Editing Humanity: The CRISPR Revolution and the New Era of Genome Editing

CHAPTER 22 CRISPR PRIME

In 1960, a tall, gentle man named Victor McKusick, a medical geneticist at Johns Hopkins University, published the first edition of a remarkable catalogue of genetic traits and disorders. Mendelian Inheritance in Man became the bible of geneticists around the world, the definitive source of information about human genetic diseases and their underlying mutations. After a dozen editions, the catalogue was moved online. It currently lists more than 7,000 discrete genetic disorders and traits. Of those, McKusick was most closely associated with Marfan syndrome, the genetics of which he first described in 1956. Thirty-five years later, McKusick’s colleagues at Hopkins identified the faulty gene alongside two other teams. I invited McKusick to write the accompanying commentary in Nature.¹ It was only fitting. In it he discussed the notion that President Abraham Lincoln might have had Marfan syndrome.

If he were still alive, the father of medical genetics would be in awe at the progress we’ve made in documenting the myriad ways in which our genetic software can be corrupted, not to mention the potential of delivering a patch to fix those errors. The Welsh geneticist Steve Jones wrote that the book of life “has a vocabulary—the genes themselves—a grammar, the way in which the inherited information is arranged, and a literature, the thousands of instructions needed to make a human being.”² The letters on the pages of our books are subject to a host of different insults—substitutions, deletions and insertions, expansions, duplications, and rearrangements. Building on McKusick’s legacy, the global genetic community has documented mutations in about one third of the total number of genes in the human genome. That number will increase.

Some genetic disorders are caused by the tiniest mutation imaginable—the swap of one letter in the genetic code for another. Sickle-cell disease results from an A shifting to a T in the beta-globin gene. Progeria, a genetic form of premature aging, is caused by a C to T substitution in the lamin A gene. Cystic fibrosis is caused by hundreds of different mutations in a single gene, the most prevalent being the loss of three bases coding for a single amino acid. Conversely, the most common mutation in patients with Tay-Sachs disease is the addition of four bases (TATC) in the beta-hexosaminidase gene. Huntington disease, fragile X mental retardation, and dozens of other disorders arise from the bizarre expansion of a tract of repetitive DNA. Other disorders arise from insertions, duplications, or deletions of longer stretches of DNA, including entire chromosomes such as trisomy 21, or Down syndrome. And there are many more subtle genetic defects, including epigenetic mutations that silence one copy of a gene, depending on which parent supplied the gene.

As genome engineers contemplate the plethora of genes and mutations that need to be corrected to treat or cure genetic disease, they will need a deluxe toolbox that will extend beyond CRISPR-Cas9. For all of the astonishing progress since 2012–13, there are strong signs that the CRISPR toolbox is receiving a major upgrade with a suite of new tools that riff on the original CRISPR gene editing machinery. Cas9 engineers a complete break in the DNA, and while the ability to stitch in the desired sequence or repair is improving rapidly, the process still lacks the requisite precision for most therapeutic applications. The classic CRISPR technology will only work therapeutically in a fraction of genetic diseases.

If you were to draw it up on a whiteboard, the Holy Grail of genome editing would be to develop a technology that can modify a single letter of the genetic code without cleaving the DNA in the process. Almost before the ink was dry on the classic CRISPR papers, researchers were studying the building blocks, seeking to modify and adapt them. That has been the goal of many investigators, but one in particular stands out in his mission to design a truly precise molecular editor. He’s a prodigiously talented scientist of Asian descent whose talent was on display in high school before enrolling at Harvard, excelling in his PhD in California, and hitting his prime at the Broad Institute. But this isn’t about Feng Zhang.

A decade older than Zhang, there are indeed some striking similarities in David Liu’s career. The son of Taiwanese parents, Liu was born and raised in Riverside, California. His mother was a physics professor, his father an engineer. Like Zhang, Liu’s scientific talent shone brightly in high school, driven by what he admits was some “immature competitiveness.” In 1990, he placed second in the national Westinghouse Science Talent Search competition, and first in his high school.

