About 15 years ago, David Liu, PhD, a young chemistry professor at Harvard University, launched an unusually profitable extracurricular activity. Every few months, he’d accompany a handful of students to Las Vegas where, using specially developed math systems for counting cards, his “blackjack ninjas” would put into action “absurd sums of money.” After taking the Sunday night JetBlue red-eye flight back to Boston in time to teach class, a weary Liu would sometimes question whether he was making the right career choice. But should he give up cards…or chemistry?

Fortunately for Harvard’s chemistry students (and several casinos), Liu decided to take a blackjack sabbatical. The field of genome editing also has cause to be grateful: in November 2013, Liu and several genomics luminaries—including Feng Zhang, PhD, a core member of the Broad Institute, and J. Keith Joung, MD PhD, professor of pathology at Harvard Medical School and a research scholar at Massachusetts General Hospital—co-founded Editas Medicine, which became the first publicly traded biotech using the now-ubiquitous CRISPR-Cas9 genome editing technology.

At the same time, however, a new gene editing technology was beginning to take shape—in Liu’s email inbox.

Alexis Komor and Nicole Gaudelli
Alexis Komor (left) and Nicole Gaudelli developed the first base editors in David Liu’s laboratory at Harvard University. [Kevin Davies]
The genesis of base editing (BE), as it is known, is a remarkable story of tenacity and serendipity. The credit belongs to two talented postdocs in Liu’s laboratory—Alexis Komor, PhD, and Nicole M. Gaudelli, PhD—who devised a pair of molecular machines that can correct the majority of known disease-related mutations in the human genome. The potential of this technology prompted Liu, Zhang, and Joung to establish new company, Beam Therapeutics, which has already raised more than $200 million to turn BE into new therapeutics.

Base editing “is not a technology that is going to completely overtake CRISPR,” says Komor, who is now assistant professor at the University of California, San Diego. But it is complementary. “Not only can base editing be used as a therapeutic to correct disease-relevant point mutations,” continues Komor, it can also “introduce these point mutations into the genome of live cells” for myriad applications.

Gaudelli, currently a senior scientist at Beam, says her company is looking to transform BE for patients, “to get it into disease-relevant sites and make it work in the real world. Base editors can be a superior tool for correction of very debilitating genetic diseases.”

The creation of these BE entities is an impressive feat, for as Liu tells GEN: “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.”

Two key papers in Nature
Two key papers in Nature heralded the first base editors. One paper (April 2016, left) described the work led by Alexis Komor on the CBE; the other (October 2017), the work led by Nicole Gaudelli on the ABE.

On paper, the first two base editors—the cytosine-to-thymine or (C-to-T) base editor (CBE) and the adenosine-to-guanosine (A-to-G) base editor (ABE), designed by Komor and Gaudelli, respectively—can address the majority of known pathogenic point mutations in the human genome. But we’re getting ahead of ourselves. The genesis of BE had little to do with precision medicine or curing disease. The real story is, if anything, much more interesting than that.

The first base editor

In 2013, Caltech chemistry PhD student Alexis Komor interviewed for a postdoctoral position in Liu’s lab. To fulfill a Caltech graduation requirement, Komor had to outline three future research projects. She decided to include one of several project ideas (“white papers”) she’d been discussing with Liu via email, a process he refers to as “mutually guided brainstorming.”

Komor’s initial idea was to evolve a ribonuclease to degrade a specific sequence of RNA. Liu suggested she think about DNA-based editors, in particular the CRISPR-associated nuclease, Cas9. “If you could program a specific A-to-G (for example) change in the human genome,” he emailed her on November 1, 2013, “you could really transform genome engineering and possibly human therapeutics.”

Komor didn’t know much about Cas9, but by the time she arrived at Harvard 10 months later, she was up to speed. One of the first people she met in the lab was Gaudelli, another recently arrived postdoc working on an unrelated project. They were to become fast friends.

