Unified metagenomic method for rapid detection of microorganisms in clinical samples

Abstract

Background: Clinical metagenomics involves the genomic sequencing of all microorganisms in clinical samples ideally after depletion of human DNA to increase sensitivity and reduce turnaround times. Current human DNA depletion methods preferentially preserve either DNA or RNA containing microbes, but not both simultaneously. Here we describe and present data using a practical and rapid mechanical host-depletion method allowing simultaneous detection of RNA and DNA microorganisms linked with nanopore sequencing.

Methods: The human cells from respiratory samples are lysed mechanically using 1.4 mm zirconium-silicate spheres and the human DNA is depleted using a nonspecific endonuclease. The RNA is converted to dsDNA to allow the simultaneous sequencing of DNA and RNA.

Results: The method decreases human DNA concentration by a median of eight Ct values while detecting a broad range of RNA & DNA viruses, bacteria, including atypical pathogens (Legionella, Chlamydia, Mycoplasma) and fungi (Candida, Pneumocystis, Aspergillus). The first automated reports are generated after 30 min sequencing from a 7 h end-to-end workflow. Sensitivity and specificity for bacterial detection are 90% and 100%, respectively, and viral detection are 92% and 100% after 2 h of sequencing. Prospective validation on 33 consecutive lower respiratory tract samples from ventilated patients with suspected pneumonia shows 60% concordance with routine testing, detection of additional pathogens in 21% of samples and pathogen genomic assembly achieve for 42% of viruses and 33% of bacteria.

Conclusions: Although further workflow refinement and validation on samples containing a broader range of pathogens is required, it holds promise as a clinically deployable workflow suitable for evaluation in routine microbiology laboratories.

Fig. 1. Metagenomics workflow.

The first step involves spinning the sample. Most of the human cells settle at the bottom, allowing for the collection of supernatant containing mainly microorganisms. The remaining human cells in the supernatant are lysed using mechanical disruption to release the DNA. A nonspecific endonuclease is added to digest the cell-free DNA and RNA present in the supernatant. DNA and RNA-containing microorganisms are extracted. After the extraction, RNA is converted into complementary DNA (cDNA) using the enzyme reverse transcriptase. The single-stranded cDNA is used as a template to synthesise a complementary strand, forming double-stranded DNA. The library preparation is performed using SQK-RPB004—Oxford Nanopore Technologies (ONT). Elements of this Figure are adapted with permission from ONT.

Fig. 2. Bacterial and Viral ROC curve analysis.

a ROC curves were constructed to establish the bacterial and viral reporting thresholds. True positive rate (sensitivity) against the false positive rate (1-specificity) for bacterial detection at different bacterial abundance thresholds (0.5%, 1%, 2%, 5%, and 10%) after 0.5 h, 2 h, and 24 h of sequencing. The y-axis represents the true positive rate, indicating how the metagenomics method detects actual cases of bacterial presence as confirmed by standard culture and molecular assays. The x-axis represents the false positive rate, reflecting the proportion of false positives among the negatives identified by the clinical assays. b This ROC curve focuses on viral read counts (1, 3, 5, and 10) after 0.5 h, 2 h, and 24 h of sequencing. Again, the y-axis shows the true positive rate, measuring how accurately the metagenomics detects viruses relative to a standard multiplex PCR assay. The x-axis shows the false positive rate. Points represent thresholds for the number of viral reads necessary to report a positive result.