As a freshman at Harvard University, Liu’s interests gravitated toward physics rather than chemistry. But that changed in December 1990 when he traveled to Stockholm as one of five top US students to attend lectures by the newly minted Nobel laureates, including Harvard chemistry professor E. J. Corey. Liu was enthralled by Corey’s work on creating new molecules, like assembling Lego blocks. Afterwards, Liu told Corey he wanted to work in his lab. He got that opportunity and eventually graduated top of his class of more than 1,600 students in 1994. Years later, Corey told the Boston Globe that Liu was “going to be a superstar.”³

Liu moved back to California to take up a PhD at Berkeley with Peter Schultz, a talented molecular biologist who was literally rewriting the genetic code. Liu spent the next few years studying methods to expand the genetic alphabet—to encode and incorporate synthetic amino acids (beyond the twenty that occur naturally in the body) into proteins. A lecture on his groundbreaking graduate work back at Harvard turned into a de facto faculty interview. Occasionally scientists shine so spectacularly during their PhD that, like a professional basketball team drafting a high school prodigy, a university will offer them a faculty position. Harvard offered Liu a professorship, bypassing the usual four to five years of postdoctoral training. It was too tempting to refuse, but he doesn’t recommend others try it. “I had no idea what I was doing,” Liu admits.⁴

In autumn 1999, at the ripe old age of twenty-six, Liu joined the ranks of Harvard’s illustrious chemistry faculty, with its seven Nobel Prizes since 1964. Having demonstrated the possibilities of performing molecular evolution on the building blocks of life, Liu decided to go for something really big: protein evolution in a test tube. During his first decade at Harvard, Liu’s lab made a name in building new technologies for molecular evolution and applying them to treat human disease. The key method is called phage-assisted continuous evolution (PACE), developed by Kevin Esvelt.⁵ In 1999, Liu even dabbled in a type of genome editing—an effort to construct a gene activator made up of a DNA (or RNA) triple helix that could be targeted to various sites in the genome to regulate genes or cut DNA. Liu admits the project “utterly failed” but even though he moved productively into other areas of research, his interest in performing chemistry on the genome stuck.⁶

To lure the top students, competing against some of the biggest names in chemistry like Schreiber, Szostak, and Whitesides, Liu promoted his annual lab open house with eye-catching posters. One showed Liu metamorphosizing into Regis Philbin, the host of Who Wants to Be a Millionaire? Others showed him in full Matrix regalia or dressed up as Spider-Man villain Doctor Octopus. The strategy apparently worked: Liu’s research took off and in 2005, he was promoted to full professor, just thirty-one years of age. That same year, he joined Doudna as one of about three hundred investigators appointed (and technically employed) by the Howard Hughes Medical Institute, one of the highest echelons in American biomedical research.

For all of Liu’s brilliance and commitment in the lab, he strives to maintain a healthy work-life balance. “Chemistry is life, but life is a lot more than chemistry,” he says.⁷ Some of his hobbies hearken back to his engineering upbringing. In the early 2000s, he built a featherweight airplane that could almost hover indoors, before drones became a phenomenon. He also built a Lego robot called the mousapult, which would entertain his cats by throwing a toy in the direction of a heat signature, using a sensor from a burglar alarm.

Liu’s most intense—and lucrative—hobby, first picked up during his student years, was blackjack. His mathematical ability to count cards (a legal activity) became a teachable moment and something bordering on an obsession. Liu started teaching a weekly course for enthusiastic students from which he cultivated a devoted squad of fourteen “blackjack ninjas.” Every few months, the young professor would lead a delegation to Las Vegas and spend the weekend gambling—sometimes running fifteen hours at a time. Liu joked he was just hoping to earn enough to buy his wife a nice pair of earrings, but his posse was known to win “absurd sums of money.”⁸

On Sunday nights, Liu took the JetBlue red-eye flight from Las Vegas to Boston, rolling up to teach his morning chemistry lecture pretty tired. He would ask himself why he was flying to casinos to gamble with students.⁹ Occasionally though, after a particularly successful trip, he wondered if being a chemistry professor was all it was cracked up to be. Eventually, his hand was forced: the MGM Grand Casino in Las Vegas banned him. But he still carries a laminated card of calculations in his wallet should opportunity knock.

