Human samples

This study was approved by a local institutional review board and ethics committee of Shanghai Ninth People’s Hospital, and it adhered to the tenets of the Declaration of Helsinki (SH9H-2019.T213-3). Human skin samples were collected from mild-to-moderate ptosis in the plastic and reconstructive surgery of Shanghai Ninth People’s Hospital (Shanghai, China) (Supplementary Table 1). All patients had signed the informed consent and agreed the collection of tissue samples for additional histological examinations. For the experiments, skin tissue was weighted, ground and added to TRIzol reagent. The total RNA in the tissues were extracted following the manufacturer’s protocols.

Isolation and identification of dermal fibroblast

Skin tissues under the axillary of mice were harvested. In brief, full-thickness skin harvested from the axillary of newborn mice is cut into pieces, about 0.1 cm2. After repeated cleaning of the debris, the pieces were placed in phosphate-buffered saline (PBS) with 1 mg/mL Collagenase I (Sigma-Aldrich, SCR103) and incubated on a rotation shaker at 37°C for 40 min. Then, the digestion was terminated with a culture medium containing the serum, the culture medium and digestion were removed, and the tissue was washed twice with PBS. Next, tissue pieces were suspended in a complete medium, which consisted of DMEM/F12 with 15% fetal bovine serum (FBS, ExCell Bio, FND500) and 1% penicillin-streptomycin (Gibco, 10,378,016) and incubated at 37°C in a 5% CO2 humidity 95% air incubator. After at least 24 h, fibroblasts “crawl” out of the tissue fragments and attach to the bottom of the cultural flasks. When DFs were at 90% confluent, they were collected and passaged until passage three. For the identification of fibroblasts, the DFs were fluorescently stained with vimentin and DAPI (Sigma-Aldrich, D9542).

Replicative senescence model of fibroblasts

DFs were cultured to the third passage with the conditional medium which was consisted of DMEM/F12 with 15% FBS and 1% penicillin-streptomycin and was changed every two days. Afterward, the conditional medium was adjusted to 10% FBS. Replicative cellular senescence used in this study is a common experimental aging model [52]. We build a replicative senescence model of mouse primary skin fibroblast by repeated passage to simulate the aging process of normal cells. In brief, murine DFs isolated from newborn skin were passaged at a 1:3 ratio and until P10 for the following experiments [53].

Extraction and purification of extracellular vesicles

Extraction of cell-derived EVs using the high-speed centrifugation method has already been reported in the literature [54,55,56]. Briefly, exosome-depleted FBS was prepared to effectively exhaust EVs by ultracentrifugation at 100,000 g for 2 h at 4°C. The supernatant was collected and subsequently filtered by a 0.22-µm filter (Millipore). Extracellular vesicles were isolated from distinct conditioned media samples. These conditioned media were obtained by subjecting dermal fibroblasts to different transfection conditions, including non-transfected cells, cells transfected with empty vectors, and cells transfected with a Gstm2 overexpression vector. The conditioned media were collected 48 h after transfection. DFs from the empty vector group (EVsNC) and Gstm2 overexpression group (EVsGstm2) cells were cultured in DMEM/F12 conditioned medium containing 10% EV-free FBS and 1% penicillin–streptomycin for 48 h and the conditioned medium was collected after cultivation. Thirdly, the conditioned medium was harvested for centrifugation at 300 g, 2000 g and 10,000 g for respectively 10 min, 10 min and 30 min to remove the dead cells and cell debris in the medium; subsequently, the supernatant was centrifuged at 100,000 g for approximately 2 h twice by ultracentrifugation (Beckman Coulter Optima L-90 K ultracentrifuge; Beckman Coulter, Fullerton, CA, USA). Finally, we used PBS to resuspend the required pellets, which were used immediately or stored at –80°C for further experiments.

Characterization of EVs

The typical morphology of the collected EVs was observed using a transmission electron microscope (TEM; HT7800, Japan). EVs were added dropwise to 200-mesh grids and incubated for 10 min at room temperature, then the grids were negatively stained with 2% phosphotungstic acid for 3 min, and the remaining liquid was removed using filter paper, and the grids were analyzed under a transmission electron microscope HT7800. The EVs particle size and concentration were measured by nanoparticle tracking analysis (NTA) at VivaCell Shanghai with Zetaview-PMX120-Z (Particle Metrix, Meerbusch, Germany) and corresponding software ZetaView (version 8.04.02, Germany). Isolated EVs were appropriately diluted with 1X PBS buffer. NTA measurement was recorded and analyzed at 11 positions. The ZetaView system was calibrated with 110 nm polystyrene particles. Temperature was maintained around 23\(^ \circ }\) and 30\(^ \circ }\). Western blots were used to evaluate the expression of the EVs markers, such as Alix (1:1000, Abcam, ab275377), CD63 (1:1000, Abcam, ab216130), CD81 (1:1000, Abcam, ab109201), and Calnexin (1:1000, Abcam, ab22595).

