Reagents and cells
Adipogenic differentiation medium, osteogenic differentiation medium, Dulbecco's Modified Eagle's medium (DMEM), and fetal bovine serum (FBS) were obtained from Nego (Shanghai, China). Antibodies against CD31, CD29, CD44, CD104, CD90, Von Willebrand factor (vWF), GAPDH, and fluorescein isothiocyanate were purchased from Sigma-Aldrich (St. Louis, MO, USA). Myofibroblasts and EPCs were obtained from Procell Life Science Co., Ltd. (Wuhan, China).
Animals and ethics statement
Animal Care and Utilization Committee at Huashan Hospital of Fudan University supervised all animal procedures (2021JS-241). Male C57BL/6 mice four weeks of age were purchased from SLAC Laboratory Animal Co. Ltd (Shanghai, China). Animals were housed individually and independently in ventilated cages at 24–26°C and constant humidity in a 12-h light/dark cycle environment. Ethics Committee in Huashan Hospital of Fudan University oversaw all experiments. All surgical procedures were performed under anesthesia with an intraperitoneal injection of 30 mg/kg sodium pentobarbital.
ADSC culture, isolation, and identification
Adipose tissue samples from mice in phosphate-buffered saline (PBS) were minced in 0.2% collagenase I (Sigma-Aldrich, St. Louis, MO, USA) and digested for 1 h at 37°C with intermittent shaking. The samples were then cleaned and cultured in DMEM (Sigma-Aldrich, St. Louis, MO, USA) containing 15% FBS (Gibco BRL, Frederick, MD, USA). The tissues samples were centrifuged at 1000 rpm for 10 min to eliminate mature adipocytes. Tissue pellets were subsequently resuspended in DMEM supplied with 15% FBS, 100 μg/mL streptomycin, and 100 U/mL penicillin and cultured at 37°C with 5% CO2. ADSCs at ~80–90% confluency were treated with 0.02% ethylenediaminetetraacetic acid/0.25% trypsin (Sigma-Aldrich, St. Louis, MO, USA) for 5 min at room temperature and then replated. Fluorescein isothiocyanate-conjugated CD90, CD29, CD105, vWF, and CD44 antibodies were used for phenotypic analysis. IgG-matched isotype was used as an internal control for all antibodies. The normoxic ADSCs were cultured in 95% air (20% O2) and 5% CO2.
Multilineage ADSC differentiation
To validate ADSC multilineage differentiation, third-passage mouse ADSCs were cultured in adipogenic differentiation medium (Sigma-Aldrich, St. Louis, MO, USA). They were then stained using Oil Red O after two weeks or cultured in osteogenic differentiation medium (Sigma-Aldrich, St. Louis, MO, USA), followed by staining with Alizarin Red after three weeks.
ADSC-Exo identification and isolation
After reaching ~80–90% confluency, the ADSCs were rinsed with PBS and then cultured in FBS-free endothelial cell growth medium-2MV supplied using 1 × serum replacement solution (PeproTech, NJ, USA) for two more days. Culture medium was collected and centrifuged first at 300 g for 10 min and then at 2000 g for 10 min to eliminate cellular debris and apoptotic cells. After centrifugation at 10,000 g for 0.5 h, the supernatant was filtered using 0.22-μm filters (Millipore, Billerica, USA). Then, 15 mL of the supernatant were transferred to Amicon Ultra-15 Centrifugal Filter Unit (100 kDa, Millipore, MMAS, USA) and centrifuged at 4000 g to achieve a volume of ~1 mL. The ultrafiltration unit was then cleaned using PBS twice and centrifuged at 4000 g to achieve a 1-mL volume. Exosome pellets were resuspended in 500 μL of PBS at 4ºC. Exosome protein content was detected using the Pierce bicinchoninic acid (BCA) Protein Assay Kit (Thermo Fisher Scientific, MA, USA). Exosomes were maintained at -80°C until further use. Western blotting (WB) and transmission electron microscopy were used to characterize the exosomes. Their size was determined using dynamic light scattering with Nanosizer (Malvern Instruments, Malvern, UK). Size distribution with particle radius (nm) were represented on the x-axis and percentage on the y-axis.
Model treatment and establishment
A bleomycin-induced murine fibrosis model was generated following a previous protocol [7]. Then, 100 μL of bleomycin (1 mg/mL) were subcutaneously injected into shaved area on the mouse’s back (1 cm2) with a 27-gauge needle. Injections were carried out daily for three consecutive weeks. Mice in the control group received 100 μL of PBS.
For treatments, bleomycin-induced SSc mice were randomly divided into PBS (controls), Exo, circ-Zfyve9-Exo, mimic+circ-Zfyve9-Exo, and si-GPX4+circ-Zfyve9-Exo groups. Animals in each group were subcutaneously injected with 100 μL of PBS with or without BMSCs (1×106) 22 days after bleomycin induction. Mice were sacrificed 3, 7, and 14 days after the first injection. Skin tissue samples within shaved area were then collected for further study. There were ≥6 mice in every group.
