Neonatal muscle-derived extracellular vesicles containing miR-542-3p rejuvenate aged skeletal muscle via a functional microneedle patch

Abstract

Age-related skeletal muscle aging can lead to sarcopenia and is closely associated with cellular senescence and mitochondrial dysfunction. Neonatal mammalian muscle exhibits a strong regenerative capacity, and neonatal muscle extracellular vesicles (NMEVs) show therapeutic potential against skeletal muscle aging. In this study, we isolated NMEVs for the first time and found that they significantly alleviated palmitic acid (PA)-induced senescence, mitochondrial dysfunction, and lipid accumulation in C2C12 cells. in vivo, we developed a bilayer microneedle (MN) system loaded with NMEVs (NMEVs@PLGA@Fucoidan-HA MN) and applied it to aged mice. The MN effectively enhanced mitochondrial function, reduced muscle aging and fibrosis, and decreased lipid deposition. Mechanistically, miR-542-3p enriched in NMEVs directly targeted and downregulated Asxl2-PPARγ, leading to reduced lipid accumulation. At the same time, it suppressed Eef1a1 to activate the AMPK pathway, thereby improving mitochondrial function and attenuating cellular senescence. Our findings demonstrate the protective role of NMEVs delivered via an innovative MN system against muscle aging, where miR-542-3p plays a central role by concurrently targeting Eef1a1 and Asxl2 to mitigate senescence and lipid dysregulation. This study reveals a novel molecular mechanism underlying the anti-aging potential of NMEVs and offers a promising therapeutic strategy for skeletal muscle aging.

Graphical abstract

Graphic abstract. (A) Isolation and purification procedures for small extracellular vesicles (sEVs) derived from the muscle tissues of 2-day-old neonatal and adult mice. (B) Preparation of NMEVs@PLGA@Fucoidan-HA and its therapeutic application in aged mice. (C) miR-542-3p encapsulated within NMEVs exerts dual targeting effects: it directly abrogates Asxl2 to subsequently repress PPARγ-mediated lipid deposition, and concurrently suppresses Eef1a1 to activate the AMPK signaling pathway, ultimately reversing cellular and organismal aging.

Highlights

• •Neonatal muscle-derived extracellular vesicles (NMEVs) are superior to adult EVs in countering cellular senescence, mitochondrial dysfunction, and lipid accumulation.
• •A bilayer microneedle patch (NMEVs@PLGA@Fucoidan-HA MNs) allows sustained local delivery of NMEVs to alleviate skeletal muscle aging in vivo.
• •NMEV-loaded miR-542-3p rejuvenates aged skeletal muscle through dual targeting: inhibiting Eef1a1 to activate AMPK signaling and suppressing Asxl2 to downregulate PPARγ expression.
• •This study provides the first proof-of-concept that NMEVs combined with microneedle technology represents a promising therapeutic strategy for age-related skeletal muscle atrophy.

1. Introduction

Skeletal muscle aging is a key hallmark of physiological aging, primarily characterized by progressive muscle mass loss and reduced regenerative capacity. This condition is closely associated with frailty and impaired quality of life in the elderly [1]. Skeletal muscle aging is caused by multiple factors, including the depletion of endogenous muscle stem cells (satellite cells), adipogenic and fibrotic infiltration, neuromuscular junction degeneration, mitochondrial oxidative stress, and chronic inflammation [2,3]. Strength training has long been recommended for the prevention and management of sarcopenia and has been shown to improve muscle strength and physical function. However, the implementation of high-intensity or progressive resistance training in routine practice may be limited by feasibility and adherence issues, especially in older adults with comorbidities or functional limitations [4]. Nutritional interventions (e.g., increased protein intake and specific supplements) serve as important adjuvant strategies, but their efficacy is often influenced by baseline nutritional status and variations in intervention protocols. Moreover, nutritional supplementation alone exerts only limited effects on muscle strength and functional recovery in patients [5,6]. Currently, pharmacological therapies for sarcopenia remain controversial and have been associated with various risks. For instance, selective androgen receptor modulators have been linked to potential risks including prostate-related adverse events, polycythemia, and debated cardiovascular risks [7]. In this context, extracellular vesicles (EVs) have emerged as a promising cell-free therapeutic strategy. EVs are critical natural mediators of intercellular communication. They encapsulate a variety of bioactive molecules (e.g., microRNAs, proteins, lipids) to mediate the transfer of functional signals between cells [8]. Compared to traditional therapies, EVs offer distinct advantages including excellent biocompatibility, low immunogenicity, ability to cross biological barriers, and capacity to simultaneously deliver multiple bioactive cargoes that can target various pathological processes [9].

EVs play an increasingly important role in regulating musculoskeletal system homeostasis [10]. Notably, EVs derived from young or neonatal tissues demonstrate superior therapeutic efficacy. For instance, EVs from young plasma can counteract osteoporosis by stimulating the osteogenic differentiation of mesenchymal stem cells [11]. Neonatal tissue-derived EVs show enhanced repair outcomes: neonatal kidney-derived EVs better ameliorate kidney injury, neonatal skin-derived EVs more effectively heal skin defects, and neonatal heart-derived EVs improve post-myocardial infarction outcomes through anti-inflammatory and pro-angiogenic effects [[12], [13], [14]]. Neonatal mammalian skeletal muscle possesses a remarkable regenerative capacity that declines with age, suggesting the presence of unique regenerative factors [15]. We therefore hypothesized that neonatal muscle-derived EVs (NMEVs) may hold therapeutic potential against skeletal muscle aging. However, the efficacy and underlying mechanisms of NMEVs in this context remain unexplored and constitute the focus of this study.

A major obstacle in EVs-based therapy is inefficient delivery. Free EVs are rapidly cleared, exhibit poor targeting to deep tissues such as skeletal muscle, and readily lose bioactivity [16]. To overcome this, various advanced delivery systems have been explored. For instance, gold nanoparticles have emerged as a promising platform for stable and efficient miRNA delivery, owing to their favorable biocompatibility and versatile surface functionalization [17]. Microneedles (MNs), particularly dissolvable ones, have also gained attention as effective tools for facilitating nanocarrier delivery by enhancing tissue penetration and retention [18,19]. Furthermore, recent innovations in MN design, such as motion-driven microneedles capable of penetrating hypertrophic scar tissue, have expanded the therapeutic potential of this platform [20]. Recent studies have employed multifunctional MNs loaded with small molecules or EVs in chronic tendon injury models. For instance, PLGA MNs loaded with diacerein nanoparticles reduced oxidative stress and promoted stem cell differentiation during tendon-bone repair [21]. However, the combination of NMEVs with dissolvable MNs for treating skeletal muscle aging has not been reported and merits further investigation.

The pathological basis of age-related skeletal muscle aging includes mitochondrial dysfunction, cellular senescence, aberrant lipid deposition, tissue fibrosis, and stem cell exhaustion [22]. Emerging evidence highlights the critical role of microRNAs (miRNAs) encapsulated in EVs as key mediators of intercellular communication and tissue homeostasis [23]. In this study, we identified miR-542-3p as a highly enriched miRNA in NMEVs through comparative sequencing. Mechanistically, miR-542-3p exerts a dual-targeting effect: it suppresses Eef1a1 to activate the AMPK signaling pathway, thereby improving mitochondrial function and attenuating cellular senescence; concurrently, it inhibits Asxl2 to downregulate PPARγ, reducing pathological lipid accumulation. This dual regulatory mechanism addresses multiple facets of muscle aging simultaneously.

Here, we developed a bilayer microneedle (MN) patch for the sustained local delivery of NMEVs. In vivo application of this system effectively alleviated skeletal muscle aging in aged mice by enhancing mitochondrial function, reducing fibrosis, and decreasing lipid deposition. Our findings reveal a novel molecular mechanism by which NMEVs counteract muscle aging and present a promising therapeutic strategy for age-related skeletal muscle aging.

2. Materials and methods

2.1. Animals

All animals (neonatal mice, 2 days old; young mice, 3 months old; old mice, 20 months old) were housed in the specific pathogen-free facility of the Experimental Animal Department at Xiangya Hospital, Central South University. They were maintained under standard conditions with a 12h light/dark cycle and had ad libitum access to food and water. This study was conducted in accordance with the guidelines for animal care and was approved by the Animal Ethics Committee of Xiangya Hospital, Central South University.

2.2. Muscle extracellular versicles (MEVs) extraction and identification

Following cervical dislocation of the mice, the gastrocnemius muscles were rapidly dissected. The tissues were rinsed three times with ice-cold PBS (pH 7.4, Thermo Fisher, USA) to remove residual blood and fascia and were then minced into approximately 1 mm³ pieces. The minced tissue was homogenized in a 10-fold volume (w/v) of PBS containing protease inhibitors (Millipore, USA) using a tissue homogenizer on an ice bath. The resulting homogeneous muscle homogenate was subsequently subjected to a series of differential centrifugations: first at 300g for 10 min (min) at 4°C to remove large tissue fragments and cell clusters, then at 2,000g for 20 min at 4°C to pellet cellular debris and nuclei, and finally at 10,000g for 30 min at 4°C to remove large organelles such as mitochondria and lysosomes. The obtained supernatant was then filtered through a 0.22 μm filter (Millipore, USA). The filtrate was transferred to ultracentrifuge tubes and centrifuged at 120,000g for 2 h at 4°C using an Optima XPN-100 Ultracentrifuge (Beckman Coulter, Brea, CA, USA). After discarding the supernatant, the pelleted EVs were resuspended in PBS. For the characterization of the isolated EVs, nanoparticle tracking analysis (NTA) by ZetaView PMX-120 (Particle Metrix, Meerbusch, Germany) was employed to determine their size distribution and concentration. Western blot analysis was performed to detect the expression of EV-positive markers (TSG101, CD9, CD81) and the negative marker Calnexin. The antibodies used are listed in Table S1. Furthermore, the morphology and diameter of the EVs were examined using a transmission electron microscopy (TEM) (JEM-1400 Flash JEOL, Tokyo, Japan).

