UCLA

Recent advances in liquid biopsy analysis have gravitated towards the utilization of cell-free DNA (cfDNA) in biofluids for cancer screening and treatment guidance. Many non-mutation aspects of cell-free DNA, such as fragment size and methylation, have demonstrated promising clinical utility for cancer detection. However, the presenting populations of cfDNA are influenced by pre-analytical steps such as DNA extraction and library preparation. We hypothesized that conventional workflows excluded a substantial portion of short fragment cfDNA in plasma. In this thesis, we detail the development of a next-generation sequencing pipeline, "Broad Range Cell-free DNA Sequencing" (BRcfDNA-Seq), which combines low-molecular weight nucleic acid extraction with a single-stranded library preparation to circumvent fragment size and strandedness limitations of conventional sequencing. In plasma, using BRcfDNA-Seq revealed the presence of ~50nt ultrashort cell-free DNA (uscfDNA) in addition to 167bp double-stranded mononucleosomal cell-free DNA (mncfDNA). Molecular and genomic analysis showcased that uscfDNA is distinct from mncfDNA in that it is single-stranded, is enriched in regulatory elements such as promoters, exons, and introns, and contains substantial G-quadruplex sequences. We examine if these unique uscfDNA features could be used as potential biomarkers to differentiate between plasma samples from non-cancer and non-small cell lung carcinoma subjects (NSCLC). We observed significant differences in functional element peaks, fragmentomics, end-motif profiles, and G-quadruplex abundance between the uscfDNA of these cohorts. Next, we investigated the methylation characteristics of uscfDNA by introducing a novel method of appending single-stranded premethylated adapters to cfDNA fragments prior to bisulfite conversion, preventing degraded genomic DNA from artificially occluding shorter cfDNA. We observed that the CpG sites of uscfDNA fragments were 15% hypomethylated compared to mncfDNA. Using a deconvolution algorithm, we inferred that uscfDNA derives from eosinophils, neutrophils, and monocytes. Later, we showed that the methylation characteristics of uscfDNA can be used to distinguish NSCLC and non-cancer subjects through differentially methylated region analysis and hypermethylated patterns of promoters, 5' UTR, and exon elements. Collectively, these studies support the uniqueness of the uscfDNA population from mncfDNA. Both genomic and epigenetic characteristics of uscfDNA demonstrate its potential clinical utility as an additional biomarker for liquid biopsy for NSCLC detection.