The MIAPaCa-2 human pancreatic cancer cell line was obtained from Japanese Collection of Research Bioresources Cell Bank (Tokyo, Japan) and maintained in minimal essential medium (MEM) supplemented with 10% (vol/vol) fetal bovine serum (both Sigma-Aldrich; Merck KGaA), 1% penicillin-streptomycin mix and 1% MEM non-essential amino acid solutions (100X; both Nacalai Tesque, Inc.) in a humidified atmosphere containing 5% CO₂ at 37°C. The doubling time of MIAPaCa-2 cells was 20–23 h (26).

The following antibodies were purchased: Anti-cytochrome c and anti-phosphorylated histone H2AX (γ-H2AX) from Cell Signaling Technology, Inc.; anti-CD63 from BD Biosciences; donkey anti-goat IgG, F(ab′)₂-horseradish peroxidase (HRP), HRP-conjugated mouse IgGκ light chain binding protein (m-IgGκ BP), anti-actin and anti-CD9 from Santa Cruz Biotechnology, Inc.; anti-SOD1 and anti-SOD2 from Merck KGaA; and rabbit anti-sheep IgG-HRP and tetramethyl rhodamine isothiocyanate (TRITC)-conjugated anti-rabbit secondary antibody from Dako (Agilent Technologies, Inc.). The following reagents were purchased: PKH-67, a lipophilic dye, and N-acetyl-L-cysteine (NAC) from Sigma-Aldrich; Merck KGaA; Hoechst 33342 and 2′,7′-dichlorodihydrofluorescein diacetate (C-H₂DCF) from Thermo Fisher Scientific, Inc.; methylene blue from FUJIFILM Wako Pure Chemical Corporation; and DAPI and wheat germ agglutinin (WGA), Alexa Fluor 594 conjugate from Invitrogen (Thermo Fisher Scientific, Inc.).

Exosomes were isolated from media-conditioned cells by ultracentrifugation, as previously described (27). Briefly, MIAPaCa-2 cells were seeded at 1.5×10⁶ cells per T75 cm² flask and irradiated after substituting the media with exosome-depleted 10% FBS cell culture media (Sigma-Aldrich; Merck KGaA). The cell culture media was centrifuged at 2,000 × g for 10 min at 4°C and the supernatant was filtered through a 0.22-µm Minisart^® syringe filter (Sartorius AG). The supernatant was centrifuged at 150,000 × g for 90 min at 4°C. The pellet was washed with PBS and again centrifuged at 150,000 × g for 90 min at 4°C and resuspended in 50 µl PBS. Then, the total amount of protein in the exosomes was measured using a Qubit™ Protein Assay kit (Invitrogen; Thermo Fisher Scientific, Inc.).

The exosomes isolated from non-irradiated (0 Gy-Exo) and 5 Gy irradiated cells (5 Gy-Exo) were evaluated by transmission electron microscopy (TEM). Briefly, 4 µl PBS suspension of isolated exosomes was loaded onto carbon-coated 200-mesh copper grids for 1 min at 25°C. Excessive fluid was removed using filter paper. The adsorbed exosomes were negatively stained with 2% uranyl acetate for 30 sec at 25°C. Finally, the air-dried exosome-containing grids were observed under a TEM microscope (JEM-1400 plus; JEOL, Ltd.) at 120 kV (magnification, ×50,000). Exosome size, concentration and distribution were analyzed by nanoparticle tracking analysis (NTA) software using NanoSight NS300 (Malvern Panalytical). The software was optimized to identify and track each particle on a frame-by-frame basis and Brownian movement was tracked and measured from frame to frame.

Exosomes were labeled with the green fluorescent dye PKH-67 using the PKH67 Green Fluorescent Cell Linker kit for General Cell Membrane Labeling (Sigma-Aldrich; Merck KGaA) as previously described (28). Briefly, 6 µg 0 and 5 Gy-Exo were labeled with 2 µM PKH-67 for 5 min at 25°C. Then, free PKH-67 was removed by centrifugation 14,000 × g for 2 min at 25°C using the VIVACON 500 ultracentrifugation device (100,000 MWCO; Sartorius Stedim Biotech; Sartorius AG). MIAPaCa-2 cells were cultured for 24 h at 37°C, after which the culture media was replaced with that containing the labeled exosomes. Following incubation at 37°C overnight, cells were gently washed twice with PBS and fixed by 4% paraformaldehyde solution (Nacalai Tesque) for 20 min at 25°C. After washing with Hanks' Balanced Salt Solution (Gibco; Thermo Fisher Scientific, Inc.), cells were incubated with 10 µg/ml WGA, Alexa Fluor 594 conjugate (Invitrogen, Thermo Fisher Scientific, Inc.) for 10 min at 25°C. Finally, samples were incubated with Hoechst 33342 (1:2,000; Invitrogen; Thermo Fisher Scientific, Inc.) for 5 min at 37°C.

