 V, and scan range m/z 100–1700. Lock masses were as follows: Purine at m/z 121.0509 and HP‐0921 at m/z 922.0098 in positive‐ion mode. The peak area value of the target component was corrected by the peak area value of reserpine, and the quantitative values for CL and MLCL were calculated based on 200 nM CL (14:0/14:0/14:0/14:0).

2.18. Metabolome Analysis

The number of the samples was 6 or 7 of 5–6 individuals for drosophilae. Frozen drosophilae were homogenized in 1500 bpm for 1 min with chloroform (1000 μL) and zirconia beads (5 mm × 2, 3 mm × 4) by Shake Master NEO (Bio medical Science). Then, immediately plunged into methanol (500 μL) containing internal standards (20 μM each of methionine sulfone; Fujifilm Wako Chemicals) for cations, MES (Dojindo, Kumamoto, Japan) and CSA (D‐camphol‐10‐sulfonic acid; Fujifilm Wako Chemicals). The solution was centrifuged at 4600× g for 15 min at 4°C, and the 300‐μL upper aqueous layer was centrifugally filtered through a Millipore 5‐kDa cutoff filter to remove proteins. The filtrate was lyophilized and dissolved in 50 μL of Milli‐Q water containing reference compounds (200 μm each of 3‐aminopyrrolidine and trimesate) prior to CE‐TOFMS analysis. All CE‐TOFMS experiments were performed using an Agilent CE‐TOFMS system No. 7 and an Agilent CE‐TOFMS system No. 9. For data acquisition, we used Master Hands ver. 2.18.0.2 [21, 22, 23].

2.19. Immunoprecipitation of Mitofilin Binding Protein

Cells stably expressing Flag‐mitofilin by the Flip‐in T‐Rex systema (Invitrogen) and the control cells with the pcDNA5/FRT/TO vector only were established. The cells (1 mL) were suspended in extraction buffer (50 mM HEPES pH 7.4/300 mM NaCl/0.2% NP‐40) and sonicated. The cell lysates were clarified by centrifugation at 12 000 r.p.m. for 30 min at 4°C. The supernatants were filtered using a Minisart Syringe Filter (Sartorius) and incubated with anti‐FLAG antibody M2 beads (40 μL, Sigma‐Aldrich) for 4 h in the presence of benzonase nuclease (10 μg/mL, Millipore) at 4°C. After washing three times with washing buffer (50 mM HEPES pH 7.4/150 mM NaCl/0.1% NP‐40) and once with PBS, the bound proteins were eluted with 40 μL of 0.1 M glycine buffer at pH 3.0. The eluted samples were neutralized by 4 μL of 1 M Tris–HCl buffer at pH 9.5 and suspended in SDS‐PAGE sample buffer. The samples were boiled for 5 min and resolved by SDS‐PAGE (MULTIGEL II mini 4/20, Cosmo Bio Co., LTD). The gel was stained using a Wako Mass silver stain kit (Fuji Film). Gel slippage was reduced by 100 mM of DTT and alkylated by 100 mM iodoacetamide. After washing, the gels were incubated with trypsin overnight at 30°C. Recovered peptides were desalted by Ziptip c18 (Millipore). Samples were analyzed by nanoLC/MS/MS systems (DiNa HPLC system KYA TECH Corporation/QSTAR XL Applied Biosystems). Mass data acquisitions were piloted by Mascot software.

2.20. Pull‐Down Assay for GST‐Mitofilin Deletion Mutants and ATP5A1 Deletion Mutants

Cell extracts were prepared from HEK293 cell by sonication for 15 s in extraction buffer (10 mM HEPES pH 7.4, 0.35 M NaCl, 0.2% NP‐40 and protease inhibitor cocktail cOmplete (Roche)). The suspension was centrifuged (12 000 g, 10 min) and the supernatant was used as cell extracts. GST‐mitofilin deletion mutant and ATP5A1 deletion mutant were constructed into pGEX4T3 (GE Healthcare) by PCR amplification and expressed in E. coli BL21 strain. GST (control), GST‐mitofilin deletion mutants and ATP5A1 deletion mutants expressed cell lysate were incubated with glutathione magnet beads 30 μL volume (MagneGST Glutathone Particles, Promega) for 1 h at 4°C and washed three times by washing buffer A (50 mM Tris–HCl pH 7.5, 0.3 M NaCl, 0.1% NP‐40). HEK293 cell extracts were incubated with GST (control), mitofilin deletion mutants and ATP5A1 deletion mutants bound magnet beads for 4 h in the presence of Benzonase nuclease (Novagen) at 4°C. After washing three times with washing buffer (0.15 M NaCl. 0.1% NP‐40, 50 mM HEPES pH 7.4), GST (control), GST‐mitofilin deletion mutant and ATP5A1 deletion mutant beads were suspended in SDS‐PAGE sample buffer and boiled for 5 min. The samples were resolved by SDS‐PAGE (SperSepAce 5%–20%, FUJIFILM) and Western blotting was performed with the ATP5A1 antibody and mitofilin antibody, respectively (anti ATP5A1; Abcam ab14748, anti mitofilin; Abcam ab190264).

2.21. Preparation of ATP Synthase‐Enriched Submitochondrial Vesicles

The isolated mitochondria were resuspended in the isolation medium to give a concentration of 100 mg of mitochondrial protein/mL. An equal volume of digitonin solution (12 mg digitonin/mL) was added, and the resultant medium was stirred slowly at 0°C for 15 min. The suspension was diluted with 3 volumes of isolation medium and centrifuged at 9000 g for 10 min. The sediment was again resuspended in 3 volumes of isolation medium and centrifuged [24]. ATP synthase‐enriched submitochondrial vesicles (SMVs) were isolated as previously described [25].

