Exosomes from human urine-derived stem cells ameliorate particulate polyethylene-induced osteolysis

Background: Wear particle-induced periprosthetic osteolysis is a common long-term complication of total joint arthroplasty, and represents the major cause of aseptic loosening and subsequent implant failure. Currently, there are no effective therapeutic options to prevent osteolysis from occurring and often need revision surgery. Exosomes are important nano-sized paracrine mediators of intercellular communications and can be directly utilized as therapeutic agents for tissue repair and regeneration. Here, we explored the therapeutic potential of exosomes from human urine-derived stem cells (USC-Exos) in preventing wear particle-induced osteolysis. Methods: USCs were characterized by ow cytometry and multiple differentiation potential. USC-Exos were identied by transmission electron microscopy (TEM), dynamic light scattering (DLS) and western blotting (WB). The impact of USC-Exos on osteoblastic differentiation of bone marrow mesenchymal stromal cells (BMSCs) and osteoclastogenesis of RAW264.7 cells were veried in vitro. The effects of USC-Exos on ultra high molecular weight PE (UHMWPE)-induced murine calvarial osteolysis model were tested to evaluate bone mass, inammation, osteogenic and osteoclastic activities. Results: USCs were positive for CD44, CD73, CD29 and CD90, but negative for CD34 and CD45. USCs were able to differentiate into osteogenic, chondrogenic, and adipogenic cells. USC-Exos exhibited a round-shaped morphology with a double-layered membrane structure and positive for CD63 and TSG101, negative for Calnexin. In vitro, USC-Exos could promote the osteogenic differentiation of BMSCs, reduces the production of proinammatory factors in macrophages and suppresses their osteoclastic abilities. In vivo, injection of USC-Exos into the center of the calvariae caused less inammatory cytokine generation and less osteolysis compared to control and signicantly enhanced the bone formation. Conclusions: Our ndings demonstrate that USC-Exos can prevent UHMWPE-induced osteolysis by inciting less inammatory, inhibiting bone resorption and stimulating bone formation. USC-Exos may represent a potential natural agent for the treatment of periprosthetic osteolysis and to obtain therapeutic exosomes for osteolytic treatment, aseptic loosening patients may just need to collect a certain volume of their own urine to harvest USCs and USC-Exos. inhibit osteoclastic differentiation of osteoclast precursor cells in vitro. Our data suggest that USC-Exos may represent a novel therapeutic tool for orthopaedic wear-debris associated osteolysis.