In his office on the third floor of the Broad Institute, besides Liu’s own art, skilled photography, and mineral collection, a visitor cannot help but notice the thirty-pound, three-foot Iron Man “Hulkbuster” replica. It’s the perfect metaphor: the ultimate shield to protect against the excesses of the Hulk’s gamma-ray-induced rampages. Like Tony Stark, Liu has a penchant for inventing cool technologies to shield humans against genetic mutations, stacking the odds and seeing how high he can fly. “He’s going to be the godfather of CRISPR 2.0,” says Gerald Joyce, director of the Genomics Institute of the Novartis Research Foundation.¹⁰ After listening to a spellbinding lecture from Liu in the Canadian Rockies in early 2020, a scientist sitting next to me whispered in my ear: “He’s a genius!”

Growing up in upstate New York, Nicole Gaudelli’s love of science and nature was nurtured by her father and grandfather. She loved going to zoos, fishing, growing crystals, and building water rockets. She thought about being a doctor, but her father suggested that she could help many more people by being a research scientist. During her PhD at Johns Hopkins, Gaudelli was captivated by a guest seminar given by Liu talking about PACE and molecular evolution. She decided to apply for a coveted position in Liu’s lab for her postdoc.

Shortly after Gaudelli arrived in the lab in 2014, she befriended a new postdoc from southern California who had just earned her PhD from Caltech. Alexis Komor was working on something completely different—a project inspired by months of email exchanges with Liu prior to her arrival. Komor had interviewed with Liu eighteen months before finishing her PhD, hoping to persuade a big-name chemist that she could flourish in his group.

Komor began emailing Liu ideas for her postdoc project (“mutually guided brainstorming” is how Liu puts it). One item was an idea she’d sketched out to fulfill a Caltech graduation requirement: she wanted to evolve a ribonuclease enzyme in the lab so that it could degrade a specific sequence of RNA. Liu liked it but suggested she think about DNA-based editors, in particular the CRISPR-associated nuclease, Cas9. On November 1, 2013, he emailed her: “If you could program a specific A-to-G (for example) change in the human genome, you could really transform genome engineering and possibly human therapeutics.”

Komor was excited but confused. “Why is he so crazy about this Cas9 thing?!” she thought.¹¹ But she kept refining her idea and by the time she arrived in Boston in September 2014, the basic idea of base editing had been born. Ironically, Liu had confused Komor and Gaudelli in the run-up to their arrival, mixing up their respective project ideas. On Komor’s first day, Liu introduced her to the rest of the male-dominated lab, and then Gaudelli. “This is Nicole. I kept confusing you. You can see why!” The pair burst out laughing: Gaudelli has dark hair and eyes, unmistakable Italian heritage. Komor is quintessential Californian, blond hair and blue eyes. They became fast friends.

Komor’s first six months were uncomfortable, far apart from her husband who was still in California finishing his PhD. She hoped Boston would be a short-term stay to complete a postdoc so she could return to her family and the California sunshine. “Technology development projects are super risky,” she told me. “At the beginning I didn’t know what I was doing!” Group meetings in which her plans were microdissected by quizzical, sometime skeptical colleagues, were the bane of her existence.