2.22. Statistical Analysis

Data are presented as mean ± standard deviation (SD). Statistical significance was determined using student or Welch t‐test or the One‐way ANOVA Tukey's test. A p‐value significance was set at p < 0.05. GraphPad Prism 9 (Version 9.5.1) (GraphPad Software, Boston, MA) was used for statistical analyses. For metabolome analysis, Metaboanalyst 6.0 was used [26].

3. Results

3.1. MA‐5 Attenuated Mitochondrial Dysfunction in the Skin Fibroblasts of BTHS Patients

We recruited four patients (numbered 1–4). Their characteristics are summarized in Table 1. The patients were clinically diagnosed with BTHS at Chiba Children's Hospital (Japan) and had no family history of the disease. The mean age of disease onset was 1.5 ± 1.5 years, and all the patients were male. Since BTHS is characterized by a reduction in mitochondrial function due to MLCL accumulation following TAZ gene mutations [3, 4], we examined the effect of MA‐5 on cellular ATP levels in BTHS fibroblasts. As shown in Figure 1A, MA‐5 (10 μM) significantly increased ATP levels in all BTHS fibroblasts. Mitochondria are the primary source of ROS, and damaged mitochondria produce excessive mitochondrial ROS (mtROS), resulting in vulnerability to oxidative stress [27]. The glutathione synthesis inhibitor BSO induces mtROS, and fibroblasts from patients with mitochondrial diseases are vulnerable to BSO [13, 28]. Therefore, we examined the effect of BSO on BTHS fibroblasts. As expected, BSO treatment reduced the viability of all BTHS fibroblasts (Figure 1B). Furthermore, MA‐5 inhibited BSO‐induced cell death in BTHS fibroblasts. Additionally, the lactate dehydrogenase (LDH) level in the culture medium containing fibroblasts from all patients with BTHS was also increased under BSO treatment, and this BSO‐induced increase in LDH levels was significantly reduced by MA‐5 (10 μM; Figure 1C). These data suggested that MA‐5 improved mitochondrial function in BTHS fibroblasts.

TABLE 1.

Characteristics of four BTHS patients enrolled in this study.

	Sex	Onset	Genome mutation	Myopathy	Cardiomyopaty	Neutropenia	Failure to thrive	Other phenotype
Patient 1	Male	0d	c.280C>T (p.R94C)		+	+		Lethargy, respiratory distress, hypotonia
Patient 2	Male	6m	c.367C>T (p.R123X)				+	Respiratory distress, heart failure, arrhythmia
Patient 3	Male	0d	c.36_57del (p.V12fs)		+		+	Neonatal asphyxia, respiratory distress, muscular hypotonia, global developmental delay, abnormal facial shape
Patient 4	Male	0d	c.589G>A (p.G197R)		+			Feeding difficulties in infancy, hepatomegaly, seizure, hypothermia, arrhythmia

FIGURE 1.

Mitochondrial abnormalities in skin fibroblasts derived from patients with BTHS. Experiments using skin fibroblasts derived from four human patients. (A) ATP levels in fibroblasts of patients with BTHS were measured. (B) Cell viability and (C) LDH levels were measured after treatment with the glutathione synthesis inhibitor BSO (n = 4). (D) OXPHOS and glycolysis results from extracellular flux analysis of fibroblasts from three patients with BTHS (n = 3). Oligo, oligomycin; FCCP, carbonyl cyanide‐4‐(trifluoromethoxy)phenylhydrazone; A/R, antimycin plus rotenone. Data are expressed as mean ± SD. Data were analyzed using one‐way ANOVA with Tukey's test. *p < 0.05, **p < 0.01, ***p < 0.001.

To clarify the mechanism of mitochondrial dysfunction in BTHS, we examined the mitochondrial bioenergetic functions governing the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) in three isolated BTHS fibroblasts. Based on bioenergetic analysis, the basal OCR, maximum OCR, and OCR for ATP production were reduced in BTHS fibroblasts compared to those in control fibroblasts (Figure 1D). Under these conditions, MA‐5 treatment did not affect the FCCP‐stimulated OCR and ECAR values, suggesting that MA‐5 did not directly affect intracellular mitochondrial respiratory activity and glycolysis [13, 14].

3.2. MA‐5 Improved iPSC‐Derived BTHS Myoblast Metabolism

Skeletal muscle symptoms such as proximal myopathy, easy fatigability, and exercise intolerance have been reported in patients with BTHS [2, 3]. However, few studies have examined skeletal muscle symptoms in BTHS models. To investigate the effectiveness of MA‐5 against BTHS myopathy, we induced myoblasts from the BTHS model iPSCs (BTH‐H, c517delG frame shift) [10]. The control iPSC line PGP1 was also used. The iPSCs were differentiated into myoblasts using an established protocol [15]. Under these conditions, myotubes could not be differentiated. However, we previously demonstrated that MA‐5 improved cell survival, ATP levels, and mitochondrial morphology and dynamics in sIBM patient myoblasts, suggesting the potential of MA‐5 as a therapeutic agent [14]. Therefore, iPS‐derived myoblasts were used for further analyses (n = 3). Immunofluorescence experiments using myogenic progenitor markers 10 days after myoblast induction revealed that 80% of the cells expressed PAX3/7, indicating muscle lineage commitment (satellite‐like cells). On day 18, after the second differentiation step, 70%–80% of the cells stained positive for MYOD1, a key regulator of myogenesis, indicating that the myogenic precursors developed into skeletal myoblasts (approximately 80% iPSC‐myoblasts, Figure 2A).

FIGURE 2.

