PRDX5 Controls Energy Production and Gene Activity During Muscle Growth and Works with PRDX3 to Slow Muscle Aging

All mice were maintained on a C57BL/6 J background, and all animal studies were approved by the Institutional Animal Care and Use Committees at Seoul National University. Prdx3^−/− and Prdx5^−/− mice have been previously described [23, 24]. Prdx3^−/− and Prdx5^−/− mice were crossed to generate Prdx3^−/−; Prdx5^−/− mice. Conditional mitochondria reporter mice, which express GFP in mitochondria upon crossing with Cre mice [25], were kindly provided by Jeremy Nathans (Johns Hopkins University School of Medicine, MD). EIIA‐Cre mice, which ubiquitously express Cre recombinase, were crossed with the conditional mitochondria reporter mice to obtain mt‐GFP mice. Subsequently, Prdx5^−/− and Prdx3^−/−; Prdx5^−/− mice were crossed with mt‐GFP mice to generate Prdx5^−/−; mt‐GFP and Prdx3^−/−; Prdx5^−/−; mt‐GFP mice, respectively. Genotyping primers are listed in Table S1.

Maximal forelimb grip strength was measured using a Grip Strength Meter (Bioseb) according to the manufacturer's guidelines. Mice were allowed to grasp a metal bar using their forelimbs only and were pulled backward horizontally. The test was performed in five consecutive trials per session, and the highest recorded value was noted. Each mouse underwent three sessions, and the average of the three sessions was used for analysis. The running endurance test was conducted using a treadmill (JD‐A‐09, Jeungdo Bio & Plant), following a previously published protocol [26]. Briefly, the treadmill speed was increased by 2–3 m/min every 3 min, and the incline was raised by 5% at 6, 12 and 21 min after the start. Work was calculated as the product of force (mass × gravitational acceleration × sin [angle]) and speed.

Snake venom from Sepedon hemachatus (Sigma) was prepared as a 10‐μM solution and injected into the right gastrocnemius (GAS) and tibialis anterior (TA) muscles of 4‐week‐old mice under isoflurane anaesthesia. A total of 80 μL was administered into the GAS and 30 μL into the TA. Cross‐sections of the GAS muscle were prepared using a CM1860 cryostat (Leica Biosystems) and stained with either haematoxylin and eosin (H&E) or immunofluorescence. Immunofluorescence staining was performed using an anti–embryonic myosin heavy chain antibody (BF‐G6, DSHB; 1:20) and an anti–laminin antibody (L9393, Sigma; 1:500). H&E and immunofluorescence staining were conducted following the previous protocol [27]. Muscle regeneration was also assessed by analysing gene expression in regenerating TA muscles.

Dual‐energy x‐ray absorptiometry (DXA) scanning was performed using the InAlyzer scanner (Medikors), and analysis was conducted with InAlyzer software (Medikors) following the previously described method by our group [28].

Primary myoblasts were isolated using the preplating method. Briefly, muscle tissues (quadriceps, GAS, and TA) from adult mice were minced and digested in a mixture of Collagenase type IV (Gibco) and Dispase II (Gibco) for approximately 1 h on a rocker at 37°C. The cell suspension was then preplated on a plastic dish for 1 h, allowing fibroblasts to adhere while myoblasts remained in suspension. The supernatant containing myoblasts was transferred to a new dish for another 1‐h preplating step, which was repeated 5 times in total. After the final step, myoblasts were plated on a 0.1% gelatin‐coated dish. Myoblasts were expanded in F‐10 medium (Gibco) supplemented with 20% horse serum (Gibco), penicillin–streptomycin (Gibco) and 2.5 ng/mL fibroblast growth factor (FGF) basic (PeproTech). Differentiation into myotubes was induced by plating myoblasts onto glass‐bottom dishes or plates coated with 1 mg/mL Matrigel (Corning) in DMEM (Cytiva) containing 5% horse serum and penicillin–streptomycin. At 24 h of differentiation, myotubes were treated with either 0.5 mM H₂O₂ (Daejung) or 0.1 μM Oligomycin A (Sigma) to analyse nuclear clustering. Fixed myotubes were stained with Phalloidin probes (Invitrogen) to label actin and DAPI to visualize nuclei.

