Role of LDHA in senescent fibroblast exosomes promoting ferroptosis via histone lactylation-mediated ACSL4 regulation in skin photoaging

Abstract

Objective

This study aimed to elucidate the mechanism of Lactate Dehydrogenase A (LDHA) in senescent fibroblast-derived exosomes during skin photoaging, focusing on the molecular pathway by which it regulates Acyl-CoA Synthetase Long-Chain Family Member 4 (ACSL4) expression through histone lactylation, thereby inducing ferroptosis and accelerating skin photoaging.

Methods

An ultraviolet B (UVB)-induced senescence model was established using human foreskin fibroblasts. Exosomes were isolated from senescent fibroblasts and characterized. Their features and uptake were assessed using Western blot, transmission electron microscopy (TEM), nanoparticle tracking analysis (NTA), and immunofluorescence. Small interfering RNA was employed to knock down LDHA and ACSL4 gene. Pharmacological inhibitors (FX11, Ferrostatin-1) and sodium lactate rescue experiments were utilized. Lactate levels, histone H3K18 lactylation modification, ACSL4 transcriptional activity, and ferroptosis markers were detected to assess the effects of ACSL4 gene lactylation on ferroptosis. The biological effects of exosomal LDHA in photoaged tissue were validated using an in vivo UVB-irradiated mouse model.

Results

UVB irradiation induced fibroblast senescence and significantly upregulated LDHA expression. Exosomes from senescent fibroblasts were effectively taken up by HaCaT cells, leading to increased lactate levels and enhanced histone H3K18 lactylation in recipient cells. LDHA knockdown or inhibition downregulated ACSL4 expression and suppressed ferroptosis, whereas exogenous lactate partially restored these effects. RNA sequencing and ChIP-qPCR results indicated that LDHA-mediated lactylation modification was enriched at the ACSL4 promoter region, enhancing its transcriptional activity. In vivo experiments further confirmed that senescent exosomes accelerated UVB-induced skin collagen degradation and ferroptosis, while LDHA intervention significantly alleviated photoaging damage.

Conclusion

Senescent fibroblast-derived exosomes deliver LDHA, promoting histone lactylation modification, which upregulates ACSL4 expression and activates the ferroptosis pathway, ultimately accelerating skin photoaging. This study reveals the coupling mechanism between metabolic signaling and epigenetic regulation in skin aging, providing new molecular targets and a theoretical basis for anti-photoaging therapy.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13072-026-00683-0.

Introduction

Skin photoaging is a widely prevalent form of skin aging globally, particularly common among Caucasian populations. Statistics indicate that approximately 80–90% of Europeans and Americans exhibit varying degrees of skin photoaging [1]. Long-term solar exposure-induced photodamage not only causes aging signs such as wrinkles and pigmentation but also significantly increases the risk of skin tumors [2]. Photoaging is considered a major trigger for photo-associated skin cancers, including cutaneous melanoma. In regions with high ultraviolet (UV) exposure, such as New Zealand, approximately 67,000 new cases of skin cancer annually are related to sun exposure, and globally, over 65,000 people die each year from skin cancers associated with photoaging [1, 3–5]. These severe epidemiological data highlight the urgency of in-depth research into the molecular mechanisms of skin photoaging and the identification of intervention targets.

UV radiation is the primary environmental factor causing skin photoaging. UV exposure induces the overproduction of reactive oxygen species (ROS) and reactive nitrogen species (RNS) within skin tissues, accompanied by the release of inflammatory mediators and direct cellular DNA damage [6]. These changes collectively lead to the accelerated degradation of extracellular matrix components, such as dermal collagen, ultimately resulting in clinical manifestations of photoaging like wrinkles and telangiectasia [6]. Furthermore, UV radiation can reduce the number of epidermal Langerhans cells, melanocytes, and dermal fibroblasts, decrease antioxidant defense capacity, and simultaneously trigger the overexpression of various pro-inflammatory cytokines and matrix metalloproteinases (MMPs), thereby exacerbating the skin photoaging process [7]. Thus, oxidative stress, inflammatory responses, and imbalance in matrix remodeling constitute the primary mechanisms of skin photoaging.

In recent years, ferroptosis, a newly discovered form of iron-dependent programmed cell death, has been proven to play a significant role in skin aging. Ferroptosis is characterized by intracellular iron accumulation and lipid peroxidation, often accompanied by massive ROS generation and abnormalities in mitochondrial morphology and function [8]. Studies have shown that ferroptosis plays a key role in ultraviolet B (UVB)-induced skin photoaging models, and inhibiting ferroptosis can alleviate UVB-induced oxidative damage and collagen degradation, thereby histologically mitigating the photoaging phenotype [9, 10]. Further mechanistic analyses indicate that intervening in ferroptosis-related pathways (e.g., knocking down the mitochondrial respiratory chain NDUFS4 subunit gene) can reduce ROS and MMP1 levels, increase type I collagen and the expression of the anti-ferroptosis molecule GPX4, suggesting enhanced cellular resistance to ferroptosis [8]. Of particular note, Acyl-CoA long-chain acyl synthetase 4 (ACSL4), a key enzyme regulating lipid composition, can catalyze the esterification of polyunsaturated fatty acids like arachidonic acid and trigger lipid peroxidation signaling, thus being regarded as an important promoter in ferroptosis and a potential therapeutic target [11].

Lactate, produced during glycolysis, was historically often considered a metabolic waste product. However, accumulating evidence indicates that lactate plays crucial signaling roles in cell survival and aging regulation. Lactate dehydrogenase A (LDHA) is the rate-limiting enzyme catalyzing the reduction of pyruvate to lactate. Its activity directly determines intracellular lactate levels, thereby influencing cellular metabolic status and signaling pathways [12]. In 2019, a study first reported that lactate can serve as a substrate for histone lysine residues, leading to a novel epigenetic modification termed histone lactylation, which directly promotes gene transcription activity on chromatin [13]. This discovery revealed a new molecular link between cellular metabolites and gene expression. Lactate-induced histone lactylation can initiate specific gene expression programs, thereby influencing cell function and fate. For instance, studies in skin microenvironment models have observed that exogenous lactate promotes H4K12 histone lactylation in macrophages, upregulating the transcription of TGF-β1 and TGF-β3, which in turn enhances collagen synthesis in adjacent fibroblasts [14]. This lactate-histone lactylation axis inhibits the activity of deacetylase HDAC3, forming a positive feedback loop that promotes collagen production, and is considered a novel mechanism against skin aging [14, 15]. Thus, histone lactylation, as a novel epigenetic mark, connects metabolic status with gene expression regulation and plays an important role in physiological and pathological processes such as skin aging.

Exosomes are nano-sized extracellular vesicles secreted by cells, carrying various bioactive components (including proteins, lipids, mRNAs, and non-coding RNAs). They can be taken up by recipient cells and mediate intercellular communication [16]. Numerous studies have shown that exosomes are closely related to aging-associated processes such as oxidative stress, inflammatory responses, and cellular senescence, and also play significant regulatory roles in skin photoaging [16]. Notably, exosomes not only transfer genetic information like nucleic acids but can also transport functional molecules such as metabolic enzymes, thereby influencing the signaling pathways and epigenetic status of recipient cells. Recent research found that exosomes derived from bone marrow mesenchymal stem cells are rich in active LDHA enzyme, which can be taken up by chondroprogenitor cells and significantly increase their lactate metabolism levels. This subsequently induces histone H3K18 lactylation, upregulates the gene expression of bone morphogenetic protein 7 (BMP7), and promotes the repair and regeneration of damaged tissue [17]. Similarly, exosomes secreted by neutrophils deliver LDHA to recipient tumor cells, increasing H3K18 lactylation levels at the GPX4 gene promoter region in the recipient cells, thereby inhibiting the occurrence of ferroptosis and enhancing the survival and proliferation capacity of tumor cells [18]. These findings reveal the critical role of exosome-mediated metabolic-epigenetic signaling in cell fate decisions, providing new perspectives for understanding intercellular communication in different tissue environments.

Although research on the mechanisms of photoaging and the regulation of ferroptosis has continuously advanced in recent years, it remains unclear whether these pathways form intersecting regulatory networks in skin photoaging. In particular, direct experimental evidence is lacking regarding whether exosome-mediated metabolic signals can trigger ferroptosis through epigenetic mechanisms, thereby accelerating skin photoaging. Based on the above background, this study focuses on this unresolved question and innovatively proposes and will validate a novel hypothesis: Exosome-carried LDHA can induce histone lactylation modification in recipient skin cells, upregulate the expression of the key ferroptosis enzyme ACSL4, promote the occurrence of ferroptosis, and ultimately accelerate the process of skin photoaging. Our research introduces this linked mechanism of exosomes–metabolic enzyme–epigenetics–cell death into the field of skin photoaging, providing new insights for revealing the deep mechanisms of photoaging and offering a novel theoretical basis and potential targets for developing anti-photoaging intervention strategies. In summary, this study hypothesizes that exosomal LDHA accelerates skin photoaging by promoting ferroptosis via histone lactylation-induced upregulation of ACSL4.

Materials and methods

Cell culture and UVB-induced senescence model

Human dermal fibroblasts (HDFs, Procell, CL-0111) and HaCaT human keratinocytes (Cell Bank of the Chinese Academy of Sciences, GNHu43) were cultured in high-glucose Dulbecco’s Modified Eagle Medium (DMEM, Gibco, 11965-092) supplemented with 10% fetal bovine serum (FBS, Gibco, 10270-106) and 1% penicillin-streptomycin mixture (Gibco, 15140-122). All cells were maintained in a humidified incubator at 37 °C with 5% CO₂ (Thermo Scientific, Model 3111). Cells were passaged at a 1:3 ratio, and the culture medium was refreshed every 2–3 days. Cells in the logarithmic growth phase were used for all experiments to ensure consistent cell status and response.

To establish a photo-induced cellular senescence model, HDFs were exposed to UVB radiation. A UVB light source with a primary peak wavelength of 312 nm (Vilber Lourmat, VL-215.M, France) was used, and the irradiation intensity was calibrated using a UV radiometer (Sigma High-Precision UV Radiometer, Model UV-340 A). Before irradiation, the culture medium was removed, and cells were gently washed twice with calcium- and magnesium-free phosphate-buffered saline (PBS, Gibco, 10010-023) to remove serum interference. UVB irradiation was then performed with the cells covered in serum-free PBS, at a distance of 15 cm from the light source. A single irradiation dose of 30 mJ/cm² was administered once daily for 5 consecutive days. Immediately after each irradiation, the PBS was replaced with fresh complete medium, and cells were returned to the incubator. Non-irradiated cells served as the negative control. Following consecutive irradiations, cells gradually exhibited enlarged, flattened morphology and reduced refractivity, indicative of senescence.

After model establishment, senescence was validated at morphological, cellular, and molecular levels. Morphological changes were first observed under an inverted microscope (Olympus IX73, Japan). Subsequently, cellular senescence was assessed using the Senescence-Associated β-Galactosidase (SA-β-Gal) Staining Kit (Cell Signaling Technology, #9860) according to the manufacturer’s instructions. Briefly, after irradiation, the medium was discarded, and cells were washed three times with PBS, fixed with 1× fixative solution at room temperature for 15 min, washed again with PBS, and then incubated with the SA-β-Gal staining working solution (containing X-gal substrate, pH 6.0) in a 37 °C incubator without CO₂ for 12–16 h. After staining, five random fields per sample were captured using an inverted microscope, and the percentage of blue-stained positive cells was calculated. A positive cell rate exceeding 50% was considered indicative of successful senescence model establishment. Cell viability was determined using the CCK-8 kit (Dojindo Laboratories, CK04-11, Japan). Cells were seeded in 96-well plates at a density of 5 × 10³ cells per well. After treatment, 10 µL of CCK-8 working solution was added to each well, followed by incubation at 37 °C for 2 h. Absorbance at 450 nm was measured using a microplate reader (Bio-Rad, Model 680).

To further verify molecular markers of senescence, total cellular protein was extracted and subjected to Western blot analysis. Proteins were extracted using RIPA lysis buffer (Solarbio, R0010, China) containing protease inhibitors (Beyotime, P1010). Protein concentration was determined by the BCA method (Thermo Fisher, 23227), and 20 µg of protein per sample was loaded for electrophoresis. Separation was performed using 12% SDS-PAGE gels, and proteins were transferred onto PVDF membranes (Millipore, IPVH00010). Membranes were blocked with 5% skim milk for 1 h at room temperature and then incubated overnight at 4 °C with the following primary antibodies: anti-p21^Cip1/Waf1^ (Cell Signaling Technology, #2947, 1:1000), anti-p16^INK4a^ (Abcam, ab211542, 1:1000), and anti-β-actin (Proteintech, 66009-1-Ig, 1:5000) as the loading control. After washing, membranes were incubated with HRP-conjugated secondary antibodies (Beyotime, A0208, 1:5000) for 1 h at room temperature. Protein bands were visualized using an ECL detection system (Thermo Scientific, 32106). Band intensity was quantified using ImageJ software, and relative protein expression levels were normalized to β-actin.

