The human low-metastasis, low endogenous IL-8 secreating level NPC cell lines CNE-2 and the CNE-2 subclones S22 and S26 have previously been established and reported (12,31), the cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum purchased from Gibco/BRL (Grand Island, NY, USA), 100 U/ml penicillin G, 100 U of streptomycin (all from Invitrogen, Carlsbad, CA, USA). Cells were incubated at 37°C in a humidified atmosphere containing 5% CO₂. Based on our previous study (12), recombinant human IL-8 (PeproTech) at a concentration of 8 ng/ml was added into cell culture medium in experimental groups. For vehicle controls, the cells were treated with equivalent amounts of BSA.

Immunoblotting was performed as described previously (12). The sources of the primary antibodies (and their concentrations) were as follows: anti-E-cadherin (1:1,000), anti-AKT (1:1,000), anti-phospho-AKT (Ser473) (1:1,000), and anti-β-actin (1:1,000) these antibodies were purchased from Cell Signaling Technology (Danvers, MA, USA); anti-DNMT1 (1:1,000) was purchased from Santa Cruz Biotechnology (Santa Cruz). Anti-rabbit peroxidase-conjugated secondary antibodies were purchased from Promega. For western blot analysis of AKT1 signaling inhibition studies, LY-294002 (the AKT1 inhibitor) (Cell Signaling Technology) (20 μM) was used one hour prior to IL-8 stimulation in an attempt to inhibit the conventional AKT1 pathway, for western blot analysis of DNMT1 inhibition studies, 5-aza-2′-deoxycytidine (the inhibitor of DNA methylation) (Sigma) (5 μM) was used in an attempt to inhibit DNMT1 function. Cells were pretreated with 5-aza-2′-deoxycytidine for 48 h then were treated with IL-8 for 24 h. Experiments were repeated in triplicate.

Total cellular RNA was extracted using the High Pure RNA kit (Roche Applied Science, Penzberg, Germany). For RT-PCR, the total RNA was quantified spectrophotometrically, and equal amounts (1 μg) were transcribed into cDNA according to the manufacturer's protocol (Roche Applied Science). The sequences of the PCR primers used for amplifications of β-actin, IL-8, DNMT1, E-cadherin were as follows: β-actin forward: 5′-CACGATGGAGGGGCCGGACTCATC-3′; β-actin reverse, 5′-TAAAGACCTCTATGCCAACACAGT-3′. IL-8 forward, 5′-CTCCAAACCTTTCCACCCC-3′; IL-8 reverse, 5′-GATTCTTGGATACCACAGAGAATG-3′. DNMT1 forward, 5′-AGCCAAATCGGATGAGTCCATC-3′; DNMT1 reverse, 5′-CCTCCTTCAGTTTCTGTTTGGG-3′. E-cadherin forward, 5′-AACAGGATGGCTGAAGGTGA-3′; E-cadherin reverse, 5′-CCTTCCATGACAGACCCCTT-3′. Fast SYBR Green Master Mix was used to determine the threshold cycle (Ct) value of each sample in the CFX96 real-time PCR detection system (Bio-Rad, CA, USA). β-actin served as the normalization gene in these studies. The relative expression levels of the target genes were given by 2ΔCt (Ct of β-actin minus the Ct of the target gene). PCR amplifications of β-actin, IL-8, E-cadherin, DNMT1 were performed under the following conditions: denaturation at 95°C for 30 sec, annealing at 58°C for 30 sec, and extension at 72°C for 30 sec; reactions were carried out for 30 cycles. The experiments were performed in triplicate.

For IL-8 overexpression studies, the IL-8 overexpressing vector (cat no. CH832510) and control pENTER vector were purchased from ViGene Biosciences Inc. (Rockville, MD, USA). These plasmids were transfected into S22, S26, CNE2 cells with X-tremeGENE HP transfection reagent according to the manufacturer's instructions (Roche Life Science). The cells were harvest 48 h after transfection for immunoblot and real time-PCR analyses.

Twenty-four hours after IL-8 overexpressing plasmids or control pENTER vector were transfected into S22, S26, CNE2 cells with X-tremeGENE HP transfection reagent, 2×10⁶ of the cells were plated into 100-mm culture plates and incubated for another 24 h in the regular medium. The medium was then replaced with a serum-free medium (10 ml) and the cells were incubated for an additional 12 h. Conditioned medium was collected and subjected to centrifugation, followed by filtration through a 0.45 μm membrane filter to remove the debris. Secreted human IL-8 concentration in the conditioned medium was then measured using a sandwich enzyme-linked immunosorbent assay (ELISA) kit (R&D, MN, USA) following the manufacturer's instructions. The experiments were performed in triplicate.

Bisulfite sequencing PCR was used to examine the methylation status of E-cadherin gene promoter. Genomic DNA from cultured cells were extracted according to the manufacturer's protocol (Qiagen 51304), followed by bisulfite conversion using EpiTect Bisulfite kit (Qiagen 59104). DNA amplification was performed via PCR using primers as follows: E-cadherin-BSP-F: 5′-TGTAGGT TTTATAATTTATTTAGAT-3′; E-cadherin-BSP-R: 5′-CTCA CAAATACTTTACAATT-3′. PCR products were cloned to pTopo TA vector (Invitrogen), and then transfected to Top10 cells for sequencing, at least ten clones from each sample were selected.

