Human serum- derived exosomes modulate macrophage inflammation to promote VCAM1- mediated angiogenesis and bone regeneration

During exogenous bone-


| INTRODUC TI ON
Exogenous bone grafts such as 3D-printed scaffolds or xenogenic bone grafts (e.g. Bio-Oss) are frequently used for bone defect repair in the clinic. 1,2 Although these bone grafts are designed to exert minimum immunogenicity, the graft recipient's immune system treats these grafts as foreign bodies and exerts inflammatory responses. 3 Macrophages are the key immune cells that actively take part in osteoimmunology during bone graft-mediated bone defect repair. 3 Based on the immunogenicity of graft material and sensitivity of immune systems, macrophages polarize to either pro-inflammatory M1 phenotype or anti-inflammatory M2 phenotype. 4 The majority of exogenous bone grafts induce M1-macrophage polarisation in vivo. 5 Although M1 macrophages are required during a very early stage of bone repair, prolongation of M1 macrophage polarisation during the early, middle and late stages of bone repair creates an inflammatory bone micro-environment that hinders bone defect repair. 3 Inflammatory mediators released from M1 macrophages inhibit migration, proliferation and differentiation of precursor cells including mesenchymal stem cells (MSCs) and endothelial cells during bone defect repair. 6,7 Therefore, regulation of bone-graft-mediated macrophage immunomodulation during bone defect repair is currently at the centre of attention.
Angiogenesis promotes bone regeneration during bone defect repair via osteogenesis-angiogenesis coupling. 8 Type-H vessels characterized by CD31 hi and EMCN hi expressing endothelial cells are mainly responsible for osteogenesis and angiogenesis coupling during bone defect repair. 9 Reports from the literature showed an inhibitory effect of M1 macrophages in angiogenesis during bone defect repair. 10 Therefore, mitigation of M1 macrophages could be beneficial to promote not only osteogenesis but also angiogenesis during bone defect repair. Various anti-inflammatory exogenous growth factors including IL-10 and IL-4 are used to inhibit M1 macrophage polarisation during bone-graft-mediated bone regeneration. 11 However, these exogenous growth factors are relatively expensive and also pose the risk of local and systemic adverse effects. 12 Therefore, novel cost-effective approaches to mitigate M1 macrophages during bone defect repair are highly demanded in clinics.
Exosomes are nano-scale vesicles secreted by cells, with a diameter of about 100 nm. 13 Exosomes encapsulate nucleic acids, proteins, lipids, amino acids and metabolites, and other components, and participate in various physiological and pathological processes. 13 Exosomes from different human cells including MSCs and M2 macrophages had shown inflammation regulation and immunomodulation potential in vitro and in vivo studies. 14,15 However, exosomes from allogenic cells exert immunogenicity, and the isolation of enough amount of exosomes from autologous cells source is time-consuming, highly expensive, and requires a highly sophisticated GMP level laboratory setup. Moreover, the use of exogenous growth factors and cell culture conditions to expand cells in vitro might exert adverse effects in vivo. Human serum contains a large number of exosomes and can be quickly and easily isolated in a simple laboratory setup that provides the opportunity of using autologous serum-Exo to treat various diseases. Therefore, it is wise to investigate the macrophage inflammation mitigation potential of serum-Exo during bone graftmediated bone defect repair.
This study aimed to (i) isolate and characterize serum-Exo from healthy individuals; (ii) investigate the effect of serum-Exo on macrophage inflammation regulation; (iii) analyse the effect of serum-Exo on macrophage inflammation-mediated angiogenesis during bone defect repair. Serum-Exo inhibited LPS-induced macrophage inflammation to promote angiogenesis in Bio-Oss-grafted bone defect repair via upregulation of VCAM1 signalling in endothelial cells, suggesting the possible application of autologous serum-Exo to mitigate exogenous bone graft-induced-immunogenicity-mediated macrophage inflammation. 100 U/mL of penicillin, and 100 μg/mL of streptomycin or in an on HUVEC. Local administration of serum-Exo during mandibular bone defect repair reduced the number of M1 macrophages and promoted angiogenesis and osteogenesis. Collectively, our results demonstrate the macrophage inflammation regulationmediated pro-angiogenic potential of serum-Exo during bone defect repair possibly via upregulation of VCAM1 signalling in HUVEC.