Introduction
Total joint arthroplasty (TJA) is a highly successful procedure in the case of severe joint diseases such as osteoarthritis, complex fractures, osteonecrosis and rheumatoid arthritis; TJA provides effect pain relief and improves the quality in patients' life with joint disease [1,2]. Approximately 1.5 million joint arthroplasties are performed annually worldwide, and the demand for these surgeries is predicted to increase substantially [3]. Despite continuing technical innovation and a new understanding of biomechanics, a substantial decline in the revision rates has not occurred over the past decades [4].
Aseptic loosening, despite excellent outcomes of the joint arthroplasties, is the most common cause of arthroplasty component failure requiring revision arthroplasty, representing close to 75% of all cases, with severe consequences for both the patient and the whole health-care system [5]. Although the pathophysiology of aseptic loosening is not completely explained, growing evidence suggests that periprosthetic osteolysis initiated by in ammatory responses to implant wear debris is responsible [6]. It is now well established that many particulates produced by prosthetic components, including PE, metal, alumina and polymethylmethacrylate (PMMA) are bioactive, and are therefore implicated in the initiation and or progression of osteolysis. However, periprosthetic osteolysis resulting from the generation of ultra high molecular weight polyethylene (UHMWPE) wear particles in arti cial components is one of the major long-term complications following total joint arthroplasty [7]. The classic paradigm of periprosthetic osteolysis begins with the generation of wear particles that stimulate several different types of cells such as macrophages and T lymphocytes to secrete a variety of proin ammatory chemokines and cytokines, such as tumor necrosis factor-a (TNF-α), interleukin (IL)-1b, and IL-6 [8, 9]. These molecules in turn stimulate osteoclast precursors, and at the same time act on other target cells, mainly osteoblasts and bone marrow stromal cells, to promote their expression of receptor activator of nuclear factor-ĸB ligand (RANKL), a critical regulator of osteoclastogenesis. Upon binding to its receptor RANK on the surface of osteoclasts and their precursors, RANKL triggers activation of multiple intracellular signaling pathways to promote osteoclast differentiation and activity whereas suppressing its apoptosis. As a result, osteoclasts are excessively formed and activated at the bone-periprosthetic sites, eventually leading to extensive bone loss [8,9]. The in ammatory reaction, bone resorption, and formation of foreign-body granulomas represent key processes leading to osteolysis [10]. Because osteoclasts are critical in the osteolytic process, compounds that speci cally target functional osteoclasts are candidate agents for the prevention and treatment of wear particle-induced osteolysis and pathological bone loss. To date, no drug therapy to prevent or inhibit periprosthetic osteolysis has been approved, and aseptic loosening of prostheses can be overcome only by surgical revision.
Stem cell biology is a promising way that mains signi cant promise in the repair of injuries. Growing evidence indicates that most transplanted stem cells do not embed into the injured sites of recipient animal models, and a single injection of stem cells is su cient to alleviate disease phenotypes and maintain therapeutic e cacy for a long time [11,12], which suggests that paracrine action may be a key mechanism underlying the bene cial effects. Stem cells are capable of releasing different types of extracellular vesicles (EVs, 100-1,000 nm in diameter), such as exosomes, microvesicles, or microparticles [13]. Recently, exosomes, 30 -150 nm in diameter, arising from the luminal membranes of multivesicular bodies (MVBs) and secreted into the extracellular milieu by most cell types through fusion with the cell membrane, have been of increased interest in the regenerative medicine eld [13,14].
Exosomes are composed of a lipid bilayer containing transmembrane proteins and enclosing cytosolic proteins and RNA. They can diffuse into the neighbouring cells or be carried via systemic transport to distant anatomic locations where they can induce signal transduction or mediate the horizontal transfer of information in speci c recipient cells [13,14]. Accumulating studies have revealed that direct treatment with these secreted vesicles can stimulate bone formation and avoid many risks associated with stem cell transplantation therapy [11,15,16]. However, the use of these stem cells is often limited by the source or genetic and epigenetic variations, and isolation of stem cells is also invasive and painful [17,18,19]. It is imperative to look for a new stem cell source from which it is easy to obtain abundant exosomes for bone remodeling and regeneration.
It has been demonstrated that a subpopulation of stem cells can be easily isolated from human urine [17,20,21,22,23], i.e. urine derived stem cells (USCs). These cells possess biological characteristics with stem cell characteristics, such as cell surface marker expression pro les, clonogenicity, cell growth patterns, expansion capacity, multipotent differentiation, proangiogenic paracrine effects and easilyinduced pluripotent stem cells [21]. A major advantage to using USCs is that these cells can be abundantly and noninvasively obtained. Chen et al used human urine-derived stem cells (USCs) to generate exosomes (USC-Exos) and found that USC-Exos could signi cantly promote osteogenesis and inhibiting osteoclastogenesis in osteoporotic mice [17]. Also USCs could serve as cell source for bone tissue regeneration. Guan et al. reported the ability of USCs seeded onto a β-tricalcium phosphate scaffold to induce bone healing in rats with femoral defects [24]. Considering that exosomes are important mediators of cell activity [13,14], it would be interesting and meaningful to explore whether USC-Exos have the ability to prevent or inhibit osteolysis in the hope of identifying an innovative therapeutic strategy for attenuating UHMWPE-induced osteolysis to some extent.
In this study, we determined the effects of USC-Exos on osteogenic activity and osteoclast formation by coculturing exosomes with rat bone marrow mesenchymal stem cells (BMSCs) as well as mouse macrophages cell line RAW264.7. In vivo, we further veri ed the potential e cacy of USC-Exos to alleviate wear-debris-induced osteolysis in a murine osteolysis model. Our data indicated that the USC-Exos could prevent osteoclast differentiation and promote bone regeneration by inhibiting the in ammatory process and enhancing osteogenesis, potentially providing new ideas for treating periprosthetic osteolysis (Scheme 1).