Exosome isolation and characterization

Exosomes were isolated from the culture supernatants of both control and UVB-induced senescent HDFs. When cells reached 80–90% confluence, they were gently washed twice with calcium- and magnesium-free PBS (Gibco, 10010-023) and then cultured for 24 h in DMEM (Gibco, 11965-092) supplemented with exosome-depleted FBS (System Biosciences, EXO-FBS-250 A-1). The conditioned medium was collected and immediately processed at 4 °C. For each isolation, approximately 30 mL of supernatant was subjected to sequential centrifugation steps: 300 × g for 10 min to remove dead cells; 2,000 × g for 20 min to remove cell debris; and 10,000 × g for 30 min to remove large vesicles and apoptotic bodies. The supernatant was then filtered through a 0.22 μm pore filter (Millipore, SLGP033RB) and ultracentrifuged at 100,000 × g for 70 min at 4 °C using a Beckman Optima L-100 K ultracentrifuge (Rotor Type 70Ti, Beckman Coulter, USA) to pellet exosomes. The pellet was resuspended in sterile PBS (pH 7.4) and subjected to a second ultracentrifugation under the same conditions to remove non-specific protein contaminants. The final exosome pellet was resuspended in 50–100 µL of sterile PBS. All samples were aliquoted and stored at -80 °C to avoid freeze-thaw cycles. To validate the isolation method, some samples were processed in parallel using a commercial exosome isolation reagent (Thermo Fisher Scientific, Total Exosome Isolation Reagent, 4478359), yielding consistent results and confirming the reliability of the protocol. Exosome protein concentration was determined using a BCA protein assay kit (Thermo Fisher Scientific, 23227), and concentrations were maintained between 1 and 2 µg/µL for all experiments.

The morphological structure and purity of the isolated exosomes were characterized using multiple techniques. Firstly, exosome morphology was examined by transmission electron microscopy (TEM, JEOL JEM-1400, Japan). A 10 µL aliquot of exosome suspension was applied to a 200-mesh carbon-coated copper grid, allowed to adsorb for 1 min at room temperature, negatively stained with 2% phosphotungstic acid, air-dried, and observed under the microscope at an accelerating voltage of 80 kV. The samples exhibited the typical cup-shaped or round bilayer membrane vesicle structure, with diameters ranging from approximately 30–150 nm. Subsequently, nanoparticle tracking analysis (NTA) was performed using a ZetaView PMX110 system (Particle Metrix, Germany) to determine particle size distribution and concentration. Samples were diluted 1:1000 in PBS before measurement. The camera shutter speed was set to 70 µs, temperature to 25 °C, and 11 fields of view were automatically captured and analyzed to calculate the average particle size (mode size) and Zeta potential, assessing particle homogeneity and dispersion.

To confirm the presence of exosome marker proteins, total exosome protein was extracted and analyzed by Western blot. Specific exosome marker proteins TSG101 (Abcam, ab125011, 1:1000), CD9 (Cell Signaling Technology, #13174, 1:1000), and CD81 (Abcam, ab79559, 1:1000) showed strong positive signals, while the endoplasmic reticulum marker Calnexin (Abcam, ab22595, 1:1000) was undetectable, indicating high exosome purity without organelle contamination. All detections were independently repeated three times with consistent results.

To assess the uptake capability of exosomes by recipient cells, exosomes were labeled with the lipophilic fluorescent dye PKH26 (Sigma-Aldrich, MINI26). Exosomes were mixed with the dye working solution (1:250 dilution) and incubated at room temperature in the dark for 5 min. An equal volume of 1% BSA was added to stop the reaction, and free dye was removed by ultracentrifugation (100,000 × g, 70 min, 4 °C). The labeled exosomes were resuspended in PBS and added to the HaCaT cell culture system at a final exosome protein concentration of 20 µg/mL, followed by incubation for 12 h at 37 °C. After incubation, cells were washed three times with PBS, fixed with 4% paraformaldehyde (Solarbio, P1110) for 15 min, and nuclei were stained with DAPI (Beyotime, C1002) for 5 min. The distribution of red fluorescent signals within the cells was observed using a laser scanning confocal microscope (Leica TCS SP8, Germany).

Gene interference and pharmacological treatments

To investigate the functional roles of LDHA and ACSL4 in exosome-mediated ferroptosis and photoaging, a combination of small interfering RNA (siRNA) and pharmacological inhibitors was employed. Target-specific siRNAs against LDHA and ACSL4 were custom-synthesized by GenePharma (Shanghai, China), desalted, purified, and their sequence specificity verified by BLAST analysis. A non-targeting scrambled siRNA sequence was used as the negative control (siNC) to exclude non-specific effects. Cell transfection was performed using Lipofectamine 3000 transfection reagent (Invitrogen, L3000015) according to the manufacturer’s instructions, preparing complexes at a ratio of 6 µL Lipofectamine 3000 to 2.5 µg siRNA per well. HDFs and HaCaT cells were seeded in 6-well plates at a density of 2.5 × 10⁵ cells per well and transfected at 70–80% confluence. Transfection was carried out in antibiotic-free medium with a final siRNA concentration of 50 nM. After 6 h of incubation with the transfection complexes, the medium was replaced with complete medium containing 10% FBS, and cells were cultured for an additional 24–48 h. The negative control group received an equivalent amount of non-targeting siRNA. Transfection efficiency was preliminarily confirmed (> 80%) using fluorescently labeled siRNA, and knockdown efficacy was verified in formal experiments by quantitative real-time PCR (qRT-PCR) and Western blot analysis of target gene mRNA and protein expression levels.

For pharmacological interventions, to dissect the causal relationship between lactate metabolism and ferroptosis signaling, the following reagents were used: the LDHA inhibitor FX11 (Selleck Chemicals, S7522), the ferroptosis inhibitor Ferrostatin-1 (Selleck Chemicals, S7243), and the metabolic rescue agent sodium lactate (Sodium L-lactate, Sigma-Aldrich, L7022). FX11 and Ferrostatin-1 were dissolved in anhydrous dimethyl sulfoxide (DMSO, Sigma-Aldrich, D2650) to prepare 10 mM stock solutions, which were then diluted to working concentrations in culture medium immediately before use, ensuring the final DMSO concentration did not exceed 0.1%. FX11 was applied at 10 µM for 24 h to inhibit lactate production and LDHA activity. Ferrostatin-1 was used at 1 µM for 24 h to block lipid peroxidation associated with ferroptosis. Sodium lactate was dissolved in serum-free DMEM and used at a final concentration of 10–20 mM for 24 h to simulate lactate accumulation and validate the reversibility of the lactate–histone lactylation–ferroptosis signaling axis. To further investigate the role of ROS in ferroptosis, cells were pre-treated with the antioxidant N-acetylcysteine (NAC, 5 mM) for 2 h before exposure to exosomes, and maintained throughout the exosome incubation period. All experiments included vehicle control groups (equivalent DMSO concentration) to account for potential solvent effects. Cell morphology and condition were monitored in real-time using an inverted microscope (Olympus IX73, Japan) during transfection and drug treatments to ensure no significant morphological damage or cell detachment occurred. To minimize batch variations, all experiments used cells in the logarithmic growth phase and were conducted with the same batch of culture media and under identical culture conditions. Each treatment was independently repeated three times (biological replicates N = 3, with technical replicates n ≥ 3).

Western blot analysis

Total protein was extracted from both cellular and animal tissue samples using RIPA lysis buffer (Solarbio, R0010, China) supplemented with a protease and phosphatase inhibitor cocktail (Beyotime, P1045, China). Samples were lysed on ice for 30 min, followed by centrifugation at 12,000 × g for 15 min at 4 °C to remove cellular debris. The resulting supernatant was collected as the protein sample. Protein concentration was determined using a BCA protein assay kit (Thermo Fisher Scientific, 23227, USA), and samples were uniformly diluted with lysis buffer based on the quantification results. Aliquots containing 20–40 µg of protein per sample were subjected to SDS-PAGE electrophoresis. Electrophoresis was performed using 10–12% separating gels on a Mini-PROTEAN Tetra Cell electrophoresis system (Bio-Rad, USA) at a constant voltage of 120 V for approximately 90 min. Subsequently, proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, IPVH00010, USA) using a Trans-Blot Turbo transfer system (Bio-Rad, USA). The transfer was conducted under ice-cold conditions at a constant current of 200 mA for 90 min.

After transfer, membranes were blocked for 1 h at room temperature to minimize non-specific binding. The blocking solution was chosen based on the target protein: 5% skim milk (Solarbio, D8340) for general cytoplasmic or membrane proteins, and 5% bovine serum albumin (BSA, Sigma-Aldrich, A7906) for histones and their modifications.

The blocked membranes were then incubated overnight at 4 °C with gentle shaking in primary antibody solutions. The primary antibodies used included: LDHA (Cell Signaling Technology, #3582, 1:1000), H3K18la (PTM Biolabs, PTM-1408, 1:1000), γ-H2AX (Cell Signaling Technology, #9718, 1:1000), p21^Cip1/Waf1^ (Abcam, ab109199, 1:1000), p16^INK4a^ (Abcam, ab211542, 1:1000), ACSL4 (Abcam, ab155282, 1:1000), GPX4 (Abcam, ab125066, 1:1000), SLC7A11 (Cell Signaling Technology, #12691, 1:1000), β-actin (Proteintech, 66009-1-Ig, 1:5000), and H3 (Abcam, ab1791, 1:2000). Histone H3 was used as the loading control for histone lactylation modifications, while β-actin served as the loading control for cytoplasmic proteins.

The following day, membranes were washed three times for 10 min each with Tris-buffered saline containing 0.1% Tween-20 (TBST, pH 7.6). They were then incubated for 1 h at room temperature with horseradish peroxidase (HRP)-conjugated goat anti-rabbit or anti-mouse secondary antibodies (Cell Signaling Technology, #7074 and #7076, respectively, diluted 1:5000). After washing, protein bands were visualized using an ECL chemiluminescence detection reagent (Thermo Fisher Scientific, 32106), and signals were captured using a Bio-Rad ChemiDoc MP imaging system. To ensure signals fell within the linear detection range, multiple exposure times (ranging from 10 to 90 s) were applied for each membrane, and non-saturated images were selected for quantitative analysis. Band intensity was quantified using ImageJ software (National Institutes of Health, version 1.53) with the “Analyze → Gel → Plot Lanes” function. The signal intensity of the target protein was normalized to its corresponding loading control to obtain the relative expression level (fold change). All experiments were independently repeated at least three times (N = 3), and the average values were used for statistical analysis. To ensure data accuracy and reproducibility, each batch of assays included positive controls (standard cell lysates from the same batch) and negative controls (no primary antibody incubation), simultaneously verifying antibody specificity and the linear detection range.

Lactate content measurement

The lactate content in cell and tissue samples was measured using a colorimetric lactate assay kit (Nanjing Jiancheng Bioengineering Institute, A019-2-1, China). For cell collection, after washing twice with serum-free PBS (Gibco, 10010-023), cells were scraped and lysed in pre-cooled saline or PBS at a ratio of 200 µL per 10⁶ cells. Tissue samples (approximately 50 mg) were homogenized in 0.5 mL of extraction buffer using a TissueLyser II (Qiagen, Germany), followed by lysis on ice for 30 min and centrifugation at 12,000 × g for 10 min at 4 °C. The resulting supernatant was collected for assay. All samples were assayed immediately after preparation or stored short-term at -80 °C if measurement on the same day was not possible.

The assay procedure was performed according to the manufacturer’s instructions. The reaction was carried out in a 37 °C water bath for 10 min, and the absorbance was measured at a wavelength of 530 nm using a microplate reader (Bio-Rad, Model 680, USA). A standard curve (y = kx + b, R² ≥ 0.99) was generated using the standard lactate solutions (0–5 mmol/L) provided in the kit, and the lactate concentration of each sample was calculated accordingly. Results are expressed as mmol of lactate per gram of protein (mmol/g protein). The total protein content of the samples was determined using a BCA protein assay kit (Thermo Fisher Scientific, 23227, USA) for normalization. To minimize experimental error, all measurements included blank wells, standard wells, positive control wells, and sample replicate wells (with technical replicates n = 3 per group). All experiments were independently repeated at least three times (N = 3). The instrument was zero-calibrated before detection, and background absorbance from cell-free samples in the reaction system was subtracted. Data from three independent experiments were averaged for subsequent statistical analysis.