Experiments with disease model hiPSC‐derived myoblasts. (A) Human iPSCs were differentiated into myoblasts via myo‐precursor cells. Pax3/7, satellite‐like cell markers. MyoD1 as the differentiated myotube. Representative immunostaining of BTHS myoblasts. Scale bars = 50 μm. (B) The ratios of MLCL/CL were measured in control, BTHS, and MA‐5‐treated BTHS myoblasts (n = 3). (C) A representative image of electron microscopy analysis. Abnormal mitochondria with concentric cristae were observed in BTHS iPS‐derived myoblasts and MA‐5 (10 μM) ameliorated the abnormal morphology. Scale bar = 500 nm. The area, circularity, and perimeter were calculated using Image J (n = 11–34). Data were analyzed using one‐way ANOVA with Tukey's test. *p < 0.05, **p < 0.01, ***p < 0.001. (D) Myoblasts from normal control and patients with BTHS treated with DMSO or MA‐5 (10 μM) for 6 h, and MA‐5 increased ATP levels (n = 5–6). *p < 0.05, Student's t‐test. (E) Mitochondrial ROS was measured using MitoSox. Increased mitochondrial ROS levels were decreased with MA‐5 treatment (10 μM; n = 5–6). (F) Cell viability assay of iPSC‐derived myoblasts after 72 h‐DMSO application as control and 72 h‐MA‐5 at 10 μM treatment under 24 h‐BSO treatment‐induced oxidative stress in each myoblast (n = 5). (G) Mitochondrial function in BTHH iPS‐derived myoblast and control (PGP1), as assessed using the cellular OCR (oxygen consumption rate, pmol/min) and glycolysis using the extracellular acidification rate (ECAR, mpH/min) from extracellular flux analyzer (n = 14–15). Data were analyzed using one‐way ANOVA with Tukey's test. *p < 0.05, **p < 0.01, ***p < 0.001. The data represent the mean ± SD.

TAZ mutations increase the MLCL/CL ratio and disrupt mitochondrial bioenergetics in BTHS [29]. To identify these characteristics, we measured the phospholipid profiles of BTHS iPSC myoblasts. The phospholipid profiles confirmed that the TAZ mutation resulted in a high MLCL/CL ratio in BTHS iPSC myoblasts (Figure 2B). We also measured the phospholipid profiles of BTHS iPSC myoblasts treated with MA‐5 (10 μM) and found that the MLCL/CL ratio was unchanged by MA‐5 treatment, indicating that MA‐5 did not mediate a change in the phospholipid profiles. In the skeletal muscles of TAZ‐knockdown mice, the number of mitochondria with various inner membrane abnormalities (swollen cristae, honeycomb‐like structures, concentric layers, and vacuoles) is increased [30]. Therefore, we examined the mitochondrial morphology in BTHS iPSC‐myoblasts using electron microscopy.

The area occupied by mitochondria within the cell, circularity assessment, and the total perimeter of the mitochondrial network were scored [31]. The results indicated that BTHS iPSC‐myoblasts had mitochondrial abnormalities with concentric cristae and enlarged mitochondria, and these morphological changes of BTHS iPSC‐myoblast mitochondria were significantly reduced by MA‐5 (Figure 2C).

Because mitochondrial dysfunction results in decreased cellular ATP levels, we examined cellular ATP levels in BTHS iPSC‐myoblasts. As shown in Figure 2D, the intracellular ATP levels were lower in BTHS iPSC‐myoblasts than those in controls. Treatment with MA‐5 (10 μM) significantly increased the ATP level in BTHS iPSC‐myoblasts (Figure 2D). Taken together, these data suggested that MA‐5 increases ATP levels and improves mitochondrial morphology without altering the cardiolipin profile.

To further assess mitochondrial function in BTHS iPSC‐myoblasts, we measured mitochondrial stress using MitoSOX (Figure 2E). We found that mitochondrial ROS levels were markedly increased in BTHS iPSC‐myoblasts compared to the control cells following BSO (100 μM) exposure. After MA‐5 treatment, mitochondrial ROS levels were significantly reduced in BTHS iPSC‐myoblasts compared to those in non‐treated BTHS iPSC‐myoblasts (BTHH), and the reduced levels in MA‐5‐treated myoblasts were comparable to those in wild‐type iPSC‐myoblasts (PGP1) (Figure 2E). Similar to BTHS fibroblasts, the viability of BTHS iPSC‐myoblasts was reduced by BSO, and BSO‐induced cell death in BTHS iPSC‐myoblasts was ameliorated by MA‐5 (Figure 2F). These data suggested the presence of mitochondrial dysfunction and the effectiveness of MA‐5 in BTHS iPSC myoblasts. Subsequently, we used an extracellular flux analyzer to probe mitochondrial function. iPSC myoblasts differentiated from BTHH cells exhibited reduced basal oxygen consumption, maximal respiration, ATP production, and proton leakage (Figure 2G). MA‐5‐treatment improved basal respiration rates. Furthermore, glycolytic pathway activity was reduced in BTHS iPSC‐myoblasts, and this reduction was not ameliorated by MA‐5. These results differ from those observed in the immature and highly glycolytic BTHS iPSC‐cardiomyocytes [10] but similar to those in our report on sIBM myoblasts [14].

