Ethics statement
The animal study was conducted in accordance with the guidelines outlined in the Care and Use of Laboratory Animals and received formal approval from the Ethical Review Board at Xiangya Hospital of Central South University. The animals were kept in specific pathogen-free (SPF) conditions, adhering to a 12/12 hour light-dark cycle at a temperature of 22°C and a humidity level of 50 − 55%. They were given unrestricted access to both food pellets and tap water.
Extraction and culture of primary cells
Primary osteocytes were extracted from the bone matrix of male C57BL/6 mice, specifically from the femurs and tibias, as previously documented [27]. Primary young osteocytes (YO) or senescent osteocytes (SO) were isolated from 2-month-old or 16-month-old mice, respectively. The age of the mice was used as the primary criterion for categorizing osteocytes as “young” or “senescent”. The femora and tibiae were dissected under aseptic conditions, with the periosteum, bone epiphyses, and bone marrow meticulously eliminated using minimum essential medium with alpha modification (α-MEM; Cat. No. PM150421, Procell) supplemented with 2% penicillin and streptomycin (P/S; Cat. No. 15,070,063, Gibco). Subsequently, the bone was crushed into small pieces, approximately 1 − 2 mm in length. These bone fragments were then incubated three times for 30 min at 37 °C in α-MEM containing 10 units mL− 1 collagenase II (Cat. No. 17,101,015, Gibco). Next, the bone pieces were subjected to alternate incubation in a solution of 5 × 10− 3 M ethylenediaminetetraacetic acid (EDTA; Cat. No. E809068, Macklin) at pH 7.4 or collagenase II. The resulting bone particles were then incubated in α-MEM supplemented with 5% fetal bovine serum (FBS; Cat. No. 12,664,025, Gibco), 5% calf serum (CS; Cat. No. A3520502, Gibco), and 2% P/S. The primary osteocytes migrated out of the bone particles within 7 days, and subsequent experiments were performed using cells at passage 0.
Primary BMSCs and primary bone marrow macrophages/monocytes were extracted from bone marrow samples of 6-week-old male C57BL/6J mice following previously published protocols [5, 46, 47]. BMSCs were cultured in α-MEM supplemented with 10% FBS and 1% P/S, while primary bone marrow macrophages/monocytes were cultured in high glucose Dulbecco’s modified eagle medium (DMEM; Cat. No. PM150210, Procell) supplemented with 10% FBS and 1% P/S. Cells in passage 1 were used in all experiments.
All cells were maintained at 37°C in a 95% humidified atmosphere with 5% CO2.
Isolation of osteocyte-derived EVs
To eliminate the possibility of EV contamination in FBS and CS, ultracentrifugation was applied to deplete EVs from both FBS and CS. The primary YO and SO cells were initially maintained for 7 days to establish the cultures. Subsequently, the culture medium was replaced with α-MEM supplemented with 5% exosome-depleted FBS, 5% exosome-depleted CS, and 1% P/S. The cells were cultured with fresh medium for 48 h, followed by a medium replacement. This process was continued over 14 days. To isolate osteocyte-derived EVs, the cell culture medium was then subjected to sequential centrifugation steps at 300 × g for 10 min, 2 000 × g for 30 min, and 10 000 × g for 30 min at 4 °C. The resulting supernatant was further filtered through a 0.22-µm filter (Cat. No. SLGP033RS, Millipore) to eliminate any remaining cellular debris. The EVs were subsequently pelleted through ultracentrifugation at 100 000 × g and 4 °C for 10 h using a Beckman Optima XPN ultracentrifuge. The EV pellets were resuspended in PBS and preserved at − 80 °C, with precautions taken to avoid multiple freeze−thaw cycles before use.
