Seventy-seven paired NSCLC tissues and adjacent non-cancerous tissues were obtained after informed consent from patients in the First Affiliated Hospital of Soochow University between 2009 and 2013. Blood specimens were obtained after informed consent from 30 randomly selected NSCLC and paired non-cancerous lung patients. Blood was isolated by centrifugation at 3,500 rpm for 20 min after blood sampling (10). The Revised International System for Staging Lung Cancer was used to determine histological and pathological diagnostics for NSCLC patients. No chemotherapy or radiotherapy was given to patients with NSCLC before tissue sampling. Tissue samples were stored at −80°C after being snap-frozen. The present study was approved by the Ethics Committee of the First Affiliated Hospital of Soochow University.

Human lung carcinoma cell lines (A549, H1650, H460, SPC-A-1, 95C, 95D, H226 and SK-MES-1) were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China), and human bronchial epithelial (HBE) cells from Shanghai Bogoo Biotechnology, Co., Ltd. (Shanghai, China). Cell lines were seeded and grown in RPMI-1640 medium (HyClone Laboratories, Inc., Logan, UT, USA), with the exception of SK-MES-1, which was seeded in MEM medium (HyClone Laboratories), with 10% heat-inactivated fetal bovine serum (FBS; Gibco, Carlsbad, CA, USA) and L-glutamine and antibiotics (Invitrogen, Carlsbad, CA, USA) in a humidified incubator containing 5% CO₂ at 37°C. Treatment with 5-aza-2′-deoxycitidine (5-Aza; Sigma-Aldrich, St. Louis, MO, USA) was used to demethylate cells in culture according to the previously described treatment protocol (10).

The human 8-oxoguanine DNA glycosydase (hOGG1) ELISA kit (cat. no. CSB-E12686h; Wuhan, Hubei, China) was used to quantitatively determine hOGG1 concentrations in serum. Briefly, standard controls and samples (100 µl) were added to each well and covered with a provided adhesive strip; cells were incubated at 37°C for 2 h. The standard and sample results were recorded in a plate layout that was provided by the manufacturer. Next, each well received 100 µl of biotin-antibody (1x), and the plate was covered with a new adhesive strip. Cells were incubated at 37°C for 1 h. If the biotin-antibody (1x) appeared cloudy, it was warmed to room temperature and mixed gently until the solution appeared uniform. Next, each well received 100 µl of HRP-avidin (1x), and the plate was covered with a new adhesive strip. Cells were incubated at 37°C for 1 h. The aspirate was washed five times, and 90 µl of TMB substrate was added to each well. Cells were incubated at 37°C for 15–30 min. Cell plates were protected from light. A total of 50 µl of stop solution was added to each well, thoroughly mixed and hOGG1 was detected using a microplate reader.

8-OHdG, 8-hydroxyguanine and its 2′-deoxynucleoside equivalent, 8-hydroxy-2′-deoxyguanosine (8-OHdG) are common byproducts of DNA damage. During the repair of damaged DNA in vivo by exonucleases, 8-hydroxy-2′-deoxyguanosine (8-OHdG) is excreted (21,22); therefore, we used an OxiSelect Oxidative DNA Damage ELISA kit (8-OHdG Quantitation; Cell Biolabs, Inc., San Diego, CA, USA) to evaluate the level of DNA damage in serum according to the manufacturers instructions (23–25). The extracted DNA was dissolved in water to reach a concentration of 1–5 mg/ml. The DNA was converted to single-stranded DNA by incubating at 95°C for 5 min, then promptly chilling on ice. The denatured DNA samples were digested to nucleosides by incubating with 5–20 units of nuclease P1 at 37°C for 2 h in 20 mM sodium acetate (pH 5.2). DNA samples were treated with 5–10 units of alkaline phosphatase at 37°C for 1 h in 100 mM Tris (pH 7.5), followed by centrifugation for 5 min at 6,000 × g. Subsequently, the supernatant was used for the 8-OHdG enzyme-linked immunosorbent assay (ELISA).

