Characterizing the Molecular Composition and Diagnostic Potential of Mycobacterium Tuberculosis Urinary Cell-free DNA Using Next-generation Sequencing

Background: Urine cell-free DNA (cfDNA) is an attractive target for diagnosing pulmonary Mycobacterium tuberculosis (MTB) infection, but has not been thoroughly characterized as a biomarker.

Methods: This study was performed to investigate the size and composition of urine cfDNA from tuberculosis (TB) patients with minimal bias using next-generation sequencing (NGS). A combination of DNA extraction and single-stranded sequence library preparation methods demonstrated to recover short, highly degraded cfDNA fragments was employed. Urine cfDNA from 10 HIV-positive patients with pulmonary TB and two MTB-negative controls was examined.

Results: MTB-derived cfDNA was identifiable by NGS from all MTB-positive patients and was absent from negative controls. MTB cfDNA was significantly shorter than human cfDNA, with median fragment lengths of ≤19-52 bp and 42-92 bp, respectively. MTB cfDNA abundance increased exponentially with decreased fragment length, having a peak fragment length of ≤19 bp in most samples. In addition, we identified a larger fraction of short human genomic cfDNA, ranging from 29 to 53 bp, than previously reported. Urine cfDNA fragments spanned the MTB genome with relative uniformity, but nucleic acids derived from multicopy elements were proportionately over-represented.

Conclusions: TB urine cfDNA is a potentially powerful biomarker but is highly fragmented, necessitating special procedures to maximize its recovery and detection.