Chlorogenic acid attenuates skin senescence and UVR-induced photoaging via the modulation of mitochondrial function Authors: Rui Li, Chujuan Hu, Ping Zhou, Ge Si, Jinjin Cui, Yingying He, Yanxing Han, Jie Zhang, Wenbin Li, Lulu Wang, Jiandong Jiang Published: 2026 ABSTRACT BackgroundMitochondrial function plays a critical role in skin aging. Chlorogenic acid (CGA), a botanical compound, has demonstrated regulatory effects on mitochondrial function and senescence inhibition. However, whether the anti-aging effects of CGA are attributable to its regulation of mitochondrial function remains unclear. There is a need to investigate the anti-aging effect and mechanism of CGA by modulating mitochondrial function, particularly its mode of action on mitochondrial function.MethodsNormal human dermal fibroblasts and human epidermal keratinocytes were used to detect the regulation of collagen I production, mitochondrial functions, and anti-aging properties of CGA and confirmed by mitochondrial transplantation. The photo-aging mouse model was established by ultraviolet (UV) radiation, followed by the treatment of CGA-gel (1 mmol/kg/d) for 14 days. The skin tissues were collected and tested.ResultsCGA promotes collagen I (Col1) production by activating the TGF-β/Smad signaling pathway, while concurrently inhibiting cellular senescence. CGA administration significantly reduced the expression of p21, senescence-associated secretory phenotype (SASP) production, and SA-β-Gal activity in skin cells. Additionally, CGA treatment notably enhanced mitochondrial function, improving disrupted mitochondrial cristae in senescent cells and boosting the oxidative phosphorylation (OXPHOS) process. ATP levels increased by approximately 40–80% following CGA treatment. Mitochondrial transplantation further confirmed CGA’s anti-aging effects are linked to mitochondrial function. CGA significantly mitigated UV-induced reduction in Col1, suppressed p21 expression and SASP production, and improved mitochondrial morphology and structure in vivo.ConclusionCGA promotes Col1 production and attenuates skin cellular senescence, with these effects being directly associated with mitochondrial regulation. Thus, CGA holds promise as a potent agent for preventing cellular senescence.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12967-026-08538-8. Introduction As the largest organ and primary physical barrier of the human body, the skin plays a pivotal role in defending against external damage. It is influenced by both intrinsic (chronological) and extrinsic aging factors, which also contribute to the aging of other tissues [1]. Intrinsic aging factors include time, genetics, and hormones, which lead to chronological and physiological changes. Additionally, antioxidant capacity and reactive oxygen species (ROS) production are linked to intrinsic aging. Extrinsic factors, such as sunlight, air pollution, and cigarette smoke, accelerate skin aging [2]. The outcomes of intrinsic and extrinsic aging differ: intrinsic aging is associated with a loss of elasticity, reduced skin thickness, and skin laxity, while extrinsic factors lead to wrinkles, irregular pigmentation, and age spots. Type I collagen (Col1), a major component of the extracellular matrix (ECM), decreases with skin cellular senescence, making Col1 degradation a key marker of aging. Col1 biosynthesis is regulated by the TGF-β/Smad signaling pathway, while its degradation is controlled by matrix metalloproteinases (MMPs) [3, 4]. Recent studies have shown that skin senescence is interconnected with aging in other organs and contributes to whole-body aging [1]. Aged skin exhibits cellular, molecular, and functional changes similar to other senescent tissues, such as inhibited cell proliferation, cell cycle arrest, and loss of replication ability [5]. The mechanisms underlying cellular senescence are complex and include DNA damage, oxidative stress, telomere attrition, and mitochondrial dysfunction. The accumulation of senescent cells results in the secretion of factors known as the senescence-associated secretory phenotype (SASP), which can induce senescence in neighboring cells [6]. Mitochondrial dysfunction is a defining characteristic of aging tissues and senescent cells, with changes in mitochondrial mass, mitochondrial DNA (mtDNA), and respiratory capacity playing a pivotal role in the aging process [7]. Oxidative phosphorylation (OXPHOS) is linked to cellular senescence in various models, influencing ATP and ROS production [8]. As a tissue with high turnover and renewal, skin aging is closely associated with mitochondrial damage [9]. Hamanaka et al. found that mitochondrial regulation affects keratinocyte differentiation in the epidermis [10]. Singh et al. demonstrated that mtDNA depletion and related mitochondrial dysfunction contribute to skin aging [11]. Mitochondrial dynamics, biogenesis, and mitophagy are all critical in the skin’s senescence process [12]. Decreased mitochondrial function and ATP content induced by UVA radiation led to insufficient Col1 formation and secretion in dermal fibroblasts [13]. Therefore, mitochondrial regulation may be directly linked to Col1 production and degradation, presenting a promising therapeutic target for anti-aging treatments in skin. Natural products have long been used to mitigate skin senescence. Resveratrol (RE), a well-known anti-aging compound, has been shown to prevent aging by activating various signaling pathways, including sirtuins, Klotho, and AMPK [14, 15]. Vitamin C (VC), a common plant-derived component, is widely used in anti-aging products due to its antioxidant properties, ROS reduction, and potential to maintain telomere length [16, 17]. The safety of these natural products has been validated in numerous studies, highlighting the promise of natural compounds in anti-aging therapies. Chlorogenic acid (CGA), a polyphenolic compound found in various plants, has demonstrated multiple bioactivities, including antioxidant, anti-inflammatory, anti-tumor, and neuroprotective effects. The anti-oxidation and anti-inflammatory effects by CGA through reducing ROS production and inflammatory cytokine release have been recognized widely [18]. Chen et al. showed that CGA could attenuate deoxynivalenol-induced oxidative stress and inflammation by inhibiting MAPK/NF-κB/NLRP3 pathway, thus reducing mtROS and inflammatory cytokine generation [19]. Recently, a Phase II clinical trial of CGA for recurrent high-grade glioma (NCT03758014) was completed, with Phase I results showing significant improvements in overall survival and a good safety profile [20]. Our previous research revealed that CGA induces cancer cell differentiation and boosts the efficiency of anti-tumor immunotherapy [21, 22]. In addition to its anti-cancer properties, CGA has been shown to exert potential anti-aging and mitochondrial regulation effects. Studies by Xue et al. and Her et al. indicated that CGA mitigates photoaging induced by UVB or UVA radiation through collagen regulation [23, 24]. Besides, He et al. found that CGA prevents UVA-induced cellular senescence through suppressing the activity of ENO1 protein to block the glycolytic pathway in skin cells [25]. CGA also appears to delay chronological aging in yeast by reducing ROS production [26]. Moreover, Han et al. demonstrated that CGA enhances mitochondrial function, stimulating thermogenesis in brown adipocytes [27]. CGA has also been reported to protect against neuronal cell death by restoring mitochondrial biogenesis and mitophagy, thereby maintaining mitochondrial homeostasis and function [28]. In this study, mitochondrial transfer was used to demonstrate that CGA directly regulates mitochondrial function, promoting Col1 production, enhancing cell renewal, reducing SASP production, and delaying skin cell senescence. These findings suggest that mitochondrial regulation plays a central role in the anti-aging effect of CGA, therefore CGA could be a safe and effective senomorphic agent for treating tissue and cellular senescence. Methods and materials Chemicals Chlorogenic acid (CGA) was acquired from the Jiuzhang Biochemical Engineering