microRNA‐19b‐3p‐containing extracellular vesicles derived from macrophages promote the development of atherosclerosis by targeting JAZF1

Abstract Atherosclerosis has been regarded as a major contributor to cardiovascular disease. The role of extracellular vesicles (EVs) in the treatment of atherosclerosis has been increasingly reported. In this study, we set out to investigate the effect of macrophages‐derived EVs (M‐EVs) containing miR‐19b‐3p in the progression of atherosclerosis, with the involvement of JAZF1. Following isolation of EVs from macrophages, the M‐EVs were induced with ox‐low density lipoprotein (LDL) (ox‐LDL‐M‐EVs), and co‐cultured with vascular smooth muscle cells (VSMCs). RT‐qPCR and western blot assay were performed to determine the expression of miR‐19b‐3p and JAZF1 in M‐EVs and in VSMCs. Lentiviral infection was used to overexpress or knock down miR‐19b‐3p. EdU staining and scratch test were conducted to examine VSMC proliferation and migration. Dual‐luciferase gene reporter assay was performed to examine the relationship between miR‐19b‐3p and JAZF1. In order to explore the role of ox‐LDL‐M‐EVs carrying miR‐19b‐3p in atherosclerotic lesions in vivo, a mouse model of atherosclerosis was established through high‐fat diet induction. M‐EVs were internalized by VSMCs. VSMC migration and proliferation were promoted by ox‐LDL‐M‐EVs. miR‐19b‐3p displayed upregulation in ox‐LDL‐M‐EVs. miR‐19b‐3p was transferred by M‐EVs into VSMCs, thereby promoting VSMC migration and proliferation. mir‐19b‐3p targeted JAZF1 to decrease its expression in VSMCs. Atherosclerosis lesions were aggravated by ox‐LDL‐M‐EVs carrying miR‐19b‐3p in ApoE−/− mice. Collectively, this study demonstrates that M‐EVs containing miR‐19b‐3p accelerate migration and promotion of VSMCs through targeting JAZF1, which promotes the development of atherosclerosis.


| INTRODUC TI ON
Atherosclerosis is regarded as a chronic disorder of the arterial wall, accounting for the poor quality of life and mortality across the globe. 1 Atherosclerosis is characterized by obstruction of the vascular lumen in the inner layer of blood vessels due to intimate lesions, namely, atheroma or atheromatous plaques, which in severity can abate tunica media. 2 Atherosclerosis is a multifocal inflammatory reaction, with bacterial and viral infections as potential risk factors. 3 Moreover, the aberrant proliferation of vascular smooth muscle cells (VSMCs) in arterial walls has been identified a crucial pathogenic factor for atherosclerosis. 4 In light of the preceding literature, it is imperative to determine the molecular mechanism in the regulation of VSMC proliferation for the treatment of atherosclerosis.
Extracellular vesicles (EVs) are signalling organelles released by various cell types, that are highly conserved in prokaryotes as well as eukaryotes. 5 An existing study demonstrated the ability of macrophages to secrete EVs, including exosomes, microvesicles, as well as apoptotic bodies. 6 Intriguingly, the participation of macrophages has been identified in atherosclerosis pathogenesis. 7 Additionally, EVs have been highlighted as novel therapeutic targets for the treatment of atherosclerosis. 6 Moreover, EVs by comprising of different DNAs, proteins, mRNAs, as well as microRNAs (miRs) play a vital role in intercellular communications in a variety of diseases. 8 Notably, miRs have been identified as a family of small (22 nucleotides) noncoding RNAs which can regulate gene expression in a post-translational manner. 9 Intriguingly, an existing study identified the involvement of miRs in the pathogenesis of atherosclerosis. 10 On the basis of the bioinformatics results obtained in the current study, miR-19b-3p was differentially expressed in macrophages-derived EVs (M-EVs).
As previously reported, miR-19b by inducing dysfunction of endothelial cells can exacerbate the development of atherosclerosis. 11 Moreover, EVs containing miR-19b-3p might be associated with PMrelated cardiovascular disease such as atherosclerosis. 12 Based on the bioinformatics analysis, zinc finger gene 1 (JAZF1) was identified as a target of miR-19b-3p. The transcription factor JAZF1 is regarded as a type of zinc finger protein binding to the nuclear orphan receptor TR4. 13 It has been reported that upregulated JAZF1 could protect ApoE −/− mice against atherosclerosis through inhibition of hepatic cholesterol synthesis in a CREB-dependent manner. 14 On the basis of the aforementioned literature, the current study proposed that M-EVs containing miR-19b-3p may affect the progression of atherosclerosis by regulation of JAZF1.