Although Doudna was a classical chemist, the field of CRISPR genome editing had evolved as a largely biological discipline. Komor and Liu brought a different skill set, and it paid off. “Single-stranded DNA is a lot more reactive than when it is double-stranded,” Komor says. When Cas9 binds to DNA, it unzips the double helix to expose a stretch of about five bases of single-stranded DNA. Here, then, was a window to perform some cool chemistry. Komor began with cytidine deaminase, an enzyme that converts cytidine (C) to uracil (U), but only works on single-stranded DNA. By tethering the deaminase to an inactive (“dead”) form of Cas9, she would create a homing machine to seek out a target DNA sequence and unspool a short stretch of DNA (without cutting the strand) upon which the cytidine deaminase could act.

After about eight months, Komor had a prototype base editor working that could convert a C:G basepair into a U:G mismatch pairing. Now she faced a new problem: the cell’s DNA repair system won’t tolerate the mismatch, so it tries to restore a natural basepair, like finding the right match in a jigsaw puzzle. Facing this U:G intermediate, Komor needed to tip the odds to favor the solution she wanted—coax the cell to repair the G, which would result in an T:A basepair.^I She had to find a trick to complete the base edit, rather than watch the cell’s DNA repair process simply undo her good work by fixing the U, reverting the base pairing back to where she’d started.

Komor’s first trick was to find a way to block an enzyme that “rips out uracils like nobody’s business.” So she fused a third component to the base editor—an inhibitor of uracil DNA glycosylase, or the ripper. It shifted the balance a bit but not as much as she hoped. And then one day, she had an epiphany while talking to a colleague in the lab kitchen. “It just came to me,” she recalls. “Oh my God, we’re working with an endonuclease!” Although she was working with a “dead” form of Cas9, it was still a nuclease that cleaves DNA. By replacing a single amino acid in the enzyme, Komor could restore a “nickase” function that would clip one strand of the double helix. By nicking the G-containing strand (leaving the U intact) she could trigger the cell’s DNA repair machinery to fix the G rather than the U nucleotide.

When Komor told Liu about her brilliant idea, he started swearing: he’d wanted to start writing up the paper for fear of being scooped. But Komor’s idea was obviously worth trying. “How quickly can you do it?” Komor spent Christmas 2015 back home in California drafting the manuscript, editing it on Christmas Day, and even foregoing her ten-year high school reunion. The Nature reviewers initially gave the paper a rough ride on technical grounds, and it was rejected. Komor worked tirelessly to rebut each of the criticisms, while Liu phoned the editor, Angela Eggleston, to appeal. The revised paper was accepted and eventually published in April 2016.¹² When I asked Komor why they chose Nature, she laughed. “Where else would we send it?!”

As Komor was developing the first C-to-T base editor (CBE), Gaudelli grew increasingly interested in her friend’s research. After much internal debate, she decided to abandon her own project, switching instead to try to develop a novel base editor that could do the reverse reaction—an A-to-G base editor. This would be a more useful setup for medical applications as about half of the known pathogenic mutations in human genes involve mutations of a G to A. (Indeed, there is a high spontaneous mutation rate involving the deamination of cytidines de novo to uracil, resulting in an erroneous T:A basepair.) Developing a system that could reverse this common source of human mutation could have a profound medical benefit. There was one small problem however: Gaudelli didn’t have any starting material.

Liu had one unbreakable rule in his lab that had endured more than fifteen years: never start a project by evolving the starting material. But Gaudelli didn’t have much choice: there was no natural enzyme that deaminates A to G in DNA.^II Undaunted, Gaudelli trained her sights on a bacterial enzyme called tadA, which works on RNA, not DNA. With nothing to lose, Gaudelli performed a slightly crazy experiment—evolution in a test tube to try to generate the desired properties.^III The first round of evolution yielded a mutation that enabled the altered enzyme to tackle single-stranded DNA instead of RNA as its substrate. The mutation was in the precise location that Gaudelli would have expected. She sent a quick slide to Liu, who started swearing again. “Holy—, this is our smoking gun,” he replied.