Science and Technology Development Co., Ltd. (Chengdu, Sichuan, China). Cell culture Normal human dermal fibroblasts (NHDF) and human epidermal keratinocyte cell line (HaCaT) were provided by BeNa Culture Collection (Henan, China) and maintained in DMEM (Gibco, USA) supplemented with 10% Fetal Bovine Serum (Gibco, USA), penicillin (100 U/mL) and streptomycin (100 µg/mL) (Invitrogen). All cells were cultured at 37 °C and 5% CO2. The NHDF and HaCaT cells were seeded respectively at density of 3 × 105 and 2 × 105 into 12-well plate and cultured overnight. Then, the cells were treated with different concentrations of CGA (0,25,50,100,200 µM) for 24 h. In determining the time-dependent effect of CGA, the NHDF and HaCaT cells were seeded respectively at density of 1.5 × 105 and 1 × 105 into 12-well plate and cultured overnight. After treatment with 200 µM CGA for 24,48,72 h, the cells were harvested for subsequent operations. Quantitative real-time PCR Total RNA of each group cells was extracted using RNAeasy™ Animal RNA Isolation Kit with Spin Column (Beyotime, R0027, China) and detected the concentration, purity of RNA. Then the RNA was reverse transcribed into cDNA by HiFiScript cDNA Synthesis Kit (CWBIO). The Quantitative Real-Time PCR (qRT-PCR) experiment was performed by UltraSYBR Mixture (Low ROX, CWBIO) and 7500 Fast Real-Time System (Applied Biosystems). The house-keeping gene GAPDH was used as the reference gene, and the mRNA expression of genes was normalized with it. And use the 2−△△Ct method to calculate the relative mRNA expression of each group of cells. The primer sequences used were shown in Supplementary Table S1. Western blot Cells were harvested after the treatment as described above, washed with PBS for 2 times and lysed using radioimmunoprecipitation assay (RIPA) buffer. Protein concentrations were determined through a bicinchoninic acid assay protein assay kit (ThermoFisher). Equal protein amounts from the supernatants were separated by polyacrylamide gel electrophoresis, proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, USA). Membranes were blocked by 5% non-fat milk powder in TBST (Beyotime, Shanghai, China) and then incubated overnight with the appropriate primary antibody. The primary antibodies for Collagen I (Abcam, #ab138492), TGF-β1 (Abcam, #ab215715), Smad3 (CST, 9523 S), p21 (proteintech, #1035S-1-AP), β-actin (CST, #4970). Membranes were next washed and incubated with the matching corresponding horseradish peroxidase (HRP) conjugated secondary antibody (Beyotime, Shanghai, China) at 1:2000 dilution in 5% non-fat milk powder in TBST. Immobilon Western Chemiluminescent HRP Substrate (Millipore, Billerica, USA) was added, and the bands were imaged with the Chemidoc XRS+ electrophoretic imaging system (Bio-Rad). Density scanning of each protein band was performed using Image J. As a control, the results were normalized to β-actin. Multi-immunofluorescence NHDF and HaCaT cells grew on cell slides and were fixed by 4% paraformaldehyde for 15 min after the treatment of different doses of CGA (0-200 µM) for 24 h. After washing with PBS three times, the cell membranes were permeabilized through 0.5% Triton X-100. Next, the samples were blocked with normal goat serum, and incubated with diluted antibody solution at 4 °C overnight. After a wash in cold PBS, the samples were incubated with diluted secondary antibodies at room temperature. After wash, the cells were stained with DAPI. Analyze the signal under a fluorescence microscope (Olympus). ATP and OCR measurement ATP Assay Kit (Beyotime, S0062, China) was used to detect the level of cellular ATP. After the treatments, cells were washed with PBS for 2 times and lysed the cells by using the lysis buffer supplemented in the kit, then collected the lysates and centrifuged them at 12,000× g for 5 min at 4 °C, the supernatant was taken as samples. Operated according to the manufacturer’s instructions, added samples, detected reagent and the standard ATP solution into a white opaque 96-well plate, used 96 Microplate Luminometer (GloMax) to measure the luminescence, and the ATP concentration was calculated based on a calibration plot made with standard ATP solution (ranging from 0.1 to 10 µM). Cellular oxygen consumption rate (OCR) test was performed using an XF96 Seahorse Analyzer (Seahorse Bioscience, North Billerica, MA, USA). Briefly, 5 × 103 HaCaT cells and 7 × 103 NHDF cells were seeded into each well of the XF96 cell culture micro-plate and cultured overnight, then treated with CGA (0-200 µM) for 24 h. The OCR value was detected with the XF Cell Mito Stress Test Kit (Seahorse Bioscience, North Billerica, MA, USA) according to the operation manual instructions. The constituent of assay medium was 1.5 µM oligomycin A, 1 µM carbonyl cyanide p-trifluoro-methoxyphenyl hydrazine (FCCP), and 0.5 µM rotenone plus 0.5 µM antimycin A (Ron/AA). Upon completion of the Seahorse XF96 Flux analysis, cells were lysed by RIPA buffer to calculate protein concentrations with a BCA assay (ThermoFisher). All OCR values that detected was normalized to the protein content (ug/µL) of the corresponding well. Electron microscopy NHDF cells were seeded in 6-cm dish at density of 1 × 106 and cultured overnight, treated with 0, 200 µM CGA for 24 h. Remove the medium and 1 mL 0.25% Trypsin-EDTA solution was added to remove remaining serum. Then, 1 mL another Trypsin-EDTA solution was added into the dish and put the dish to incubator for about 1 min. Tapped the bottom of the dish softly until all the cells detached in the solution, terminated the detach and collected the solution. Then the collections were centrifuged at 1000× g for 5 min and washed 1 time with PBS. 1 mL 2.5% glutaraldehyde at 4 °C was added into the tubes and made the cell mass suspend in the glutaraldehyde. The samples were fixed in 4 °C overnight. Determination of ROS generation Cells were harvested and incubated in 10 µM DCFH-DA (Beyotime, S0033M, China) solution diluted in serum-free medium for 20 min at 37°C, 5% CO2, turned the tubes upside down per 3–5 min. Then, the cells were washed three times with serum-free medium to remove DCFH-DA that did not enter the cells. The level of ROS was measured using a FACSVerse flow cytometer under the FITC channel and analyzed with FlowJo software. Mitochondria transplantation The sufficient NHDF and HaCaT cells were cultured for the isolation of mitochondria. After the treatment by different doses of CGA (0, 200 µM) for 24 h, cells were collected and washed 1 time with PBS, then, the mitochondria were isolated by Cell Mitochondria Isolation Kit (Beyotime, C3601, China). The detection of mitochondrial membrane potential (JC-1 staining) was employed to identify the activity of isolated mitochondria. The isolated mitochondria were resuspended in PBS and the protein concentrations were determined through a Detergent Compatible Bradford Protein Assay Kit (Beyotime, P0006C, China) and the isolated mitochondria (10 µg) was added to the cultured cells for 24 h. To confirm the success of transplantation, the isolated mitochondria from NHDF cells were labeled by MitoTracker (Invitrogen, M22425, USA), and then co-cultured with recipient cells. After 2-hour incubation at 37 °C, 5% CO2, cell membranes were stained with DiO (Yeasen, 40725ES10, China) and nuclei with DAPI (Beyotime, C1006, China). All staining procedures were performed according to the manufacturers’ instructions and visualized by confocal microscopy. Detection of mitochondrial membrane potential (JC-1 staining) Enhanced mitochondrial membrane potential assay kit with JC-1 (Beyotime, C2003S, China) was used to detect mitochondrial membrane potential change. By following the instructions, NHDF cells grew in DMEM were incubated with freshly prepared JC-1 staining solution at 37 °C and protection from light for 20 min, then washed 2 times with JC-1 staining buffer. Photographs were taken using a fluorescence inverted microscope. ELISA Collected the culture medium of NHDF cells and centrifuged at 1000 g for 20 min, 4 °C. The supernatant was taken as samples. Used the COL I ELISA Kit (Mlbio, ml057630, China) to detect