DNA methylation affects metastasis of renal cancer and is associated with TGF-β/RUNX3 inhibition

Experiments with animals involved were conducted following the guidelines for animals approved by the Committee on the Ethics of Animal Experiments of QiLu Hospital of Shandong University and Central Hospital of Zibo. The metastatic or primary renal cancer cells (1 × 10⁴) were suspended in 50 µL of culture medium and injected into the left foreleg of the BALB/c nude male mice (purchased from Charles River, Beijing, China) aged at 6–8 weeks. As indicated, some metastatic renal cancer cells were treated with 5-aza-2′-deoxycytidine (10 mM, Sigma, St. Louis, MO, USA) for 2 days and then were injected to the left foreleg of the BALB/c nude male mice with the untreated cells as the control. The animals were monitored daily until death, and the survival days after inoculation were calculated. Animals used for tumor size analysis were euthanized at 2 months post inoculation, and tumors were dissected with size determined using a caliper. Then the tumor tissues were used for methylation and gene expression analysis.

All experiments complied with the regulations and guidelines and were approved by the ethics committee of QiLu Hospital of Shandong University and Central Hospital of Zibo. All renal cancer patients involved in this study provided written informed consent to participate in this study (Table 1 and Additional file 1: Table S1). The patients underwent the surgery to remove the renal cancer tissue, which were subjected to extraction of DNA and proteins for methylation analysis and Western blot.

Primary renal patient cancer tissues and metastatic disease tissues (from the bone, lymph, liver and lung of patients) were cut to small pieces by spring scissors, followed by collagenase treatment. After that, the mixture was filtered by 100 and 40 µM cell strainer sequentially, and centrifuged at 400g for 10 min. Finally, the cell pellet was suspended in Dulbecco’s Modified Eagle Medium (DMEM) culture media supplemented with 10% fetal bovine serum (FBS) and Pen/Strep (GIBCO-Invitrogen, 1:100). Renal cancer cells were maintained in DMEM culture media with 10% FBS at 37  °C in a humidified incubator with 5% CO₂.

Targeted quantitative RT-PCR arrays were performed to measure expression of genes with CpG islands in their promoters. The objective was to investigate the methylation regulated genes associated with metastasis, rather than the entire database of human genes.

Quantitative methylation specific PCR was performed as previously described [17]. In brief, purified DNA samples from patient cancer tissues or xenograft tumors were bisulfite-converted with EZ-96 DNA Methylation-GoldTM Kit (Zymo Researh). 10 ng bisulfite-converted DNA was then used for pre-amplification with gene specific primers and probed with 1× Taqman Universal PCR master mix (ThermoFisher Scientific) in 15 µL reaction for 15 cycles. Next, 2 µL pre-amplification product was used for qMSP analysis in 10 µL reaction system with the same primers as in pre-amplification and 1× Taqman Universal PCR master mix for 50 cycles in a 7900HT Fast RealTime PCR System (Applied Biosystems). All qMSP reactions were run in triplicates. The Alu-element based normalization assay for qMSP was conducted as quantity and quality control. RUNX3 and TGF-β detection limit was examined by decreasing amounts of methylated control DNA mixed into unmethylated control DNA. Primers used in this study were as follows: RUNX3 sense 5′-GGC AAT GAC GAG AAC TAC-3′, antisense 5′-GGA GAA TGG GTT CAG TTC-3′; TGF-β sense 5′-CAC CCG CGT GCT AAT GG-3′, antisense 5′-ATG CTG TGT GTA CTC TGC TTG AAC-3′; β-actin sense 5′-AGA GCT ACG AGC TGC CTG AC-3′, antisense 5′-AGC ACT GTG TTG GCG TAC AG-3′.

Equal amounts of protein samples (20 μg) were loaded in an 8–12% SDS–polyacrylamide gel and then were transferred to a polyvinyl-iodine fluoride membrane (Biorad, Hercules, CA, USA). After blocking (5% non-fat dry milk), membranes were incubated with primary antibodies overnight at 4 °C (rabbit anti-RUNX3, Abcam ab11905, 1:3000, rabbit anti-TGF-β, R&D systems AB-100-NA, 1:1000, mouse anti-β-actin, R&D systems MAB8929, 1:1000) followed by 1 h incubation with corresponding secondary anti-mouse (1:5000, Sigma) or anti-rabbit (1:5000, Sigma) antibodies conjugated to HRP. Chemiluminescent visualization was performed by SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific, Waltham, MA, USA) with β-actin as the internal loading control.

All data shown were repeated in at least three independent experiments and are presented as mean ± SD. Student’s t tests were used to analyze the differences between two groups. P < 0.05 was considered significant.

In this study, we utilized metastatic and non-metastatic/primary renal cancer tissues from patients. Metastatic renal cancer tissues were obtained from 67 patients (Additional file 1: Table S1) and primary renal cancer tissues were obtained from 53 patients (Table 1). The level of CpG methylation was elevated in patients with metastatic renal cancer, while the level of global methylation was decreased in metastatic renal cancer. It indicated that high CpG methylation was related to metastatic feature of renal cancer.