2.3. Cell culture and treatment

The C2C12 cell line was procured from Procell Life Science & Technology Co., Ltd (Wuhan, China). Cells were maintained in high-glucose Dulbecco's Modified Eagle Medium (DMEM, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% penicillin-streptomycin (Thermo Fisher, USA). Palmitic acid (PA, MCE, USA) was used to induce cellular senescence in C2C12 cells. PA is a well-established inducer of cellular senescence and has been widely used to mimic lipotoxic stress–induced aging phenotypes [24,25]. To determine the optimal concentration, C2C12 cells were treated with different concentrations of PA (0, 50, 100, 200, and 400 μM). Cell viability was evaluated using the CCK-8 assay, and senescence was assessed by measuring p21 protein expression via Western blot analysis.

To determine the optimal concentration of NMEVs for subsequent experiments, C2C12 cells were treated with palmitic acid (PA, 200 μM) to induce cellular senescence. Concurrently, cells were incubated with increasing concentrations of NMEVs (0, 50, 100, and 200 μg/mL) for 24 h. The expression of the senescence marker p21 was analyzed by Western blotting. For EVs treatment, C2C12 cells were incubated with EVs at a final concentration of 100 μg/mL for 24 h prior to subsequent assays. For the in vitro EVs uptake experiment, EVs were labeled using the PKH67 Kit (Sigma-Aldrich, USA) following the manufacturer's protocol. Briefly, C2C12 cells were seeded in 24-well plates at a density of 100,000 cells per well and then incubated with 100 μg of PKH67-labeled EVs for 24 h. Internalization of the labeled EVs was subsequently visualized by immunofluorescence microscopy. The miR-542-3p mimic, miR-542-3p inhibitor, and plasmids constructs for Asxl2 and Eef1a1 overexpression were obtained from Hanbio Biotechnology Co., Ltd (Shanghai, China). Transfections were performed according to the manufacturer's instructions. The working concentrations were 50 nM for the miR-542-3p mimic, 100 nM for the miR-542-3p inhibitor, and 500 ng/mL for all plasmids. All subsequent experiments were conducted 24 h post-transfection.

2.4. Reactive oxygen species (ROS) assay

Intracellular ROS levels were measured using DCFH-DA (Beyotime Biotechnology, China). Briefly, DCFH-DA was diluted in Hank's Balanced Salt Solution (HBSS, Thermo Fisher, USA) at a ratio of 1:1000. After trypsinization and centrifugation, the cells were resuspended in 1 mL of DCFH-DA/HBSS solution per well and incubated at 37°C for 20 min. The cells were then washed twice with cold PBS, followed by centrifugation at 1200 rpm for 5 min each time to remove the supernatant. DCF fluorescence was detected using a BD FACSAria III flow cytometer (BD Biosciences, San Jose, CA, USA), and data were analyzed with FlowJo software (TreeStar, USA).

2.5. Quantitative reverse transcription polymerase chain reaction (qRT-PCR)

For total RNA extraction: After removing medium and washing with PBS, cells were lysed directly in TRIzol (Invitrogen, USA). Chloroform (1:5 ratio to TRIzol) was added, and the mixture was vortexed vigorously, incubated at room temperature for phase separation, then centrifuged at 12,000g for 15 min at 4°C. The upper aqueous phase was mixed with equal-volume isopropanol, incubated at room temperature, and centrifuged at 12,000g for 10 min at 4°C to pellet RNA. The pellet was washed with 75% ethanol, dissolved in RNase-free water, and concentration was measured via NanoDrop One/OneC Microvolume UV-Vis Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). cDNA was synthesized using a Reverse Transcription System (Promega, USA), and real-time qPCR was performed with GoTaq® qPCR Master Mix (Promega, USA). Target gene expression was analyzed by the 2^(-ΔΔCt) method, with GAPDH as the internal reference. U6 was used as the internal reference gene for microRNA expression detection.

For mtDNA extraction: Cells were digested with pre-cold lysis buffer, followed by differential centrifugation: 1,000g for 5 min at 4°C (supernatant collected), then 12,000g for 10 min at 4°C to pellet crude mitochondria. The pellet was resuspended and re-centrifuged at 1,000g to remove residual nuclei, then at 12,000g to obtain purified mitochondria. Mitochondria were treated with DNase I (Thermo Fisher, USA) at 37°C to eliminate nuclear DNA contamination, and the reaction was stopped by centrifugation. The pellet was lysed, mixed with protein precipitation solution, and centrifuged to further remove nuclear DNA. The supernatant underwent chloroform extraction; the aqueous phase was mixed with isopropanol and a co-precipitant, then centrifuged. The DNA pellet was washed with 70% ethanol, dissolved in nucleic acid buffer, and purified mtDNA was obtained. Subsequent reverse transcription and PCR steps followed the above protocol, with β2-MG as the internal reference. Primer sequences are listed in Table S2.

2.6. Western blot

MEVs and C2C12 cells were lysed using RIPA buffer (Beyotime Biotechnology, China), followed by centrifugation at 12,000g for 20 min at 4°C to collect the protein supernatant. Protein concentration was determined using a BCA protein assay kit (Thermo Fisher, USA), and samples were normalized to a uniform concentration of 2 μg/μL. Subsequently, 5X loading buffer (Beyotime Biotechnology, China) was added, and the proteins were denatured by heating at 95°C for 5 min. Equal amounts of protein (10 μL per lane) were loaded into the wells of an SDS-polyacrylamide gel. Electrophoresis was performed at a constant voltage of 120 V to separate the proteins. The proteins were then transferred from the gel onto a PVDF membrane (Millipore, USA) via wet electroblotting at 300 mA for 90 min. The PVDF membrane was blocked with 5% non-fat milk in Tris Buffered Saline with Tween 20 (TBST) (Millipore, USA) buffer for 2 h at room temperature and then incubated with respective primary antibodies overnight at 4°C. After removing the primary antibody solution, the membrane was washed three times with TBST (5 min per wash). Then incubate the membranes with corresponding HRP-conjugated secondary antibodies for 1.5 h at room temperature, followed by TBST buffer washing for three times. Immunoreactive bands were visualized using an enhanced chemiluminescence (ECL) reagent (CST, USA), and band intensities were quantified with ImageJ software by normalizing the signal of the target protein to that of internal control. Information for all antibodies used was provided in Table S1.

2.7. Immunofluorescent staining

After washing frozen muscle sections or C2C12 cell coverslips with PBS, the samples were fixed 4% paraformaldehyde (PFA, Beyotime Biotechnology, China) for 10 min. After three PBS washes, the samples were permeabilized with 0.1% Triton X-100 in PBS for 15 min, followed by blocking with 3% bovine serum albumin (BSA, Thermo Fisher, USA) for 30 min. Subsequently, the samples were incubated with appropriately diluted primary antibodies overnight at 4°C. The following day, the primary antibody solution was removed, and the samples were washed three times with PBST (0.1% Tween-20). Then, cover the sections or coverslips with the diluted secondary antibody and incubate at room temperature for 1.5 h. After washing with PBST for three times, cell nuclei were stained with 4′,6-Diamidino-2′-phenylindole (DAPI, Vector Laboratories, USA). Finally, the samples were imaged using a Zeiss Apotome 3.0 fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany). All antibodies used were listed in Table S1.

2.8. Transmission electron microscope (TEM)

For cell samples, the culture medium was discarded and replaced with pre-cooled 2.5% glutaraldehyde fixative (MCE, USA). After sample fixation overnight at 4°C, cells were scraped from the plate using a cell scraper. The cell suspension containing fixative was transferred to a centrifuge tube and centrifuged at 1500 rpm for 5 min. The supernatant was then removed, and the pellet was fixed again with 2.5% glutaraldehyde. For muscle tissue samples, the tissue was first trimmed into small pieces of approximately 1 mm³. These pieces were then fixed with pre-cooled 2.5% glutaraldehyde. Subsequent steps for both sample types included post-fixation with 1% osmium tetroxide for 1 h at room temperature, washing with ultrapure water, and gradual dehydration through a graded ethanol series (30%, 50%, 70%, 90%, and 100% ethanol, 10 min per step) to remove water progressively while minimizing osmotic shock and preserving ultrastructural integrity. The samples were then transferred to absolute ethanol twice for complete dehydration before resin infiltration. The samples were then embedded in resin. The embedded samples were placed in an oven to allow the resin to polymerize completely and harden. Ultrathin sections with a thickness of 50–70 nm were cut using an ultramicrotome. The sections were mounted on copper grids and stained sequentially with uranyl acetate (20 min for nucleic acid and membrane contrast) and lead citrate (10 min for protein and cytoplasmic contrast). Finally, the stained sections were examined and imaged under a transmission electron microscope (JEM1400, Japan).

2.9. EdU staining

C2C12 cell proliferation was assessed using the EdU-594 Detection Kit (Elabscience, China). C2C12 cells in the logarithmic growth phase were seeded into 24-well plates and cultured until approximately 70% confluent. The EdU working solution was diluted to 10 μM using serum-free medium. The old medium was removed, and the cells were incubated with the EdU working solution in serum-free medium for 2 h in a cell culture incubator. After discarding the EdU-containing medium, the cells were fixed with 4% PFA for 15 min. Following washing with pre-cold PBS, the cells were permeabilized with 0.3% Triton X-100 and blocked with 5% BSA solution. Subsequently, the cells were incubated with the prepared Click Additive Solution at room temperature for 30 min in the dark. The Click reaction solution was then removed, and the cells were washed three times with washing buffer (3–5 min each). Nuclei were counterstained with DAPI. Finally, images were acquired using a Zeiss Apotome 3.0 fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany).