Several rounds later, Gaudelli had evolved a potent A-base editor, or ABE. She was also able to demonstrate the ability to modify mutations in genes responsible for hereditary diseases including hemochromatosis and sickle-cell disease. Like Komor, Gaudelli’s base-editing exploits also earned her a first-author paper in Nature.¹³ By now, rival journal editors were visiting Liu to solicit hot papers like Gaudelli’s. It sailed through peer review over a long weekend. Researchers around the world immediately jumped on the base editing bandwagon.¹⁴

Looking back, Gaudelli took an almost ludicrous gamble but she pays tribute to the nurturing environment in Liu’s lab and her “Hulkbuster” of a boss, who “just makes you feel invincible.”¹⁵ She could have had any faculty position she wanted, but she elected to join Beam Therapeutics, a new biotech company Liu cofounded with his comrades in arms, Feng Zhang and Keith Joung. Gaudelli started to think about the friends and family base editing might eventually help. “What if one of those people was my father? My grandfather? What if that was a hypothetical child of mine?”

A friend of Liu’s, a pediatric oncologist at Stanford named Agnieszka Czechowicz, came up with the company’s name. She texted Liu her suggestion—Beam—which evokes a laser, a precision technology. “It also happens to stand for ‘Base Editing And More’,” she pointed out.

“What’s the ‘more’?” Liu asked.

“I’m sure you’ll figure it out,” she replied.¹⁶

Several Liu and Zhang postdocs, including Fei Ann Ran, have followed Gaudelli’s path to Beam’s facility in a building next to Novartis’s R&D headquarters in the former Necco candy factory. In 2019, CEO John Evans took Beam public, raising a tidy $180 million, which will help them build a new headquarters in the heart of Kendall Square.

In less than five years, base editing had evolved from a speculative postdoc proposal to a pair of landmark Nature papers, rapid uptake in labs around the world, and a public biotech company. The creation of base editors is an impressive feat of chemistry. As Liu told me: “These molecular machines have to search the genome for a single target position, open up the DNA, perform chemical surgery directly on a base to rearrange the atoms—then do nothing else [except] defend the edit from the cell’s fervent desire to undo them.”¹⁷

The first two base editors offer a means to edit “all the easy mutations,” says Komor. Liu anticipates “there’ll be a library of base editors and you’ll pull out the book that matches exactly what you need.” That choice will be influenced by the desired edit, the sequence context, off-target effects, and so on. In March 2020, Liu underscored that prediction: working in collaboration with Jennifer Doudna’s group, another postdoc, Michelle Richter, unveiled a new-and-evolved version of the A base editor that was six hundred times more active than Gaudelli’s original.¹⁸ Just as with CRISPR-Cas9, base editors are prone to cutting at off-targets. But scientists are working fast to improve their specificity. To keep this in perspective, note that in each of your 10 trillion cells, the genome is constantly mutating. Hundreds of times a day, a C is mutated to a U, which if left unchecked, would become a C-to-T mutation.

It will be years before a base editing drug is available, with many hurdles to climb to get there. Komor, who is now a university professor in her beloved Southern California, sees great promise in not just treating symptoms, but “curing the disease.” That’s what I’d expect her to say, but Liu’s group has already used base editing to correct a mouse model of a rare but devastating genetic disease. Progeria, or Hutchinson-Gilford progeria syndrome, is a dominantly inherited disease caused by a single-base mutation in the laminA gene that results in extreme premature aging. The mutant protein, called progerin, wreaks damage in the aorta and other tissues. Affected children rarely live beyond fifteen years of age.

Liu partnered with NIH director Francis Collins, who years earlier had developed a mouse model of the disease carrying the human progeria mutation. Liu’s team delivered the base editor via a pair of AAV vectors, using a molecular Velcro to splice the components together once inside the cell. The ABE corrected the mutation and squashed production of progerin. In the treated mice, cells regained their normal shape, and the aorta was restored to near-normal health. Stunningly, the treated mice look healthy and live longer than the progeria mice. How much longer Liu couldn’t exactly say—for the good reason they were still alive. “We’re really excited,” Liu said, moving forward “carefully but quickly” to advance this revolutionary treatment from mice to boys and girls.¹⁹

