The human lung cancer carcinoma cell lines, A549, H460 and 11-18, the human monocytic cell line, THP-1, the Burkitt's lymphoma B cell line, Raji, human primary umbilical vein endothelial cells (HUVECs) and the human umbilical vein cell line, EA.hy926, were all purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). All the cells, apart from the HUVECs and EA.hy926 cells were maintained in RPMI-1640 medium (HyClone, Logan, UT, USA). The HUVECs were maintained with F-12K medium containing endothelial growth factors and the EA.hy926 cells was maintained in DMEM supplemented with 10% fetal bovine serum (FBS), with 100 U/ml penicillin and 100 µg/ml streptomycin (all from HyClone) at 37°C in a humidified 5% CO₂ atmosphere.

To determine the optimal concentration, the cells were incubated with celecoxib (cat. no. C251000; Toronto Research Chemicals, Toronto, ON, Canada) at 0, 25, 50, 75 or 100 µM for 16 h, and 75 µM for 0, 2, 4, 6, 8 or 12 h. To perform further experiment with minimally affected cell growth inhibition, 75 µM of celecoxib was treated in lung cancer cells for 16 h.

Peripheral blood mononuclear cells (PBMCs) from human blood were isolated by Ficoll-Paque (Amersham Life Science, Buckinghamshire, England) density-gradient centrifugation. Monocyte purification was achieved using a monocyte isolation kit (Miltenyi Biothec, Bergisch Gladhach, Germany). PBMCs were combined with FcR blocking reagent and biotin-antibody cocktail and incubated for 10 min at 4°C. Subsequenlty, anti-biotin MicroBeads were added and incubated for 15 min at 4°C. The cells were centrifuged at 300 × g, 4°C for 10 min after washing and resuspended with buffer. The cell suspension was loaded into columns and rinsed with buffer. The cells that passed through were unlabeled cells, representing the enriched monocyte fraction. This study was approved by the Institutional Bioethics Review Board at the College of Medicine, Inje University, Busan, Korea and all donors provided informed consent for the use of their samples in this study.

Cell viability was examined by WST-1 assay (Takara, Shiga, Japan). The cells were seeded in 96-well plates (5×10³ cells/well) and incubated for 24 h. The cells were treated with various concentrations of exosomes overnight. WST-1 then was added to each well, and the plates were incubated for 1 h and the absorbance was measured using a microplate reader (Multiskan FC model; Thermo Scientific, Waltham, MA, USA) at a wavelength of 450 nm.

Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Briefly, the harvested cells were mixed with TRIzol reagent and chloroform. The mixture was centrifuged at 12,000 × g for 15 min at 4°C. The supernatant was discarded and washed with 75% ethanol in DEPC-treated water. The pellet containing RNA was resuspended with DEPC-treated water. RT-premix (Bioneer, Daejeon, Korea) was used for cDNA synthesis from 2 µg of total RNA. The mixture of RNA and oligo(dTs) was combined with RT-premix and incubated for 1 h at 42°C and 5 min at 95°C. cDNA was amplified using primers the following primers: COX-2 forward, GCT CCT ACC CAC GCA GAT TT and reverse, AGA CGC CAT TTG GAT TGG GT; and β-actin forward, ATC CAC GAA ACT ACC TTC AA and reverse, ATC CAC ACG GAG TAC TTG C and Tenuto PCR premix (Enzynomics, Daejeon, Korea) in a PCR Thermal Cycler (Takara). PCR products were analyzed by gel electrophoresis in TAE buffer through 1% agarose gels. The densities of bands were analyzed using ImageJ 1.51K software [National Institutes of Health (NIH), Bethesda, MD, USA].