For live imaging, myoblasts were transfected with a Map 7 or Lifeact plasmid tagged with mCherry using FuGENE HD (Promega) or Lipofectamine 3000 (Invitrogen) transfection reagent to label microtubules or actin filaments, respectively. The pCMV‐Map 7 plasmid was obtained from the Korea Human Gene Bank, Medical Genomics Research center, KRIBB, Korea. The pCMV‐mCherry‐Lifeact plasmid was kindly provided by Jin Man Kim (Seoul National University, Korea). pCMV‐Rhot1 and pCMV‐Trak1 plasmids (Origene; MR209606 and MR209421, respectively) were transfected using Lipofectamine 3000 (Invitrogen) reagent. For siRNA transfection, Lipofectamine RNAiMAX (Invitrogen) transfection reagent was used. Predesigned siRNAs (Bioneer) were used for siCtrl (SN‐1002), siRhot1 (59 040‐1) and siPrdx5 (54 683‐1). The sequences for siTrak1 (Bioneer) are as follows: Sense: GAUGACACAGGUGACCACA; Antisense: UGUGGUCACCUGUGUCAUC. All transfections were performed according to the manufacturer's protocol.

Single myofibers were isolated from extensor digitorum longus (EDL) muscles by digesting the tissue in 2 mg/mL Collagenase type 1 (Sigma) solution for approximately 1 h at 37°C. Digested tissues were then gently triturated using a wide‐bore pipette to release individual myofibers. Fibre immunostaining was performed following a previously published protocol with slight modifications [29]. Anti‐PAX7 antibody (DSHB; 1:50) and conjugated alpha‐bungarotoxin (Invitrogen; 1:100) were used to label satellite cells and neuromuscular junctions, respectively.

Myoblasts, myotubes and muscle cross‐sections were imaged using the Zeiss LSM 800 confocal microscope (Zeiss) with or without Airyscan mode. Live imaging of myoblast differentiation was performed using Zeiss Axio Observer Z1 (Zeiss), Zeiss LSM 800 confocal microscope (Zeiss), Zeiss Elyra 7 with Lattice SIM (Zeiss), or Zeiss Lattice SIM 5 (Zeiss) and processed using the Zeiss Zen blue edition software (Zeiss). Tetramethylrhodamine (TMRM) and mitochondrial GFP intensities within defined regions of interest (Figure 4) were quantified using the Zeiss Zen blue edition software (Zeiss). TMRM signal intensity was normalized to the corresponding mitochondrial GFP signal. Mitochondrial GFP intensity profiles were generated using the Fiji software.

Mitochondrial membrane potential in myotubes or muscle cross‐sections was assessed after staining with 100 nM TMRM (Invitrogen) for 30 min at 37°C, following the manufacturer's protocol. Mitochondrial oxygen consumption rate (OCR) was measured using the Seahorse XFe96 Analyzer (Agilent) following the manufacturer's protocol and our previously described method [30]. After basal readings were recorded, 1.5 μM oligomycin, 0.5 μM FCCP and 0.5 μM rotenone/antimycin A were sequentially injected. All reagents were provided in the Cell Mito Stress Test Kit for XFe/XF Analyzers (Agilent). Seahorse data were normalized to protein content, which was measured using the Bradford assay. H₂O₂ production by isolated mitochondria was assessed using the Amplex Red Hydrogen Peroxide/Peroxidase Assay Kit (Invitrogen), following the manufacturer's instructions. Briefly, GAS tissues were minced and treated with 0.3 mg/mL Nagarse (Sigma) for 3 min on ice, followed by centrifugation at 700 × g for 10 min, repeated twice to remove debris. Mitochondria were pelleted by centrifugation at 8000 × g for 10 min. Isolated mitochondria (30 μg) were incubated with Amplex Red reagent and HRP solution at room temperature, and fluorescence was measured using a Spark 10 M microplate reader (Tecan).

Total RNA was extracted using the QIAzol Lysis Reagent (Qiagen) and the AccuPrep Universal RNA Extraction Kit (Bioneer). cDNA was synthesized using the PrimeScript RT Reagent Kit (Takara) according to the manufacturer's guidelines. Quantitative reverse transcriptase polymerase chain reaction (qRT‐PCR) was performed using the TB Green Premix Ex Taq II (Takara) on a QuantStudio 3 Real‐Time PCR system (Applied Biosystems). Relative gene expression was determined using the 2^{(−∆∆Ct)} method, normalized to 18S rRNA. For the evaluation of mitochondrial DNA (mtDNA) content, qRT‐PCR was performed on purified DNA using primers for Cytb and Rplp0 (36B4). A complete list of primers used in this study is provided in Table S2.