3.3. RNA‐Seq Analysis of BTHS iPSC‐Myoblasts

To elucidate the mechanism of MA‐5 function, we performed RNA‐seq analysis on the control and BTHS iPSC (BTHH) myoblasts with or without treatment with MA‐5 (10 μM). The principal component analysis (PCA) plot revealed the phenotypic changes induced by BTHH and their modification by MA‐5 treatment (Figure 3A). We subsequently conducted a pathway analysis based on the Mouse Genome Informatics (MGI) phenotype (Figure 3B) and Gene Ontology (GO) analysis (Figure 3C) of genes that were upregulated in BTHH group and downregulated following MA‐5 treatment with Enrichr (https://maayanlab.cloud/Enrichr/). The analysis was performed using 74 overlapping genes identified through two criteria: genes with p < 0.05 and logFC > 0 in the BTHH vs. control comparison and genes with FDR p < 0.05, and logFC < 0 in the MA‐5 vs. BTHH comparison. The top‐ranked pathways were consistent with muscle‐related changes observed in the clinical Barth syndrome phenotype. The MGI Mammalian Phenotype pathway analysis revealed significant changes in both skeletal and cardiac muscle‐related pathways (Figure 3B), and muscle contraction and cell diffence (Figure 3C). These findings suggest that MA‐5 normalizes the abnormal muscle formation and contractility induced by BTHH, BTHS iPSC‐myoblasts.

FIGURE 3.

RNA‐seq analysis of BTHS iPS‐derived myoblasts. (A) PCA plot of differentially expressed genes among control, BTHH myoblasts, and BTHH myoblast+MA‐5. (B, C) MGI phenotype pathway analysis and Gene Ontology (GO) analysis with 74 overlapping genes identified through two criteria: Genes with p‐value < 0.005 and logFC > 0 in BTHH vs. control comparison, and genes with FDR p < 0.05 and logFC < 0 in MA‐5 vs. BTHH comparison. (D) Expression of ER stress‐related genes measured using RT‐PCR. For detecting effects on the PERK pathway, ATF4, CHOP, GPR94 and BIP (GPR78) expression levels were measured (upper panel). For detecting effects on the PERK and IRE1α pathways, the expression of XBP1 and its splice form were detected using quantitative real‐time PCR (RT‐PCR) (lower panel) (n = 3–4). (E) Western blot analyses of two ER stress‐related protein pPERK and pIRE1 expressions (n = 4). The relative expression was compared with total protein expression. (F) Gene Ontology (GO) analysis of 94 overlapping genes identified through two criteria: Genes with p‐value < 0.005 and logFC < 0 in BTHH vs. control comparison, and genes with FDR p < 0.05 and logFC > 0 in MA‐5 vs. BTHH comparison. Data were analyzed using one‐way ANOVA with Tukey's test. *p < 0.05, **p < 0.01, ***p < 0.001. Data are expressed as mean ± SD.

Figure 3C shows numerous GO terms related to protein ubiquitination. The production of unfolded proteins that require ubiquitination is generally considered a consequence of endoplasmic reticulum (ER) stress and is associated with mitochondrial dysfunction [32, 33]. The ER plays a pivotal role in protein folding and calcium homeostasis in many mammalian cell types, including skeletal muscles. ER stress‐induced unfolded protein response (UPR) pathways play important roles in the regulation of skeletal muscle mass and metabolic functions under various conditions [34]. The UPR is divided into three branches that may be triggered by specific ER transmembrane proteins such as protein kinase RNA‐activated (PKR)‐like ER kinase (PERK), inositol‐regulating enzyme 1 (IRE1), and activating transcription factor 6 (ATF6) [16, 35]. However, the ER stress status of BTHS muscle is not well characterized. Therefore, we measured the expression of genes related to endoplasmic reticulum stress using a previously reported method [16]. In iPS‐BTHS myoblasts, the expression levels of ATF4, CHOP, and GPR94, which are related to the PERK pathway, significantly increased, and this increase was ameliorated by MA‐5 treatment (Figure 3D, upper panels). Activation of IRE1α splices XBP 1 mRNA, which can be detected using quantitative real‐time PCR (qRT‐PCR) [16]. The levels of total, spliced, and unspliced XBP 1 in BTHS iPSC‐myoblasts were also decreased by MA‐5 treatment (Figure 3D, lower panels).

We also determined the expression levels of PERK and IRE1 α using western blotting (Figure 3E). High expression of pPERK and pIRE1α was detected in BTHH‐iPS‐derived myocytes and the induced ER stress was ameliorated by MA‐5 treatment. These data further suggested that MA‐5 rescued ER stress in BTHS iPSC‐myoblasts.

We next performed GO analyses of the 94 genes that were downregulated in the BTHH group and upregulated in the BTHH+MA‐5 group (using a cutoff of p < 0.05 and logFC > 0 or logFC < 0, genes list in Dataset S1). Among these, genes related to mitochondria were desccribed (Table S2). CHRNA7, encoding the α7 nicotinic acetylcholine receptor, is expressed on the mitochondrial outer membrane, where its activation attenuates Ca²⁺ uptake and cytochrome c release, thereby modulating mitochondrial permeability and early apoptotic events [36]. The 18 kDa translocator protein (TSPO) resides in the outer mitochondrial membrane and is involved in mitochondrial cholesterol import, porphyrin/heme biosynthesis, and the modulation of cell survival (including apoptosis) under stress [37]. The ER‐membrane selenoprotein (SELENOS) has an indirect protective role for mitochondria: it combats oxidative stress and prevents unwarranted apoptosis, as silencing SELENOS disrupts Ca²⁺ homeostasis, causes ROS accumulation and loss of mitochondrial membrane potential, and triggers cell death [38].

Extracellular factors also influence mitochondrial fate: the APOE gene product (apolipoprotein E), particularly the APOE4 isoform, impairs mitochondrial metabolism by downregulating the PGC‐1α/SIRT3 pathway and reducing neuronal ATP production [39]. At the level of gene‐expression machinery, TYW5 (tRNA wybutosine synthase 5) contributes to mitochondrial protein synthesis and biogenesis through its interaction with mitochondrial methionyl‐tRNA synthetase [40].