K E Y W O R D S
angiogenesis, bone regeneration, human serum exosomes, macrophage inflammation regulation, VCAM1 endothelial cell special medium (ECM) at 37°C in a humidified 5% CO 2 /95% air atmosphere.

| Isolation and characterisation of serum-Exo
A total of 40 healthy volunteers that is, 20 males and 20 females were recruited to collect the blood sample. Inclusion criteria for the volunteers were as follows: (i) healthy individuals with age 20-30 years, (ii) no history of smoking, (iii) blood platelets count >10 5 /μL cells. Persons with a history of infectious diseases, systemic diseases, and haematological disorders and under medication such as Asprin in the last 3 months were not included in this study. Five milliliters of blood was collected from each volunteer in a vacutainer tube without anticoagulants and placed vertically for 30 min at room temperature. The clotted blood was centrifuged at 1000 rpm for 10 min and serum was collected. Serum samples were centrifuged at 20,000 g for 45 min. The supernatant was transferred to an ultracentrifuge tube (Beckman).
Samples were then ultracentrifuged at 110,000 g at 4°C for 60 min. 16 The obtained serum-Exo was dissolved in phosphate-

| Flow cytometry
After the RAW264.7 cells were cultured under the 6-well plates after stimulation by lipopolysaccharide (LPS; Sigma) for 24 h. The cell suspension was incubated with the blocking antibody CD16 (Abcam) and the anti-mouse CD86 (PE; Abcam) and sorted in a flow cytometry machine as soon as possible.

| Conditioned medium preparation
To harvest the conditioned medium (CM), RAW264.7 were incubated with 100 ng/mL LPS for 24 h to induce macrophage inflammation (M1 macrophages) and then substituted for a fresh culture medium with or without containing 50 μg/mL serum-Exo for 24 h.
Then, the medium was replaced with a serum-free endothelial cell culture medium (ECM). At 6 h, the cell culture supernatant was collected and centrifuged at 1000 rpm for 5 min to remove detached cells and cellular debris. At this time, processed supernatant was mixed with ECM serum-free medium at a ratio of 1:1 and configured into the CM with or without serum-Exo pre-treatment, which was M1 macrophage-CM and serum-Exo-treated M1 macrophage-CM, respectively. CM was stored at −80°C for future experiments.
In the wound healing experiment, HUVEC was inoculated on a 24-well plate to scrape the joined cells with a 200 μL pipette tip after fusion and take pictures. Images were taken under a light microscope after 6 or 12 h of incubation in M1 macrophage-CM or serum Exo-treated M1 macrophage-CM and the wound healing rate was determined.
In the transwell migration assay, a different conditioned medium containing 10% foetal bovine serum (600 μL) is added to the lower chamber of the 24-well transwell insert (8 μm pore size, Corning Costar), and the HUVEC suspension containing 5% foetal bovine serum (200 μL) was seeded into the upper chamber at a density of 1 × 10 5 cells/well. Migrated cells were analysed accordingly.
In the Matrigel tube formation assay, Matrigel and ECM for freeserum in a 1:1 ratio (10 μL/well, BD, USA) were added to precooled ibiTreat plates and polymerized at 37°C for 30 min to form a thin gel layer. HUVEC (1 × 10 5 cells/well) were suspended in a conditioned medium supplemented with 5% FBS and seeded onto ibiTreat plates containing the mixture above. After being incubated for 2.5 h, the capillary-like structures were observed under a light microscope, and pictures of 3-5 visual fields/well were taken. The total length of the tubular structure was analysed by using the Angiogenesis Analyser plug-in for ImageJ software (NIH).