All values were presented as mean ± standard error of the mean (SEM). Statistical analysis for comparisons involving more than 2 groups was performed using one‐way ANOVA followed by Tukey's post hoc test, while comparisons between 2 groups were analysed using Student's t test. A p value of less than 0.05 was considered statistically significant. All statistical analyses were conducted using the SPSS Statistics 26 (IBM) software.

To analyse nuclear and mitochondrial positioning during myotube differentiation, we isolated primary myoblasts from mitochondrial reporter (mt‐GFP) mice, which express green fluorescent protein (GFP) in the mitochondrial matrix, and induced differentiation into myotubes. Myoblast fusion was observed as early as 6 h into differentiation, with notable nuclear centration occurring by 24 h (Figure 1A). By 48 h, elongated myotubes with evenly distributed nuclei and mitochondria were observed (Figure 1A and Movie 1). At 24 h, when nuclei and mitochondria began to actively disperse along the myotube, the expression of Mrf4 and Myh4, mid‐ to late‐stage myogenic markers, sharply increased, whereas the expression of Myh3, an early myogenic marker, decreased (Figure 1B). A similar expression pattern of myogenic markers was detected in regenerating muscle in vivo following snake venom‐induced injury to the TA muscle (Figure 1C). Mitochondrial OCR significantly increased during myogenesis, as indicated by elevated basal and maximal respiration, ATP production, and spare respiratory capacity at 48 h compared to 0 h of myogenic induction (Figure 1D).

Next, we examined the expression patterns of Prdx genes in skeletal muscle tissue and in vitro myotubes. In the quadriceps muscle of 10‐week‐old mice, Prdx3 and Prdx5 were the two most abundantly expressed Prdx isoforms (Figure 2A). Furthermore, a comparison of Prdx gene expression levels in myoblasts, in vitro myotubes and TA muscle revealed that only Prdx3 and Prdx5 showed an increasing (Prdx3) or a stable (Prdx5) expression pattern as myoblasts differentiated into myotubes and matured into muscle tissue, whereas the expression of other Prdx genes significantly decreased (Figure 2B). A similar trend was observed during in vitro myogenesis, where only Prdx3 and Prdx5 displayed increasing or stable expression levels (Figure 2C). Among all Prdx genes, Prdx3 and Prdx5 were also the only ones to show increased expression during in vivo muscle regeneration induced by snake venom (Figure 2D). These findings suggest that Prdx3 and Prdx5 may have distinct and functionally significant roles in myogenesis and muscle regeneration.

To investigate the role of PRDX3 and PRDX5 in myogenesis, we isolated myoblasts from wild‐type mt‐GFP (WT; mt‐GFP), Prdx3^−/−; mt‐GFP, Prdx5^−/−; mt‐GFP and Prdx3^−/−; Prdx5^−/−; mt‐GFP mice and induced differentiation into myotubes. After 48 h of differentiation, Prdx5^−/−; mt‐GFP and Prdx3^−/−; Prdx5^−/−; mt‐GFP myotubes exhibited impaired nuclear and mitochondrial distribution, leading to aggregated nuclei and mitochondria, while WT; mt‐GFP and Prdx3^−/−; mt‐GFP myotubes displayed an even distribution of nuclei and mitochondria (Figure 3A, Movies 2, 3). Quantification of myotubes containing clustered nuclei revealed that Prdx5^−/− myotubes exhibited more than twice the percentage of clustered nuclei compared to WT myotubes (Figure 3B). Notably, nuclear clustering was not further increased in Prdx3^−/− or Prdx3^−/−; Prdx5^−/− myotubes compared to WT and Prdx5^−/− myotubes, respectively, indicating that the impaired nuclear distribution is specifically caused by Prdx5 deficiency rather than Prdx3 deficiency (Figure 3B). Next, we investigated whether myonuclear distribution is impaired in mature myofibers of Prdx5^−/− mice in vivo (Figure 3C). Myofibers were isolated from EDL muscles of 48‐week‐old mice and stained for neuromuscular junctions (NMJs) and satellite cells. Since myonuclei are naturally clustered in NMJs [31], we avoided these regions for analysis. Additionally, we separately labelled satellite cells using a PAX7 antibody to effectively distinguish them from myonuclei. Analysis of myofibers revealed a dramatic, 4‐fold increase in the percentage of myofibers with clustered myonuclei in Prdx5^−/− myofibers (Figure 3D), indicating that Prdx5 deficiency disrupts myonuclear distribution both during in vitro myogenesis and in mature myofibers in vivo.

