Exercise based Intervention For Metabolic Inflexibility Linked With Lipid Storage Myopathy Using Innovative CRISPR Etf-QO Mutant Knock-in Models

Abstract

Multiple acyl-CoA dehydrogenase deficiency (MADD) is a mitochondrial lipid storage myopathy characterized by impaired fatty acid β-oxidation, mitochondrial dysfunction, and progressive neuromuscular and cardiac disease. MADD is most commonly caused by pathogenic variants in electron transfer flavoprotein dehydrogenase (ETFDH), which encodes electron transfer flavoprotein–ubiquinone oxidoreductase (Etf-QO), a critical redox enzyme that transfers electrons from acyl-CoA dehydrogenases to the mitochondrial electron transport chain. Defective Etf-QO activity disrupts electron flow, promotes reactive oxygen species (ROS) production, and impairs cellular energy metabolism, linking abnormal lipid oxidation to oxidative stress–mediated tissue damage. To investigate the role of redox imbalance in MADD pathogenesis, we generated CRISPR/Cas9 knock-in Drosophila melanogaster models carrying patient-relevant Etf-QO missense mutations (L127R, S296C, and L399F; corresponding to human L138R, S307C, and L409F) within conserved FAD- and ubiquinone-binding domains. Mutant flies developed progressive locomotor impairment, reduced muscle performance, and marked lipid droplet accumulation in skeletal muscle, cardiac tissue, and fat bodies, indicating systemic defects in mitochondrial lipid utilization. Cardiac analyses demonstrated reduced fractional shortening, prolonged heart period, and increased arrhythmia index, consistent with metabolic cardiomyopathy associated with mitochondrial oxidative stress. In vivo respirometry revealed significantly decreased oxygen consumption, reflecting impaired oxidative phosphorylation. At the molecular level, mutant flies exhibited elevated ROS levels and ATP depletion, accompanied by increased expression of AMPK, PGC-1α, and Tfam, suggesting activation of energy stress signaling and compensatory mitochondrial biogenesis. Importantly, endurance exercise significantly improved locomotor and cardiac function while reducing lipid accumulation and oxidative stress. Together, these findings establish a redox-centered in vivo model of MADD and identify oxidative stress as a major driver of disease pathology and a potential therapeutic target.

Introduction

Lipid Storage Myopathies (LSMs) are a heterogeneous group of inherited metabolic disorders characterized by abnormal accumulation of lipids within muscle cells, leading to progressive skeletal muscle weakness and cardiomyopathy^{1, 2}. These disorders arise primarily from disruptions in mitochondrial fatty acid β oxidation (FAO), lipid trafficking, or electron transfer processes required for oxidative metabolism³. Impairments in lipid trafficking and electron transfer affect the use of fatty acids as an energy source, resulting in ectopic lipid accumulation, reduced Adenosine Triphosphate (ATP) production, and progressive dysfunction of high energy tissues such as skeletal muscle and heart⁴. Clinically, LSM spectrum ranges from fatal neonatal and congenital forms to progressive adult onset types⁵. Among LSMs, MADD, also known as Glutaric Acidemia Type II, represents a paradigmatic disorder that links impaired mitochondrial FAO to systemic metabolic dysfunction⁶. MADD is caused by pathogenic variants in genes encoding electron transfer flavoprotein subunit A (ETFA), electron transfer flavoprotein subunit B (ETFB), or most commonly ETFDH⁷. ETFDH functions as a key mitochondrial enzyme that transfers electrons from multiple acyl CoA dehydrogenases to the ubiquinone pool of the respiratory chain⁸. Disruption of this step compromises FAO across short, medium, and long chain substrates, leading to accumulation of lipid intermediates, impaired oxidative phosphorylation, and cellular energy stress⁹. Mutations in ETFA and ETFB are mostly associated with severe neonatal onset form characterized by dysfunction in mitochondrial energy metabolism and early mortality¹⁰. In contrast, mutations in ETFDH often result in late onset MADD with partially retained enzymatic activity⁶. Late-onset MADD patients commonly present with progressive proximal muscle weakness, recurrent rhabdomyolysis, myalgia, and exercise intolerance and cardiomyopathy in some case-reported study^11–14. Because FAO is a dominant energy source during prolonged exercise, defects in electron transfer exacerbate metabolic stress and increase susceptibility to oxidative damage¹⁵. Cardiac tissues subjected to oxidative stress exhibit increased ROS, DNA damage signaling, and impaired structural integrity¹⁶. Although exercise intolerance is common in late onset MADD, available clinical observations indicate that it reflects reduced metabolic flexibility rather than a fixed inability to perform physical activity¹⁷. Many patients recover exercise capacity once metabolic stability is achieved, particularly those with riboflavin responsive variants¹⁸. Physical activity activates AMP-activated protein kinase (AMPK), a key energy sensor that stimulates peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α)^{19, 20}. However, the molecular and physiological consequences of exercise in the context of ETFDH dysfunction remain poorly defined.

Current therapeutic options for MADD are limited and largely supportive. Riboflavin (flavin) supplementation represents the most effective treatment for a subset of patients, particularly those harboring flavin responsive ETFDH missense mutations²¹. Riboflavin is thought to enhance FAD binding, stabilize mutant ETFDH protein, and partially restore electron transfer function²². Nevertheless, not all late onset mutations respond to riboflavin, and many patients continue to experience progressive muscle weakness and cardiomyopathy despite treatment. Dietary fat restriction, carnitine supplementation, and coenzyme Q10 have shown variable benefit, underscoring the need for improved mechanistic understanding and complementary therapeutic strategies²³. Progress in understanding MADD pathogenesis has been limited by the lack of genetically precise and physiologically scalable in vivo models that capture patient specific mutations and age dependent disease progression. Models incorporating patient relevant ETFDH missense mutations remain scarce, as most existing systems rely on acute knockdown, non physiological overexpression, or pharmacological inhibition which cannot capture the genotype phenotype relationships characteristic of late onset MADD. In this context, genetically tractable systems that preserve endogenous regulation while permitting longitudinal analysis are critically needed.

