Image created by Dr. Michael J. Miller
Bloodstream infections are a persistent threat to hospitalized patients, and those caused by fungi are particularly dangerous. Candida species are responsible for most invasive fungal infections worldwide and rank among the leading causes of hospital-acquired bloodstream infections overall. Because different fungal species can respond differently to antifungal drugs, doctors strive to correctly identify the fungus responsible for a given infection before initiating treatment.
However, conventional diagnosis relies on blood cultures, in which a patient’s blood is incubated until microorganisms grow to detectable levels. This process, together with subsequent identification tests, can take several days to complete. During this waiting period, clinicians often treat patients with broad antifungal drugs or wait longer than ideal to start more targeted therapies.
To address this challenge, a research team led by Professor Hiroki Takahashi from the Medical Mycology Research Center, Chiba University, Japan, developed a workflow to more rapidly identify fungal pathogens. Their study, published online in the journal Microbiology Spectrum, was co-authored by Dr. Isato Yoshioka, Dr. Momotaka Uchida, Professor Akira Watanabe, and Dr. Takashi Yaguchi, all from Chiba University.
The proposed workflow comprises three main steps, performed on a blood culture sample collected while it is still incubating and before the automated system flags it as positive. First, the researchers selectively break down human cells and then degrade human DNA using an enzyme called benzonase, without affecting fungal and bacterial DNA. This greatly increases the proportion of microbial DNA versus human DNA in the sample. Afterwards, the researchers use PCR-based whole-genome amplification to make many copies of DNA fragments from across the genomes present in the sample, generating enough genetic material for sequencing.
The amplified DNA is then loaded onto a portable device for nanopore sequencing, a technique that determines the sequence of DNA molecules as they pass through nanoscale pores. Unlike sequencing approaches that require a completed run before results can be assessed, nanopore sequencing generates DNA sequence data in real time as the run progresses. The resulting sequences are then investigated against a custom-built reference database containing genetic information from a wide range of microorganisms, including fungal and bacterial pathogens associated with bloodstream infections. “I have a background in sequence analysis and genomics, and I saw an opportunity to apply my expertise to the important clinical challenge of fungal infections,” explains Prof. Takahashi. “I was particularly motivated by the possibility of using modern genomic technologies to improve our understanding, diagnosis, and ultimately treatment of these infections.”
In tests using 48 clinical blood culture samples representing eight fungal species, the proposed workflow achieved species-level identification within approximately seven hours with remarkably high accuracy. Notably, the researchers also found that the method could identify a range of fungal pathogens, such as Candida albicans, Nakaseomyces glabratus, Candida parapsilosis, Candida tropicalis, and Cryptococcus neoformans. It even detected mixed infections in some samples, involving those with two fungal species or both fungi and bacteria.
A key feature of this novel approach is its speed, as it can identify fungal pathogens directly from blood culture samples before the cultures turn positive in conventional culture systems. “Our method may enable clinicians to initiate appropriate antifungal treatment earlier, potentially improving outcomes for patients with life-threatening fungal bloodstream infections,” remarks Prof. Takahashi. Moreover, because the workflow captures the full genome of the pathogen, it can also detect genetic variants in genes associated with drug resistance.
The researchers note that further work will be needed to determine the best timing for collecting samples during incubation and to improve detection when heavy bacterial growth masks fungal signals in mixed infections. With further validation, their strategy could contribute to the development of rapid diagnostic procedures for life-threatening fungal infections, reducing delays in effective therapy and ultimately improving care for affected patients.
Reference
Yoshioka I, Uchida M, Kusuya Y, Shibata S, Arai T, Yaguchi T, Watanabe A, Takahashi H. 2026. Random PCR-based nanopore whole-genome sequencing enables pre-positivity detection of fungal bloodstream infections. Microbiol Spectr 14:e00478-26. https://doi.org/10.1128/spectrum.00478-26
Abstract
Fungal bloodstream infections (BSIs), particularly those caused by Candida species, represent a major clinical challenge associated with substantial morbidity and mortality. Timely initiation of appropriate antifungal therapy is critical for patient survival; however, conventional diagnostic methods rely on blood culture positivity and downstream identification procedures, resulting in delays that limit early, targeted treatment. In this study, we developed and validated an optimized workflow for rapid identification of fungal pathogens directly from blood culture samples collected prior to positivity in automated blood culture systems using real-time nanopore sequencing. The workflow consists of selective depletion of host genomic DNA, microbial DNA extraction, and species identification via random PCR-based whole-genome amplification, followed by sequencing on the MinION platform and real-time taxonomic classification using a custom database. Host DNA was effectively removed using saponin-mediated cell lysis and benzonase treatment, enabling enrichment of microbial DNA. Using this approach, species-level identification was achieved within 1.5 h from sequencing initiation, and accurate identification was possible with as few as 4,000 sequencing reads. The total turnaround time of the workflow was approximately 7 h. Validation using 48 clinical blood culture samples demonstrated high concordance with routine clinical diagnoses, and the method reliably detected both monomicrobial and polymicrobial infections, including fungal–fungal and fungal–bacterial co-infections. Genome-wide sequencing data enabled detection of candidate resistance-associated variants in clinically relevant genes. Overall, this workflow provides a clinically actionable solution for early diagnosis of fungal BSIs, with the potential to reduce diagnostic delays, inform antifungal selection, and improve patient outcomes.