DNA Methylation mediated down-regulating of MicroRNA-33b and its role in gastric cancer

Abstract

The discovery of microRNAs (miRNAs) provides a new and powerful tool for studying the mechanism, diagnosis and treatment of human cancers. Currently, down-regulation of tumor suppressive miRNAs by CpG island hypermethylation is emerging as a common hallmark of cancer. Here, we reported that the down-regulation of miR-33b was associated with pM stage of gastric cancer (GC) patients. Ectopic expression of miR-33b in HGC-27 and MGC-803 cells inhibited cell proliferation, migration and invasion, which might be due to miR-33b targeting oncogene c-Myc. Moreover, enhanced methylation level of the CpG island upstream of miR-33b in GC patients with down-regulated miR-33b was confirmed by methylation-specific PCR (MSP) amplification. Furthermore, re-introduction of miR-33b significantly suppressed tumorigenesis of GC cells in the nude mice. In conclusion, miR-33b acts as a tumor suppressor and hypermethylation of the CpG island upstream of miR-33b is responsible for its down-regulation in gastric cancer.

Figure 1. The expression analysis of miR-33b in GC tissues and its relationship with clinicopathological factors.

(A) miR-33b was detected in 150 GC patients by q-RT-PCR. Data was presented as log 1.5 of fold change of GC tissues relative to adjacent normal regions. (B) Relative miR-33b expression levels in primary GC tissues and adjacent normal regions. (C) The down-regulated miR-33b level in stage IV cases is significantly lower than that in stage I, II and III cases. (D) GC patients with metastasis have a lower level of miR-33b than those without metastasis (p < 0.01). All data are shown as mean ± SD.

Figure 2. Overexpression of miR-33b in GC cells inhibits cell proliferation, migration and invasion.

(A) The expression levels of miR-33b in GC cell lines (HGC-27, MGC-803, SGC-7901 and MKN-45) were abnormally down-regulated compared with four paired GC samples. “C” represents clinic GC tissue, “N” represents the matched adjacent normal tissue. (B) The expression levels of miR-33b were examined by real-time PCR after transfection with 50 nmol/L of miR-33b mimics or scramble or no transfection. (C) The cell growth of MGC-803 and HGC-27 cells at day 0, 1, 2, 3, 4 post transfection which was detected by CCK-8 assay. (D) MGC-803 and HGC-27 cells were not transfected or transfected with 50nmol/L of miR-33b mimic or scramble for 24 h, then wounds were made. The relative ratio of wound closure per field was shown. (E) MGC-803 and HGC-27 cells were not transfected or transfected with 50 nmol/L of miR-33b mimic or scramble for 24 h, and transwell invasion assay was performed. The relative ratio of invasive cells per field is shown. Magnification for identification of migration and invasion is ×10. Bar, 100 μm. All data are shown as mean ± SD. ^∗p < 0.05; ^∗∗p < 0.01, ^∗∗∗p < 0.001.

Figure 3. miR-33b targets c-Myc in GC cells and patients.

(A) Schematic representation of c-Myc 3′ UTRs showing putative miR-33b binding sites. (B) Western blot analysis of c-Myc expression in MGC-803 and HGC-27 cells transfected with scramble oligonucleotide or miR-33b mimics (Top). Gray value of c-Myc expression was also shown (Bottom). (C) Western blot analysis of c-Myc expression and relative level of miR-33b in 10 pairs of GC tissues. (D) Statistical expression of c-Myc in GC samples. (E) Expression of miR-33b was negatively correlated with c-Myc in GC sample.

Figure 4. Down-regulation of miR-33b in gastric cancer cells is associated with hypermethylation of miR-33b upstream region.