Materials And Methods
Particle generation UHMWPE particles (made available from Shamrock Tianjin TEDA, China) in the size range of 0.5 -5 µm were used to model wear debris particles generated in vivo and in vitro. The particles had a mean diameter of 3.8 µm. Macrophage in ammatory response to wear particles has been shown to depend on the size and shape of the wear particles with particles in the size range of 0.5 -5 µm eliciting a strong in ammatory response [25].
The particles were rinsed in 70% ethanol and 100% ethanol respectively, and were sterilized with cobalt 60γ -radiation in a vacuum. PE are white micronized powder and have a density of 0.97 g/cm 3 , thus the particles were suspended in absolute ethanol rstly. The particle suspension was determined to be endotoxin free using a BIOENDO Endotoxin Quantitation Kit (Xiamen Bioendo Technology Co., Ltd., Xiamen, China). The size distribution and Zeta potential of UHMWPE was measured by dynamic light scattering (DLS) with a Nanosizer™ instrument (Malvern Instruments, Malvern, UK).
Mouse macrophage culture and co-culture of macrophages and PE particles RAW 264.7 mouse macrophage cells were cultured in high glucose DMEM (Shanghai BasalMedia Technologies Co., Shanghai, China) supplemented with 10% FBS (Gibco) and 1% penicillin-streptomycin (Gibco) at 37℃ in a humidi ed 95% air/5% CO 2 atmosphere. Because UHMWPE density is less than water, PEs oat on the surface of the cell culture media, presenting a challenge to a conventional experimental set-up. A number of studies have used various techniques to con ne UHMWPE to the bottom of well plates such as inverted cell culture, but the system have several disadvantages: 1) it increases the probability of cell contamination when changing the cell culture media; 2) most importantly the experimental methods is not simple [26,27]. We have therefore suspended sterilized UHMWPE in culture medium, then boasted and aspirated repeatedly the medium with a pipette for 30 minnutes until the color of the medium became pink.
Isolation, culture and identi cation of human USCs The isolation and characterization of USCs were performed as described previously [17,23]. The process of collection and isolation of urinary cells was showed in Fig. S1. Brie y, the mid-stream of human urine samples (50 mL) were collected into a sterile 50 mL tube containing 500 μL Antibiotic-Antimycotic (100 x stock; Gibco, Grand Island, USA) from a healthy male donor with age of 30 years. Then transferred the 50ml urine into four 15ml tube, after centrifugation at 400 x g for 10 min, the supernatant was aspirated and only 1 mL was left in the bottom of tube. 10 mL sterile PBS was added to the tube, after centrifugation at 200 x g for another 10 min, the supernatant was carefully discarded, leaving only 0.2 mL plus the pellet. All procedures were performed at room temperature. The cell pellet was resuspended in 1 mL primary medium (Lonza, USA) and transferred into one wells of a 12-well plate. Cells were maintained at 37 °C in a humidi ed atmosphere containing 5% CO 2 . Check the plates under the microscope (Leica DMI6000B, Solms, Germany) daily and continue to incubate until cells reach 80-90% density. Earlypassage USCs (p2-6) were used for the downstream experiments.

Isolation and identi cation of USC-Exos
USCs-Exos were isolated from culture medium using a protocol modi ed from a previous study [17,29].
The process of isolation and characterization of USCs-Exos was showed in Fig. S2. After reaching 70-80% con uence, USCs were washed with PBS and incubated with freshly prepared complete medium (Lonza, USA) containing exosome depleted fetal bovine serum (Shanghai VivaCell Biosciences Ltd, Shanghai, China) for 48 h. The conditioned medium of USCs was collected and centrifuged at 300 ×g for 10 min, 2000 ×g for 30 min, and 4000 ×g for 30 min. Then the supernatant was ltered using a 0.22 µm lter (Millipore, Billerica, USA). 15 mL supernatant was added to an Amicon Ultra-15 Centrifugal Filter Unit (10 kDa; Millipore) and centrifuged at 4000 ×g to about 1 mL. One-fth volume of Exoquick Exosome Precipitation Solution (System Biosciences, USA) was added to the ultra ltration liquid and mixed well by inverting. After incubation for 12 h, the mixture was centrifuged at 1500 ×g for 30 min and the supernatant was removed by aspiration. The exosome pellets were resuspended in 100-200 μL PBS according to the volume of exosome pellets. All procedures were performed at 4 ºC. The protein content of exosomes was determined by the BCA Protein Assay Kit (Multi Sciences LTD., Hangzhou,China). Exosomes were stored at -80 °C or used for the downstream experiments.

Exosomes uptake assay
To determine whether RAW264.7 can uptake USC-Exos, we stained exosomes with PKH26 uorescent dye (Sigma) according to the manufacturer's instruction. Exosomes (800 µg) diluted in PBS were added to 1.0 ml Diluent C. In parallel, 4μl PKH26 dye was added to the exosome solution for 5 minutes at room temperature while avoiding light. Cells were seeded on glass coverslips pretreated with TC (Solarbio) and cultured for 24 h to reach 80 -90% con uency. We then co-culture a total of 25μL (100 ug) USC-Exos with RAW264.7 in serum-free media. Images were captured with uorescence microscope (Leica, Germany).

Osteoclastic differentiation assay
The osteoclast progenitor RAW264.7 cells were plated in 48-well culture plates at 1.0 × 10 4 cells per well and incubated overnight. After that, the medium were randomized into the following treatment conditions 1) Vehicle group (DMEM + PBS); 2) UHMWPE + RANKL + vehicle group ( DMEM + UHMWPE + RANKL + PBS); 3) UHMWPE + RANKL + USC-Exos group ( DMEM + RANKL + PBS). The medium was changed to fresh complete DMEM or DMEM + UHMWPE containing 100 ng·mL-1 RANKL (ProteinTech, Chicago, USA) and 300 μg·mL-1 USC-Exos or vehicle (PBS). The culture medium was changed every 2 days. After 7 days of induction, the cells were washed with PBS and xed in 4% PFA for 10 min. Osteoclasts were stained using a commercially TRAP Kit (Sigma) and then quanti ed using an inverted microscopy (Leica). The numbers of TRAP + osteoclasts (> 3 nuclei) in each well were counted under a microscope (Leica). After 4 days of induction, the CM were harvested and centrifuged at 2000 ×g for 10 min to collect the supernatant, which was stored at -80 °C or used for ELISA.

