TRIzol was purchased from Invitrogen Life Technologies (Carlsbad, CA, USA) and M-MLV reverse transcriptase was purchased from Promega Corporation (Madison, WI, USA). MIC A/B and NKG2D mouse anti-human monoclonal antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA), hypoxia-inducible factor 1-α (HIF-1α) rabbit anti-human polyclonal antibody was purchased from Boster Systems Inc. (Pleasanton, CA, USA) and glyceryl trinitrate (GTN) transdermal patches were purchased from Schwarz Pharma (Brussels, Belgium).

A PANC-1 cell line was obtained from the Shanghai Institute of Biological Sciences (Shanghai, China) and maintained in a monolayer culture in Dulbecco’s modified Eagle’s medium supplemented with 20% fetal bovine serum (FBS). The study was approved by the ethics committee of Central South University (Changsha, China).

All mice used were handled in compliance with the guide for the care and use of laboratory animals. A total of 32 nude mice were divided into two groups; a GTN and a placebo group. The mice were subcutaneously injected with 2×10⁷ cells and tumor size was measured using a caliper. The tumor volume was calculated using the following formula: Tumor volume (mm³) = [(width)² × length]/2. When tumors had grown to a volume of 150 mm³ the GTN group was treated with GTN transdermal patches (dose, 7.3 μg/h) while the mice in the placebo group were administered untreated patches.

Tumor samples were fixed with 10% paraformaldehyde and embedded in paraffin. Samples were cut into sections (thickness, 4-μm) using a Leica-RM2135 rotary microtome (Leica, Mannheim, Germany) and adhered to microscope slides. The sections were dewaxed and washed three times with phosphate-buffered saline (PBS). For non-specific blocking, sections were incubated in 10% normal goat serum for 30 min at 37°C and incubated overnight with the primary polyclonal rabbit anti-human HIF-1α and monoclonal mouse anti-human MIC A/B antibodies at 4°C (Santa Cruz Biotechnology, Inc.). Following three washes with PBS, the sections were incubated in secondary mouse polyclonal anti-rabbit and goat anti-mouse antibodies conjugated to biotin and horseradish peroxidase (HRP) marked polyclonal anti-biotin for 30 min at 37°C (Santa Cruz Biotechnology, Inc.). Next, the samples were incubated in freshly prepared 3,3′-diaminobenzidine and counterstained with hematoxylin. Finally, samples were observed under an optical microscope (CX41, Olympus Corporation, Tokyo, Japan).

The positive reaction for the MIC A/B protein results in is yellow or brown granules, located in the cell membrane and/or the cytoplasm. Briefly, each slice was randomly selected and 10 were viewed under a high-power field, the following staining intensity score was used: 0, negative staining, all cancer cells without staining; 1, weak positive staining, found in <20% of cells. In a high power field the staining is markedly strong in the cells compared with the interstitial tissue; 2, strong positive staining, staining was markedly stronger in all cancer cells compared with the interstitial tissue. MIC A/B expression was subsequently divided as follows: 0–1 points, low expression group; 2 points, high expression group.

The positive result for the HIF-1α protein is the observation of brown particles in the nucleus or cytoplasm. The entire slice was viewed under a light microscope, (magnification, ×400) were randomly observed in 10 visual fields. The staining score used was as follows: (−), no staining, or <1% nuclear staining; (+), 1–10% nuclear staining and/or weak cytoplasmic staining; (+), 10–50% nuclear staining and/or distinct cytoplasmic staining; (+++): >50% nuclear staining and/or strong cytoplasmic staining. (−)/(+) Low expression; (+)/(++) high expression.

All RNAs were extracted with TRIzol. RT was conducted using 1 μg total RNA and the following oligo (dT) primers: Forward, 5′-AGGTACATCTGGATGGTCAG-3′ and reverse, 5′-TTGTCTTCATGGATCTCACA-3′ for homo-MIC A, yielding an amplified fragment of 232 bp; forward, 5′-CTTCGTTACAACCTCATGGT-3′ and reverse, 5′-ATATGAGTCAG GGTCCTCCT-3′ for homo-MIC B, yielding an amplified fragment of 227 bp; and forward, 5′-ACCACAGTCCATGCCATCAC-3′ and reverse, 5′-TCCACCACCCTGTTGCTGTA-3′ for homo-GAPDH, yielding an amplified fragment of 450 bp. A total of 40 cycles of PCR were performed at an annealing temperature of 54°C.