What about more common diseases? Verve Therapeutics, a genome editing start-up, collaborated with Gaudelli’s team at Beam to test a one-shot strategy in two monkey models of heart disease. Verve used a lipid nanoparticle to deliver the ABE to the liver of crab-eating macaques to inactivate a pair of known genes that regulate cholesterol. CEO Sek Kathiresan reported a dramatic lowering of “bad” LDL cholesterol and triglycerides in animals targeted at the PCSK9 and ANGPTL3 genes, respectively.²⁰ The results need to be confirmed and extended in humans, which is some years away. But base editing could help realize Kathiresan’s dream of a “one-and-done genome editing medicine for heart disease,” providing an alternative to chronic statins and reduce the 18 million cardiovascular deaths each year.

From Archimedes in the bath to Isaac Newton’s bruised head, there are many legendary aha moments in science history. Perhaps the most bizarre episode belongs to the late Kary Mullis, who recalled the invention of the polymerase chain reaction in a quite extraordinary Nobel lecture. It’s too good not to relive here:

One Friday night I was driving, as was my custom, from Berkeley up to Mendocino where I had a cabin far away from everything off in the woods… As I drove through the mountains that night, the stalks of the California buckeyes heavily in blossom leaned over into the road. The air was moist and cool and filled with their heady aroma… EUREKA!!!!… EUREKA again!!!!… I stopped the car at mile marker 46.7 on Highway 128. In the glove compartment I found some paper and a pen… “Dear Thor!” I exclaimed. I had solved the most annoying problems in DNA chemistry in a single lightning bolt… We got to my cabin and I started drawing little diagrams… with the aid of a last bottle of good Mendocino County cabernet, I settled into a perplexed semiconsciousness… The first successful experiment happened on December 16th. I remember the date. It was the birthday of Cynthia, my former wife… There is a general place in your brain, I think, reserved for “melancholy of relationships past.” It grows and prospers as life progresses, forcing you finally, against your grain, to listen to country music.²¹

By comparison, Andrew Anzalone’s story of the genesis of “prime editing”—fueled by caffeine not cabernet, ambling around the streets of Lower Manhattan rather than speeding through the Napa night—could use a little work in the dramatic license stakes. But his ideas, formed in 2017 before leaving Columbia University for a position in Liu’s lab, were crucial in devising a new genome editing platform. A physician-scientist rather than a chemist, Anzalone was inspired by Liu’s base editing exploits, but sensed an opportunity to go further. “The base editors were really good for making four possible base changes but they couldn’t address the other eight base changes or the small indels,” he said.^22IV

In October 2019, Liu unveiled a new genome editing technology developed by Anzalone and other members of the lab that riffed on base editing, expanding the repertoire of potential DNA alterations. “This is the beginning of an aspiration to make any DNA change in any position of a living cell or organism,” Liu said.²³ I was sitting in the audience of four hundred rapt scientists at the Cold Spring Harbor Laboratory for Liu’s first public presentation on prime editing, just ten days before the study was published in (no surprise) Nature.²⁴ “Prime editing is somewhat complicated,” Liu admitted. But it works, and the meticulous four-step sequence offers important advantages.

There are more than 75,000 known disease-causing mutations in the human genome—about half of those are point mutations—but most can’t be targeted by CRISPR-Cas9 or base editing. While base editing’s strength is its ability to make a class of base substitutions known as transitions, they only account for four of the dozen possible base changes. The CBE would in principle fix 14 percent of known point mutations; the ABE accounted for a higher fraction, some 48 percent.

