MiR-150-5p-modied Bone Marrow Mesenchymal Stem Cells Derived Exosomes Ameliorate Osteonecrosis of Femoral Head by Promoting Endothelial Cell Angiogenesis

Background: Osteonecrosis of femoral head (ONFH) is a common ischemic disease that induces femoral head necrosis. The role of exosomes and miRNA in ONFH has been elucidated, however, whether miRNA-modied exosomes improve the therapy of ONFH is not clear. Methods: We screened ONFH-related miRNAs by RNA sequencing in plasma exosomes of ONFH patients and healthy donors. The key miRNA was overexpressed in bone marrow mesenchymal stem cells (BMSC) exosomes. The regulatory functions of miRNA-modied BMSC exosomes in vascular endothelial cells were illustrated through angiogenesis assay and scratch assay. Results: We identied 9 differently expressed miRNAs (DEmiRNAs) in plasma exosomes between ONFH and healthy groups, with 6 up-regulated and 3 down-regulated miRNAs. Function and pathway analysis revealed DEmiRNAs were primarily involved in angiogenesis, cell migration, focal adhesion. Moreover, miR-150-5p was declined in ONFH exosomes and regulated multiple angiogenesis-related pathways. The miR-150-5p-overexpressed BMSC exosomes were successfully obtained and transported miR-150-5p to endothelial cells. Moreover, the miR-150-5p-modied BMSC exosomes promoted the angiogenesis and migration of endothelial cells. Conclusion: Our results elucidate the exosomal miRNA expression proles in ONFH, and miR-150-5p-modied BMSC exosomes protect against ONFH by promoting angiogenesis, suggesting a new molecular knowledge for the clinical application of ONFH. via the transfer of specic miRNA. Our study aimed to identify exosomal miRNAs benecial for ONFH, and construct the miRNA-overexpressed BMSC exosomes to verify its therapeutic ecacy in ONFH. We identied differentially expressed miRNAs in exosomes from ONFH patients and healthy donors through RNA-seq. The functions of key miRNA-modied BMSC exosomes were illustrated by matrix-gel based in vitro angiogenesis assay and scratch assay. We look forward to exploring a novel molecular for the study basis and clinical application of ONFH.