The cells were harvested, rinsed with phosphate-buffered saline (PBS) (Welgene, Gyeongsan, Korea) and lysed in lysis buffer (Invitrogen) supplemented with protease inhibitor cocktail (Sigma, St. Louis, MO, USA). The cells were incubated on ice for 10 min and centrifuged for 10 min at 15,000 × g. Protein quantity was determined by BCA assay (Thermo Scientific) and then separated on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gels overlaid with a 4% polyacrylamide stacking gel. The proteins were transferred onto a nitrocellulose membrane after separation. The membranes were blocked with 5% skim milk for 1 h at room temperature and washed 3 times with Tris-buffered saline containing 0.1% Tween-20 (TBST, pH 7.4) for 10 min each. Each membrane was incubated with primary antibodies to COX-2 (1:1,000 dilution, cat. no. sc-12282; Cell Signaling Technology, Beverly, MA, USA), CD63 (1:500 dilution, cat. no. sc-15363) or β-actin (1:2,000 dilution, cat. no. sc-47778) (both from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) in 0.1% TBST at 4°C overnight. The membranes were then washed with 0.1% TBST 3 times for 10 min and incubated with secondary antibody [anti-rabbit IgG [1:5,000 dilution, cat. no. 7074; Cell Signaling Technology) for COX-2 and CD63; anti-mouse IgG (1:5,000 dilution, cat. no. 7076; Cell Signaling Technology) for β-actin] for 1 h at room temperature. Following incubation, each membrane was washed with 0.1% TBST 3 times. β-actin or CD63 was detected as an internal standard or exosome marker, respectively. Signals were detected by chemiluminescence (Fuji Photo Film Co., Tokyo, Japan). The densities of the bands were analyzed using ImageJ 1.51K software (NIH).

Exosomes were isolated using an Exospin exosome purification kit (Cell Guidance Systems, St. Louis, MO, USA) according to the manufacturer's instructions. The H460 and 11-18 cells were cultured in 150-mm dishes and cell culture supernatant was collected by centrifugation at 300 × g for 10 min. The supernatant was transferred into new tubes and centrifuged at 20,000 × g at 4°C for 30 min in a high-speed centrifuge. The collected supernatant was transferred into new tubes and buffer A (provided with the kit) was added. The mixture was incubated at 4°C for overnight after mixing well. The following day, the mixture was centrifuged at 20,000 × g at 4°C for 1 h. Exosome-containing pellets were resuspended with PBS following the aspiration of the supernatant. Spin columns were prepared and spun down at 50 × g for 30 sec. Pellets containing exosomes were loaded into the tops of columns and centrifuged at 50 × g for 60 sec. The columns were placed into 1.5-ml tube and rinsed with PBS or medium applied at the top and centrifuged at 50 × g for 60 sec.

The purified exosomes were used to examine cell viability, the transport of COX-2, and the production of cytokines in cells uptaking exosomes. To examine cell viability, various concentration (100, 250 or 500 µg/ml) of exosomes were incubated with the THP-1 cells for 24 h. For the measurements of the COX-2 concentration and the production of cytokines, exosomes (75 or 150 µg/ml) were used to treat the THP-1, human primary monocytes, HUVECs, EA.hy926 or Raji cells for 16 h.

Exosomes were purified using an Exospin kit and resuspended in a small volume of PBS. Exosomes were fixed with 2.5% glutaraldehyde solution, washed and post-fixed in 1% osmium tetroxide. The pellets were then dehydrated through an ethanol series followed by acetone, and then embedded in epoxy resin. Ultrathin sections were contrasted with uranyl acetate and lead citrate. Transmission electron microscopy JEM-1200 EX II (Jeol, Tokyo, Japan) was used to view these sections at 80 kV.

To measure the concentrations of PGE₂ and VEGF, cell culture supernatants were examined using an EIA kit (Cayman Chemical Co., Ann Arbor, MI, USA) and an ELISA kit (R&D Systems, Minneapolis, MN, USA) according to the instructions provided by the respective manufacturers. The kit contains the required solutions and all antibodies, and concentrations and dilution ratios followed the instructions. Briefly, PGE₂ standard solution and culture supernatants were added to plates with goat polyclonal anti-mouse IgG. PGE₂ AChE tracer and PGE₂ monoclonal antibodies (provided with the kit) were added followed by incubation for 18 h at 4°C. Ellman's reagent (provided with the kit) was added to the wells after rinsing 5 times with washing buffer and the plates were incubated for 1 h. The absorbance was measured with a microplate reader (Multiskan FC model, Thermo Scientific) at 405 nm.