Osteogenic differentiation assay
BMSCs were isolated from 4-week-old male C57BL/6 mice. Brie y, bone marrow cells were obtained from the femurs and tibia of mice. After rinsing by centrifugation, cells were resuspended in α-MEM (Gibco, Grand Island, USA) containing 10% FBS (Gibco), 1% penicillin (100 U/mL, Gibco) and streptomycin (100 μg/mL, Gibco). BMSCs were plated in 48-well plates at 1.0 × 10 5 cells per well. After 24 h, the cells were cultured in osteogenesis induction medium (Cyagen Biosciences Inc, Santa Clara, USA) + UHMWPE (1.0 mg/ml) treated with or without USC-Exos (100 μg/ml). The medium was changed every two days. 4 days later, the conditioned media were obtained and centrifuged at 2000 × g for 10 min to remove cellular debris. Then the supernatant was assayed with an Alkaline phosphatase assay kit and Calcium assay kit according to the manufacturer's protocol (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). After 12 days of induction, the cells were washed with PBS, xed in 4% PFA for 10 min, and then stained with Alizarin Red S (ARS) solution (Cyagen Biosciences Inc, Santa Clara, USA). An inverted microscope (Leica DMI6000B, Solms, Germany) was used for imaging.
Western blotting 5× protein-loading buffer (5×; Beyotime Biotechnology, Jiangsu, China) was added directly to the lysates of USCs or exosomes and heated at 95 ºC for 5 min. Next protein extracts were loaded and resolved in 12% SDS-PAGE. The protein sample was run at 120 V for 45 minutes and transferred onto polyvinylidene di uoride membranes (Immobilon P, Millipore, USA) for 1.5 hours at 100 mA. The membranes were incubated with primary antibodies at 4 ºC overnight and then incubated with the secondary antibodies at room temperature for 1 h. Primary antibodies and dilutions were used as follows: anti-CD63 (sc-5275, 1:500, Santa), TSG101 (ab125011, 1:1000, Abcam) and Calnexin (ab22595, 1:1000, Abcam). All the secondary antibodies (1:5000) were obtained from Cell Signaling Technology. The immunoreactive bands were visualized using enhanced chemiluminescence reagent (Thermo Fisher Scienti c, Waltham, USA) and imaged by the ChemiDoc XRS Plus luminescent image analyser (Bio-Rad).

UHMWPE-induced calvarial osteolysis model and surgical treatment
Animal care and experimental procedures were approved by the Department of laboratory Animal Management Committee of Central South University (No. 2020sydw0972). Thirty female BALB/c mice (8 -10 weeks old) were purchased from Hunan SJA Laboratory Animal Company (Changsha, China). Each mouse weighed 20-24 g at the beginning of the experiment. The in vivo calvaria experiments were performed with reference to similar previous experiments [32] with a little modi cation. Brie y, mice were anesthetized by intraperitoneal injection of ketamine and xylazine following the standard protocol. Following anesthesia, skin on the calvariae was shaved, disinfected and then incised with a sharp scalpel along the middle line. Subsequently, 30 mg of UHMWPE particles were evenly spread on the surfaces of the bilateral parietal bones, followed by closure of the surgical skin incision using 4-0 Prolene sutures. A total of 30 mice were randomly assigned to three experimental groups of 10 each: Vehicle control (sterile PBS), UHMWPE group, UHMWPE + USC-Exos group. After subcutaneously injected to the center of the calvariae once postoperation, USC-Exos injections continued to be given once a week for another three times, and the control mice were local injected with PBS (the vehicle of USC-Exos). The mice were sacri ced in a carbon dioxide chamber at day 28 and their calvariae were isolated and xed with 4% PFA for 2 days. After xation, calvarial samples were washed three times with PBS and then stored in PBS at 4 °C until further analysis.
In Vivo Fluorescent Imaging for Biodistribution of USC-Exos USC-Exos (800 µg) were incubated with DiR (D12731 ,Thermo Fisher Scienti c) for 20 min in a nal volume of 500 µl at 37℃ while avoiding light, then washed with PBS, and the samples were then processed as with the PKH26 dye in front. The BALB/c mice were anaesthetized by intraperitoneal injection of pentobarbital (50 mg/kg) and USC-Exos -DiR (n = 3) or PBS (n = 3) was subcutaneously injected to the center of the calvariae. Fluorescence images were obtained after day 1, day 3, day 6, day 8, day 10, day 13 and day 15 by a uorescence tomography imaging system (FMT-4000; PerkinElmer, USA). 15 days after the injection, the mice were sacri ced for ex vivo tissue distribution analysis.