the secretion of collagen I by NHDF. All steps were performed according to the manufacturer’s instructions and the absorbency of the samples was detected by microplate reader in 450 nm wavelength. The activities of superoxide dismutase were determined by the Total SOD assay kit with WST-8 (Beyotime, S0101M, China). After treatment with different doses or time of CGA, both HaCaT and NHDF cells were harvested and lysed with lysate provided in the kit; then, they were centrifuged at 12,000 g for 5 min, 4 °C. Take the supernatant for determination. All steps were performed according to the instructions of manufacturer. At last, the absorbency of the samples was detected by microplate reader in 570 nm wavelength. SA-β-gal staining NHDF cells grew in DMEM with different CGA concentrations (0-200 µM) were stained by using Senescence β-Galactosidase Staining Kit (Beyotime, C0602, China). Cells grew on plates were washed with PBS 2 times, fixed in fixative for 15 min at room temperature, and washed with PBS 3 times. Then, cells were incubated overnight at 37 °C without CO2 in a freshly prepared staining buffer following the instructions. Photographs were taken by using an ordinary optical microscope (Olympus) at 40×, 100× and 400× magnification, and calculated the average value of cells positive for the SA-β-gal activity for each sample. Animal model Four-week-old female C57BL/6 N mice were purchased from Vital River (Beijing, China) and housed in a temperature-controlled room at 20 ± 1 °C on a 12 h light/dark cycle with free access to food and water, all mice acclimated for 2 weeks under the housing conditions. Dorsal skin (4 cm × 2 cm) was depilated using a hair removal cream, and mice were then randomly divided into four groups (n = 6 per group): Control; UV Model; NS-Gel; CGA-Gel, 6 mice per group. The gel was made of Poloxamer 407 (P407) at a concentration of 23%. CGA gel was, added CGA into the solvent and mixed with P407, and the final concentration of CGA in the CGA-gel was 100 mM. The dose of CGA-gel treated group was 1 mmol/kg/d and the dose of UVA and UVB were respectively 1.9 J/cm2 and 180 mJ/cm2. An hour after the complication of irradiation applied the gel and CGA gel to the mice. The treatment was performed every day and lasted almost 14 days. After sacrificing the mice, the serum and the skin tissue samples were collected. H&E and masson staining Skin tissues were fixed in paraformaldehyde and embedded in paraffin, and the paraffin sections were immunostained overnight with primary antibodies followed by secondary antibodies. Longitudinal cross sections of the mice skin tissue from these blocks were obtained and deparaffinization was performed by warming the slides at 65 °C for 30 min. The slides were then immersed in xylenes and next immersed in 100%, 95% then 70% ethanol for 15 min each. The slides were washed under running tap water and then incubated in a 1% hydrogen peroxide/methanol solution for 10 min. The slides were rinsed with distilled water and rinsed in PBST (Phosphate-Buffered Saline containing 0.05% Tween-20). Statistical analysis Statistical analysis was performed using GraphPad Prism 6 (CA). All data represent biological replicates (n) and were expressed as the mean ± SEM unless otherwise stated. Student’s t-test and One-way ANOVA test were used, unless otherwise stated. We considered p < 0.05 to be statistically significant. Results CGA enhanced Col1 production with TGF-β1/Smad3 pathway upregulation Col1 expression, both at the mRNA and protein levels, was assessed as a marker of skin cellular senescence. Compared with untreated controls, CGA significantly enhanced Col1 expression in NHDF and HaCaT cell lines (200 µM, 24 h; p < 0.001, 0.001, respectively) (Fig. 1A). This effect was further confirmed at the COL1A1 and COL3A1 mRNA levels (p < 0.01, 0.05, respectively) (Fig. 1B). Col1 secretion in the extracellular space was also increased in NHDF cells treated with CGA (200 µM, 24 h; p < 0.05) (Fig. 1C). Furthermore, immunofluorescence analysis confirmed a marked upregulation of Col1 expression in the NHDF cell line following CGA treatment (Fig. 1D). The TGF-β/Smad signaling pathway, a well-known regulator of Col1 production, has been shown to influence Col1 expression in skin. Therefore, the expression of TGF-β1 and Smad3 was evaluated in NHDF and HaCaT cell lines. As shown in Fig. 1E, CGA treatment significantly stimulated the expression of TGF-β1 and Smad3 in a dose-dependent manner in both NHDF and HaCaT cells (vs. untreated, TGF-β1, p < 0.001, 0.05; Smad3, p < 0.001, 0.001, respectively). These results suggest that the enhancement of Col1 expression and production by CGA is linked to its potential role in inhibiting skin cellular senescence. Fig. 1CGA enhanced production of Col1 in skin cells. Normal human dermal fibroblasts (NHDF) and human epidermal keratinocyte cell line (HaCaT) were treated with CGA (0,25,50,100,200 µM) for 24 h at 37 °C in 5% CO2. A. The protein expression of Col1 was tested by using Western blot, and the results were normalized to that of β-actin. The density scanning was also performed. B. The mRNA expressions of COL1A1 and COL3A1 were evaluated by qRT-PCR, the results were normalized to GAPDH. C. The level of Col1 secreted by NHDF cells was tested by ELISA and normalized to 0 µM group. (D) The representative images of immunofluorescent staining for Col1 in NHDF cells. Scale bars, 50 μm. (E) The protein expression of TGF-β1 and Smad3 were detected by Western blot assay. β-actin was used as an endogenous reference. Data are presented as mean ± SEM. Three independent experiments were performed under identical condition. Significant differences are indicated: * p < 0.05, ** p < 0.01, *** p < 0.001. CGA treated vs. 0 µM group by One-way ANOVA test CGA enhanced production of Col1 in skin cells. Normal human dermal fibroblasts (NHDF) and human epidermal keratinocyte cell line (HaCaT) were treated with CGA (0,25,50,100,200 µM) for 24 h at 37 °C in 5% CO2. A. The protein expression of Col1 was tested by using Western blot, and the results were normalized to that of β-actin. The density scanning was also performed. B. The mRNA expressions of COL1A1 and COL3A1 were evaluated by qRT-PCR, the results were normalized to GAPDH. C. The level of Col1 secreted by NHDF cells was tested by ELISA and normalized to 0 µM group. (D) The representative images of immunofluorescent staining for Col1 in NHDF cells. Scale bars, 50 μm. (E) The protein expression of TGF-β1 and Smad3 were detected by Western blot assay. β-actin was used as an endogenous reference. Data are presented as mean ± SEM. Three independent experiments were performed under identical condition. Significant differences are indicated: * p < 0.05, ** p < 0.01, *** p < 0.001. CGA treated vs. 0 µM group by One-way ANOVA test CGA treatment inhibited the skin cellular senescence in vitro To assess the anti-aging effects of CGA, a senescence-associated β-galactosidase (SA-β-gal) assay was performed, where senescent cells appear blue. As shown in Fig. 2A and Supplementary Fig. S1, CGA treatment significantly reduced the number of senescent cells (blue-stained) in a dose-dependent manner (25–200 µM, 24 h) in the NHDF cell line. Senescent cells are characterized by proliferation inhibition and cell cycle arrest, with p21, a key regulator of cell cycle progression, often upregulated. To explore this, the expression of p21 was measured in both NHDF and HaCaT cell lines. CGA treatment significantly downregulated p21 expression at the protein level (200 µM, 24 h; p < 0.001, 0.001, respectively) (Fig. 2B). A similar reduction was observed at the p21 mRNA level in both cell lines (Fig. 2C). Immunofluorescence analysis of p21 in these cell lines also confirmed the protein expression results (Fig. 2D and Supplementary Fig. S2). Cell cycle arrest, a hallmark of cellular senescence, was evaluated by flow cytometry. Compared to solvent controls, the percentage of cells in the G0/G1 phase decreased from 84.3% to 82.7% (CGA 50 µM) or 81.3% (CGA 200 µM) in NHDF cells, and from 52.2% to 51.1% (CGA 50 µM) or 47.6% (CGA 200 µM) in HaCaT cells. Conversely, S-phase