Reduction of Human DNA Contamination in Clinical Cerebrospinal Fluid Specimens Improves the Sensitivity of Metagenomic Next-Generation Sequencing.

Abstract

Metagenomics next-generation sequencing (mNGS) is increasingly available for the detection of obscure infectious diseases of the central nervous system. However, human DNA contamination from elevated white cells, one of the characteristic cerebrospinal fluid (CSF) features in meningitis patients, greatly reduces the sensitivity of mNGS in the pathogen detection. Currently, effective approaches to selectively reduce host DNA contamination from clinical CSF samples are still lacking. In this study, a total of 20 meningitis patients were enrolled, including 10 definitively diagnosed tuberculous meningitis (TBM) and 10 definite cryptococcal meningitis (CM) cases. To evaluate the effect of reduced human DNA in the sensitivity of mNGS detection, three specimen-processing protocols were performed: (i) To remove human DNA, saponin, a nonionic surfactant, was used to selectively lyse white cells in CSF followed by DNase treatment prior to the extraction of DNA; (ii) to reduce host DNA, CSF was centrifuged to remove human cells, and the supernatant was collected for DNA extraction; and (iii) DNA extraction from the unprocessed specimens was set as the control. We found that saponin processing significantly elevated the NGS unique reads for Cryptococcus (P < 0.01) compared with the control but had no effects for Mycobacterium tuberculosis (P > 0.05). However, detection of centrifuged supernatants improved the NGS unique reads for both TBM and CM compared with controls (P < 0.01). Our results demonstrate that the use of mNGS of centrifuged supernatants from clinical CSF samples in patients with TBM and CM is a simple and effective method to improve the sensitivity of pathogen detection.