Transfection of DFs

For plasmid transfections, 0.8 × 106 P3 DFs were seeded in 6-well plates and cultured for 24 h. Cells were transiently transfected with polyethyleneimine (PEI) (polysciences, 23966-100) using a cDNA/PEI ratio of 1:3. In the empty vector group (EVsNC), the transfection medium was composed of Opti-MEM™ (Gibco, 31985-070) with 2 µg pcDNA3.1-CD63-L7Ae, 2 µg pcDNA3.1-nanoluc-C/Dbox, and 12 µL PEI. In EVsGstm2 groups, the transfection medium was composed of Opti-MEM™ with 2 µg pcDNA3.1-CD63-L7Ae, 2 µg pcDNA3.1-Gstm2-C/Dbox and 12µ PEI. After 6 h of incubation, the transfection medium was replaced with a complete medium. The cell culture supernatant was collected after 48 h culture, and EVs were separated by differential centrifugation as previously described.

Measurement of GST activity

Senescent fibroblasts treated respectively by PBS, EVsNC, and EVsGstm2 were carefully collected and used to measure GST activity via using a commercial GST activity assay kit (Sangon Biotech, D799612-0100) according to the instruction protocol. Briefly, senescent fibroblasts were homogenized in 500 µL of extraction solution (provided in the kit) with an ultrasonic cell crusher on ice followed by centrifugation (8000 g/min) at 4°C for 10 min. The supernatant was collected and various reagent were added in a 96-well plate in order. After a 5s incubation at room temperature and a 5 min incubation at 37°C, absorbance at 340 nm was individually recorded with a microplate reader. The GST activity of each sample was subsequently estimated.

Determination of the cellular GSH levels

The cellular GSH content was measured using a commercially available GSH assay kit (Nanjing Jiancheng Bioengineering Institute, China). Cells were seeded to 90 cm2 dishes in triplicates and treated with PBS, EVsNC and EVsGstm2. After 48 h, the following operations were conducted strictly according to the manufacturer’s instructions. GSH content was determined with a spectrophotometer at 405 nm.

Internalization of EVs

According to the manufacturer’s instructions, EVs were labeled with CM-Dil red fluorescent membrane linker dye (Beyotime, C1035) as previously described [57]. In brief, 50 µg EVs were labeled by CM-Dil stock solution (5 µL, 1 mg/mL) and incubated at room temperature for 5 min. After incubation, unbound CM-Dil was removed via ultrafiltration centrifugation at 100,000 g for 2 h at 4°C; then, the EVs were resuspended in PBS and repeatedly washed with PBS thrice. DFs were incubated with Dil-labeled EVs (50 µg/mL) for 24 h. Then, DFs were washed three times with PBS and fixed in 4% paraformaldehyde solution. The nucleus was stained by DAPI, F-actin of DFs were stained with 50 µg/mL phalloidin-FITC (YEASEN, 40736ES75), and the images of cellular uptake of EVs were captured by fluorescence microscopy.

In vivo bio-distribution of EVs

Purified EVs (200 µg/mouse) labeled with Dil were suspended in PBS as previously described and injected with a needleless syringe on the back. The mice were treated with light shielded for 6 h and were sacrificed. Their dorsal skins were harvested, and longitudinal frozen Sect. (10 μm) were made. Sections were counter-stained with DAPI before imaging. Finally, the frozen sections were photographed with a fluorescence microscope.

Senescence-associated β-galactosidase (SA-β-gal) staining

The expression of senescence-associated β-galactosidase (SA-β-Gal) was measured by using the SA-β-gal staining kit (Beyotime, C0602) [58]. When the cells’ confluence reached approximately 70%, they were used for experiments such as fixation and staining. After washing with PBS, fixed DFs were incubated with β-gal staining mixture overnight at 37°C. β-gal-positive cells were expressed as the ratio of β-Gal positive cells to total cells.

RNA extraction and quantitative real-time PCR (qRT-PCR) analysis

Total RNA was extracted from DFs or HaCaT cells with Trizol universal reagent (Tiangen, Beijing) following the manufacturer’s protocol. The total RNA (500 ng) of each sample was reversely transcribed into cDNA by HiScript III RT SuperMix for qPCR (+ gDNA wiper) (Vazyme, R323-01), real-time qPCR was performed with ChamQ Universal SYBR qPCR Master Mix (Vazyme, Q711) to quantify gene expression level. The relative gene expression levels were calculated by the 2–ΔƊCt method. The primer sequences used in this study are listed in Supplementary Table 2.

