用于癌症甲基化生物标志物绝对定量的无参照PCR检测方法

Background: DNA m5C methylation is a valuable resource for disease biomarkers widely applied in diagnosis. However, current qPCR methods rely on reference assays of genes or repetitive elements (e.g., ACTB, ALU-C4) for data normalization, which may bring inaccuracy and false negatives. To overcome these limitations, we developed a reference-free qPCR technology, MeRFPCR, that enabled accurate and absolute quantification of m5C methylation without external normalization. Methods: MeRFPCR utilized uncomplementary sense and antisense DNA strands yielded from C-to-T conversion in bisulfite treatment, allowing the design of two sets of primers and fluorescent probes targeting the methylated CpGs in the sense strand and the adjacent CpG-free region in the antisense strand. An absolute methylation level was quantified as the ratio of the methylated signal to the CpG-free signal following PCR amplification, eliminating the need for external reference normalization. We evaluated specificity, sensitivity, and reproducibility across multiple loci, and benchmarked MeRFPCR against quantitative methylation-specific PCR (qMSP) with external reference normalization and bisulfite pyrosequencing. The assay was further applied to detect cancer methylation biomarkers in both tissue and liquid biopsy samples. Results: MeRFPCR preserved robustness under challenging conditions, yielding reproducible methylation percentages across technical replicates (CV <5%) with inputs as low as 2 ng and fragment sizes ranging from 80 to 200 bp. In 20 matched tumor and adjacent normal tissues from colorectal cancer patients, SEPT9 methylation measured by MeRFPCR showed excellent concordance with pyrosequencing (R² = 0.97) and distinguished all tumors from normal tissues. In contrast, qMSP showed lower concordance (R² = 0.73) and failed to resolve tumor-normal differences in several cases (sensitivity 90%). Of note, the workflow of MeRFPCR is faster, easily accessible and more cost-effective compared with pyrosequencing. In addition, MeRFPCR accurately detected methylated SEPT9 in circulating cell-free DNA from as little as 300 µL of plasma in metastatic CRC patients. These findings support the technical accuracy and advantages of MeRFPCR resulting from the reference-free methods, highlighting its specific diagnostic potential in compromised samples and across laboratory settings where reference-based assays are limited. Conclusion: The MeRFPCR technique is a fast, accurate, cost-effective, and reference-free method for absolute quantification of DNA methylation. It holds promise for reducing batch effects across platforms and improving detection in challenging clinical samples, including liquid biopsy and stool DNA. These findings highlight the potential of MeRFPCR as a basic and valuable tool for facilitating noninvasive detection of cancer methylation biomarkers and broader epigenetic applications.