Micro-CT analysis
μCT was used to quantitatively evaluate the bone mass. The cranium were dissected from mice, xed for 48 hours in 4% paraformaldehyde (PFA) and analyzed by high-resolution μCT (Skyscan 1176; Skycan, Aartselaar, Belgium). The voltage, current, and resolution were set to 50 kV, 400 μA, and 8.88 μm per pixel, respectively. For quantitative analysis, 100 μCT slices in the center of each calvaria were used, and region of interest (ROI) was selected as previously described. 32 CTAn software (SkyScan, Aartselaar, Belgium) was utilized to generate three-dimensional (3D) images and calculate the percentage of bone volume out of total tissue volume (BV/TV). 3D μCT images were visualized using Mimics v10.01 software (Materialise, Leuven, Belgium). The number of pores and percentage of porosity were counted by ImagePro plus 6.0 software.

Histological and immunohistochemical analysis
After micro-CT scanning, the calvaria samples were decalci ed in 10% EDTA (pH 7.4) (Amresco, Solon, OH, USA) for 5 days and embedded in para n. Tissue coronal sections at the middle level (5 μm) were stained with hematoxylin & eosin (H&E) to evaluate the degree of bone erosion. Three separate sections per specimen were evaluated in a blinded fashion. The regions containing the discontinuous and nonosseous tissues were considered as the osteolytic areas, and selected for histomorphometric analyses. The periosteum thickness (mm) were measured and quanti ed. Osteoclast-like cells were identi ed by histochemical tartrate-resistant acid phosphatase (TRAP) staining using a commercial kit (Sigma-Aldrich, St. Louis, MO) to identify the the number of osteoclasts. The percentage of osteoclast surface per bone surface (OCs/BS, %) were determined for each sample. The stained sections were examined under a light microscope (Olympus CX31; Olympus Optical Co., Tokyo, Japan) and digital photomicrographs were captured and analyzed using a computerized image analysis system with Image-Pro Plus software, version 6.0 (Media Cybernetics, Silver Spring, MD, USA).
For immuno uorescence staining for OCN and TNF-α, the sections were rehydrated and heated in a microwave in citrate buffer (0.01 M; pH 6.0) for 15 min to retrieve the antigen. The sections were then blocked in 1.5% normal goat serum for 30 min at room temperature, incubated with the primary antibodies anti-OCN (1:100; Abcam) or anti-TNF-α (1:500; Abcam)overnight at 4°C, and then incubated with the respective secondary antibodies (1:250; Abcam) at room temperature for 1 h while avoiding light. The color was developed by adding 3,30-diaminobenzidine tetrahydrochloride (DAB). In negative control sections, the primary antibody was replaced with 1.5% normal goat serum. Images were acquired with a uorescence microscope (Leica). Relative staining intensity or positively stained cell number were measured in three random visual elds per section, three sequential sections per mouse and three mice in each group.

Safety examination
To determine the immunogenicity of the USC-Exos, the blood was collected after anesthesia and subsequently analyzed by hematology and blood chemistry tests. The safety of the USC-Exos was further evaluated by examining the main organs via histological sectioning and H&E staining. Samples from untreated mice were used as controls.

Enzyme-linked immunosorbent assay (ELISA)
The concentrations of TNF-α and IL-6 in the CM were measured using a commercial Mouse TNF-α ELISA Kit (ab208348, Abacam) and a Mouse IL-6 ELISA Kit (ab222503, Abacam) according to the protocol provided by the manufacturer. The optical density of each well was determined using a microplate reader (Bio-Rad 680, Hercules, USA) set to 450 nm. Wavelength correction was set to 570 nm. The protein concentration for each sample was calculated according to the standard curve.

Statistical analysis
Data are presented as mean ± standard deviation (SD). Statistical analysis of multiple-group comparisons was analyzed by one-way analysis of variance (ANOVA), followed by the Bonferroni post hoc test to assess the signi cance of differences between two groups. Analyses were performed using Prism 8.0 software. Differences were judged to be statistically signi cant when P ≤ 0.05.

