Hospital fungus copies its genome four times over and grows tolerant of a key drug

A study in Nature Communications reports that the hospital fungus Candida auris can grow more tolerant of a frontline antifungal drug while copying its whole genome, four times over or more ^[1].

The paper, published October 15, 2026, comes from researchers at the University of Exeter and City St George's, University of London ^[1]. It describes a way a fungal population can slip past treatment even when the usual lab test does not call it resistant.

That makes it a warning, not yet a verdict. The evidence comes from two patients and from experiments in laboratory culture, and the authors do not show that this sequence of changes caused treatment to fail in either person ^[1].

Two patients, two infections that would not clear

Candida auris is a yeast that infects people who are already very ill. The study describes it as a World Health Organization critical-priority fungal pathogen and puts the mortality of invasive infection at up to 45% ^[1]. The US Centers for Disease Control and Prevention adds that it can colonize patients without causing symptoms and persist on surfaces, which helps it move through health care settings ^[2].

The researchers followed serial isolates from two patients with bloodstream infections in intensive care units at St George's and King's College hospitals in London. In both, the infection did not clear during treatment with echinocandins, the class of antifungal drugs that is a common first choice because most C. auris isolates resist fluconazole ^[1].

The obvious suspect was missing. Echinocandin resistance is usually tied to mutations in two genes, FKS1 and FKS2, and neither patient's isolates carried them. Routine susceptibility tests did not classify the isolates as resistant either, even though the infections persisted or came back ^[1].

What the fungus did in the lab

To see what else might be going on, the team recreated some of the variation in culture. They exposed clinical isolates to the echinocandin anidulafungin at eight micrograms per millilitre, a level above the minimum concentration that stops growth ^[1][3].

The survivors did not look alike. The cells produced different forms, including small-colony and large-colony variants ^[1][3]. Both had higher minimum inhibitory concentrations, the lowest dose of a drug that halts growth, so they could take more of it ^[1][3].

Both were also polyploid, meaning they carried more than the usual two sets of chromosomes. Some had four or more copies of the genome ^[1][3]. A fungus that copies everything, extra chromosomes and all, gets a larger and more varied pool of genetic material to draw on under pressure.

The large-colony variants carried more mutations in a gene called CRZ1 than the original clinical isolates or the small-colony variants did. CRZ1 sits in a stress-tolerance pathway linked to calcineurin, a protein the cell uses to cope with threats. The mutations were associated with further increases in drug tolerance ^[1][3].

The team also saw changes in the cell wall, specifically in the content of beta-1,3-glucan, the structural sugar that echinocandins attack. That suggests the cell wall was being remodeled alongside the genomic changes ^[1][3].

Why a clean test result may not mean much

The finding matters because of what it says about testing. A susceptibility test reads how an isolate behaves at the start. If a small subpopulation of cells is already copying its genome and picking up CRZ1 mutations under drug exposure, a test run on the starting isolate may not show it.

That reading is the paper's logic, and it is a hypothesis about a mechanism, not a measured rate in patients. The study does not say how often polyploid cells or CRZ1 mutations arise in people receiving echinocandins beyond these two cases ^[1].

Outcomes in the two patients were grim, though the paper does not tie them to the fungus. Both later received treatment based on amphotericin B, another antifungal, in combination with other drugs, and both died. Neither death was directly attributed to the C. auris bloodstream infection ^[1].

The researchers point to combination antifungal treatment as one avenue for further work, but the experiments do not yet show whether it can prevent or suppress the tolerant cells ^[1].

The bigger story, and where it stops

The study also notes that six C. auris clades, or distinct genetic lineages, emerged near-simultaneously across the world ^[1]. It is tempting to ask whether a trick like genome copying helps explain that. The study does not show it, and it would be wrong to say it does ^[1].

The scale of the problem is not in doubt. In the United States alone, CDC surveillance counted 13,507 clinical cases and 27,853 screening cases, which find colonized patients, during 2022 to 2024, with reported counts rising each year ^[2]. That is US data and not a global estimate.

What the London work adds is a concrete place to look. If tolerance can grow inside a patient from cells that a standard test never flagged, then "susceptible" on a lab report is a weaker promise than it sounds, in a disease where the patients have little margin.

The next questions are practical. Researchers will want to know how common these polyploid cells and CRZ1 mutations are across more patients, whether they cause clinically meaningful treatment failure or merely mark a wider process of adaptation, and whether routine testing could catch the evolving subpopulations before an infection turns persistent ^[1].