Cell viability and senescence detection

Cell viability was assessed using the Cell Counting Kit-8 (CCK-8, Dojindo, CK04-11, Japan). HDFs and HaCaT cells were seeded in 96-well plates at a density of 5 × 10³ cells per well and cultured for 24 h before respective treatments. After treatment, 10 µL of CCK-8 working solution was added to each well, and the plates were incubated in a 37 °C, 5% CO₂ incubator for 2 h. Following incubation, the absorbance at 450 nm (OD₄₅₀) was measured using a microplate reader (Bio-Rad, Model 680, USA). Wells without cells served as the background control, and three technical replicates were set up for each condition. Cell survival rate is expressed as the percentage ratio of the absorbance of the treatment group to that of the control group, or relative proliferative activity was calculated after normalization. For cell death inhibitor analysis, recipient HaCaT cells were pretreated with Ferrostatin-1 (Fer-1, 10 µM), Z-VAD-FMK (20 µM), Necrostatin-1 (30 µM), or Chloroquine (20 µM) for 24 h prior to incubation with UVB-induced exosomes. Cell viability was measured using the CCK-8 following the manufacturer’s protocol.

Cellular senescence levels were detected using the Senescence-Associated β-Galactosidase (SA-β-Gal) staining method with a cellular senescence detection kit (Cell Signaling Technology, #9860, USA), following the manufacturer’s protocol. After cells were seeded in 6-well plates and treatments completed, they were gently washed twice with PBS (pH 7.4), fixed with 1× fixative solution for 15 min at room temperature, washed again with PBS, and then incubated with freshly prepared SA-β-Gal staining working solution (containing X-gal substrate, pH 6.0). The plates were incubated in a humidified, CO₂-free incubator at 37 °C in the dark for 12–16 h. After incubation, the staining solution was discarded, and cells were washed three times with PBS. Images were then observed and captured using an inverted microscope (Olympus IX73, Japan). Five non-overlapping fields per well were randomly selected for counting positive cells. Cells were considered positive if their cytoplasm exhibited distinct blue-green precipitation. The percentage of positive cells was calculated as: (Number of SA-β-Gal positive cells / Total number of cells) × 100%.

All experiments were conducted with three independent replicates (biological replicates N = 3, with technical replicates n ≥ 3 each time). Results are presented as mean ± standard deviation (SD). To avoid observer bias, the analysis of staining results was independently performed by two investigators under blinded conditions, and their average counts were used.

Assessment of ferroptosis markers

Lipid peroxidation levels were assessed using the lipid peroxidation-sensitive fluorescent probe C11-BODIPY 581/591 (Thermo Fisher Scientific, D3861, USA). After treatments, cells were washed twice with PBS and incubated with 10 µM C11-BODIPY working solution at 37 °C under 5% CO₂ in the dark for 30 min. Following incubation, cells were washed twice with PBS, collected by centrifugation (1,000 × g, 5 min), and resuspended in 500 µL PBS for flow cytometric analysis. Analysis was performed on a BD FACSAria II flow cytometer (BD Biosciences, USA) with an excitation wavelength of 488 nm. Fluorescence signals were collected in the FITC channel (oxidized form, Em 530 ± 15 nm) and PE channel (reduced form, Em 590 ± 20 nm). The ratio of oxidized-to-reduced fluorescence reflects the degree of intracellular lipid peroxidation. A minimum of 1 × 10⁴ events were acquired per sample, and data were analyzed using FlowJo v10 software (TreeStar, USA). Positive controls (RSL3, 1 µM, Selleck, S8155) and negative controls (Ferrostatin-1, 1 µM, Selleck, S7243) were included to validate assay specificity.

Intracellular free divalent iron (Fe²⁺) levels were measured using the FerroOrange probe (Dojindo Laboratories, F374, Japan). After washing with PBS, cells were incubated with 1 µM FerroOrange working solution at 37 °C in the dark for 30 min, followed by two PBS washes. Fluorescence intensity was observed under a fluorescence microscope (Olympus IX73, Japan) or recorded using a microplate reader at Ex/Em = 543/580 nm. Signals were normalized to total protein content or cell number, and results are expressed as Relative Fluorescence Units (RFU/10⁶ cells). Malondialdehyde (MDA) levels and Superoxide Dismutase (SOD) activity were measured using commercial kits from the Nanjing Jiancheng Bioengineering Institute (MDA: A003-1-2; SOD: A001-3-2, China). MDA was quantified by the thiobarbituric acid (TBA) method, measuring absorbance at 532 nm. SOD activity was determined using the xanthine oxidase method, measuring absorbance at 550 nm. Three technical replicates were set up per sample group. Data were normalized to protein concentration and expressed as nmol MDA per mg protein (nmol/mg protein) or units of SOD activity per mg protein (U/mg protein). Protein expression of Glutathione Peroxidase 4 (GPX4) and the cystine/glutamate antiporter SLC7A11 was analyzed by Western blot, as described in Sect.  2.4. Antibodies used were: GPX4 (Abcam, ab125066, 1:1000), SLC7A11 (Cell Signaling Technology, #12691, 1:1000), with β-actin (Proteintech, 66009-1-Ig, 1:5000) serving as the loading control.

All assays were independently repeated three times (N = 3). Each experiment included positive induction controls, negative inhibition controls, and vehicle control groups to verify the stability and specificity of the results. Data are presented as the mean values from independent experiments, and results normalized to the control group were used for subsequent statistical analysis.

RNA sequencing and bioinformatic analysis

Total RNA was extracted from cells of the Exo and Exo + siLDHA groups using TRIzol reagent (Invitrogen, 15596026, USA) according to the manufacturer’s instructions. Three independent biological replicates were prepared per group. RNA sample purity was assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA), requiring A₂₆₀/A₂₈₀ ratios between 1.8 and 2.1 and A₂₆₀/A₂₃₀ ratios greater than 2.0. RNA integrity was evaluated with an Agilent 2100 Bioanalyzer (Agilent Technologies, USA), and only samples with RNA Integrity Number (RIN) greater than 7.0 were used for library construction. RNA libraries were prepared using the NEBNext Ultra™ RNA Library Prep Kit for Illumina (New England Biolabs, E7770, USA). Following mRNA enrichment (using Oligo(dT) magnetic beads) or rRNA depletion, RNA was fragmented into 200–300 bp pieces. Subsequent steps included cDNA synthesis, end repair, adenylation, adapter ligation, and PCR amplification. The concentration and fragment size distribution of the final libraries were quality-controlled using Qubit 2.0 (Thermo Fisher Scientific) and Agilent 2100 Bioanalyzer. Qualified libraries were subjected to high-throughput paired-end sequencing (PE150, paired-end 150 bp) on an Illumina NovaSeq 6000 platform, performed by Novogene Co., Ltd. (Beijing, China). The sequencing depth was at least 6 Gb of raw data per sample.

Raw sequencing reads underwent quality control using FastQC (v0.11.9). Adapter sequences and low-quality bases (Q < 20) were trimmed using Trimmomatic (v0.39). The resulting clean reads were aligned to the human reference genome hg38 (GENCODE release 42) using the STAR aligner (v2.7.9a), achieving an overall alignment rate exceeding 95%. Transcript expression levels were normalized as Fragments Per Kilobase of transcript per Million mapped reads (FPKM) or Transcripts Per Million (TPM). Differential expression analysis was performed using DESeq2 (v1.38.3) in the R environment (v4.3.1). Genes with an absolute log2 fold change (|log₂FC|) ≥ 1 and a False Discovery Rate (FDR) < 0.05 were considered differentially expressed genes (DEGs). DEGs were visualized using volcano plots and hierarchical clustering heatmaps to display gene expression profile changes. Functional enrichment analysis, including Gene Ontology (GO) classification and Kyoto Encyclopedia of Genes and Genomes (KEGG, release 2023) pathway enrichment, was performed using the ClusterProfiler package (v4.8.1). Among the enriched pathways, particular focus was placed on the ferroptosis-related pathway (KEGG pathway: hsa04216). Key regulatory molecules significantly upregulated in the ferroptosis signaling cascade were identified by comparing DEG sets and sorting by Fold Change. Among these, ACSL4 was identified as a core target gene. All sequencing data were uploaded to the National Genomics Data Center (CNCB-NGDC) or the Gene Expression Omnibus (GEO) public database after internal validation. The accession numbers will be released upon formal publication.

Chromatin immunoprecipitation (ChIP-qPCR)

Chromatin Immunoprecipitation (ChIP) assays were performed using the EZ-ChIP™ Chromatin Immunoprecipitation Kit (Millipore, 17–371, USA). Cells were washed twice with PBS and cross-linked with 1% formaldehyde (Sigma-Aldrich, F8775) for 10 min at room temperature to fix DNA-protein interactions. The cross-linking reaction was quenched by adding 125 mM glycine (Sigma-Aldrich, G7126) for 5 min. Cells were washed three times with PBS, collected, and lysed using lysis buffer containing protease inhibitors (Beyotime, P1045). The resulting chromatin was sonicated using a Bioruptor Plus sonicator (Diagenode, Belgium) set to high power mode, with cycles of 30 s on/30 seconds off, repeated 10–15 times, to obtain DNA fragments of approximately 200–500 bp. The efficiency of sonication was verified by 1.5% agarose gel electrophoresis. Samples were centrifuged at 12,000 × g for 10 min at 4 °C to remove debris, and the supernatant was used for the immunoprecipitation reaction.

Approximately 50 µg of chromatin protein was used per immunoprecipitation reaction. Immunoprecipitation was performed using either an anti-H3K18la antibody (PTM Biolabs, PTM-1408, 5 µg per reaction) or normal rabbit IgG (Cell Signaling Technology, #2729, 5 µg per reaction) as a control. Antibodies were incubated with Protein A/G Magnetic Beads (Thermo Fisher Scientific, 26162) overnight at 4 °C with rotation to form immunocomplexes. The immunocomplexes were sequentially washed twice with low salt wash buffer, high salt wash buffer, LiCl wash buffer, and TE buffer to remove non-specifically bound material. Bound chromatin was eluted using 1% SDS elution buffer (containing 0.1 M NaHCO₃) and reverse-cross-linked at 65 °C for 8 h to release DNA. The reverse-cross-linked products were sequentially digested with RNase A (Thermo Fisher Scientific, EN0531) and Proteinase K (Thermo Fisher Scientific, AM2546), and the DNA was purified using a ChIP DNA Purification Kit (Zymo Research, D5205, USA).

The purified DNA samples were used to quantitatively analyze the enrichment of histone H3K18 lactylation modification at the ACSL4 promoter region (− 500 bp to + 100 bp). qPCR reactions were set up using SYBR Green Master Mix (Takara, RR420A, Japan) and performed on a QuantStudio 6 Flex Real-Time PCR System (Applied Biosystems, USA). The specific primers for the ACSL4 promoter region were: forward 5′-AGCTGTTCTTCTTGAAGCTG-3′ and reverse 5′-TGCTGAATCCTCTGTTTCCTT-3′. The qPCR program consisted of an initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s. Amplification specificity was confirmed by melt curve analysis. Data analysis was performed using the percent input method, where %Input = 2^(Ct < sub> Input</sub > − Ct < sub> IP</sub> ) × dilution factor. The fold enrichment over the IgG control was also calculated to assess binding specificity. Three technical replicates were set up for each sample, and three independent biological replicates (N = 3) were performed.

Dual-luciferase reporter assay

To verify the regulatory effect of LDHA-mediated histone H3K18 lactylation on ACSL4 transcriptional activity, the promoter region of the ACSL4 gene (− 1000 bp to + 200 bp relative to the transcription start site) was cloned into the multiple cloning site (MluI/BglII) of the pGL3-Basic vector (Promega, E1751, USA), resulting in the pGL3-ACSL4-Luc reporter plasmid. The sequence of the recombinant plasmid was verified by Sanger sequencing (Tsingke Biotechnology, Beijing, China). The empty pGL3-Basic vector served as the negative control, and the pRL-TK vector (Promega, E2241, USA), expressing Renilla luciferase, was used as an internal control for normalization.