UHMWPE microparticle characterization
Morphology, size distribution and Zeta potential of UHMWPE was determined by scanning electron microscopy (SEM) and DLS (Fig. 1A, 1B, 1C). A representative SEM image of the UHMWPE depicts the characteristic irregular shape and rough surface of these particles (Fig. 1A). On the basis of size distributions, over 90% of the particles were in the size range of 0.5 -5 μm (Fig. 1B). The Zeta potential is "slightly negative" (Fig. 1C), this "slightly negative" Zeta potential is ideal for cytophagy. The curve of endotoxin were as presented in Fig. 1D, and the detection of the UHMWPE endotoxin concentration at 2 h is 0.052 EU/mL.

Identi cation of USCs and USC-Exos
Single, small, compact"rice-grain" like cells was observed 2-3 days after initial seeding (Fig. S3). A few days after being placed in a well a single cell formed a cluster of cells that appeared small, compact and uniform (Fig. S3). Cell colonies were observed in the USCs cultured plates approximately 3 to 10 days after initial plating (Fig. S3). After 12 days, the broblast-like cells retained a robust proliferation capability and reached 80%-90% con uence (Fig. S3). Most of the adherent cells exhibited a broblastlike morphology ( Fig. 2A). Fig. 2B reveal that after culturing in speci c induction medium, USCs demonstrated to differentiate into an osteogenic, chondrogenic or adipogenic lineage as indicated by the positive staining for Alizarin Red, toluidine blue and Oil Red O, respectively. Moreover, the CCK-8 assay showed that the cells underwent a rapid growth phase from day 3 to day 7 and the growth slowed down after day 7 (Fig. 2C). In addition, ow cytometry results showed that USCs were highly positive for MSC surface markers including CD29, CD44, CD73 and CD90, but negative for CD34 and CD45 (Fig. 2D). The data indicated that cells isolated from human urine and maintained under speci c culture conditions were classi ed as MSC [17,20,33,34]. The morphology of exosomes was observed using TEM (Fig. 2E), which revealed that the particles were a sphere-shaped morphology with a double-layered membrane structure, similar to previously described [13,17,35]. Second, the size distribution of the extracellular particles measured by DLS demonstrated that these particles predominantly ranged from 30 nm to 100 nm (Fig. 2G), which was corresponding with the previously reported exosomes size distributions [13,17,35]. Finally, western blotting results showed that USCs-Exos were positive for exosomal speci c markers CD63 and TSG101, and were negative for Calnexin (Fig. 2F).

Internalization of exosomes by RAW 264.7
To investigate whether exosomes could enter into the cytoplasm of RAW 264.7, we incubated the labeled exosomes with RAW 264.7. Fluorescence microscopy analysis (Fig. 3B) showed that the USC-Exos labelled with PKH-26 (the red dots) was internalized by the RAW 264.7. At 24h post incubation, a large number of exosomes have been internalized and distributed in the perinuclear region.
USC-Exos inhibit osteoclast formation and downregulate in ammatory cytokines 1 mg/ml UHMWPE were cultured with RAW 264.7, and results showed that: (1) RAW 264.7 cells are capable to phagocyte particle (dark dots), as shown in Fig. 3A; (2) the concentration of particles did not seriously affect the cell viability (Fig. 3C).
To examine the effects of USC-Exos on osteoclastogenesis and the production of in ammatory cytokines, osteoclast precursor cell line RAW264.7 cells were induced by the osteoclastogenic factor cells cultured with RANKL and UHMWPE exhibited a large amount of TRAP + mononuclear preosteoclasts and multinucleated osteoclasts, whereas USC-Exos treatment inhibited osteoclast formation and caused the accumulation of preosteoclasts (Fig. 3D, 3E). As evidenced by ELISA, the concentration of TNF-α and IL-6 in CM from RNAKL + UHMWPE induced cells was much higher than that in the un-induced group, and USC-Exos treatment resulted in a further decrease of TNF-α and IL-6 production (Fig. 3F). Our results suggest that USC-Exos inhibit osteoclast formation by decreasing the concentrations of in ammatory cytokines.

USC-Exos enhance osteogenic activities
To evaluate the effects of USC-Exos on osteoblasts in vitro, BMSCs were cultured in osteogenesis induction medium supplemented with USC-Exos or PBS. The concentration of ALP and Ca 2+ were analysed to determine the osteoblastic differentiation of BMSCs after 3 days of induction (Fig. 3G). As illustrated in Fig. 3H, the USC-Exos enhanced matrix mineralization and the concentration of ALP, Ca 2+ were signi cantly higher than the vehicle group. Alizarin Red S (ARS) staining revealed that USC-Exos markedly enhanced the calcium nodule formation of MSCs (Fig. 3H). The results of the concentration ALP, Ca 2+ and ARS staining showed that USC-Exos promoted osteoblastic differentiation of BMSCs.