Student's t-test was used to compare two independent groups of data. A P-value <0.05 was considered statistically significant.

Our previous study showed that IL-8 is lowly expressed in low-metastasis S26 cells (12). In the present study, we investigated whether IL-8 could enhance the expression of DNMT1 by transient transfection of IL-8 plasmid. After transient transfection of IL-8 plasmid in S22, S26, and CNE-2 cells, real-time PCR analysis showed that IL-8 mRNA levels were increased greatly 48 h later (Fig. 1A), and ELISA analysis showed high levels of IL-8 were secreted into the culture medium by the cells (Fig. 1B). Then immunoblotting of the nuclear extracts showed an increase in DNMT1 protein expression 48 h after transient transfections (Fig. 1C). Next, we used exogenous recombinant human IL-8 to verify this finding. Consistently, DNMT1 protein level was increased in the S22, S26, CNE-2 cells following IL-8 treatment with 8 ng/ml for 24 h (Fig. 1D). We next analyzed DNMT1 mRNA expression using real-time PCR, and found that DNMT1 mRNA levels were not significantly altered in these cells (Fig. 1E), suggesting that IL-8-induced accumulation of DNMT1 protein was not due to mRNA overexpression.

DNMT1 protein stability can be regulated via various post-translational modifications. It has been reported that AKT1 activity can stabilize DNMT1 protein (26,27). Our previous study showed that IL-8 can stimulate AKT1 pathway in NPC cells (12), therefore, we wanted to know whether the IL-8-induced accumulation of DNMT1 protein is partly due to AKT1 signaling. After transient transfection of IL-8, immunoblotting of nuclear extracts showed an increase in p-AKT1 protein expression in S22, S26, and CNE-2 cells (Fig. 2A). Then S22, S26, CNE-2 cells were treated with IL-8 at 8 ng/ml, and total AKT1 and phosphorylated AKT1 levels were tested by immunoblotting. The results showed that treatment of cells with IL-8 lead to activation of the AKT1 pathway as expected. To confirm that the increase of DNMT1 is AKT1-dependent, we treated cultured S22, S26, CNE-2 cells with the AKT1 inhibitor LY294002 and then measured DNMT1 protein levels. LY294002 blocked pAKT1 activation triggered by IL-8 and resulted in reduction of total DNMT1 (Fig. 2B). To confirm DNMT1 protein stabilization, we treated S26 cells with the protein synthesis inhibitor cycloheximide to block the synthesis of DNMT1. Immunoblot analysis showed that DNMT1 levels gradually decreased in 10 h following the addition of cycloheximide to culture media without IL-8. However, the degradation of DNMT1 was slowed down by the addition of IL-8 to the medium, indicating that IL-8 stabilizes DNMT1 protein, an effect persisting for ≥12 h (Fig. 2C). Therefore, we conclude that IL-8 upregulates DNMT1 protein level through activating AKT1 pathway and enhancing DNMT1 stabilization.

Our previous study shows that overexpression of IL-8 in S26 cells can induce EMT with downregulation of E-cadherin (12). In the present study, we continued to explore whether this downregulation of E-cadherin level is partly due to upregulation of DNMT1. Immunoblot analysis showed that IL-8 inhibited the expression of E-cadherin in S26 cell lines and also decreased the transcription of E-cadherin (Fig. 3A and B). As inhibiting AKT1 reduced the overall DNMT1 protein levels, we expected that treating the cells with LY294002 would result in upregulated E-cadherin expression. Then, S26 cells were treated with IL-8 for 24 h, while in the presence of LY-294002, an inhibitor of the AKT pathway, which was added one hour before the addition of IL-8, immunoblot analysis showed that LY-294002 blocked IL-8-induced pAKT1 activation and resulted in reduction of DNMT1 as well as an increase of E-cadherin protein level. Many studies have demonstrated that promoter methylation of E-cadherin is an important mechanism contributing to its downregulation (33). We speculate that methylation of the E-cadherin promoter may play an important role in IL-8 mediated decrease of this gene. In addition, while in the presence of 5-aza-2′-deoxycytidine, an inhibitor of DNA methylation, which was added 48 h before IL-8, and IL-8-induced downregulation of E-cadherin was reversed (Fig. 3C). The RNA level of E-cadherin also showed consistent changes (Fig. 3D), suggesting that the effect of IL-8 on E-cadherin expression is transcriptionally regulated via DNMT1.

Next, in order to verify that the downregulation of E-cadherin was partly due to its promoter methylation, the bisulfite sequencing PCR of E-cadherin promoter analysis was performed, the methylated cytosine residues (ranging from −280 to +40 of E-cadherin promoter) in each clone were represented by a solid spot as depicted (Fig. 4). In total, 17 CpG methylation sites were detected. Bisulfite sequencing PCR assay of the E-cadherin promoter showed that the promoter CpG island methylation status of E-cadherin in IL-8 treated cells was markedly higher than control cells (Fig. 4). In addition, when LY-294002 or 5-aza-2′-deoxycytidine was added for 48 h together with IL-8 administration, the methylation status of E-cadherin reversed to the control level. These data demonstrated that IL-8 could induce hypermethylation on the specific CpG sites of E-cadherin promoter.