Epigenetic mechanisms involved in differential MDR1 mRNA expression between gastric and colon cancer cell lines and rationales for clinical chemotherapy

Abstract

Background: The membrane transporters such as P-glycoprotein (Pgp), the MDR1 gene product, are one of causes of treatment failure in cancer patients. In this study, the epigenetic mechanisms involved in differential MDR1 mRNA expression were compared between 10 gastric and 9 colon cancer cell lines.

Methods: The MDR1 mRNA levels were determined using PCR and real-time PCR assays after reverse transcription. Cytotoxicity was performed using the MTT assay. Methylation status was explored by quantification PCR-based methylation and bisulfite DNA sequencing analyses.

Results: The MDR1 mRNA levels obtained by 35 cycles of RT-PCR in gastric cancer cells were just comparable to those obtained by 22 cycles of RT-PCR in colon cancer cells. Real-time RT-PCR analysis revealed that MDR1 mRNA was not detected in the 10 gastric cancer cell lines but variable MDR1 mRNA levels in 7 of 9 colon cancer cell lines except the SNU-C5 and HT-29 cells. MTT assay showed that Pgp inhibitors such as cyclosporine A, verapamil and PSC833 sensitized Colo320HSR (colon, highest MDR1 expression) but not SNU-668 (gastric, highest) and SNU-C5 (gastric, no expression) to paclitaxel. Quantification PCR-based methylation analysis revealed that 90% of gastric cancer cells, and 33% of colon cancer cells were methylated, which were completely matched with the results obtained by bisulfite DNA sequencing analysis. 5-aza-2'-deoxcytidine (5AC, a DNA methyltransferase inhibitor) increased the MDR1 mRNA levels in 60% of gastric cells, and in 11% of colon cancer cells. Trichostatin A (TSA, histone deacetylase inhibitor) increased the MDR1 mRNA levels in 70% of gastric cancer cells and 55% of colon cancer cells. The combined treatment of 5AC with TSA increased the MDR1 mRNA levels additively in 20% of gastric cancer cells, but synergistically in 40% of gastric and 11% of colon cancer cells.

Conclusion: These results indicate that the MDR1 mRNA levels in gastric cancer cells are significantly lower than those in colon cancer cells, which is at least in part due to different epigenetic regulations such as DNA methylation and/or histone deacetylation. These results can provide a better understanding of the efficacy of combined chemotherapy as well as their oral bioavailability.

Figure 1

MDR1 mRNA expression in gastric cancer cell lines. The level of MDR1 mRNA expression was determined by RT-PCR, and normalized by that of mRNA β-actin, which was used as a control for RNA. The cDNA reverse-transcribed from the mRNA was amplified separately with each primer pair for MDR1 and β-actin genes. Aliquots of each PCR reaction mixture were separated on 7% polyacrylamide gel. The gel was dried and exposed on X-ray film overnight.

Figure 2

MDR1 mRNA expression in the colon cancer cell lines. The same methodology reported in Figure 1 was used.

Figure 3

MDR1 mRNA expression in gastric and colon cancer cell lines. The level of MDR1 mRNA expression was determined by real-time RT-PCR, and normalized by that of mRNA β-actin, which was used as a control for RNA. The cDNA reverse-transcribed from the mRNA was amplified separately with each primer pair for MDR1 and β-actin genes.

Figure 4

Comparison of Pgp expression and function in gastric and colon cancer cell lines. (A) Comparative sensitivity of Colo320HSR (colon, highest), SNU-668 (gastric, highest) and SNU-C5 (colon, no expression) to paclitaxel; (B) Effects of Pgp inhibitors on the sensitivity of Colo320HSR, SNU-668 and SNU-C5 to paclitaxel (IC10 concentration; 50 μM, 0.3 nM and 0.5 nM, respectively). Sensitivity to paclitaxel was determined using MTT assay in the presence or absence of the Pgp inhibitors (cyclosporin A, verapamil and PSC833 of 0.8 μM each). *, P <0.05 versus the control.

Figure 5

The quantification PCR-based methylation analysis of gastric and colon cancer cell lines. (A) CpG sites and Hpa II/Msp I sites in the human MDR1 promoter region. Top: The CpG sites are represented by the short vertical bars. The positions of exons 1 and 2 are indicated as closed boxes. The position corresponding to these fragments are indicated. Middle: Hpa II/Msp I recognition sites are represented by short vertical bars. Bottom, PCR primers used in methylation analysis. (B) Representative methylation status of the MDR1 promoter region by quantification PCR-based methylation analysis in SNU-5 (gastric) and HT-29 (colon). 1: MN, Never-methylated Hpa II/Msp I site at the triosephosphate isomerase gene promoter region (negative control) (240 bp); 2: MC, the positive control primer pair (240 bp); 3: MS1, Hpa II/Msp I site 1 (121 bp); 4: MS2, Hpa II/Msp I site 2 (206 bp).

Figure 6

Activation of MDR1 mRNA expression by 5AC and/or TSA in gastric cancer cells. The expression level is reported as the ratio of MDR1/β-actin. The total RNA was isolated after treatment with 2.5 μM 5AC for 96 hr and/or 100 ng/ml TSA for 48 hr. RT-PCR was performed using the same methodology reported in Figure 1.

Figure 7

Activation of MDR1 mRNA expression by 5AC and/or TSA in various colon cancer cell lines. RT-PCR assay after treating the cells with 5AC and/or TSA using the same method described in Figure 6. The MDR1/β-actin ratio obtained through 35-cycle PCR after TSA in combination with 5AC, and alone in SNU-C5 and HT-29 expressing no MDR1 mRNA, respectively, was omitted in the histogram.