Drosophila melanogaster (commonly referred to as fruit-fly) has emerged as a powerful model organism for studying metabolic diseases affecting skeletal muscle and heart, owing to its high conservation of mitochondrial pathways, well characterized striated muscle and cardiac physiology^24–29. Importantly, the fly ortholog of ETFDH, ETF ubiquinone oxidoreductase (Etf-QO), retains conserved FAD and ubiquinone binding domains and mutations that are disrupted by MADD-associated mutations^{8, 30}. In this study, we generated the first CRISPR knock in Drosophila models carrying clinically relevant missense mutations in Etf-QO, the conserved ortholog of human ETFDH, to model late onset MADD under endogenous regulation. Using locomotor and muscle performance assays, lipid imaging, metabolic and gene expression analyses, and high resolution cardiac physiology, we perform phenotypic and molecular characterization of the mutation. We further assess impact of moderate exercise as a non pharmacological modifier and show that it improves skeletal muscle and cardiac performance while reducing lipid accumulation and oxidative stress, supporting adaptive mitochondrial remodeling as a key response to ETFDH dysfunction. Overall, these findings identify conserved disease phenotypes and mechanisms in LSM and indicate that structured endurance exercise produces measurable functional benefits under Etf-QO mutation backgrounds.

Materials and Methods

1. Drosophila Stocks and Standard Environmental Conditions

CRISPR knock-in Drosophila models were developed to explore the effects of mutations in the Etf-QO gene associated with MADD. The specific point mutations introduced were Etf-QO^L127R, Etf-QO^S296C, and Etf-QO^L399F associated with FAD and ubiquinone (UQ) binding domains, recapitulating mutations observed in human cases of late-onset MADD. The point mutations were generated with the help of Rainbow Transgenic Flies, Inc. through microinjection of sgRNA-target plasmids and donor templates containing the desired mutations into nos-Cas9 embryos. Following multiple rounds of genetic crosses using balancer chromosomes, flies carrying the targeted mutations were identified via PCR and confirmed by sequencing. Assays were performed at two points: mid-age (3 weeks old) and old age (7 weeks old). As previously described^{24, 27, 31, 32}, flies were housed in controlled environmental chambers set to 23°C and 50% humidity under a light-dark cycle lasting 12 hours each. Adult flies were collected immediately after eclosion, sorted by sex, and housed in groups of about 10 per vial. They were maintained in cornmeal diet fed ad-libitum and moved to vials with fresh food every 4 to 7 days until the experimental endpoints.

2. Locomotor and Skeletal Muscle Performance Analysis

To evaluate the functional effect of MADD-associated mutations on muscles, flight ability and negative geotaxis behavior was assessed in adult flies at 3 weeks and 7 weeks.

a. Flight Assay

The flight assay was adapted from established protocols to assess age- and genotype-dependent motor performance^{24, 33}. Briefly, groups of 15–30 adult Drosophila were gently released into the center of a vertically oriented Plexiglass flight chamber illuminated from above. Individual flies were scored based on their directional flight response: upward (score = 6.0), horizontal (4.0), downward (2.0), or flightless (0.0). Flight Index (FI) was calculated for each cohort, representing the average flight performance of the group. Comparisons were made across genotypes and age groups, with all experiments conducted in parallel with age-matched control lines. Detailed information on fly age, genotype, experimental conditions, number of cohorts, total flies tested, and cohort-wise FI values is provided in the Source Data file.

b. Geotaxis Assay

As described in previous studies^{24, 34}, flies were transferred to a fresh vial containing 8–12 flies per trial. A custom-built 3D-printed device was used to hold up to 12 vials simultaneously. This setup included a Raspberry Pi Camera precisely aligned with the vials, connected to a Raspberry Pi 4B and monitor. After a 2-minute acclimation period, the vial was tapped three times to initiate a negative geotaxis response. Climbing behavior was recorded on video for subsequent analysis using a Faster R-CNN deep learning model to detect vial boundaries. Machine learning algorithms were then used to quantify climbing distance, velocity, and performance indices using methods developed by our group³⁵. 3–5 cohorts per genotype and age were included in the assay totaling 50–60 flies.

3. Cytological Analysis

Similar to our recently published protocol^{31, 36}, fly thoraces were fixed in 4% paraformaldehyde (PFA) in PBS for 15 minutes at room temperature, followed by three 10-minute PBS washes. Samples were embedded in OCT, frozen at −20 °C, and cryosectioned at 20 μm for heart and 30 μm for thorax. Sections were taken at enough depth to reveal the IFMs of thorax. Flies were embedded by orienting them on their dorsal side for the former and lateral side for the latter(sagitally). Each slide included samples from all genotypes and experimental groups within a cohort to minimize inter-cohort and inter-slide staining variability. Slides were washed (3 × 10 min in PBS) to remove residual OCT and then stained with Phalloidin-488 and Lipid-Spot-610 for 30 minutes. It was followed by (3 × 5 min) PBS wash. Slides were mounted in the Vectashield mounting media with DAPI. Immunofluorescence imaging was collected using an Olympus BX-63 microscope at 10× magnification, capturing FITC, TRITC, and DAPI channels. Fluorophores were selected based on their emission spectra and minimal spectral overlap to ensure precise signal discrimination. Imaging was performed on randomly selected, non-overlapping fields within each sample. To ensure representativeness, 3–5 samples per experimental group were imaged and analysis was done with the help of CellSens software by quantifying lipid intensity.

4. ROS Measurement

ROS levels were assessed with the help of previously published methods³² using Dihydroethidium (DHE) staining. 3-week male flies were dissected to isolate thoraces and then submerged in Schneider’s medium to maintain tissue hydration. Thoraces were then embedded in OCT without sucrose or fixation to preserve native redox state and cryosectioned onto slides. Sections were washed once in PBS for 10 minutes, followed by incubation with a 10 μM DHE working solution prepared in Schneider’s medium for 15 minutes at room temperature in the dark. After a 10 minute PBS wash, sections were mounted in Vectashield medium and imaged immediately under identical acquisition settings.

5. Cardiac Function Assay

Ex vivo physiological analysis of semi-intact Drosophila hearts was performed following established protocols^{28, 37}. Heart contractions were recorded using direct immersion optics paired with a high-speed digital camera (Hamamatsu Flash 4) capturing at 200 frames per second. Thirty-second videos of beating hearts were acquired using HC Image software (Hamamatsu). Cardiac function was evaluated from high-speed video recordings using the Semi-Automated Optical Heartbeat Analysis (SOHA) software, which quantifies parameters such as heart rate, heart period, systolic and diastolic diameters and intervals, rhythmicity index, and FS³⁸. The software also generates mechanical-mode traces for detailed visualization of heart dynamics³⁸. 25–35 hearts from male and female (data not shown) flies at 3 and 7 weeks of age were analyzed for each genotype, including controls and Etf-QO mutants.