HaCaT cells were seeded in 24-well plates at a density of 1.5 × 10⁵ cells per well and transfected at 70–80% confluence. Transfection was performed using Lipofectamine 3000 transfection reagent (Invitrogen, L3000015, USA) according to the manufacturer’s instructions. Cells were co-transfected in serum-free medium with pGL3-ACSL4-Luc (500 ng) and pRL-TK (50 ng). Intervention groups were simultaneously treated with corresponding siRNA (e.g., siLDHA) or drugs (e.g., FX11, sodium lactate) to assess changes in transcriptional activity. After 6 h, the medium was replaced with complete medium containing 10% FBS, and cells were cultured for an additional 24 h.

Luciferase activity was measured using the Dual-Luciferase^® Reporter Assay System (Promega, E1910, USA) following the manufacturer’s protocol. Cells were harvested and lysed in Passive Lysis Buffer (PLB) with shaking at room temperature for 15 min, followed by centrifugation (12,000 × g, 1 min). The supernatant was collected for assay. Luminescence signals were read on a GloMax Discover multimode microplate detection system (Promega, USA). Firefly luciferase activity was measured first, followed by Renilla luciferase activity. Three technical replicates were set up per sample group. The Firefly luciferase signal was normalized to the Renilla luciferase signal to calculate the relative luciferase activity (Firefly/Renilla ratio).

All experiments were independently repeated at least three times (N = 3). Relative promoter activity was calculated by normalizing to the control group results to ensure data comparability and reproducibility.

In vivo photoaging animal model

Male C57BL/6 mice (8 weeks old, weight 20 ± 2 g, n = 8 per group) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (China; License No. SCXK(Jing)2021-0006). All animals were housed in a barrier-level animal facility under specific pathogen-free (SPF) conditions, maintained at a constant temperature (22 ± 2 °C), relative humidity (50 ± 10%), and a 12-hour light/dark cycle, with free access to sterile food and water. All animal experiments were approved by the Animal Ethics Committee of Shandong University (Approval No.: 2024-ACUC-0176) and strictly conducted in accordance with the “Regulations for the Administration of Affairs Concerning Experimental Animals” of China and the ARRIVE 2.0 guidelines.

Mice were randomly divided into three groups (n = 8 per group): Control group, Exosome group, and Exosome + siLDHA group. Prior to experimentation, the dorsal hair was gently shaved using an electric shaver and allowed to recover naturally for 24 h. The photoaging model was induced by UVB irradiation using a UVB light source with a primary peak wavelength of 312 nm (Vilber Lourmat, VL-215.M, France). The irradiation intensity was calibrated to 180 mJ/cm² using a UV radiometer (Sigma, UV-340 A). Irradiation was performed at a distance of 20 cm, three times per week for 8 consecutive weeks. Non-irradiated areas (head and abdomen) were shielded during irradiation to prevent systemic damage. Control group mice were exposed to the same environmental conditions for identical durations but did not receive UVB irradiation. Exosomes were derived from in vitro cultured HDFs and from senescent fibroblasts treated with siLDHA. All exosomes were purified by ultracentrifugation (100,000 × g, 70 min, 4 °C) and resuspended in PBS before use. The injection dose was 100 µg of exosome protein per mouse per injection, in a volume of 100 µL. The Exosome group and Exosome + siLDHA group received injections via tail vein or multiple subcutaneous injections on the back, twice weekly for 8 weeks. The Control group received an equal volume of PBS. Exosome solutions were stored short-term at 4 °C and used within 72 h of preparation to ensure bioactivity.

At the end of the experiment, mice were anesthetized (sodium pentobarbital, 50 mg/kg, intraperitoneal injection) and euthanized. Dorsal skin tissues were collected. A portion of the tissue was fixed in 4% paraformaldehyde for 24 h, followed by dehydration, paraffin embedding, and sectioning into 4 μm thick slices for Hematoxylin and Eosin (H&E) staining and Masson’s trichrome staining (Solarbio, G1340). Another portion of fresh skin tissue was rapidly frozen in liquid nitrogen and stored at -80 °C for protein extraction and biochemical assays. The expression levels of key proteins including LDHA, γ-H2AX, H3K18la, ACSL4, GPX4, and SLC7A11 were analyzed by Western blot. Lactate content, Fe²⁺ concentration, MDA levels, and lipid peroxidation (ROS) were measured using the methods described in Sect.  2.5 and 2.7, respectively. Throughout the experimental period, mouse body weight, hair condition, and general activity were closely monitored. No significant drug toxicity or excessive skin damage due to UVB irradiation was observed. All animal experiments ensured a minimum sample size meeting statistical power requirements (power ≥ 0.8), and euthanasia was performed according to established ethical principles post-experimentation.

Histology and immunohistochemistry

Skin tissues were fixed in 4% paraformaldehyde (Solarbio, P1110, China) for 24 h, followed by sequential dehydration through a graded ethanol series, clearing in xylene, and embedding in paraffin wax. Continuous sections of 4 μm thickness were cut using a microtome (Leica RM2235, Germany) and baked at 60 °C for 2 h. Sections were deparaffinized in xylene twice (10 min each), rehydrated through a graded ethanol series (100%, 95%, 85%, 75%, 5 min each), and rinsed with distilled water before staining. Routine Hematoxylin and Eosin (H&E) staining and Masson’s trichrome staining were performed according to the kit manufacturer’s instructions (Solarbio, G1340, China). For H&E staining, sections were stained with hematoxylin for 5 min, rinsed with water, differentiated in 1% acid alcohol for 5 s, blued in alkaline solution for 30 s, and counterstained with 0.5% eosin for 3 min. Masson’s trichrome staining involved nuclear staining with Weigert’s iron hematoxylin for 10 min, cytoplasmic staining with Ponceau S-acid fuchsin for 5 min, differentiation with phosphomolybdic acid for 5 min, and collagen fiber staining with aniline blue for 5 min. After staining, sections were dehydrated, cleared, and mounted with neutral balsam. Tissue structure and collagen deposition were observed and photographed under an optical microscope (Olympus BX53, Japan).

Immunohistochemistry (IHC) was performed according to standard protocols. After deparaffinization and rehydration, antigen retrieval was carried out by microwave heating in sodium citrate buffer (pH 6.0, Solarbio, P0081, China) at 95 °C for 15 min, followed by natural cooling to room temperature. Sections were washed three times with PBS and incubated with 3% H₂O₂ at room temperature for 10 min to block endogenous peroxidase activity. Subsequently, sections were blocked with 5% goat serum (Beyotime, C0265, China) at room temperature for 30 min to reduce non-specific binding. Sections were then incubated with respective primary antibodies in a humidified chamber at 4 °C overnight. Primary antibodies included: LDHA (Cell Signaling Technology, #3582, 1:200), γ-H2AX (Cell Signaling Technology, #9718, 1:200), H3K18la (PTM Biolabs, PTM-1408, 1:200), and ACSL4 (Abcam, ab155282, 1:200). The next day, after washing three times with PBS, sections were incubated with an HRP-conjugated goat anti-rabbit secondary antibody (ZSGB-BIO, PV-6001, China) at room temperature for 30 min. Color development was performed using a DAB substrate kit (ZSGB-BIO, ZLI-9018, China) for 2–5 min, and the reaction was stopped by rinsing with distilled water. Nuclei were counterstained with hematoxylin for 30 s. Sections were then dehydrated through a graded ethanol series, cleared in xylene, and mounted.

Images were captured under an Olympus BX53 optical microscope (Japan) using identical exposure conditions. Five non-overlapping fields were randomly selected per section. Positive signals appeared as brownish-yellow deposits in the cytoplasm or nucleus. Quantitative image analysis was performed using Image-Pro Plus 6.0 software (Media Cybernetics, USA) to measure the positive area and mean optical density (MOD). Results are expressed as integrated optical density (IOD) per unit area (µm²). At least three independent tissue sections from different mice were analyzed per group.

Statistical analysis

All experiments were independently repeated at least three times (biological replicates N = 3), with each experiment including a minimum of three technical replicates (n ≥ 3). Data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software, USA) or SPSS Statistics 27.0 (IBM, USA). Data were first tested for normality (Shapiro-Wilk test) and homogeneity of variances (Levene’s test). For comparisons between two groups meeting both normality and homogeneity of variances, a two-tailed Student’s t-test was used. Comparisons among multiple groups were analyzed by one-way analysis of variance (ANOVA), followed by Tukey’s multiple comparison test for post-hoc analysis. If data did not meet the assumptions of normality or equal variances, the non-parametric Kruskal-Wallis test was employed. The significance level was set at p < 0.05, and all tests were two-tailed. Error bars in statistical graphs represent SD. For key results, exact p-values, sample sizes (n), and effect sizes are reported in the figure legends. When performing multiple comparisons, the Benjamini-Hochberg method was applied where necessary to control the False Discovery Rate (FDR). No samples meeting the experimental criteria were excluded during data analysis, and all statistical methods were pre-specified during the experimental design phase to ensure the reproducibility and statistical reliability of the results.

Results

Expression and characterization of LDHA in exosomes from UVB-induced senescent foreskin fibroblasts

Identification of UVB-induced senescence in human foreskin fibroblasts and associated exosomal LDHA with epigenetic clues

To initially confirm the successful establishment of the model, we compared typical senescence phenotypes between control and UVB-treated HDFs. Figure 1A shows that after UVB exposure, cells transitioned from a spindle shape to a larger, flatter morphology, with apparent SA-β-Gal-positive blue staining in the cytoplasm; these concurrent morphological and staining changes align with classical senescence markers. To avoid reliance on a single indicator, we further examined cell cycle inhibition-related proteins. In Fig. 1B, the bands for p21 and p16 were markedly stronger in the UVB group compared to the control, corroborating the SA-β-Gal results in Fig. 1A and indicating that UVB sufficiently induces a robust senescence program rather than a transient stress response.

Fig. 1.

Identification of UVB-induced senescence in human foreskin fibroblasts and their senescence-associated exosomes. A SA-β-galactosidase staining shows that after UVB irradiation, foreskin fibroblasts transitioned from a spindle shape to a flatter, enlarged morphology, with markedly enhanced blue positive signals, indicating a significant increase in cellular senescence levels. B Western blot analysis of the expression of senescence markers p21 and p16. Protein levels of both were significantly upregulated in the UVB group, indicating successful induction of cellular senescence by UVB. C Transmission electron microscopy (TEM) images showing the morphology of exosomes. Exosomes from both the control (Exo) and senescent (Exo-UV) groups exhibited the typical bilayer membrane vesicle structure, with diameters of approximately 30–150 nm. D Nanoparticle tracking analysis (NTA) showed that the particle size distribution of both Exo and Exo-UV peaked at around 120 nm, consistent with the typical size of exosomes. E Western blot analysis of exosomal marker proteins TSG101, CD9, and CD81, and the endoplasmic reticulum contamination marker Calnexin. Results showed positive signals for TSG101, CD9, and CD81, but negative for Calnexin, indicating high purity of the isolated exosomes. F Western blot analysis of exosomal LDHA protein expression. LDHA levels were higher in Exo-UV than in control exosomes, suggesting altered exosomal metabolic cargo under senescent conditions. G After uptake by HaCaT cells, PKH26 membrane dye-labeled exosomes showed red fluorescent signals distributed around the cell periphery and in the cytoplasm, indicating good membrane integrity and uptake capability of the exosomes. Scale bar = 50 μm. ***P < 0.001 compared to the control group

After confirming cellular senescence, we systematically characterized the physicochemical properties and purity of senescence-associated exosomes (Exo-UV). Transmission electron microscopy revealed that vesicles from both groups exhibited the typical cup-shaped/round bilayer membrane structure (Fig. 1C), while nanoparticle tracking analysis (NTA) showed particle size distributions concentrated in the hundred-nanometer range (Fig. 1D). These morphological and size-related physical evidences complement each other, indicating that the isolated particles fall within the typical size range for exosomes, excluding larger extracellular vesicles or apoptotic bodies. Besides, NTA and TEM analysis revealed that Exo-UV maintained a similar size distribution and morphological appearance compared to Control-Exo, with no significant differences in average diameter observed between the two groups. These data indicate that the elevated LDHA protein levels detected in Exo-UV are not a consequence of altered vesicle size, but rather reflect a specific enrichment of this metabolic enzyme within the exosomal cargo under UVB-induced stress.