The splicing factor LUC7L, which is primarily nuclear but is also found in mitochondria [41], is likely involved in metabolic programming, similar to its paralog LUC7L2, which modulates the balance between glycolysis and oxidative phosphorylation. Noncoding RNAs further underscore cross‐talk with mitochondria: the long non‐coding RNA (lncRNA), NEAT1 helps maintain mitochondrial homeostasis, as NEAT1 depletion leads to elongated mitochondria and enhanced retention of transcripts for key respiratory chain components [42], while the imprinted lncRNA KCNQ1OT1 promotes intrinsic mitochondrial apoptosis by sequestering a microRNA (miR‐296‐5p) that normally represses the pro‐apoptotic factor BAX, thereby upregulating BAX‐mediated cell death signaling [43]. The atypical cadherin FAT4 (homolog of Drosophila Fat) has been linked to the control of mitochondrial complex I activity and the cellular metabolic state [44]. According to the Human Protein Atlas (https://www.proteinatlas.org) [45], ZNF26 is mainly localized to the mitochondria. In addition, it is localized to the nuclear bodies. Although some of these differentially expressed genes are related to mitochondrial function, the relatively small number of DEGs warrants caution in drawing broad conclusions about overall mitochondrial involvement.

Pathway analyses were done with the 94 genes listed in Table S2 with enrichr (https://maayanlab.cloud/Enrichr/) (Figure 3F). Among these pathways, CoA transferases, highlighted by CHRNA7 and A PO E, play crucial roles in mitochondrial fatty acid utilization and energy production [46] and are also involved in ketone body metabolism. For other pathways listed in Figure 3F, the nitrogen compound metabolic process pathway showed improvement with MA‐5 treatment. Most amino acid metabolism occurs in the mitochondria, and some of these processes are directly linked to mitochondrial energy production [47, 48]. Thus, the results suggested improvement in some mitochondrial functions.

3.4. MA‐5 Improved Cardiac Symptoms and Movement in Drosophila With BTHS

BTHS affects the heart and skeletal muscles [29, 49] and our data suggest that MA‐5 improves muscle function in BTHS. To clarify the effects of MA‐5 in vivo, we used a Drosophila model. TAZ mutations in Drosophila generate a BTHS‐related phenotype with a triad of abnormal cardiolipins, pathological mitochondria, and motor weakness, leading to mitochondrial myopathy [18] and a reduced ability to climb against gravity [50]. We administered MA‐5 for 10 days after eclosion [18, 50] and examined the results. Similar to BTHS iPSC myoblasts, TAZ mutations resulted in altered MLCL/CL ratios in the whole body of flies, but no change was observed after MA‐5 administration (Figure 4A).

FIGURE 4.

In vivo experiments with Drosophila TAZ‐knockout mutant. w ¹¹¹⁸ flies were used as wild type (WT). Homozygous TAZ ^−/− flies were used as BTHS model. (A) The ratios of MLCL/CL in TAZ mutants were measured (n = 6 per group). (B) Climbing assay results and representative images of three groups (n = 10). (C) A representative transmission electron microscope (TEM) image (left panel) and mitochondrial image analysis (right panel) from skeletal muscle of Drosophila. TAZ deficiency also shows increase in the number of giant mitochondria with an onion‐like appearance [18, 51] (n = 6–11). Scale bars, 5 μm. (D) Heart rate change in BTHS model flies. Heartbeat plotted as a line graph (left panel). The heart rate was changed according to the copy number of TAZ gene (middle panel). Increased heart rate in BTHS model flies was reduced under MA‐5 treatment (right panel) (n = 5–8). The normal group refers to the group that received a normal diet with vehicle treatment. Welch's t‐test or one‐way ANOVA with Tukey's test was used to calculate significance. *p < 0.05, **p < 0.01, ***p < 0.001. n = 5 (WT), 5 (TAZ ^+/− ), 5 (TAZ ^−/− ) and 8 (TAZ ^−/−  + MA‐5) biologically independent samples. Data are expressed as mean ± SD.

To determine the effect of MA‐5 on mitochondrial myopathy, the climbing assay was performed. The number of TAZ^‐/‐ Drosophila that could climb to a certain height within a set time decreased; however, the climbing activity was significantly improved by MA‐5 administration (Figure 4B), Movie S1 (Wild type), Movie S2 (TAZ^‐/‐), Movie S3 (TAZ^‐/‐ + MA‐5).

To explore the changes in Drosophila muscle, we observed the mitochondrial ultrastructure in dorsal indirect flight skeletal muscles using electron microscopy (Figure 4C). TAZ ^−/− muscles contain many structurally abnormal mitochondria that primarily affect the internal cristae membrane, with swelling and disruption of cristae [18]. TAZ deficiency also resulted in an increase in the number of giant mitochondria with an onion‐like appearance [18, 51] (Figure 4C, middle panel). Following MA‐5 administration, the number of large vesicles inside the mitochondria significantly decreased (Figure 4C, right panel). These data suggest that MA‐5 ameliorates the pathological changes to mitochondrial structure in skeletal muscles of TAZ ^‐/‐ Drosophila.

The arrhythmia associated with BTHS manifests as tachy‐ or bradycardia [29, 49]. Drosophila with a mutation in the dystrophin gene or those in other disease models can exhibit dilated cardiomyopathy and an increased heart rate [52, 53, 54]. Therefore, we measured the heart rate of TAZ^‐/‐ Drosophila. Figure 4D shows a graph of the heart rate analysis, detailing the time and amplitude.

Despite a previous report stating that TAZ^‐/‐ Drosophila have a normal heart rate [18], Compared with wild type (Movie S4), we found that the heart rate of TAZ ^+/− mutant (Movie S5) and TAZ ^−/− mutant (Movie S6) Drosophila was higher than that of wild‐type flies (Figure 4D, left). This indicated that the heart rate increased with the number of mutant alleles present. Under these conditions, MA‐5 reduced tachycardia in TAZ ^−/− Drosophila (Figure 4D, right); during a 5‐s period, the heart rates were 15 beats for wild‐type (WT), 20 beats for TAZ ^−/−, and 18 beats for TAZ ^−/− + MA‐5 flies (Movies S7). These data also suggest that MA‐5 improves not only muscle activity but also cardiac symptoms in Drosophila, suggesting a possible improvement in cardiomyopathy in BTHS.