Western blotting

DFs or HaCaT cells were washed three times by a cold PBS. Samples from different treatments were prepared with RIPA buffer (Beyotime, P0013B) containing 1mM phenylmethanesulfonyl fluoride (PMSF, Beyotime, ST2573) and protease inhibitor (Beyotime, P1005). Protein concentrations were measured by BCA protein assay Kit (Thermo Fisher Scientifific, 23,225). For each sample, 30 µg protein denatured was separated by SDS-PAGE and electrophoretically transferred to a PVDF membrane. The blots were blocked with 5% BSA in TBST solution for 60 min, incubated overnight with primary antibodies at 4°C and incubated with the secondary antibody for 2 h at room temperature. Finally, immunoreactive bands were detected by ECL reagents (Keygen Biotech, SQ202L). Band signal intensities were quantified using the ImageJ software (version 1.46r, USA).

5-Ethynyl-2’-deoxyuridine (EdU) staining

For EdU staining, EdU Staining Proliferation Kit (Beyotime, C0078S) was used. In brief, cells were seeded in a 12-well plate and EdU solution was added for 4 h incubation. Then, fixative solution and permeabilization buffer were added for 20 min. Then, reaction mix was added to fluorescently label EdU. The nuclei are stained with Hoechst for 15 min. Finally, DFs were viewed and photographed under the fluorescence microscope.

Scratch assay

To assess the cell migration of DFs and HaCaT cells, the scratch assay was performed by scratching with the tip of a 10 µL pistol. After different treatments, the cells and the wound healing status were observed and photographed at 24 h, 36–72 h under a microscope. Using Image J software, we measured the migrated area.

Measurement of intracellular ROS

The level of intracellular ROS was examined by the ROS assay kit (Nanjing Jiancheng Bioengineering Institute, E004-1-1). Briefly, cells with different treatments were seeded in six-well plates. After the medium was removed, cells were rinsed using 37°C pre-warmed PBS and incubated with 1 mL of 10 µM DCFH‐DA (1:1000, no serum) for 20 min at 37°C. The dye solution was washed away, and cells were washed using 37°C DMEM without serum three times. The stained ROS signals were then photographed using a fluorescent microscope and the absorbance was measured at 485 nm by microplate reader.

Lipid peroxidation assay

Malondialdehyde (MDA) assay kits (Beyotime, China, S0131S) were used in accordance with the manufacturer’s guidelines to measure the level of MDA in aging dermal fibroblast (DFs) and HaCaT cells. Aging cells were seeded at 1 × 105 cells per well into 12-well plates and incubated with the indicated treatments in an incubator of 5% CO2 at 37°C. 30 µl of cell lysis buffer (Beyotime, China, P0013) was added to the cells and incubated for 10 min. The cells were then detached using a cell scraper and transferred to a 1.5 mL tube along with the buffer. The tubes were vortexed every 10 min, and this process was repeated three times. The supernatant lysate from each tube was collected by centrifugation at 12,000 g, 4°C for 10 min. Subsequently, 100 µl of the lysate sample was mixed with 200 µl of malondialdehyde solution and incubated for 15 min at 100°C, protected from light. After cooling to room temperature, the mixtures were centrifuged at 1000 g, 25°C for 10 min. Next, 200 µl of the supernatant from each tube was transferred to a 96-well plate and immediately measured for absorbance at OD 532 nm. Additionally, a standard curve was simultaneously performed according to the manufacturer’s instructions.

In vitro co-culture assay

DFs and HaCaT cells were co-cultured by indirect transwell co-culture. 4 × 105 DFs were seeded in the upper compartment and treated with PBS, EVsNC and EVsGstm2 for 48 h. Then, the supernatant was replaced with fresh medium and 4 × 105 HaCaT cells were seeded in the lower compartment of a transwell membrane. Then, the cells were incubated for another 48 h.

Small interference RNA transfection

Small interfering RNA (siRNA) was purchased from Genomeditech (Shanghai) Co.,Ltd. The HaCaT cells were transfected with NACA siRNA (siNACA) (50 nM) or negative control siRNA (si-NC) (50nM) in mediation of Lipofectamine™ 2000 Transfection Reagent (Invitrogen Inc., Carlsbad, CA, USA). HaCaT cells in each group were seeded in a 12-well plate and cultured in an incubator at 37°C with 5% CO2 until 60% confluence. According to the operation manual of Lipofectamine 2000 Transfection Reagent, cell transfection was performed. The knockdown efficiency was confirmed at 48 h post transfection by qRT-PCR.