To rule out organelle contamination and confirm exosomal identity, we detected marker proteins of membrane and endosomal origin. Figure 1E shows that TSG101, CD9, and CD81 were positive in both Exo and Exo-UV samples, while the endoplasmic reticulum marker Calnexin was absent. This combination of a “positive triad + negative internal reference” provides biochemical evidence for high-purity exosomes, consistent with the aforementioned physical characterization. Notably, during initial comparison of exosomal cargo, we observed higher levels of LDHA protein in Exo-UV compared to control exosomes (Fig. 1F). To further confirm that the accumulation of LDHA in recipient cells is derived from exosomal transport, the exosome secretion inhibitor GW4869 was employed to treat senescent fibroblasts. Western blot analysis revealed that GW4869 treatment significantly reduced LDHA levels within the isolated exosomes. Consistently, upon incubation with these secretion-inhibited exosomes, the protein levels of LDHA in recipient HaCaT cells were markedly decreased (Figure S1). These findings demonstrate that the elevated LDHA in recipient cells is primarily dependent on exosomal shuttling rather than endogenous induction. LDHA is a key rate-limiting enzyme in lactate metabolism. Its enrichment in exosomes provides a clue from the perspective of metabolic reprogramming in the source cells: under senescence, fibroblasts not only upregulate their own lactate metabolism but may also transmit this “metabolic instruction” to recipient cells via vesicles, creating the substrate conditions for subsequent histone lactylation and transcriptional remodeling. This phenomenon is logically consistent with the epigenetic changes we later observed in recipient cells (see the next section for mechanistic validation). Finally, to ensure the biological accessibility of these vesicles, we labeled exosome membranes with PKH26 and co-incubated them with HaCaT cells. Figure 1G shows red fluorescent particles distributed around the cell periphery and within the cytoplasmic region, indicating that both baseline exosomes (Exo) and senescence-associated exosomes (Exo-UV) could be effectively taken up by keratinocytes. Notably, no discernible difference in uptake efficiency was observed between the two groups, indicating that the subsequent phenotypic changes were independent of the total amount of exosome entry. The internalization of exosomes by recipient cells represents a fundamental prerequisite for their functional effects. It is suggested that the cargoes enriched in Exo-UV, such as LDHA, may contribute to the subsequent increase in intracellular lactate and epigenetic modifications following their uptake.

Morphological, biochemical, and physical evidence collectively demonstrate that UVB can stably induce senescence in human foreskin fibroblasts, yielding vesicles with typical exosomal characteristics and high purity. Furthermore, these vesicles are enriched with LDHA under senescent conditions and can be taken up by keratinocytes. These findings lay a solid foundation for subsequent mechanistic studies on how senescent fibroblasts transmit metabolic signals via exosomes to influence epigenetic transcription programs.

Upregulated LDHA in senescent fibroblast exosomes leads to lactate accumulation and promotes histone H3K18 lactylation

To investigate the role of exosomes from UVB-induced senescent fibroblasts in regulating lactate metabolism and histone lactylation, we compared the effects of control exosomes (Exo), senescent fibroblast-derived exosomes (Exo-UV), and LDHA-knockdown senescent exosomes (Exo-UV + siLDHA) on cellular metabolic status.

First, Western blot analysis revealed that Exo-UV treatment significantly increased LDHA protein levels in both fibroblasts-derived exosomes and recipient HaCaT cells, whereas the Exo-UV + siLDHA group showed a marked decrease in LDHA expression (Fig. 2A). This result indicates that exosomes can mediate the transcellular transport of LDHA, and LDHA enriched in exosomes secreted by senescent cells can be effectively taken up by recipient cells. The upregulation of LDHA implies activation of the lactate production pathway, providing substrate support at the metabolic level for subsequent histone lactylation.

Fig. 2.

Senescent fibroblast exosomal LDHA promotes lactate accumulation and enhances histone H3K18 lactylation. A Western blot analysis of LDHA protein expression in fibroblast-derived exosomes and recipient HaCaT cells. Exo-UV treatment significantly upregulated LDHA, whereas LDHA-knockdown exosomes (Exo-UV + siLDHA) markedly reduced LDHA levels. B Lactate quantification analysis showed that Exo-UV significantly increased lactate concentration in both fibroblasts and HaCaT cells, while the Exo-UV + siLDHA group showed decreased lactate levels, indicating that exosomal LDHA possesses activity promoting lactate generation. C Western blot analysis of histone H3K18 lactylation modification. Exo-UV treatment significantly enhanced the H3K18la signal, while the signal was weakened in the Exo-UV + siLDHA group, with no significant change in total H3 protein. ***P < 0.001 compared to the control group (Exo); ###P < 0.001 compared to the Exo-UV group

To verify this hypothesis and investigate whether the epigenetic regulatory function of Exo-UV is a generalized phenomenon in the skin microenvironment, we utilized both HaCaT cells (representing the epidermis) and primary fibroblasts (representing the dermis) as recipient cells. We further examined changes in intracellular lactate levels (Fig. 2B). The results showed that Exo-UV treatment significantly increased lactate concentration in recipent fibroblasts and HaCaT cells (p < 0.001), while in the Exo-UV + siLDHA group where LDHA was knocked down, lactate levels decreased significantly compared to the Exo-UV group (p < 0.001). This trend is consistent with the variations in LDHA protein expression, suggesting that exosome-derived LDHA remains functionally active upon delivery to recipient cells and contributes to intracellular lactate accumulation. As a primary substrate for histone lactylation, the elevated lactate level not only reflects a metabolic shift but also represents a potential mechanism for initiating epigenetic transcriptional regulation. Based on these findings, we further detected the level of histone H3K18 lactylation (H3K18la) to verify whether exosome-mediated metabolic reprogramming is accompanied by epigenetic modifications. To identify the most sensitive epigenetic markers in this process, we screened several well-characterized lactylation sites, including H3K18la, H3K9la, H3K14la, H3K23la, and H4K27la. Western blot analysis revealed that H3K18la exhibited the most significant and consistent upregulation in recipient cells following Exo-UV treatment, whereas changes at other sites were less pronounced (Figure S2). Western blot results (Fig. 2C) showed that Exo-UV treatment led to a significant increase in H3K18la levels in both fibroblasts and HaCaT cells, while the H3K18la signal was markedly weakened in the Exo-UV + siLDHA group, independent of changes in total H3 protein. This result indicates that LDHA in senescent fibroblast exosomes not only promotes lactate accumulation but also drives the enhancement of histone lactylation modifications in recipient cells, with H3K18la being the predominant modification site.

UVB-induced senescent fibroblasts can establish a metabolism–epigenetics coupled pathway between donor and recipient cells by secreting LDHA-enriched exosomes. Exosomal LDHA promotes increased lactate production, subsequently upregulating H3K18la levels, thereby potentially regulating the expression program of downstream genes. This “LDHA–Lactate–Histone Lactylation” axis provides a key clue for understanding the signaling mechanism of senescence-associated exosomes in skin photoaging.

Effects of senescent fibroblast exosomal LDHA on skin photoaging and ferroptosis in in vivo and in vitro models

Senescent fibroblast exosomal LDHA exacerbates photoaging phenotypes and triggers ferroptosis in an in vitro model

Under three conditions–control exosomes, senescent-derived exosomes, and LDHA-knockdown senescent exosomes – we simultaneously assessed damage signals, proliferative activity, and the ferroptosis pathway in foreskin fibroblasts and HaCaT cells. We first observed that exposure to senescent exosomes in recipient cells markedly enhanced the band intensity of the DNA damage marker γ-H2AX, while concurrently upregulating the two senescence suppressor proteins p16 and p21 (Fig. 3A, all increases compared to control exosomes, p < 0.001). When exosomal LDHA was knocked down, the intensity of these bands generally decreased to an intermediate level, suggesting that the activation of DNA damage and the senescence program is closely related to LDHA in the exosomes. Corresponding with the molecular readouts, cell viability significantly decreased in the CCK-8 assay, and LDHA knockdown partially restored viability (Fig. 3B, Exo-UV vs. Exo p < 0.001, Exo-UV+siLDHA vs. Exo-UV p < 0.001). At the morphological level, SA-β-Gal staining showed a significant increase in the proportion of senescence-positive cells, which was also attenuated by LDHA knockdown (Fig. 3C, Figure S3), thus forming a consistent evidence chain across morphology, function, and molecules.

Fig. 3.

Senescent fibroblast exosomal LDHA exacerbates photoaging phenotypes and induces ferroptosis activation in vitro. A Western blot analysis of DNA damage marker γ-H2AX and senescence-associated proteins p16 and p21 in foreskin fibroblasts and HaCaT cells. Exo-UV treatment significantly enhanced the levels of γ-H2AX, p16, and p21, while LDHA-knockdown exosomes (Exo-UV + siLDHA) reduced the expression of these markers, suggesting a close association between exosomal LDHA and cellular senescence/DNA damage. B Cell viability determined by CCK-8 assay. Cell viability was significantly reduced in the Exo-UV group and partially restored in the Exo-UV + siLDHA group, indicating LDHA involvement in exosome-mediated inhibition of cell activity. C SA-β-galactosidase staining showing changes in the proportion of senescence-positive cells. The Exo-UV group exhibited more blue-positive cells, which was attenuated by LDHA-knockdown exosomes. D Flow cytometric analysis of lipid peroxidation levels. Exo-UV significantly increased the proportion of lipid ROS-positive cells, which was markedly decreased in the Exo-UV + siLDHA group, indicating that LDHA promotes lipid peroxidation accumulation. E Measurement of intracellular free Fe²⁺ content showed that Exo-UV treatment significantly increased Fe²⁺ levels, which was partially reversed by LDHA-knockdown exosomes. F The end product of lipid peroxidation, MDA, was significantly increased in the Exo-UV group and decreased after LDHA knockdown. G Antioxidant enzyme SOD activity was significantly reduced in the Exo-UV group and partially restored by LDHA knockdown. H Western blot analysis of the ferroptosis inhibitory axis proteins GPX4 and SLC7A11. Both were downregulated in the Exo-UV group, and their expression levels were increased by LDHA-knockdown exosomes. Data represent results from at least three independent experiments. Error bars indicate standard deviation (SD). ***P < 0.001 compared to the control group (Exo); ###P < 0.001 compared to the Exo-UV group

Beyond metabolic alterations, we also assessed the impact on lipid oxidation and iron homeostasis, C11-BODIPY flow cytometry readings indicated a higher proportion of lipid ROS-positive populations in both cell types under the influence of senescent exosomes, and LDHA knockdown significantly reduced this proportion (Fig. 3D, Exo-UV vs. Exo p < 0.001, Exo-UV+siLDHA vs. Exo-UV p < 0.001). Synchronized with the oxidative load, intracellular free Fe²⁺ content significantly increased, and LDHA knockdown partially reversed this increase (Fig. 3E, all p < 0.001). The end product of lipid peroxidation, MDA, peaked in the senescent exosome group and decreased significantly after LDHA knockdown (Fig. 3F, Exo-UV vs. Exo p < 0.001, Exo-UV+siLDHA vs. Exo-UV p < 0.001). The antioxidant defense showed changes in the opposite direction: SOD activity was inhibited in the senescent exosome group, and LDHA knockdown brought about a considerable recovery (Fig. 3G, Exo-UV vs. Exo p < 0.001, Exo-UV+siLDHA vs. Exo-UV p < 0.001). On the key inhibitory axis, the protein levels of GPX4 and SLC7A11 were downregulated by senescent exosome treatment, and LDHA knockdown led to their partial restoration (Fig. 3H).

We further established LDHA-knockout fibroblasts and extracted exosomes to evaluate its effects on cell phenotypes. As shown by western blot assay, exosomes from UV-treated and LDHA-knockout fibroblasts exhibited reduced LDHA level compared with the control UV-treated fibroblasts (Figure S4A). After treatment with LDHA-knockout exosomes, recipient cells exhibited lower LDHA protein level (Figure S4B) and higher cell viability (Figure S4C). Similar with the siLDHA treatment, treatment with LDHA-knockout exosomes downregulated DNA damage marker γ-H2AX and senescence suppressor proteins (Figure S4D), repressed cell senescence (Figure S4E). Simultaneously, compared with the control exosomes from UV-treated fibroblasts, LDHA-knockout exosomes alleviated the lipid ROS accumulation (Figure S4F), reduced Fe2 + level and MDA production (Figure S4G and H), upregulated SOD activity (Figure S4I). Western blot results further indicated upregulated level of ferroptosis suppressors GPX4 and SLC7A11 (Figure S4J). Furthermore, treatment with exosomes formation inhibitor (GW4869) reduced the level of LDHA in HaCaT cells under co-culture with the senescent fibroblasts (Figure S5A), downregulated the level of cell senescence (Figure S5B-C), as well as repressed ferroptosis phenotype (Figure S5D-H).

To confirm that the observed ferroptosis was driven by excessive ROS, we employed the antioxidant NAC as a scavenger. As expected, NAC significantly reduced the lipid ROS-positive cell population and MDA content in Exo-UV-treated cells (Figure S6A-C). Moreover, NAC decreased the Fe²⁺ overload (Figure S6D) while restoring the protein levels of GPX4 and SLC7A11 (Figure S6E). These results indicate that scavenging ROS can effectively mitigate the ferroptotic damage mediated by exosomal LDHA.