Because PRDXs are well‐known H₂O₂ scavengers and have been reported to prevent ROS accumulation [2, 32], we treated developing myotubes with H₂O₂ to assess its effect on nuclear clustering. Treatment with 0.5 mM H₂O₂ at 24 h of myogenesis significantly increased nuclear clustering in WT myotubes at 48 h, reaching a level comparable to that observed in Prdx5^−/− myotubes (Figure 3E, F). However, H₂O₂ treatment further exacerbated nuclear clustering in Prdx5^−/− myotubes (Figure 3E, F), suggesting that although H₂O₂ accumulation may contribute to the myonuclear clustering phenotype in Prdx5^−/− myotubes, it is not the sole cause.

We next examined whether Prdx5 deficiency affects mitochondrial function in myotubes and muscle tissue. To assess mitochondrial function, we used TMRM dye to measure mitochondrial membrane potential. At 48 h of differentiation, mitochondrial membrane potential, indicated by TMRM fluorescence intensity normalized to endogenous mitochondrial GFP intensity, was significantly reduced in Prdx5^−/−; mt‐GFP and Prdx3^−/−; Prdx5^−/−; mt‐GFP myotubes compared to WT; mt‐GFP myotubes (Figure 4A, B). Furthermore, mitochondrial membrane potential was significantly lower in both EDL and soleus muscle cross‐sections of 48‐week‐old Prdx5^−/−; mt‐GFP mice compared to WT; mt‐GFP mice (Figure 4C, D). Notably, endogenous mitochondrial GFP signal appeared reduced in muscle cross‐sections of Prdx5^−/−; mt‐GFP mice, implying a decrease in mitochondrial content (Figure 4D). Consistently, quantification of the mitochondrial DNA‐to‐nuclear DNA ratio revealed a reduction in mitochondrial quantity in both the EDL and soleus muscles of Prdx5^−/− mice compared to WT mice (Figure 4E). We also measured mitochondrial OCR in differentiated myotubes, which revealed a significant reduction in ATP production in Prdx3^−/− and Prdx5^−/− myotubes (Figure 4F). Basal respiration, maximal respiration, ATP production and spare respiratory capacity were further reduced in Prdx3^−/−; Prdx5^−/− myotubes, indicating an additive effect of dual deficiency on mitochondrial function (Figure 4F). Additionally, staining myotubes with ATP‐Red dye, which specifically labels mitochondrial ATP, confirmed that mitochondrial ATP production was diminished in Prdx5‐deficient myotubes (Movie 4). Mitochondrial morphology was also abnormal in Prdx5^−/− and Prdx3^−/−; Prdx5^−/− myotubes, exhibiting a fragmented appearance (Movie 4). The expression of mitochondrial fusion markers, particularly Mfn2, showed a decreasing trend in Prdx5‐deficient myotubes, while mitochondrial fission markers remained unchanged (Figure S1).

To investigate whether mitochondrial dysfunction induces nuclear clustering in myotubes, we treated developing myotubes with Oligomycin A, an ATP synthase inhibitor. Treatment with 0.1 μM oligomycin A at 24 h of differentiation dramatically increased nuclear clustering by 48 h (Figure S2), suggesting that mitochondrial ATP production is essential for proper nuclear distribution and that mitochondrial dysfunction partially contributes to impaired nuclear spreading in Prdx5‐deficient myotubes.

