CRISPR Fingers Drug-Resistant Microbes in a 'FLASH'
May 21, 2019
FOR IMMEDIATE RELEASE
SAN FRANCISCO — May 21, 2019 — A research team led by scientists at UC San Francisco and the Chan Zuckerberg Biohub has developed a new CRISPR-based diagnostic tool, dubbed FLASH, that can rapidly identify any drug-resistant microbes that may be present in body fluids or other patient-derived samples, even when the populations of those pathogens are so small that they elude standard detection methods.
Described in an open-access paper published May 22 in Nucleic Acids Research, FLASH — short for “Finding Low Abundance Sequences by Hybridization” — can provide greater amounts of clinically actionable data more rapidly than the current crop of technologies used to diagnose the cause of, and best course of treatment for, a given patient’s infection.
Because FLASH is relatively easy to use and inexpensive to implement, it may one day make tools for microbial detection based on gene sequencing available to medical professionals who now lack access to these approaches, said senior author Emily Crawford, Ph.D., a scientist at the CZ Biohub Infectious Disease Initiative and adjunct assistant professor of microbiology and immunology at UCSF.
For decades, the only option available to clinicians who wanted to identify the infectious agents afflicting their patients was culturing — growing microbes derived from patient samples in Petri dishes. Even under ideal circumstances, it’s a slow process, and when working to halt the progress of an infection and curb the spread of drug-resistant pathogens, time is of the essence. At best, the culture-based methods that are still standard at many hospitals can identify drug-resistant bugs in 48 to 72 hours. But some microbes grow extremely slowly — for example, testing for drug-resistant tuberculosis can take more than six weeks — and many microbes can’t be cultured at all.
In recent years, a technique known as metagenomic sequencing has begun to supplement culture-based detection methods in research settings. This approach involves sequencing all the DNA found in a patient-derived sample — which is usually a mixture of human DNA and the DNA of the various microbes, helpful or harmful, that can be found in the human body — and then using powerful computational techniques to sort the results by species.
Metagenomic sequencing can even detect microbial DNA sequences that confer resistance to antibiotics and other antimicrobial drugs. If assembled early enough, this information can guide treatment, improve patient outcomes and minimize the spread of antimicrobial-resistant (AMR) pathogens.
“Metagenomic sequencing is a game changer for infectious disease diagnostics, giving clinicians the ability to search for all pathogens at once rather than taking the typical ‘one test, one bug’ approach,” Crawford said. “But this approach can be expensive and requires advanced computational resources that are not yet widely available to medical professionals. FLASH has the potential to make sequencing-based AMR gene profiling accessible to the growing number of hospital labs and public health departments around the globe that now have some access to high-throughput sequencing.”
In contrast to standard metagenomic sequencing — which sequences all DNA, whether human or microbial, in a particular sample — FLASH limits its scope to clinically informative AMR regions. The overwhelming majority of a given microbe’s genome, which provides no diagnostic or clinically actionable information, can then be excluded from sequencing, drastically reducing the amount of time and computational power needed to analyze the resulting data.
To create these pared-down sequences, FLASH employs the CRISPR-Cas9 gene-editing system to isolate DNA sequences that are known to confer AMR. The CRISPR complex is loaded with guide RNAs that target known AMR sequences. Then, Cas9 — the “scissors” part of the complex — cuts the DNA on either side of the target, which separates AMR sequences from the rest of the microbial genome. Once the informative DNA fragments have been sequenced, the data can help clinicians identify the pathogens that are present and determine which antimicrobial drugs may vanquish the pathogen and which are likely to be powerless to stop them.
To demonstrate FLASH’s effectiveness, the researchers used the tool to detect and characterize drug-resistant bacteria from two different sources: bacteria grown in the lab and bacteria found in respiratory fluid obtained from four hospitalized patients. When they compared the results from the two techniques, they found that FLASH was able to provide the same clinically actionable information — in some cases more — as standard metagenomic sequencing, but it was able to do so with 1,000 times greater efficiency.
Nor is FLASH limited to bacterial diagnostics. The technology was also able to detect different strains of drug-resistant Plasmodium, the single-celled parasite that causes malaria, in dried blood spots obtained from patients suffering from the disease.
Though FLASH isn’t the first CRISPR-based microbial diagnostic tool — SHERLOCK and DETECTR, developed at Broad Institute and UC Berkeley, respectively, came earlier — so far it is the only one that can be multiplexed, which means that FLASH can be programmed to detect and reveal the precise identity, not just the presence, of thousands of AMR sequences at once. This unique ability should allow clinicians to identify the full spectrum of AMR genes from a patient sample in a single round of sequencing. Ultimately, doctors will be able to use this data to personalize treatment for individual patients.
FLASH may also prove versatile enough to be used in applications well beyond the scope of microbial diagnostics.
“The hunt for antimicrobial resistance is just one example of how FLASH can be used,” Crawford said. “We’re already working with collaborators to use FLASH to study the genetic variation that exists in viruses and human gut microbiome sequences. Soon, I hope others will be using FLASH to find mutations in cancer.”
Authors: Co-first authors on the paper are Jenai Quan of CZ Biohub and UCSF, Charles Langelier of UCSF and Alison Kuchta of UCSF; additional authors include Joshua Batson, Amy Lyden, Aaron McGeever, Lara Pesce Ares, Katherine A. Travisano, Rene Sit and Norma F. Neff of CZ Biohub; Noam Teyssier, Saharai Caldera, Jordan Wilheim, Maxwell Murphy, Jennifer L. Smith, Adam Bennett, Roly Gosling and Carolyn S. Calfee of UCSF; Boris Dimitrov and Ryan King of the Chan Zuckerberg Initiative; Roberto Amato of the Wellcome Sanger Institute; Davis R. Mumbengegwi of the University of Namibia; Peter M. Mourani of the University of Colorado School of Medicine and Children’s Hospital Colorado; Eric D. Chow of UCSF; Peter S. Kim of the CZ Biohub and Stanford University School of Medicine; Bryan Greenhouse and Joseph L. DeRisi of CZ Biohub and UCSF.
Funding: Research was supported by UCSF CTSI Catalyst grant C27552C-01-135119; NHLBI grant K23HL138461; NIGMS grant 5T32GM007546; NIH NHLBI grant 5R01HL124103; NHLBI grants R01HL110969, K24HL133390 and R35HL140026; the Chan Zuckerberg Initiative; and the Chan Zuckerberg Biohub.
Conflicts: Two U.S. Provisional Patent Applications relating to this technology have been filed.