6. ATP Measurement

ATP content in the fly thorax was quantified based on published methods^{34, 39}. 8–10 flies from each group were dissected and the thoraces were placed in cold PBS in ice. Tissues were homogenized in RIPA buffer containing protease inhibitor using a pellet pestle while keeping samples chilled throughout. The homogenate was then centrifuged at 16,000 × g for 10 minutes at 4 °C, and the clear supernatant was transferred to a fresh, pre chilled tube for downstream ATP quantification and protein measurement. Luciferin and luciferase reagents were prepared according to the manufacturer’s instructions. ATP standards were generated from the kit stock to produce a multi point standard curve on the same plate. Samples and standards were loaded into opaque 96 well plates, the reaction mix was added, and luminescence was read at 560 nm using Synergy LX Multi-Mode Reader (Agilent BioTek Instruments). Each group was assayed in technical triplicate, with three biological replicates per genotype. ATP values were interpolated from the standard curve and normalized to protein concentration for each sample. All readings were collected under identical instrument settings, and background signal was subtracted using reagent only wells

7. Gene Expression Analysis

Gene expression analysis was performed using Real Time-qPCR as previously reported by our group for Drosophila tissues³¹. The flies were dissected to isolate thorax and flash frozen in liquid nitrogen and conserved in −80 °C until use. Total RNA was extracted from each biological replicate containing 3–4 thoraces using Zymo Quick-RNA MicroPrep Kit (R1051, Zymo Research, Irvine, CA, USA). Briefly, tissues were homogenized in 150 μL of lysis buffer, incubated at room temperature, and passed through spin columns for purification. Genomic DNA was removed by DNase I treatment (15 min), and RNA was eluted in RNase/DNase-free water. RNA concentration and purity were determined using a Synergy LX Multi-Mode Reader (Agilent BioTek Instruments). Complementary DNA (cDNA) was synthesized from 500 ng of total RNA using iScript RT Supermix (Bio-Rad, #1708840). Reverse transcription was performed under the following conditions: priming at 25 °C for 5 min, reverse transcription at 46 °C for 20 min, and enzyme inactivation at 95 °C for 1 min. qPCR was conducted using 5 ng of cDNA, 200 ng of gene-specific primers, and SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, #1725275) on a CFX Opus Real-Time PCR System (Bio-Rad). Expression levels were normalized to Rpl11 (60S ribosomal protein). Relative gene expression was calculated using the 2^−ΔΔCt method, normalized to reference genes and control samples. All reactions were performed in technical duplicates. Genes with corresponding forward and reverse primers used in this study are: Cpt1-F: TTGCCATCACCCATGAGGG; Cpt1-R: CAATCGCTTTTTCCAGGAACG; SdhA-F: TCGGTGGACAGAGTCTGAAGT; SdhA-R: GCAGCGAGTGACCAGTACG; Dsrebp-F: AGTCGCCGCTTCTCGTCTA; Dsrebp-R: TGTATGGTGGCTGTTGGTTGG; Mef2-F: GAAGCCGAAACGGACTACACA; Mef2-R: GTTGTCGCCGTAAGATCCCG; Duox-F: ATGGCTGGTACAATAACCTGGC; Duox-R: AACCCCATCCGAATAGGAGGG; Grim-F: CAATATTTCCGTGCCGCTGG; Grim-R: CGTAGCAGAAGATCTGGGCC; Ter94-F: AGTCGCGGTGTCCTTTTCTAC; Ter94-R: GGACCCTTGACTGAGATGAAGTT; OPA1-F: CGAGGAGTTCCTACTTGCCG; Opa1-R: GTATCGCAGCTTGAGGGCTC; Pgc-1α (Srl)-F: TATGGAGTGACATAGAGTGTGCT; Pgc-1α (Srl)-R: CCACTTCAATCCACCCAGAAAG; Ampkα-F: TGGGCACTACCTACTGGG; Ampkα-R: ATCTGGTGCTCGCCGATCTT; EtfQO-F: TGCTGTCTTCGGTAGCACTG; EtfQO-R: GTTGATGTTCTGCTTGGGGT; ATPsynD-F: AGTACGAAGCCCTGAAGGTG; ATPsynD-R: CAGAGACTTGAGATGGGCGAT; ATPsynC-F: CCACAGATCAGGTCATTCCAGA; ATPsynC-R: CGAATACTGTTCCGATACCAGC; Tfam-F: AACACCCAGATGCAAAACTTTCA; Tfam-R: GACTTGGAGTTAGCTGCTCTTT; Hid-F: CACCGACCAAGTGCTATACG; Hid-R: GGCGGATACTGGAAGATTTGC; Reaper-F: TGGCATTCTACATACCCGATCA; Reaper-R: CCAGGAATCTCCACTGTGACT; Cox5a-F: CAAAAGGCCACCCTCTATCCC; Cox5a-R: GCATCAATGTCTGGCTTGTTGAA; Etfqo_1 F: TGCTGTCTTCGGTAGCACTG; Etfqo_1 R: GTTGATGTTCTGCTTGGGGT; Etfqo_2-F: TGAAAATGCACAGAGTTCGGAG; Etfqo_2-R: ATAGTGGGTGGTTATCCTGGG.

8. Exercise Protocol in Drosophila

As depicted in the schematic in the result section, 3 days old flies were randomly assigned to exercise or sedentary cohorts and subject to negative geotaxis-based endurance training system, which exploits the innate climbing response of flies⁴⁰. We designed and 3D-printed an exercise setup which holds up to 10 Drosophila vials simultaneously. The flies were housed in standard vials and placed in the exercise apparatus, which allows an operator to repeatedly drop vials to a flat surface, prompting the flies to climb upward. Each session consisted of 15–20 drops per minute for 15 minutes daily, five days per week. All exercise sessions were conducted at 1 PM, corresponding to approximately ZT6 under the 12:12 light–dark cycle. The exercise duration and drop frequency was selected based on empirical observation of tolerance and effect of survival on the mutant flies. After each session, flies were returned to fresh food vials. All experiments were performed in parallel across genotypes and sexes.

9. Real-time Respiration

Oxygen consumption was measured to assess mitochondrial function using a fluorescence-based optical microplate system (Loligo Systems, #SY25200), which enables real-time, non-invasive monitoring of respiration⁴¹. Briefly, flies were transferred into 80 μL glass chambers mounted on a 24-well microplate placed atop a 24-channel optical fluorescence oxygen reader, and the entire setup was maintained at 37 °C in a temperature-controlled incubator. Oxygen concentration was recorded at defined intervals for up to 60 minutes, and blank wells containing no flies served as controls for baseline correction. Respiration rates were calculated as the decline in oxygen concentration (mmol/L per unit time) after subtracting blank values.