Identification of YO-EVs and SO-EVs
The concentration of EVs was estimated based on the total protein content, determined using a bicinchoninic acid (BCA) protein assay kit (Cat. No. E-BC-K318-M, Elabscience). Nanoparticle tracking analysis (NTA) with a ZetaView PMX 110 (Particle Metrix) was used to analyze the distribution and size of EVs, while Hitachi H-7650 transmission electron microscope (TEM; Hitachi) was used to identify the morphologies of EVs. To assess the effectiveness and stability of each batch of EVs, cellular functional tests and NTA analysis were employed. Flow cytometry on a FACSCANTO II (BD Biosciences) was employed to resolve EV surface marker proteins, including CD63, CD81, and TSG101, as previously described [48, 49], and FlowJo software (Tree Star Inc) was used to analyze the results. The CD63 polyclonal antibody (Cat. No. 25682-1-AP, Proteintech), CD81 monoclonal antibody (Cat. No. SC7637, Santa Cruz Biotechnology), TSG101 polyclonal antibody (Cat. No. 14497-1-AP, Proteintech) were used as primary antibodies. Alexa Fluor 488-conjugated goat anti-rabbit IgG (Cat. No. 111-545-144, Jackson ImmunoResearch) and Alexa Fluor 594-conjugated donkey anti-mouse IgG (Cat. No. 715-585-151, Jackson ImmunoResearch) were used as secondary antibodies.
Osteogenesis and adipogenesis differentiation assay
BMSCs were seeded at densities of 5.0 × 104 cells or 1.0 × 105 cells per well in 48-well plates. Then, the cells were cultured in either osteogenic induction medium (Cat. No. MUBMD-90,021, Cyagen Biosciences) or adipogenic induction medium (Cat. No. MUBMD-90,031, Cyagen Biosciences) to induce osteogenic or adipogenic differentiation, respectively. To evaluate the effects of different treatments, 50 µg/mL of YO-EVs, SO-EVs, EVs from YB-OC-CM, EVs from SB-OC-CM, YOsi−Control-EVs, YOsi−Tpm1-EVs, or 5 µL of concentrated osteoclast culture medium cocultured with young bone slices or senescent bone slices (YB-OC-CM or SB-OC-CM) were added to the above induction medium. During longer culture periods for the Alizarin Red S (ARS) and Oil Red O (ORO) staining, the culture medium was changed every 2 to 3 days. The solvent group (PBS-treated) was set as positive control to ensure that any observed effects were specifically due to the EVs. To assess the direct influence of tropomyosin 1 (TPM1) overexpression on osteoblastic and adipogenic differentiation, we constructed a lentiviral vector to express TPM1 under the EF1A promoter (Fig. S9A). This vector was used to transfect BMSCs, which were then subjected to osteoblastic or adipogenic differentiation protocols.
Two days after induction, the cells were collected and processed for quantitative real-time PCR (qRT−PCR) to examine the expression of osteogenic or adipogenic genes. Alkaline phosphatase (ALP) activity was measured by staining the cells with an ALP stain kit (Cat. No. 40749ES60, Yeasen) or using an ALP assay kit (Cat. No. A059-2-2, Nanjing Jiancheng) at 3 days after induction. In order to identify the presence of mineralized nodules or lipid droplets, the cells were subjected to staining with Alizarin Red S (ARS) solution (G1452; Solarbio) at 7 days following osteogenic induction, or Oil Red O (ORO) solution (G1262; Solarbio) at 15 days following adipogenic induction. F-actin polymerization was revealed by TRITC-labeled phalloidin (Cat. No. 40734ES75, Yeasen). The stained cells were photographed with an inverted microscope (DMI6000B, Leica). The quantification of ALP+, ARS+, and ORO+ areas was conducted using Image-Pro Plus 6 software.
Osteoclastogenesis differentiation assay
Bone marrow macrophages/monocytes were seeded at a density of 1.0 × 104 cells per well in 48-well plates. Then, the cells were cultured in complete medium supplemented with 100 ng mL− 1 receptor activator for nuclear factor κB ligand (RANKL; Cat. No. 315 − 11, Peprotech) as an osteoclastogenesis induction factor. To assess the effects of different treatments, 50 µg/mL of YO-EVs, or SO-EVs was added to the above induction medium. During longer culture periods for the tartrate-resistant acid phosphatase (TRAP) staining, the culture medium was changed every 2 to 3 days. Two days after induction, the cells were collected and subjected to qRT−PCR to examine the expression of osteoclastic genes. The formation of multi-nucleated and large-spread mature osteoclasts was detected by staining the cells with a tartrate-resistant acid phosphatase (TRAP) staining kit (Cat. No. 387 A, Sigma − Aldrich). TRAP+ cells with more than three nuclei were counted as osteoclasts and photographed with an inverted microscope (DMI6000B, Leica).