cells increased from 5.96% to 7.01% (CGA 50 µM) or 8.32% (CGA 200 µM) in NHDF cells, and from 25.9% to 29.3% (CGA 50 µM) or 31.2% (CGA 200 µM) in HaCaT cells, suggesting a proliferative tendency (Fig. 2E). The accumulation of SASP factors was examined using quantitative real-time PCR and multi-color immunofluorescent staining. As shown in Fig. 2F, CGA treatment significantly reduced the mRNA levels of IL-6 and IL-1β in NHDF cells (200 µM, 24 h; p < 0.01, 0.05). Multi-color immunofluorescence staining further confirmed the attenuation of IL-6 and IL-1β expression following CGA intervention (Fig. 2G). These results demonstrate that CGA effectively inhibits cellular senescence in skin and may influence Col1 production. Fig. 2CGA inhibited the senescence of skin cells in vitro. Normal human dermal fibroblasts cell line NHDF and human epidermal keratinocyte cell line HaCaT were treated with CGA (0,25,50,100,200 µM) for 24 h at 37 °C in 5% CO2. A. The SA-β-gal positive cells in NHDF cell line were detected using a β-galactosidase staining kit as described in materials and methods, and cells in blue were considered as positive for SA-β-gal staining which indicated the senescence cells. B. Protein expression of p21 was tested by Western blot analysis and the results of p21 were normalized to β-actin as density ratio. C. The mRNA expression of p21 (CDKN1A) was evaluated by qRT-PCR, GAPDH was employed as an endogenous reference. D. Representative images showed the level of p21 in different treatment groups by immunofluorescent staining. Scale bars, 20 μm. E. Cell cycle analysis of NHDF and HaCaT cells treated with CGA (50 or 200 µM for 24 h) was detected by flow cytometry. Compared with the control. F. The mRNA expressions of IL-6 and IL-1β were evaluated by qRT-PCR, the results were normalized to GAPDH. G. The representative images of multi-color immunofluorescent staining for IL-6 and IL-1β in NHDF cells. Scale bars, 20 μm. All data are expressed as mean ± SEM. The experiments were done 3 times. * p < 0.05, ** p < 0.01, *** p < 0.001. vs.0 µM by One-way ANOVA test CGA inhibited the senescence of skin cells in vitro. Normal human dermal fibroblasts cell line NHDF and human epidermal keratinocyte cell line HaCaT were treated with CGA (0,25,50,100,200 µM) for 24 h at 37 °C in 5% CO2. A. The SA-β-gal positive cells in NHDF cell line were detected using a β-galactosidase staining kit as described in materials and methods, and cells in blue were considered as positive for SA-β-gal staining which indicated the senescence cells. B. Protein expression of p21 was tested by Western blot analysis and the results of p21 were normalized to β-actin as density ratio. C. The mRNA expression of p21 (CDKN1A) was evaluated by qRT-PCR, GAPDH was employed as an endogenous reference. D. Representative images showed the level of p21 in different treatment groups by immunofluorescent staining. Scale bars, 20 μm. E. Cell cycle analysis of NHDF and HaCaT cells treated with CGA (50 or 200 µM for 24 h) was detected by flow cytometry. Compared with the control. F. The mRNA expressions of IL-6 and IL-1β were evaluated by qRT-PCR, the results were normalized to GAPDH. G. The representative images of multi-color immunofluorescent staining for IL-6 and IL-1β in NHDF cells. Scale bars, 20 μm. All data are expressed as mean ± SEM. The experiments were done 3 times. * p < 0.05, ** p < 0.01, *** p < 0.001. vs.0 µM by One-way ANOVA test The anti-aging effect of CGA is similar to that of resveratrol and vitamin C Several natural products, including RE, metformin, and VC, have been shown to possess anti-aging properties. To compare CGA’s anti-aging effects, RE (100 µM, RE) and VC (100 µM, VC) were used as positive controls [29]. SA-β-gal staining revealed similar inhibitory effects on cellular senescence across the CGA (200 µM), VC (100 µM), and RE (100 µM) groups, compared to untreated cells (24 h; p < 0.01, 0.05, 0.05, respectively) (Fig. 3A). Next, p21 expression was assessed at both the mRNA and protein levels to compare the anti-aging effects of CGA and the positive controls. CGA (200 µM) and VC (100 µM) significantly downregulated p21 mRNA expression (vs. untreated, 24 h; p < 0.01, 0.01, respectively), whereas RE (100 µM) treatment upregulated p21 expression (Fig. 3B). Interestingly, the protein levels of p21 were reduced by both CGA (200 µM) and RE (100 µM) treatments (vs. untreated, 24 h; p < 0.01, 0.01, respectively) (Fig. 3C), with the different regulatory effects of RE warranting further investigation. The expression of SASP factors, IL-6 and IL-1β, was assessed by qRT-PCR, showing a significant decrease in IL-6 mRNA levels following treatment with CGA (200 µM), VC (100 µM), and RE (100 µM) (vs. untreated, 24 h; p < 0.01, ns, ns, respectively). A similar trend of downregulation was observed for IL-1β mRNA expression (Fig. 3D). These results suggest that the anti-aging effects of CGA are comparable to those of RE and VC. Fig. 3The comparable anti-aging effect of CGA, VC and RE. Normal human dermal fibroblasts cell line NHDF was treated with CGA (200 µM), resveratrol (100 µM) and vitamin C (100 µM) for 24 h at 37°C in 5% CO2, DMSO was used as the solvent control. A. The SA-β-gal positive cells in NHDF cell line were detected using a β-galactosidase staining kit as described in materials and methods, blue-stained cells were considered as SA-β-gal positive, indicating senescent cells B. The mRNA expression of CDKN1A was evaluated by qRT-PCR, the results were normalized to GAPDH. C. Protein level of p21 was tested by Western blot analysis and the results were normalized to β-actin as density ratio. D. The mRNA expression of IL-6 and IL-1β was evaluated by qRT-PCR, the results were normalized to GAPDH. All data are expressed as mean ± SEM (n = 3). * p < 0.05, ** p < 0.01, *** p < 0.001. vs. untreated by One-way ANOVA test The comparable anti-aging effect of CGA, VC and RE. Normal human dermal fibroblasts cell line NHDF was treated with CGA (200 µM), resveratrol (100 µM) and vitamin C (100 µM) for 24 h at 37°C in 5% CO2, DMSO was used as the solvent control. A. The SA-β-gal positive cells in NHDF cell line were detected using a β-galactosidase staining kit as described in materials and methods, blue-stained cells were considered as SA-β-gal positive, indicating senescent cells B. The mRNA expression of CDKN1A was evaluated by qRT-PCR, the results were normalized to GAPDH. C. Protein level of p21 was tested by Western blot analysis and the results were normalized to β-actin as density ratio. D. The mRNA expression of IL-6 and IL-1β was evaluated by qRT-PCR, the results were normalized to GAPDH. All data are expressed as mean ± SEM (n = 3). * p < 0.05, ** p < 0.01, *** p < 0.001. vs. untreated by One-way ANOVA test Improvement of mitochondrial function in skin cellular senescence by CGA administration To investigate the relationship between cellular senescence and mitochondrial dysfunction, mitochondrial function was assessed following CGA treatment. Transmission electron microscopy (TEM) revealed improved mitochondrial morphology in NHDF cells treated with CGA, showing less disruption of cristae and more intact, orderly packed cristae structures compared to the untreated group (Fig. 4A). Oxygen consumption rate (OCR) analysis was performed to assess mitochondrial stress, with results showing an increase in both basal and maximal OCR in NHDF cells treated with CGA (vs. untreated; 25–200 µM, 24 h), indicating enhanced OXPHOS in the mitochondrial respiratory chain (Fig. 4B). Correspondingly, total ATP content in both NHDF and HaCaT cell lines was significantly increased by CGA in a dose-dependent manner, showing 1.44- and 1.81-fold changes, respectively (vs. untreated; 200 µM; p < 0.001, p < 0.05) (Fig. 4C). Time-dependent ATP content analysis revealed a more pronounced increase in NHDF cells at 48 and 72 h (vs. untreated; 200 µM; p < 0.001, p < 0.001), whereas no significant increase was observed in HaCaT cells (Fig. 4D). The lack of sustained ATP elevation in HaCaT cells may be related to the potential for excessive proliferation in keratinocytes, which could be associated with skin disorders. Superoxide dismutase (SOD) activity, a common marker of cellular senescence, was significantly enhanced by CGA treatment in NHDF cells, showing a 1.57-fold increase (vs. untreated; 200 µM; p < 0.001) (Fig. 4E). Notably, CGA treatment also reduced ROS production (Fig. 4F), despite the positive correlation between mitochondrial ROS production and enhanced OXPHOS. This reduction in ROS might be attributed to the unique chemical structure of CGA, which may interact with mitochondrial superoxide anions (O2.