Introduction
Osteonecrosis of femoral head (ONFH), also known as avascular necrosis of the femoral head, is a complicated skeleton disorder that shows a high disability rate [1]. ONFH is characterized by a femoral head cells apoptosis and accompanied by diminished function of the hip joint [2]. ONFH is a progressive disease which occurs primarily in young adults and may lead to the deformation or even collapse of the femoral head [3]. Various macroscopic risks and causative factors are closely linked to the occurrence of ONFH, including trauma disrupting, blood dyscrasias, corticosteroids usage, hyperlipidemia, excessive alcoholism, and other miscellaneous [4]. Currently, treatments for ONFH is based on condition develops through different stages to perform operatively and non-operatively. Conservative therapy includes pharmacologic agents, biophysical treatments, with femoral head replacement procedures (FHRP) and femoral head sparing procedures (FHSP) for advanced deterioration [1,5]. However, ONFH is still a major problem for no complete cure possibilities. ONFH originates from the cellular level and carries a poor prognosis, a precise and effective therapy in clinical is needed.
Bone mesenchymal stem cells (BMSCs), nonhematopoietic pluripotent stromal cells, which can differentiate into a variety of cell types as well as can promote cell regeneration [6,7]. Previous studies have reported that the failure of bone repair and reconstruction during osteonecrosis is mainly caused by abnormal osteogenesis and adipogenesis of BMSCs [8]. At present, in the treatment of femoral head necrosis, mesenchymal stem cells (MSCs) have received more applications. The autologous implantation or ex vivo expanded of BMSCs has emerged to effectively delay or avoid early-stage ONFH deterioration [9]. Betaine, which attenuates the ethanol-induced inhibition of mineralization of hBMSCs and osteogenesis, is potential pharmacotherapy for ONFH with alcohol induction [10]. Inherently, the therapeutic ability of BMSCs depends to a great extent on their secreted regulatory carriers, such as exosomes. Exosomes are a type of small bilayer membrane-bound nanovesicles (30 -150 nm) that deliver proteins, lipids, and nucleic acids (miRNA, tsRNA, etc) between cells [11], and play essential roles in many biological processes. Exosomes derived from BMSCs can modulate cell survival and tissue repair [12]. However, the effect of MSCs-derived exosomes on ONFH remains unclear.
MicroRNAs (miRNAs) are endogenous small RNAs of less than 22 base pairs, which participate in the post-translational regulation of gene expression [13]. MiR-15b improves ONFH through inhibits the BMSCs-based osteogenic differentiation of via targets Smad7 [14]; MiR-144-3p inhibits the BMSCs cell activity and osteogenic differentiation by targeting FZD4, indicating that MiR-144-3p may mediate ONFH progression and may serve as a new target [15]. Previous ndings showed that miRNAs involve in the mediation of various biological processes via exosomes. The decline of microRNA-224-3p in BMSC exosomes resulted in the upregulation of FIP200 and thereby potentiates angiogenesis and endothelial cell proliferation, invasion, and migration in traumatic ONFH [16]. Liao et al. found miR-122-5p overexpressed exosomes impaired the development of ONFH by inhibiting SPRY2 expression via the RTK/Ras/MAPK signaling pathway [17]. These ndings testify that MSCs-derived exosomes can be conducive to the mend of ONFH via the transfer of speci c miRNA.
Our study aimed to identify exosomal miRNAs bene cial for ONFH, and construct the miRNAoverexpressed BMSC exosomes to verify its therapeutic e cacy in ONFH. We identi ed differentially expressed miRNAs in exosomes from ONFH patients and healthy donors through RNA-seq. The functions of key miRNA-modi ed BMSC exosomes were illustrated by matrix-gel based in vitro angiogenesis assay and scratch assay. We look forward to exploring a novel molecular for the study basis and clinical application of ONFH.

Human plasma specimens
In the current study, three paired plasma samples (ONFH patients and healthy individuals) were used for sequencing, an itional nine paired plasma samples were used for PCR validation. The blood specimens were placed in BD vacutainer K2 EDTA blood collection tubes and mixed lightly. Next, the anticoagulanttreated blood samples were performed to centrifuge separates at 1000 rpm for 15 min at 20 °C to removes the blood cell debris and impurities. The upper layer containing plasma was then obtained from each tube and transferred into sterile EP tubes by the pipette. Taking care not to touch the bottom sediment during this step. At last, each plasma was labeled and stored at -80 °C for subsequent experiments. None of the ONFH patients received any therapy before blood collection. Each participant was signed written informed consent before the experiment. This study was performed with the approval of the Human Ethics Committee of the First A liated Hospital of Fujian Medical University Isolation and identi cation of exosomes The ExoQuick™ Plasma Prep and Exosome Precipitation Kit (SBI, EXOQ5TMA-1, Japan) were used to isolate exosomes from plasma. Brie y, plasma samples were incubated for 1 h at -20 ℃ and 4 ℃, respectively. After ultra-centrifuged for 15 min (13,000 rpm), partial cells and their debris were removed from plasma samples. Add 5 µL SBI Thrombin Reagent to the supernatant and mixed. Then centrifuge isolation was conducted at 10,000 rpm for 5 min at 4 °C to help dissolve the brin. After refrigerated 30 min at 4°C, discarded the supernatant, the exosomes pellet were resuspended with 1X PBS (20 μg exosomes per 1 mL PBS) and stored at − 80 °C. The exosomes were negatively stained with the 3% (w/v) sodium phosphotungstate solution, and a photograph was captured using an LVEM5 TEM (Delong America, Montreal, QC, Canada). NTA (NanoSight; Malvern Panalytical, Worcestershire, UK) was carried out to the diameter and size distribution of exosomes.