Anzalone wondered if he could build on this technology to engineer any single-base change, transitions and transversions.²⁵ What he developed is a new system that moves scientists closer to a true search-and-replace function for DNA, regardless of the letter or its location. Naturally, the system would start with the programmable single-guide RNA, which directs Cas9 to the stretch of DNA to be edited. But what if, instead of providing the replacement sequence via a DNA template, he used the same guide RNA molecule to supply the edit? The system would have two programmable elements—the target site and the edit itself. Naturally the replacement sequence would have to be converted from the RNA guide to DNA, but luckily there’s a very well-known enzyme for that—reverse transcriptase (RT).^V

The extended guide RNA, renamed the pegRNA (prime editing-gRNA), specifies the target as well as the desired edit. Similar to the construction of the base editors, Anzalone fused RT to dead Cas9, then figured out a scheme to coax the edited DNA copy into the target sequence. The first step is to engineer the pegRNA and RT enzyme to copy the pegRNA strand into DNA. This results in a flap of DNA that needs to be stitched into the double helix. (Helpfully, cells have enzymes called “flap nucleases” that help in this process.) Finally, the method introduces a nick into the complementary strand, which is then repaired to fully match the edited strand.

Anzalone’s approach didn’t get off to an auspicious start. When he fused RT to Cas9, he achieved zero editing. But further tests with a batch of RT variants soon resulted in some positive results. If the original CRISPR approach is a molecular scissors and base editing is a more precise pencil eraser, Liu describes prime editing as a word processor, capable of performing a search-and-replace function on any typo in the DNA alphabet. It also carries out some classes of insertions and deletions (indels), including those responsible for the most common form of cystic fibrosis (a three-base deletion) and Tay-Sachs disease (a four-base insertion), respectively.

After submitting their report to Nature, Liu and Anzalone had little trouble attending to the three anonymous referees’ comments. Figuring out the identity of one of them was easy: few reviewers sprinkle words like “quixotic” in their reports. Urnov effusively recommended publication, citing the increased flexibility on editing sites, no more PAM deserts, no DNA donor, few off-targets, and the early data on correcting DNA in neurons.

In their paper, the Liu group showcased the full range of prime editing’s prowess—175 different edits, including 100 point mutations of all possible types; repair of known disease mutations in human cells; insertions and deletions in the forty to eighty base range; and simultaneously deleting two bases while converting a G-to-a-T a few bases away. That’s like watching Lionel Messi slalom his way through the opposition defense and around the goalkeeper before tapping the ball into the net. Or as Sharon Begley put it, “the genome equivalent of a pool shark’s banking the 9 ball off the 7 and sinking the 1, 5, and 6.”²⁶

Liu’s Cold Spring Harbor unveiling of prime editing was pretty much as close to a drop-the-mic moment as I’ve witnessed at a science meeting. He had given the meeting organizers a deliberately vague summary of his talk for the program book to preserve the element of surprise. I shook my head in astonishment as he flashed a pie chart showing the categories of human disease mutations, and said calmly that prime editing could in theory address 89 percent of them. Thirty minutes later, Liu wrapped up his talk by acknowledging Anzalone, with a smile: “I’m really looking forward to seeing what Andrew can do in the second year of his postdoc!” Doudna, who was sitting in the front row, later declared, “I literally had chills running down my spine” as she savored the latest power upgrade to the CRISPR toolbox.

The media reaction to prime editing was extraordinary, even overshadowing Google’s claim of “quantum supremacy” published the same week. Commentators and journalists gushed about this gorgeous new “CRISPR 3.0” technology. The breakthrough even caught Elon Musk’s attention, who retweeted a New Scientist story. Urnov was much in demand, obligingly dashing off a different analogy for each reporter who called. For Scientific American, prime editing was like a new breed of dog. For STAT, it was a new superhero joining the Avengers. For Genetic Engineering & Biotechnology News, it was a college sports star preparing to join the professional leagues. “We all hope of course it will be like Alex Morgan or Aaron Rodgers in this regard—and we should know soon.”²⁷

While Urnov speculated that prime editing could be part of an immunotherapy clinical program within a couple of years, a few commentators took issue with Liu’s eyebrow-raising estimate of 89 percent mutations that were potentially fixable. Liu is a fastidious scientist who is not inclined to hype his results—because he doesn’t need to. In a subsequent talk a month later in Barcelona, he politely but firmly pushed back. “This is a smart audience,” he told 1,500 gene therapy experts. “You know the difference between correcting a mutation and actually treating patients.”²⁸

