RPMI-1640 medium was purchased from HyClone, GE Healthcare Life Sciences (Logan, UT, USA). Fetal bovine serum (FBS), siRNA targeting E2F1, scrambled siRNA and Lipofectamine 2000 were purchased from Invitrogen Co. (Carlsbad, CA, USA). Penicillin and streptomycin were purchased from Luye Pharmaceutical Co., Ltd. (Yantai, Shandong, China). Protein A/G beads, normal mouse IgG and a mouse anti-E2F1 monoclonal antibody (sc-251) were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). A chromatin immunoprecipitation (ChIP) assay kit (17–295) and ChIP grade mouse anti-E2F1 monoclonal antibody (17–10061) were purchased from Merck Millipore (Billerica, MA, USA). The goat anti-ADAM12 polyclonal antibody (AF1025a) was purchased from Abgent Co., Ltd. (Suzhou, China). Goat anti-mouse and rabbit anti-goat secondary antibody with HRP, and the DAB coloring kit were purchased from Beijing Zhongshan Jinqiao Biotechnology Co., Ltd. (Beijing, China). A dual-luciferase analysis kit (E1980) was purchased from Promega Corporation (Madison, WI, USA). FastDigest enzymes including NheI, BglII, KpnI and MluI were purchased from Thermo Fisher Scientific Co. (Waltham, MA, USA). A gel extraction kit was purchased from Takara Technology Co. (Dalian, China).

Human SCLC cell lines H1688 and H446, and human lung adenocarcinoma cell line A549 were preserved by our laboratory (Shandong Province Key Laboratory of Tumor Molecular Biology, Binzhou Medical University). All cells were cultured at 37°C in humidified 95% air and 5% CO₂ in RPMI-1640 medium supplemented with 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin. Forty SCLC tissue samples were obtained from Yantai Affiliated Hospital of Binzhou Medical University from January to December 2013. All samples were biopsy samples, and all patients voluntarily provided informed consent. The present study was approved by the Medical Ethics Committee of Binzhou Medical University (no. 2013007). The patient information is listed in Table I.

ChIP was conducted according to the manual provided with the ChIP assay kit (13). In brief, 5×10⁷ cells were fixed using 1% formaldehyde and were subsequently incubated in SDS lysis buffer. Ultrasound was used to fragment the genomic DNA, and the sample was pretreated with protein A/G beads, and then centrifugation (2,000 rpm, 4°C). The protein beads were then removed. The resulting sample was divided into two parts. One part was incubated overnight with the ChIP grade mouse anti-E2F1 monoclonal antibody (4 µg), and the other with normal mouse IgG (4 µg). On the following morning, the protein A/G beads were added and incubated for 2 h at 4°C. The resulting antigen-antibody-protein bead complexes were reverse crosslinked in the presence of salt at a high concentration (5 M, NaCl). DNA fragments were purified and sequenced (BGI Co., www.genomics.cn). The data processing was reported in our previous study (13).

IHC was carried out according to the protocols of our laboratory (8,13). In brief, the sections were dewaxed and rehydrated in a series of alcohols to water. Antigens were retrieved by heating the sections in citrate buffer (0.01 M, pH=6.0) for 45 min at 95°C in a boiler. The sections were subsequently incubated with the primary antibodies overnight at 4°C. The dilution of primary antibodies was 1:50 for E2F1 and 1:200 for ADAM12. On the following day, all the sections were incubated with a secondary horseradish peroxidase (HRP)-conjugated antibody, and a brown color reaction was developed using the DAB kit. Sections were counterstained with hematoxylin, differentiated, dehydrated, cleared and mounted in neutral gum. The IHC scores were assessed according to our previous studies (8,13).

RT-qPCR was performed according to our previous studies (8,13). The primers for each target gene are listed in Table II.

Western blotting was performed according to our previous studies (8,13). Cells were lysed, and 50 µg of protein was loaded and separated on 10% polyacrylamide gels (70 V for 30 min; 140 V for 70 min; 180 V for 10 min). Proteins were subsequently transferred to NC membranes by wet transfer (300 mA for 150 min), which were blocked with 5% skimmed milk powder, prior to incubation with the primary antibodies overnight at 4°C. On the following morning, the membranes were washed four times (three times in TBS buffer, one time in TBST buffer), incubated with HRP-conjugated secondary antibodies (goat anti-mouse for E2F1 and rabbit anti-goat for ADAM12, 1:5,000), washed four times and exposed to X-ray film. The dilution of primary antibodies was as follows: 1:100 for E2F1, 1:200 for ADAM12.

siRNAs targeting E2F1 and a scramble- control siRNA were used to assess the relevance of E2F1 to the expression of ADAM12 (13). Cells (1×10⁵) were cultured into a 6-well plate and transfected with Lipofectamine 2000. The procedure was performed according to previously described methods (8,13).

Genomic DNA was extracted from the H1688 cells, and the fragments (Chr10: 128010444–128011026, located in the intron of ADAM12, named seq0; Chr10: 128076927–128078127, located in the promoter of ADAM12, named seq1; Chr10: 128086195–128086876, located in the upstream 20 kb from transcription start site of ADAM12, named seq2) pulled down by E2F1 were amplified by PCR using primers with the sequences shown in Table III. These PCR fragments and the pGL3-basic promoter-less vector were digested with FastDigest NheI, BglII for seq1, KpnI and MluI for seq0 and seq2. The digested fragments were extracted using a gel extraction kit and ligated using T4 DNA ligase to generate three luciferase reporter constructs. The luciferase reporter vector containing a mutant E2F1 binding site was constructed by overlap PCR using the primers listed in Table III. These three luciferase reporter vectors were digested with FastDigest KpnI and MluI, and the seq0 and seq2 fragments were ligated into the digested seq1 vector using T4 DNA ligase to generate the fusion luciferase reporter constructs containing seq0-seq1 and seq2-seq1. All the constructs were verified by sequencing (BioSune Co., Jinan, China).

