Clinical samples and cell culture
The NPC biopsy specimens and serum samples (n = 110) were collected from Nanfang Hospital between 2007 and 2015. All samples were pathologically reassessed. The clinical characteristics of the study participants were presented in Additional file 1: Table S1. None of the patients had received radiotherapy or chemotherapy before biopsy and blood sampling. The clinical classification for NPC was based on the seventh edition UICC/AJCC staging system. All patients were treated with a uniform protocol of image-guided intensity modulated radiotherapy or/and cisplatin-based concurrent chemotherapy following induction chemotherapy according to National Comprehensive Cancer Network (NCCN) Guidelines. Healthy donors were recruited, and normal biopsies of the nasopharynx were obtained from every donor and used as healthy control. Informed consent was obtained from all individuals, and the research protocols were approved by the Ethics Committee of Nanfang Hospital. The approval number is NCT01171235. The median follow-up was 60 months. The 5-year overall survival rate was 71.8%.
Human NPC cell lines (5-8F and CNE1), HUVECs and human embryonic kidney 293 T cells (HEK293T) were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). An immortalized nasopharyngeal epithelial cell NP69 was kindly provided by cancer center of Sun Yat-sen University. NPC cell lines were cultured in RPMI1640 supplemented with 10% heat-inactivated exosome depleted fetal bovine serum (Gibco) at 37 °C in a humidified atmosphere containing 5% CO2. HUVEC cells were cultured in endothelial basal medium (EBM-2) supplemented with 10% fetal bovine serum. NP69 was cultured in Keratinocyte-SFM (Invitrogen) supplemented with bovine pituitary extract (BD Biosciences). HEK293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen). The PDK1 inhibitor OSU-03012 was purchased from Selleckchem (Houston, TX).
Exosomes isolation, transmission electron microscopy and nanoparticle tracking analysis
Exosomes were isolated from the plasma of healthy controls or NPC patients and from cell culture medium by differential centrifugation as previously described [13]. Briefly, collected culture supernatants or plasma samples were differentially centrifuged at 300×g, 1200×g, and 10,000×g for 3 h, and the supernatant was filtered and ultracentrifuged at 110,000×g for 3 h (all steps were performed at 4 °C). Exosomes were collected from the pellet and resuspended in phosphate-buffered saline (PBS).
Exosomes were prepared and fixed as described previously [14]. The samples were observed with a transmission electron microscope (H-7650, HITACHI, Japan) at an acceleration voltage of 80 kV.
Measurements for nanoparticle tracking analysis (NTA) were performed using the Nanosight NS3000 system (Nanosight, Amesbury, United Kingdom) equipped with a blue laser (405 nm). Exosomes were illuminated by the laser, and their movement under Brownian motion was recorded in 9 s sample videos which were analyzed with NTA analytical software (version 2.3, Nanosight). The capture settings and analysis settings were done manually according to the manufacturer’s instructions.
Cy3-labeled pre-miR miRNA trasnfection, exosomes labeling and confocal microscopy
Pre-miR miRNA precursor (hsa-miR-9, Ambion, Austin, TX) was labeled with Label IT siRNA Tracker Cy3 Kit, according to the manufacturer’s instructions (Mirus, Madison, WI, USA). 5-8F cells were transfected with 100 nM of Cy3-labeled pre-miR miRNA precursor using HiPerFect (Qiagen, Dusseldorf, Germany). After incubation for a day, the culture medium was collected and used for exosome preparation. The exosomes including Cy3-miR-9 were labeled using the green lipophilic fluorescent dye PKH67 (Sigma-Aldrich, St. Louis, MO, USA). The PKH67-labeled exosomes including Cy3-miR-9 were incubated with cultured HUVEC cells. After incubation, HUVEC cells were washed and visualized under confocal microscopy (FV10i, Olympus, Japan). 4′6-diamidino-phenylidole (DAPI, Abbott, USA) was used for nuclear staining.
Migration assay and tube formation assay
The migration ability of HUVEC cells was tested with BD Transwell invasion assay inserts, according to the manufacturer’s protocol. HUVEC cells were placed in the upper chamber of a Transwell coated with Matrigel. The chamber was placed in a 24-well culture dish. After incubation, the cells that did not invade through the pores were removed by a cotton swab. Cells on the lower surface of the membrane were fixed with 4% paraformaldehyde, stained with 0.5% crystal violet and counted.
The formation of capillary-like structures was analyzed as described previously [14]. HUVEC cells (2 × 104 cells/well) were plated on top of Matrigel (300 μL/well) and treated with exosomes (600 μL of exosome fraction/well) derived from 5-8F/con or 5-8F/miR-9 cells. The total tube area was quantified as mean relative tube length obtained from image analysis of five random microscopic fields using Image J software (http://rsb.info.nih.gov/nih-image).