| Ethical approval
This study was conducted with the approval of the ethics committee of The First Affiliated Hospital of Harbin Medical University. Great efforts were made to minimize the suffering of the experimental animals used in the study.

| Cell treatment
The RAW264.7 macrophage line provided by American Type Culture Collection (ATCC, VA, USA) was cultured in Dulbecco's modified Eagles Medium (DMEM) containing a combination of 10% (v/v) EVsdepleted fetal bovine serum (FBS), 1% (v/v) P-S and 1 mM sodium pyruvate. The mouse arterial VSMC cell line purchased from Beijing Solarbio Science & Technology Co., Ltd. was cultured in DMEM containing 10% (v/v) FBS and 1% (v/v) P-S.
The original culture medium was discarded, after which the cells were detached with 0.25% trypsin solution. Then, culture medium was added to terminate the digestion, and the adherent cells were transferred into suspension using a pipette and then transferred into other bottles containing culture medium for subsequent culture.
Macrophages were used as negative control (NC), or treated with miR-19b-3p mimic or miR-19b-3p KD. Lentiviral overexpression vectors and interference plasmid vectors (Invitrogen) were subsequently constructed. During transfection, the macrophages (5.0 × 10 7 cells/ml) were seeded into a six-well plate with 2 ml per well. Upon attain 50% cell confluency, the cells were incubated with the prepared lentivirus supernatant (concentration higher than 10 7 TU/ml). After 24 h of infection, the solution was renewed with a complete medium. Upon attaining 80% confluency, the complete medium was renewed with serum-free medium for a regimen of incubation for 24 h. Next, the EVs were separated by differential centrifugation. The expression of miR-19b-3p in the isolated EVs was determined by reverse transcription-quantitative polymerase chain reaction (RT-qPCR); the results of which were compared with those of the uninfected macrophages.

| Isolation and purification of EVs
Macrophages in each group were cultured in serum without EVs.
Upon attaining 70%-80% macrophage confluency, the medium was centrifuged for 10 min at 300 g to isolate the supernatant. Next, the supernatant was centrifuged at low speed (2000 g, 10 min) to remove any cell debris. Subsequently, the supernatant was centrifuged at 10,000 g for 30 min, and then ultracentrifuged at 100,000 g for 120 min (Optima L-100XP, Beckman Coulter Inc.). Finally, the outer coating body was washed twice with an abundant amount of phosphate-buffered saline (PBS), followed by centrifugation for 120 min at 100,000 g. The EVs underwent isolation and purification by sucrose density gradient. The isolated EVs were re-suspended in 0.25 M sucrose and loaded with sucrose layers of 2, 1.3, 1.16, 0.8 and 0.5 M in gradient, followed by centrifugation at 100,000 g through ultracentrifugation for 2.5 h. The components with different densities were extracted from the brim of the sample tube and analysed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis separation (SDS-PAGE). The suspension containing EVs was rinsed twice with an abundant volume of PBS, and centrifuged at 100,000 g at 4°C for 120 min. The concentrated EVs were preserved at −80°C for subsequent experimentation.

| Transmission electron microscopy (TEM)
The EVs derived from the macrophages were subjected to co-culture with VSMCs and purification. The EVs were fixed using 5% glutaraldehyde (in the dark) in 0.1 M phosphate buffer at 4°C. The EVs (20-40 μm) separated from the re-suspended droplets were placed on a copper mesh specifically used for electron microscopy (Hitachi H-7650), followed by counter-staining with 20 μl of 2% phosphotungstic acid for 10 min. All samples were analysed under electron microscopy at 100 KV.