H1688, H446 and A549 cells were transfected with 0.5 µg luciferase reporter vector, 0.3 µg E2F1 expression vector or pcDNA3.1 and 0.02 µg pRL-TK Renilla reniformis luciferase. Luciferase activity was analyzed with dual-luciferase assay kits according to the instruction manual.

All the data were analyzed using SPSS 17.0 software (SPSS, Inc., Chicago, IL, USA) and paired t-tests were used to assess the significance of the differences in expression among the groups. P<0.05 was considered to indicate a statistically significant difference.

Our previous results found that ADAM12 and E2F1 were highly expressed in SCLC tissue samples, respectively (8,13). NF-κB induced the expression of ADAM12 (11,12) and p65 was highly expressed and regulated by E2F1 in the SCLC samples (13), which indicated that E2F1 may be a significant factor for promoting the expression of ADAM12. Since the tissues used for detection in our previous studies were from differential hospitals (8,13), it was unconvincing that E2F1 may regulate the expression of ADAM12. In order to solve this issue, an additional 40 SCLC tissue samples were obtained. IHC results revealed that negative/weak (<10%), moderate (10–60%) and strong (>60%) positive expression was noted in 2, 3, 12 and 22 cases for E2F1; and in 5, 4, 11 and 19 cases for ADAM12 (Table IV). These results were consistent with our previous studies (8,13). The positive expression (including moderate and strong staining) of E2F1 and ADAM12 was 85 and 75%, respectively. In the same tissue samples, ADAM12 was highly expressed in the tissue samples for which anti-E2F1 staining was strong-positive (Fig. 1). The section incubated with phosphate-buffered saline (PBS) instead of the primary antibody was considered as the negative control, and the expression of vascular endothelial growth factor receptor (VEGFR) in SCLC has been reported and was considered as the positive control (14), and the expression levels of E2F1 and ADAM12 were negative in normal alveolar epithelial cells (Fig. 1).

As shown in Fig. 1, E2F1 may regulate the expression of ADAM12. siRNA targeting E2F1 was transfected into H1688 and H446 cells, and E2F1 expression was significantly decreased at the transcription and translation levels (Fig. 2). Meanwhile, ADAM12 expression was also significantly reduced (Fig. 2). This result showed that E2F1 knockdown significantly inhibited the expression of ADAM12 at the mRNA and protein levels, thus indicating that E2F1 was able to regulate the expression of ADAM12 at the transcription level.

As ADAM12 may be regulated by E2F1, ChIP-to-seq was employed to discover the E2F1 binding sites to explore the detailed mechanism. A total of 4,700 genes regulated by E2F1 were identified by ChIP-to-seq (data not shown; these data will be reported in a subsequent study). There were three E2F1 binding sites in the ADAM12 gene (Table V). They were the following: Chr10: 128010444–128011026, located in the intron of ADAM12, named seq0; Chr10: 128076927–128078127, located in the promoter of ADAM12, named seq1; Chr10: 128086195–128086876, located in the upstream 20 kb from the transcription start site of ADAM12, named seq2. Since there were three E2F1 binding sites, the enrichment fold was calculated. The results showed that the enrichment fold of seq0, seq1 and seq2 by E2F1 was 2.3, 14.9 and 1.9 as determined by qPCR (Fig. 3), indicating that the ability of E2F1 binding seq1 was stronger than seq0 and seq2.

After considering the above-mentioned results, we speculated that E2F1 could directly regulate ADAM12 expression and tested this using dual-luciferase reporter constructs. The seq1 fragment was cloned and incorporated into the pGL3 basic vector (known as the wild-type ADAM12 promoter). The wild-type ADAM12 promoter was transfected into H1688, H446 and A549 cells, and this signifi-cantly promoted the luciferase activity. When co-transfected with the E2F1 expression vector, the luciferase activity was significantly increased (Fig. 4A). One putative E2F1 binding site (gttccGCGGtaataaga) was identified by MatInspector software (http://www.genomatix.de/) in the seq1 fragment (Fig. 4B). An E2F1 binding site mutant luciferase reporter (known as the mut-ADAM12 promoter) was constructed. Following transfection into H1688, H446 and A549 cells, overexpression of E2F1 did not significantly promote the activity of the mut-ADAM12 promoter, indicating that E2F1 stimulated the expression of ADAM12 via the E2F1 binding site in the ADAM12 promoter region (Fig. 4C).

ChIP-to-seq results showed that there were three E2F1 binding sites in the ADAM12 gene (Table V), in which seq1 as a promoter could recruit E2F1 to drive ADAM12 expression. Seq0 and seq2 also bind E2F1, but the function of seq0 and seq2 in the regulation of ADAM12 is unknown. Seq0 and seq2 fragments were cloned into the pGL3 basic vector. After transfection into H1688, H446 and A549 cells, there was no luciferase activity, indicating that seq0 and seq2 fragments could not promote the expression of luciferase (Fig. 5A). We next ascertained whether seq0 or seq2 has enhancer function. Subsequently, the seq0-seq1 and seq2-seq1 fusion luciferase expression constructs were analyzed. Compared with the seq1 fragments, the luciferase activity of fusion fragment vectors was on average decreased, but there was no statistically significant difference. When co-transfected with E2F1, the luciferase activity of the fusion fragments was significantly decreased, indicating that seq1 and seq2 repressed the ADAM12 promoter (Fig. 5B).