Preparation of osteoclast resorption culture medium with bone slices
In order to prepare osteoclastic bone resorption culture medium (OC-CM), various culture plates were used to host young or senescent bone slices without periosteum or bone marrow. Following previous reports [16], primary macrophages/monocytes were seeded onto bone slices at a density of 1.0 × 104 cells per well, and this mixture was cultured in osteoclastic induction medium for 7 days. This permitted the development of mature osteoclasts, which were capable of initiating bone resorption during this time. Afterward, fresh osteoclastic induction medium with EV-depleted FBS was introduced to the culture, replacing the older medium. This new medium was collected after two days to obtain the conditional culture media. To demonstrate the resorptive activity of osteoclasts on bone slices, we performed scanning electron microscopy (SEM; S-3400, Hitachi) analysis after a 14-day culture period to observe resorption lacunae on bone slices. Young bone slices, senescent bone slices, or a mixture without bone was used to create the respective conditional medium, YB-OC-CM, SB-OC-CM, or OC-CM. These different types of OC-CM were either concentrated 10-fold or underwent EV purification using ultracentrifugation. The concentrated OC-CM, as well as any collected EVs, were stored at − 80 °C until use.
Inhibition of Tpm1
Mouse Tpm1 siRNA and control siRNA were procured from RiboBio. The specific siRNA serial numbers and sequences employed in this study included the following: si-m-Tpm1-001, siG2107140843153170, GAAGGAAGACAAATATGAA; si-m-Tpm1-002, siG2107140843154262, GACGTAGCTTCTCTGAACA; si-m-Tpm1-003, siG2107140843155354, CAAGCACATTGCTGAAGAT. For the transfection of osteocytes, a concentration of 50 µM siRNA was utilized, followed by the utilization of Lipofectamine 3000 transfection reagent (Cat. No. L3000008, Thermo Fisher), in agreement with the manufacturer’s protocol.
Animals and treatments
C57BL/6 mice were purchased from Hunan SJA Laboratory Animal Co. Ltd. and were allowed to adapt to their environment for one week. In order to investigate the effects of YO-EVs and SO-EVs, 3-month-old or 15-month-old male C57BL/6 mice were randomly divided into three groups: solvent, YO-EVs, and SO-EVs, with n = 5 per group. To study the effects of Tpm1, animals were randomly divided into three groups: solvent, YOsi−Control-EVs, and YOsi−Tpm1-EVs, with n = 5 per group. The solvent control groups were treated with an equal volume of PBS. The intervention for these groups was conducted by tail vein injection. Four weeks after the treatment, bone and blood samples were collected from mice and processed for subsequent experiments.
µCT analysis
All harvested right femora were fixed in 4% paraformaldehyde for 48 h before measurement by µCT scanning. A vivaCT80 scanner (SCANCO Medical AG) was utilized with a voltage of 50 kV, a current of 400 µA, and a resolution of 18 μm per pixel. After scanning, bone reconstruction and visualization were performed using CT Analyser 1.11.0.0, µCTVol 2.2.0.0, and Dataviewer 1.4.3. Trabecular parameters were assessed at the distal femoral metaphysis by defining a region of interest (ROI) starting from 0.15 mm below the distal epiphyseal growth plate and extending proximally for 0.4 mm, allowing for measurement of parameters, including trabecular bone volume fraction (Tb. BV/TV), trabecular bone thickness (Tb. Th), trabecular number (Tb. N), trabecular separation (Tb. Sp). Cortical parameters were measured in a region that constituted 5% of the femoral length at the femoral mid-diaphysis, with the cortical bone thickness (Ct. Th) being assessed.
Cellular internalization of EVs
For the EV internalization experiments, YO-EVs and SO-EVs were labeled with 1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate (DiI; Cat. No. 40726ES10, Yeasen) according to the manufacturer’s instructions. Following the removal of redundant dye through utilization of an Amicon 0.5 mL Ultracentrifugal Filter equipped with a 100 kDa membrane (Cat. No. UFC510024, Millipore), the labeled EVs were subjected to a 3-hour incubation period with BMSCs at a temperature of 37 °C. Subsequently, the treated cells underwent a rinsing process with PBS and were subsequently fixed in a 4% paraformaldehyde solution for a duration of 15 min. The stained cells were then subjected to two rounds of PBS washing before being mounted on coverslips using a commercially available antifade mounting medium containing 4’,6-diamidino-2-phenylindole (DAPI; Cat. No. H-1200-10, Vectorlabs). The internalization of the red DiI-labeled EVs by the cells was observed using a Zeiss ApoTome fluorescence microscope.