| Western blot assay
The surface markers tumour susceptibility gene 101 (TSG101), CD9, CD63 and endoplasmic reticulum protein marker calenxin were identified by western blot assay. The suspension of EVs was concentrated, after which the protein content was determined using BCA Kits (23227, Thermo Fisher Scientific Inc.). SDS-PAGE gel was prepared and protein denaturation and electrophoresis were performed, after which the protein was transferred onto a membrane and the concentration of cell marker protein was determined.
Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody was regarded as an internal reference. The relative content of the target protein was expressed as the gray value of the target protein band to that of the internal reference protein band. The key antibod-

| Nanoparticle tracking analysis (NTA)
A total of 20 μg EVs were dissolved in 1 ml PBS for 1 min with uniform distribution of EVs. Next, the particle size distribution of EVs was directly observed and measured using a NanoSight nanoparticle tracking analyzer (Malvern Instruments Co., Ltd.). where the X-axis was indicative of the distribution of estimated particle size (nm), and the Y-axis was indicative of the relative percentage. The particle size distribution was plotted based on the estimated particle size (nm) on the X-axis and the concentration (particles/ml) on the Y-axis. Additionally, the particle size distribution diagram based on the intensity was plotted in strict accordance with the particle size (nm) on the X-axis and the intensity (a.u.) on the Y-axis.

| Tracing experiment of VSMCsinternalized EVs
VSMCs and internalized EVs were labelled using the membrane dye PKH67 (Invitrogen) and were rinsed and re-suspended in serum-free medium. Next, VSMCs were seeded on a single-layer glass chassis (Cellvis) and co-cultured with the PKH67-labelled vesicles for periods of 0, 1, 6 and 24 h, followed by three rinses with PBS, fixation in 4% paraformaldehyde, staining with DiI and ultimately observation under confocal microscopy (Leica TCS Sp8).

| 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay and 5-ethynyl-2′deoxyuridine (EdU) staining
The effect of M-EVs on cell proliferation was examined by cell counting and EdU staining (Riobio). Briefly, a concentration of 2 × 10 3 cells/ well (4 replicates in each group) were seeded into a 96-well plate and cultured in a medium containing miR-19b-3p inhibitor with or without M-EVs (8 μg/well). M-EVs were used as control. Subsequently, 10 nm EdU was added for incubation for 12 h. The cells were fixed in 4% paraformaldehyde and stained in strict accordance with the provided manufacturer's instructions. Cell proliferation was observed under fluorescence microscopy.

| Scratch test
In the scratch test, 3 × 10 5 cells/well (three replicates in each group) were placed in a 12-well plate and for growth until confluency. The monolayer was scratched with the tip of the pipette and rinsed with serum-free medium to eliminate the isolated cells. Subsequently, the cells were cultured in EVs-deficient medium containing miR-19b-3p inhibitors with or without M-EVs (80 μg/well), where M-EVs were used as control. The VSMCs were photographed at 0, 12 and 24 h after injury. Wound closure was calculated as follows: migration area (%) = (A 0 -A n )/A 0 × 100, where A 0 represents the initial wound area and A n represents the wound area at the time of measurement.

| miR sequencing
The miRNA microarray data of the EVs were retrieved from the GEO database (GSE13 2646 and GSE11 4318). GSE13 2646 included 12 samples (six control groups and six experimental groups), while GSE11 4318 included nine samples (six control groups and three experimental groups). The 'limma' package of R language was used to analyse the differential expression of several genes. DEmiRNAs were obtained with logFC > 1, p < 0.05 as screening threshold. R language 'ggplot2' software package was used to plot the volcano map, and the intersection of DEmiRNAs was processed.
Initially, the total RNA content was isolated from control EVs and ox-LDL-EVs samples using Trizol reagent (Takara). The total RNA content was separated on 15% three boric acid ethylene diamine tetraacetic acid (EDTA) (TBE) polyacrylamide gel (Invitrogen), and a strip corresponding to miR (18-30 nt) was removed. Next, the miR was reversely transcribed into complementary DNA (cDNA) and amplified. The miRs were sequenced using Illuminutesa HiSeq TM 2000.
After that, Novoalign software (v2.07.11) was adopted to process the sequencing results. Next, the paired t-test with double-sided distribution was adopted to identify the difference of each miR in the M-EVs and M-EVs samples, and used Bonferroni method to correct multiple determination. Parts per million (CPM) was used for calculation. Fold changes of each miR (mean CPM of each miR in control EVs/CPM of each miR in ox-LDL-EVs) and p values were calculated.
The p value was used to calculate the error detection rate (FDR) of each miR, and was further used as a functional filter to identify significant miRs with fold change ≥2 or ≤0.05 and FDR <0.05. The MEV software was utilized to analyse the expression data.