These readouts point to a consistent mechanistic picture. Exosomes secreted by senescent fibroblasts carry high abundance LDHA. Recipient cells, upon uptake, exhibit enhanced lactate production and elevated oxidative stress, thereby triggering lipid peroxidation, iron ion accumulation, and failure of the antioxidant system, ultimately manifesting as suppression of the GPX4/SLC7A11 axis and increased susceptibility to ferroptosis. Downregulating the level of LDHA in donor exosomes concurrently produced mitigating effects across multiple dimensions including DNA damage, proliferative activity, lipid peroxidation, iron load, and antioxidant capacity. This consistency supports the position of LDHA as an upstream driver of the pathogenic activity of senescent exosomes.

Senescent fibroblast exosomal LDHA promotes skin photoaging and ferroptosis activation in vivo

In a UVB-induced skin photoaging mouse model, we further validated the physiological function of LDHA in senescent fibroblast exosomes and its role in skin tissue ferroptosis. Mice were randomly divided into Control, Exo, and Exo-siLDHA groups, receiving injections of PBS, senescent fibroblast exosomes, or LDHA-knockdown exosomes, respectively. Macroscopically, the Exo group showed significantly deepened wrinkles on the dorsal skin, with rough surface and reduced elasticity, whereas the Exo-siLDHA group exhibited a significant alleviation in wrinkle severity (Fig. 4A, Exo vs. Control p < 0.01, Exo-siLDHA vs. Exo p < 0.01). This indicates that exosomal LDHA contributes to the deterioration of skin phenotype during photoaging.

Fig. 4.

Senescent fibroblast exosomal LDHA promotes skin photoaging phenotypes and induces ferroptosis response in vivo. A Mouse skin wrinkle scoring results. The Exo group showed deepened dorsal skin wrinkles and rough surface, while the Exo-siLDHA group exhibited significantly reduced wrinkle severity, suggesting exosomal LDHA accelerates photoaging progression. B Western blot analysis of LDHA and DNA damage marker γ-H2AX expression in skin tissue. Exo treatment significantly enhanced the levels of both LDHA and γ-H2AX, while their expression was markedly reduced in the Exo-siLDHA group. C Lactate quantification results showed that lactate content in skin tissue was significantly higher in the Exo group than in the control, and LDHA-knockdown exosomes significantly reduced lactate accumulation. D Western blot detection of histone H3K18 lactylation modification. Exo treatment significantly enhanced the H3K18la signal, while the modification level decreased in the Exo-siLDHA group, with stable total H3 expression. E H&E staining showed significant epidermal thickening and loose dermal collagen arrangement in the Exo group, while the skin structure became more compact and inflammatory cell infiltration was reduced in the Exo-siLDHA group. F Masson’s trichrome staining results indicated fragmented and sparse blue-stained dermal collagen fibers in the Exo group, while collagen deposition was significantly improved in the Exo-siLDHA group. G Flow cytometric analysis of lipid peroxidation levels. Lipid ROS was significantly elevated in the Exo group and decreased in the Exo-siLDHA group, suggesting LDHA promotes lipid peroxidation accumulation. H–I Detection of ferroptosis-related metabolic indicators. MDA and Fe²⁺ content were significantly increased in the Exo group (H), and both levels were significantly decreased in the Exo-siLDHA group (I). J Western blot analysis of the ferroptosis inhibitory axis proteins GPX4 and SLC7A11. Both were significantly downregulated in the Exo group and partially restored by LDHA-knockdown exosomes. Data represent results from at least three independent experiments. Error bars indicate SD. **P < 0.01 compared to the control group (Exo); ##P < 0.01 compared to the Exo-UV group

Molecular level detection further supported this observation. Western blot results showed that Exo treatment significantly upregulated LDHA and the DNA damage marker γ-H2AX in skin tissue, while their expression significantly decreased in the Exo-siLDHA group (Fig. 4B), suggesting that exosomal LDHA can also induce DNA damage accumulation and promote tissue aging in vivo. Consistent with this, lactate quantification analysis revealed a significant increase in lactate concentration in the skin tissue of the Exo group, while LDHA-knockdown exosomes significantly reduced lactate levels (Fig. 4C, Exo vs. Control p < 0.01, Exo-siLDHA vs. Exo p < 0.01), further indicating that exosomal LDHA drives enhanced local lactate metabolism in vivo. To investigate whether this metabolic signal affects the epigenetic state, we detected the level of histone H3K18 lactylation. The results showed that Exo treatment significantly enhanced H3K18la modification in skin tissue, while LDHA-knockdown exosomes markedly reduced this modification level (Fig. 4D), indicating a positive association between upregulated lactate metabolism and enhanced epigenetic lactylation also exists in vivo.

Histological analysis further revealed the impact of exosomal LDHA on skin structure. H&E staining showed significant epidermal thickening in the Exo group, with loose dermal fiber arrangement and enlarged collagen gaps, whereas the skin structure in the Exo-siLDHA group tended to be more intact, with denser collagen distribution (Fig. 4E). Masson’s trichrome staining results similarly indicated fragmented and sparse blue-stained dermal collagen fibers in the Exo group, while collagen deposition recovered after LDHA knockdown (Fig. 4F). These results collectively demonstrate that exosomal LDHA can promote typical photoaging pathological changes in vivo, including skin structure degeneration and collagen degradation.

Regarding the mode of cell death, we assessed the role of exosomal LDHA by detecting key ferroptosis indicators. Flow cytometry detection showed that Exo treatment significantly increased the lipid peroxidation level in skin tissue cells, while the lipid ROS proportion decreased in the Exo-siLDHA group (Fig. 4G, Exo vs. Control p < 0.01, Exo-siLDHA vs. Exo p < 0.01). Biochemical assay results further indicated that MDA content and Fe²⁺ concentration were significantly elevated in the skin of the Exo group, and both levels markedly decreased after LDHA knockdown (Fig. 4H-I, both p < 0.01), suggesting that exosomal LDHA promoted lipid peroxidation and iron accumulation. Concurrently, Western blot analysis showed that the ferroptosis inhibitory axis proteins GPX4 and SLC7A11 were significantly downregulated in the Exo group skin, and were partially restored in the Exo-siLDHA group (Fig. 4J). These data consistently point, from the three dimensions of oxidative metabolism, metal ion balance, and the antioxidant network, to the conclusion that exosomal LDHA amplifies the ferroptosis response in vivo through lactate accumulation and epigenetic lactylation modification, thereby accelerating the skin photoaging process.

Senescent fibroblast exosomes establish a pro-aging pathway of “upregulated lactate metabolism – enhanced histone lactylation – activated ferroptosis” in vivo through the enrichment of LDHA. Knocking down LDHA not only improved skin structure and function but also attenuated lactate accumulation, enhanced epigenetic modification, and amplified ferroptosis signaling in photoaged tissue, suggesting that LDHA is a key molecular link mediating skin photoaging and ferroptosis via senescent exosomes.

Mechanism of senescent fibroblast exosomal LDHA in epigenetically regulating key ferroptosis molecules via histone lactylation

Senescent fibroblast exosomal LDHA remodels the transcriptional landscape of recipient cells and points to key ferroptosis target genes

When comparing the transcriptomic differences between the exosome group and the exosome + siLDHA group in foreskin fibroblasts, the overall expression profile first presented a typical volcano plot distribution. Figure 5A shows a large number of genes were significantly upregulated or downregulated. The magnitude of change and the number of genes both suggest that the exosomal cargo can drive global transcriptional reprogramming in recipient cells, and this reprogramming was markedly attenuated after inhibiting LDHA in the exosomes, indicating that LDHA is a major component of this transcriptional drive. Enrichment analysis of the differentially expressed genes for biological processes revealed that nutrient and chemical stress response, oxidative stress response, organelle disassembly, autophagy regulation, and hypoxia response were among the top terms (Fig. 5B). These terms collectively depict homeostatic imbalance and stress accumulation under metabolic pressure, consistent with the physiological context of sustained damage in photoaged tissue. Further pathway enrichment results clearly highlighted ferroptosis, accompanied by coordinated enrichment of networks such as lipid metabolism, mTOR and autophagy, and lifespan regulation, pointing to a metabolism-stress coupling framework with membrane lipid peroxidation as the core node, and the directionality is consistent with the aforementioned functional results (Fig. 5C). HALLMARK gene set analysis supplemented the inflammatory and injury background. The parallel activation of TNFα via NF-κB, hypoxia, p53, and inflammation-related pathways in Fig. 5D implies that the cells are under chronic stress and inflammatory amplification, signals known to lower the antioxidant threshold and increase susceptibility to ferroptosis. To evaluate the contribution of different cell death pathways, we measured HaCaT cell viability after senescent exosome exposure in the presence of Fer-1 (ferroptosis inhibitor), Z-VAD-FMK (apoptosis inhibitor), Necrostatin-1 (necroptosis inhibitor), or Chloroquine (autophagy inhibitor) and measured cell survival using the CCK-8 assay. Only Fer-1 significantly rescued cell viability, whereas other inhibitors had rather slight effects (Figure S7). These results further support that ferroptosis, rather than apoptosis, necroptosis, or autophagy, is the primary pathway mediating exosome-induced cell death in our model.

Fig. 5.

Senescent fibroblast exosomal LDHA remodels the transcriptional landscape of recipient cells and points to key ferroptosis target genes. A Volcano plot displaying the distribution of differentially expressed genes (DEGs) between the Exo and Exo + siLDHA groups. Most upregulated and downregulated genes showed significant stratification, indicating that exosomal cargo drives transcriptional reprogramming in recipient cells, which is significantly attenuated by LDHA knockdown. B GO functional enrichment analysis showed that DEGs were concentrated in biological processes such as nutrient and chemical stress response, oxidative stress response, autophagy regulation, and hypoxia response, indicating the involvement of exosomal LDHA in regulating cellular stress homeostasis. C KEGG pathway enrichment results further revealed that the ferroptosis pathway was significantly upregulated in the Exo group, accompanied by coordinated activation of metabolic pathways such as mTOR, autophagy, and lipid metabolism, suggesting a close relationship between exosomal LDHA and ferroptosis regulation. D HALLMARK gene set analysis showed parallel activation of signals including TNFα/NF-κB, hypoxia, p53, and inflammatory response, implying that exosomal LDHA places recipient cells in a state of chronic inflammation and oxidative damage, thereby lowering the threshold of the antioxidant defense. E RNA sequencing results showed that the ferroptosis inhibitor GPX4 was expressed significantly higher in the Exo + siLDHA group than in the Exo group, suggesting LDHA knockdown can partially restore antioxidant defense. F The ferroptosis promoter ACSL4 was significantly upregulated in the Exo group and downregulated in the Exo + siLDHA group, indicating that LDHA pushes up ACSL4 expression through coupling of metabolic and epigenetic layers. G The cystine transporter SLC7A11 expression increased after LDHA knockdown, consistent in direction with the change in GPX4, further supporting LDHA’s position at an upstream regulatory point of ferroptosis susceptibility. Data represent results from at least three independent experiments. Error bars indicate SD. *P < 0.05 compared to the control group (Exo)

At the candidate target level, we focused on the core molecules of ferroptosis execution and inhibition, and presented expression validation with box plots. Figure 5E shows that GPX4, which inhibits ferroptosis, was significantly higher in the exosome + siLDHA group than in the exosome group, consistent with the biological trend of suppressed lipid peroxidation after LDHA impairment, suggesting a partial restoration of the antioxidant defense. In contrast, ACSL4, which promotes the esterification of polyunsaturated fatty acids and amplifies membrane lipid peroxidation sensitivity, was upregulated in the exosome group and downregulated in the exosome + siLDHA group, indicating that LDHA drives the expression of this amplifier through both metabolic substrate supply and downstream regulation (Fig. 5F). SLC7A11, which also belongs to the inhibitory axis, increased when LDHA was knocked down, aligning in direction with GPX4, further supporting LDHA’s position at an upstream regulatory point of ferroptosis susceptibility (Fig. 5G).

The global shift in the transcriptome, the coordinated enrichment of biological processes and pathways, and the inverse changes in GPX4 decrease and ACSL4 increase together outline a sensitization pathway initiated by exosomal LDHA. This pathway connects enhanced lactate metabolism with oxidative stress, inflammatory signals, and membrane lipid peroxidation, and highlights ACSL4 as a key ferroptosis effector at the gene level, thus identifying it as the priority target for subsequent causal validation and mechanistic experiments.