−), a predominant form of ROS [30]. Mitochondrial membrane potential (Δψm), a key bioenergetic parameter associated with cellular senescence, was assessed using JC-1 staining. In senescent cells, a loss of Δψm is typically observed, with an increase in JC-1 monomer fluorescence (green) and a decrease in JC-1 aggregate fluorescence (red). In contrast, CGA treatment enhanced the aggregate-to-monomer fluorescence ratio in a dose-dependent manner, indicating improved Δψm (vs. untreated; p < 0.001, p < 0.001) (Fig. 4G), reflecting restored mitochondrial function. Additionally, the expression of PGC-1α and TFAM, which are markers of mitochondrial biogenesis, were detected in the NHDF and HaCaT cells. As the results, CGA treatment significantly enhanced TFAM expression (Supplementary Fig. S3). Telomere length, a classical hallmark of aging, was measured alongside mtDNA copy number, and both of which showed improvements following CGA treatment in the NHDF cells (Fig. 4H). These results collectively suggest that CGA improves mitochondrial function, which may partially explain its anti-aging effects. Fig. 4CGA enhanced mitochondrial function in skin cellular senescence. Normal human dermal fibroblasts cell line NHDF and human epidermal keratinocyte cell line HaCaT were treated with CGA (0,25,50,100,200 µM) for 24 h at 37 °C in 5% CO2. (A) NHDF cells were treated with 200 µM CGA for 24 h. Mitochondria detected by TEM and representative images are shown. (B) Mitochondrial stress test of NHDF cells after intervened by CGA. Oxygen consumption rate (OCR) was calculated by the Seahorse XF-96 analysis. (C) The total ATP level was detected by using the ATP assay kit and measuring the luminescence in NHDF and HaCaT cells after exposure to CGA. The ATP level was influenced through a dose-dependent fashion after CGA administration in NHDF cells. (D) Total ATP level was detected after 200 µM CGA treating NHDF and HaCaT cells for 0,24,48,72 h and the level was influenced through a time-dependent fashion in NHDF cells. (E) Levels of SOD in NHDF cells were analyzed by ELISA method and it was influenced through a dose-dependent fashion after CGA administration. G. Changes in mitochondrial membrane potential detected by the JC-1 probe method after 50, 200 µM CGA administration. Green: low Δψm, JC-1 monomer; Red: high Δψm, JC-1 aggregate. H. The telomere length and the mtDNA copy number were detected in NHDF cell line by qRT-PCR under CGA (0,25,50,100,200 µM) treatment for 24 h. All data are expressed as mean ± SEM (n = 3). * p < 0.05, ** p < 0.01, *** p < 0.001. vs.0 µM or 0 h by One-way ANOVA test CGA enhanced mitochondrial function in skin cellular senescence. Normal human dermal fibroblasts cell line NHDF and human epidermal keratinocyte cell line HaCaT were treated with CGA (0,25,50,100,200 µM) for 24 h at 37 °C in 5% CO2. (A) NHDF cells were treated with 200 µM CGA for 24 h. Mitochondria detected by TEM and representative images are shown. (B) Mitochondrial stress test of NHDF cells after intervened by CGA. Oxygen consumption rate (OCR) was calculated by the Seahorse XF-96 analysis. (C) The total ATP level was detected by using the ATP assay kit and measuring the luminescence in NHDF and HaCaT cells after exposure to CGA. The ATP level was influenced through a dose-dependent fashion after CGA administration in NHDF cells. (D) Total ATP level was detected after 200 µM CGA treating NHDF and HaCaT cells for 0,24,48,72 h and the level was influenced through a time-dependent fashion in NHDF cells. (E) Levels of SOD in NHDF cells were analyzed by ELISA method and it was influenced through a dose-dependent fashion after CGA administration. G. Changes in mitochondrial membrane potential detected by the JC-1 probe method after 50, 200 µM CGA administration. Green: low Δψm, JC-1 monomer; Red: high Δψm, JC-1 aggregate. H. The telomere length and the mtDNA copy number were detected in NHDF cell line by qRT-PCR under CGA (0,25,50,100,200 µM) treatment for 24 h. All data are expressed as mean ± SEM (n = 3). * p < 0.05, ** p < 0.01, *** p < 0.001. vs.0 µM or 0 h by One-way ANOVA test Mitochondrial transplantation demonstrated that the anti-aging effect of CGA was associated with mitochondrial regulation To investigate the direct anti-aging effects of CGA through mitochondrial regulation, mitochondria treated with CGA were isolated and transferred to untreated NHDF and HaCaT cell lines via co-culture in vitro (Fig. 5A). The activity of isolated mitochondria was testified by the detection of mitochondrial membrane potential (JC-1 staining) (Supplementary Fig. S4). Mitochondria were specifically labeled with MitoTracker, and the nuclei were stained with DAPI. Confocal microscopy was used to observe mitochondrial transfer after two hours (Fig. 5B). Following mitochondrial transfer, ATP content was significantly increased in both cell lines (vs. Ctrl; p < 0.001, p < 0.01, respectively) (Fig. 5C). In NHDF cells, the activity of SOD was upregulated by mitochondrial transfer, with the effect being more pronounced following CGA-treated mitochondrial transfer (vs. Ctrl; p < 0.05) (Supplementary Fig. S5). To assess the impact of mitochondrial transfer on cellular senescence and collagen production, related changes were examined. As shown in Fig. 5D, the expression of p21 mRNA and protein was notably downregulated following transfer of CGA-treated mitochondria in both NHDF and HaCaT cell lines (vs. Ctrl; p < 0.001, p < 0.01; p < 0.05, p < 0.01, respectively). Moreover, mRNA levels of the representative SASP factor IL-6 were significantly reduced after mitochondrial transfer in the CGA-treated group (vs. Ctrl; p < 0.001, p < 0.001, respectively). A similar trend was observed in IL-1β expression in HaCaT cells (vs. Ctrl; p < 0.01). Interestingly, while IL-1β expression increased following mitochondrial transfer in NHDF cells, CGA-treated mitochondria did not induce this upregulation (Fig. 5E). The SA-β-gal assay further confirmed the anti-aging effect of CGA-treated mitochondrial transfer (Fig. 5F and Supplementary Fig. S6). Additionally, the secretion of Col1 in NHDF cells was assessed. Mitochondrial transfer resulted in increased Col1 secretion, with a more significant increase observed in the CGA-treated mitochondrial transfer group (vs. Ctrl; p < 0.05) (Fig. 5G). The expression of key proteins in the TGF-β/Smad signaling pathway, which regulates Col1 synthesis, was enhanced following CGA-treated mitochondrial transfer in both NHDF and HaCaT cell lines (Col1, vs. Ctrl; p < 0.05, p < 0.05; TGF-β1, p < 0.05, p < 0.01; Smad3, p < 0.05, p < 0.01, respectively) (Fig. 5H and I). In conclusion, CGA-treated mitochondrial transfer effectively improves cellular senescence, with mitochondrial regulation appearing to play a central role in CGA’s anti-aging effects on skin. However, the precise molecular mechanisms require further investigation. Fig. 5The regulation of mitochondrial function by CGA was demonstrated through mitochondrial transplantation. The sufficient NHDF and HaCaT cells were cultured for the isolation of mitochondria. After the treatment by different doses of CGA (0, 200 µM) for 24 h, the mitochondria were isolated and added to the untreated corresponding cells for 24 h. (A) The flow diagram of mitochondrial transplantation. (B) The transfer results were observed by confocal microscopy after 2 h. (C) The content of ATP was evaluated by using the ATP assay kit and measuring the luminescence in NHDF and HaCaT cells after the mitochondrial transplantation. (D) The mRNA expression of CDKN1A was evaluated by qRT-PCR, the results were normalized to