But how could prime editing be used therapeutically? With protein and RNA components made up of thousands of atoms, the prime editing molecular machinery is too bulky to fit into the standard AAV vector. But Anzalone was able to use a lentivirus to perform prime editing in mouse cortical neurons. The scarcity of PAM sites is not a big issue because of prime editing’s greater flexibility with regard to the edit location. The prime editing window is much longer than traditional Cas9 editing, and the system has a lower rate of off-target effects. The system appeared safer than Cas9, for reasons that make sense: whereas CRISPR-Cas9 has just one base-pairing event (when the guide RNA aligns with the target sequence), prime editing has two additional pairing steps—the binding of the pegRNA to the target site and the pairing of the flap to the original site—which offer additional opportunities to reject an off-target sequence. “If any of those three pairing events fail, prime editing can’t proceed,” Liu said.

As with the earlier CRISPR genome editors, delivery will be a big challenge, but Liu was confident he could deliver prime editors into animals, for example by using a pair of AAVs (as used in the progeria mouse model). On the safety question, Liu stressed that all genome editors have off-target effects—chemical binding is an imperfect process, just as all prescription drugs have some sort of side or off-target effect. “Each platform has complementary strengths. All will have roles in basic research and therapeutics,” Liu said, mindful that prime editing could steal some thunder from his earlier platform companies, Editas Medicine and Beam Therapeutics. Indeed, by the time prime editing was announced, Liu’s latest company, Prime Medicine, had already been formed on paper, with funding from a Google fund and F-Prime, with some rights licensed to Beam.²⁹

Prime editing won’t be the last word in genome editing technology. I could mention Homology Medicines, which is patching in a full gene delivered by a virus to genomic targets to treat phenylketonuria without CRISPR. Or an Israeli start-up called TargetGene Biotechnologies, which modestly claims it is developing “the world’s best therapeutic genome editing platform.” Or Tessera Therapeutics which touts “gene writing” as the route “to cure thousands of diseases at their source.”³⁰ In July 2020, Liu unveiled another impressive riff on base editing, turning a bacterial toxin into a precise gene editor that could be delivered to mitochondria to edit mtDNA. Liu’s team used TALEs rather than CRISPR as the guide.³¹

Whether it takes five years or fifty, it seems inevitable that we will be able to engineer bespoke variants into the genome precisely and safely.³² The prospect of rewiring the genetic code to cure deafness or diabetes, sickle-cell or schizophrenia, is getting closer all the time.³³ But why stop there?

I. In this example, the C base editor targets a C:G basepair and deaminates the C to a U, resulting in a U:G mismatch. The cell’s DNA repair processes seek to repair the mismatch in one of two ways: either by switching the U back to a C; or by fixing the G to an A, thus creating a U:A, or T:A basepair. The goal was to push the system toward the latter, resulting in a C:G to T:A substitution.

II. Deamination of A actually yields inosine (I), but this is read as G.

III. Liu later found out that five groups had conducted the same initial experiment, fusing an RNA adenine deaminase, replacing our cytidine deaminase, but all five saw no editing. “No-one else made the crazy-sounding decision that we did to go ahead and evolve one.”

IV. With four bases, each in theory mutable to three other bases, there are a total of 12 possible base substitutions. The C and A base editors developed by Liu’s lab account in total for four substitutions (CBE catalyzes C-to-T and G-to-A; the ABE catalyzes the reverse substitutions, A-to-G or T-to-C), known as transitions.

V. When Anzalone googled “Cas9-RT fusion,” he learned that Cas1, which plays a role in capturing viral sequences in the CRISPR defense system, has a natural RT activity. This shows, not surprisingly, how nature uses similar concepts to prime editing but for different purposes.