In order to identify the differentially expressed genes regulated by methylation between metastatic and primary renal cancer, gene array was performed. Figure 1a showed the heatmap of differentially expressed genes identified in metastatic renal cancer compared with primary renal cancer in the order of most changed genes on the top. As shown, most differentially expressed genes were downregulated, with RUNX3 and TGF-β being the top 2 down-regulated genes. Furthermore, the decreased expressions of RUNX3 and TGF-β in metastatic renal cancer compared with the primary renal cancer were confirmed by Western blot (Fig. 1b).

Next, we compared the changes in methylation levels of RUNX3 and TGF-β in metastatic renal cancer tissues, with primary renal cancer tissues as the control. As shown in Fig. 2a, elevated CpG methylation levels of RUNX3 and TGF-β were detected in metastatic cancer tissues by qMSP, while reduced global DNA methylation levels of RUNX3 and TGF-β were observed compared with primary renal cancer tissues by quantitative methylation real-time PCR (Fig. 2b). It showed 7 to ninefolds of CpG methylation level, and the global methylation levels were reduced to ~ 30% in metastatic renal cancer, as compared to primary renal cancer tissues.

To investigate the roles of RUNX3 and TGF-β in metastatic renal cancer, we established two murine models, metastatic and primary renal cancer xenograft models by injection of 1 × 10⁴ corresponding renal cancer cells into the left forelegs of mice, respectively. About 3 months later, tumors had been formed and developed into late stages of the diseases. Mice were observed daily until they died of illness and the numbers of survival days were monitored. Figure 3a showed that compared with primary renal cancer xenograft model, the survival time of mice with metastatic renal cancer xenografts was significantly shortened, indicating that metastatic renal cancer has stronger lethality. 2 months post inoculation, animals with renal cancer xenografts formed clear tumors but no death observed. For calculating the sizes of the tumors, mice were sacrificed 2 months post inoculation. Figure 3b showed that the volumes of metastatic xenografts were significantly larger than the non-metastatic xenografts, indicating that the metastatic renal cancer had stronger proliferation and growth rate. Importantly, the expressions of RUNX3 and TGF-β were significantly lower in metastatic renal cancer model than those in non-metastatic renal cancer model. These results suggested that the RUNX3-TGF-β pathway played a role in the metastasis of renal cancer.

Then we examined the methylation levels for RUNX3 and TGF-β in murine metastatic renal cancer xenograft model. The CpG methylations of RUNX3 and TGF-β were increased in the metastatic renal cancer model (Fig. 4a), while the global methylations of RUNX3 and TGF-β were decreased (Fig. 4b), consistent with the human cancer tissue results. As both RUNX3 and TGF-β play a tumor suppressor role in cancer development, their elevated CpG methylations inhibited their expression in metastatic cancer model thereby contributing to higher degree of malignancy.

As our previous data showed that the high CpG methylation levels of RUNX3 and TGF-β played an important role in the metastasis ability of renal cancer, we next investigated whether decreased level of methylation by methylated transferase inhibitor methyltransferase (5-aza-2′-deoxycytidine) has an effect in the murine metastatic renal cancer xenograft model. 1 × 10⁴ metastatic renal cancer cells with 5-aza-2′-deoxycytidine treatment or untreated control were injected into the left foreleg of mice. Figure 5a showed that the CpG methylation levels of RUNX3 and TGF-β genes were decreased by 5-aza-2′-deoxycytidine treatment in metastatic renal cancer animal model. As expected, the expressions of RUNX3 and TGF-β were increased in 5-aza-2′-deoxycytidine treated metastatic renal cancer model by immunoblotting (Fig. 5b). Compared with the untreated group, the survival time of the mice inoculated with methylation inhibited metastatic renal cancer cells were significantly prolonged (Fig. 5c). At 2 months after inoculation, the sizes of the tumors were calculated. Figure 5d showed that the tumor size of the mice inoculated with 5-aza-2′-deoxycytidine treated metastatic renal cancer cells was significantly lower than that of the non-treated metastatic renal cancer cells. These results strongly indicated that inhibition of methylation was the key to inhibit the development of renal cell cancer through elevation of TGF-β/RUNX3 pathway. In addition, another DNA methylation inhibitor, SGI-1027, was selected to verify the effects of DNA methylation on suppressing tumor (Additional file 1: Figure S1), and the results confirmed that demethylation suppressed the development of renal cancer. Furthermore, we also found that the down-regulated TGF-beta/RUNX3 exhibited no significant effects on CpG methylation and global methylation (Additional file 1: Figure S2), which further suggested that the effect of methylation on the development of renal cancer was associated with TGF-beta/RUNX3 pathway inhibition.