10. Statistical Analysis

Statistical analyses were performed using GraphPad Prism version 10 and Python. Assays were analyzed using either one-way or two-way ANOVA, depending on the experimental design. Post hoc comparisons were conducted using Tukey’s, Dunnet’s or Dunn’s multiple comparisons as appropriate. Data are presented as mean ± standard deviation (SD), and statistical significance was defined as follows: p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****). Raw data along with detailed statistical comparisons across genotypes and age groups are provided in the Source Data file.

Results

1. CRISPR Knock-In Etf-QO Mutants Exhibit Impaired Skeletal Muscle Performance and low oxygen consumption

We generated CRISPR knock in Drosophila models carrying three Etf QO missense mutations L127R, S296C, and L399F within conserved FAD and ubiquinone binding domains, recapitulating late onset MADD variants. These lines were produced by microinjecting sgRNA plasmids and donor templates into nos Cas9 embryos (SI1A), followed by balancer based screening and sequence confirmation (SI1 B–D). To assess skeletal muscle performance, we evaluated skeletal muscle performance using flight ability of the flies at 3 and 7 weeks. All Etf-QO mutants showed reduced flight performance compared with controls, with a stronger deficit at 7 weeks (Fig. 1C, D). Across the mutants, L127R showed the largest decline, S296C displayed an intermediate reduction, and L399F was the least affected, and this order was consistent across sexes and ages. To complement these functional readouts, we measured whole body oxygen consumption as an index of metabolic capacity. Oxygen consumption measured at 3 weeks was reduced in all three male mutants and in L127R and S296C females (SI2 A, B). The direction of change matched the severity trend observed in flight assays, with L127R showing the largest reduction respiration. As sustained flight in Drosophila relies heavily on mitochondrial ATP production, the lower respiration rates provide a metabolic correlate to the reduced flight performance.

Figure 1. Progressive decline in flight ability accompanied by increased lipid deposition in Etf-QO mutants.

Domain architecture of electron transfer flavoprotein-ubiquinone oxidoreductase (Etf-QO) with sites of Drosophila (black) and human (red) missense mutation (A). The protein contains two FAD-binding domains (blue), two UQ-binding domains (red), and two 4Fe–4S clusters (yellow) located at the N- and C-terminal regions. Domain arrangement is shown in linear order from N-terminus to C-terminus. Schematic representation of the flight assay setup (B). Flies were released in a vertical flight chamber, and their FI (FI) was calculated and quantified based on their landing positions. The mutants showed diminished flight performance compared to the age-matched controls across both sexes which worsened with age (C, D). n=80–90 flies, 3–5 cohorts. Procedure for cryoimaging IFM muscle (E). Visualization of IFMs in Drosophila (F). Thorax and abdomen were isolated by dissecting away wings, legs, and head. Following fixation, tissues were embedded in OCT, cryosectioned to obtain thorax sections, and then subjected to sequential washes and fluorescence staining. Thorax sections were sagittally cryosectioned and stained for visualizing muscle architecture and lipid deposition in IFM (G). Lipid droplets (red puncta) were more evident in significantly greater intensity in most of the mutants compared to control in both 3-wk and 7-wk flies (H, I). Phalloidin (green) and DAPI (blue) mark the cytoskeleton and nuclei respectively. Quantification of lipid intensity (n=10–20 images, 5 males/group) confirmed these effects. Two-way ANOVA with Tukey’s multiple comparisons test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Scale bar: 100 μm (10X), 20 μm(60X). Black and red asterisks denote comparison across genotypes and age respectively. All the raw data associated with this figure are presented in the Source Data file.

2. Mutant Flies Exhibit Ectopic Lipid Deposition

Lipid imaging of thoracic tissue revealed increased lipid droplet intensity in all Etf-QO mutants relative to controls at both 3 and 7 weeks (Fig. 1E–I). L127R and S296C showed the highest lipid burden across ages, while L399F displayed a milder increase that was most evident at 3 weeks (Fig. 1H–I, SI 3). Both male and female flies showed similar genotype dependent trends, although the intensity appeared slightly elevated in females especially in 7-week L399F (SI3). Because statistical comparisons between males and females were not performed, these observations are presented descriptively. Elevated lipid signal was also confirmed at higher magnification (60×) (Fig. 1I). Thoracic sections were obtained at comparable depths to consistently expose the IFMs. Overall, we observed that ectopic lipid deposition is present across Etf-QO mutation backgrounds, with the strongest signal in L127R and S296C and a comparatively mild effect in L399F.

3. Etf-QO Mutations Affects Locomotor Performance

Unlike flight, which is driven by the IFMs of the thorax, climbing in Drosophila is primarily done through leg musculature and neurons. These muscle groups are also distinct in that flight muscles generate high power, oscillatory output whereas leg muscles perform occasional whole body locomotion. All Etf-QO mutants showed reduced climbing performance measured with geotaxis assay, compared with controls (Fig. 2B–G). L127R consistently exhibited the strongest impairment in both sexes. S296C showed reduced climbing in males at 3 weeks and in females at 7 weeks, while L399F displayed milder deficits. More specifically, the distance climbed at t = 6 seconds was used as a summary metric (Fig. 2 F, G) for comparing climbing activity across and showed similar genotype dependent trend observed in full geotaxis traces. Several deficits in control as well as L127R and L399F were significantly higher in older females indicating age and sex dependent variation in locomotor output. Together, these results show that Etf-QO mutations impair leg driven locomotor behavior in addition to the flight related deficits reported above.

Figure 2. Etf-QO mutations impair locomotion.

Schematic of the geotaxis setup with Raspberry Pi Camera(A). Mutants (Etf-QO ^L127R, Etf-QO ^L399F) showed reduced climbing performance across both sexes (B-E) One-way ANOVA with Dunn’s multiple comparisons test, *p<0.05, **p<0.01, ***p<0.001. Distance covered at half duration (t=6 sec) showed significant deficiency Etf-QO L127R females in mid-age as well as old-age (G). Similar trend was observed in male mutants (F). n=50–60 flies, 3–5 cohorts. Two-way ANOVA with Tukey’s multiple comparisons test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, black and red asterisks denote comparison across genotypes and age respectively. All the raw data associated with this figure are presented in the Source Data file.