Images were captured using a confocal microscope (LSM700; Carl Zeiss AG) equipped with an oil immersion objective lens (magnification, ×40). Images were analyzed with ZEN 2012 (Carl Zeiss AG) and processed using ImageJ software ver.1.51 (National Institutes of Health) (29).

Cells were exposed to 5 or 8 Gy 150 kV X-rays delivered at 0.57 Gy/min using an MBR-1505R2 generator (Hitachi Ltd.). The beam was filtered through a 1-mm aluminum board and the accuracy of irradiation was checked, as previously described (30).

Cell survival following irradiation was evaluated by performing a colony forming assay in the presence or absence of 10 or 20 µg exosomes generated following 5 Gy irradiation (5 Gy-Exo). Cells were reseeded into 6-well cell culture plates (Corning, Inc.) at a density of 200–4,000 cells/well and incubated for 7–10 days (7–10 cell cycles) at 37°C. The number of seeded cells was different in each group [control, 100; 5 Gy, 500; 5 Gy + 5 Gy-Exo (5 µg/ml), 1,000 and 5 Gy + 5 Gy-Exo (10 µg/ml), 2,000 cells/ml] depending on the dose of irradiation. At the end of each experiment using non-irradiated exosomes (0 Gy-Exo) or 5 Gy-Exo, the cells were fixed with a solution of 10% methanol and 20% acetic acid for 30 min and stained at 25°C. With methylene blue for 30 min as previously described (31). Colonies (≥50 cells) were counted and the surviving fraction was determined based on the number of colonies per seeded cell.

Intracellular ROS levels were determined using the oxidation-sensitive fluorescent probe dye C-H₂DCF (Invitrogen; Thermo Fisher Scientific, Inc.) as described previously (32). Cells were seeded in 6-well plates (1.5×10⁵ cells/well) overnight at 37°C and treated with 5 Gy radiation in the presence or absence of 10 µg/ml 5 Gy-Exo and 1 mM NAC (Sigma Aldrich; Merck KGaA) for 24 h at 37°C, as previously described (33). After washing twice with FBS-free media (MEM; Sigma Aldrich; Merck KGaA) the cells were stained with 50 µM C-H₂DCF for 1 h at 37°C. The nuclei of cells were then stained with Hoechst 33342 (1:2,000; Invitrogen; Thermo Fisher Scientific, Inc.) for 5 min at 37°C. The fluorescence of C-H₂DCF was visualized using a fluorescence microscope (magnification, ×20) (BZ-9000; Keyence Corporation).

Induction of DNA damage was investigated by detecting γ-H2AX foci using immunocytochemistry, as described previously (34). Cells were seeded on 35-mm dishes and treated with 10 µg/ml 5 Gy-Exo and 1 mM NAC for 24 h and/or 5 Gy irradiation at 25°C. Then, the cells were fixed in 4% paraformaldehyde in PBS for 20 min, permeabilized with 0.1% Triton X-100 in PBS for 5 min and blocked in 5% BSA (Sigma-Aldrich; Merck KGaA) in PBS for 60 min at 25°C. The cells were incubated with rabbit anti γ-H2AX antibody (1:200; Cell Signaling Technology, Inc.) overnight at 4°C. The cells were then incubated with TRITC-conjugated secondary antibody (1:20; Dako; Agilent Technologies, Inc.) for 90 min at 25°C. The nuclei were stained with DAPI (1:300; Invitrogen; Thermo Fisher Scientific, Inc.) for 10 min at 25°C. The stained cells were observed using a fluorescence microscope (magnification, ×40). The number of cells expressing nuclear γ-H2AX foci were then counted manually in 100 cells of each treatment group, as previously described (35).

The expression levels of CD9, CD63 and cytochrome c were analyzed. Briefly, 3 µg exosomes (0 and 5 Gy-Exo) were separated by 8 (CD9), 12 (CD63) or 15% (cytochrome c) SDS-PAGE gels in non-reducing conditions. A total of 30 µg whole cell lysate was separated using RIPA buffer supplemented with protease inhibitor cocktail (Nacalai Tesque) previously described (36) in reducing conditions [boiling for 5 min at 95°C and addition of reducing agent, 5% 2-Mercaptoethanol (FUJIFILM Wako Pure Chemical Corporation) at 25°C] and transferred to a PVDF membrane. Membranes were blocked using 3 (cytochrome c) or 5% non-fat milk for 30 min at 25°C, incubated with anti-CD9 (1:500), anti-CD63 (1:500) or anti-cytochrome c (1:1,000) antibodies overnight at 4°C and washed three times in Tris-buffered saline with 10% Tween-20 (TBS-T). Subsequently, membranes were incubated for 1 h with m-IgGκ BP-HRP (1:4,000) at 25°C.