3.5. Metabolomic Analysis of Drosophila With BTHS

To further clarify the mechanism by which MA‐5 improves muscle strength, metabolomic analysis of Drosophila was performed [23]. PCA plots of the metabolomes of WT flies (control) vs. BTHS (TAZ ^−/−) vs. BTHS+MA‐5 are shown in Figure 5A.

FIGURE 5.

Metabolomic analysis of BTHS Drosophila. (A) PCA plots of metabolomic analysis among WT with normal fed, TAZ^−/− with normal fed (BTHS model), and TAZ^−/− with MA‐5 fed. (B) Heatmap showed the metabolites that were significantly upregulated in the BTHS group and downregulated in the MA‐5 treatment group, extracted using p‐values (data were analyzed using one‐way ANOVA with Tukey's test) (C) Enrichment analysis with metabolites listed in Figure 5B. (D) Heatmap showed the metabolites significantly downregulated in the BTHS group and upregulated in the MA‐5 treatment group, extracted by p‐value (one‐way ANOVA with Tukey's test). (E) Enrichment analysis with metabolites listed in Figure 5D.

The PCA shows a slight effect by MA‐5, shifting the mutant Drosophila metabolomes towards the WT group. Metabolites showing convex patterns (increased in BTHS and decreased with MA‐5 treatment) were extracted and visualized in the heatmap shown in Figure 5B, extracted using p‐values (calculated using one‐way ANOVA and Tukey's test). Figure 5C shows the results of enrichment analysis of these metabolites. While only five pathways showed an FDR p < 0.05, the changes in amino acid metabolism and glutathione pathways likely reflected the improvement of BTHS‐induced mitochondrial dysfunction under MA‐5 treatment.

Metabolites showing concave patterns (decreased in BTHS and increased with MA‐5 treatment) are summarized in Figure 5D and extracted using p‐values (calculated using one‐way ANOVA and Tukey's test). Consistent with previous in vitro experiments, the results indicated that MA‐5 improved the BTHS‐reduced ATP production and increased NAD⁺ levels, suggesting enhanced antioxidant function. Furthermore, an enrichment analysis of these metabolites was performed (Figure 5E). Pathways crucial for life‐sustaining energy production in mitochondria, including the TCA cycle, oxidative phosphorylation, fatty acid β‐oxidation, and nicotinamide metabolism, showed comprehensive improvement. Because a metabolic switch from fatty acid oxidation to glycolysis occurs in Barth syndrome knock‐in mice [11, 55], activation of the beta‐oxidation pathway is consistent. Additionally, mitochondrial amino acid metabolism was improved. These changes in amino acid metabolism are consistent with and support the improvements observed in the TCA cycle and oxidative phosphorylation [48]. Even without affecting cardiolipin metabolism, MA‐5 demonstrated comprehensive metabolic improvements through mitofilin‐mediated recovery of mitochondrial function.

3.6. Protein Interaction of Mitofilin and ATP Synthase

MA‐5 binds to mitofilin in the mitochondrial cristae, changing the cristae structure and promoting ATP synthase dimerization, resulting in changes in structural dynamics that promote ATP synthesis without altering the electron transport chain [13]. We also reported that MA‐5 increases supracomplex formation via ATP synthase [13]. Several reports have discussed the possibility that dimerization and oligomerization of this complex inflict a certain curvature on cristae membranes [56]. However, this raise questions about the localization of mitofilin and ATP synthase. Mitofilin is an essential component of the MICOS complex located at the base of the cristae [57], whereas rows of ATP synthase dimers shape the rims and tips of the cristae [58]. Therefore, mitofilin and ATP synthase complexes are thought to be spatially separated. To clarify this, we conducted the following experiments. First, we immunoprecipitated mitofilin to determine its binding protein signatures. As shown in Figure 6A, recombinant mitofilin immunoprecipitated ATP5A and ATP5B as binding proteins, indicating that mitofilin and ATP synthase interacted with each other (Figure 6A). Based on the above results, we performed a database analysis to identify proteins that interact with mitofilin by searching the BioGRID4.4, Harmonize3, and STRING databases. The BioGrid4.4 (https://thebiogrid.org/116185/summary/homo‐sapiens/immt.html) and Harmonize3 (https://maayanlab.cloud/Harmonizome/gene/IMMT) databases revealed that mitofilin possibly interacted with ATP5A1, ATP5B, and ATP5C1. The STRING (https://string‐db.org/cgi/network?taskId=bnQdG65ia9p7&sessionId=bwTcFKcb7ZLB) database search revealed that mitofilin may interact with ATP5F1C (Figure 6B). These data further suggest an interaction between mitofilin and ATP synthase.

FIGURE 6.

(A) Immunoprecipitation of mitofilin binding proteins. (B) Database search for mitofilin binding protein using the STRING database. (C) Western blotting of the submitochondrial vesicle (SMV) fraction. Left panel; IMMT antibody, Right panel Total OXPHOS Rodent WB antibody Cocktail. (D) GST pull down assay. Deletion mutants of mitofilin (left panel) and ATleP5A1 (right panel) were constructed, and the binding of GST‐mitofilin deletion mutants and ATP5A1 deletion mutants was examined. MTS, mitochondrial targeting sequence. (E) Effect of MA‐5 on the interaction between mitofilin and ATP5A1. CBB, Coomassie Brilliant Blue.