| RT-qPCR
Total RNA was extracted from the cell lines and frozen tissue samples using the Trizol reagent (15596-018, Solabio) in strict accordance with the provided instructions. To measure the mRNA expression, PrimeScript™ RT-PCR kits (Takara) were used to synthesize cDNA from the total RNA content. To determine the miR expression, PrimeScript™ miRNA RT-PCR kits (b532451, SANGON) were used for reverse transcription following the provided instructions.
SYBR Premix Ex Taq TM (Takara) was used for real-time RT-qPCR on LightCycler 480 system (Roche Diagnostics GmbH). With β-actin and U6 serving as internal references, the mRNA and miR expression was standardized. The primers for amplification were designed and purchased by General Biotechnology Co., Ltd. The primer sequences are shown in Table S1. The relative quantitative method (2 −△△Ct method) was employed for calculating the relative expression of the target genes.

| Dual-luciferase reporter gene assay
The targeting relationship between miR-19b-3p and JAZF1 was verified by a combination of the biological prediction website and the luciferase reporter method. The binding sites between miR-19b-3p and JAZF1 were analysed, and the fragment sequences containing the action site were identified. The JAZF1 3'-UTR sequence containing predicted that the miR-19b-3p binding site was inserted into the pGL3 basic vector (Promega) for the synthesis of the firefly/Renilla luciferase report vector pGL3-basic-JAZF1-3'-UTR-wild type (WT); the mutation was pGL3-basic-JAZF1-3'-UTR-mutation type (MUT).
Next, the HEK-293 cells were seeded into a 24-well plate and incubated for 24 h to attain 50%-60% cell confluency. Subsequently, Lipofectamine 2000 was used to co-transfect with NC-mimic/miR-19b-3p mimic and pGL3-basic-JAZF1-3'-UTR-WT, or NC-mimic/ miR-19b-3p mimic and pGL3-basic-JAZF1-3' -UTR-MUT into the cells. In addition, all groups were transfected with 10 ng of PRL TK Renilla luciferase. After 24 h of transfection, the cell lysate was collected to measure the luciferase activity. According to the provided manufacturer's instructions, the relative luciferase activity was determined using the dual-luciferase reporter gene assay system (E1910; Promega) and normalized to Renilla luciferase activity.
The feeding period spanned over 12 weeks, with ad libitum access to food and water. During the experiment, several parameters such as the body weight, blood glucose, cholesterol, LDL, high-density lipoprotein (HDL), triglyceride and uric acid in serum were measured.

| Histological staining
After 12 weeks, the intact aorta was carefully separated and divided as (+ +), 50%-75% as (+ + +) and >75% as (+ + + +). In the negative group, the number of positive cells was less than 25%; in the positive group, the number of positive cells was more than 25%.

| Immunofluorescence
Macrophages, VSMCs and EVs grew into monolayers with confluency on a cell culture glass coverslip. The cells were rinsed with cold PBS, fixed with paraformaldehyde, and then permeated in Tris buffer saline. The cells were sealed at 4°C overnight along with the required primary and secondary antibodies. The main antibody used for macrophages was CD68 (1:1000, ab213363, Abcam), for VSMCs was α-smooth muscle actin (α-SMA) (1:200, 55135-1-AP, Proteintech Group) and for EVs was CD9 (1:1000, ab223052, Abcam); the secondary antibody was IgG (1:5000, ab6721). Fluorescence images of cells were documented under an inverted Leica fluorescence microscope (IX71) or Leica SP-8 confocal microscope system. U6 probe and a scrambled sequence were used as positive and negative controls respectively. SPSS 21.0 (SPSS, IBM) was used to analyse the experiment data.

| Statistical analysis
The measurement data, from three independent experiments were expressed as mean ± standard deviation. An unpaired t-test was adopted for comparing data between two groups. One-way analysis of variance (ANOVA) was conducted for comparing data between multiple groups, followed by Tukey's post hoc test. In all statistical references, a value of p < 0.05 was indicative of a statistically significant difference.

| M-EVs are internalized by VSMCs
The RAW264.7 macrophages were cultured with EV-depleted medium, treated with ox-LDL or without ox-LDL (50 μg/ml) for 48 h, after which the supernatant was isolated for separation and pu-  Figure 1E). These results suggest that VSMCs can essentially internalize M-EVs.