(A) Schematic illustration of the CpG islands upstream of miR-33b gene within the chr 17p11.2 segment. (B) Effect of 5-Aza-CdR and TSA on miR-33b expression in MGC-803 and HGC-27 gastric cancer cell lines. There is a 1.5-fold increase of miR-33b level after treatment of 5′-AZA (1 μM) also both AZA and TSA, a 1.2-fold increase after treatment of TSA (300 nM) in MGC-803 cells, a 3-fold increase after treatment of 5′-AZA (1 μM), a 2-fold increase after treatment of TSA (300 nM) and 2.5-fold increase after treatment of both AZA and TSA in HGC-27 cells. (C) The analysis of SREBF-1 expression in GC cells treated with AZA and/or TSA. There is an obviously increase of SREBF-1 level after treatment of 5′-AZA (1 μM) (p < 0.01) also both AZA and TSA (p < 0.01), a 8-fold increase after treatment of TSA (300 nM) (p < 0.05) in HGC-27 cells, a 12-fold increase after treatment of 5′-AZA (1 μM) (p < 0.01), a 8-fold increase after treatment of TSA (300 nM) (p < 0.05) and 6-fold increase after treatment of both AZA and TSA (p < 0.05) in MGC-803 cells. All data are shown as mean ± SD. ^∗p < 0.05; ^∗∗p < 0.01. (D) The methylation level of the CpG island 1 was decreased in GC cells treated with 5′-AZA, suggesting that CpG island 1 was hypermethylated in MGC-803 and HGC-27 cell lines while CpG island 2 and 3 didn’t have obvious changes in these cells. (E) CpG island 1 methylation analysis using qMSP in GC cells. Similar to the results of MSP, the methylation level was decreased in GC cells treated with 5′-AZA. M, methylated state; U, unmethylated state; Methylated control, Methyltransferase treated normal lymphocytes DNA.

Figure 5. The methylation analysis of miR-33b regulatory regions in GC cases.

(A) The methylation status in CpG island 1/2/3. The methylation degree of CpG island 1 in down-regulation group was obviously higher than that in up-regulation group (p < 0.01); The methylation degree of CpG island 2 in down-regulation group was obviously higher than that in up-regulation group (p < 0.05). The unmethylated state of CpG island 3 is higher than methylated state in this island. M, methylated state; U, unmethylated state. All data are shown as mean ± SD. ^∗p < 0.05; ^∗∗p < 0.01. (B) The methylation status of CpG island 1 analyzed by using qMSP. Similar to the results of MSP, the methylation degree in down-regulation group was higher than that in up-regulation group (p < 0.05). (C) Correlation analysis of methylation index with pM stage or pTNM in GC tissues. The methylation index in stage IV cases is higher than that in stage II and III cases (p < 0.05). The methylation index in GC patients with metastasis is also higher than those without metastasis (p < 0.05). All data are shown as mean ± SD. ^∗p < 0.05.

Figure 6. miR-33b suppresses gastric tumorigenicity in vivo.

(A) Photographs of the mice injected with scramble control and miR-33b mimics. (B) Tumors formed in the two group of nude mice. 5 × 10⁶ HGC-27 cells were subcutaneously injected into nude mice and miR-33b mimics (or scramble control) were directly into the tumor every 7 days. Tumors were harvested after 5 weeks. (C) Graph representing tumor volumes at the indicated days during the experiment for the miR-33b mimics group and scramble group. Five mice in each group. (D) Tumor weight averages between the scramble and miR-33b mimics groups at the end of the experiment (day 35). (E) H&E staining and immunohistochemistry analysis of Ki-67 expression in tumors from xenograft mice of the two group. (F) Nude mice were injected with HGC-27 cells infected with Lenti-33b or Lenti-scr through the lateral tail vein. Five weeks after injection, the mice were killed and the livers were dissected for microscopic histology. (G) The numbers of liver metastases in mice injected with Lenti-33b-infected HGC-27 cells were significantly lower than those in mice injected with Lenti-scr-infected cells. (H) Histological analysis of sections from livers of the mice injected with HGC27 cells treated by either Lenti-scr (control) or Lenti-33b. Images shown in the top-right panel represent magnified view of boxed region indicated in the panel. Data are mean ± SD and representative of three independent experiments. (^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001).