Inside the cell, mitochondria are actively transported along microtubules by kinesin and dynein motor proteins [11]. In particular, kinesin motor proteins mediate the anterograde transport of mitochondria towards microtubule plus ends, a key process in mitochondrial distribution along myotubes. Mitochondria are linked to kinesin motor proteins via Miro1, a mitochondrial outer membrane protein, and Milton, an adaptor protein that connects kinesin to Miro1 (Figure 5A). During in vitro myotube differentiation, the expression of Rhot1, which encodes Miro1, significantly increased at 24 and 48 h compared to 0 h (Figure 5B). Similarly, the expression of Trak1, which encodes Milton, exhibited an increasing trend during in vitro myotube formation and was significantly elevated in regenerating muscle tissue in vivo at 7 and 14 days after injury (dpi) (Figure 5B). Notably, the expression of both Rhot1 and Trak1 was significantly downregulated in Prdx5^−/− and Prdx3^−/−; Prdx5^−/− myotubes at 48 h (Figure 5B). However, Rhot1 and Trak1 expression levels were not further reduced in Prdx3^−/−; Prdx5^−/− myotubes compared to Prdx5^−/− myotubes, suggesting that PRDX5 specifically regulates their expression levels. Furthermore, treatment with H₂O₂ did not decrease Rhot1 or Trak1 expression in myotubes, indicating that their downregulation in Prdx5‐deficient myotubes is unlikely to be an indirect effect of elevated H₂O₂ levels (Figure S3A). The expression of Kif5b, which encodes a kinesin subunit, was mildly downregulated in Prdx5^−/− myotubes (Figure S3B).

To determine whether mitochondrial transport plays a significant role in nuclear distribution during myotube formation, we performed siRNA‐mediated knockdown of Prdx5, Rhot1 or Trak1 in WT; mt‐GFP myoblasts simultaneously with myogenic induction (Figure 5C). After confirming successful gene knockdown (Figure S3C) and impaired mitochondrial distribution and transport along microtubules (Figure 5C and Movie 5), we assessed nuclear distribution in myotubes at 48 h, which revealed a significant increase in nuclear clustering upon Prdx5, Rhot1 or Trak1 downregulation (Figure 5D). Trak1 knockdown led to the most dramatic increase in myonuclear clustering, with approximately 90% of myotubes exhibiting clustered nuclei, while Rhot1 knockdown resulted in a similar level of nuclear clustering as Prdx5 knockdown (Figure 5D). Prdx5, Rhot1 or Trak1 knockdown also significantly reduced the expression of Mrf4, a myogenic maturation marker, with the most pronounced reduction observed upon Trak1 knockdown (Figure S3D). Conversely, overexpression of Rhot1 alone or in combination with Trak1 significantly attenuated nuclear clustering in Prdx5^−/− myotubes, indicating that restoring mitochondrial transport promotes proper nuclear distribution (Figure 5E and Figure S3E). Collectively, these findings suggest that mitochondrial transport via Miro1 and Milton is essential for normal myotube development and nuclear spreading and that PRDX5 facilitates this process by promoting the expression of Rhot1 and Trak1 (Figure 5F).

To investigate the impact of PRDX5 deficiency on muscle regeneration in vivo, we injected the GAS and TA muscles of WT or Prdx5^−/− mice with snake venom and analysed muscle tissues at 3, 7, 14 and 28 dpi (Figure 6A). Analysis of muscle regeneration marker expression revealed impaired muscle regeneration in Prdx5^−/− mice, with significantly decreased Mrf4 and Myh3 expression at 3 dpi and reduced Myh4 expression at 7 and 14 dpi (Figure 6B). The delayed maturation of regenerating myofibers in Prdx5^−/− muscles was further evidenced by a delayed reduction in Myh3, which encodes embryonic myosin heavy chain (eMHC), at 7 dpi (Figure 6B). Consistent with the gene expression data, histological analysis at 3 and 7 dpi demonstrated impaired muscle regeneration, as indicated by the defective formation of healthy, centrally nucleated regenerating myofibers (Figure 6C, D and Figure S4A, B). Notably, Prdx5^−/− mice exhibited regenerating myofibers with aggregated nuclei, similar to what was observed during in vitro myogenesis (Figure 6D). At 28 dpi, histology further revealed persistent defects in regeneration in Prdx5^−/− muscles, characterized by persistent regions of fibrosis and calcification (Figure S4C). To further examine regenerating myofibers in three dimensions, we performed whole‐mount staining of venom‐injured EDL muscles (Figure S4D). Unlike in WT mice, regenerating myofibers (eMHC‐positive) in Prdx5^−/− mice exhibited disrupted nuclear distribution and abnormal mitochondrial morphology, characterized by a fragmented or swollen appearance (Figure 6E). Gene expression analysis revealed that Rhot1 and Trak1 expression levels were significantly decreased in regenerating muscle tissues of Prdx5^−/− mice at 3, 7 and 28 dpi, indicating impaired mitochondrial transport (Figure 6F). These findings demonstrate that PRDX5 promotes muscle regeneration in vivo by facilitating proper mitochondrial transport and nuclear distribution.