Small RNA sequencing and bioinformatics analysis
Illumina sequencing was performed for plasma exosomal small RNA sequencing according to the instructions of the Multiplex Small RNA Library Prep Kit (Illumina, USA). The concrete operations were brie y described as follows. Total RNA from plasma exosomes was rstly extracted using the Trizol method (Qiagen, Germany), measured and quanti ed by NanoDrop 2000 (Thermo, USA). For each sample, 3' adaptor was connected to RNA (200 ng), and reverse primer hybridization and 5' adaptor connection were disposed of orderly. Previous products were synthesized into cDNA and then enriched by PCR. The sequencing of screened DNA fragments was performed using the HiSeq platform (Illumina, USA). The original sequence data (ONFH patients and healthy participants) were ltered and mapped to reference genome, the internationally recognized algorithm DESeq2.0 was adopted to select the differentially expressed miRNAs (DEmiRNAs) with the threshold of p-value < 0.05 and |Log Fold Change| > 1. Subsequently, the hierarchical clustering of DEmiRNAs was analyzed by using MEV software and plotted the heatmap. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontology terms (GO) annotation analysis were used to infer the functional roles of DEmiRNAs. Additionally, RNAHybrid and Miranda were used for the target gene prediction of miRNAs and the miRNA-mRNAs interaction network analysis was constructed using Cytoscape software 3.6.1 (https://cytoscape.org/).

Isolation and culture of cells
Bone marrow was respectively obtained from SD rat using a sterile syringe with heparin, followed by quick mixed with heparin. Next, the fat cells were removed from the BMSCs with a centrifuge separates (1000 rpm for 20 min), and then rinsed the sediment three times with Dulbecco s Modi ed Eagle Medium (DMEM), and cultured the BMSCs in RPMI-1640 media (GIBCO) with additional 10% fetal bovine serum (FBS). Human umbilical vein endothelial cells (hUVEC) (Procell Life Science&Technology Co., Ltd) were incubated in modi ed ECM medium (added 10% FBS, P/S, and 1% ECGS). All cells were incubated at 37 ℃ with 95% air and 5% CO 2 . After 72 h of incubation, the whole medium was replaced, and the culture medium changed every 2-3 d. When cell con uence reached about 80%, cell subculture was performed.

Cell transfection
BMSCs were prepared in a 6-well plate for miRNA transfection. Brie y, 100 pmol of synthetic miR-150-5p mimics or negative control (NC) RNA were mixed with 250 μL serum-free Opti-MEM (Gibco, Grand Island, NY, USA) and incubated for 5 min at room temperature, respectively. Meanwhile, 5 μL Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA) was diluted with 250 μL serum-free medium Opti-MEM at room temperature for 5 min. The above two reagents were mixed and incubated at room temperature for 20 min and added to the wells containing BMSCs. The 6-well plate was placed in a 37 ℃ incubators for 6 -8 h and refreshed the medium. Subsequent tests were performed after an additional 24 -48 h incubation.

Internalization of exosomes
The exosomes derived from BMSCs were extracted and ultra-centrifuged (120,000 x g) at 4 ℃ for 3 h after PBS dilutions using a Beckman tabletop ultracentrifuge. After washed with 100 ml medium, the exosomes' precipitation was resuspended in 700 ml diluent C. The exosomes were labeled by PKH67 Fluorescent with PKH67 Fluorescent Cell Connection Kit Sigma-Aldrich according to manufacturer protocol. Exosomes solution (250 ml) was mixed with diluted PKH67 dye and incubated for 4 min, and neutralize the excess PKH67 dye with bovine serum albumin (1%, 4.2 ml). The bye-labeled exosomes were centrifuged at 120,000 x g for 3 h (4 ℃) and washed the precipitation with 1 x PBS. The HUVEC cells (1.3 x 10 4 cells /well) were placed on 8-well slides and cultured for 24 h. After 1 x PBS washing, the medium containing PKH67-labeled exosomes was added and incubated for 48 hours. The cells were washed and incubated with a 4% paraformaldehyde solution at room temperature for 10 minutes. After 1 x PBS washing, the nuclear staining was employed by ProLong Gold Antifade Reagent containing DAPI (Thermo Fisher Scienti c, USA). The internalization of exosomes in HUVEC cells was observed with a confocal laser-scanning microscope (LSM510, Carl Zeiss, Germany).