Total RNA of cells and tissues were extracted by adding 1.0 ml RNAiso Plus (Takara Bio, Osaka, Japan) according to the manufacturers protocol. The RNA concentration was measured using a NanoDrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA) and synthesis of cDNA was performed using Reverse Transcriptase M-MLV (Takara Bio) with reverse transcriptase. The sequences of qRT-PCR for hOGG1 and GAPDH were as follows: hOGG1, forward, 5′-ATCGTACTCTAGCCTCCACTCC-3′ and reverse, 5′-GTCAGTGTCCATACTTGATCCGC-3′; GAPDH, forward, 5′-TGCACCACCAACTGCTTAGC-3′ and reverse, 5′-TGCACCACCAACTGCTTAGC-3′. qRT-PCR was performed using SYBR Premix ExTaq™ (Takara Bio) according to the manufacturers instructions on an ABI StepOnePlus Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). The PCR program was 50°C for 2 min, 95°C for 10 min, followed by 45 cycles of 95°C for 15 sec and 60°C for 1 min.

Quantitative methylation analysis of DNA was performed using MassARRAY EpiTYPER assays (Sequenom, Inc., San Diego, CA, USA) according to the protocol recommended by the manufacturer (26). This system uses matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry in combination with RNA base-specific cleavage (Mass Cleave). After bisulfite modification, genomic DNA was amplified using MassArray primers. PCR products were introduced to a T7 promoter sequence by the Beijing Bio-Miao Biotechnology, Co., Ltd. (Beijing, China). Next, RNA products were transcribed in vitro using T-base-specific cleavage, in which small RNA fragments were obtained. The molecular weight of each fragment was detected by flight mass spectrometry (MALDI-TOF) and methylation levels were analyzed using EpiTyper software. PCR amplification bias was controlled for by using DNA methylation standards (0, 20, 40, 60, 80 and 100%) and data was normalized by correction algorithms based on an R statistical computing environment.

To construct a plasmid containing the hOGG1 promoter, we used pGL3 basic vector (Promega, Madison, WI, USA). Briefly, an 88-bp fragment containing the predicted Sp1 target site (positions +322–327) was chosen for the luciferase assay. The wild-type and mutated fragment was directly synthesized (Genewiz, Suzhou, China) and subcloned into the pGL3 basic vector to generate pGL3-wild-type (WT: tggtccttgtctgggCGgggtctttgggCGtCGaCGaggcctggt tctggg taggCGgggctactaCGgggCGgtgcctgctgtggaa) and pGL3-mutant plasmid (Mut: tggtccttgtctgggCGgggtctttgggCGtCG aCGaggcctggttctgggtaggCGgggctactaCTggAATgtgcctgctgt ggaa). Subsequently, A549 and SPC-A1 cells were plated in a 24-well plate and cotransfected with wild-type plasmid, mutated plasmid, or pRL-TK plasmid using Lipofectamine 2000 (Life Technologies, Carlsbad, CA, USA). After 48 h, cells were collected, and luciferase activities were measured by the Dual-Luciferase reporter assay kit (Promega). Each experiment was performed in triplicate.

ChIP assay was carried out as previously described (20). Briefly, immunoprecipitation was performed using 5 µg anti-Sp1 antibody (Cell Signaling Technology, Beverly, MA, USA). Purified ChIP DNA was subjected to PCR, using primers specific for the hOGG1 promoter region (positions +247 to +398) encompassing the putative Sp1-binding site. Specific ChIP primers used for PCR were as follows: forward, 5-TAAGGGTCGTG GTCCTTGTC-3 and reverse, 5-TGGAGGCTAGAGTACGA TGC-3.

hOGG1 gene encodes a DNA glycosylase that catalyzes the excision and removal of 8-OH-dG adducts (27). A previous report has shown a decrease of hOGG1 in the brain of Alzheimers patients (28), which caused the accumulation of 8-oxoG in the mitochondrial DNA of neurons and calpain-dependent neuronal loss (12). In addition, an association between the hOGG1 gene and lung cancer risk has been reported (27,29). Here, we detected the level of hOGG1 and 8-OH-dG using ELISA assays. As illustrated in Fig. 1A, the level of hOGG1 was lower in NSCLC serum than in paired normal serum. Furthermore, we detected 8-OH-dG and found that the level of 8-OH-dG was higher in NSCLC serum than in paired normal serum (Fig. 1B).

