Prime Medicine raises $315 million for search-and-replace CRISPR gene editing

Prime Medicine has raised $315 million to commercialize an invention from David Liu’s lab at Broad Institute of MIT and Harvard. The start-up is developing therapies based on a new class of CRISPR gene-editing tools called prime editors, which some scientists say could be the most versatile form of gene editing yet.

Liu and his postdoc Andrew Anzalone quietly founded the firm in 2019 before publishing their work on prime editors that October (Nature 2019, DOI: 10.1038/s41586-019-1711-4). The pandemic largely delayed Prime’s operations until the second half of 2020, CEO Keith Gottesdiener says. In the months since, Prime has grown to about 50 employees—most of them scientists—with plans to expand to about 100 by the end of 2021, he adds, buoyed by its newly disclosed $115 million series A and $200 million series B financings.

The earliest versions of CRISPR gene editing relied on using the enzyme Cas9 to make double-strand breaks in DNA—an action better suited to turning genes off than to fixing them. In 2016, Liu’s lab debuted the first base editor, which uses a modified Cas9 and other proteins to change a single nucleotide of DNA into another nucleotide. Two versions of base editors can make four kinds of nucleotide changes, but eight nucleotide changes were still out of reach.

Prime editors overcome that limitation. The editors are made from a reverse transcriptase enzyme—which transcribes RNA into DNA—tethered to a modified Cas9. The complex uses a special guide RNA template to search for a particular stretch of DNA and replace it with a new stretch encoded in the guide RNA. In 2019, Liu’s lab showed that prime editors could add, remove, or swap sequences of DNA.

“David Liu’s creativity and technical mastery of protein engineering are astounding,” says gene therapy expert Terence Flotte, dean at the University of Massachusetts Medical School. “Prime editing allows one to treat the genomic DNA as almost a blank canvas of possibilities.”

Prime Medicine makes similarly bold claims. “We estimate that prime editors can address more than 90% of disease-causing mutations and even crack multiple mutations at once,” Gottesdiener says. “And it also seems to work in just about every organ or type of cell that we look at.” Delivering the large prime-editing complex into the human body may pose a challenge, but Prime could benefit from its partnership with Liu’s base-editing company, Beam Therapeutics, which has a head start on designing delivery vessels for its editors.

Prime won’t yet disclose specific diseases it is working on, but Gottesdiener says the company has active programs to use the technology in eye, liver, and neuromuscular indications and in the blood-forming hematopoietic stem cells.

Prime’s hefty funding has come just weeks after Intellia Therapeutics published promising, albeit early, clinical data from a study using CRISPR gene editing to treat a rare, genetic liver disease in a human. That data, plus early results from a sickle cell study from Crispr Thera­peutics and Vertex Pharmaceuticals, demonstrate “the potential power of these approaches in treating, if not curing, real unmet medical needs,” says Nessan Bermingham, who is currently CEO of Triplet Therapeutics and is also cofounder and former CEO of Intellia. He expects to see “continued growth” in the gene-editing industry.

Most gene-editing insiders don’t think that prime editing will replace earlier forms of gene editing anytime soon. Fyodor Urnov, a gene-editing researcher at the University of California, Berkeley, says that with more than 5,000 distinct genetic diseases, “there is ample room for many technologies to succeed.”

“So far, when prime editing and base editing can do the same thing, base editing has proven to be much more efficient,” says Sekar Kathiresan, CEO of Verve Therapeutics, which is developing base-editing therapies for heart disease. “The primary advantage of prime editing is that it can do things that base editing cannot do at all and that Cas9 nuclease usually cannot do well. For the foreseeable future, different tools will be best at different tasks and have different efficiencies for those tasks.”