Matrigel plug assay
Matrigel plug assays were performed as described elsewhere [15]. Briefly, 4 × 106 HUVEC cells were incubated with exosomes derived from 5-8F/miR-9 cells cells or 5-8F/con cells. The cell suspensions were mixed with 500 μl of Matrigel-reduced growth factor (BD Matrigel 356,230) at a ratio of 1:1. Then the cell suspensions were injected subcutaneously in the dorsal region of nude mice (female, 7–8 weeks old, BALB/c). Plugs were recovered 2 weeks later and then embedded in OCT, cryostat sectioned, and stained by hematoxylin and eosin. The vessel area was quantified as mean relative tube length obtained from image analysis of six random microscopic fields using Image J software. For all the experiments, animal handling and experimental procedures were approved by the Animal Experimental Ethics Committee of Southern Medical University.
Dual luciferase reporter assay
A 300-bp fragment of MDK 3′ untranslated regions (UTR) including wild type (wt) or mutant (mt) miR-9 binding sites were cloned downstream of the firefly luciferase gene in pGL3 vector (Promega). For reporter assays, HUVEC cells were firstly co-transfected with wt or mt MDK 3′ UTR vector and the control vector pRL-CMV [(cytomegalovirus) coding for Renilla luciferase, Promega) using lipofectamine 2000 reagent (Invitrogen). Then the cells were incubated with exosomes extracted from miR-9-overexpressing NPC cells or control cells. Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA).
Western blot
The cells were lysed in lysis buffer (Roche, Penzberg, Germany), and equal amounts of protein were separated by SDS-PAGE. The exosome pellets isolated from the same amount of culture medium (10 mL) were lysed in 200 μL of lysis buffer (Roche), and the same amounts of lysate were loaded in each lane of the gels. Immunoblots were probed with antibodies directed against MDK (rabbit monoclonal anti-Midkine, Abcam, Cambridge, MA, USA), P70S6K P-Ser424 (rabbit polyclonal anti-S6 K1, Abcam), or PDK1 P-Ser241 (rabbit polyclonal anti-PDK1, Abcam). CD63 (rabbit monoclonal anti-CD63, Abcam) and TSG101 (rabbit monoclonal anti-TSG101, Abcam) were used as exosomal markers.
The human and mouse AKT Pathway Phosphorylation Array and MAPK Pathway Phosphorylation Array were performed according to the manufacture’s instructions (RayBiotech, Norcross, GA, USA).
Immunohistochemistry (IHC) and in situ hybridization (ISH)
For quantification of microvessel density (MVD) in NPC tumor samples, the number of blood vessels staining positive for CD31 (1:100 dilution, Abcam, Cambridge, MA, USA) was recorded in ten random fields at 200 magnification. The expression of MDK in NPC tumor samples was examined with IHC as previously described [16]. The antibody was purchased from Abcam (rabbit monoclonal anti-Midkine, Abcam). In situ detection of miR-9 on formalin-fixed paraffin-embedded samples was essentially performed with a miRCURY LNA™ miR-9 detection probe, using an ISH optimization kit (Exiqon, Vedbaek, Denmark) [10]. A scramble-miR probe (Exiqon) was performed as negative control.
Plasmids, transfection and lentivirus transduction
The lentiviral vector for delivery of pre-miR-9 was described previously [10]. The vector for cDNA delivery of MDK was purchased from GeneCopoeia. The production, purification, and titration of lentivirus were performed as described by Tiscornia et al. [17]. Transient transfection of miR-9 mimic or inhibitor (Ambion, Thermo Fisher Scientific) was performed using Lipofectamine 2000 reagent (Invitrogen), and transient transfection of miR-9 mimic or control into exosomes was performed using Exo-fect exosome transfection kit (System Biosciences). To inhibit MDK expression, siRNA against MDK (Invitrogen) was used. Transient transfection of siMDK was performed using Lipofectamine 2000.
RNA isolation, reverse transcription and quantitative real-time PCR
Isolation of cellular and exosomal miRNAs was performed using TRIzol reagent (Invitrogen) and Total Exosome RNA and Protein Isolation Kit (Invitrogen) respectively as described previously [14]. MiR-9 expression in both cells and exosomes was performed using a TaqMan microRNA reverse transcription kit and TaqMan microRNA assay kit (Applied Biosystems, Foster City, CA) according to the manufacturer’s instructions. To measure the mRNA levels of MDK and pri-miR-9, total RNA was reversely transcribed using ImProm-II Reverse Transcription System (Promega). Quantitative real-time PCR (qPCR) was performed using SYBR Green PCR master mix (Applied Biosystems). All samples were normalized to internal controls and fold changes were calculated through relative quantification [18].
Statistical analysis
All statistical analyses were conducted using SPSS 19.0 software (SPSS, Chicago, IL, USA). Data were presented as Mean ± SD of at least three independent experiments. Two-tailed Student’s t test was used for comparisons of two independent groups. Comparisons of multiple independent groups were analyzed using One-way ANOVA followed by a Student-Newman-Keuls test. Survival curves were plotted by the Kaplan-Meier method and compared by the log-rank test. The relationships between CD31 expression and miR-9 or MDK expression were explored by Spearman’s correlation. A two-sided P < 0.05 was considered statistically significant.