Building the first BE molecular machine was a multistep construction challenge.
Komor selected cytidine deaminase, an enzyme that converts cytosine to uracil but works only on single-stranded DNA. She coupled the deaminase to an inactive (“dead”) form of the Cas9 nuclease, which was used to seek out a target sequence in DNA and unspool a short stretch of about five base pairs.

After about six months, she had a prototype BE working, but she faced two major hurdles: to get the BE working in mammalian cells, and then override nature’s DNA repair processes, which would detect the U mismatch and revert the mutation. “[We knew the cell would] change it to something else,” Komor recalls. “The question was, how [could] we force the cell to go in the AT direction rather than just back to a CG base pair?”

Adding a third component—an inhibitor of uracil DNA glycosylase to squelch the natural repair process—helped tip the balance, but not as much as she hoped. Then one day, while talking to a colleague in the break room, she had an epiphany. “It just came to me…we’re working with an endonuclease! We’ve inactivated [Cas9] so all it does is bind [DNA],” but replacing one key amino acid would restore a nickase function that would clip one strand of the double helix.” Komor suddenly saw a way to nick the C-containing strand (leaving the uracil intact) to trigger DNA repair of the C rather than the uracil nucleotide.

When Komor told Liu about her brilliant idea, he started swearing: he’d wanted to start writing up the paper. Komor spent Christmas 2015 at home in southern California revising the manuscript, even foregoing her high school reunion. The paper was eventually accepted and published by Nature in April 2016.

ABE evolution

With Komor’s success in engineering the transition of one pyrimidine (C) to another (T), Gaudelli became increasingly distracted about the complementary transition of purines, A to G. After much debate, she abandoned her original project and committed to developing a base editor that would in principle reverse the most common point mutation found in genetic diseases. There was just one problem: there was no naturally occurring adenine deaminase that worked on DNA for Gaudelli to begin building the new BE.

At a crossroads, Gaudelli decided to “double down [and] do the crazy thing!” She broke a golden rule in the Liu lab and set out to evolve a DNA-based ABE. To start, she selected tadA, an adenosine deaminase that works on transfer RNA. Using error-prone PCR, she built gigantic libraries to screen for the desired activity. To her delight, the most significant mutation in the enzyme occurred at the precise residue making contact with the hydrozyl group of the ribose sugar. “[It’s] the one thing that’s different between what I’m trying to make and what I started with,” Gaudelli recalls thinking. She sent a quick slide to Liu—who began swearing again. “Holy —, this is our smoking gun,” he emailed back.

Gaudelli’s initial attempts at editing turned out negative except one, at a sequence that serendipitously resembled the natural enzyme substrate. The evolved tadA enzyme deaminates adenine on the exposed, single-strand-DNA, forming inosine, which is read as a G after DNA replication. After further refinement, the paper describing ABE was also published in Nature in November 2017.

(CBE) and adenosine base editing (ABE)
Running the bases: Molecular reactions that underlie cytosine base editing (CBE) and adenosine base editing (ABE). (Top) The CBE catalyzes the transition of cytosine to uridine. It consists of a single-stranded cytidine deaminase linked to Cas9 nickase (Cas9n). The uracil glycosylase inhibitor (UGI) domain prevents the U:G mismatch from being repaired back to a C:G. The nickase clips the opposite non-edited strand so it is the U that ultimately gets repaired to T. (Bottom) The ABE catalyzes the transition of adenosine to guanine. An evolved adenosine deaminase is linked to Cas9n. The single-stranded target A is deaminated while the non-edited strand is nicked by Cas9n, resulting in the A ultimately being repaired to G. [A. Komor]
“We have all the easy mutations” now, says Komor. The harder task of developing BEs for transversions—a purine to a pyrimidine or vice versa—will require some “creative thinking.”

“In the end, there won’t be one uber-base editor that fits everybody’s needs,” predicts Liu. “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.