We next isolated mitochondria from rat livers and examined the presence of mitofilin in the submitochondrial vesicle (SMV) fractions. Western blotting revealed that mitofilin protein was present in the SMV (Figure 6C). These data further suggested that mitofilin colocalizes with ATP synthase. Furthermore, to clarify the protein–protein interactions between mitofilin and ATP5A1, we created deletion mutants of mitofilin and ATP5A1 with a GST‐tag and pulled down the cell extract, followed by detection using ATP5A1 and mitofilin antibodies, respectively (Figure 6D). Mitofilin mainly interacted with ATP synthase in the 377–590 amino acid region (Figure 6D, left panel). ATP5A1 interacted with mitofilin in the 137–430 amino acid region, which corresponds to the ATPase domain (Figure 6D, right panel). MA‐5 treatment enhanced the interaction between ATP synthase and mitofilin (Figure 6E). These results indicate that mitofilin and ATP synthase exist in close proximity and interact with each other.

4. Discussion

4.1. Mechanism of Action

Treatments for BTHS have been limited to heart transplantation for cardiac symptoms or supportive therapy to alleviate the burden [4]; there have been no reports of specific treatments for BTHS myopathy. In BTHS, the inhibition of cardiolipin remodeling disrupts the mitochondrial structure and formation of the ATP synthase complex, resulting in impaired ATP synthesis [59]. CL promotes membrane curvature to optimize the electron transport chain, and serves as a proton trap on the outer leaflet of the inner mitochondrial membrane to allow rapid lateral diffusion of protons to ATP synthase, with minimal changes in protons, thereby facilitating ATP production [59].

Recently, the tetrapeptide SS‐31, which acts on CL, was developed as a therapeutic drug for mitochondrial diseases [59]. SS‐31 regulates mitochondrial dynamics by binding to CL, promoting the assembly of the ATP synthase complex, and increasing the efficiency of ATP synthesis [59]. SS‐31 also reverses mitochondrial fragmentation in fibroblasts from patients with mitochondrial cardiomyopathy [60]. This implies that, for treatment, not only are drugs that reduce oxidative stress, lipid supplementation [29], and gene therapy [61] required, but also drugs that change the morphology and/or dynamics of mitochondrial cristae and promote ATP production, even under low proton conditions. Here, we found that mitofilin and ATP synthase co‐localized and that MA‐5 facilitated their binding (Figure 6). It is reported that mitofilin interacts with ATP subunits e (Su e/Atp21p) and g (Su g/Atp20p) [62, 63]. Su e and Su g are critical components for dimerization and higher‐order oligomerization of ATP synthase [64]. Dimer formation induces a tight bend in the mitochondrial crista membrane, resulting in a higher surface density of protons in curved membrane regions, local pH differences, and the promotion of ATP synthesis [65]. This conserved arrangement generates a local proton gradient in the cristae space, which explains how the dimer rows help optimize mitochondrial ATP synthesis and provide a functional role for mitochondrial cristae [66]. Su e and Su g are also related to the stable formation of the ATP synthase supracomplex, and highly oligomeric supracomplexes show efficient substrate channeling [67]. Greggio et al. showed that exercise increases the individual components of the skeletal muscle mitochondrial electron transport chain and supracomplex, revealing a novel adaptive mechanism for increased energy demand [68]. Thus, the interaction of the MA‐5‐mitofilin complex facilitates the oligomerization of ATP synthase and increases ATP production, which is beneficial in BTHS fibroblasts, iPS‐derived myoblasts, and Drosophila models. The difference is that SS‐31 is an injectable drug, whereas MA‐5 is a drug that can be administered orally. Considering these results, MA‐5 alone could potentially treat the cardiac and skeletal muscle dysfunctions associated with BTHS.

4.2. Cardiac Availability

Arrhythmias such as tachycardia or bradycardia are commonly reported in patients with BTHS [29, 49]. In addition, ventricular tachycardia is frequent [49, 69]. In the Drosophila model, dystrophin or other disease model mutations cause dilated cardiomyopathy and an increased heart rate, which are compensatory responses to dilated cardiomyopathy [52, 53, 54]. Therefore, we used the Drosophila model to explore the cardiac effects of MA‐5. In an original report on TAZ ^‐/‐ mutants, the flies did not exhibit any cardiac symptoms [18, 70]. However, our observations revealed that TAZ ^‐/‐ Drosophila exhibited tachycardia, which was more severe in homozygotes than in heterozygotes. MA‐5 also suppressed tachycardia. Following a report that heart rate may indicate the severity of dilated cardiomyopathy, the reduction in heart rate by MA‐5 suggests its effectiveness in ameliorating dilated cardiomyopathy.

4.3. ER Stress Reduction

In BTHS, cardiolipin deficiency in the mitochondria leads to an increase in dysfunctional mitochondria [71, 72], and this mitochondrial dysfunction results in ER stress [72, 73, 74]. Therefore, TAZ‐deficient cells have an increased sensitivity to certain ER stress‐mediated and non‐ER stress‐mediated apoptotic triggers [75]. ER stress was detected by RNA‐seq and western blot analyses in BTHS myoblasts, and MA‐5 ameliorated ER stress (Figure 3F,G). In addition, an increase in antioxidants such as taurine induced by MA‐5 treatment may also mitigate ER stress [76, 77] (Figure 5C).