Ex vivo biodistribution of EVs
To assess the EV distribution in vivo, YO-EVs and SO-EVs were purified and labeled using 1,1-dioctadecyl-3,3,3,3-tetramethylindotricarbocyaine iodide (DiR; Cat. No. 40757ES25, Yeasen) lipophilic dye according to the manufacturer’s instructions. After a 30-minute incubation period, unincorporated DiR was eliminated using Amicon 0.5 mL Ultracentrifuge Filters equipped with 100 kDa membranes. To evaluate whether EVs could effectively reach the bone, equal amounts of DiR-labeled YO-EVs and DiR-labeled SO-EVs were injected intravenously via the tail vein. After a 12-hour interval, the bone specimens were retrieved and analyzed using a fluorescence tomography imaging system (Cat. No. FMT4000, PerkinElmer) to quantify the fluorescence intensity.
EV tracer transgenic mice
Cd63loxp−mCherry−loxp−eGFP EV tracer transgenic mice were created utilizing the CRISPR/Cas9 homologous recombination technique. Cas9 mRNA and gRNA were generated via in vitro transcription. To generate the Cd63loxp−mCherry−loxp−eGFP EV tracer transgenic mice, a homologous recombination donor vector was constructed. This vector consisted of a 3.4 kb 5’ homologous arm, loxp-mCherry-loxp-eGFP, and a 4.0-kb 3’ homologous arm, which was inserted at the termination codon of the Cd63 gene. Fertilized eggs from C57BL/6J mice were then microinjected with Cas9 mRNA, gRNA, and the donor vector to produce the F0 generation mice. The identification of homologous recombinant F0 generation mice was accomplished through the utilization of long fragment PCR. Dmp1Cre;Cd63loxp−mCherry−loxp−eGFP mice, which have a tracer specific to osteocyte-derived EVs, were obtained by breeding homogenously recombinant F0 generation mice with Dmp1Cre mice. The genotyping of the Dmp1Cre;Cd63loxp−mCherry−loxp−eGFP mice is shown in Supplementary Fig. S2. The genotyping analysis of hemizygous and wild-type mice was performed by PCR, with the Dmp1Cre allele producing a 324 bp PCR product in hemizygous mice and a 167 bp PCR product in wild-type mice. Similarly, the Cd63loxp−mCherry−loxp−eGFP allele produced a 754 bp PCR product in hemizygous mice and a 488 bp PCR product in wild-type mice.
Femora harvested from Dmp1Cre;Cd63loxp−mCherry−loxp−eGFP mice were used for fluorescence detection of mCherry and eGFP signals using a fluorescence microscope. The eGFP antibody (Cat. No. ab290, Abcam) was used to stain eGFP.
