A CRISPR Fingerprint of Pathogenic Candida auris Fungi Is Changing Precision Diagnostics

In recent years, the pathogenic yeast Candida auris has emerged as one of the most troubling threats inside hospitals and long-term care facilities. This fungus is not just another opportunistic pathogen. It spreads easily, survives for long periods on surfaces, and has shown an alarming ability to develop resistance to multiple antifungal drugs. Now, a new study from researchers at Harvard’s Wyss Institute introduces a CRISPR-based diagnostic approach that could significantly improve how quickly and accurately these infections are detected and managed.

Candida auris infections are especially dangerous for patients who are already medically vulnerable. Individuals undergoing chemotherapy, organ transplant recipients, people living with HIV, and patients who rely on invasive medical devices such as ventilators, feeding tubes, or catheters face the highest risks. Once the fungus enters the bloodstream or vital organs, it can trigger life-threatening conditions that resemble severe bacterial or viral infections.

Although antifungal medications exist, treatment has become increasingly complicated. Many strains of C. auris have developed antimicrobial resistance (AMR), meaning standard drugs may not work. In some cases, doctors must switch treatments after discovering resistance, while in the worst scenarios, infections can become nearly impossible to treat. This makes fast and precise diagnostics not just helpful, but critical.

Why Current Diagnostics Are Not Enough

Traditional diagnostic methods for Candida auris are far from ideal. Samples collected from patients often need to be shipped to one of a small number of specialized laboratories. The testing process can take several days to a full week, all while patients may require immediate treatment. These methods are also expensive, rely on complex laboratory equipment, and require highly trained personnel.

Even more concerning is that many existing tests provide only a simple yes-or-no answer. They struggle to quantify how much of the pathogen is present or to detect multiple strains with different resistance profiles within the same patient sample. This lack of detailed information limits a clinician’s ability to choose the most effective therapy.

Enter dSHERLOCK, a Digital CRISPR Diagnostic

To address these challenges, the Wyss Institute research team developed a next-generation diagnostic tool called dSHERLOCK, short for digital SHERLOCK. This system builds on earlier CRISPR-based diagnostic technologies but adds several critical improvements that make it both faster and more informative.

At its core, the system integrates CRISPR-mediated nucleic acid detection with single-molecule microarray technology. CRISPR allows the test to recognize specific genetic sequences from Candida auris with single-nucleotide precision, while the microarray platform enables thousands of individual reactions to occur simultaneously.

The result is a test that can detect the presence of C. auris from simple swab samples in about 20 minutes and accurately quantify the level of fungal colonization within 40 minutes. This is a dramatic improvement over existing workflows that can take days.

Detecting Antifungal Resistance in Real Time

One of the most powerful aspects of dSHERLOCK is its ability to identify resistance-causing mutations. By fine-tuning the CRISPR detection mechanism, the researchers designed assays that respond differently depending on whether specific resistance mutations are present.

Distinct fluorescent signals are generated at different rates depending on the genetic makeup of the fungal targets. These kinetic signatures allow the system to distinguish between strains resistant to commonly used antifungal drug classes, particularly azole and echinocandin medications.

This approach makes it possible to detect mixed populations of Candida auris within a single patient sample, something that most current diagnostics cannot do. For clinicians, this means a clearer picture of the resistance landscape and better guidance for selecting treatment strategies.

Machine Learning Turns Data into Clinical Insight

Each dSHERLOCK microarray contains roughly 18,000 individual compartments, many of which capture single target molecules. This generates a massive amount of fluorescent data that reflects not only whether the pathogen is present, but also how abundant it is and what resistance mutations it carries.

To make sense of this complexity, the team developed a machine learning-based analytical pipeline. This system interprets the evolving fluorescence patterns and translates them into results that can be easily understood by hospital staff. The output provides quantitative information on pathogen load as well as the presence and proportion of resistant strains.

Importantly, the researchers streamlined the assay into a one-pot reaction, simplifying the process and making it more practical for clinical settings.

Collaboration Driven by an Urgent Need

The project was sparked by a real-world public health crisis. After multiple outbreaks of treatment-resistant Candida auris in New York hospitals, the New York State Department of Health issued an urgent call to accelerate diagnostic development. The Wyss Institute team responded, receiving funding and collaborating closely with the Wadsworth Center Mycology Laboratory, which provided patient surveillance samples for validation.

This collaboration ensured that the technology was tested under realistic clinical conditions rather than purely theoretical ones.

Why Candida auris Is Such a Serious Threat

Candida auris stands out among fungal pathogens for several reasons. It can persist on hospital surfaces for weeks or even months, making infection control extremely difficult. It is often resistant to standard disinfectants, and outbreaks can spread rapidly within healthcare facilities.

Globally, C. auris has been reported on multiple continents, often emerging independently in different regions. This suggests that it is particularly adept at adapting to modern healthcare environments. Rapid diagnostics like dSHERLOCK could play a key role in containing outbreaks before they spiral out of control.

A Platform with Broader Potential

While this study focuses on Candida auris, the implications go much further. The researchers emphasize that the dSHERLOCK platform can be adapted to other pathogens by changing the CRISPR detection components. This opens the door to rapid, quantitative diagnostics for a wide range of infectious diseases, including those with emerging drug resistance.

By combining CRISPR engineering, single-molecule detection, and machine learning, the study demonstrates how multiple cutting-edge technologies can converge to meet urgent clinical needs.

A Step Forward for Precision Diagnostics

The development of dSHERLOCK represents a significant advance in precision infectious disease diagnostics. It delivers speed, sensitivity, and detailed genetic information in a single assay, addressing many of the shortcomings of current diagnostic methods.

As antimicrobial resistance continues to rise and hospital-acquired infections remain a major concern, tools like this could reshape how clinicians detect, monitor, and respond to dangerous pathogens. While further work will be needed to scale and deploy the technology widely, this study provides a compelling blueprint for the future of rapid diagnostics.

Research paper:
https://www.nature.com/articles/s41551-025-01597-0