The human lung cancer A549 and H460 cell lines were purchased from the American Type Culture Collection (ATCC) and were maintained in RPMI-1640 medium supplemented with 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin (all HyClone; Cytiva). The human NK NK-92MI cell line (ATCC) was maintained in α Minimum Essential Medium (HyClone; Cytiva) supplemented with 2 mM L-glutamine, 0.2 mM inositol, 20 mM folic acid and 15% FBS. Cells were grown in a humidified incubator at 37°C with 5% CO₂. NK-92MI cells were maintained with optimal cell density between 2×10⁵ and 8×10⁵ cells/ml. A549 and H450 cells were maintained between 5×10³ and 5×10⁴ cells/cm².

A549 or H460 cells (1×10⁴ cells/cm²) were initially placed in a culture dish. After 8 h of incubation at 37°C, the attached cells were treated with various concentrations (0, 25, 50, 75, 100 and 200 µM) of celecoxib (Toronto Research Chemicals) at 37°C overnight. To examine time-dependent increase of ULBP-1, A549 and H460 cells were treated with 75 µM celecoxib for 2, 4, 6, 8, 10, 12 and 24 h. To investigate the expression levels of NKG2D ligands, cells were treated with sublethal concentrations (50 or 75 µM) of celecoxib. For inhibition of PI3K or JNK pathway, cells were preincubated with 10 µM LY294002 (EMD Millipore), a PI3K inhibitor, or 10 µM SP600125 (EMD Millipore), a JNK inhibitor, for 1 h at 37°C before celecoxib treatment. Cells were further treated with celecoxib as aforementioned without removing the medium containing the inhibitors.

Cell viability was measured using the WST-1 assay (Takara Bio, Inc.) according to the manufacturer's protocol. A549 and H460 cells (5×10³/well) were seeded in 96-well plates and incubated at 37°C for 24 h. Cells were treated with various concentrations of celecoxib overnight, as aforementioned. The next day, 10 µl of WST-1 was added to each well, and plates were incubated at 37°C for 1 h. Absorbance was measured using a microplate reader (Pierce; Thermo Fisher Scientific, Inc.) at a wavelength of 450 nm.

Total RNA was extracted from A549 and H460 cells using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol and was used to synthesize cDNA at 42°C for 1 h using oligo dT primers and AccuPower^® RT-premix (Bioneer Corporation). cDNA was amplified using Tenuto PCR premix (Enzynomics Co., Ltd.) and a PCR thermal cycler (Takara Bio, Inc.) with the following thermocycling conditions: 5 min at 95°C (pre-denaturation), 30 cycles of 20 sec at 94°C, 10 sec at 55°C (for MICA/B) or 65°C (for ULBPs), 30 sec at 72°C and 5 min at 72°C (final extension). The following primer pairs were used: ULBP-1 forward, 5′-GCCAGGATGTCTTGTGAGCA-3′ and reverse, 5′-CAGTGGTGAGTAGACAGGCG-3′; ULBP-2 forward, 5′-CCCTGGGGAAGAAACTAAATGTC-3′ and reverse, 5′-ACTGAACTGCCAAGATCCACTGCT-3′; ULBP-3 forward, 5′-GAGGCTCAGACTGGAACTGG-3′ and reverse, 5′-GCCTCTTCTTCCTGTGCATC-3′; MICA forward, 5′-CAGACTGCCTGCAGGAACTA-3′ and reverse, 5′-TTTCTTCTTACAACAACGGACATA-3′; MICB forward, 5′-CGGACAGACTTTCCATAT GTTT-3′ and reverse, 5′-TCCAACAACAATAAATAAGTG ATG-3′; and β-actin forward, 5′-CATCGTGATGGACTCCGGTGAC-3′ and reverse, 5′-TCAGGTAGTCAGTCAGGTCC-3′. The PCR products were analyzed via 1% agarose gel electrophoresis in TAE buffer with ethidium bromide and a UV illuminator. The densities of the bands were measured using ImageJ v1.53 software (National Institutes of Health).