| Aortic ring assay
Thawed matrigel (100 μL) was transferred into the central 10 wells of a 48-well plate on ice. Subsequently, the plate is transferred to 37°C and placed for 30 min to allow it to solidify. The rat aorta was placed on the laid matrigel and then added 100 μL matrigel to the rat aorta, allow to solidify at 37°C for 30 min. M1 macrophage-CM or serum-Exo-treated M1 macrophage-CM (200 μL) were added to each test well. After a further 6-8 days, all plates were fixed with 4% paraformaldehyde, which was washed extensively with water.
Explants and their outgrowths were photographed under an optical microscope, allowing the entire well to be imaged.

| Quantitative real-time PCR
Total cellular RNA was isolated from RAW264.7 cells or HUVEC with SteadyPure Quick RNA Extraction Kit (RightGene) and then reversetranscribed into cDNA with PrimeScript® RT reagent kit (Takara).
The housekeeping gene GAPDH was used for normalisation. The primers used in this study are shown in Table 1.

| Western Blot
The protein sample from cells or serum-Exo (10 μg

| mRNA sequencing
Sequencing was performed by Nanjing Paisennuo Gene Technology Co., Ltd. After quality control of the original data, the high-quality sequencing data were compared with the designated reference genome. The expression values were calculated by the StringTie tool, TA B L E 1 Primers used for RT-qPCR analysis.

| Animal experiment design
Bio-Oss (Geistlich) were used as serum-Exo carriers in our study.
In total, 50 μg of serum-Exo combined with 5 mg of Bio-Oss were

| Statistical analysis
All experiments were repeated three times and all results are expressed as the mean ± SD. Statistical analysis was performed with GraphPad Prism 5.0 (GraphPad Software). The t-test was used for statistical analysis between two groups, and the one-way anova was used for statistical analysis between three or more groups. The test level was α = 0.05. Differences were considered statistically significant at p < 0.05.

| Isolation and characterisation of serum-Exo
Serum-Exo was isolated from human serum by ultrafiltration centrifugation combined with ultracentrifugation. TEM analysis showed that serum-Exo isolated from human serum bore a cup-shaped morphology ( Figure 1A). Size and concentration detection of serum-Exo by NTA showed that the serum-Exo had a narrow size distribution, with a mean particle diameter of 101.49 nm with a range of 50-150 nm ( Figure 1B). Serum-Exo markers CD9, CD63, CD81 and TSG101 were abundantly expressed in serum-Exo ( Figure 1C). The protein concentration of isolated serum-Exo stock was 7.0 μg/μL.

RAW264.7 cells
To study the uptake of isolated serum-Exo, we treated serum-Exo with PKH26, a fluorescent dye with long aliphatic tails that are incorporated into the lipid membrane of serum-Exo. 18 RAW264.7 cells were incubated with PKH26-labelled serum-Exo for 24 h. We observed the presence of PKH26-positive granules in the cytoplasm of RAW264.7 cells by confocal laser microscopy, suggesting that RAW264.7 cells uptake the serum-Exo ( Figure 1D).

| Serum-Exo promoted RAW264.7 proliferation
We selected 50 μg/mL as the optimal anti-inflammatory concentration of serum-Exo by RT-qPCR analysis and used it in subsequent experiments ( Figure S1A). CCK-8 assays revealed that serum-Exo significantly promoted the proliferation of macrophages cultured for day 1, 2 and 3 ( Figure 2A).

| Serum-Exo-treated M1 macrophage-CM promoted angiogenic differentiation of HUVEC
Serum-Exo did not directly affect the angiogenic differentiation of HUVEC cells ( Figure S1B). HUVEC cultured with serum-Exo- These results indicate that serum-Exo-modulated macrophage inflammation promotes angiogenesis.  Figure S2). Among these the differentially expressed genes were involved in NF-ĸB, TNF, rheumatoid arthritis, and chemokine signalling pathways as well as many other biological processes related to angiogenesis ( Figure 5A,B). GSEA analysis showed that the highest scores of