2.10. BODIPY staining

For C2C12 cells, serum-free cell culture medium was used to prepare a 10 μM BODIPY 493/503 dye (MCE, USA) working solution. C2C12 cells were cultured on glass coverslips until they reached 60% confluence. The coverslips were then removed from the culture dish and washed three times with PBS (5 min per wash). Subsequently, the cells were covered with 100 μL of the BODIPY dye working solution and incubated at room temperature for 30 min. After incubation, the cells were washed three times with PBS and stained with DAPI for nuclei visualization.

For muscle tissue sections, a 5 μM BODIPY 493/503 dye working solution was prepared in PBS. Frozen sections were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 solution. The sections were then covered with the BODIPY working solution and incubated at room temperature for 30 min. After washing with PBS, the nuclei were stained with DAPI. All samples were observed and photographed using a Zeiss Apotome 3.0 fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany).

2.11. Oil Red O staining

Lipid droplets in C2C12 cells were stained using the Oil Red O staining kit (Beyotime Biotechnology, China). Briefly, the culture medium was removed, and the cells were washed once with PBS. After fixation with 4% paraformaldehyde for 10 min, the cells were washed twice with PBS (3 min per wash). The cells were then covered with staining wash solution for 20 s (s), which was subsequently removed. An appropriate amount of Oil Red O working solution was added to cover the cells, followed by incubation at room temperature for 15 min. After removal of the staining solution, the cells were washed with PBS. Nuclei were counterstained with DAPI, and images were acquired using a Zeiss Apotome 3.0 fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany).

2.12. Oxygen consumption rate (OCR)

Oxygen Consumption Rate Fluorometric Assay Kit (Elabscience, China) was used to detect OCR. C2C12 cells in the logarithmic growth phase (following respective treatments) were first adjusted to a concentration of 5 × 10⁵ cells/mL. A 100 μL cell suspension was seeded into each well of the assay plate, with blank wells (containing no cells) included for background correction. The plate was incubated at 37°C under 5% CO₂ for 24 h to allow cell attachment and stabilization. After removal of the culture medium, each well was treated with 100 μL of probe working solution (prepared in SFM as per kit instructions). The plate was then returned to the incubator for 30 min to facilitate probe uptake and reaction. Subsequently, two drops of sealing solution were added to each well. Fluorescence intensity was measured using a microplate reader (Thermo Fisher, Waltham, USA) with an excitation wavelength of 405 nm and an emission wavelength of 675 nm. Fluorescence intensity was detected every 10 min for a total monitoring duration of 2 h, and the changes in fluorescence intensity over time were recorded. For analysis, fluorescence intensity was plotted against time, and the linear phase of the resulting kinetic curve was identified. The OCR of C2C12 cells in each group was evaluated based on the slope of this linear segment, with a steeper slope indicating a higher OCR.

2.13. Complex V/ATP detection

To assess cellular oxidative phosphorylation and ATP production capacity, the Mitochondrial Respiratory Chain Complex V Detection Kit (Solarbio, China) and the ATP Assay Kit (Beyotime Biotechnology, China) were employed. Briefly, C2C12 cells cultured in dishes were detached using trypsin, and the digestion was stopped by adding serum-containing medium. The cell suspension was collected and centrifuged at 1000 g for 5 min at 4°C. After discarding the supernatant, the cell pellet was resuspended in ice-cold PBS. For ATP concentration measurement, the cell pellet was lysed in pre-chilled cell lysis buffer, incubated on ice for 10 min, and then centrifuged at 12,000 g for 10 min at 4°C. The resulting supernatant was collected, and ATP levels were determined according to the manufacturer's instructions. For Complex V activity analysis, the cell pellet was treated with ice-cold mitochondrial extraction buffer, incubated on ice for 15 min, and centrifuged at 600g for 10 min at 4°C. The supernatant was further centrifuged at 12,000g for 20 min at 4°C. The resulting mitochondrial pellet was collected, and Complex V activity was measured as specified by the manufacturer's instructions. Finally, a microplate reader (Thermo Fisher, Waltham, USA) was used to detect both Complex V activity and ATP concentration in the samples.

2.14. Fabrication and characterization of double-layer microneedles

The bilayer microneedles (MNs) were fabricated using a polydimethylsiloxane (PDMS) mold (Henan Weina Pentium Biotechnology Co., Ltd., Xinxiang, Henan, China). The mold contained a 20×20 array of conical cavities, each with a depth (needle length) of 800 μm and a tip diameter of approximately 15 μm. The center-to-center spacing (pitch) between adjacent cavities was 550 μm. The dimensions of the mold were verified by scanning electron microscopy (SEM) prior to use. Prior to fabrication, the PDMS mold was sterilized by soaking in 75% ethanol for 1 h followed by 30 min of UV irradiation and then air-dried under aseptic conditions. For the outer layer, A 10% fucoidan (Sigma-aldrich, USA) solution was prepared in deionized water and homogenously mixed with the fluorescent dye DiL (Beyotime Biotechnology, China). This DiL-fucoidan solution was poured into the PDMS mold. A matching male mold, with precisely spaced stainless-steel spacers controlling the outer layer thickness, was pressed into the filled female mold. The assembly was then vacuum-degassed and centrifuged to ensure complete filling of the needle cavities. The outer layer was allowed to dry and solidify, after which the male mold was carefully removed. For the inner layer of the microneedles, poly (lactic-co-glycolic acid) (PLGA, 75:25 PLA: PGA) was dissolved in dimethyl sulfoxide (Millipore, USA). Isolated NMEVs (1 μg) were fluorescently labeled with DiO (Beyotime Biotechnology, China) and uniformly mixed into the PLGA solution (10 μl, final concentration: 100 μg/mL) to form the PLGA-DiO-EVs mixture. This mixture was then carefully pipetted into the mold cavities atop the pre-formed outer layer. The mold was again subjected to vacuum degassing and centrifugation to settle the mixture into the needle tips and remove excess solution. Subsequently, a 25% hyaluronic acid solution (MCE, New Jersey, USA) was pipetted into the mold to form the backing layer. The entire MN structure was then dried at 37°C under vacuum for 5 h to ensure complete evaporation of DMSO and minimize any potential solvent residue, while also allowing the PLGA to fully solidify and form the inner core of the needles. The resulting bilayer MNs were gently peeled from the mold using sterile tweezers and stored sealed at 4°C.

The macroscopic and microscopic structures of the MNs were examined using a Nikon camera and scanning electron microscopy (SEM, ZEISS, Oberkochen, Germany). To further evaluate the mechanical strength of the MNs, axial compression tests were performed using a Universal Testing System (Instron, Norwood, MA, USA). The microneedle was securely fixed, and the stage was adjusted in height and levelness to ensure perfect alignment between the needle tip axis and the indenter movement axis. The indenter was then lowered at a slow pre-contact speed of 0.005 mm/s. Contact between the indenter and the needle tip was determined when the axial force reached a range of 0.001 to 0.002 N, as monitored in real-time by a force sensor. Subsequently, axial compression was applied at a speed of 1.1 mm/s, with data on pressure, axial deformation, and testing time recorded every 0.04 s until needle fracture occurred.

To examine the bilayer structure, fluorescence-labeled microneedles were sectioned longitudinally (50 μm thickness) using a blade. The sections were placed on glass slides, and the spatial distribution and structural integrity of the DiL (red) and DiO (green) signals were visualized using LSM 880 Confocal Laser Scanning Microscope (ZEISS, Oberkochen, Germany).

To determine the degradation profile, microneedle patches were first weighed to obtain the initial dry weight (W₀). The samples were then incubated in artificial tissue fluid at 37°C in a constant-temperature shaker (60 rpm). At predetermined time points (3, 7, 14, and 28 days), the microneedles were collected, gently rinsed with deionized water to remove residual salts, and blotted with filter paper to remove excess surface liquid. The samples were subsequently freeze-dried to constant weight to eliminate absorbed moisture before weighing (W_t). The degradation rate was calculated according to the following equation: Degradation rate (%) =(W₀−W_t)/W₀×100.

To evaluate in vivo insertion of MN into muscle tissue, mice were anesthetized, and the microneedles were vertically inserted into the gastrocnemius muscle tissue to a depth of 800 μm. After 10 min, the microneedles were withdrawn. The muscle tissue from the insertion site was then harvested, fixed, embedded in paraffin, and sectioned (10 μm thickness). Hematoxylin and eosin (H&E) staining (Solaibao, China) was performed to examine the morphology of the microchannels formed by the microneedles in the tissue.

2.15. Characterization of NMEVs release and bioactivity

For EV release studies, PLGA hydrogels loaded with 50 μg of NMEVs were incubated in 800 μL of PBS at 37°C with gentle shaking. At predetermined time points (1, 3, 7, 14, and 28 days), the entire supernatant was collected and replenished with an equal volume of fresh PBS. Collected samples were concentrated via ultracentrifugation (120,000g, 2h, 4°C), and EV protein concentration was determined using a BCA assay (Thermo Fisher, USA). The cumulative release percentage was calculated as follows: Cumulative Release (%) = ∑[(C_n−Cctrl_n)×V]/M×100 (where C_n = experimental EVs concentration at each time point, Cctrl_n = blank control concentration, V = 800 μL, M = initial 50 μg NMEVs). To evaluate EV integrity post-release, EVs collected at day 3 were examined by transmission electron microscopy (TEM). For bioactivity assessment, released EVs were labeled with PKH67 (Sigma-Aldrich, USA) according to the manufacturer's instructions and incubated with C2C12 cells for 24h. Cellular uptake was visualized using a Zeiss Apotome 3.0 fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany).

