Hypermethylation at 45S rDNA promoter in cancers

Abstract

The ribosomal genes (rDNA genes) encode 47S rRNA which accounts for up to 80% of all cellular RNA. At any given time, no more than 50% of rDNA genes are actively transcribed, and the other half is silent by forming heterochromatin structures through DNA methylation. In cancer cells, upregulation of ribosome biogenesis has been recognized as a hallmark feature, thus, the reduced methylation of rDNA promoter has been thought to support conformational changes of chromatin accessibility and the subsequent increase in rDNA transcription. However, an increase in the heterochromatin state through rDNA hypermethylation can be a protective mechanism teetering on the brink of a threshold where cancer cells rarely successfully proliferate. Hence, clarifying hypo- or hypermethylation of rDNA will unravel its additional cellular functions, including organization of genome architecture and regulation of gene expression, in response to growth signaling, cellular stressors, and carcinogenesis. Using the bisulfite-based quantitative real-time methylation-specific PCR (qMSP) method after ensuring unbiased amplification and complete bisulfite conversion of the minuscule DNA amount of 1 ng, we established that the rDNA promoter was significantly hypermethylated in 107 breast, 65 lung, and 135 colon tumour tissue samples (46.81%, 51.02% and 96.60%, respectively) as compared with their corresponding adjacent normal samples (26.84%, 38.26% and 77.52%, respectively; p < 0.0001). An excessive DNA input of 1 μg resulted in double-stranded rDNA remaining unconverted even after bisulfite conversion, hence the dramatic drop in the single-stranded DNA that strictly required for bisulfite conversion, and leading to an underestimation of rDNA promoter methylation, in other words, a faulty hypomethylation status of the rDNA promoter. Our results are in line with the hypothesis that an increase in rDNA methylation is a natural pathway protecting rDNA repeats that are extremely sensitive to DNA damage in cancer cells.

Fig 1. Evaluation of PCR amplification bias and PCR amplification efficiency.

(A) Control samples with defined methylation levels ranging from 0% to 100% were used as templates in qPCR. Standard curves obtained by plotting the CT values against the methylation level for the MSP primer set (rDNA Me) were linear, and that for the MIP primer set (Ref) remained constant throughout. (B) A serial dilution of human methylated DNA ranging from 5 pg to 1000 pg was bisulfite-treated and used as the template for qPCR. The CT values have a strong linear relationship with the DNA input for both primer sets (R² > 0.99); however, the amplification efficiency (E value) of the reference amplicon strongly differs from that of the rDNA amplicon (24% difference). (C) The ΔCt value, calculated by subtracting the CT value of the MSP primer set from that of the MIP primer set for each sample, proportionally increases as the DNA input decreases, indicating the Pfaffl formula to be suitable for relative quantification. (D) Control samples with defined methylation levels ranging from 0% to 100% were used as templates in qPCR. Using the Pfaffl formula, recovered rDNA methylation levels were identical to the input methylation levels, demonstrating the ability to accurately identify rDNA methylation levels using the designed primer sets. Simple linear regression was used in statistical analysis. Number of observations for each assay was ≥ 3.

Fig 2. An excessive DNA input resulted in enormous variation in rDNA methylation level and remaining DNA unconverted.

rDNA methylation levels enormously varied between different DNA input amounts of human genomic DNA (hDNA) (1000 ng and 1 ng) (A) and fully methylated human DNA (hmDNA) (1000 ng, 50 ng, 5 ng, and 1 ng) (B) for bisulfite conversion. (C) PCR products were amplified with the primers specific to the native rDNA sequences. The DNA amount of 1 μg from hmDNA and hDNA was treated with bisulfite and used as templates for PCR. The DNA amount of 1 μg from hmDNA (1) and hDNA (2) were bisulfite treated, digested with restriction enzymes and used as templates for PCR. The DNA amount of 100 pg from hDNA (3) and hmDNA (4) were digested with restriction enzymes and used as templates for positive control. Restriction enzyme reactions (RE1-HpaII, RE2-MspI) and PCR reactions without DNA (NC1, NC2) were used as templates for negative control. The Mann-Whitney U test (A) and the unpaired t-test (B) were used in statistical analysis. (ns) nonsignificant; (*) p < 0.05; (***) p < 0.001; (****) p < 0.0001.

Fig 3. rDNA methylation level in cancer.

rDNA methylation profile in tumour samples of breast cancer (BC), lung cancer (LC) and colorectal cancer (CRC) as compared with their corresponding adjacent tissue samples (AD). Methylation assessments were performed on 1 ng of DNA converted by bisulfite. The Wilcoxon matched-pairs signed-rank test was used in statistical analysis. (****) p < 0.0001.