4. Cardiac Performance Deficits in Etf-QO mutants

Figure 3 A outlines the semi intact heart preparation and cardiac parameters extraction from high speed videos recordings. Only male hearts were quantified for all cardiac measurements. At 3 weeks, Etf-QO mutants showed cardiac abnormalities relative to controls, including higher Arrythmia Index (AI) across all three mutant lines in both age groups (Fig. 3 D). AI increased significantly in all lines at 7 weeks, with a greater rise in the Etf QO mutants compared with controls. Fractional Shortening (FS) was also significantly reduced in all the mutant lines at 3 weeks and in L127R and S296C at 7 weeks with 7-week flies exhibiting significantly higher impairment compared to 3-week flies across all lines (Fig. 3 I). Diastolic and systolic diameters were smaller in L127R at 3 weeks, and by 7 weeks diastolic diameter was significantly lower in all three mutants (Fig. 3G). Diastolic diameter remained comparable between 3 week and 7 week cohorts, whereas systolic diameter decreased in L127R, S296C, and control (Fig. 3G, H). This pattern indicates that the age related decline in FS and contractile performance is driven primarily by reduced systolic displacement rather than diastolic dilation. Conduction timing parameters showed a comparable pattern among genotypes at 3 weeks, and heart rate declined in all lines by 7 weeks (Fig. 3B). In addition to these cardiac changes, abdominal sections surrounding the heart showed greater lipid accumulation in the fat bodies of 3 week male mutants, with quantitatively elevated levels in L127R and S296C (Fig. 3 J, K), indicating that systemic lipid dysregulation parallels the observed cardiac defects.

Figure 3. MADD-induced LSM disrupts cardiac performance and lipid accumulation in the fat bodies.

Schematic of procedure for visualizing cardiac physiology in Drosophila (A). Decreased heart rate and elongated diastolic interval observed in 3-wk male mutants compared to age-matched controls (B, E), which worsen with age. AI (D) was higher in Etf-QO ^L399F at 3 wk, with significant worsening observed at 7 wk in all three mutants. Two-way ANOVA with Tukey’s multiple comparisons test. *p<0.05, **p<0.01, ****p<0.0001. Black and red asterisks denote comparison across genotypes and age, respectively. n=25–30 flies per genotype and age. Higher lipid accumulation was observed in 3-week male mutants in the fat bodies region surrounding the heart (J) and exhibited significantly higher intensity in two mutants (K). One-way ANOVA with Dunnett’s multiple comparisons test, *p<0.05, ***p<0.001, ****p<0.0001, n=8 images, 5 3-week males per genotype. Scale bar: 20 μm. All the raw data associated with this figure are presented in the Source Data file.

5. Etf-QO Mutations Increase ROS with Reduced ATP and Alter Mitochondrial and Stress Transcripts

DHE staining of IFM tissue showed higher ROS levels in all Etf QO mutants compared with controls (Fig. 4 A, B). Cellular ATP was significantly reduced in L127R and S296C and relatively preserved in L399F (Fig. 4C). Duox transcripts were elevated across all mutants (Fig. 4E), consistent with increased oxidative burden. Energy stress and mitochondrial remodeling pathways were broadly activated. Srl/PGC 1α and AMPKα were upregulated in all three lines (Fig. 4 G, H). Opa1 (Fig. 1F) and Tfam (Fig. 4I) were significantly higher in L127R and L399F, with minimal change in S296C, suggesting allele specific differences in mitochondrial fusion and mtDNA maintenance. Cox5 showed a slight upward trend across all backgrounds (Fig. 4M). ATP synthase subunits showed similar responses. AtpsynD (Fig. 4D) remained unchanged, while AtpsynC (SI4 A) was downregulated in L127R, consistent with highly reduced ATP content in this line. Additional metabolic genes showed genotype specific effects. Mef2 (SI4 B), SdhA (SI4 D), and Cpt1 (SI4 C) were significantly reduced in L127R. Ter94 (SI4 E) remained broadly comparable among lines, though it trended slightly higher in L127R, indicating mild activation of protein quality control pathways. Dsrebp showed a non significant upward trend in all mutants (Fig. 4J). The Etf QO transcript level remained unchanged across the alleles, indicating that the observed phenotypes result from the functional consequences of the missense mutations rather than alterations in Etf QO gene expression.

Figure 4. Etf-QO Mutations Elevate ROS Levels and Modify Mitochondrial and Stress-Related Gene Expression.

Quantification of ROS using DHE staining (A, B). ATP content was measured from thoraces of adult Drosophila melanogaster to assess mitochondrial bioenergetic function at the tissue level (C). Relative mRNA expression levels of mitochondrial, lipid metabolism and stress-response genes assessed (B-N). Expression normalized to Rpl11; data represent mean ± SD from three biological and two technical replicates obtained from thorax samples of 3-week-old males. One-way ANOVA with Dunnetťs multiple comparisons test, *p<0.05,**p<0.01,***p<0.001,****p<0.0001. All the raw data associated with this figure are presented in the Source Data file.

6. Exercise Mitigates Physiological and Metabolic Deficits and Induces Compensatory Transcriptional Response

Only 3 week males were used to assess the effect of exercise. Exercised cohorts showed higher FI than their sedentary counterparts in all three Etf QO mutant lines, whereas control flies exhibited no change, indicating that exercise mitigates mutation linked deficits rather than enhancing baseline performance (Fig. 5B). In contrast, geotaxis outcomes showed minimal improvement with exercise (Fig. 6). L127R displayed a slight upward shift in climbing performance (Fig. 6E–F) in both males and females but this change was not statistically significant, and S296C and L399F showed no clear improvement (Fig. 6E–F). These findings indicate that the exercise regimen has minimal effect on leg driven locomotor behavior although it produced measurable benefits in high power flight performance. In parallel with the improvement in flight ability, the 15 minute daily training protocol reduced lipid accumulation in dorsal longitudinal muscles across all mutants (Fig. 5C–D). The reduction was also evident at higher magnification (SI5). DHE imaging revealed a downward trend in ROS across exercised groups, with S296C showing a significant reduction (Fig. 5E–F), suggesting partial alleviation of oxidative stress. Transcriptional analysis showed coordinated activation of mitochondrial remodeling pathways. Tfam and AMPKα increased in all exercised groups, and PGC 1α was significantly elevated in L127R (Fig. 5G–I), supporting engagement of an AMPK PGC 1α–dependent compensatory program. Together, these findings indicate that moderate endurance exercise reduces lipid and ROS burden, improves muscle performance, and activates mitochondrial biogenesis pathways specifically under Etf QO dysfunction, consistent with functional and metabolic rescue rather than enhancement of baseline physiology.

Figure 5. Moderate daily exercise intervention showed beneficial effect in mitigating lipid deposition and improving flight ability.