Biodistribution of USC-Exos in murine calvarial osteolysis model
To investigate the local biodistribution of USC-Exos, we rst employed the mouse model of calvarial bone resorption, and then stained exosomes with DiR and injected them into the center of calvariae. The mice were imaged using the IVIS at indicated time and we observed that the uorescence intensity of DiRlabled USC-Exos gradually decrease within 2 weeks, but the uorescence could still be detected locally in the calvariae at the end point (Fig. 4A, 4B). Then the major organs including the calvariae were taken out and observed. We found that DiR-labled USC-Exos cannot be detected in the lungs, hearts,livers, spleens, and kidneys of the treated animals, but can be detected in the calvariae (Fig. 4C). These data indicate that the local injection of USC-Exos could sustain about 2 weeks.

USC-Exos prevents UHMWPE-induced osteoclastic bone resorption
As seen in Fig. 5A, the gross pathology veri ed that UHMWPE particle developed a pronounced in ammatory response in the calvariae. To explore the biological effect of USC-Exos on pathological osteolysis, we examined how USC-Exos affected UHMWPE-induced osteolysis in a murine calvarial model. By performing micro-CT scans and 3D reconstruction, we determined that extensive bone resorption occurred in the UHMWPE group, which was observed as extensive surface erosions on the calvaria, when compared with the negative control (Sham; PBS injection) ( Figure. 5B). However, treatment with USC-Exos suppressed UHMWPE induced osteolysis (Fig. 5B). Quanti cation of bone parameters con rmed that USC-Exos signi cantly increased the BV/TV (Fig. 5C) and decreased the number of pores and percentage of porosity (%) (Fig. 5D, 5E).
Histological assessment and histomorphometric analysis further con rmed that USC-Exos treatment protected against UHMWPE-induced bone loss. The H&E staining showed that sections in the shamgroup exhibited few osteolytic changes. In the vehicle group, osteolysis had clearly occurred, whereas the USC-Exos-treated groups exhibited reduced osteolysis (Fig. 6A), and the area of eroded surface and thickness of periosteum were lower than in the UHMWPE-induced group, respectively (Fig. 6B, 6C). Furthermore, in accord with the micro-CT quantitation, histomorphometric analysis demonstrated that USC-Exos reduced the erosion surface. TRAP staining revealed that the number of multinucleated osteoclasts in the injection site was increased in response to the presence of UHMWPE particles, which was indicated by the presence of osteoclasts that lined the eroded bone surface. However, in USC-Exostreatment groups, the number of osteoclasts and the percentage of osteoclast surface relative to bone surface decreased (Fig. 6D, 6E, 6F), which indicates that USC-Exos treatment inhibited osteoclast formation during UHMWPE induced osteolysis in vivo. The results of immunohistochemical staining are shown in Fig. 6D. We found that the osteogenic markers OCN were highly expressed in the USC-Exos treatment group (Fig. 6G). In addition, decreased amounts of TNF-α expression were quanti ed in the USC-Exos group (Fig. 6H).

Safety of USC-Exos
Major concerns about the application of exosomes from human urine sample are immunogenicity. The safety of the USC-Exos was evaluated by testing the function of red blood cells and performing a histological examination of the major organs (Fig. 7A). The proportions of white blood cells and platelets showed no signi cant differences after USC-Exos treatment (Fig. 7B). Moreover, the proportions of differential white blood cells (including neutrophils, lymphocytes, monocytes, eosinophils, basophils) were calculated and showed no signi cant differences. As shown in Fig. 7C, the concentration of hemoglobin in the blood (HGB), the mean corpuscular hemoglobin (MCHC) level and the hemoglobin distribution width (HDW) were quite similar among the groups, indicating that the USC-Exos did not in uence the function of red blood cells. H&E staining of the brain, heart, lung, liver, spleen and kidneys further revealed no obvious pathological changes among the groups. (Fig. 7D).