qRT−PCR analysis
Total RNA was extracted from cells, bone matrix, and bone marrow samples using TRIzol reagent (Cat. No. 15,596,026, Invitrogen), and cDNA was synthesized from 1.0 µg of total RNA using the Revert Aid First Strand cDNA synthesis kit (Cat. No. K1622, Thermo Scientific). The FTC-3000 real-time PCR system (Funglyn Biotech) was then used to perform quantitative real-time PCR (qRT−PCR) analysis using GoTaq® qPCR Master Mix (Cat. No. A6002, Promega). To assess the relative mRNA expression levels, the relative standard curve method (2–△△CT) with GAPDH as an mRNA normalization control was employed. The following primer sequences were used for qRT−PCR: Dmp1: forward, 5′-GAAAGCTCTGAAGAGAGGACGG-3′, and reverse, 5′-CCTCTCCAGATTCACTGCTGTC-3′; Col1a1: forward, 5′-GCTCCTCTTAGGGGCCACT-3′, and reverse, 5′-CCACGTCTCACCATTGGGG-3′; P16: forward, 5′-TGTTGAGGCTAGAGAGGATCTTG-3′, and reverse, 5′-CGAATCTGCACCGTAGTTGAGC-3′; P21: forward, 5′-TCGCTGTCTTGCACTCTGGTGT-3′, and reverse, 5′-CCAATCTGCGCTTGGAGTGATAG-3′; Runx2: forward, 5′-GACTGTGGTTACCGTCATGGC-3′, and reverse, 5′-ACTTGGTTTTTCATAACAGCGGA-3′; Pparg: forward, 5′-TCGCTGATGCACTGCCTATG-3′, and reverse, 5′-GAGAGGTCCACAGAGCTGATT-3′; Ctsk: forward, 5′-GCGGCATTACCAACAT-3′, and reverse, 5′-CTGGAAGCACCAACGA-3′; Tpm1, forward, 5′-AACGGTGACGAACAACTTGAA-3′, and reverse, 5′-GGAAGTCATATCGTTGAGAGCG-3′; and Gapdh: forward, 5′-CACCATGGAGAAGGCCGGGG-3′, and reverse, 5′-GACGGACACATTGGGGGTAG-3′.
Enzyme-linked immunosorbent assay (ELISA)
The serum levels of osteocalcin (OCN) and cross-linked C-telopeptide of type I collagen (CTX-I) were assessed using mouse ELISA kits E-EL-M0864c and E-EL-M3023 (Elabscience). All ELISA were conducted in accordance with the manufacturer’s instructions, and the protein concentration for each sample was calculated based on the standard curve.
Histomorphometry and immunostaining
Freshly harvested femora were first fixed in 4% paraformaldehyde for 48 h prior to undergoing decalcification in 10% EDTA at pH 7.4 for 7 days. Following decalcification, samples were embedded in paraffin and sliced into longitudinally oriented sections with a thickness of 5 μm. Subsequently, the sections were subjected to staining for osteocalcin (OCN), perilipin (PLIN), and TRAP to identify osteoblasts, adipocytes, and osteoclasts, respectively. Appropriate secondary antibodies were then incubated, followed by countered with DAPI. Images were acquired using either an optical microscope (Olympus CX31) or a fluorescence microscope (Zeiss ApoTome). The abundance of osteoblasts and osteoclasts was quantified as the number per millimeter of the bone surface, denoted as N. OBs/BS/mm and N. OCs/BS/mm. Similarly, the number of adipocytes per square millimeter of marrow tissue was quantified and recorded as N. AdCs/Ar/mm2.
Three-point bending test
The strength of the femur at the midshaft location was measured by subjecting bones to a three-point bending test on a mechanical-testing machine (Cat. No. 3343, Instron). This involved utilizing two end-support points and one central loading point, with the span length between the support points constituting 60% of the total bone length. Bones were subjected to a constant loading speed of 0.155 mm per second until failure occurred. The maximum load (N) was recorded using load−deformation curves obtained through biomechanical measurements.
Proteomic analysis
YO-EVs and SO-EVs (n = 3 for each group) were subjected to label-free quantitative proteomic analysis, which was performed by Jingjie PTM BioLab. Using the criteria of YO-EVs/SO-EVs > 1.5 or < 0.67 and P < 0.05, differentially enriched proteins were identified. To determine the enrichment of functions, the Gene Ontology (GO) database was employed for analysis. The proteins that exhibited a YO-EVs/SO-EVs ratio greater than 1.5 were subjected to mapping onto Gene Ontology (GO) terms. Subsequently, the GO terms that showed significant enrichment were determined using a threshold of P < 0.05. These enriched GO terms were then categorized into three distinct groups: biological process (BP), cellular component (CC), and molecular function (MF).
Statistical analysis
The results are expressed as the mean ± standard deviation (SD), and the sample size (n) for each statistical analysis is specified in the Fig. legends. The statistical significance of differences between various treatments was assessed using either the two-tailed Student’s t-test or one-way ANOVA with Bonferroni post-test. The data analyses were performed utilizing GraphPad Prism 9.1 software. A statistically significant level was set at P < 0.05, P < 0.01, P < 0.001, and P < 0.0001, which were denoted by “*”, “**”, “***”, and “****”, respectively.