Schematic for implementation of exercise regimen (A). Exercise (15 min/day for 2.5 weeks) improved flight performance compared in 3-week males compared to the sedentary controls. (B). n=80–90 flies, 3–5 cohorts. Higher lipid deposition observed in IFM of Etf-QO mutants was also attenuated as evidenced by visible decrease in lipid intensity (C, D) in IF images quantified using CellSense Imaging Software, Olympus Life Science. Lipid reduction was accompanied by decrease in oxidative stress as evidence by DHE staining (E, F) along with upregulation of key energy-stress-responsive genes Tfam, Srl and Ampk (G-I) n=15–25 images from 5 male flies per group at 3 weeks of age. Two-way ANOVA with Tukey’s multiple comparisons test **p

genotypes and exercise conditions, respectively. Scale bar: 100 μm. All the raw data associated with this figure are presented in the Source Data file.

Figure 6. Effect of moderate exercise on locomotor dysfunction linked with Etf-QO mutations.

Exercise exhibited mixed results with modest improvement in climbing performance (A-D) and distance climbed at half duration in Etf-QO ^L399F mutant flies (E-F). Data represent mean ± SD from 3–5 cohorts (n = 80–90 flies per genotype). One-way ANOVA with Dunn’s multiple comparisons test, ***p<0.001 (A-D). Two-way ANOVA with Tukey’s multiple comparisons test

**p<0.01, ***p<0.001. All the raw data associated with this figure are presented in the Source Data file.

7. Exercise Improves Cardiac Parameters in Etf-QO mutants

Exercised cohorts showed improved rhythm and contractility within each genotype. AI (Fig. 7C) was lower and FS (Fig. 7H) was significantly higher in all four lines compared with their sedentary cohorts. Systolic and diastolic diameters (Fig. E, F) in exercised groups was associated with a tendency toward smaller end systolic diameter with little to modest change in end diastolic diameter, which is consistent with the observed increase in FS. Heart rate was lower and heart period was longer in exercised groups, reflecting improved contraction efficiency (Fig. 7 A, B). Together, these panel level readouts indicate an exercise associated shift toward more regular rhythm and stronger contraction across the three Etf-QO mutant lines (Fig. 7A–H).

Figure 7. Exercise induced improvements in cardiac function in EtfQO mutants and a working model illustrating the underlying metabolic and mitochondrial responses.

Sedentary mutants exhibited cardiac dysfunction characterized by elevated AI and reduced FS which was significantly improved in exercised cohorts as evidenced by decreased AI (C) and increased FS (H) indicating enhanced pumping efficiency. Working model summarizing the relationship between ETFDH-linked dysfunction, mitochondrial stress and exercise-associated adaptive responses (I). ETFDH linked mutations are associated with lipid accumulation which can lead to elevated oxidative stress, impaired respiration, and reduced ATP production, contributing to cardiac and muscle contractile dysfunction. The low energy stress activates AMPK signaling leading to induction of PGC 1α–dependent mitochondrial biogenesis and remodeling, which can counteract lipid and ROS mediated mitochondrial dysfunction. Exercise enhances this adaptive response by synergizing with the Ampk-Pgc 1α alpha axis. Solid arrows indicate experimentally observed associations, whereas dashed arrows denote proposed or inferred relationships. n = 25–30 hearts per genotype from 3-week-old male flies. Two-way ANOVA with Tukey’s multiple comparisons test, *p<0.05,**p<0.01,***p<0.001,*8**p<0.0001, black and red asterisks denote comparison across genotypes and exercise conditions, respectively. All the raw data associated with this figure are presented in the Source Data file.

Discussion

1. Modeling Human ETFDH Mutations in Drosophila

Clinically, MADD presents with a spectrum of symptoms, including progressive muscle weakness, lipid accumulation in the skeletal and cardiac muscle, and in some cases, cardiomyopathy and metabolic crises^{7, 42, 43}. To model MADD phenotypes in vivo, we generated CRISPR knock-in Drosophila lines carrying patient-relevant missense mutations, L127R, S296C, and L399F, in the Etf-QO gene. Our CRISPR knock in mutations reflect well documented pathogenic contexts of human ETFDH variants (Fig. 1A). The L127R fly variant maps to the conserved FAD binding domain position p.Leu138Arg (c.413T>G), reported as pathogenic in ClinVar⁴⁴ and consistently associated with late onset MADD featuring progressive proximal weakness, exercise intolerance, myalgia, elevated CK, and episodic rhabdomyolysis^{13, 45}. The S296C fly variant lies within the FAD domain and parallels human p.Ser307Cys (c.920C>G), classified likely pathogenic⁴⁶ and situated in the riboflavin responsive mutational spectrum^47–49. The L399F fly variant corresponds to the UQ binding domain and aligns with human p.Leu409Phe (c.1227A>C), designated pathogenic⁵⁰ and linked to variable severity with occasional cardiac involvement, often improved by riboflavin and CoQ10^51–53. Consistent with these human patterns, the three fly lines exhibited progressive locomotor (Fig. 1C–D) and cardiac dysfunction (Fig. 3 D, I), lipid accumulation (Fig. 1 G–I, SI3 A–C), and oxidative stress (Fig. 4 A–B). Moreover, L127R mutant showed significant reduced climbing ability across the cohorts (Fig. 2 B–G). Other mutant S296C showed impaired locomotor ability at 7 weeks for female (Fig. 2 E) and 3 weeks for male (Fig. 2 B) suggesting differential effects of the mutations due to allele specific disruption of Etf QO enzymatic efficiency and variable thresholds of metabolic stress tolerance across muscle groups. L127R lies within the conserved FAD binding domain of Etf QO adjacent to a 4Fe–4S cluster that supports electron transfer from upstream acyl CoA dehydrogenases. Missense mutation at this position seems to more strongly disrupt cofactor interaction and electron transfer efficiency. The precise structural basis behind this requires further study. Immunofluorescence imaging revealed ectopic lipid accumulation in thoracic (Fig. 1 E–G) and abdominal tissues, mimicking the lipid accumulation seen in human tissues⁵⁴. Cardiac assessments further demonstrated functional impairments such increased AI (Fig. 3 D), and reduced FS (Fig. 3 I), potentially due to metabolic or mitochondrial dysregulation. In fact, consistent with our findings impaired contractile performance associated with mitochondrial cardiomyopathy reported in MADD⁴². Overall, the L127R mutation produced the most severe phenotypes across muscular, metabolic, and cardiac domains, suggesting that specific mutations may differentially impact Etf-QO function and disease severity^{55, 56}. This mirrors the clinical heterogeneity observed in MADD patients, where genotype-phenotype correlations are increasingly recognized⁴³. While not a complete representation of the human condition, the model offers a platform for exploring the disease in a controlled setting.