Matrix-gel based in vitro angiogenesis assay
HUVECs were suspended in a serum-free ECM culture medium, seeded in a 24-well plate, and incubated with miR-150-5p-exosomes (miR-Exo), NC-exosomes (NC-Exo), and PBS for 24 h. The angiogenesis analysis was employed on ice, adding a 200 ul cooled Matrigel Matrix (10m /ml; Becton Dickinson, USA) and incubated at 37 °C for 1 h. HUVECs (6 × 10 4 per group) were seeded at coagulated Matrigel and incubated for 18 h at 37 °C. The blood vessels formation of each group was observed with an inverted microscope (NIKON, Japan).

Scratch assay
The cells were seeded in a 6-well plate and formed adherent cells. The monolayer cell was scraped in a straight line to create a "scratch" by a p200 pipette tip when cell density reached 90-100% con uent.
Then the cells incubated with BMSCs-derived exosomes (10 mg/mL). Images of the scratch were acquired at 0 h and 48 h. The migration distance of the scratched area and the node numbers were observed, and the multiple visual elds of cell migrated were randomly selected and photographed and performed a comparison between groups.
Reverse transcription-quantitative polymerase chain reaction Total RNA was extracted as the previous method. After the detection of RNA quality, purity, and content, 1ug RNA was reverse-transcribed into cDNA with a PrimeScriptTM RT reagent kit with gDNA Eraser kit (TaKaRa, Tokyo, Japan). The Reverse transcription-quantitative polymerase chain reaction (qRT-PCR) reaction was performed with the SYBR Green PCR kit (Toyobo, Osaka, Japan) on Applied Biosystems 7300 real-time PCR system (Applied Biosystems, Foster City, CA). The reaction conditions were: 10 min of pre-denaturation at 95 ℃, 45 cycles of denaturation at 95 ℃ for 15 s, and annealing at 60 ℃ for 60 s. The internal references in qRT-PCR were β-actin and U6. Three independent experiments were conducted.
The relative expression of each factor was evaluated using the 2 -ΔΔCt method. The RT primer and special PCR primers were listed in Additional le 1.

Statistical analysis
All of these experiments were repeated three times. Statistical analyses were assessed with the SPSS v.21.0 software (IBM, USA). The comparisons of means among the two groups were evaluated by Student's t-test. One-way ANOVAs were performed for multiple comparisons. For all tests, p < 0.05 was considered to be statistically signi cant.

Characterization of plasma exosomes derived from ONFH patients
To comprehensively characterize the morphologies and structures of exosomes derived from plasma of ONFH patients and healthy donors, we performed TEM and NTA methods. Both TEM and NTA showed that there was no signi cant difference in the plasma exosomes between ONFH patients and healthy donors (Fig 1A and B). TEM image indicated plasma particles were identi ed as two-layer membrane structures, mainly presented with a morphology of round or cup-shaped (Fig 1A). NTA revealed that the mean size of particles was 121.9 and 125.0 nm, with an average concentration of 5.4 × 10 11 and 5.2 × 10 11 particles/mL in ONFH patients and healthy donors, respectively ( Fig 1B). All these results unequivocally con rmed that exosomes were successfully isolated from human plasma.