The expression levels of SOD1 and SOD2 in MIAPaCa-2 cells were analyzed. Briefly, cells were seeded at 1.0×10⁵ cells/well in 6-well plates and incubated overnight at 37°C and subjected to irradiation with 8 Gy or addition of 30 µg 8 Gy-Exo at 25°C. Proteins was collected using RIPA buffer supplemented with protease inhibitor cocktail (Nacalai Tesque) at 24 h after treatment. Quantification of proteins was proceeded using Qubit™ Protein Assay kit (Invitrogen; Thermo Fisher Scientific, Inc.). A total of 15 µg protein/lane was separated by 10% SDS-PAGE and transferred to a PVDF membrane. Membranes were blocked using 5% BSA (Sigma-Aldrich; Merch KGaA) blocking buffer for 40 min at 25°C, incubated with anti-SOD1 and anti-SOD2 (both 1:1,000) antibodies overnight at 4°C and then washed three times in 10% TBS-T. Subsequently, membranes were incubated for 1 h with rabbit anti-sheep IgG-HRP (1:2,000) secondary antibody at 25°C. The secondary antibodies were visualized with ECL™ Prime Western Blotting Detection Reagents (GE Healthcare) using a gel imaging system with preconfigured Image Lab Touch software 5.2.1 (ChemiDoc Touch MP; Bio-Rad Laboratories, Inc.). Subsequently, the PVDF membranes were incubated with anti-actin (1:5,000) antibody overnight at 4°C. After washing with TBS-T three times, membranes were incubated for 1 h with donkey anti-goat IgG, F(ab′)₂-HRP secondary antibody at 25°C, which was visualized as aforementioned. The intensity of each signal was analyzed using ImageJ software ver.1.51 (National Institutes of Health) and the ratios of SOD1, SOD2 and actin levels were calculated.

Following 48 h transfection with miR-6823-5p-mimic, the expression levels of SOD1 were analyzed, as aforementioned. Protein concentrations were determined using a Qubit™ Protein Assay kit.

Total RNA was extracted from exosomes using Toray's 3D-Gene RNA extraction reagent from a liquid sample kit (Toray Industries, Inc.). Comprehensive miRNA expression analysis was performed using a 3D-Gene miRNA Labeling kit and a 3D-Gene Human miRNA Oligo Chip Ver. 21 (Toray Industries, Inc.) according to the manufacturer's protocol to detect 2,565 human miRNA sequences. The expression levels of each miRNA were expressed as the background-subtracted signal intensity of all miRNAs in each microarray. Any signal intensity in both the duplicate spots at >1.5 SD of the background signal intensity was considered a valid measurement. The raw data are available in the Gene Expression Omnibus database (GSE163133).

miRNAs from exosomes isolated from cells irradiated with either 5 or 8 Gy were visualized in the form of a heat map using R software (version 3.5.3; R-project.org) (37) and heatmap.2 from the gplots package (version 3.0.1.1; CRAN.R-project.org/pakage=gplots) (38). The heatmap presents Z-score values for miRNAs with ratios of expression values between control (exosomes from non-irradiated cells) and 5 or 8 Gy-Exo <0.5 or >1.5. In addition, TargetScan (targetscan.org/vert_72/) (39) and miRTarBase (mirtarbase.cuhk.edu.cn/php/index.php) (40) were searched for targets of these miRNAs that result in increased ROS levels. miRNAs data were then proceeded hierarchical clustering with Euclidean distance and complete linkage.

In order to investigate the effect of miRNAs on intracellular ROS levels, mirVana™ miRNA (miR-6823-5p) mimics and negative control (Sigma-Aldrich; Merch KGaA) were used. MIAPaCa-2 cells (2.2×10⁵ per well) were seeded in 24-well plates and transfected with 5 nM miR-mimic or negative control using HiPerFect Transfection Reagent (Qiagen GmbH) for 48 h at 25°C. Cells were harvested at 37°C and the expression levels of mRNA and protein were examined 48 h after transfection at 25°C. ROS levels and DNA damage in transfected cells were assessed as aforementioned. The mimic sequences were as follows: miR-6823-5p, 5′-UCAGGGUUGGUAGGGGUUGCU-3′; siRNA control 1, 5′-GGUUCGUACGUACACUGUUCA-3′; and siRNA control 2, 5′-CGGUACGAUCGCGGCGGGAUAUC-3′.

