Base editing is the modality that rewrites a single DNA letter without breaking both strands, and US11214780B2 - "Nucleobase editors and uses thereof," issued January 4, 2022 to the President and Fellows of Harvard College (inventors include David R. Liu and Alexis Komor) - claims the architecture that makes it possible.
The load-bearing limitation is the fusion. The claim recites a nucleobase editor in which a deaminase enzyme is tethered to a Cas protein. The Cas component supplies programmable targeting via a guide RNA; the deaminase does the chemistry, converting one base into another. Crucially, the Cas component is impaired so it nicks or does not cut - there is no double-strand break.
Why is the absence of a cut the whole point? Double-strand breaks invite error-prone repair and unwanted insertions or deletions. A base editor sidesteps that by chemically editing the base in place. The CPC tagging captures the dual nature: C12N 9/22 (Cas endonuclease class) for targeting, C12Y 305/04005 (a deaminase enzyme class) for the chemistry.
The genus-versus-species framing is foundational here because this is close to a genus-defining claim for the platform. The claimed editor is a class of deaminase-Cas fusions, not one specific construct. That breadth is what makes the patent valuable and contested: any base-editing program has to ask whether its specific editor falls within the claimed fusion architecture.
For the landscape, the Harvard / Broad base-editing estate became one of the most strategically important clusters in gene editing precisely because it claims the architecture rather than a single editor. Companies built on adenine or cytosine base editing - Beam Therapeutics among them - operate downstream of foundational fusion claims like this one, which is why the licensing around the platform is so closely watched.