2.16. In vivo imaging for EVs tracing

The purified neonatal mouse muscle extracellular vesicles (NMEVs) suspension was first labeled with the DiR fluorescent dye (MCE, USA). Subsequently, the labeled DiR-NMEVs were used for the fabrication of microneedles (denoted as DiR-NMEVs@PLGA@Fucoidan-HA). The prepared microneedles were inserted into the right gastrocnemius muscle region of mice. In vivo fluorescence imaging was performed using a Xenogen IVIS Imaging System (Caliper Life Sciences, Hopkinton, USA) at the following time points: before treatment, and at 1 day, 7 days, 14 days, 28 days, and 56 days post-treatment.

2.17. Masson staining

Masson's Trichrome Staining Kit (Beyotime Biotechnology, China) was used to detect muscle tissue fibrosis. The muscle tissue was fixed, dehydrated, and embedded in paraffin. Sections were cut at a thickness of 6 μm. The paraffin sections were deparaffinized in xylene and rehydrated through a graded ethanol series. The staining procedure was performed as follows: nuclei were stained with hematoxylin for 5 min, followed by staining with Ponceau-Acid Fuchsin solution for 10 min. After washing, the sections were treated with Light Green staining solution for 1 min and rinsed with distilled water. Differentiation was then carried out using an acidic differentiation solution for approximately 1 min, after which the solution was removed. Finally, the stained sections were dehydrated through an ethanol series (70%, 90%, and 100% absolute ethanol), cleared in xylene, and mounted with coverslip.

2.18. Exosomal miRNA microarray assay

Total RNA was first isolated from both NMEVs and AMEVs using the PureLink™ RNA Mini Kit (Invitrogen, USA). The purity of the extracted RNA was assessed with a NanoDrop One/OneC Microvolume UV-Vis Spectrophotometer (Thermo Fisher, Waltham, MA, USA), while its concentration was accurately quantified using the Qubit™ RNA HS Assay Kit on a Life Invitrogen Qubit® 3.0 Fluorometer (Thermo Fisher, Waltham, MA, USA). RNA integrity was precisely determined with the Agilent 4200 TapeStation system (Agilent Technologies, Santa Clara, CA, USA). Subsequently, small RNA sequencing libraries were constructed. Briefly, specific adapters were ligated to the 3′ and 5’ ends of the small RNAs. The adapter-ligated RNA was then reverse-transcribed into complementary DNA (cDNA). Following this, cDNA fragments of approximately 150 bp, corresponding to the small RNA inserts plus adapters, were size-selected and purified via gel electrophoresis. The quality and quantity of the final constructed libraries were confirmed using KAPA Library Quantification Kits (KAPA Biosystems, USA) and the Agilent 4200 TapeStation system. Qualified libraries from different samples were pooled in appropriate ratios and subjected to high-throughput sequencing on an Illumina NovaSeq platform.

2.19. Luciferase reporter assay

To verify the direct targeting relationships between miR-542-3p and Asxl2, as well as between miR-542-3p and Eef1a1, a dual-luciferase reporter assay was performed. The plasmids for Asxl2, Eef1a1, and the miR-542-3p mimic were constructed by Hanbio Co., Ltd, China. The specific procedures were as follows: The wild-type 3′ untranslated regions (3′UTRs) of the Asxl2/Eef1a1 genes (containing the mmu-miR-542-3p binding site) and their corresponding mutant 3′UTRs were cloned into pSI-Check2 plasmids (Hanbio Co., Ltd.) respectively. The constructed plasmids were named Asxl2-3′UTR-wt (wild-type 3′UTR plasmid of Asxl2), Asxl2-3′UTR-mut (mutant 3′UTR plasmid of Asxl2), Eef1a1-3′UTR-wt (wild-type 3′UTR plasmid of Eef1a1), and Eef1a1-3′UTR-mut (mutant 3′UTR plasmid of Eef1a1). The above-constructed plasmids (either Asxl2-3′UTR-wt/Asxl2-3′UTR-mut or Eef1a1-3′UTR-wt/Eef1a1-3′UTR-mut) were co-transfected into 293T cells (a human renal epithelial cell line) together with miRNAs (miR-542-3p-negative control (NC) or miR-542-3p-mimic). Forty-eight hours (h) after transfection, operations were conducted in accordance with the manufacturer's instructions of the Dual-Luciferase Reporter Assay Kit (Beyotime Biotechnology Co., Ltd.) to detect the direct targeting relationships between miR-542-3p and Asxl2, and between miR-542-3p and Eef1a1. Finally, a microplate reader (Thermo Fisher, Waltham, USA) was used to measure the luciferase activity of cells in each group.

2.20. Statistical analysis

All statistical analyses were performed using Prism 9 software (GraphPad Software Inc., Boston, MA, USA). Tissue samples were collected at 2 months post-treatment for analysis. Data are presented as mean ± standard deviation (SD). A P-value <0.05 was considered statistically significant. For comparisons between two groups, an unpaired Student's t-test was used; for comparisons among multiple groups, one-way analysis of variance (one-way ANOVA) was conducted, followed by Tukey's post hoc test.

3. Results

3.1. Isolation and characterization of neonatal and adult muscle-derived extracellular vesicles

Extracellular vesicles (EVs) were isolated from skeletal muscle tissues of neonatal and adult mice by ultracentrifugation. Nanoparticle tracking analysis (NTA) showed that both neonatal muscle EVs (NMEVs) and adult muscle EVs (AMEVs) were mainly distributed within 50-200 nm (Fig. S1A–B). Transmission electron microscopy (TEM) confirmed typical EV morphology with a bilayer membrane (Fig. S1C–D). Western blotting verified EV markers TSG101, CD9, and CD81 in both NMEVs and AMEVs (Fig. S1E). PKH67 labeling further demonstrated efficient uptake of EVs by C2C12 cells (Fig. S1F), supporting subsequent functional assays.

3.2. NMEVs, but not AMEVs, attenuated palmitic acid-induced senescence in C2C12 cells

To establish a robust in vitro senescence model, C2C12 cells were treated with increasing concentrations of palmitic acid (PA; 0, 50, 100, 200, and 400 μM). Cell viability gradually decreased from 0 to 200 μM, whereas 400 μM caused marked cytotoxicity (Fig. S2A). Consistently, p21 protein increased from 0 to 200 μM but declined at 400 μM, likely due to extensive cell death masking senescence features (Fig. S2B–C). Therefore, 200 μM PA was selected for subsequent experiments. To establish the optimal working concentration of NMEVs, we treated PA-induced senescent C2C12 cells with increasing concentrations of NMEVs (0, 50, 100, and 200 μg/mL) and assessed p21 expression by Western blotting. As shown in Fig. S3, NMEVs suppressed p21 expression in a concentration-dependent manner. However, no statistically significant difference was observed between the 100 μg/mL and 200 μg/mL groups. Therefore, 100 μg/mL was selected as the optimal concentration for all subsequent experiments.

Using this model, C2C12 cells were exposed to 200 μM PA and concurrently treated with NMEVs or AMEVs (Fig. 1A). PA markedly increased senescence markers p16, p21, and p53, whereas NMEVs significantly reversed these changes; AMEVs showed no comparable effect (Fig. 1B–C’’’). In parallel, PA elevated the DNA damage marker γH2AX, while NMEVs reduced γH2AX staining intensity (Fig. 1D–D’). Immunofluorescence further showed that NMEVs lowered the proportion of p16-and p21-positive cells induced by PA (Fig. 1E–F’). Because oxidative stress is a key upstream driver of DNA damage and senescence, intracellular ROS was assessed using DCFH-DA; PA increased ROS levels, which were effectively suppressed by NMEVs but not AMEVs (Fig. 1G–G’). Together, these data indicate that NMEVs effectively counteract PA-induced senescence-associated phenotypes.

Fig. 1.

NMEVs effectively attenuated palmitic acid-induced senescence in C2C12 cells. (A) Schematic diagram of cell culture and treatment. (B) qRT-PCR analysis of the expression of senescence markers p53, cdkn1a, and cdkn2a in each group (n = 3). (C-C‴) Western blotting for the expression of senescence markers p53, cdkn1a (p21), and cdkn2a (p16) in each group with relative quantification (n = 6). (D-D′) Immunofluorescence staining for the DNA damage marker γH2AX with quantitative analysis (Scale bar, 100 μm; n = 6 for each group). (E-E′) Immunofluorescence staining for p16 with quantitative analysis (Scale bar, 100 μm; n = 6). (F-F′) Immunofluorescence staining for p21 with quantitative analysis (Scale bar, 100 μm; n = 6). (G-G′) Flow cytometry analysis of relative reactive oxygen species (ROS) levels (n = 6). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001; ns, not significant. Rel. fold, relative fold; T.Ar, total area.

3.3. NMEVs restored mitochondrial homeostasis and reduced lipid deposition in C2C12 cells after PA treatment

Given that mitochondrial dysfunction and ectopic lipid accumulation are closely linked to stress-induced senescence, we next examined whether NMEVs modulate these processes. PA significantly reduced ATP production, while NMEVs robustly restored ATP levels (Fig. 2A). Consistently, PA decreased mitochondrial DNA (mtDNA) content, including mtDNA-encoded genes (MT-CO1, MT-ND1, MT-CO3) and the control region D-loop; these reductions were significantly alleviated by NMEVs (Fig. 2B–E). Functional assays further showed that PA impaired mitochondrial respiratory capacity, as reflected by reduced Complex V activity and decreased oxygen consumption rate (OCR), both of which were rescued by NMEVs (Fig. 2F–G).