Function and pathway analysis of exosomal DEmiRNAs
To further investigate the roles of DEmiRNAs, we predicted target genes for these miRNAs, and the function and pathway of the target genes were analyzed by GO and KEGG analysis. Using RNAhybrid and Miranda, we identi ed 1,156 target genes for the DEmiRNAs (Additional le 3). GO enrichment analysis indicated 914 targeted genes were signi cantly enriched in the GO terms. Biological process analysis revealed DEmiRNAs mainly involved in the function of angiogenesis, phosphorylation, transcription regulation, signal transduction, and cell migration. For instance, numerous targeted genes enriched in the GO terms of "neuron migration", "phosphorylation", "regulation of transcription, DNA-templated", "sprouting angiogenesis", "intracellular signal transduction" (Fig 3A, Additional le 4). Cell composition analysis showed DEmiRNAs were mainly associated with neuronal cell body, cytoplasm, terminal bouton, growth cone (Additional le 5). Molecular function analysis showed DEmiRNAs were related to doublestranded DNA binding, protein binding, kinase activity, nucleotide binding (Additional le 6). A total of 365 targeted genes were mapped to terms in KEGG database. We found that targeted genes were primarily involved in the signaling pathway of "Focal adhesion", "ErbB signaling pathway", "mTOR signaling pathway", "Rap1 signaling pathway", and "Ras signaling pathway" (Fig 3B).

Exosomal miR-150-5p is down-regulated in ONFH and involved in angiogenic signals
To identi ed the key exosomal miRNAs related to ONFH, two DEmiRNAs were selected for qRT-PCR validation in 9 ONFH patients and healthy controls, based on their large abundance and high fold changes. In accordant with the RNA-seq results, qRT-PCR results exhibited that miR-150-5p was signi cantly decreased in the ONFH exosomes compared to the healthy control exosomes (Fig 4A). The expression of miR-452-5p was upregulated in the ONFH exosomes, but the difference was not signi cant. Therefore, miR-150-5p was screened for further study. Molecular regulatory (Fig 4B) network demonstrated that miR-150-5p regulates multiple angiogenesis-related pathways by targeting the mRNAs. For instance, miR-150-5p was involved in the vital signaling of angiogenesis, VEGF signaling pathway, by targeting PRKCB and AKT2. Moreover, the target genes of miR-150-5p participate in the TGFbeta signaling pathway, MAPK signaling pathway, HIF-1 signaling pathway, PI3K-Akt signaling pathway, and mTOR signaling pathway, which been proved to regulate angiogenesis [18,21]. The interruption of blood supply to the proximal femur is the main cause of osteonecrosis, indicating that exosomal miR-150-5p might be involved in the treatment of ONFH by regulating angiogenesis.
BMSC exosomes promote endothelial cell angiogenesis Next, we investigated whether exosomes overexpressed miR-150-5p could regulate angiogenesis. BMSC exosomes have been demonstrated to display therapeutic effect on various diseases including osteonecrosis [22], so we chose BMSC exosomes as miR-150-5p transporter. We rst evaluated the action of BMSC exosomes in angiogenesis. As expected, the uorescent tracer technique revealed BMSC exosomes derived from health rats could be internalized by HUVECs (Fig 5A). Moreover, we explored the angiogenesis potential by a matrix-gel based in vitro angiogenesis assay. The results displayed that BMSC exosomes derived from the healthy rats remarkably increased the tube formation of HUVECs, compared to PBS (Fig 5B and C). Collectively, BMSC exosomes were internalized by endothelial cells and promoted angiogenesis. miR-150-5p modi ed BMSC exosomes enhance endothelial cell angiogenesis To obtain the miR-150-5p overexpressed exosomes, we transfected the healthy BMSCs with miR-150-5p mimics and NC RNA. qRT-PCR veri ed that the BMSC exosomes with highly expressed miR-150-5p were successfully obtained (Additional le 7). Interestingly, the uorescent tracer technique showed BMSC exosomes successfully transferred the miR-150-5p to the cytoplasm of HUVECs (Fig 6A). BMSC exosomes with miR-150-5p overexpression dramatically increased the tube forming capacity of HUVECs, compared to the control exosomes ( Fig 6B). Furthermore, miR-150-5p overexpression promoted the enhanced effect of BMSC exosomes on the migration capability of HUVECs (Fig 6C). Taken together, these results indicated that miR-150-5p modi ed BMSC exosomes mediated the ONFH progression via promoted the angiogenesis of HUVECs.