| ox-LDL-M-EVs promote migration and proliferation of VSMCs
Next, we determined the effect of ox-LDL-M-EVs on the migration and proliferation of VSMCs. The co-culture system of VSMCs and macrophages is shown in Figure 2A. a proliferation rate of 28% as determined measured by EdU staining ( Figure 2E). These results demonstrate that ox-LDL-M-EVs can stimulate VSMCs migration and proliferation.

| miR-19b-3p is upregulated in ox-LDL-M-EVs
To identify the potential molecular targets for ox-LDL-M-EVsmediated VSMC migration and proliferation, we searched and analysed the miRNA microarray data (GSE13 2646 and GSE11 4318) of EVs was analysed from the GEO database. As shown in Figure 3A  Overall, we identified upregulation of miR-19b-3p in ox-LDL-M-EVs.

| M-EVs promotes VSMC migration and proliferation by transferring miR-19b-3p
To determine the mechanism by which M-EVs deliver miR-19b-3p into VSMCs, we used Cy3 labelling (red fluorescence) to track the transport of miR-19b-3p between cells, and used the Transwell co-culture device to further investigate the transfer mechanism ( Figure 4A). RT-qPCR results demonstrated that the expression pattern of miR-19b-3p was increased in VSMCs co-cultured with ox-

LDL-M or ox-LDL-M-EVs, while its expression pattern was reduced upon treatment with an inhibitor of secretion of EVs, GW4869 in
VSMCs co-cultured with ox-LDL-M ( Figure 4B). The preceding data is indicative for chief transfer of miR-19b-3p in the form of EVs.
To further verify our conjecture, we infected miR-19b-3p mimic or miR-19b-3p-KD into ox-LDL-M-EVs, followed by incubation of VSMCs; after which the alterations in cell migration and proliferation were determined in each group. After treatment with miR-19b-3p mimic, the migration and proliferation of VSMCs were significantly increased ( Figure 4C,D). Moreover, the migration and proliferation of VSMCs

| M-EVs carrying miR-19b-3p promote VSMC migration and proliferation by targeting JAZF1
In order to elucidate the molecular mechanism of miR-19b-3p involved in VSMC migration and proliferation, a combination of TargetScan, miRWalk and miRBase was adopted to predict the target gene, which identified JAZF1 as one of the downstream target genes of miR-19b-3p ( Figure 5A) and the presence of a binding site between JAZF1 and miR-19b-3p ( Figure 5B). In order to verify whether JAZF1 is the target gene of miR-19b-3p, dual-luciferase reporter gene assay was adopted to analyse and determine the binding site between miR-19b-3p and JAZF1. The results showed that miR-19b-3p could directly target 3'UTR of JAZF1 ( Figure 5C). To further determine the regulatory effect of miR-19b-3p on VSMCs, the VSMCs were treated with miR-19b-3p mimic, mimic NC, miR-19b-3p inhibitor or inhibitor NC, followed by comparison of the expression pattern of JAZF1 at the mRNA and protein levels. After overexpression of miR-19b-3p (miR-19b-3p mimic), the expression pattern of JAZF1 was lowered, while it was increased in response to knockdown of miR-19b-3p (miR-19b-3p-kd) was increased ( Figure 5D,E). These results demonstrated that miR-19-3p can inhibit the mRNA and protein levels of JAZF1.
To study the role of JAZF1 in VSMC migration and proliferation, VSMCs ( Figure S4). Meanwhile, we also labelled the aorta in mice  The measurement data were expressed as mean ± standard deviation. Unpaired t-test was used for comparing data between two groups. One-way ANOVA was conducted for comparing data between multiple groups, followed by Tukey's post hoc test

| DISCUSS ION
Atherosclerosis is regarded as a chief contributor to cardiovascular disease and Peripheral artery disease. 15 In the current study, we ex-   Besides, an existing study elicited the ability of an increased plasma expression of miR-19b-3p in the early stage of acute myocardial infarction to serve as potential biomarker for the diagnosis of myocardial infarction. 20 Mechanistically, we found in the current study that miR-19b-3p inhibited JAZF1 to promote VSMC migration and proliferation. It is noteworthy that the regulation relationship between miR-19b-3p and JAZF1 has been rarely reported. In this study, our dual-luciferase gene reporter assay confirmed that miR-19b-3p could

CO N FLI C T O F I NTE R E S T
None.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request from the authors.