2. Progressive Neuromuscular Decline Reflects Clinical MADD Trajectory

In our Drosophila models, we observed a clear decline in motor performance that recapitulated the clinical trajectory of late-onset MADD. Flight ability was markedly reduced across all alleles and worsened with age, whereas geotaxis defects were present but less pronounced (Fig. 1C–D, Fig. 2B–G). FI worsened with age, mirroring the progressive motor decline seen in human patients⁵⁵. Sustained flight in flies relies heavily on mitochondrial ATP production, and the combined reduction in FI and oxygen consumption (SI2) is consistent with a progressive energetic constraint in indirect flight muscle (IFM). These patterns indicate that tissues with high energetic demand are disproportionately affected when electron transfer capacity is compromised and underscore the critical role of Etf-QO gene in maintaining mitochondrial energy homeostasis in high-demand tissues such as skeletal muscle¹. The gradual decline in motor output is consistent with accumulating metabolic stress in muscle, including impaired fatty acid oxidation, reduced ATP availability, and accompanying redox imbalance, all of which were observed in the mutant backgrounds.

3. Ectopic Lipid Deposition in Etf-QO Mutations

A hallmark of MADD is the pathological accumulation of neutral lipids in muscle tissues¹. In our Etf-QO mutations model, we observed a pronounced level of lipid droplet in the thoracic tissues in the mutants compared to the age- and sex-matched control (Fig. 1 E–G). These findings reflect clinical observations in MADD patients, where muscle biopsies often reveal vacuolated fibers filled with lipid droplets, especially in late-onset cases⁵⁷. This phenotype worsened with age although it was statistically insignificant (Fig. 1 H–I). Mechanistically, lipid accumulation likely results from a bottleneck in mitochondrial β-oxidation caused by ETFDH dysfunction, which impairs electron transfer to the respiratory chain and leads to the buildup of acyl-CoA intermediates⁵⁵. These intermediates are then diverted into cytoplasmic lipid droplets (Fig. 1 G–I), a compensatory but ultimately pathological response⁵⁸. In human LSM cases, lipid overload has been implicated in sarcomeric disorganization and muscle fiber degeneration⁵⁹. This deterioration may arise from altered lipid–protein interactions at the sarcomere or from ROS mediated damage to cytoskeletal components, both of which are exacerbated by mitochondrial dysfunction⁵⁹. Our mutants show both lipid accumulation and elevated ROS (Fig. 4A–B), suggesting that defective fatty acid utilization and redox stress act together to exacerbate muscle vulnerability.

4. Cardiac Dysfunction Linked with Mitochondrial Impairment

Mitochondrial dysfunction is evident in Etf QO mutants from significantly reduced thoracic ATP in L127R and S296C (Fig. 4C) together with a compensatory transcriptional program indicative of energy stress and remodeling, including increased AMPKα and PGC 1α (Fig. G, H) and allele specific elevations in Opa1 and Tfam (Fig. 4 F, I). These mitochondrial abnormalities align with the cardiac phenotype as mutants exhibited hallmark features of cardiomyopathy such as elevated AI and reduced FS (Fig. 3 D, I), together with age-related changes in end diastolic interval (Fig. 4 E). Collectively, these data indicate that compromised mitochondrial energetics contribute to cardiac dysfunction in this model.

Clinically, cardiac involvement is a recognized component of MADD, encompassing cardiomyopathy, arrhythmias, and reduced cardiac output ⁶⁰. Our findings in Drosophila are congruent with reports that defects within electron transfer components precipitate bradycardia, rhythm instability, and diminished contractile efficiency in mitochondrial disorders ^{61, 62}. Mechanistically, impaired electron transfer between ETF and the respiratory chain can reduce ATP production and increase oxidative stress, processes expected to exacerbate Ca²⁺ handling deficits and sarcomeric dysfunction ^{61, 63}. Importantly, this work provides one of the first in vivo demonstrations of cardiac dysfunction in a genetically engineered Drosophila model of patient relevant ETFDH mutations ⁶⁴.

5. Redox Imbalance and Compensatory Pathways in Disease Pathogenesis

Oxidative stress can play a major role in pathogenesis and progression of metabolic disorders, including MADD, diabetes, obesity, and metabolic syndrome⁶⁵. All three mutant lines exhibited significantly elevated ROS levels, as evidenced by increased DHE fluorescence intensity (Fig. 4A–B), accompanied by concordant upregulation of the ROS generating enzyme Duox (Fig. 4E). This oxidative stress likely arises from impaired electron transfer within the mitochondrial respiratory chain due to ETFDH dysfunction, which leads to electron leakage and the generation of superoxide radicals⁶⁶. The accumulation of ROS can damage mitochondrial DNA, proteins, and lipids, further exacerbating mitochondrial dysfunction and contributing to tissue degeneration⁶⁶. These findings are consistent with previous studies in patient-derived cells, where ETFDH mutations have been associated with increased oxidative stress and mitochondrial damage⁶⁶. Our data extends these observations by providing in vivo evidence of redox imbalance in a genetically tractable Drosophila model, enabling precise dissection of the temporal and tissue-specific dynamics of oxidative stress in LSM.

At the molecular level, redox imbalance observed in Etf-QO mutants may act as both a consequence and a driver of disease progression. On one hand, impaired FAO leads to the accumulation of incompletely oxidized lipid intermediates, which can disrupt mitochondrial membrane potential and promote ROS generation⁶⁶. On the other hand, chronic oxidative stress can impair mitochondrial biogenesis, damage contractile proteins, and trigger apoptotic pathways, thereby accelerating muscle and cardiac degeneration⁶⁷. Transcript analysis suggests mutants exhibit a coordinated compensatory program spanning mitochondrial biogenesis, dynamics, and quality control. For instance, Spargel (Srl/PGC 1α) is strongly upregulated in all three lines (Fig. 4 G) and Ampk (Fig. 4 H) is likewise increased, supporting activation of the Ampk-Pgc 1α axis to enhance mitochondrial biogenesis and oxidative capacity under stress. Downstream, Tfam is up in L127R and L399F (Fig. 4 I), and Opa1 (Fig. 4F) showed a similar pattern, indicating engagement of mtDNA maintenance and inner membrane fusion/remodeling respectively. These responses are known to stabilize mitochondrial networks and sustain contractile function in muscle and heart ^68–71. Cox5 exhibited a uniform upregulated trend across mutants (Fig. 4 M) which can be interpreted as attempts to sustain electron flux and preserve oxidative phosphorylation despite upstream limitations imposed by ETFDH dysfunction^{72, 73}. Markers of tissue maintenance and survival further reflected disease burden. Mef2 expression was reduced in L127R and S296C (SI4 B), consistent with impaired myogenic maintenance. Apoptotic regulators Hid was up in L399F and Reaper trended upward in all three lines (Fig. 4 K, L). Dsrebp (lipogenic transcription factor) also trended upward in all mutants, matching the lipid droplet accumulation we observed histologically (Fig. 4 J).

