Enriching and Base-Resolution Profiling of 5-Methylcytosine in Degraded Clinical Samples

PROJECT SUMMARY The epigenome, centered around 5-methylcytosine (5mC), carries ample molecular signatures for understanding, detecting, and tracking tumorigenesis. However, 5mC profiling in clinical samples, such as formalin-fixed, paraffin-embedded (FFPE) tissues and cell-free DNA (cfDNA), is hindered by their limited quantity and degraded DNA. To address these challenges, this proposal describes development and application of the first glycosylase-based epigenome-sequencing platform. With superior sensitivity and fidelity, glycosylase- implemented epigenome sequencing (Glimp-seq) is poised to uncover the molecular underpinnings of tumorigenesis in low-volume degraded samples. Glimp-seq harnesses cascade enzymatic reactions involved in mammalian active demethylation—5mC oxidization by ten-eleven translocation dioxygenase (TET) and 5-carboxylcytosine (5caC) excision by thymine DNA glycosylase (TDG)—to convert 5mC into abasic sites. These abasic sites are subsequently functionalized by a thymine mimic to facilitate 5mC-specific enrichment and direct mutation detection (5mC-to-T). Glimp-seq represents a major advancement over current epigenome profiling methods, including bisulfite sequencing (BS- seq), enzymatic methyl sequencing (EM-seq), and TET-assisted pyridine borane sequencing (TAPS), by employing gentle, yet efficient, enzymatic conversions that preserve DNA integrity. As the specificity of the cascade TET/TDG reactions has been extensively optimized during natural evolution, Glimp-seq achieves unparalleled specificity for 5mC, reducing the risk of false positive detection. Unlike existing methods, Glimp-seq uniquely supports sequence-independent 5mC-specific enrichment, thereby increasing detection sensitivity. The resulting superior sensitivity and fidelity make Glimp-seq particularly suited for 5mC profiling in degraded clinical samples. In Aim 1, we will develop and rigorously benchmark Glimp-seq, using both synthetic and genomic DNA samples. We will compare Glimp-seq with BS-seq, EM-seq, and TAPS, and demonstrate its superiority using quantitative metrics. In Aim 2, we will apply Glimp-seq to two sets of clinical samples: ~140 FFPE cancer tissues from patients of prenatal exposure to diethylstilbestrol (known as DES daughters), which provides a unique opportunity to investigate synthetic estrogen-induced epigenetic inheritance; and cfDNA from 50 colorectal cancer patients, to showcase non-invasive detection of cancer-associated 5mC signatures. Successful implementation of Glimp-seq will transform the clinical practice of epigenome profiling by facilitating mechanistic studies using FFPE samples and by expediting early cancer diagnosis through cfDNA analysis. By freely distributing Glimp-seq reagents and protocols, we aim to support cancer researchers both locally and globally, fostering continuous advancements in understanding, detecting, and tracking cancer.