GAPDH. Protein level of p21 was tested by Western blot analysis and the results of p21 were normalized to β-actin as density ratio. (E) The mRNA expression of IL-6 and IL-1β were evaluated by qRT-PCR, the results were normalized to GAPDH. (F) The SA-β-gal positive cells in NHDF cell line were detected using a β-galactosidase staining kit, and cells in blue were considered as positive for SA-β-gal staining which indicated the senescence cells. (G) The secretion of collagen I in NHDF cells was analyzed by ELISA method. H-I. Protein level of Col1, TGF-β1, Smad3 were tested by Western blot analysis and the results were normalized to β-actin as density ratio. All data are expressed as mean ± SEM of 3 independent experiments. * p < 0.05, ** p < 0.01, *** p < 0.001. vs. Ctrl group by One-way ANOVA test The regulation of mitochondrial function by CGA was demonstrated through mitochondrial transplantation. The sufficient NHDF and HaCaT cells were cultured for the isolation of mitochondria. After the treatment by different doses of CGA (0, 200 µM) for 24 h, the mitochondria were isolated and added to the untreated corresponding cells for 24 h. (A) The flow diagram of mitochondrial transplantation. (B) The transfer results were observed by confocal microscopy after 2 h. (C) The content of ATP was evaluated by using the ATP assay kit and measuring the luminescence in NHDF and HaCaT cells after the mitochondrial transplantation. (D) The mRNA expression of CDKN1A was evaluated by qRT-PCR, the results were normalized to GAPDH. Protein level of p21 was tested by Western blot analysis and the results of p21 were normalized to β-actin as density ratio. (E) The mRNA expression of IL-6 and IL-1β were evaluated by qRT-PCR, the results were normalized to GAPDH. (F) The SA-β-gal positive cells in NHDF cell line were detected using a β-galactosidase staining kit, and cells in blue were considered as positive for SA-β-gal staining which indicated the senescence cells. (G) The secretion of collagen I in NHDF cells was analyzed by ELISA method. H-I. Protein level of Col1, TGF-β1, Smad3 were tested by Western blot analysis and the results were normalized to β-actin as density ratio. All data are expressed as mean ± SEM of 3 independent experiments. * p < 0.05, ** p < 0.01, *** p < 0.001. vs. Ctrl group by One-way ANOVA test CGA-gel administration improved skin senescence in photoaging animal model UV radiation, a major extrinsic factor contributing to skin senescence, was used to establish a photoaging animal model to accelerate skin aging. To investigate the direct anti-aging effects of CGA, a CGA-containing gel was developed for topical application. The animal model creation and gel production process are depicted in Fig. 6A. UV radiation induced progressively severe skin lesions over time, while CGA-gel (1 mM/kg/d) treatment significantly protected the skin from UV-induced damage (Fig. 6B). Masson staining was conducted to assess collagen organization in the tissues. As shown in Fig. 6C, compared to the control group, collagen fibers in the model group were more disorganized. Analysis of collagen fiber abundance and density showed that UV radiation significantly reduced both parameters, while the CGA-gel treatment reversed this effect. (vs. UV; n = 6, p < 0.001). In the CGA-gel-treated group, collagen organization was notably improved, and both collagen abundance and density were restored. Additionally, the expression of Col1 in skin tissue was assessed at the protein level. CGA-gel treatment significantly promoted Col1 expression compared to the UV radiation group (vs. UV; n = 3, p < 0.001) (Fig. 6D). H&E staining results (Fig. 6E and Supplementary Fig. S7) showed that the epidermal thickness was increased in the UV radiation group, while CGA-gel treatment effectively reduced epidermal thickness to levels comparable to the control group. UV radiation also caused the infiltration of multinucleated giant cells (MGC) in the skin, which may be associated with inflammatory responses. However, the infiltration of MGCs was significantly reduced in the CGA-gel-treated group. Moreover, the body weight curve has shown the safety of CGA-gel (Supplementary Fig. S8). These results suggest that CGA treatment effectively reverses UV-induced skin damage and may exert anti-aging effects in this photoaging model. Fig. 6CGA-gel administration inhibited skin senescence caused by UV in photoaging animal model. Female C57BL/6 N mice were depilated by using hair remover cream for the fixed-area of the mice’s dorsal skin and randomly divided the mice into four groups: Control; UV Model; NS-gel; CGA-gel (1 mM/kg/d). The dose of UVA and UVB were respectively 1.9 J/cm2 and 180 mJ/cm2. An hour after the completion of irradiation applied the gel and CGA gel to the mice. The treatment was performed every day and lasted almost 14 days. A. Schematic diagram of the process of modeling and drug administration. B. Representative photographs of mice’s dorsal skin surfaces from the Control, UV Model, NS-gel and CGA-gel. Each group was exposed to UV irradiation for 2, 7 and 14 days. C. Representative micrographs of Masson staining skin tissue sections from each group of mice after UV irradiation for 14 days. The regions of interest (ROI) are boxed in white, and their magnified photos are shown below. Collagen fibers were stained blue and compared the areas of collagen fibers of each group. Scale bars, 100–20 μm. Collagen fiber abundance and density was analyzed by ImageJ software (n = 6). D. The protein level of Col1 in skin tissues was tested by Western blot analysis and β-actin was used as an endogenous reference (n = 3). E. Epidermal thickness and histopathological changes in the skins of each group were observed by H&E staining. ROIs are boxed in blue, and their magnified photos are shown below. Scale bars, 100–20 μm. Data are presented as mean ± SEM (n = 3 or 6). ***p < 0.001, vs. UV group by One-way ANOVA test CGA-gel administration inhibited skin senescence caused by UV in photoaging animal model. Female C57BL/6 N mice were depilated by using hair remover cream for the fixed-area of the mice’s dorsal skin and randomly divided the mice into four groups: Control; UV Model; NS-gel; CGA-gel (1 mM/kg/d). The dose of UVA and UVB were respectively 1.9 J/cm2 and 180 mJ/cm2. An hour after the completion of irradiation applied the gel and CGA gel to the mice. The treatment was performed every day and lasted almost 14 days. A. Schematic diagram of the process of modeling and drug administration. B. Representative photographs of mice’s dorsal skin surfaces from the Control, UV Model, NS-gel and CGA-gel. Each group was exposed to UV irradiation for 2, 7 and 14 days. C. Representative micrographs of Masson staining skin tissue sections from each group of mice after UV irradiation for 14 days. The regions of interest (ROI) are boxed in white, and their magnified photos are shown below. Collagen fibers were stained blue and compared the areas of collagen fibers of each group. Scale bars, 100–20 μm. Collagen fiber abundance and density was analyzed by ImageJ software (n = 6). D. The protein level of Col1 in skin tissues was tested by Western blot analysis and β-actin was used as an endogenous reference (n = 3). E. Epidermal thickness and histopathological changes in the skins of each group were observed by H&E staining. ROIs are boxed in blue, and their magnified photos are shown below. Scale bars, 100–20 μm. Data are presented as mean ± SEM (n = 3 or 6). ***p < 0.001, vs. UV group by One-way ANOVA test CGA treatment attenuates p21 and SASP expression in photoaging model To evaluate senescence markers in the UV radiation-induced photoaging model following CGA-gel treatment, skin tissues were collected for analysis. p21 expression at both the protein and mRNA levels was assessed. As shown in Fig. 7A, UV radiation significantly increased p21 expression compared to the control group, confirming successful induction of skin senescence. However, in the CGA-gel-treated group, p21 expression was markedly reduced at both mRNA and protein levels (vs. UV; p < 0.05, p < 0.01, respectively). Real-time PCR analysis