Total RNA for RT-qPCR was obtained from cell samples using a mirVana™ miRNA Isolation kit (Invitrogen; Thermo Fisher Scientific, Inc.) and reverse transcribed at 25°C for 30 min, 37°C for 2 h, 85°C for 5 min and 4°C for 10 min to cDNA using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems; Thermo Fisher Scientific, Inc.). RT-qPCR was performed using EagleTaq Universal Master Mix (ROX) (Roche Diagnostics) and TaqMan^® Gene Expression Assays (Applied Biosystems; Thermo Fisher Scientific, Inc.) using following TaqMan probe to SOD1 (cat. no. Hs00533490_m1). The thermocycling conditions were as follows: initial denaturation at 50°C for 2 min and 95°C for 10 min, followed by 95°C for 15 sec and 40 cycles at 60°C for 1 min. Expression data of SOD1 was acquired and analyzed by the 2^−ΔΔCq method (41) using a Thermal Cycler Dice^® Real Time System III (Takara Bio, Inc.). All data were normalized to GAPDH (cat. no. Hs02786624_g1).

Data are presented as the mean ± SEM of three independent experimental repeats. Differences between the means were compared using one-way ANOVA, followed by post hoc Tukey's test. P<0.05 was considered to indicate a statistically significant difference.

The purity, quality, and morphology of exosomes were analyzed. According to the result of measurement using Qubit Protein Assay kit, ~1.67 µg/ml exosomes were collected from the culture medium. The isolated exosomes were shaped in the form of closed, round vesicles with a diameter of ~100 nm. The morphological features of 0 and 5 Gy-Exo were similar, as shown by TEM (Fig. 1A). CD9 and CD63 expression was observed in both whole cell lysate and exosomes, while cytochrome c expression was detected only in the whole cell lysate (Fig. 1B). Distribution profiles for 0 and 5 Gy-Exo revealed peaks at 107 and 101 nm, respectively, as assessed using NanoSight (Fig. 1C; Videos S1 and S2). These data indicated that the exosomes were successfully isolated from the culture media supernatant without contamination due to cellular components.

The uptake of 5 Gy-Exo was increased compared with that of 0 Gy-Exo (Fig. 2A). Cells were irradiated and treated in the presence or absence of 5 Gy-Exo (Fig. 2B). Cells irradiated with 5 Gy-Exo showed an increased radiosensitizing effect in an exosome concentration-dependent manner (Fig. 2C; Table SI).

Intracellular ROS levels increased following addition of 0 or 5 Gy-Exo or irradiation with 5 Gy; the highest levels were observed following combined irradiation and exosome treatment. Irradiation (5 Gy) combined with the addition of 0 Gy-Exo resulted in ROS levels similar to those for the irradiation-alone treatment group (Fig. 3). Intracellular ROS levels were significantly decreased following addition of the ROS scavenger NAC and these effects were dose-dependent (Figs. S1 and S2).

Addition of both 0 and 5 Gy-Exo increased the number of γ-H2AX foci/cell, while combined irradiation and 5 Gy-Exo treatment significantly increased the number of γ-H2AX foci further. Irradiation (5 Gy) and treatment with 0 Gy-Exo resulted in numbers of γ-H2AX foci similar to that of the irradiation-alone group (Fig. 4). NAC treatment resulted in fewer γ-H2AX foci/cell, indicating that DNA damage was induced by an increase in ROS levels and these effects were dose-dependent (Figs. S2 and S3).

The differential expression levels of miRNAs in 5 and 8 Gy-Exo compared with 0 Gy-Exo were stratified using a heat map (Fig. 5A). A total of six up- and five downregulated microRNA in both 5 and 8 Gy-Exo were identified (Fig. 5B). TargetScan and miRTarBase were used to identify potential targets of these miRNAs that may be associated with increased ROS levels. miR-6823-5p was identified as a potential candidate for SOD1 inhibition and subsequent analysis confirmed the inhibitory effect of miR-6823-5p (Fig. 6).

Cells were treated with 8 Gy irradiation or addition of 8 Gy-Exo, then the expression levels of antioxidant enzymes including SOD1 and SOD2 were analyzed. SOD1 expression notably decreased following addition of 8 Gy-Exo, but SOD2 expression levels did not change (Fig. 6A). Using TargetScan, the complementary sequence site of SOD1 was found to correspond with miR-6823-5p (Fig. 6B). SOD1 expression levels decreased following transfection with an miR-6823-5p-mimic (Fig. 6C). Additionally, the relative cDNA level of SOD1 in cells transfected with miR-6823-5p-mimic significantly decreased compared with that of non-transfected cells (non-siRNA), although that of cells transfected with siRNA control 1 and 2 did not change significantly. (Fig. 6D). Taken together, these results suggest that miR-6823-5p in exosomes derived from irradiated cells may contribute to decreased SOD1 expression levels.

ROS levels and DNA damage, which significantly increased in cells transfected with miR-6823-5p mimic, decreased significantly following the addition of NAC (Figs. 7 and 8). These results confirmed that miR-6823-5p increased intracellular ROS levels, leading to DNA damage in MIAPaCa-2 cells.