Skip to Main Content

The revolutionary genetic technique better known for its potential to cure thousands of inherited diseases could also solve the challenge of Covid-19 diagnostic testing, scientists announced on Tuesday. A team headed by biologist Feng Zhang of the McGovern Institute at MIT and the Broad Institute has repurposed the genome-editing tool CRISPR into a test able to quickly detect as few as 100 coronavirus particles in a swab or saliva sample.

Crucially, the technique, dubbed a “one pot” protocol, works in a single test tube and does not require the many specialty chemicals, or reagents, whose shortage has hampered the rollout of widespread Covid-19 testing in the U.S. It takes about an hour to get results, requires minimal handling, and in preliminary studies has been highly accurate, Zhang told STAT. He and his colleagues, led by the McGovern’s Jonathan Gootenberg and Omar Abudayyeh, released the protocol on their website.


Because the test has not been approved by the Food and Drug Administration, it is only for research purposes for now. But minutes before speaking to STAT on Monday, Zhang and his colleagues were on a conference call with FDA officials about what they needed to do to receive an “emergency use authorization” that would allow clinical use of the test. The FDA has used EUAs to fast-track Covid-19 diagnostics as well as experimental therapies, including remdesivir, after less extensive testing than usually required.

For an EUA, the agency will require the scientists to validate the test, which they call STOPCovid, on dozens to hundreds of samples. Although “it is still early in the process,” Zhang said, he and his colleagues are confident enough in its accuracy that they are conferring with potential commercial partners who could turn the test into a cartridge-like device, similar to a pregnancy test, enabling Covid-19 testing at doctor offices and other point-of-care sites.

“It could potentially even be used at home or at workplaces,” Zhang said. “It’s inexpensive, does not require a lab, and can return results within an hour using a paper strip, not unlike a pregnancy test. This helps address the urgent need for widespread, accurate, inexpensive, and accessible Covid-19 testing.” Public health experts say the availability of such a test is one of the keys to safely reopening society, which will require widespread testing, and then tracing and possibly isolating the contacts of those who test positive.


The scientists are also in discussions with the Gates Foundation about distributing the testing kits. “It could be especially valuable in places where there is not the capacity for centralized testing,” Gootenberg said.

Other labs are also developing CRISPR-based Covid-19 tests. Last month scientists at the University of California, San Francisco, and Mammoth Biosciences reported that theirs takes roughly 40 minutes, compared with four to six hours for the traditional swab tests (which use a biochemical reaction called PCR and require numerous reagents). But that test appears to produce false negatives slightly more often than the PCR test. It also requires two steps, similar to one that Zhang developed in February; that adds complexity and increases the chances that samples will cross-contaminate.

Zhang’s February test is being used by hospitals in Thailand. But with his colleagues, he has spent the last two months creating the one-step version.

“The key advance for STOP is the simplification into a single-step reaction, which prevents cross contamination during step-to-step liquid transfers,” he said. “That makes it suitable for point-of-care use.”

Existing “home” Covid-19 diagnostic tests such as that from LabCorp allow people to collect the sample — that is, swab themselves — at home. The actual test must be done at a lab. Currently approved point-of-care Covid-19 diagnostic tests, such as Abbott ID NOW and Cepheid GeneXpert, require specialized and expensive instrumentation and technical expertise, limiting widespread use. Abbott’s has also been plagued by false negatives, meaning it sometimes fails to detect the coronavirus.

STOP stands for “Sherlock Testing in One Pot.” It builds on Zhang’s 2017 CRISPR invention, called Sherlock. Like the better-known, genome-editing versions of CRISPR, Sherlock starts with a guide molecule that homes in on a specific stretch of DNA — or RNA, which constitutes the genome of the new coronavirus. An enzyme called AapCas12b, from the bacterium Alicyclobacillus acidophilus, cuts the RNA in a way that generates a fluorescent signal detectable much as home pregnancy tests detect the presence of a pregnancy-related hormone.

For STOP, the scientists targeted their guide molecule at the coronavirus’s N gene, which codes for the nucleocapsid, or shell, that encloses its genome. The PCR test developed by the Centers for Disease Control and Prevention also targets the N gene.

The MIT scientists tested STOP on nasopharyngeal swabs from Covid-19 patients and on saliva samples from healthy people to which the new coronavirus was added. STOP had 100% specificity — it never “found” the coronavirus if the microbe wasn’t really there — and 97% sensitivity, meaning if the virus was present, the test missed it only 3% of the time. The existing PCR test misses up to 30% of cases, scientists have said.

The scientists have prepared enough material so far for 10,000 tests, and are making it free to researchers who want to evaluate its potential diagnostic use.

  • Academic scientists developing cheaply deliverable Covid tests with hyper-fast results, based on CRISPR technique, in multiple nations simultaneously, means there must be merit to the method. The FDA must check and act swiftly, this would solve the biggest bottleneck (dire shortage of tests and speed) in short time. Bravo to the scientists !

  • The penultimate paragraph of this article includes the following sentence: “STOP had 100% specificity — it never “found” the coronavirus if the microbe wasn’t really there — and 97% sensitivity, meaning if the virus was present, the test missed it only 3% of the time.”

    To find a microbe that is not really there is a false positive. If a test never finds a microbe that’s not there, then its false positive rate is 0% and its true positive rate is 100%. According to Wikipedia, the true positive rate is called the “sensitivity” of a test, not the “specificity,” as the cited sentence claims.

    To miss finding a microbe that is actually present is called a false negative. If a test misses finding a microbe that is actually present 3% of the time, then its false negative rate is 3% and therefore it’s true negative rate is 97%. According to Wikipedia, the true negative rate is called the “specificity” of the test, not the “sensitivity” of the test, as the cited sentence claims.

    Here’s the Wikipedia link:

    If Wikipedia is correct, then it appears that this article is using the terms “sensitivity” and “specificity” incorrectly. I have noticed that many articles in STAT and other high-quality publications use these terms incorrectly or ambiguously. I think it would be very helpful if STAT published an article clarifying once and for all what the correct usage of these terms actually is. And for bonus points, an authoritative discussion of the terms “positive predictive value” and “negative predictive value” would also be nice.

  • Isn’t science wonderful! CRISPR could be the answer to many health issues.
    I remember when early genome studies were being done and the uproar over its potential for misuse was prominent in the news. Everything that exists can be used for good or evil. Humanity will benefit from the good use of scientific discoveries.
    I wish you much success on approval from the FDA.

  • For every dollar spent on ramping up testing in the US, how many dollars in additional GDP do we get back in return? Congress needs this number, companies developing and supplying the tests need the funding, and the world needs to safely return to work ASAP.

  • Is there anything that can’t won’t be done using the defense mechanism of these single cell entities?

    But bravo.

    One small query: RNA is cut causing a fluorescent …is DNA cut as well.

  • Re near universal testing which could be accomplished by this test refer to the writer Michael Connelly’s protagonist Bosch “Everybody matters or nobody matters.” Don’t give up.
    Thank you
    Julie B. Koerner

Comments are closed.