Total proteins were extracted from the tumor tissues and the protein concentration was determined using the Bradford method (15). A total of 50 μg protein was loaded per lane. Following electrophoresis, all proteins were transferred to a cellulose acetate membrane (Koch Membrane Systems, Inc., Wilmington, MA, USA) and the membrane was blocked with 5% non-fat milk in PBS overnight at 4°C. The following day, the blot was incubated with a primary monoclonal mouse anti-human antibody (1:1,000; Santa Cruz Biotechnology, Inc.) for 2 h at 37°C and washed three times with PBS. Next, the blot was incubated with a HRP-conjugated goat anti-mouse secondary antibody (1:2,000; Santa Cruz Biotechnology, Inc.) for 1 h at 37°C and X-ray films were used to observe the bands of the western blot. The proteins were analysed using Scion Image version 4.0.3.2 (Scion Corporation, Frederick, MD, USA). The protein content was observed relative to that of the GAPDH control band.

Two non-overlapping-epitope antibodies were used. Plates were coated with the primary mouse anti-human MIC A/B antibody overnight at 4°C, blocked with 5% FBS for 2 h at 37°C and washed with PBS. Next, samples were added and incubated for 1 h at 37°C. The plates were washed with PBS and incubated with the secondary mouse anti-human antibody conjugated with HRP for 1 h at 37°C. Finally, the substrate was added and the absorbance was measured at a wavelength of 450 nm using ELISA (DU640, Beckman Coulter, Miami, FL, USA).

Blood was collected from the eye of the mice. Antibodies conjugated with fluorochrome were added and incubated for 20 min at room temperature in the absence of visible light. Next, hemolysin was added and the samples were centrifuged (JM-6, Beckman Coulter) at 1,000 × g for 5 min to obtain the precipitates. Following three washes with PBS, the cells were resuspended with 1% paraformaldehyde and analyzed using a BD FACSCalibur^TM flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).

All statistical analyses were performed using the SPSS 13.0 statistical software package (SPSS, Inc., Chicago, IL, USA). Differences between groups were compared using a single factor analysis of variance and Student’s t-test. P<0.05 was considered to indicate a statistically significant difference.

The mice were sacrificed following 28 days of treatment. It was found that the tumors of the GTN group were markedly smaller than those of the placebo group (Figs. 1 and 2; Table I) and the inhibition ratio of tumor growth increased in a time-dependent manner (Fig. 3 and Table II).

As shown in Table III, HIF-1α expression was lower in the GTN group (8/16 mice) than that of the placebo group (14/16 mice). Notably, 12/16 mice from the GTN group highly expressed MIC when compared with the placebo group, whereby only two mice expressed MIC. Furthermore, immunohistochemical analysis revealed that HIF-1α expression was lower in the GTN group and MIC expression was higher, when compared with that of the placebo group (Fig. 4). Furthermore, HIF-1α and MIC were found to be negatively correlated (Table IV).

To investigate the effect of GTN on MIC A/B expression, total RNA and protein was extracted from the tumors. No significant difference in MIC A/B mRNA expression was identified between the GTN and placebo groups (Fig. 5 and Table V; P>0.05). However, at the protein level, MIC A/B expression was significantly higher in the GTN group compared with that of the placebo group (Fig. 6 and Table VI; P<0.01).

As noted above, the same mRNA levels, but different protein levels of MIC A/B were identified between the two groups. To assess whether cellular MIC was transferred outside of the membrane, the levels of sMIC in serum were investigated. ELISA revealed that the placebo group exhibited higher levels of sMIC than the GTN group and the two groups exhibited significantly higher sMIC levels when compared with those of the normal mice (Fig. 7 and Table VII).

Blood samples were collected from the two groups to analyze the levels of NKG2D. Flow cytometry revealed that the GTN and normal groups exhibited significantly higher levels of NKG2D when compared with the placebo group (Figs. 8 and 9; Table VIII). In addition, NKG2D and sMIC were found to be negatively correlated (data not shown).