The Code Breaker: Jennifer Doudna, Gene Editing, and the Future of the Human Race

When the coronavirus was declared a public health emergency in the United States on January 31, 2020, it meant the FDA could speed up the review of diagnostic tests. However, this had a strange unintended consequence. In a public health emergency, universities and research labs cannot use the tests they have devised unless they seek emergency use authorization. As it turned out, getting public use authorization would be a slow, painful process. Worse, the first batch of the CDC’s tests, which received FDA approval in February, did not work properly.

In a typical coronavirus test, RNA, if present, is extracted from a patient’s mucus—collected through a long swab—and then reverse transcribed into DNA. This DNA is amplified into a million strands using a polymerase chain reaction (PCR), making it easy to spot. When the CDC’s first batch of tests proved faulty, the University of Washington stepped in. Alex Greninger, an assistant director of virology at the university, had already been working on a diagnostic test. By February, he had a working test ready, but getting authorization for the test proved tricky.

However, Anthony Fauci, the head of infectious diseases at the National Institutes of Health, stepped in and urged the FDA to hasten approvals to universities. Rapid testing was the absolute need of the hour. On February 29, the FDA allowed university labs to begin testing while they awaited authorization. Within a few weeks, Greninger was testing 2,500 samples a day. Meanwhile, Eric Lander, always eager to deploy science in public interest, sanctioned a testing lab on the premises of the Broad Institute.

Doudna and her colleagues brainstormed whether they should use the trusted (but slow) PCR method or invent a new tech based on CRISPR. The urgency of the pandemic mandated that they go ahead with both, focusing especially on the PCR method. To command the operation, Doudna chose Moscow-born Fyodor Urnov, who had led her lab’s efforts in devising a treatment for sickle-cell anemia. A quasi-renaissance man, Urnov was also interested in both entrepreneurship and literature. Urnov was joined by postdoc researcher Enrique Lin Shiao and Jennifer Hamilton, who had taught Isaacson how to use CRISPR in a human cell.

The team scrambled for equipment from the now-deserted campus. Spending over half a million dollars on supplies, Urnov, Hamilton, and Lin Shiao procured a contraption called the Hamilton STARlet. Robotic pipettes extract small amounts from patient samples, place them onto plates the size of iPhones with 96 wells, and load the plates into the machine’s chamber, where each sample is doused with reagents to extract the RNA. The machine issues a barcode to each sample to eliminate mixing and errors. Since federal permissions were still unclear and private labs were taking a week to return tests, there was a huge demand for Berkeley’s testing, especially for the poor and homeless. On April 6, 2020, the first batch of tests was delivered to quarantined firefighters.

At IGI’s March 13, 2020, meeting, Urnov raised a pertinent issue: If bacteria can use CRISPR to detect attacking viruses, can CRISR be made to detect coronavirus RNA in human cells?

Two groups had already presented papers outlining a viable CRISPR method for COVID-19 diagnosis. One was associated with Mammoth Biosciences, which had Doudna on its board of scientific advisors. Mammoth was founded by Doudna’s doctoral students Janice Chen and Lucas Harrington. In Doudna’s lab, Chen and Harrington had discovered the that Cas12a enzyme could also be engineered to target and cut DNA. However, Cas12a didn’t just stop there; it went into a cutting frenzy, slicing any single-strand DNA available. Doudna’s husband Jamie Cate suggested that this property could be harnessed into a diagnostic tool. So, Chen and Harrington combined Cas12a with a “reporter molecule,” which was a fluorescent signal connected to a bit of DNA. When Cas12a found a targeted sequence of DNA, it also attacked the reporter molecule. The glowing signal from the chopping indicated that the enzyme had found its targeted sequence, which could be a virus, a bacteria, or even cancer cells. This system was dubbed DETECTR.

The other group was affiliated with Sherlock Biosciences, founded by Zhang and two of his protégés, Omar Abudayyeh and Jonathan Gootenberg. Zhang’s proposed detection system—with the acronym SHERLOCK—was based on the newly detected Cas13 enzyme, which behaved much like Cas12a, but with RNA. In his work with Cas13, Zhang speculated that it was an evolutionary method to have the infected cell commit suicide: The cell cut up its own RNA and died to prevent the infection from spreading.

Though Mammoth and Sherlock released their papers around the same time, there was much less wrangling over patents this time round. With the coronavirus crisis looming, both Doudna and Zhang knew it was imperative to reprogram their diagnostic tools to detect coronavirus infections.

In January 2020, Zhang began getting calls for help about the coronavirus from Chinese academics and even the Chinese consulate in New York. Zhang decided to reconfigure his SHERLOCK tool to test for COVID-19. Since he did not have access to real COVID-19 samples so early on, Zhang used the system on a synthetic version. With Abudayyeh and Gootenberg’s help, he soon devised a simple method that could deliver results in an hour. All it required was a small device that kept a constant temperature as it amplified the virus’s genetic material. By February 14, Zhang’s lab had posted a white paper on the process—called STOP, short for SHERLOCK Testing in One Pot—and invited any lab to use the process freely. Meanwhile, Sherlock Biosciences began making COVID-19 test kits based on Zhang’s method. By late 2020, the company was working with manufacturers to develop small machines that could give results in less than an hour.

Around the same time, Chen and Harrington were also approached to develop a test by a board member of Mammoth Biosciences. Because of Mammoth’s ties with IGI, Chen and Harrington could access resources at the University of San Francisco’s hospital. Thus, they could test their DETECTR-based tool on real COVID-19 samples, unlike Zhang. Further, the Mammoth test relied on the Cas12a enzyme, which targets DNA. Theoretically, this should have been made it less efficient than Sherlock, which uses RNA-destroying Cas13. However, both tests required RNA to be converted to DNA to be amplified. SHERLOCK further had to convert this amplified DNA into RNA to be detected, adding an extra step.