revealed a significant decrease in IL-6 and IL-1β mRNA expression, which were upregulated by UV radiation, in the CGA-gel group (Fig. 7B, vs. UV; p < 0.001, p < 0.05, respectively). Multi-color immunofluorescence staining further confirmed the reduced expression of these cytokines in the skin tissues (Fig. 7C). Electron microscopy revealed severe mitochondrial degeneration with cristae loss in the skin of photoaging mice, while CGA-gel treatment restored more organized mitochondrial morphology (Fig. 7D). These results demonstrate that CGA-gel effectively downregulated the senescence marker p21 and inhibited SASP production (IL-6 and IL-1β). Thus, CGA-gel administration may offer an effective approach to mitigate skin senescence, warranting further investigation. Fig. 7CGA-gel treatment decreased p21 and SASP expression in photoaging animal model. The skin tissues were collected from mice in Control, UV Model, NS-gel and CGA-gel group. And the total RNA and protein were extracted. (A) The mRNA expression of CDKN1A in skin tissues was evaluated by qRT-PCR (n = 6), the results were normalized to GAPDH. The protein level of p21 in skin tissues was tested by Western blot analysis and the results of p21 were normalized to β-actin as density ratio (n = 3). (B) The mRNA expression of IL-6 and IL-1β in skin tissues were evaluated by qRT-PCR (n = 6), the results were normalized to GAPDH. (C) Representative multi-color immunofluorescence images stained of IL-6 (red) and IL-1β (green) in skin tissues (n = 6). Scale bars, 50 μm. (D) Mitochondria in skin tissues of Ctrl, UV and CGA-gel group were detected by TEM and representative images are shown. Data are presented as mean ± SEM (n = 3 or 6). * p < 0.05, ** p < 0.01, *** p < 0.001, vs. Ctrl or UV group by One-way ANOVA test CGA-gel treatment decreased p21 and SASP expression in photoaging animal model. The skin tissues were collected from mice in Control, UV Model, NS-gel and CGA-gel group. And the total RNA and protein were extracted. (A) The mRNA expression of CDKN1A in skin tissues was evaluated by qRT-PCR (n = 6), the results were normalized to GAPDH. The protein level of p21 in skin tissues was tested by Western blot analysis and the results of p21 were normalized to β-actin as density ratio (n = 3). (B) The mRNA expression of IL-6 and IL-1β in skin tissues were evaluated by qRT-PCR (n = 6), the results were normalized to GAPDH. (C) Representative multi-color immunofluorescence images stained of IL-6 (red) and IL-1β (green) in skin tissues (n = 6). Scale bars, 50 μm. (D) Mitochondria in skin tissues of Ctrl, UV and CGA-gel group were detected by TEM and representative images are shown. Data are presented as mean ± SEM (n = 3 or 6). * p < 0.05, ** p < 0.01, *** p < 0.001, vs. Ctrl or UV group by One-way ANOVA test Discussion Senescence is a complex and prolonged process regulated by various factors. Due to their unique anti-aging effects and minimal toxicity [31], natural products are frequently studied for their potential in anti-aging therapies. This study demonstrates that CGA promoted the synthesis of Col1 in both NHDF and HaCaT cell lines. Significant improvements in cell cycle arrest, p21 expression, and SASP production were observed, suggesting a delay in senescence. Additionally, CGA treatment significantly enhanced mitochondrial functions, including mitochondrial morphology, ATP levels, OXPHOS, and Δψm, as confirmed by mitochondria transplantation assays. CGA administration also restored skin collagen levels and downregulated senescence marker expression in a photoaging mouse model. Together, these findings suggest that CGA attenuates skin aging, at least in part, through modulation of mitochondrial function. Previous studies by Xue et al. demonstrated that CGA reversed the decrease in the TGF-β-Smad2/3 pathway, which regulates Col1 synthesis, and reduced ROS accumulation in NHDF cells exposed to UV radiation [23]. In the present study, CGA enhanced Col1 synthesis even without UV stimulation in vitro, indicating that the collagen protective effect of CGA is not limited to photoaging intervention, but also exerts regulatory effects on intrinsic aging processes such as replicative senescence. Senescent cells are characterized by elevated SA-β-gal activity [32], and CGA treatment significantly suppressed SA-β-gal expression in NHDF cells. CGA also decreased p21 expression and alleviated cell cycle arrest in both NHDF and HaCaT cells. Since SASP factors, such as IL-1β and IL-6, are closely associated with skin senescence and chronic inflammation [33], the accumulation of SASP factors can recruit inflammatory cells and exacerbate the accumulation of senescent cells [34]. Consistent with previous reports showing anti-inflammatory effects of CGA in LPS-stimulated macrophages [35], CGA reduced IL-1β and IL-6 expression in vitro and in vivo, supporting the idea that its anti-inflammatory activity contributes to its anti-aging effects. The relationship between TGF-β/Smad pathway and cellular senescence is complicated. In certain tissue and cell types, activation of TGF-β could induce p21 and SASP-mediated cellular senescence [36]. However, in skin tissue, TGF-β pathway may exert opposing effects [37]. Based on the findings of this study, we hypothesize that CGA exhibits functional selectivity in activating the TGF-β/Smad pathway in skin cells. Specifically, as a canonical pathway mediating Col1 synthesis, CGA-induced activation of this pathway promotes Col1 production, thereby mitigating the age-related decline in Col1 synthesis in skin cells. Together with CGA’s ability to improve mitochondrial function and the energy metabolic environment, these actions collectively abrogate the potential pro-aging effects of the TGF-β pathway. Therefore, CGA might attenuate cellular senescence, and activate TGF-β/Smad pathway, result in collagen increase. Mitochondrial dysfunction is a hallmark of cellular aging and is characterized by impaired respiration, ATP depletion, and excessive ROS generation [7]. In this study, CGA administration significantly improved mitochondrial cristae integrity and organization. The restoration of intact and orderly cristae provides a better spatial arrangement for respiratory chain complexes, thereby enhancing respiratory capacity [38]. Thus, the OCR in NHDF cells was measured after CGA treatment, with increases in both basal and maximal OCR, indicating improved OXPHOS and mitochondrial function. Furthermore, ATP content was significantly elevated by CGA in a dose- and time-dependent manner in NHDF cells. Given that Col1 biosynthesis demands a substantial supply of ATP [39], the collagen synthesis-promoting effect of CGA may be associated with its capacity to enhance mitochondrial function and facilitate ATP production. In contrast, HaCaT cells did not exhibit a time-dependent increase in ATP levels. As keratinocytes, HaCaT cells’ continuous ATP increase might disrupt skin barrier function and potentially contribute to skin diseases [40]. Since upregulation of OXPHOS leads to ROS production, CGA treatment reduced ROS levels despite the upregulation of OXPHOS in NHDF cells in this study. This effect may be attributed to CGA’s molecular structure, particularly the caffeic acid component, which has been shown to interact with O2.