We next focused on the analysis of Prdx3^−/−; Prdx5^−/− mice and found that, unlike Prdx5^−/− mice, they exhibited significant reductions in muscle mass throughout the body as early as 10 weeks of age (Figure 7A). This muscle mass loss was accompanied by an increase in Fbxo32 and Trim63 expression, which encode the muscle‐specific E3 ubiquitin ligases Atrogin1 and MURF1, respectively, in the soleus and EDL muscles of Prdx3^−/−; Prdx5^−/− mice at 10 weeks of age (Figure 7A). In addition to reduced muscle mass, grip strength and treadmill running performance were significantly decreased in Prdx3^−/−; Prdx5^−/− mice compared to WT or Prdx5^−/− mice at 9–11 weeks of age (Figure 7A). The muscle atrophy and functional reduction observed in Prdx3^−/−; Prdx5^−/− mice at a young age reflect a sarcopenic phenotype, suggesting that the combined deficiency of PRDX3 and PRDX5 accelerates muscle aging. While Prdx5^−/− mice did not exhibit muscle mass or strength loss at a young age, their running performance was significantly impaired at 47–50 weeks of age, reflecting mitochondrial dysfunction in muscle (Figure 7A). Fat mass was also significantly increased in Prdx5^−/− mice at 47–50 weeks of age (Figure S5).

To examine whether excessive H₂O₂ accumulation drives the muscle aging phenotype observed in Prdx3^−/−; Prdx5^−/− mice, we isolated live mitochondria from muscle tissues and performed Amplex Red assay to measure H₂O₂ production levels (Figure 7B). Compared to mitochondria from WT muscle, those from Prdx3^−/−; Prdx5^−/− muscle showed a significant increase in H₂O₂ production (Figure 7C). Furthermore, H₂O₂ treatment of myotubes significantly upregulated the expression of muscle proteolysis markers Fbxo32 and Trim63 (Figure 7D), demonstrating that excessive H₂O₂ accumulation contributes to the muscle aging phenotype in Prdx3^−/−; Prdx5^−/− mice. To investigate additional mechanisms driving Atrogin1 and MURF1 expression beyond elevated oxidative stress, we evaluated the expression of Ppargc1a (PGC‐1α), a known suppressor of Atrogin1 and MURF1 transcription [33] (Figure S6A). Ppargc1a expression was significantly reduced in Prdx3^−/−; Prdx5^−/− muscles, supporting the possibility that its downregulation contributes to the elevated expression of muscle proteolysis genes in the double knockout muscles. However, Ppargc1a expression was also significantly decreased in the soleus of Prdx5^−/− mice at 10 weeks of age, despite no significant increase in Atrogin1 or MuRF1 levels (Figure 7A and Figure S6A). This suggests that a threshold level of PGC‐1α may exist, below which the induction of these proteolytic genes is triggered.

Next, we analysed publicly available datasets to assess changes in Prdx gene expression during aging and under pathological muscle conditions. Among the Prdx genes, Prdx3 and Prdx5 were the two most dramatically downregulated genes in aged mouse soleus and aged rat GAS (Figure 7E). Notably, Prdx3 and Prdx5 were the only downregulated Prdx genes in aged rat GAS, whereas all other Prdx genes were upregulated at 22 and 24 months of age compared to 8 months of age (Figure 7E), highlighting a strong association between muscle aging and reduced PRDX3 and PRDX5 levels in rodent muscles. Prdx3 and Prdx5 levels also showed a decreasing trend in mouse models of Duchenne muscular dystrophy (DMD) and dysferlinopathy‐associated accelerated muscle degeneration (Figure S6B). Trak1 expression was also significantly reduced in aged mouse soleus and GAS, mirroring the expression pattern of Prdx5 (Figure S6C). In human vastus lateralis muscle, PRDX2 and PRDX5 expression levels were significantly decreased in sarcopenic individuals compared to young or age‐matched healthy individuals. Among them, PRDX5 showed the most significant decrease, highlighting its potential importance in human muscle aging (Figure 7E). Collectively, these findings suggest that PRDX3 and PRDX5 work together to prevent excessive H₂O₂ accumulation, thereby mitigating muscle proteolysis, muscle atrophy and strength loss, and ultimately preserving both muscle quantity and quality, and that their loss may lead to accelerated muscle aging (Figure 7F).