Strand-specific RNA-Seq library and RNA-Seq
Total RNA samples were extracted from skin tissue with or without exosome treatment using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). Then, ~3-μg total RNA samples were processed using VAHTS Total RNA-Seq (H/M/R) Library Prep Kit (Illumina, Vazyme Biotech Co., Ltd, Nanjing, China) to eliminate ribosomal RNA and preserving other RNA, such as mRNA. RNA samples were treated with RNase R (Epicenter, 40 U, 37°C for 3 h) and purified using TRIzol. The RNA-Seq library was created utilizing KAPA Stranded RNA-Seq Library Prep Kit (Roche, Basel, Switzerland), which was subjected to deep sequencing via Illumina HiSeq 4000.
Quantitative real-time polymerase chain reaction (qPCR)
Total RNA samples were extracted from cells using a TRIzol reagent kit (Invitrogen, Carlsbad, CA, USA) following the standard protocol. Then, cDNA samples were synthesized and amplified using a TaqMan miRNA reverse transcription kit. Primers used to assay circ-Zfyve9 expression were as follows: forward, 5′-GGCATCACTGCAGAGCG-3′ and reverse, 5′-CTTATTCCACTTTGTATCC-3′. GPX4 primers were as follows: forward, 5′-TGTGCATCCCGCGATGATT-3′ and reverse, 5′-CCCTGTACTTATCCAGGCAGA-3′. GAPDH primers were as follows: forward, 5′-AGGTCGGTGTGAACGGATTTG-3′ and reverse, 5′-GGGGTCGTTGATGGCAACA-3′. Finally, miR-135 primers were as follows: forward, 5′-GAAATGGTTTTGAAGTCG-3′ and reverse, 5′-CGTTCCAGAGGCTCTAGTT-3′. Then, qPCR was conducted using a TaqMan human miRNA assay kit. The 2−ΔΔCT method was used to calculate relative expression fold changes. U6 and GAPDH were used as internal references.
Luciferase reporter assay
The entire GPX4 and circ-Zfyve93′-UTR fragment containing miR-135 binding sites, which were inserted into downstream firefly luciferase reporter gene in the pmirGLO plasmid (Promega Corp., Madison, WI, USA), was amplified. GPX4 and circ-Zfyve9 sequences interacting with the miR-135 seed region were mutated using site-directed gene mutagenesis kit (Beyotime Institute of Biotechnology, Beijing, China). They were then cloned into the luciferase reporter plasmid. The constructed luciferase reporter plasmids were named pmirGLO-GPX4-3′-UTR-WT, pmirGLO-circ-Zfyve9-WT, pmirGLO-GPX4-3′-UTR-MUT, and pmirGLO-circ-Zfyve9-MUT. For the luciferase reporter assay, the 293T cells were co-transfected using 10-ng pmirGLO control vectors, 20-ng constructed luciferase reporter plasmids, and 25 nM miR-NC or miR-135 using Lipofectamine 2000 (Invitrogen). Luciferase function was detected utilizing dual luciferase reporter system (Promega) two days post-transfection. Renilla luciferase was used for normalization.
Tubule formation assay
In vitro neovascularization was assayed in human fibrin matrices. Following treatment, serum-starved EPCs in endothelial basal medium were seeded onto six-well plates coated with Matrigel (105 cells/well, BD Biosciences, Franklin Lakes, NJ, USA), which were incubated at 37°C for 0.5 days. Tubular structures that formed in the Matrigel were observed and photographed using phase-contrast microscopy. The newly formed tube lengths in ten randomly selected fields per well were measured.
Western blotassay
Cell and tissue lysates were centrifuged at 12,000 rpm and 4°C following protease inhibitor addition. Protein concentration was determined using Pierce BCA kit (Thermo Fisher). Proteins were separated using 10% SDS-PAGE and then transferred to PVDF membranes. Primary antibodies for CD63 (1:600), CD81 (1:600), TGF-β (1:1000), α-SMA (1:1000), collagen I (1:500), and anti-GAPDH (1:1000, Sigma-Aldrich) were used to assay the protein expression. Horseradish peroxidase-conjugated secondary antibody (1:1000, Abcam, USA). The ECL kit (Millipore, Burlington, MA, USA) was used.
Histological examination
Skin tissue samples from each group were fixed in 10% formalin solution and embedded in paraffin. Thin 5-μm sections were stained using Masson’s trichrome, followed by CD31 staining to establish histopathological angiogenesis alternations. The sections were visualized using an Axiophot light microscope (Zeiss, Oberkochen, Germany) or ECLIPSE E600 fluorescence microscope (Nikon, Tokyo, Japan).
ROS production analysis
ROS production in skin scar tissue samples was measured using 2',7'-dichlorofluorescin diacetate (Molecular Probes, Eugene, OR, USA). Lipidosome ROS production in cells was detected using a C11 BODIPY581/591 Lipid Peroxidation Sensor (MKBio, Shanghai, China).
Statisticalanalysis
All data analyses were carried out using GraphPad Prism (CA, USA). Overall survival analysis was performed using Kaplan-Meier curves and log-rank tests. Two-tailed Student’s t tests were used to explore the significance between two groups, and one-way analysis of variance with the post hoc Bonferroni test was used for ³3 groups. Correlations were determined using the Pearson correlation test. Data were expressed using means ± standard deviation. A P value of <0.05 indicated a statistically significant difference.