How do base editors and prime editors differ from a standard CRISPR nuclease?

Why

Anzalone, Koblan and Liu group the tools into four classes with different capabilities. A nuclease is efficient at knocking a gene out, because random indels usually disrupt the reading frame, but precise changes need homology-directed repair, which is inefficient in cells that are not dividing. Base editors fuse a catalytically impaired Cas to a deaminase.

Cytosine base editors change C to T, and adenine base editors change A to G, within a small window of the protospacer. They are efficient and avoid double-strand breaks, but they cannot make every change, and nearby bases in the window may be edited too. Prime editors fuse a nickase to a reverse transcriptase and use a prime editing guide RNA that both targets the site and carries the template for the new sequence.

Anzalone and colleagues showed it can make all 12 base-to-base changes and small insertions and deletions. It is more flexible but usually less efficient and needs more optimization. The choice follows the edit.

For a knockout, a nuclease is simplest. For a known point mutation, a base editor often wins. For an insertion or a change a base editor cannot reach, prime editing is the option.

All three need verification of the actual edit at the DNA, RNA and protein level.