The ultimate vision of this proposal is to develop a technology platform for the highly specific, rapid, and robust detection of cancer-associated DNA methylation biomarkers for diagnostics and research. Epigenetic alterations are known to be a crucial mode of regulation in cancer. DNA methylation in particular has been known to be dysregulated in cancer for decades and specific loci of DNA methylation have been sought as non-invasive biomarkers for cancer in blood, stool and other samples. However, a major barrier to progress has been the lack of highly sensitive, specific, and quantitative methods for detecting DNA methylation at specific sites in DNA. Nearly all established methods require bisulfite treatment of the DNA to convert unmethylated cytosines to uracil (leaving 5-methylcytosine, 5mC, intact), followed by DNA sequencing or quantitative PCR. However, bisulfite treatment is harsh and damages the DNA; obtaining a high efficiency of conversion is typically associated with degradation of >80% of the DNA into poorly amplified fragments. This, when combined with PCR-amplification-based methods, introduces limitations in the sensitivity of detecting methylated DNA at specific sites. Furthermore, the reduced sequence complexity of bisulfite-treated DNA? essentially reducing the four-letter genetic code to a three-letter one?increases the risk of spurious amplification and reduced specificity in quantitative PCR-based approaches. We here propose an approach to DNA methylation detection and quantification that is conceptually simple, yet takes advantage of sophisticated and elegant advances in single-molecule imaging science. The approach is based on using total internal reflection fluorescence microscopy to detect the repeated binding and release of sequence-specific fluorescently tagged probes to immobilized target DNA molecules on the surface of a glass slide. Following bisulfite conversion, this unique approach of repetitive probing (i.e., fingerprinting) of single molecules can distinguish between methylated and unmethylated target genes with exquisite specificity (>99.999%) and at the single-molecule level, enabling counting each specific biomarker molecule while avoiding the problem of spurious priming seen with PCR. In contrast to established PCR-based methods, no enzymatic manipulation of the analyte DNA is required with this approach, which is expected to improve sensitivity for highly fragmented input DNA. In order to validate this approach and assess its potential for application to DNA loci important in cancer diagnostics, we propose to develop and benchmark a direct single-molecule DNA methylation assay using kinetic fingerprinting with oligonucleotide probes against loci that are commonly hypermethylated in colorectal cancer, using both synthetic samples and patient blood specimens. By pursuing this work, we will maximize the likelihood of success in developing a transformative new approach for sensitive, specific, PCR amplification-free quantification of cancer-relevant DNA methylation at specific CpG loci.
Altered DNA methylation is a promising blood-based biomarker of several types of cancer, yet existing methods to quantify methylation require harsh treatment of the DNA with a chemical called bisulfite and amplification with the polymerase chain reaction (PCR), which leads to reduced accuracy for detecting rare tumor DNA. We will optimize our recently pioneered approach, single-molecule kinetic fingerprinting, for the detection of cancer-associated DNA methylation with high sensitivity and specificity while avoiding the errors associated with PCR. The ultimate goal of this study is to provide clinicians and researchers with a more rapid and accurate alternative to existing technologies for the earlier detection of colorectal and other cancers via abnormal methylation found on tumor DNA.
