Since 2013, CRISPR has enjoyed celebrity status as the revolutionary gene-editing technology that could change everything. So you may be wondering—why haven't you heard of gene editing actually making an impact on human disease? People might disagree on how okay it would be to choose the eye color of your offspring, but there are lots of editing applications most would agree we should try, like on devastating illnesses known to be caused by genetic mutations, such as cystic fibrosis or sickle cell anemia.
The answer, in short, is that while CRISPR is a great tool for targeting and knocking out genes, when it comes to making precise changes, it could use a little help.
Now it may have gotten some. In an announcement today from Harvard, scientists say they are launching a company called Beam Therapeutics, which will be the first to pursue therapies using a new, more exact technique called base editing. It could be the first step to translating this type of gene-editing technology into treatments for human illness that are caused by small genetic mutations.
As a refresher: CRISPR is a defense that some bacteria have to target and cut the DNA of invading viruses. The labs of Jennifer Doudna at UC Berkeley and Feng Zhang at MIT and the Broad Institute showed around the same time that this system could be used to cut a piece of human DNA at a desired spot. Gene editing suddenly became way easier than it used to be, and the applications were enticing: cancer treatments, malaria-free mosquitoes, and other potential cures for genetic diseases.
But while the community at large debated over the ethics of editing germ line cells or designer babies, scientists were trying the possibilities a reality. "At a recent conference, it was pointed out that the field is hard at work, trying to make possible what most people—whose experience with genome editing is watching movies or reading casual pieces about the field—think is already possible," says David Liu, the Director of the Merkin Institute for Transformative Technologies in Healthcare at at the Broad Institute of Harvard and MIT.
In the past two years, Liu has been publishing his research on another method of gene editing: Rather than cutting out sections of DNA, this technique can edit a single base pair, offering more precision than the original CRISPR model.
Your DNA is composed of billions of nucleobases: adenine, cytosine, thymine, and guanine, called A, G, T, and C for short. A is always paired with T, and C is always paired with G. A “point mutation” is when just one base pair has changed. More than 60,000 human genetic changes are currently associated with disease, Liu tells me. The majority of these—currently about 33,000—are point mutations.
“If we want to go in and correct those genetic mutations, we need an efficient way of making precise single base changes,” says Holly Rees, a graduate student in Liu’s lab.
In the CRISPR process that's well known, an enzyme called Cas-9 cuts both strands of the DNA helix at a specified point. This cutting process is very exact, but the repair job isn’t. In response to the cuts, a cell can make small random insertions or deletions around the cuts during the process of joining together the two dangling ends of DNA. The good news: this often leads to the gene that you’ve cut being knocked out—it’s gone. But what if you don’t just want to disrupt a gene, you want to fix it instead?
Scientists have tried to give a cell a piece of DNA along with CRISPR—the DNA they want to be inserted. When they put it in the cell, sometimes in the repair process, that piece of DNA will be used. But this rarely works—at best, only about 10 to 15 percent of the time, Rees says. “And that's in cultured cells in a dish, under optimized conditions,” she points out. Not only are you not reliably getting the replacement that you want, you also usually get those random insertions or deletions as well—leading potentially to more problems.
“The useful analogy I like to make is that CRISPR/cas9 is like molecular scissors, while base editors are like pencils,” Liu says. “And it's important to point out that for some applications, scissors are the best tool for the job, while for other applications, like fixing a single letter in DNA, a pencil is best.”
The C base editor is made of three different components, all linked together with floppy pieces of protein. The first part is called a deaminase enzyme (deamination means it removes an amino acid) and this is the enzyme that actually converts the C to a T.
The deaminase enzyme is attached to a modified Cas-9; the same enzyme from CRISPR that is good at finding and sitting at specific positions within the genome. Once Cas-9 gets to its target, instead of cutting, it unravels the DNA, and allows the deaminase enzyme to do its job. Meanwhile, it "nicks" the unedited strand of DNA, which causes the cell’s repair process to change the G to an A (rather than changing back the newly made T). The third protein attachment is another that stops the cell from undoing the correction.
It’s a bit like when the Transformers combine together to make one big machine—the C editor is made of three pieces that each have their own functions, but together join to make a very reliable edit from C to T, leading to the cell to fix the G to an A. This final result is that base editor changed the C:G base pair into a T:A base pair.
In 2017, Liu and his collaborators published another base editor called ABE, which changes an A to a G, and its T to C. This was where things got crazy: Changing an A to a G was going to be complicated, because there wasn’t an enzyme already out there that did that. (In the C editor from 2016, the deaminase enzyme that changed C to T already existed in nature, so the team could assemble the pre-existing pieces together.) “We made the decision that we would go ahead and try to evolve the first known example of a DNA adenine deaminase enzyme, even though there doesn't exist an example in nature,” Liu tells me. “That was a really non-trivial decision because the lack of a known example in nature that does that reaction made success not obvious at all. We certainly considered it to be a fairly high-risk project. Normally we shy away from projects in which the first step requires evolving a whole other protein that doesn't exist in nature, because there's quite a bit of risk associated with step one.”