Immunoprecipitation experiments with mitofilin (Figure 6A), ATP5A, ATP5B, and MICOS components, namely the sarco/endoplasmic reticulum calcium ATPase ATP2A2 (SERCA2) and ribophorin 1, all of which are present in the ER, were identified simultaneously. Ribophorin‐1 serves as one of the subunits of the oligosaccharyltransferase (OST) complex located in the ER [78]. ATP2A2 (SERCA2) is the only key player that actively transports cytoplasmic Ca²⁺ into the sarcoplasmic reticulum or the ER and exerts a pivotal role in maintaining the high Ca²⁺concentration of the ER [79, 80]. In addition, Ikeda et al. found that treatment with miclxin (a mitofilin/MIC60 inhibitor) or knockdown of mitofilin induced an increase in the expression levels of phosphorylated eIF2α, ATF4, and CHOP, indicating that dysregulation of mitofilin contributed to the activation of the ER stress response and mitochondrial stress [81]. These data further suggest that mitofilin and the ER interact with each other and that MA‐5 decreases ER stress by enhancing the function of mitofilin. The improvement in ER stress by MA‐5 also shows its therapeutic potential in other diseases involving ER stress originating from mitochondrial dysfunction, such as Alzheimer's disease [82], Parkinson's disease [83] and amyotrophic lateral sclerosis [84].

4.4. Limitations

This study was designed to highlight the effects of MA‐5 on skeletal muscle molecular modifications in BTHS using an internal validity paradigm. However, we were unable to differentiate the BTHS iPSC‐derived myoblasts into myotubes due to unknown technical issues.

ATP synthase and respiratory chain complexes are localized within the mitochondrial cristae and specifically, ATP synthase forms dimer rows at the cristae ridges [85], whereas the proton pumps of the electron transport chain are distributed across the adjacent, continuous planar regions of the inner membrane [86]. In addition, cristae junctions, which form the highly curved neck regions of the cristae, are thought to function as selective entry gates into the cristae space [87]. Moreover, MICOS is essential for the formation of cristae junctions and for maintaining the overall architecture of the inner mitochondrial membrane [88]. Among MICOS components, mitofilin/IMMT/Mic60 plays a pivotal role by shaping the cristae junctions and forming contact sites with the outer membrane [88]. Importantly, its ability to generate positive membrane curvature supports its proposed function at the rims of crista junctions, where such curvature must be stabilized to preserve cristae structure [88]. Taken together, these findings demonstrate a spatial distinction in the localization of cristae‐associated proteins: mitofilin/IMMT/Mic60 is enriched at the crista junctions, whereas ATP synthase is predominantly located at the tips or rims of the cristae. However, this compartmentalization appears inconsistent with our observation of interaction between ATP synthase and mitofilin/IMMT/Mic60. Salewskij et al. reported that under normal conditions, ATP synthase is largely restricted to cristae, as evidenced by orthogonal trajectories along crista membranes. Interestingly, metabolic changes increase the mobility of ATP synthase, leading to its redistribution to other regions of the inner membrane, such as the inner boundary membrane [89], indicating that ATP synthase responds rapidly and reversibly to metabolic conditions through not only functional but also spatial and structural reorganization. Therefore, further investigation using advanced techniques such as immunoelectron microscopy and cryo‐electron microscopy is warranted to elucidate the mechanisms underlying these spatial dynamics.

5. Conclusion

In conclusion, we showed that MA‐5, a mitochondria‐homing drug, increased ATP production and reduced ER stress in BTHS skin fibroblasts and myoblasts derived from patient‐derived iPS cells, and improved exercise performance and heart rate in a Drosophila BTHS model. These data suggest the potential of MA‐5 as a novel treatment for BTHS. Currently, a phase I trial of MA‐5 has been completed in Japan, and the design of a phase II trial has been approved by the Pharmaceuticals and Medical Devices Agency in Japan. Therefore, MA‐5 has the potential to treat BTHS in the near future.

Author Contributions

Yoshiyasu Tongu, Takafumi Toyohara, and Takaaki Abe participated in the conception and design of the study. Yoshiyasu Tongu, Erina Kuranaga, and Takaaki Abe wrote the manuscript. Kei Murayama, Yoshiyasu Tongu, Tetsuro Matsuhashi, and Yoshitsugu Oikawa performed BTHS patient's fibroblast collection and experiment. Shinichiro Kanno performed immunoprecipitaion and GST pull down experiments. Yoshiyasu Tongu, Kyohei Sato, Yoshitsugu Oikawa, Hitomi Kashiwagi, Akihiko Muto, and Kazuhiro Igarashi isolated mitochondria and purified them into SMV. Tomoko Kasahara, Yoshiyasu Tongu, Hidetaka Tokuno, and Takehiro Suzuki analyzed RNA‐seq and metabolome data. Ryota Akimoto, Sayaka Sekine, and Erina Kuranaga performed Drosophila experiments. Yoshiyasu Tongu and Momoka Suzuki discussed and performed iPS experiments. Takamasa Ishikawa and Shinji Fukuda performed cardiolipin analysis. Hitomi Kashiwagi, Shun Watanabe, Shun Itai, Takehiro Suzuki, and Tomoyoshi Soga performed in vitro experiments. Fundings were acquired by Takaaki Abe, Tomoko Kasahara, and Takafumi Toyohara. Takaaki Abe oversaw data collection and interpretation. Yoshiyasu Tongu, Takafumi Toyohara, and Takaaki Abe contributed to writing the discussion. All authors read and approved the final manuscript.

Disclosure

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1.

Movie S1.

Movie S2.

Movie S3.

Movie S4..

Movie S5.

Movie S6.

Movie S7.

Dataset S1.

Data S1.

Text S1.

Tongu Y., Kasahara T., Matsuhashi T., et al., “Mitochondria‐Homing Drug Mitochonic Acid 5 Improves Barth Syndrome Myopathy in a Human‐Induced Pluripotent Stem Cell Model and Barth Syndrome Drosophila Model,” The FASEB Journal 39, no. 12 (2025): e70739, 10.1096/fj.202401856RRR.

Data Availability Statement

The data of RNA‐seq are deposited as PRJNA1181915.