| Serum-Exo promoted mandibular bone defect repair
Serum-Exo was carried by Bio-Oss in vivo to examine its therapeutic effects. In total, 10 SD rats with mandibular defects were split into two groups (Bio-Oss and Bio-Oss + serum-Exo; n = 5).
After 6 weeks, the mandibles with defects were harvested and scanned using micro-CT. Figure 6A shows a 2D image of the bone defect area surrounded by blue boxes in each group. Compared to defects implanted with only Bio-Oss, defects implanted with Bio-Oss + serum-Exo healed better. There was a marked increase in new bone formation in the Bio-Oss + serum-Exo-treated group when compared to the Bio-Oss group, particularly in terms of new bone thickness by 2.3-fold ( Figure 6B). Compared to Bio-Oss, the Bio-Oss + serum-Exo-treated group improved BV/TV and BS/TV ratio by 1.2, and 1.4-fold, respectively, according to 3D reconstruction analysis ( Figure 6A). According to Haematoxylin & Eosin staining, Bio-Oss + serum-Exo showed greater bone regeneration and thicker cortical bones than Bio-Oss alone, and the latter group has more infiltrating inflammatory cells ( Figure 6B). In the BioOss+serum-Exo group, Masson staining showed more collagen deposition and blood vessels punctate with collagen ( Figure 6C). Moreover, com- tion. 24 While another study showed that rat-derived serum-Exo inhibits macrophage inflammation. 25 This study found that serum-Exo from healthy donors was efficiently internalized by macrophages.
Moreover Newly formed blood vessels play an important role in bone regeneration via osteogenesis-angiogenesis coupling. 8 Newly formed type-H vessels characterized by CD31 hi and EMCN hi are mainly responsible for osteogenesis-angiogenesis coupling during bone regeneration. 9 However, effective angiogenesis during bone regeneration is still a challenging task in clinical practice. Numerous studies in regenerative medicine have shown that inflammation inhibits angiogenesis during tissue regeneration. 10 The proliferation and migration of HUVEC are key processes required for angiogenesis. 26 We found that serum-Exo-treated M1 macrophage-CM Bone graft-mediated bone regeneration is a complex process that involves the activity of osteogenic and angiogenic precursor cells as well as immune cells such as macrophages. 30,31 MSCs-derived exosomes had shown bone regenerative potential via regulating the survival and activity of these cells. 32,33 In this study, local application of serum-Exo promoted bone graft-based rat mandibular bone defect repair. Enhanced VCAM1 expression and angiogenesis and alleviated macrophage inflammation were observed during bone defect repair in serum-Exo-treated rats. Representative Western blot images (F), and quantification (G) of angiogenic differentiation markers (n = 3). CPD323: VCAM1 inhibitors. Significant difference between the groups, *p < 0.05, **p < 0.01, and ***p < 0.001. serum-Exo-treated M1 macrophage-CM has better angiogenic potential ( Figure S1B). However, the key factor present in the serum-Exo that mitigates macrophage inflammation should be further investigated. Moreover, the mechanism of serum-Exo-treated M1 macrophage-CM-induced VCAM1 in endothelial cells should also be explored.
Despite tremendous progress in deciphering the mysteries of exosomes over the past few decades, challenges in efficient exosome isolation remain unresolved. 34 Although ultracentrifugation has been the 'gold standard' for exosome separation due to its high processing capacity, high levels of protein aggregate, and lipoprotein contamination in exosome samples prepared through this method greatly compromises their quantification and functional analysis. 35 The development of standardized methods for exosome isolation for clinical application is an urgent task. Different storage temperatures and storage times affect the stability, size distribution, and number of particles of exosomes, as well as the cellular uptake and biodistribution of exosomes. 36

ACK N O WLE D G E M ENTS
We would like to thank all the teachers in the laboratory of the Affiliated University for providing us with a pathology experiment platform.

CO N FLI C T O F I NTER E S T S TATEM ENT
The author reports no conflicts of interest in this work.

DATA AVA I L A B I L I T Y S TAT E M E N T
The part of raw data generated or analyzed during this study is included in this published article and its supplementary information files. The raw data presented in tables are available from the corresponding author upon reasonable request.