Senescent fibroblast exosomal LDHA epigenetically regulates key ferroptosis molecules via histone lactylation

To verify whether exosomal LDHA drives transcriptional remodeling at the epigenetic level through lactate supply, we first measured intracellular lactate levels. Lactate content was high in both cell types treated with Exo, siLDHA caused a significant decrease, and supplementing sodium lactate in this context effectively rescued the lactate level, bringing it close to the range of the Exo group (Fig. 6A, Exo vs. Exo+siLDHA p < 0.01, Exo+siLDHA+Lactate vs. Exo+siLDHA p < 0.01). At the molecular level, the band intensity of H3K18 lactylation changed in the same direction as lactate: strongest in the Exo group, significantly weakened by siLDHA, and partially restored by sodium lactate (Fig. 6B). This indicates that exosomal LDHA directly affects the availability of histone lactylation through metabolic substrate supply.

Fig. 6.

Mechanism of senescent fibroblast exosomal LDHA epigenetically regulating the key ferroptosis molecule ACSL4 via histone lactylation. A Lactate content measurement showed that intracellular lactate levels were significantly elevated in the Exo group, significantly decreased after LDHA knockdown, and partially restored by sodium lactate (NaLa) supplementation on the basis of Exo + siLDHA, indicating that LDHA influences intracellular lactate homeostasis through metabolic substrate supply. B Western blot analysis of histone H3K18 lactylation modification. The H3K18la signal was strongest in the Exo group, significantly weakened by siLDHA, and partially restored after sodium lactate supplementation, suggesting a positive correlation between lactate accumulation and histone lactylation. C ChIP-qPCR results showed that the enrichment of H3K18la at the ACSL4 promoter region was highest in the Exo group, significantly reduced by siLDHA, and markedly recovered after sodium lactate supplementation, demonstrating that lactylation modification has site-specific regulatory function on this promoter. D Dual-luciferase reporter assay showed that ACSL4 promoter activity was highest in the Exo group, significantly inhibited by LDHA knockdown, and clearly recovered after exogenous lactate rescue, supporting the functional role of lactylation in ACSL4 transcriptional activation. E qPCR detection results showed that ACSL4 mRNA was highly expressed in the Exo group, significantly downregulated by siLDHA, and recovered after sodium lactate supplementation, a trend consistent with changes in histone lactylation. F Western blot validation of ACSL4 protein expression changes. Expression was highest in the Exo group, significantly reduced by siLDHA, and partially restored after NaLa supplementation, further confirming that LDHA regulates the transcription-translation process of ACSL4 via lactylation. **P < 0.01 compared to the control group (Exo); ##P < 0.01 compared to the Exo-UV group

To confirm whether lactylation achieves functional enrichment on the target gene promoter, we performed ChIP-qPCR on the ACSL4 promoter region. The enrichment of H3K18la at this site peaked in the Exo group, was significantly reduced by siLDHA, and showed a clear recovery after sodium lactate supplementation, with consistent trends in both fibroblasts and HaCaT cells (Fig. 6C, Exo vs. Exo+siLDHA p < 0.01, Exo+siLDHA+Lactate vs. Exo+siLDHA p < 0.01). This suggests that lactylation is not global noise but participates in ACSL4 transcriptional regulation in a site-specific manner. If this site enrichment has transcriptional function, it should be corroborated at the reporter gene level. The dual-luciferase assay showed that ACSL4 promoter activity was higher in the Exo group, significantly decreased by siLDHA, and exogenous lactate restored most of the activity deficit (Fig. 6D, Exo vs. Exo+siLDHA p < 0.01, Exo+siLDHA+Lactate vs. Exo+siLDHA p < 0.01). Correspondingly, ACSL4 transcription and protein expression showed concordant changes: both qPCR and Western blot showed the highest level in Exo, the lowest in siLDHA, and a recovery after lactate rescue (Fig. 6E-F, consistent in both fibroblast and HaCaT cell lines, key comparisons p < 0.01).

These results collectively outline a coherent mechanistic chain. LDHA carried by senescent exosomes increases lactate supply in recipient cells, promotes enhanced lactylation at the H3K18 site, and leads to functional enrichment on the ACSL4 promoter, resulting in increased promoter activity and upregulated transcriptional expression. Disrupting this pathway by knocking down LDHA in exosomes caused simultaneous decreases in lactate and H3K18la, followed by suppression of ACSL4 transcriptional activity and expression; exogenous lactate could largely rescue the entire effect chain from metabolism to epigenetics to transcription, proving that the LDHA—Lactate—H3K18la—ACSL4 axis has a causal sequential relationship in both skin-related cell types.

Senescent fibroblast exosomal LDHA amplifies ferroptosis and exacerbates photoaging-related damage via the ACSL4 axis

In both foreskin fibroblast and HaCaT models, we set up four groups: Control, siACSL4, siACSL4 combined with senescent fibroblast exosomes, and siACSL4 combined with exosomes plus LDHA inhibitor, to dissect the causal relationship between exosomal LDHA and ACSL4 in the ferroptosis pathway. Starting from damage and pathway initiation molecules, silencing ACSL4 reduced the γ-H2AX band intensity while maintaining a low endogenous damage background. Adding senescent exosomes simultaneously enhanced both γ-H2AX and ACSL4 bands, suggesting that exosomes can restart the DNA damage and lipid remodeling program even in the silenced background. Further addition of the LDHA inhibitor markedly weakened this enhancement, indicating that this restorative effect depends on the LDHA activity in the exosomes, with consistent trends in both cell lines (Fig. 7A, key comparisons showed significant differences, p < 0.01). At the functional level, corresponding to the molecular readouts, the CCK-8 assay showed increased cell viability in the siACSL4 group, decreased viability after exosome treatment, and significant recovery after LDHA inhibition (Fig. 7B, decrease for Exo relative to siACSL4 p < 0.001, recovery after LDHA inhibition p < 0.01). These results indicate that the absence of ACSL4 can temporarily increase cell tolerance, whereas exosome-carried LDHA can break this tolerance, and inhibiting LDHA rescues the cells from the suppressed state.

Fig. 7.

Senescent fibroblast exosomal LDHA amplifies ferroptosis and exacerbates skin photoaging-related damage via the ACSL4 axis. A Western blot analysis of LDHA, γ-H2AX, and ACSL4 expression. ACSL4 silencing reduced γ-H2AX levels, while senescent exosome (Exo) treatment re-upregulated both proteins, which was partially reversed by the LDHA inhibitor Oxamate (Oxa), suggesting that exosomes drive ACSL4 upregulation and DNA damage accumulation through LDHA activity. B CCK-8 assay detecting cell viability. The siACSL4 group showed enhanced viability, the Exo group showed a significant decrease, and viability recovered after co-treatment with the LDHA inhibitor, indicating that exosome-induced cell damage depends on the LDHA-ACSL4 pathway. C Flow cytometric analysis of lipid peroxidation levels. ACSL4 silencing reduced the proportion of lipid ROS-positive cells, Exo treatment significantly increased it, and LDHA inhibition attenuated this increasing trend, indicating that exosome-promoted lipid peroxidation depends on both LDHA and ACSL4 activity. D Fe²⁺ content measurement showed that the siACSL4 group had the lowest iron ion levels, the Exo group showed a significant increase, and the LDHA inhibitor brought Fe²⁺ concentration down, suggesting that iron homeostasis imbalance is related to the synergistic action of the LDHA-ACSL4 axis. E MDA content detection further verified the lipid oxidation state. ACSL4 silencing decreased MDA, the Exo group showed a significant increase, and Oxa inhibition of LDHA led to partial recovery, a trend consistent with lipid ROS. F Western blot analysis of the ferroptosis inhibitory axis proteins GPX4 and SLC7A11. Exo treatment led to significant downregulation of both, and LDHA inhibition restored their expression, indicating that the LDHA-ACSL4 axis promotes ferroptosis activation by suppressing the antioxidant defense. Data represent results from at least three independent experiments. Error bars indicate SD. *, *** indicate significant difference compared to the Control group; #, ### indicate significant difference compared to the siACSL4 + Exo group; ^, ^^ indicate significant difference compared to the siACSL4 + Exo + Oxa group (*p < 0.05, **p < 0.01, ***p < 0.001; #p < 0.05, ##p < 0.01, ###p < 0.001; ^p < 0.05, ^^p < 0.01)

To confirm whether these biochemical changes culminated in cell death, we evaluated the core phenotype of ferroptosis, C11-BODIPY flow cytometry readings showed that the lipid peroxidation ratio was low in the siACSL4 group, significantly increased after exosome addition, and was pulled down again to an intermediate level by LDHA inhibition, indicating that exosomes amplify membrane lipid peroxidation on an ACSL4-dependent basis, and this amplification requires the metabolic support of LDHA (Fig. 7C, increase for Exo relative to siACSL4 p < 0.01, decrease after LDHA inhibition p < 0.01). Iron homeostasis indicators showed a consistent direction: free Fe²⁺ decreased in the siACSL4 group, significantly increased after exosome treatment, and decreased again with LDHA inhibition, with consistent trends in both cell lines (Fig. 7D, p < 0.001 and p < 0.01). The change in the lipid peroxidation end product MDA was completely concordant with lipid ROS and Fe²⁺: decreased by ACSL4 silencing, increased by exosomes, and decreased after LDHA inhibition, suggesting that the membrane lipid oxidation chain is synchronously driven and relieved at both molecular and metabolic levels (Fig. 7E, p < 0.001 and p < 0.01).

Molecular evidence for the inhibitory axis further closed the loop. Western blot showed that GPX4 and SLC7A11 remained at relatively high levels under siACSL4 conditions, were downregulated after exosome addition, and were partially restored by LDHA inhibition, demonstrating that the collapse of the antioxidant defense is closely related to metabolic input and is regulated by the LDHA-ACSL4 linkage (Fig. 7F, key comparisons p < 0.01). Consistent with the transcriptomic profiling, this result confirmed that Exo-UV treatment not only upregulated ACSL4 but also led to a decrease in the defense proteins GPX4 and SLC7A11. Importantly, the expression of these markers was partially restored upon LDHA inhibition. However, considering that our ChIP-qPCR results identified a direct enrichment of H3K18la at the ACSL4 promoter, we focused on ACSL4 as the most direct epigenetic target of exosomal LDHA, while the alterations in GPX4 and SLC7A11 likely represent a broader, synergistic response to the disturbed redox homeostasis.

Integrating the above readouts, a clear mechanistic chain can be delineated. Senescent fibroblast exosomes, via LDHA, enhance lactate supply and metabolic flux, upregulate ACSL4-mediated polyunsaturated fatty acid esterification, thereby making membrane lipids more sensitive to peroxidation, concurrently with Fe²⁺ accumulation and GPX4-SLC7A11 suppression, ultimately pushing recipient cells into a ferroptosis-susceptible state. Silencing ACSL4 can weaken this pathway, while inhibiting LDHA simultaneously removes the dual push on lipid peroxidation and iron load from the metabolic source, thus producing consistent counteracting effects across morphological, functional, and molecular levels.

Discussion

This study aimed to elucidate the mechanism by which LDHA in exosomes released by senescent fibroblasts promotes iron-dependent cell death (ferroptosis) and accelerates skin photoaging by inducing histone lactylation to regulate the expression of ACSL4. Our experimental results demonstrate that in a UV-induced skin photoaging environment, fibroblasts undergo glycolytic metabolic reprogramming upon entering senescence, leading to a significant increase in LDHA expression and lactate production [19]. Senescent cells transmit LDHA-enriched cargo to neighboring cells via exosomes, inducing histone lysine lactylation modifications in recipient cells, which subsequently upregulates ACSL4 gene transcription. This transcellular metabolic-epigenetic signaling pathway may be a significant driver of the skin photoaging process [20].

Our findings align with the recent understanding of the role of lactate and histone lactylation in epigenetic regulation. Lactate can serve as a substrate to directly modify histone lysine residues, forming the novel epigenetic mark of histone lactylation, thereby activating gene transcription on chromatin [21]. Unlike classical acetylation, histone lactylation increases during cellular metabolic disturbances and specifically promotes the expression of a set of target genes. For instance, during the late stage of bacterial infection-induced M1 macrophage polarization, endogenous lactate accumulation triggers histone lactylation to upregulate the wound healing-related gene Arg1, facilitating homeostasis recovery post-inflammation [22]. Our results indicate that a similar “lactate clock” mechanism exists in the skin aging environment. As fibroblasts senesce and glycolysis enhances, excessively produced lactate enters neighboring cell nuclei via exosomes, catalyzing histone lactylation to initiate the transcriptional program of specific genes. Notably, we demonstrated that ACSL4 is a key downstream target of the lactate metabolism-epigenetic axis. Evidence for this exists in other disease models. A recent study by Wu et al. [23] found that a high-lactate microenvironment in the joints of rheumatoid arthritis patients led to increased global histone lactylation levels in synovial fibroblast-like cells and upregulated NFATc2 gene expression via H3K9 lactylation, promoting inflammatory cell migration and disease progression. Similarly, in our photoaging model, LDHA-mediated histone lactylation selectively enhanced ACSL4 expression, suggesting that lactate-mediated epigenetic regulation may play a role in promoting specific gene expression across different cell types and pathological states.