Fig. 2.

NMEVs alleviated palmitic acid-induced mitochondrial dysfunction and lipid deposition. (A) Relative ATP synthesis rates in each group (n = 6). (B) qRT-PCR analysis of MT-CO1 expression in each group (n = 3). (C) qRT-PCR analysis of MT-ND1 expression in each group (n = 3). (D) qRT-PCR analysis of MT-CO3 expression in each group (n = 3). (E) qRT-PCR analysis of D-loop expression in each group (n = 3). (F) Mitochondrial complex V activity in C2C12 cells of each group (n = 6). (G) Measurement of oxygen consumption rate (OCR) in C2C12 cells of each group (n = 4). (H-H′) Transmission electron microscopy (TEM) assessment of mitochondrial quantity with quantitative analysis (Scale bar, 500 nm; n = 3). (I-I′) Western blotting for PGC-1α expression in each group with relative quantification (n = 6). (J-J′) Immunofluorescence staining for SDHA with quantitative analysis (Scale bar, 100 μm; n = 6). (K-K′) Immunofluorescence staining for EdU with quantitative analysis (Scale bar, 100 μm; n = 6). (L-L′) Representative images of BODIPY staining in each group with quantitative analysis (Scale bar, 20 μm; magnified Scale bar, 5 μm; n = 6). (M-M′) Representative images of Oil Red O staining in each group with quantitative analysis (Scale bar, 20 μm; magnified scale bar, 5 μm; n = 6). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001; ns, not significant.

To evaluate mitochondrial quality, TEM revealed an increased proportion of abnormal mitochondria after PA treatment, which was markedly reduced by NMEVs (Fig. 2H–H’). Mechanistically, PA decreased PGC-1α protein expression, whereas NMEVs prevented this downregulation (Fig. 2I–I’), suggesting preserved mitochondrial maintenance. In agreement with enhanced mitochondrial metabolism, NMEVs increased SDHA staining, indicating improved TCA cycle/oxidative capacity (Fig. 2J–J’). As a functional consequence, NMEVs increased EdU incorporation, supporting improved proliferative competence under PA stress (Fig. 2K–K’).

Because mitochondrial dysfunction often coincides with lipid metabolic imbalance, we assessed lipid deposition. PA induced lipid accumulation by BODIPY and Oil Red O staining; NMEVs significantly reduced lipid deposition (Fig. 2L–M’). Collectively, NMEVs not only suppress senescence markers but also simultaneously restore mitochondrial homeostasis and mitigate lipid accumulation, providing basis for the anti-senescence effect.

3.4. Development and characterization of NMEVs-loaded bilayer microneedles for targeted delivery

To enable localized and sustained delivery of NMEVs in vivo, we fabricated a bilayer microneedle (MN) patch, named NMEVs@PLGA@Fucoidan-HA MNs (Fig. 3A). Hyaluronic acid (HA) served as the base; the inner layer consisted of PLGA mixed with NMEVs, and the outer layer was fucoidan. The morphology of MNs was observed using a Nikon camera and SEM. The MN patch formed a 20×20 conical array with needle length of 800 μm, tip-to-tip spacing of 550 μm, and tip diameter of approximately 15 μm (Fig. 3B–C). Dual fluorescence labeling (outer layer DiL; inner layer DiO) confirmed a clear bilayer architecture, as observed by confocal microscopy (Fig. 3D).

Fig. 3.

Preparation and characterization of NMEVs@PLGA@Fucoidan-HA microneedles. (A) Preparation process of NMEVs@PLGA@Fucoidan-HA MNs. (B) General appearance of NMEVs@PLGA@Fucoidan-HA MNs (Scale bar, 5 mm). (C) Scanning electron microscopy images of NMEVs@PLGA@Fucoidan-HA MNs (Scale bar, 200 μm; magnified Scale bar, 50 μm). (D) Longitudinal section of DiO/DiI-labeled bilayer structure of microneedles (Scale bar, 200 μm). (E) Mechanical strength testing of NMEVs@PLGA@Fucoidan-HA MNs. (F) H&E staining showing microneedle penetration into muscle tissue (Scale bar, 200 μm). (G) Representative images of microneedle degradation at different time points (Scale bar, 200 μm). (H) Degradation efficiency of microneedles at different time points (n = 3). (I-J) in vivo imaging showing microneedle degradation and release at different time points with quantitative analysis (n = 3). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001.

Compression testing showed a tolerable force of approximately 30 N, indicating sufficient mechanical strength for muscle penetration (Fig. 3E). Hematoxylin-eosin (HE) staining of muscle tissue after MN insertion confirmed successful penetration (Fig. 3F). Degradation assessment in artificial tissue fluid revealed time-dependent degradation, with approximately 30% by day 3 and 60% by day 7 (Fig. 3G and H). In vivo fluorescence tracking demonstrated sustained EVs release from MNs for up to 28 days (Fig. 3I and J), supporting their use for long-term intervention in age-associated muscle decline.

To characterize release kinetics, PLGA hydrogels loaded with 50 μg NMEVs were incubated in 800 μL PBS. Supernatants were collected and replaced with fresh PBS at days 1, 3, 7, 14, and 28. EV protein concentration was measured by BCA assay, and cumulative release was calculated as described in Section 2.15. EVs release followed a sustained profile, reaching approximately 85% by day 28, consistent with the in vivo fluorescence data (Fig. S4A). We further assessed whether the fabrication and release process affects EVs integrity and bioactivity. Released EVs collected after complete MNs dissolution were examined by TEM, which confirmed retention of the characteristic cup-shaped bilayer structure (Fig. S4B). Additionally, PKH67-labeled released EVs were readily internalized by C2C12 cells, demonstrating preserved functional integrity (Fig. S4C). These findings validate the suitability of our MNs system for sustained, functional NMEVs delivery.

3.5. NMEVs@PLGA@Fucoidan-HA microneedles alleviated skeletal muscle aging in aged mice

We next tested whether sustained delivery of NMEVs via MNs could mitigate skeletal muscle aging in vivo 2 months post-treatment. H&E staining of major organs (heart, liver, spleen, kidneys, intestines, and lungs) revealed no histological abnormalities after 2 months of microneedle treatment (Fig. S5), confirming the biocompatibility of the MNs fabrication process. To assess whether microneedles can deliver NMEVs in vivo, we administered PBS, free-PKH26, or PKH26-labeled NMEVs loaded onto microneedles, respectively. The results demonstrated that microneedles effectively delivered NMEVs into muscle tissue (Fig. S6A).

Compared with young mice, the skeletal muscle of aged mice exhibited typical aging characteristics, including reduced myofiber diameter, a shift from fast-twitch (MYH1+) toward slow-twitch fibers (MYH7+), and reduced PAX7+ muscle stem cells. MN treatment significantly reversed these aging-associated changes, increasing PAX7+ cells and restoring fast-twitch fiber proportion (Fig. 4A–B’). Consistently, MNs reduced senescence/DNA damage markers (p16, p21, γH2AX) and the atrophy marker Fbx32 (Fig. 4C–F’’’). MNs also improved tissue architecture by reducing fiber disorganization (WGA staining), fibrosis (Masson's trichrome staining), and lipid deposition (BODIPY staining) (Fig. 4G–I’).

Fig. 4.

NMEVs@PLGA@Fucoidan-HA microneedles (MNs) counteracted skeletal muscle aging. (A-A‴) Immunofluorescence staining for Myh1 and Myh7 with quantitative analysis (Scale bar, 50 μm; n = 6). (B-B′) Immunofluorescence staining for Pax7 with quantitative analysis (Scale bar, 50 μm; magnified Scale bar, 5 μm; n = 6). (C-C′) Immunofluorescence staining for p16 with quantitative analysis (Scale bar, 20 μm; magnified Scale bar, 5 μm; n = 6). (D-D′) Immunofluorescence staining for p21 with quantitative analysis (Scale bar, 20 μm; magnified Scale bar, 5 μm; n = 6). (E-E′) Immunofluorescence staining for γH2AX with quantitative analysis (Scale bar, 20 μm; magnified Scale bar, 5 μm; n = 6). (F-F‴) Western blotting for p16, p21 and Fbx32 expression in each group with relative quantification (n = 6). (G-G′) Immunofluorescence staining for WGA with quantitative analysis (Scale bar, 50 μm; n = 6). (H-H′) Masson's trichrome staining for assessment of fibrotic area in each group with quantitative analysis (Scale bar, 100 μm; magnified Scale bar, 20 μm; n = 6). (I-I′) Representative images of BODIPY staining in each group with quantitative analysis (Scale bar, 50 μm; magnified Scale bar, 10 μm; n = 6). (J) Relative ATP synthesis rates in each group (n = 6). (K) qRT-PCR analysis of MT-CO1 expression in each group (n = 3). (L) qRT-PCR analysis of MT-ND1 expression in each group (n = 3). (M) qRT-PCR analysis of MT-CO3 expression in each group (n = 3). (N) qRT-PCR analysis of D-loop expression in each group (n = 3). (O-O′) Scanning electron microscopy (SEM) assessment of mitochondrial morphology in muscle tissues of each group with quantitative analysis (Scale bar, 1 μm; n = 3). (P-P′) Immunofluorescence staining for WGA and SDHA with quantitative analysis (Scale bar, 50 μm; n = 6). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001.