To coat the plates, VEGF capture antibody (provided with the kit) was added to the wells followed by incubation overnight. Each plate was then washed and blocked for 1 h. Standard solutions and culture supernatants were added after washing. VEGF detection antibody (provided with the kit) was then added followed by incubation for 2 h, followed by the addition of streptavidin-HRP and incubation for 20 min in indirect light and washing of each plate three times. Substrate solutions were added followed by incubation for 20 min in indirect light. Stop solution was added and the optical density of each well was determined using a microplate reader (Multiskan FC model; Thermo Scientific) set to 450 nm.

Tube formation assay was conducted in 96-well plates coated with 50 µl of growth factor-reduced Matrigel (BD Biosciences, San Jose, CA, USA) according to the manufacturer's instructions. The EA.hy926 cells (1.0×10⁴ cells/well) were treated with filtered THP-1 supernatant (following incubation with exosomes carrying high concentrations of COX-2) or control medium for an additional 16 h. Tube formation was photographed under a microscope (iRIS; Logos Biosystems, Anyang, Korea). The numbers of tubes in an identical area were manually counted and statistically analyzed using the Student's t-test. A P-value <0.05 was considered to indicate a statistically significant difference.

The densities of all the bands were measured using ImageJ 1.51K software (NIH), and the relative ratio to the internal standard was calculated. Two-sided Student's t-tests were used to evaluate differences between 2 groups. For comparison of viabilities between more than 2 groups, data were analyzed using one-way ANOVA (with Tukey's multiple comparison test). The data from ELISA and tube formation data are presented as the means ± SD. A P-value <0.05 was considered to indicate a statistically significant difference.

Cytotoxicity was examined when the lung cancer cell lines were treated with celecoxib. The A549, H460 and 11-18 lung cancer cell lines were incubated with various concentrations of celecoxib. The concentration of celecoxib leading to the 50% suppression of cell viability was estimated to be ~50 µM in the H460 and 11-18 cells, but >100 µM in the A549 cells. We observed that higher concentrations of celecoxib led to a greater decrease in the viability of the lung cancer cells (data not shown).

To observe the effects of celecoxib on COX-2 expression in lung cancer cells following treatment with celecoxib at various concentrations and treatment durations, the 3 lung cancer cell lines, A549, H460 and 11-18, were treated with 25, 50, 75 or 100 µM of celecoxib for 6 h. COX-2 expression was then detected by western blot analysis. The expression of COX-2 was elevated in a dose-dependent manner (Fig. 1A), as well as in a time-dependent manner when the cells lines were treated with 75 µM of celecoxib for 2, 4, 6, 8 or 12 h. A marked increase in COX-2 expression was observed after 4 h (Fig. 1B). COX-2 mRNA expression was immediately increased following treatment with celecoxib (Fig. 1C).

The H460 cells were cultured in medium supplemented with exosome-free FBS, and exosomes were then isolated. Transmission electron microscopy revealed that purified exosomes were ~100 nm in size and had a dish-like shape (Fig. 2A).

The H460 and 11-18 cells were incubated in exosome-free FBS with 75 µM of celecoxib for 16 h, and cell lysates and exosomes were then prepared from cells and supernatants, respectively. The expression level of COX-2 was detected by western blot analysis. COX-2 expression was increased in the cell lysates and exosomes collected from the celecoxib-treated cells compared to the control cells. The concentration of COX-2 in exosomes from the 11-18 cells was much higher than that from the H460 cells. CD63 was used as an exosome marker and β-actin was used as internal standard for cell lysates (Fig. 2B).