Discussion
Total joint arthroplasty (TJA) is commonly used to treat the most severe joint diseases. However, periprosthetic osteolysis and aseptic loosening following TJA is a complex problem that requires meticulous evaluation and preoperative planning. Wear particles and the biological effects of wear debris have a major effect on biological parameters in the periprosthetic environment, which are currently among the most serious problems in patients undergoing arthroplasty surgery. PE particles produced by an arti cial joint have been increasingly recognized as having the potential to induce in ammation and osteolysis [36,37,38], which would directly result in failure of arti cial joint replacement. Revision TJA is related to surgical and implant demands as bone defects need to be repaired and joint stability has to be restored. In addition, revision surgery is associated with increased patient morbidity, and great economic burden for the health care system [39,40]. In this study, we found that USC-Exos could effectively promoted osteoblast formation and inhibit osteoclast formation in vitro. We also noted that USC-Exos induced lower in ammatory cytokines compared to the control. In vivo, we demonstrated that local injection of USC-Exos could effectively alleviate bone loss in mice, as de ned by increased trabecular and cortical bone mass, enhanced osteogenic activities and reduced osteoclast formation. Our ndings suggest a promising future prospect for autologous USC-Exos to be used as a new treatment for periprosthetic osteolysis patients. That is, to obtain therapeutic exosomes for osteolytic treatment, aseptic loosening patients may just need to collect a certain volume of their own urine to harvest USCs.
Over the past years, the application of pluripotent stem cells for bone tissue engineering has received much attention among researchers and clinicians [41,42,43]. USCs is one of the most promising stem cell resource and have therapeutic potential for many bone diseases [24,42]. Recent studies have demonstrated that transplantation of stem cells contributes to tissue repair and regeneration not by the direct differentiation into the parenchymal cells, but rather by the paracrine effects to stimulate endogenous cells participating in tissue regeneration [11,12,44,45]. Exosomes are crucial paracrine mediators obtained from most cell types in culture and biological uids including plasma, milk, amniotic uid, pleural effusions. All extracellular vesicles bear surface molecules that allow them to be targeted to the recipient cells. Once attached to target cell, extracellular vesicles can induce intracellular signaling via receptor-ligand interaction or can be internalized by endocytosis and/or phagocytosis or even fuse with the target cell's membrane to deliver their contents into its cytoplasm, thereby modifying the physiological state of the recipient cell [13].  [42]. These results provided evidences that exosomes could be used to stimulate bone repair in bone diseases. A large number of studies have reported the function of urine-derived stem cells on tissue regeneration [17], whereas few studies have directly utilized USCs to isolate extracellular vesicles for therapeutic uses. Herein, we harvested exosomes from human USCs and veri ed that USC-Exos could ameliorate bone loss of mouse osteolysis model, indicating that USC-Exos might be utilized as a new non-operative therapeutic interventions for the treatment of aseptic loosening.
The in ammatory reaction, bone resorption, and formation of foreign-matter granulomas around the bone-prosthesis site is the principal pathophysiological mechanism of aseptic loosening [10], while monocytes/macrophages represent the critical cell type closely coupled with this pathophysiological process. RAW 264.7 cell is also frequently employed in in vitro models for evaluating cellular phenomena related to exposure to wear debris [9]. RAW 264.7 is a mouse monocyte/macrophages cell line that represents an ideal osteoclasts precursor, as its expression pro le is very close to the osteoclasts, and also because these cells are able, once properly stimulated by RANKL cytokines only, to resorb bone [48]. BMSCs are the major source of osteoblasts that contribute to bone formation [49]. In this study, we demonstrated that USC-Exos could be internalized by RAW264.7 and remarkably inhibited the osteoclast differentiation of RAW264.7 as evidenced by decreased number of osteoclasts, while accelerated osteoblastic differentiation of BMSCs as de ned by ARS staining and increased ALP activity and matrix mineralization. In addition, an adequate blood supply is important for bone formation and maintenance [50] and Chen at al. previously con rmed that USC-Exos are able to transfer proangiogenic proteins to endothelial cells and thereby promote angiogenesis [17]. Therefore, USC-Exos may be a feasible and effective therapeutic candidate to treat or prevent wear debris-associated osteolysis and aseptic loosening.
At present, major concerns regarding the application of USC-Exos are immunogenicity and safety.
Nevertheless, if we can test the immunogenicity of USC-Exos, then it will be useful to evaluate their safety when the use of allogeneic or heterogeneic USC-Exos is required. The safety of the USC-Exos was evaluated by testing the function of red blood cells and performing a histological examination of the major organs. In our study, we found that human USC-Exos did not evoke obvious immune responses in recipient mice between USC-Exos-treated mice and vehicle-treated control mice. However, we did not test the levels of antibodies against USC-Exos in these mice after the long-term use of these heterogeneic USC-Exos. Future studies are required to comprehensively and systematically assess the immunogenicity of USC-Exos.

Conclusions
In conclusion, our results demonstrate that USC-Exos are able to ameliorate bone loss in osteolysis mice. The potential mechanism is the inhibition of in ammatory cytokines and the regulation of function properties of osteoblasts and osteoclasts, as USC-Exos can promote osteogenic differentiation of BMSCs and inhibit osteoclastic differentiation of osteoclast precursor cells in vitro. Our data suggest that USC-Exos may represent a novel therapeutic tool for orthopaedic wear-debris associated osteolysis.

Availability of data and materials
The data generated or analyzed during this study are included in this article, or if absent are available from the corresponding author upon reasonable request.

Ethics approval and consent to participate
This study was approved by the Department of laboratory Animal Management Committee of Central South University (No. 2020sydw0972) and was conducted according to all current ethics guidelines.

Consent for publication
Not applicable.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.