A team at the Broad Institute of Harvard and MIT has developed a genome editing method that could, in principle, correct 90% of the roughly 75,000 currently known genomic changes that are associated with genetic diseases.

For now, the method is limited to use in cell culture. Lead investigator David Liu, Thomas Dudley Cabot Professor of the Natural Sciences at Harvard University, told reporters that it "is the beginning rather than the end of a long-standing aspiration in the molecular life sciences" to be able to edit the genome at will.

But Liu also noted that the two other widely used genome editing methods, CRISPR/Cas9 and base editing, have been used in hundreds of cell types as well as organisms "from bacteria to insects to fish to primates."

Furthermore, in the paper describing prime editing, "we report the use of two delivery techniques that are pretty widespread already" for CRISPR/Cas9 and base editing. Both those techniques, transfection and lentiviral delivery, were successful in delivering prime editing's components to cells.

Liu is the senior author of the paper, which appeared online in Nature on Oct. 21.

Prime editing joins CRISPR/Cas9 and base editing in the genomic editor's kit. CRISPR/Cas9, which edits by making double-stranded cuts in DNA, has been likened to molecular scissors, which are repaired through insertions and deletions. Base editors, which can directly correct four of 12 possible point mutations, are akin to a pencil. (See BioWorld, Oct. 26, 2015.)

In that framework, Liu said, "you can think of prime editors to be like word processors, capable of searching for target DNA sequences and precisely replacing them."

The idea behind the work, first author Andrew Anzalone told reporters, was to develop "another way of genome editing that didn't involve DNA breaks," which are useful to disrupt genes, but make precise editing challenging.

Instead, prime editors consist of two key components, an engineered protein and an engineered RNA.

The prime editor protein is a fusion protein, made up of a disabled Cas9, which can target specific DNA sequences but can no longer make double-stranded DNA cuts, and a reverse transcriptase, which can make DNA from an RNA template.

It is combined with a prime editor guide RNA or pegRNA, that both targets the desired DNA site and provides an RNA template of the desired edits.

The reverse transcriptase "directly copies, letter by letter, the part of the pegRNA that encodes the edited DNA sequence into the target DNA site." That is followed by replacement of the original DNA sequence on both DNA strands.

The team demonstrated that their method was able to make roughly 175 different edits, including all 12 possible point mutations between DNA bases, and precise insertions of up to 44 and deletions of up to 80 base pairs. They also showed that they were able to remove the four-base insertion that is often behind Tay-Sachs disease and the point mutation that underlies sickle cell anemia.

They used five different cell types in their paper, including the HEK293T cell line, a frequently used cell line in genome editing research studies, and primary mouse cortical neurons, which are nondividing cells and so can be challenging to edit.

Liu said that CRISPR/Cas9, base editing and prime editing "each have complementary strengths and weaknesses," and that all three "have or will have roles in basic research and in applications such as human therapeutics and agriculture."

One weakness that is shared by the three methods, and probably any and all future genome editing techniques that might be developed, is that "all have off-target effects, because the nature of chemistry, the nature of binding, is imperfect."

However, his team was pleasantly surprised to find that prime editing led to fewer off-target effects than Cas9 at known Cas9 off-target sites.

The team looked at 16 known off-target sites that are associated with four target sites they investigated, and found that while CRISPR/Cas9 genome editing led to off-target activity at all 16 events, prime editing led to demonstrable off-target effects at only four sites, and off-target editing of a frequency greater than 1% in only one of those sites – even though prime editing, too, uses a Cas9 guide RNA.

Liu said that this specificity, which he called "remarkable," was likely due to the fact that for successful prime editing, three independent DNA pairing events need to occur before any prime editing happens. In both CRISPR/Cas9 and prime editing, the Cas9 guide RNA has to recognize the targeted DA site. But for prime editing, the pegRNA primer must recognize its binding site, and the DNA synthesized by the reverse transcriptase needs to pair with the original genomic DNA.