Exosomes were isolated from celecoxib-treated and untreated 11-18 cells. Cell viability was determined using the monocytic leukemia cell line, THP-1, to confirm that there was no change in the cell response by substances derived from other cells. The THP-1 cells were seeded on 96-well plates and then 100, 250 and 500 µg/ml of exosomes were added. WST-1 was added following incubation for 24 h and detected using a microplate reader. In all wells, the suppression of cell viability was not observed following treatment with exosomes from the control and celecoxib-treated cells (P>0.05) (Fig. 3A).

As a small quantity of celecoxib carried in exosomes can directly affect THP-1 cells, the THP-1 cells were treated with celecoxib at concentrations of 0.05, 0.5, 5 and 50 µM for 16 h and the concentration of COX-2 was detected by western blot analysis (Fig. 3B). Direct incubation with celecoxib did not increase COX-2 protein expression in THP-1 cells.

To determine whether exosomes can transfer COX-2 to other cells, exosomes isolated from the supernatant of the 11-18 cells were added to THP-1 cells at 75 and 150 µg/ml. Following incubation with exosomes, RNA was extracted from the THP-1 cells and RT-PCR was performed. The same mRNA expression level of COX-2 was observed in all samples (Fig. 3C). However, different protein levels of COX-2 were observed, as shown by western blot analysis. COX-2 expression was higher in the cells treated with exosomes from the celecoxib-treated 11-18 cell supernatant. A greater concentration of COX-2 was detected in the cells treated with 150 µg/ml of exosomes than in those treated with 75 µg/ml (Fig. 3D). Thus, COX-2 protein appeared to have been transferred to the THP-1 cells. Exosomes did not transfer COX-2 mRNA, and did not stimulate the THP-1 cells.

To determine whether COX-2 concentration can be altered in other cells through exosomes, several types of cells were treated with exosomes. Human monocytes were isolated from human blood. PBMCs were isolated from blood by Ficoll-Plaque using density-gradient centrifugation and monocytes were collected using a monocyte isolation kit. Exosomes were added to human monocytes for 16 h and COX-2 protein expression was detected by western blot analysis. As in the monocytic leukemia cell line, THP-1, the concentration of COX-2 increased in the human primary monocytes following treatment with exosomes from the celecoxib-treated 11-18 cell supernatant (Fig. 4A).

To examine whether exosomes from lung cancer cells transfer COX-2 to the same lung cancer type, exosomes from the 11-18 cells were added to the 11-18 cells. However, the concentration of COX-2 was not altered (data not shown).

Exosomes isolated from the celecoxib-treated 11-18 cell supernatant were added to Raji cells (a Burkitt's lymphoma B cell line), EA.hy926 (a human umbilical vein cell line) and HUVECs (a human primary umbilical vein endothelial cell line). The concentration of COX-2 was not higher in the Raji cells, but differed in the EA.hy926 cells and HUVECs (Fig. 4B).

To measure the function of COX-2 transferred from exosomes, a PGE₂ EIA assay was performed. The cell culture supernatant collected following treatment of the THP-1 cells with exosomes from celecoxib-treated 11-18 cells for 16 h revealed that PGE₂ production increased following incubation with the exosomes (Fig. 5A). As PGE₂ stimulates angiogenic VEGF production in monocytes (29), VEGF ELISA was performed using the supernatant of THP-1 cells treated with exosomes. The production of both PGE₂ and VEGF was higher in the THP-1 cells treated with exosomes than in the controls (Fig. 5A).

Following incubation with exosomes from the celecoxib-treated lung cancer cells, conditioned medium of THP-1 was added to the EA.hy926 cells. In response to VEGF in medium, tube formation was observed in the EA.hy926 cells, as in HUVECs. The THP-1 cells incubated with exosomes from the celecoxib-treated 11-18 cells produced angiogenic factors, which triggered more tube formation in the EA.hy926 cells than in the control exosomes (Fig. 5B). Tube formation counts were measured and analyzed, as shown in Fig. 5C.