This study provides an in-depth analysis of the functions of key ferroptosis factors, such as ACSL4 and GPX4, in skin aging, and compares and discusses our results in the context of recent literature. ACSL4 is a fatty acid long-chain acyl-CoA synthetase recently established as an essential component in the ferroptosis process [24]. Research by Doll et al. [25] demonstrated through CRISPR screening that ACSL4 deficiency confers significant resistance to ferroptosis in cells, suggesting that ACSL4 enriches polyunsaturated fatty acids in the cell membrane, predisposing them to peroxidation, which is a crucial prerequisite for ferroptosis execution. Our data similarly show that ACSL4 is upregulated in photoaged skin tissue, likely due to epigenetically activation mediated by senescent cell-derived exosomes. The elevation of ACSL4 enriches skin cell membranes with peroxidation-susceptible lipids, thereby exacerbating lipid peroxidation and cellular damage under UV-induced high oxidative stress conditions [26]. Recent studies on photoaging also support the involvement of ferroptosis; during UVB-induced skin photoaging, intracellular iron ion and lipid peroxidation levels increase, and ferroptosis plays a key role [27]. Teng et al. [8] found that knocking down the mitochondrial respiratory chain subunit NDUFS4 significantly alleviated ferroptosis hallmark molecular changes, upregulated intracellular GPX4 and iron storage protein levels, decreased the collagen-degrading enzyme MMP1, and the ferroptosis inhibitor Ferrostatin-1 significantly mitigated UVB-induced skin aging damage. These results resonate with our findings: enhanced ferroptosis accelerates collagen degradation and cellular functional decline, promoting the pathological progression of photoaging, and ACSL4 is a key promoter triggering ferroptosis in this context.

In contrast to ACSL4, GPX4 is a major antagonist of ferroptosis. GPX4 utilizes glutathione to reduce membrane lipid peroxides, preventing irreversible oxidative damage to the cell membrane [28]. We observed relatively low GPX4 function in photoaged skin, which may be related to excessive lactylation regulation and an iron-overloaded environment. When histone lactylation upregulates ACSL4 and promotes ferroptosis, the antioxidant defense line of GPX4 is easily overwhelmed, leading to cumulative cell death. The importance of GPX4 has been demonstrated in other skin aging-related pathologies. Zhang et al. [29] reported that high GPX4 expression in skin fibroblasts of scleroderma (systemic sclerosis) patients conferred resistance to ferroptosis, and inhibiting GPX4 increased fibroblast susceptibility to ferroptosis, potentially alleviating pathological tissue fibrosis. Correspondingly, during healthy aging, if GPX4 function declines, cells become more susceptible to ferroptosis. The NDUFS4 study also indicated that maintaining levels of anti-ferroptosis molecules like GPX4 under UVB irradiation helps reduce skin aging damage [8]. Therefore, the enhanced ferroptosis resulting from the “LDHA-histone lactylation-ACSL4” axis proposed in this study is likely achieved by suppressing GPX4-mediated antioxidant defense and enhancing lipid peroxidation. This inference is consistent with the mode of action of ferroptosis in various skin diseases and aging; when GPX4 is inactivated or depleted, uncleared lipid peroxides and iron ions within the cell jointly trigger the ferroptosis pathway [30, 31]. The combination of ACSL4 upregulation and relative GPX4 deficiency observed in our UV photoaging model is precisely the trigger for ferroptosis. The resulting cellular damage and matrix degradation can partially explain the phenotypes of photoaged skin, such as wrinkling and loss of elasticity.

It is worth emphasizing that this study reveals a new dimension of intercellular communication in the mechanism of skin aging: the regulatory role of exosomes in transmitting metabolic and epigenetic information between different cells. The traditional view holds that senescent cells influence the surrounding microenvironment through secretions like pro-inflammatory cytokines and proteases (collectively known as the Senescence-Associated Secretory Phenotype, SASP), accelerating tissue aging. However, our findings expand the focus to exosome-mediated signal transduction. Exosomes contain active cargo such as proteins, lipids, and nucleic acids, which can be efficiently endocytosed by recipient cells, thereby altering their functional state [32]. Recent review articles indicate that exosomes participate in regulating various aging-related processes, including oxidative stress, inflammatory responses, and cellular senescence itself [33]. Our discovery further complements the role spectrum of exosomes in skin photoaging. Exosomes released by senescent fibroblasts carry not only regulatory factors like miRNAs but also functional proteins such as the metabolic enzyme LDHA. This makes exosomes carriers of “metabolic information” between cells, allowing the high-lactate metabolic state of donor cells to be transmitted to neighboring cells via exosomes, thereby inducing epigenetic changes (histone lactylation) and functional alterations (ferroptosis susceptibility) in the latter. One might question the biological “benefit” of such harmful communication that induces ferroptosis in healthy neighbors. We propose that this process represents a “Senescence-Induced Bystander Effect”—a pathological amplification mechanism rather than a beneficial adaptation. In the context of chronic UVB exposure, senescent fibroblasts fail to be cleared and instead create a pro-aging niche. The exosomal transfer of LDHA acts as a metabolic stress signal that “contaminates” healthy keratinocytes, spreading the damage across the skin tissue. While inducing death in compromised cells might originally serve tissue homeostasis, its persistent activation under chronic stress leads to maladaptive loss of healthy cells, ultimately contributing to skin thinning and barrier dysfunction.

This mechanism echoes some reports in the tumor microenvironment. For example, one study found that neutrophil-derived exosomes delivered LDHA into breast cancer cells, increasing histone H3K18 lactylation levels in recipient cells, thereby influencing the chromatin state of the GPX4 gene and regulating tumor cell ferroptosis and proliferative capacity [18]. These parallels suggest that exosome-mediated LDHA transport may have universal biological significance; under different pathological environments, metabolic enzymes carried by exosomes can potentially regulate gene expression in recipient cells, thereby influencing cell fate decisions [34, 35]. In skin tissue, fibroblasts engage in close paracrine communication with keratinocytes, melanocytes, etc. Our study suggests that exosomes may be an important part of this network. Senescent fibroblasts transmitting pro-aging signals to surrounding cells via exosomes might not only lead to epigenetic changes and functional impairment in neighboring normal fibroblasts but could also affect the proliferation, differentiation of epidermal cells, or melanin metabolism, exacerbating the photoaging phenotype at multiple levels. This provides new avenues for anti-photoaging interventions; for instance, blocking exosome production or release, or clearing senescence-promoting cargo within exosomes, could become potential strategies for future skin aging interventions.

Our study has several limitations that warrant further improvement in the future. Firstly, while this research primarily revealed the molecular mechanism of LDHA promoting ferroptosis via histone lactylation in cell and mouse models, validation in human skin tissue is insufficient. The correlation between LDHA/lactate levels and ACSL4 expression and ferroptosis markers in actual human photoaged skin needs confirmation through clinical sample testing. Secondly, we focused on the pathway of LDHA-mediated histone lactylation regulating ACSL4, but the cellular effects induced by lactate could be multifaceted. For example, lactate can also act on cell membrane receptors such as G-protein coupled receptors (e.g., GPR81) to trigger signaling cascades, or influence cellular NAD⁺/NADH ratios thereby regulating other epigenetic factors (e.g., acetyltransferases and deacetylases). These potential alternative or synergistic pathways were not explored in this study; future research could undertake a more comprehensive investigation. Furthermore, identifying the specific enzymes responsible for histone lactylation and its precise sites remains an unresolved scientific question. While several histone acetyltransferases are known to potentially possess lactyltransferase activity [36], determining which specific enzyme mediates lactylation at the ACSL4 promoter region in skin cells requires further clarification using mass spectrometry analysis and gene knockout experiments. Moreover, ACSL4 is not the sole regulator of ferroptosis; the ferroptosis pathway involves numerous other components working in concert, including hepcidin, lipid peroxidases (e.g., 15-LOX), the cysteine supply pathway, and the Nrf2 antioxidant pathway [37]. While we focused on changes in ACSL4 and GPX4 to explain the mechanism of photoaging, contributions from other pathways cannot be ruled out. For instance, we observed that senescent cells themselves exhibit some tolerance to ferroptosis, and literature reports that senescent cells can avoid ferroptosis by sequestering iron in lysosomes. This suggests that senescent cells might, on one hand, avoid their own death, while on the other hand, promote ferroptosis in surrounding cells via exosomes and other means, thus creating a kind of “bystander effect,” exacerbating tissue degenerative changes. For this interesting hypothesis, we currently lack direct evidence; future studies could utilize co-culture models with different reporter genes to track the impact of senescent cells on ferroptosis in neighboring cells. Finally, the models used in this study primarily involved fibroblasts, whereas skin photoaging is a complex process involving multiple cell types. Keratinocytes and immune cells also play important roles in UV-induced inflammation and photoaging; whether they are similarly affected by the exosomal LDHA-lactylation pathway warrants in-depth investigation. Furthermore, although our study emphasizes ACSL4 as the dominant downstream target of the LDHA-lactylation axis, it is plausible that other genes are simultaneously regulated by this epigenetic mark. Histone lactylation has been shown to influence a spectrum of genes involved in inflammation and metabolism in different contexts. While our functional rescue experiments with ACSL4 knockdown demonstrate that it is the major driver of the observed ferroptosis phenotype, future studies employing genome-wide ChIP-seq and RNA-seq are necessary to fully map the “lactylation landscape” in skin photoaging and identify additional genes that may synergistically contribute to tissue degeneration. Future research should integrate single-cell sequencing and spatial multi-omics technologies to map the exosome-mediated intercellular communication network at the whole-skin level, providing a comprehensive understanding of the mechanisms underlying skin photoaging.

Conclusion

This study systematically reveals the key regulatory role of senescent fibroblast-derived exosomes in skin photoaging. The results indicate that UV-induced fibroblast senescence is accompanied by increased glycolytic flux and lactate accumulation. Exosomes act as carriers of metabolic signals, transmitting LDHA-enriched cargo to neighboring skin cells, thereby inducing histone H3K18 lactylation modification in recipient cells. This epigenetic change significantly upregulates the expression of the ferroptosis promoter ACSL4, enhances lipid peroxidation and iron ion accumulation, and is accompanied by the downregulation of the anti-ferroptosis molecules GPX4 and SLC7A11, ultimately leading to exacerbated cellular damage and skin structure degeneration. Intervention by inhibiting LDHA activity or silencing ACSL4 effectively suppressed the occurrence of these molecular events, alleviated UVB-induced ferroptosis and photoaging phenotypes, and validated this metabolism–epigenetics–cell death causal chain.

This study establishes, for the first time from the perspective of exosome-mediated transcellular metabolic communication, the molecular axis of LDHA–histone lactylation–ACSL4–ferroptosis, providing a new theoretical framework for understanding the metabolic basis and epigenetic regulation of skin photoaging. The findings not only uncover the unique mechanism by which exosomes transmit metabolic and death signals between skin cells but also offer new potential targets for developing intervention strategies against photoaging. Future work should validate the expression patterns of LDHA and histone lactylation in photoaged human skin using clinical samples and further explore the spatiotemporal characteristics and systemic effects of exosome regulation.

In summary, senescent fibroblast exosomes upregulate ACSL4 expression through LDHA-mediated histone lactylation modifications, induce ferroptosis, and accelerate the skin photoaging process. This study reveals the coupling mechanism between metabolic signaling and epigenetic regulation in skin aging, providing a new theoretical basis and experimental evidence for targeting exosomes and their metabolic components to delay photoaging (Fig. 8).

Fig. 8.

Mechanism of LDHA-mediated histone lactylation in senescent fibroblast-derived exosomes promoting ferroptosis and skin photoaging

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Author contributions

Tingting Yan, Lixia Huang and Xuebiao Peng wrote the main manuscript text and Jingjing Tao, Yunling Yan, Yiping Zhong prepared figures. All authors reviewed the manuscript.

Funding

There was no funding for this study.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.