Discussion
ONFH is a progressive pathologic disease characterized by dysfunction of endothelial cells, lipid metabolic disorders, and bone mass loss [23,24]. Our previous study has revealed BMSC-derived exosomes were involved in osteonecrosis by regulating osteogenesis [25], however, whether miRNAmodi ed exosomes can enhance the therapeutic effect remain unknown. Moreover, the expression pro le of plasma exosomes from ONFH has not been reported. In this study, we rst revealed the exosomal miRNA expression pro les of ONFH and identi ed 9 exosomal DEmiRNAs between ONFH and healthy donors. Exosomal miR-150-5p was signi cantly decreased in the ONFH plasma and was predicted to be involved in the angiogenesis-related signaling pathways. Moreover, BMSC exosomes carrying miR-150-5p induced the migration and tube formation of endothelial cells.
Exosomes themselves or exosomes carrying target molecules are being actively investigated as therapeutic agents [26]. Compared with other drug carriers, exosomes can effectively enter other cells and deliver functional molecules with minimal immunological clearance [27]. Especially exosomes derived from stem cells have exerted a therapeutic effect on various diseases, including cancer, cardiovascular disease, graft-versus-host disease, and osteonecrosis [28,30]. Emerging evidence showed stem cellderived exosomes involved in angiogenesis. For instance, human urine-derived stem cell exosomes alleviate ONFH by reversing the GC-induced inhibition of endothelial angiogenesis [31]. HIF-1α modi ed rabbit BMSC exosomes induced neovascularization by promoting the HUVEC cell ability, migration, and tube formation [32]. BMSCs have been shown to transfer miR-29b-3p into brain microvascular endothelial cells and enhance angiogenesis, resulting in amelioration of ischemic brain injury [33]. In accordant with previous studies, our results indicated miR-150-5p modi ed BMSC exosomes promoted the migration and tube formation of HUVECs.
Previous reports revealed that miR-150-5p promotes angiogenesis in extravillous trophoblast cells and synovial broblasts [34,35]. However, the mechanism by which miR-150-5p regulates angiogenesis remains unclear. In this study, we showed that the target genes of miR-150-5p were involved in VEGF, PI3K-Akt, MAPK, and HIF-1 signaling pathway, which can regulate angiogenesis and were related to ONFH. VEGF is a key regulator of angiogenesis, VEGF overexpressed adipose stem cells can provide rapid angiogenesis and osteogenesis in inhospitable avascular environments of ONFH [36]. Desferrioxamine ameliorates glucocorticoid-induced ONFH by inducing angiogenesis via HIF-1α/VEGF pathway [37]. Exosomes secreted by induced pluripotent stem cell-derived mesenchymal stem cells mediate a protective effect in ONFH by inducing local angiogenesis via the activation of the PI3K/Akt signaling pathway [38]. Induction of ONFH caused increased expression of angiogenic responses-related genes, which were implicated in HIF-1, PI3K-Akt, and MAPK signaling pathways [39]. Mesenchymal stem cellsderived exosomal miR-122-5p protects against ONFH progression by triggering the RTK/Ras/ MAPK signaling pathway [17]. Here, our ndings indicated that miR-150-5p-overexpressed BMSC exosomes may ameliorate ONFH by promoting angiogenesis via these signaling pathways.

Conclusions
Our results rst provided the miRNAs expression pro les of ONFH plasma exosomes, and identi ed miR-150-5p was signi cantly decreased in the plasma exosomes of ONFH patients. The molecular regulatory network suggested that miR-150-5p was involved in the angiogenesis-related signaling pathways.
Moreover, the overexpression of miR-150-5p in BMSC exosomes attenuated the angiogenesis of endothelial cells. This study expands the knowledge of BMSC exosomes and provides a new therapeutic target for ONFH.