hOGG1 mRNA levels were significantly lower in NSCLC tissues compared with adjacent non-cancerous lung tissues (P=0.034; Fig. 1D). No significant differences were observed in hOGG1 mRNA levels between NSCLC tissues classified by various clinicopathological characteristics (Table I). Moreover, a public data set (GSE19188) containing 91 NSCLC tissues and 65 normal lung tissues showed that hOGG1 mRNA expression was downregulated in human NSCLC tissues (P=0.002; Fig. 1C).

As shown in Fig. 2A, hOGG1 mRNA levels were significantly lower in A549, H1650, H460, 95C, 95D, H226 and SK-MES-1 cells compared with control HBE cells, except for SPC-A-1. Our previous study supported the idea that DNA methylation could be epigenetically responsible for inactivation of tumor suppressor genes in NSCLC, and the methylation of the hOGG1 gene promoter region occurs frequently in NSCLC (10,19). Therefore, to determine whether methylation of hOGG1 gene promoter is an alternative mechanism underlying inactivation of hOGG1 mRNA expression, we detected the mRNA expression after using demethylating agent 5-Aza on NSCLC cell lines. As illustrated in Fig. 2B, after 5-Aza treatment, hOGG1 mRNA expression was increased in NSCLC cell lines (A549, H460, SPC-A1, 95D and SK-MES-1); therefore, we suggest that hOGG1 expression is silenced by DNA methylation.

It is known that CpG islands are located −200 to −1,000 bp from the transcription start site of a gene. Based on this knowledge, we used Methyl Primer Express^® software to identify potential CpG sites in the hOGG1 promoter and observed a GC-rich region (Fig. 3A). Furthermore, by using the MassARRAY EpiTYPER application, methylation levels of CpG sites in the hOGG1 gene were observed in 10 paired NSCLC tissues and adjacent non-cancerous lung tissues (Tables II–IV). In the present study, we detected three separate regions (position −1000 - −643, −463 - +34 and +35 - +412), including 68 CpG sites, in the hOGG1 promoter. Several CpG sites were detected between positions −1000 - −643 and −463 - +34, and differences in methylation levels were detect (P<0.05). However, we found that these regions have no transcriptional binding sites or have lower frequency of methylation. Consequently, we expanded the sample size to 25-paired tissues to detect potential CpG sites within the third area (position +35 - +412). We observed significantly higher methylation of CpG site-3 in NSCLC patients compared with the control group (Fig. 3B; Table V). Notably, the methylation level of +322 - 327 site (T/N) was inversely correlated with hOGG1 mRNA level (T/N) in 25 paired tissues (P=0.0104; Fig. 4C).

Because methylation of individual CpG dinucleotides may contribute to cancer development (20,30), we postulated that site-specific CpG methylation could alter the expression of hOGG1 in NSCLC. We found one putative functional CpG site at position +322–327 in the proximal promoter region of hOGG1 that was located within a transcription factor Sp1-binding sequence (5-CGGGGCGG-3) using TRANSFAC, TFSEARCH and Methyl Primer Express^® software (Fig. 4A). We utilized ChIP analysis to examine whether Sp1 binds to the hOGG1 proximal promoter region at the +322–327 CpG site, and found that Sp1 was recruited to the +322–327 CpG site in A549 cells (Fig. 4B). Collectively, the results demonstrated that Sp1 may be target hOGG1 5-promoter containing the +322–327 CpG site and thereby upregulate hOGG1 expression in NSCLC cells. Subsequently, we synthesized segments of the hOGG1 promoter that contained a wild-type and a mutant Sp1 binding site for use in luciferase reporter constructs; these constructs were transfected into A549 and SPC-A1 cells. The mutant Sp1 binding site construct displayed a significant decrease in luciferase activity compared with the wild-type construct (Fig. 4C). In summary, our results suggest that the methylation of the +322–327 site in the hOGG1 promoter represses Sp1 binding and regulates the expression of hOGG1 in NSCLC cells.