−, a predominant form of ROS, thus reducing intracellular ROS levels [30]. Moreover, mitochondrial cristae structural abnormalities induced by pathological factors can also enhance ROS production. The repairing mitochondrial cristae architecture performed by CGA might reduce endogenous ROS generation. The restoration of Δψm by CGA provides further evidence of mitochondrial improvement during the anti-aging process. CGA also increased TFAM expression and mtDNA copy number, suggesting a potential involvement of mitochondrial biogenesis in its anti-aging effects. Because telomere shortening and mtDNA alterations are recognized markers of aging and senescence [41], the observed increases in telomere length and mtDNA copy number further support the protective role of CGA against cellular aging. Mitochondrial transplantation aims to transfer healthy mitochondria from donor cells to senescent cells to replace damaged or dysfunctional mitochondria [42]. Previous studies have suggested that mitochondrial transplantation could serve as a potential therapeutic strategy in combating neurodegenerative diseases and cardiovascular aging [43–45]. In this study, mitochondrial transplantation was employed to investigate the direct regulation of mitochondria by CGA in the anti-aging process. Compared to untreated mitochondrial transplantation, CGA-treated mitochondria exhibited more pronounced anti-aging effects. This finding suggests that the anti-aging efficacy of mitochondrial transplantation may depend not only on the presence of exogenous mitochondria but also on their functional quality. Both CGA-treated and untreated mitochondria reduced IL-6 levels in recipient cells, suggesting that mitochondrial transplantation broadly alleviates cellular inflammatory stress. However, only untreated mitochondria induced IL-1β expression, whereas CGA-treated mitochondria did not. One possible explanation is that mitochondrial transfer transiently perturbs intracellular energy homeostasis, which may activate the NLRP3 inflammasome and subsequently promote IL-1β production [46]. Previous studies have shown that STING activation can enhance NLRP3 expression, while CGA may reduce inflammasome activation through modulation of mitochondrial inflammatory signaling, potentially involving the STING-NLRP3 axis [47, 48]. Therefore, mitochondria derived from CGA-treated cells may carry fewer pro-inflammatory signals, resulting in reduced NLRP3 activation and lower IL-1β production after transplantation. In contrast, HaCaT cells inherently contain abundant IL-1β and pro-IL-1β [49]. The activation of NLRP3 in these cells requires potent extracellular damage signals; alterations solely in intracellular mitochondria are insufficient to activate NLRP3 and induce further IL-1β release [46]. Concurrently, mitochondrial transplantation suppresses the senescence of HaCaT cells, leading to the downregulation of IL-1β. Notably, IL-6 displayed a distinct expression pattern from IL-1β, showing downregulation in both mito and c-mito groups. IL-6 expression is largely regulated by sustained NF-κB signaling. Although mitochondrial transplantation may transiently disturb intracellular metabolism, such perturbations may be insufficient to maintain robust NF-κB activation [50]. Thus, IL-6 exhibited downregulation in both cell lines within the mito and c-mito groups. Interestingly, mitochondrial transplantation increased Col1 expression in NHDF cells regardless of whether the donor mitochondria were derived from untreated or CGA-treated cells, whereas in HaCaT cells, a significant increase was observed only following transplantation of CGA-treated mitochondria. NHDF cells are primary source of Col1 and are highly responsive to improvements in mitochondrial bioenergetics. Therefore, even untreated mitochondria may provide sufficient metabolic support to enhance collagen synthesis. In contrast, HaCaT cells exhibit a more limited capacity for collagen production, which means the beneficial effects of mitochondrial transfer may more dependent on mitochondrial quality. Taken together, our results support the notion that CGA attenuates cellular senescence, at least in part, through regulating mitochondrial function. The UV-induced photoaging mouse model has been widely employed in studies on anti-aging. It was found that UV radiation disrupted collagen beneath the epidermis, worsening skin lesions. Cellular senescence served as the driver of skin photoaging and the bridge of disruption of ECM and UV radiation, and the SASP produced by senescent cells accelerate photoaging [51]. The elevated levels of ROS and SASP secreted by senescent cells trigger inflammatory responses and inhibit the clearance of these senescent cells by immune system [52]. Spatial analysis has found that the senescent cells are more likely to be clustered, especially in photoaged skin, which is related to decreased collagen [53]. In photoaging model, a CGA-containing gel was formulated using poloxamer to facilitate administration. This in vivo efficacy is particularly noteworthy because it demonstrates that CGA can exert its anti‑senescence and pro‑collagen effects via topical administration, a clinically practical route for skin aging intervention. The results showed that CGA-gel increased collagen content and restored a more organized collagen structure, providing in vivo validation of the anti-senescent and pro-collagen effects observed in cultured cells. The gel formulation may serve as a promising approach in skin aging therapy. In recent years, novel senolytic and senomorphic drugs have garnered significant attention due to their potential therapeutic benefits in aging-related diseases, such as Bcl-2 family protein inhibitors, antibiotics, and natural polyphenols, have shown promising effects in the nervous and cardiovascular systems [54]. Resveratrol (RE) and Vitamin C (VC), well-known natural anti-aging compounds [14, 55] exert their effects through mechanisms like antioxidant activity, ROS reduction, and activation of various signaling pathways. This study compared the anti-aging effects of CGA, RE, and VC on skin senescence. The results demonstrated similar effects across these compounds, suggesting that CGA may hold significant potential for anti-aging applications. Given their distinct mechanisms of action, a combination therapy involving CGA, RE, and VC may offer a synergistic approach. Our previous studies demonstrated that CGA induces differentiation and suppresses ATP production in tumor cells [21]. However, in the present study, we found that CGA upregulates ATP levels in normal cells. These opposing effects may be attributed to the distinct mechanisms of action of natural products in different cellular contexts. Normal cells require energy to maintain self-renewal, and a reduction in ATP leads to senescence occurrence [56]. In contrast to normal cells, ATP levels are crucial for tumor cell proliferation and metastasis, and a decrease in ATP level should inhibit tumor progression [57]. Senescent and tumor cells both exhibit a disordered state, a condition indicative of increased intracellular entropy. Since mitochondria are the primary energy suppliers of cells, they may act as entropy regulators in both tumor and senescent cells. Our previous study demonstrated that CGA regulates tumoral mitochondria through inhibiting mitochondrial ACAT1 [58]. This work proposes a novel hypothesis regarding CGA’s direct regulation of mitochondrial function in anti-aging, a concept that has been preliminarily supported by our findings. The dual effects of CGA could be explained by the mechanism of “Negative entropy” (Neg-entropy)—a process that facilitates the transition from a disordered state to a normal, ordered state [59, 60]. The regulation of mitochondria could play a vital role in reducing entropy, and further exploration is needed to fully understand this process. In conclusion, CGA promotes Col1 production and attenuates skin cellular senescence, by combining mitochondrial functional analyses, mitochondrial transplantation, and in vivo validation, this study provides evidence that mitochondrial regulation represents an important mechanism underlying the anti-aging effects of CGA (Fig. 8). Further research will focus on elucidating the detailed mechanisms underlying mitochondrial regulation and the anti-aging effects of CGA. As a safe and efficacious natural product, CGA holds potential for development into novel topical formulations for skincare. In clinical dermatology, its ability to promote collagen synthesis could be leveraged to accelerate skin injury repair and wound healing. Moreover, these properties position CGA as a valuable active ingredient for anti-aging cosmeceuticals. Fig. 8CGA promotes Col1 production and attenuates skin cellular senescence by modulating mitochondrial function CGA promotes Col1 production and attenuates skin cellular senescence by modulating mitochondrial function Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2