6. Exercise as a Non-Therapeutic Intervention of LSM

A major translational insight from this study is the therapeutic potential of low-to-moderate exercise in mitigating the pathological features of LSM caused by ETFDH mutations. In our models of MADD, a 15-minute daily exercise regimen led to significant improvements in locomotor performance and cardiac function, as evidenced by enhanced flight ability (Fig. 5B) as well as improved cardiac contractility and rhythmicity (Fig. 7 A–H). These functional gains were accompanied by a marked reduction in lipid droplet accumulation as revealed by immunofluorescence imaging (Fig. 5C–D) suggesting that physical activity is beneficial in curtailing lipid overload. Exercise was associated with a reduced oxidative burden (Fig. 5 E–F) along with elevated expression of mitochondrial remodeling–related genes such as Srl (PGC 1α) and Tfam (Fig. 5G–I). Together, these findings support a framework in which exercise engages adaptive programs linked to mitochondrial remodeling that may help buffer lipid and ROS associated mitochondrial dysfunction, rather than directly reversing the underlying genetic defect. These observations are summarized in a working model (Fig. 7I). ETFDH linked mutations drive lipid accumulation, which associates with reduced respiration and ATP synthesis and culminates in impaired contractile performance. In parallel, mutations increase ROS and oxidative stress, contributing to mitochondrial dysfunction and further exacerbating contractile deficits. (Fig. 7I). Exercise acts on this network by inducing energy stress that activates AMPK, which in turn upregulates PGC 1α and promotes mitochondrial biogenesis and remodeling. This adaptive program is proposed to lower lipid burden and oxidative stress and to partially restore the mutation-linked dysfunction (Fig. 7I).

Our study is consistent with previous works in mammalian and Drosophila models⁷⁴, where exercise has been shown to activate AMPK signaling, enhance mitochondrial biogenesis, and improve muscle endurance and metabolic flexibility in the context of mitochondrial dysfunction⁶⁷. Memme et al. in 2021 also demonstrated that endurance exercise promotes mitochondrial remodeling and improves muscle oxidative capacity in models of mitochondrial myopathy⁷⁵. Our findings hold significant clinical significance, especially given the limited efficacy of current treatments for MADD, such as riboflavin supplementation and dietary fat restriction⁴². However, clinical translation should be variant-informed and individualized. Late onset MADD commonly presents with exercise intolerance which improves in many cases after riboflavin therapy^{13, 53, 54}. These observations support moderate, supervised physical activity as a complementary strategy once biochemical stabilization is achieved. We infer that moderate aerobic activity can amplify the same mitochondrial biogenesis/oxidative remodeling programs observed in our fly data, provided intensity and duration are carefully tailored to avoid decompensation. In contrast, ETF A/ETF B variants and ETFDH genotypes with null or severely destabilizing alleles are more often associated with severe neonatal or early onset disease with minimal residual activity, in which exercise is unlikely to be beneficial and may pose rhabdomyolysis risk^{14, 76}.

Conclusion

This study presents an innovative Drosophila as a model for investigating the pathogenesis and therapeutic modulation of LSMs, specifically MADD caused by ETFDH mutations in progressive manner. We introduce clinically relevant missense mutations (L127R, S296C, and L399F) into the conserved FAD and ubiquinone-binding domains of the fly Etf-QO gene. We successfully recapitulated hallmark clinical features of MADD, including progressive neuromuscular decline, ectopic lipid accumulation, cardiac dysfunction, and mitochondrial stress. Unlike previous models that rely on acute knockdown or pharmacological inhibition, our CRISPR-engineered lines carry stable, patient-relevant mutations, enabling longitudinal assessment of disease progression and gene-environment interactions. The age-dependent worsening of motor and cardiac phenotypes, coupled with lipid dysregulation and myofibrillar disorganization, mirrors the clinical trajectory of late-onset MADD and provides mechanistic insights into how mitochondrial dysfunction drives tissue degeneration.

A key translational finding of this work is the demonstration that moderate, daily exercise significantly improves muscle performance and cardiac function in Etf-QO mutants. These benefits were accompanied by reduced lipid accumulation and improved locomotor performance suggesting that exercise promotes mitochondrial remodeling and metabolic reprogramming. Furthermore, the observed reduction in oxidative stress markers highlights the potential of exercise to restore redox balance, positioning it as a promising non-pharmacological intervention for mitochondrial lipid disorders. Our findings not only validate the Drosophila model as a powerful platform for dissecting the molecular underpinnings of ETFDH-related myopathies but also open avenues for high-throughput screening of therapeutic compounds and genetic modifiers. This innovative Drosophila model demonstrates that even short bouts of moderate exercise can yield significant cardiac benefits, reinforcing its utility for studying gene-environment interactions and non-pharmacological interventions in mitochondrial disease. This model offers a unique opportunity to explore combinatorial interventions that integrate exercise, dietary modulation, and antioxidant therapies.

Limitations

While exercise improved locomotor and cardiac function and reduced lipid burden, the underlying molecular mechanisms such as AMPK activation, mitochondrial biogenesis, and antioxidant responses were inferred but not directly measured. Our findings demonstrate that a short daily negative geotaxis–based exercise improves LSM phenotypes in the Drosophila model; however, direct translation to human therapy is challenging. The 15 minute regimen represents a substantial proportion of the fly’s lifespan and metabolic capacity, whereas human exercise prescriptions must account for differences in body size, energy expenditure, and cardiovascular physiology. Therefore, these results should be interpreted as proof of concept for the potential benefits of moderate physical activity rather than a direct recommendation for duration or intensity in patients. Future work should include unbiased transcriptomics and lipidomics to define pathways and lipid species linked with LSM progression and exercise response, which is beyond the scope of this study. Studies in mammalian models and, ultimately, clinical trials are needed to establish optimal exercise parameters for ETFDH related myopathies. Additional efforts should test combinatorial strategies, including dietary modulation and circadian based interventions, to enhance translational relevance.

Availability of data and material:

All the data associated with the manuscript has been presented as source data.