Given our in vitro findings on mitochondrial rescue, we assessed mitochondrial indices in vivo. MNs increased ATP production and mtDNA content (MT-CO1, MT-ND1, MT-CO3, D-loop), reduced abnormal mitochondria (SEM readouts), and enhanced oxidative phosphorylation capacity (SDHA staining) (Fig. 4J–P’). Finally, because the MN outer layer contains fucoidan with anti-inflammatory potential, we validated that fucoidan suppressed LPS-induced IL-1β and IL-6 in macrophages without impairing C2C12 viability (Fig. S6B–C). In aged muscle, pro-inflammatory macrophage infiltration (CD45⁺, CD68⁺, CD86⁺) was elevated and was reduced by MNs treatment (Fig. S7). Together, these data demonstrate that MNs provide a biocompatible platform enabling sustained NMEV delivery and multi-dimensional improvement of aged muscle phenotypes, consistent with the anti-senescence/mitochondrial/lipid benefits observed in vitro.

To verify the indispensability of the bilayer structure, we performed comparative experiments in aged mice using free NMEVs, monolayer MNs (PLGA/NMEVs without fucoidan), and bilayer MNs, all administered at equivalent NMEV dosages. Compared with free NMEVs, monolayer MNs did not reduce the expression of inflammatory markers (IL-1β, IL-6, and TNF-α), yet they significantly decreased senescence markers (p21) and atrophy marker (Fbx32), lowered lipid deposition (BODIPY staining), and improved mitochondrial function (SDHA staining). Strikingly, bilayer MNs consistently outperformed all other control groups, exerting the strongest anti-inflammatory effects (reduction in IL-1β, IL-6, and TNF-α), the most potent suppression of senescence (p16/p21) and atrophy (Fbx32) markers, markedly reduced lipid deposition (BODIPY staining), and the most pronounced restoration of mitochondrial function (SDHA staining) (Fig. S8).

3.6. miRNA profiling identified miR-542-3p as a key NMEVs-enriched mediator of anti-senescence effects

To explore the molecular mechanism underlying the anti-senescence effect of NMEVs on C2C12 cells, we collected NMEVs and AMEVs for microRNA sequencing (Fig. 5A–B). Among the top 12 most significantly differentially expressed microRNAs, miR-542-3p showed the most pronounced difference between NMEVs and AMEVs (Fig. 5C). We then transfected C2C12 cells with miR-542-3p mimic or inhibitor, together with NMEVs. qRT-PCR results confirmed that the mimic significantly enhanced miR-542-3p expression, while the inhibitor suppressed it (Fig. 5D).

Fig. 5.

NMEVs reduced C2C12 senescence and lipid accumulation by enriching miR-542-3p to stabilize mitochondrial function. (A) PCA plot showing sample homogeneity of AMEVs and NMEVs (n = 3). (B) Heatmap showing the top 20 significantly upregulated and downregulated microRNAs. (C) qRT-PCR analysis of the expression of the top 12 significantly upregulated microRNAs (n = 3). (D) qRT-PCR analysis of miR-542-3p expression in C2C12 cells after transfection with NMEVs, mimic, or NMEVs + inhibitor (n = 3). (E-E′) Immunofluorescence staining for p16 with quantitative analysis (Scale bar, 50 μm; n = 6). (F-F′) Immunofluorescence staining for p21 with quantitative analysis (Scale bar, 50 μm; n = 6). (G-G′) Immunofluorescence staining for γH2AX with quantitative analysis (Scale bar, 50 μm; n = 6). (H-H′) Immunofluorescence staining for EdU with quantitative analysis (Scale bar, 50 μm; n = 6). (I-I′) Immunofluorescence staining for SDHA with quantitative analysis (Scale bar, 50 μm; n = 6). (J-J‴) Western blotting for p16, p21 and PGC-1α expression in each group with relative quantification (n = 6). (K) Relative ATP synthesis rates in each group (n = 6). (L) qRT-PCR analysis of MT-CO1 expression in each group (n = 3). (M) qRT-PCR analysis of MT-ND1 expression in each group (n = 3). (N) qRT-PCR analysis of MT-CO3 expression in each group (n = 3). (O) qRT-PCR analysis of D-loop expression in each group (n = 3). (P-P′) Representative images of BODIPY staining in each group with quantitative analysis (Scale bar, 20 μm; magnified Scale bar, 10 μm; n = 6). (Q-Q′) Representative images of Oil Red O staining in each group with quantitative analysis (Scale bar, 20 μm; magnified Scale bar, 10 μm; n = 6). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001; ns, not significant.

Functionally, miR-542-3p mimic phenocopied NMEVs by suppressing PA-induced senescence markers (p16, p21, γH2AX) and increasing EdU-positive cells, whereas miR-542-3p inhibition attenuated the protective effects of NMEVs (Fig. 5E–H’, Fig. 5J–J’’). Moreover, the miR-542-3p mimic recapitulated the ability of NMEVs to promote mitochondrial homeostasis and reduce lipid deposition, as evidenced by elevated expression of SDHA, enhanced ATP production, upregulation of mitochondrial DNA content (MT-CO1, MT-ND1, MT-CO3, D-loop), increased expression of PGC-1α, and reduced staining areas in BODIPY and Oil Red O assays. Conversely, the miR-542-3p inhibitor markedly counteracted these beneficial effects of NMEVs on mitochondrial function and lipid deposition (Fig. 5H–H’, Fig. 5K–Q’, Fig. 5J and. J’’’). These findings suggest that miR-542-3p, enriched in NMEVs, is a key mediator of their anti-senescent, mitochondrial-stabilizing, and lipid-reducing functions.

3.7. Asxl2 and Eef1a1 were validated as direct downstream targets of miR-542-3p

To identify downstream mediators, we integrated multiple miRNA target prediction tools and obtained 9 candidate targets (Asxl2, Eef1a1, Lrrc59, Gabarap, Ap3d1, Arhgap5, Kcmf1, Pten, and Ube2e1) (Fig. 6A). Upon miR-542-3p mimic transfection, qRT-PCR analysis revealed a significant reduction in the expression levels of additional sex combs like 2 (Asxl2) and eukaryotic translation elongation factor 1 alpha 1 (Eef1a1), while the other candidate genes remained largely unaffected (Fig. 6B–J). Western blotting further confirmed the suppressive effect of miR-542-3p on Asxl2 and Eef1a1 protein expression (Fig. 6K–K’’). Dual-luciferase assays demonstrated direct binding of miR-542-3p to the 3′UTRs of Asxl2 and Eef1a1 (Fig. 6L–M’). Notably, both Asxl2 and Eef1a1 were upregulated in aged mouse muscle compared with neonatal muscle (Fig. 6N–N’), consistent with their potential involvement in muscle aging. Thus, Asxl2 and Eef1a1 were identified as direct, physiologically relevant targets of miR-542-3p.

Fig. 6.

Asxl2 and Eef1a1 served as downstream target genes of miR-542-3p. (A) Prediction of downstream target genes of miR-542-3p using multiple target gene prediction software. (B) qRT-PCR analysis of Asxl2 expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (C) qRT-PCR analysis of Eef1a1 expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (D) qRT-PCR analysis of Lrrc59 expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (E) qRT-PCR analysis of Gabarap expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (F) qRT-PCR analysis of Ap3d1 expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (G) qRT-PCR analysis of Arhgap5 expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (H) qRT-PCR analysis of Kcmf1 expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (I) qRT-PCR analysis of Pten expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (J) qRT-PCR analysis of Ube2e1 expression in C2C12 cells after transfection with miR-542-3p mimic (n = 3). (K-K″) Western blotting for Asxl2 and Eef1a1 expression in C2C12 cells after transfection with miR-542-3p mimic, with relative quantification (n = 3). (L-L′) Dual-luciferase reporter assay verifying the direct targeting binding relationship between miR-542-3p and Asxl2 (n = 3). (M-M′) Dual-luciferase reporter assay verifying the direct targeting binding relationship between miR-542-3p and Eef1a1 (n = 3). (N-N″) Western blotting for Asxl2 and Eef1a1 expression in neonatal and aging muscle tissues (n = 3). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001; ns, not significant.

3.8. miR-542-3p reduced lipid deposition through the Asxl2–PPARγ axis

Because lipid accumulation was robustly suppressed by NMEVs/miR-542-3p, we examined whether Asxl2 mediates this effect. An Asxl2 overexpression plasmid (Asxl2^OE) was co-transfected with miR-542-3p mimic into C2C12 cells. Western blot results showed that miR-542-3p mimic significantly downregulated the PA-induced increase in Asxl2 expression, while the Asxl2^OE restored Asxl2 expression (Fig. 7A–A’). PPARγ levels were higher in aged muscle than neonatal muscle (Fig. 7B–C), suggesting a lipidogenic activation state in aging. In vitro, miR-542-3p mimic attenuated PA-induced PPARγ upregulation, indicating negative regulation of the Asxl2-PPARγ axis (Fig. 7D–D’). Importantly, Asxl2 overexpression significantly abrogate the lipid deposition-reducing effect of the miR-542-3p in both BODIPY and Oil Red O assays (Fig. 7E–F’). These data support a model in which miR-542-3p alleviates lipid deposition at least in part via inhibition of the Asxl2-PPARγ pathway.