Nicole Gaudelli, then a postdoctoral fellow in Liu’s lab, led the evolution, beginning with an enzyme that could deaminate adenine, but in RNA instead of DNA. She created massive libraries of millions of different RNA processing enzymes in bacteria that they hoped would eventually act on DNA.
Then she introduced an antibiotic so that the cells could only survive if their base editors worked on DNA and could make a mutation in the antibiotic genes. "It was a really well-designed system but it took a lot of very large plates of bacteria to work,” Rees says.
I ask Liu: If we can evolve enzymes to do whatever we want, why use pre-existing enzymes in the first place? Can't we just evolve whatever we might need? He assures me that this is not an easy feat. There’s a saying in their group: “Nobody is that lucky,” meaning it wasn’t luck that brought forth the right enzyme from Gaudelli’s evolutions. “The way that one succeeds in these evolution experiments is to be insightful enough and creative enough to create a situation where the solutions make themselves apparent,” Liu says.
It took seven rounds of evolution, and three or four different evolution systems before Gaudelli found one that worked. Evolving an enzyme to do a job can come with unforeseen side effects, traits the enzyme inherits that you might not want it to, or unexpected limitations that come from it evolving in the lab.
Once they had a working enzyme that could change an A to a G, they attached it to the modified Cas-9 and had their second base editor. In 2017, they tried ABE against hereditary hemochromatosis, a liver condition that's one of the most common genetic disorders in Caucasian people, and caused by a point mutation. In cells that were cultured in the lab but from a patient with the disease, they found that ABE could correct the mutation in the cells.
Liu says that many other labs have used base editing to install or correct mutations in a variety of organisms, like bacteria, fungi, rice, wheat, corn, tomatoes, insects, fish, frogs, mice and even in embryos. “Those studies have established that base editing is really a robust and widely applicable technology, and we've been really excited by the fact that it's has been so quickly adopted by other researchers,” he says.
Between both base editors, the four kinds of changes that base editing can make could, in theory, correct more than 60 percent of known pathogenic point mutations. Because of this, and the precision base editing offers, Liu says he feels an obligation to get it to patients soon. As with CRISPR, delivery will be an important discussion going forward. It’s one thing to show in the lab that gene editing works, but how to actually get it safely and effectively into the human body, to the exact area you want to treat?
“The base-editors are very large protein and RNA machines that have to somehow not only get into the cell, but actually get into the cell nucleus,” Rees says.
John Evans, the CEO of Beam, says that they can follow the lead of other gene-editing delivery methods, and that they’ll be trying them all: removing cells, treating them and putting them in to the body (called ex-vivo delivery), or ways to get the editors directly into the body through nanoparticles or viral vectors.
Liu and Evans say that since there are so many diseases that could benefit from base editing, they have some guidelines for which they’ll tackle first. Do they believe that a single base change will reverse some of the consequences of the disease? Is the genetic sequence itself a good fit for a base editor to bind to and make the edit? And, is there a delivery method that they can get to the target tissue with? “We hope to fulfill our obligation to patients by trying to develop as many promising therapeutics as possible," Liu says. "But we also will have a very structured way of going about identifying the most promising opportunities from this pretty large venue.”
Beam will also investigate another base editor, developed by Feng Zhang around the time that ABE was published. It edits A bases, but in RNA. The concept is similar: a deaminating enzyme is attached to a CRISPR enzyme, but this time Cas-13, an enzyme that acts on RNA. RNA is the middleman between our genes, and the proteins that they produce. By editing the RNA, Zhang showed you could change the gene expression, or what protein is made, without changing the bases of the DNA itself.
Liu says this could be especially useful when you want to make a temporary change. “For some acute applications, in which a transient rather than a permanent edit is desirable, RNA-based editing could prove to be an ideal approach," Liu tells me. "For other applications in which a one-time permanent correction of the genome is desired, such as the many lifelong genetic diseases caused by point mutations that we've already hinted at, DNA based editing could be ideal. So together, these technologies we view as perfectly complementary, and together they make the development of base editing especially exciting.”
It’s too early to say just when the base editors will be tried in people, but Evans says he’s optimistic they can get it to patients relatively quickly. Because base editing is so precise, the researchers are hopeful that it will be easier to monitor, test, and develop than its predecessor— at least for point mutations. It won't replace traditional CRISPR, but just add another dimension to our gene editing capabilities, so that we can perhaps start to see it make a difference in people soon.
Sign up for our newsletter to get the best of Tonic delivered to your inbox.