Whole cell lysates from A549 and H460 cells (4×10⁵/60-mm culture dish) were prepared in RIPA lysis buffer (Invitrogen; Thermo Fisher Scientific, Inc.) supplemented with a protease inhibitor cocktail (Sigma-Aldrich; Merck KGaA). Total protein in the cell lysates was determined using the BCA assay, and sample proteins (10 µg/lane) were separated via 10% SDS-PAGE. Separated proteins were transferred onto nitrocellulose membranes after separation by electrophoresis. Membranes were blocked with 5% skimmed milk at room temperature for 2 h and washed three times with TBS-0.1% Tween (TBST). Primary antibodies against ULBP-1 (1:1,000; cat. no. ab176566, Abcam) and β-actin (1:2,000; cat. no. sc-69879, Santa Cruz Biotechnology, Inc.) were used. Following incubation with primary antibodies in TBST at 4°C overnight, unbound antibodies were washed away with TBST. Bound primary antibodies were visualized via incubation of blots with HRP-conjugated anti-rabbit secondary antibodies (1:5,000; cat. no. 7074; Cell Signaling Technology, Inc.) or HRP-conjugated anti-mouse secondary antibodies (1:5,000; cat. no. 7076; Cell Signaling Technology, Inc.) at room temperature for 1 h, and the signal was detected using chemiluminescence detection reagents (Amersham; Cytiva). The densities of bands were measured using ImageJ software version 1.53 (National Institutes of Health).

For flow cytometry, 1×10⁶ cells of A549 and H460 were incubated with anti-ULBP-1 antibody (1 µg/ml; rabbit IgG; cat. no. ab176566, Abcam) in PBS containing 0.1% FBS for 30 min on ice. Cells were washed and incubated with Alexa flour 488-conjugated secondary antibody for 30 min on ice (1:1,000; cat. no. AP132JA4; Sigma-Aldrich; Merck KGaA). Labeled cells were washed twice and resuspended in PBS. Flow cytometry data were obtained using a MACSquant flow cytometer (Miltenyi Biotec GmbH) and analyzed by MACSquant analysis software version 2.6.1517 (Miltenyi Biotech, GmbH) embedded in the machine. For fluorescence microscopy, labeled cells were collected onto glass slides using a Cytospin centrifuge (Hanil Science Industrial Co., Ltd.). Cells were analyzed via fluorescence microscopy (magnification, ×20) using the iRiS digital cell imaging system (Logos Biosystems, Inc.).

NK cell cytotoxic activity was determined using a calcein-AM assay. Briefly, target cells (A549 and H460) were washed twice with PBS and incubated with 5 mM calcein-AM (Sigma-Aldrich; Merck KGaA) in serum-free RPMI medium for 10 min at 37°C. Cells were then treated with/without celecoxib, as aforementioned. Labeled target cells were distributed into the wells of U-bottom microtiter plates at a concentration of 1×10⁴ cells/well. NK-92MI cells were used as effector cells and were added at various effector:target (E:T) ratios (0:1, 2.5:1, 5:1 and 10:1) in quadruplicate. Target cells in complete RPMI medium alone were used to determine spontaneous calcein-AM retention. Maximal lysis was determined by solubilizing three wells of target cells in lysis buffer (0.1% Triton X-100; Sigma-Aldrich; Merck KGaA). After incubation at 37°C for 8 h, assays were analyzed with 490/520 nm (excitement/emission) using a fluorescence reader. Percent specific cytotoxicity was calculated as calcein release relative to maximal lysis.

Results are expressed as the mean ± standard deviation. Statistical analysis was performed using GraphPad Prism 5.01 (GraphPad Software, Inc.). For comparisons among >2 groups, data were analyzed using a one-way ANOVA with Bonferroni correction to compare each two pairs between the indicated concentrations and control. In the NK cytotoxicity assay, Student's unpaired two-sided t-test was used to compare the lysis percentage between control and each celecoxib-treated group, or a one-way ANOVA followed by the post hoc Bonferonni test was used to compare multiple groups. P<0.05 was considered to indicate a statistically significant difference.