Fig. 7.

miR-542-3p mitigated PA-induced lipid deposition by inhibiting the Asxl2-PPARγ pathway. (A-A′) Western blotting for Asxl2 expression after PA induction, followed by transfection with miR-542-3p mimic and Asxl2 overexpression plasmid (Asxl2^OE), with relative quantification (n = 3). (B) Schematic diagram illustrating Asxl2 regulation of lipid storage via PPARγ. (C-C′) Western blotting for PPARγ expression in neonatal and aging muscle tissues with relative quantification (n = 3). (D-D′) Western blotting for PPARγ expression after PA induction, followed by transfection with miR-542-3p mimic and Asxl2 overexpression plasmid (Asxl2^OE), with relative quantification (n = 3). (E-E′) Representative images of BODIPY staining in each group with quantitative analysis (Scale bar, 20 μm; magnified scale bar, 5 μm; n = 6). (F-F′) Representative images of Oil Red O staining in each group with quantitative analysis (Scale bar, 20 μm; magnified scale bar, 5 μm; n = 6). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001; ns, not significant.

3.9. miR-542-3p ameliorated mitochondrial dysfunction and senescence by suppressing Eef1a1 and activating AMPK

We next investigated how miR-542-3p improves mitochondrial function and senescence. We observed elevated Eef1a1 expression in muscle tissues of aged mice (Fig. 6N–N’’). Co-transfection of an Eef1a1 overexpression plasmid (Eef1a1^OE) with miR-542-3p mimic confirmed that miR-542-3p negatively regulates Eef1a1, and that Eef1a1^OE effectively restored Eef1a1 expression (Fig. 8A–A’). Because AMPK is a central regulator of mitochondrial homeostasis, we assessed its activation state. Phosphorylated AMPK (p-AMPK) was reduced in aged muscle (Fig. 8B–C’). In vitro, miR-542-3p mimic enhanced PA-induced AMPK activation, whereas Eef1a1 overexpression attenuated this activation (Fig. 8D–D’), suggesting that miR-542-3p activates AMPK via Eef1a1 suppression (see Fig. 9).

Fig. 8.

miR-542-3p suppressed Eef1a1 to ameliorate PA-induced mitochondrial dysfunction and cellular senescence. (A-A′) Western blotting for Eef1a1 expression after PA induction, followed by transfection with miR-542-3p mimic and Eef1a1 overexpression plasmid (Eef1a1^OE), with relative quantification (n = 3). (B) Schematic diagram illustrating Eef1a1 regulation of lipid storage via AMPK. (C-C′) Western blotting for AMPK and p-AMPK expression in neonatal and aging muscle tissues with relative quantification (n = 3). (D-D′) Western blotting for AMPK and p-AMPK expression after PA induction, followed by transfection with miR-542-3p mimic and Eef1a1 overexpression plasmid (Eef1a1^OE), with relative quantification (n = 3). (E-E″) Representative images of p16 and p21 staining in each group with quantitative analysis (Scale bar, 50 μm; n = 6). (F) Relative ATP synthesis rates in each group (n = 6). (G) qRT-PCR analysis of MT-CO1 expression in each group (n = 3). (H) qRT-PCR analysis of MT-ND1 expression in each group (n = 3). (I) qRT-PCR analysis of MT-CO3 expression in each group (n = 3). (J) qRT-PCR analysis of D-loop expression in each group (n = 3). (K) Mitochondrial complex V activity in C2C12 cells of each group (n = 6). (L-L′) Representative images of SDHA staining in each group with quantitative analysis (Scale bar, 50 μm; n = 6). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001; ns, not significant.

Fig. 9.

Graphic abstract. (A) Isolation and purification procedures for small extracellular vesicles (sEVs) derived from the muscle tissues of 2-day-old neonatal and adult mice. (B) Preparation of NMEVs@PLGA@Fucoidan-HA and its therapeutic application in aged mice. (C) miR-542-3p encapsulated within NMEVs exerts dual targeting effects: it directly abrogates Asxl2 to subsequently repress PPARγ-mediated lipid deposition, and concurrently suppresses Eef1a1 to activate the AMPK signaling pathway, ultimately reversing cellular and organismal aging.

To determine whether AMPK activation is required for the anti-senescent effects of NMEVs and miR-542-3p, we treated cells with NMEVs or miR-542-3p mimic in the presence or absence of the AMPK inhibitor Dorsomorphin (Compound C, MCE, USA) at a concentration of 10 μM. Dorsomorphin treatment significantly abrogated the NMEVs- and miR-542-3p mimic-induced upregulation of SDHA and reversed their inhibitory effects on senescence markers p16 and p21 (Fig. S9). These results demonstrate that AMPK activation is a critical downstream mediator of these beneficial effects.

Functionally, Eef1a1^OE reversed the miR-542-3p mediated reduction in senescent cells (Fig. 8E–E’’). Eef1a1^OE also impaired mitochondrial rescue by miR-542-3p, including decreased ATP production (Fig. 8F and K), reduced mtDNA content (MT-CO1, MT-ND1, MT-CO3, D-loop) (Fig. 8H–J), and diminished oxidative phosphorylation capacity as indicated by reduced SDHA intensity (Fig. 8L–L’). Together, these results indicated that miR-542-3p alleviated PA-induced mitochondrial dysfunction and cellular senescence by inhibiting Eef1a1, thereby relieving its suppression on AMPK.

4. Discussion

Skeletal muscle aging poses a major threat to the health and quality of life of the elderly [26]. Current interventions predominantly rely on exercise, nutritional support, or symptomatic management, yet there remains a lack of directly translatable therapies capable of reversing mitochondrial dysfunction and cellular senescence [3]. Here, we identify neonatal mouse muscle-derived extracellular vesicles (NMEVs) as a potent therapeutic agent against muscle aging. To enable targeted and sustained delivery, we engineered a bilayer dissolvable microneedle (MN) patch loaded with NMEVs. Mechanistically, we uncovered a previously unknown role for NMEVs-enriched miR-542-3p, which simultaneously targets Eef1a1 and Asxl2 to restore mitochondrial function, reduce senescence, and limit lipid accumulation. This aligns with the established notion that mitochondrial dysfunction is a core driver of skeletal muscle aging [7,27]. Together, our findings establish NMEVs-loaded microneedles as a promising and translatable strategy for combating age-related skeletal muscle degeneration.

Skeletal muscle aging is a complex, multifactorial process driven by interconnected pathological alterations that form a vicious cycle. Skeletal muscle is primarily composed of terminally differentiated multinucleated myofibers, along with mononuclear cell populations such as satellite cells, fibro-adipogenic progenitors (FAPs), endothelial cells, and immune cells [28]. Disruption of the FAP-immune balance promotes fibrosis and fat infiltration, while satellite cell depletion compromises regeneration [29]. Aging also induces progressive loss of cellular proteins and organelles, leading to myofiber atrophy [30]. Accordingly, we observed elevated atrophy markers (e.g., Fbx32) in aged muscle, an effect partially reversed by NMEVs. Enhancing mitochondrial function has emerged as a promising therapeutic strategy. Recent studies show that activating the PI3K/Akt/mTOR pathway can improve mitochondrial function and counteract muscle aging [31]. Notably, NMEVs act through a distinct mechanism: miR-542-3p relieves Eef1a1-mediated inhibition of AMPK, thereby restoring mitochondrial homeostasis. Senescent cells contribute to a state of chronic, low-grade inflammation throughout the body via the senescence-associated secretory phenotype (SASP), which accelerates the exhaustion of muscle stem cells and impairs tissue regenerative capacity [32]. Thus, reducing senescent burden may preserve the satellite cell pool. In line with this, NMEVs decreased p16/p21 expression, increased PAX7+ cells, and reduced fibrosis. Collectively, our findings highlight a multi-targeted capacity of NMEVs to combat muscle aging by stabilizing mitochondrial function. Specifically, NMEVs reversed PA-induced senescence in C2C12 cells and mitigated mitochondrial defects (e.g., reduced ATP, oxidative phosphorylation, and mtDNA content) and lipid accumulation. These observations suggest that NMEVs improve the muscle microenvironment by disrupting the vicious cycle of mitochondrial damage, ROS accumulation, and DNA damage/senescence.

Extracellular vesicles (EVs) are emerging as promising therapeutic tools in the musculoskeletal system due to their capacity to shuttle bioactive molecules and mediate intercellular communication [33,34]. EVs derived from adult mesenchymal stem cells (e.g., adipose, umbilical cord, bone marrow) have shown efficacy in treating muscle injury, osteoporosis, and tendon disorders [[35], [36], [37]]. For instance, tendon stem cell-derived EVs significantly reduced heterotopic ossification and promoted tendon healing after injury [38]. More recently, EVs from young tissues have gained attention for their superior regenerative potential. Young plasma-derived EVs, for example, modulate mitochondrial metabolism via PGC-1α, reducing frailty and extending lifespan in mice [39]. They also mitigate brain injury and promote functional recovery after intracerebral hemorrhage by regulating the p53/SLC711/GPX4 pathway [40]. Collectively, these studies highlight "young EV-mediated reprogramming of mitochondrial and stress pathways" as a transferable anti-aging strategy. Our research advances this concept further. Neonatal tissues exhibit exceptional regenerative capacity, driven by rapid proliferation and unique developmental signals. After tendon injury, neonatal mouse tendons activate the PI3K-Akt pathway to enhance cell proliferation and stemness, leading to superior repair [41]. Similarly, neonatal cardiomyocytes upregulate HMGB2 to promote glycolysis-driven proliferation and heart repair after infarction [42]. These observations led us to hypothesize that NMEVs retain neonatal developmental signals and can re-activate repair and metabolic homeostasis programs in aged muscle. Recent studies support the promise of neonatal tissue-derived EVs in regeneration. Li et al. reported that EVs from neonatal mouse heart tissue exhibited stronger pro-proliferative, anti-apoptotic, and pro-angiogenic activities than those from adult hearts [14]. Others have reported that neonatal skin- and kidney-derived EVs accelerate wound healing and improve renal function after injury [12]. However, evidence for neonatal tissue EVs in combating natural skeletal muscle aging remains limited, and a clear mechanistic link from a specific miRNA to target genes and functional phenotypes is lacking. Our study fills this gap. We demonstrate that NMEVs possess unique advantages over adult muscle-derived EVs (AMEVs) in reversing muscle aging. By preserving neonatal developmental cues, NMEVs stabilize mitochondrial function, thereby reducing cellular senescence and lipid deposition.