Not surprisingly, academic interest in base editing is surging. The nonprofit organization Addgene has already distributed more than 6000 BE plasmids to more than 1000 investigators around the world. And Synthego offers chemically modified, synthetic single-guide RNAs that are fully compatible with base-editing nucleases. “They can be used with purified base-editing nucleases as ribonucleoproteins, with base-editing mRNAs or with plasmid-based base-editing nucleases,” says chief science officer Kevin Holden, PhD.

Positively beaming

Liu considered licensing his BE intellectual property to Editas or one of the other gene editing companies, but he ultimately decided to build a new firm. We wanted every type of editing technology to be given the full attention of a company so that it has the best possible chance of benefiting patients,” Liu says.

The Beam team
The Beam team: (left to right) CEO John Evans and co-founders David Liu, Keith Joung, and Feng Zhang.

Unlike the contentious (and expensive) CRISPR-Cas9 patent saga, the BE intellectual property (IP) estate appears relatively clean. “I think everybody’s goal should be for the IP estate to be as boring as possible, because we all want to spend as much time and effort moving technologies to patient benefit rather than fighting over IP,” says Liu. “We were fortunate that it evolved out of one institution rather than appearing on multiple coasts,” concurs Beam Therapeutics CEO John Evans.

Naming the start-up was a breeze. Liu mentioned to his friend, Agnieszka Czechowicz MD, PhD, a pediatric oncologist at Stanford, that he was trying to come up with a company name. Five minutes later, she texted 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.

Beam has grown from 10 to 70 employees in less than 12 months. “For many years, we sought to develop methods that would enable us to robustly induce targeted single-base changes,” says Beam co-founder Joung. “It’s exciting now to have the capability to make precise base alterations with such high efficiencies.”

Ten small steps

Evans won’t disclose which specific diseases Beam is focused on (“I have a long list of things I won’t tell you”) but confides there are about 10 programs under active investigation. “In cells, at least, we can see therapeutic levels of editing,” he reveals. The successful Series B round ($135 million) will provide resources so that “as many programs as deserve to go forward can,” at least until Beam moves into heavier phases of manufacturing and clinical development.

Beam has the distinct advantage that it won’t be spending years screening for the ideal small molecule. “In Phase I, we’re not just testing safety and dose, we’re also testing efficacy and biomarkers,” Evans says. “It sets up a relatively fast path for getting approval.”

For now, Beam is agnostic on delivery technology. “We’ve made an important choice: we’re investing in all delivery modalities in parallel,” Evans says. That includes ex vivo approaches for targeting blood cells and T cells as well as in vivo programs, using either lipid nanoparticles (to target liver) or viral vectors, including adeno-associated virus, for other tissues including eye, muscle, and the central nervous system.

And while delivery of these molecular machines won’t be straightforward, as Evans points out, “It’s still a single protein, a fusion protein. It will fold into that multidomain [structure]. We need to deliver a guide and the editor.” And BE can be effective without the creation of indels “in either dividing or nondividing cells” such as neurons or retinal cells.

Evans is closely watching neurological gene therapy companies like AveXis and Voyager Therapeutics. “Anywhere they can go, we want to come with a base editor inside to do something very different,” he declares.

Hanging over the entire genome editing field is the alarm over the CRISPR baby controversy that erupted in late 2018. (See Davies, K. GEN January 2019.) “It would be unfortunate if the outrage from that episode impeded legitimate nongermline and nonclinical research on genome editing, because the field has so much promise,” Liu says.

Despite that rich promise, Evans and his colleagues marked the Series B infusion in modest fashion, with cake, balloons, and a visit from Liu, during which he reflected on the early days of base editing. A proper Vegas-style celebration might be in the cards one day, but there’s still a great deal more to do.

Previous articleSigns, Portents, and Immuno-Oncology Biomarkers
Next articleCrafting a More Efficient CAR T-Cell Industry