Author Date

2024-03-15

Degree Name

Department

Physiology and Developmental Biology

College

Life Sciences

Defense Date

2024-02-29

Publication Date

2024-03-15

First Faculty Advisor

Timothy Jenkins

First Faculty Reader

Jonathon Hill

Honors Coordinator

Roy Silcox

Keywords

Cell-free DNA, Bisulfite-conversion, Enzymatic-conversion, DNA Fragmentation, Seminal cfDNA, Femto Pulse

Abstract

This project aims to investigate the DNA fragmentation produced by enzymatic bisulfite conversion compared to DNA fragmentation produced by traditional chemical bisulfite conversion techniques for the analysis of cytosine methylation in seminal cell-free DNA (cfDNA). Traditional chemical bisulfite (HSO₃^-) conversion methods are known to introduce biases and damage DNA, potentially compromising methylation analysis accuracy. In contrast, enzymatic conversion, utilizing biological molecules, is hypothesized to offer more precise visualization with reduced fragmentation (Lambert et al., 2019). While previous studies have analyzed methylation conversion efficiency in genomic DNA, this project specifically examines how these conversion techniques impact cfDNA.

20 samples of seminal cfDNA were bisulfite-converted, 20 samples of cfDNA were enzymatically converted, and 20 samples of cfDNA were not converted and used as a control. Agilent Femto Pulse DNA analysis was used to compute DQN (DNA quality number) scores, length of sequences, and fragmentation values within each sample. Using these data, bioinformatic analyses were ran on DNA fragmentation using the R programming language. Results indicate that bisulfite DNA conversion contributes to significantly more fragmented sequences than enzymatic conversion in cfDNA. ANOVA and Tukey’s test analyses were conducted on DQN scores, revealing a statistically significant difference in DQN values between bisulfite-converted cfDNA and enzymatically converted cfDNA samples (p = 0.048). Analyses indicate that the average base pair length found in enzymatic-converted cfDNA were significantly longer (p=0.0044) than samples that were bisulfite converted. These outcomes suggest that enzymatic conversion decreases bias when converting cfDNA samples compared to using alternative methylation conversion techniques.