MicroRNAs are key mediators of EV biological functions [43]. For instance, tendon stem cell-derived EVs deliver miR-21a-5p to regulate the PDCD4/AKT/mTOR pathway, promoting BMSC-mediated tendon-to-bone healing [44]. Similarly, young plasma-derived EVs carry miR-142-5p, which targets ZFPM2 to enhance BMSC osteogenic differentiation and ameliorate osteoporosis [11]. These examples show that EV-enriched miRNAs display tissue and age specificity and can be functionally validated, which is the approach we used in this study. By comparing miRNA profiles of NMEVs and AMEVs, we identified miR-542-3p as the top candidate and subsequently validated its anti-aging effects. Previous studies have shown that miR-542-3p suppresses tumor cell proliferation and metastasis in cancers such as hepatocellular carcinoma and neuroblastoma [[45], [46], [47]]. Within the musculoskeletal system, miR-542-3p mitigates methotrexate-induced osteopenia by targeting sFRP-1 and Smurf2, and improves osteoporosis by regulating MSC osteogenic differentiation via sFRP-1 inhibition [48,49]. These studies suggest that miR-542-3p can reshape cell proliferation/stress and lipid metabolism-related phenotypes, but its role and direct targets in skeletal muscle aging remain unclear. Using dual-luciferase reporter assays, we confirmed that miR-542-3p directly binds to the 3′UTRs of Eef1a1 and Asxl2. While this establishes a canonical miRNA–target interaction, we acknowledge that miRNAs can also regulate gene expression through non-canonical mechanisms, including mRNA degradation [5,50]'UTR binding [51], cooperate with RNA-binding proteins [52], ceRNA networks [53], or nuclear transcriptional regulation [54]. To fully map the regulatory landscape of miR-542-3p in muscle aging, future studies should employ approaches such as AGO2-RIP-seq, CLIP-seq, and integrated transcriptome-proteome analysis. Notably, our study is the first to uncover a protective role for miR-542-3p in muscle aging and to elucidate its dual-target mechanism involving the Eef1a1-AMPK and Asxl2-PPARγ pathways. This dual targeting is particularly significant, as it simultaneously addresses two core pathological modules in aged muscle: energy metabolism/mitochondrial homeostasis and aberrant lipid deposition. Eef1a1 is a translation elongation factor involved in cytoskeletal organization, apoptosis, and stress responses [55]. We found that Eef1a1 is upregulated in aged muscle and that its overexpression suppresses AMPK activation-a central energy sensor and regulator of autophagy and mitochondrial function [56,57]. By inhibiting Eef1a1, miR-542-3p relieves this suppression, restoring AMPK activity, enhancing mitochondrial ATP production, reducing ROS, and alleviating senescence, thereby breaking the "mitochondrial dysfunction-ROS-DNA damage/senescence" vicious cycle [7,27]. Asxl2 is an epigenetic regulator of lipid metabolism. Asxl2 knockout mice exhibit reduced lipid accumulation [58], and myeloid-specific Asxl2 deletion confers resistance to high-fat diet-induced obesity [59]. Studies have also shown that Asxl2 positively regulates PPARγ, which plays a key role in lipid deposition [58]. While most previous studies focused on systemic metabolism or immune cells, these findings point to a conserved Asxl2-PPARγ axis driving lipid accumulation. We extend this concept to skeletal muscle aging, showing that Asxl2 and PPARγ are elevated in aged muscle and that miR-542-3p reduces ectopic lipid deposition by suppressing this axis. This dual-targeting regulatory mechanism offers significant advantages in combating mitochondrial dysfunction, cellular senescence, stem cell exhaustion, and lipid deposition in aged muscle, further supporting its potential as a broad-spectrum therapeutic target for age-related diseases.

A major barrier to clinical translation of EV-based therapies is inefficient delivery. Free EVs are rapidly cleared by the reticuloendothelial system, exhibit poor targeting to deep tissues like skeletal muscle, and easily lose bioactivity due to degradation [60]. These limitations highlight the need for a delivery system that enhances local retention and prolongs effective EVs exposure. To realize the therapeutic potential of NMEVs, we employed a microneedle (MN) patch. MNs are minimally invasive, painless arrays of micron-scale needles that enable localized, sustained release with reduced systemic exposure [61]. Since sarcopenia is a chronic condition requiring repeated intervention, MNs are well suited for long-term management owing to their repeatability, minimal pain, and sustained local drug delivery. In tendinopathy, hydrogel MNs loaded with cell-free fat extract promote tendon repair by rescuing mitochondrial damage [62]. Similarly, MSC-EV-loaded MNs enhance angiogenesis and reduce inflammation in diabetic wound healing [63]. These studies not only validate the feasibility of EV-loaded MNs but also suggest that sustained local EV delivery may support mitochondrial repair and anti-inflammatory effects. Current therapeutic MNs targeting muscle are primarily focused on cardiovascular diseases like myocardial infarction [64]. For instance, gelatin-based biocompatible MNs patches loaded with human umbilical cord MSCs-derived EVs, delivering miR-29b mimics to fibroblasts, significantly reduced inflammation, infarct size, and fibrosis post-myocardial infarction [65]. However, MN systems specifically designed for skeletal muscle applications remain limited. In contrast, most existing skeletal muscle MN applications are limited to flexible sensors for electromyographic monitoring rather than therapeutic intervention [66]. To address this gap, we designed a dissolvable bilayer MN patch for aged skeletal muscle, with localized deep-tissue delivery and prolonged release. The bilayer design contributes to a controlled degradation and release profile. The fucoidan outer layer dissolves relatively rapidly, while the PLGA core degrades more slowly to sustain EV delivery. In vitro, degradation reached ∼30% at day 3 and ∼60% at day 7, reflecting outer layer dissolution and partial PLGA erosion. In vivo, fluorescence signals persisted at 28 days, confirming sustained EVs release under physiological conditions. Such a sustained release profile may be particularly beneficial for aged skeletal muscle, where regenerative capacity is impaired and prolonged therapeutic support is often required [67,68]. PLGA degradation products (lactic and glycolic acids) are endogenous metabolites with established safety, posing no local or systemic risk [69]. Consistent with this, we observed no histological abnormalities in major organs following MN treatment, further supporting the translational potential of this system.

Our study has several limitations. First, in vivo experiments were only conducted in mice; future studies should validate the efficacy and safety of the NMEV-MN system in larger animal models (e.g., canine or non-human primates) to provide more sufficient evidence for clinical translation. Second, while we focused on miR-542-3p, NMEVs likely contain other bioactive components (e.g., proteins, lipids, other microRNAs) that may act synergistically with miR-542-3p. Future multi-omics analyses (proteomics, lipidomics, metabolomics) could uncover additional regulatory molecules and pathways. Finally, the precise mechanisms by which Eef1a1 regulates AMPK and Asxl2 modulates PPARγ in skeletal muscle remain to be fully elucidated. Future studies using Co-IP/MS and metabolic flux analysis could clarify whether Eef1a1 directly interacts with AMPK or acts via energy status. For Asxl2, ChIP-seq targeting histone modifications would determine whether it directly controls PPARγ transcription. Addressing these questions will complete the downstream signaling network.

5. Conclusion

In summary, this study introduces a novel therapeutic strategy for skeletal muscle aging using neonatal muscle-derived extracellular vesicles (NMEVs) delivered via a bilayer microneedle patch. We demonstrate that NMEVs outperform adult muscle-derived EVs in attenuating cellular senescence, mitochondrial dysfunction, and lipid accumulation. Mechanistically, NMEVs are enriched with miR-542-3p, which exerts dual anti-aging effects by suppressing Eef1a1 to activate AMPK and inhibiting Asxl2 to downregulate PPARγ. This coordinated regulation of two key pathways establishes a molecular framework for NMEVs-based therapy and offers a promising translatable approach for combating skeletal muscle aging.

Building on our findings, several exciting directions warrant further investigation. First, the dual-target anti-aging mechanism of miR-542-3p could guide the design of engineered EVs-for example, by loading miR-542-3p mimics into EVs from more accessible sources (e.g., mesenchymal stem cells) to develop scalable therapeutics. Second, given that lipid accumulation and mitochondrial dysfunction are common features of various muscle atrophy conditions, the NMEVs-MN platform may be extended to other muscle atrophy disorders, such as Duchenne muscular dystrophy, corticosteroid-induced myopathy, or anterior cruciate ligament injury-induced muscle atrophy. Third, the bilayer microneedle design provides a versatile platform for combinatorial therapy-for instance, co-delivery with senolytics to achieve synergistic anti-aging effects.

Ethics approval and consent to participate

This study was conducted in accordance with the guidelines for animal care and was approved by the Animal Ethics Committee of Xiangya Hospital, Central South University (No. 202511197).

CRediT authorship contribution statement

Feifei Yuan: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft. Yong Chen: Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization. Wenjie Li: Formal analysis, Investigation. Lin Zhang: Formal analysis, Investigation. Ruixue Du: Data curation, Software. Sheng Xu: Investigation, Validation. Zhiyu Hu: Formal analysis, Software, Visualization. Tao Zhang: Funding acquisition, Methodology, Resources. Ling Qin: Resources, Supervision, Writing – review & editing. Hongbin Lu: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing. Chengjun Li: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

All data in the current study are available from the corresponding author on reasonable request.