Cell viability was determined using a WST-1 assay after celecoxib treatment of the lung cancer A549 and H460 cell lines. Lung cancer cells were incubated with various concentrations of celecoxib (0, 25, 50, 100 and 200 µM) overnight. Cell viabilities of A549 or H460 cells were significantly decreased when treated with ≥100 µM celecoxib; therefore, 50 µM of celecoxib, which exhibited minimal cell cytotoxicity, were the concentrations used in subsequent experiments (Fig. 1).

Expression levels of NKG2D ligands, including ULBP-1, ULBP-2, ULBP-3, MICA and MICB, were examined after treatment of A549 and H460 cells with celecoxib (Fig. 2A). ULBP-1 mRNA expression increased when A549 and H460 cells were treated with 25, 50 and 75 µM of celecoxib. ULBP-2 expression significantly increased in A549 cells treated with celecoxib compared with the control. In contrast, ULBP-2 at 75 µM of celecoxib treatment slightly decreased in H460 cells compared with the control, but this was not statistically significant (Fig. 2B). Celecoxib treatment also increased ULBP-3 in A549 cells compared to the control, however, this was not observed in H460 cells. Transcript levels of other NKG2D ligands, namely MICA and MICB, did not significantly change after celecoxib treatment of A549 and H460 cells (Fig. 2B). Protein expression of ULBP-1 was further investigated as its mRNA expression distinctly. Other ULBPs proteins were also investigated and ULBP-2 protein was also increased on celecoxib-treated A549 cells (data not shown), but the main focus was on ULBP-1 which demonstrated distinct differences. It was further observed that the celecoxib-induced increase in ULBP-1 expression was dose- (Fig. 3A) and time-dependent (Fig. 3B). ULBP-1 expression was significantly increased in both A549 and H460 cells treated with ≥25 µM celecoxib (Fig. 3A). Additionally, celecoxib significantly increased ULBP-1 protein expression after 6 h of treatment (Fig. 3B). The increase in ULBP-1 expression was also investigated via flow cytometry analysis (Fig. 4A) and fluorescence microscopy (Fig. 4B). ULBP-1 was more strongly expressed by A549 than H460 cells after celecoxib treatment (Fig. 4A). Fluorescence microscopic images revealed abundant ULBP-1 in the cytoplasm after celecoxib treatment (Fig. 4B).

Cells were preincubated for 1 h with LY294002 (10 µM) as a PI3K inhibitor or SP600125 (10 µM) as a JNK inhibitor, and were then treated with a sublethal concentration (50 µM) of celecoxib. ULBP-1 expression was determined via western blotting. SP600125 significantly attenuated the induction of ULBP-1 expression after celecoxib treatment of A549 and H460 cells (P<0.05; Fig. 5A). Additionally, LY294002 significantly attenuated ULBP-1 induction in A549 (P<0.05), but not H460 cells (P=0.071) (Fig. 5B).

Finally, the present study investigated whether the celecoxib-induced increase in ULBP-1 expression resulted in increased NK cell-mediated lysis. Celecoxib-treated A549 and H460 cells were co-cultured with various ratios of NK-92MI cells, and then NK cell toxicity was measured using a calcein assay. NK-91MI cells were significantly cytotoxic to celecoxib-treated A549 cells at 5:1 and 10:1 E:T ratios, while NK cell-mediated toxicity of celecoxib-treated H460 cells was only observed at a 10:1 E:T ratio (Fig. 6A). Both SP600125 and LY294002 significantly attenuated NK cell susceptibility of A549 cells after celecoxib treatment (P<0.05; Fig. 6B). Additionally, SP600125, but not LY249002, significantly attenuated NK cell-mediated toxicity of celecoxib-treated H460 cells (P<0.05 and P=0.060, respectively; Fig. 6B).