Animals

All animal procedures were approved by the institutional Animal Care Committee and Use Committee of the Tongji University for Laboratory Animal Medicine (TJBB06123101). Three young male ICR mice (1-month old, body weight 20 ± 2 g) and 33 aging male ICR mice (12-month-old, body weight 50 ± 5 g) were provided by Shanghai Laboratory Animal Research Center (Shanghai, China). All mice were housed in plastic cages on a 12 h light/dark cycle and were allowed to drink water and eat freely.

In vivo wound healing experiment

In order to detect the effect of EVsGstm2 on wound healing [59], Treatment: 9 mice were randomly selected from 18 aging male ICR mice and were divided into three groups randomly: (a) 3 ICR mice treated with PBS (NC group); (b) 3 ICR mice treated with EVsNC (EVsNC group); (c) 3 ICR mice treated with EVsGstm2 (EVsGstm2 group). These mice were anaesthetized with an intraperitoneal injection of 2% pentobarbital sodium (60 mg/kg), shaved the hair with an electric razor, and a depilatory cream on the backs. The skin was sterilized with betadine and 70% alcohol. Then, full-thickness wounds were prepared on the back using a 6-mm biopsy punch. Immediately, the wound area was surrounded by a plasticine ring (inner diameter: 6 mm; height: 2 mm), which was then imaged using a digital camera. Mice in each group were administered subcutaneously PBS, EVsNC and EVsGstm2 at day 0, 4, 8 and 12, the wound area was observed and photographed to evaluate the wound healing simultaneously. The wound area and healed wound percentage were measured with Image J software, and wound healing were simulated with photoshop and Image J software (version 1.46r, USA).

In vivo natural aging and treatment

12-month-old mice were used for the natural aging model. 9 mice (n = 9) were randomly selected from 18 aging male ICR mice and were divided into three groups randomly. Treatment: 9 ICR mice were randomly divided into three groups of five mice each: (a) 3 ICR mice treated with PBS; (b) 3 ICR mice treated with EVsNC; (c) 3 ICR mice treated with EVsGstm2. In this experiment, PBS, EVsNC and EVsGstm2 were delivered by mini electric nano mesotherapy at days 0, 4, 8, 12. Needleless dermopressurs injector-EVs delivery consisted of one-time injections in 4 different sites evenly on the whole dorsal skin.

Histological analysis

The histological analysis of the skin condition of each group of mice was assessed by H&E and Masson staining. Briefly, mouse skin tissues were embedded in paraffin and horizontally cut into 8 μm slices. According to the manufacturer’s protocol, these sections were stained with H&E and Masson trichrome after deparaffinization with xylene and hydration [60].

Immunohistochemistry (IHC) analysis

IHC analysis was performed with standard procedures described previously [61]. According to manufacturers’ instructions, the following antibodies were used at indicated dilutions: primary Collagen I antibodies (1:200, Abclonal, A1352) and primary MMP2 antibodies (1:200, Servicebio, GB11130).

Immunofluorescence staining

As previously described, immunofluorescence staining was performed with 4% PFA-fixed DFs and PFA-fixed paraffin-embedded tissue slices [62]. PFA-fixed paraffin-embedded tissue slices as previously described. Primary antibodies were: Vimentin (1:200, Abcam, ab92547), GSTM2 (1:200, Abclonal, A13496), and CD31(1:200, Abclonal, A0378, 1:300). All Alexa Fluor secondary antibodies (Molecular Probes) were used at dilutions of 1:200. Samples were counterstained with ProLong Gold anti-fading containing DAPI to label cell nuclei. Afterward, all stained cells were examined and photographed using fluorescent microscopy. The quantification of CD31 staining was conducted by a blinded manner, where the observer was unaware of the experimental conditions. At least five random images were captured for each experimental sample. The percentage of CD31-positive area was calculated using Image J software, and the mean percentage of CD31-positive area was determined for each field of view (n = 5).

RNA sequencing (RNA-seq)

The raw data of RNA-seq were trimmed to remove adapters and low-quality reads through Trim Galore! (version 0.6.4_dev). Hisat2 (version 2.2.1) was used to align the clean reads to the mouse genome (mm10). FeatureCounts (version 2.0.1) was then applied to quantify the reads mapped to the genes. DEseq2 (version 1.28.1) was used to do the differential expression analysis. Genes with a fold change (FC) > 1 and adjusted P-value < 0.05 were considered as differentially expressed genes (DEGs). Gene Set Enrichment Analysis (GSEA) was conducted by clusterProfiler (version 3.16.1) R package.

Statistical analysis

All data were analyzed with GraphPad Prism (version 8.3.0, USA) and calculated as mean ± standard deviation (mean ± SD). Statistical analyses were conducted using the student’s t-test or Analysis of Variance (ANOVA). Three biological replicates were performed for each experiment. Statistical significance was determined by a p-value less than 0.05.