Before base editors and prime editors, the specificity problem in CRISPR had a blunter solution, and US10544433B2 - "Using RNA-guided FokI nucleases (RFNs) to increase specificity for RNA-guided genome editing," issued January 28, 2020 to The General Hospital Corporation - claims one version of it. The inventors (Joung, Tsai) fuse the FokI cleavage domain to a catalytically dead RNA-guided protein and require that two such fusions bind adjacent sites before a cut occurs.
The claim limitation that matters is dimerization. FokI cuts DNA only as a dimer; a single monomer sitting at an off-target site does nothing. By forcing the system to recruit two separately-targeted guides, the claimed method roughly squares the recognition requirement. An off-target match at one guide site is tolerated because the second guide must also match nearby and in the correct orientation.
Read against the CPC tags - C12N 15/907, C12N 9/22 - this is squarely genome-editing machinery, not a therapeutic-use claim. That distinction matters for scope: a machinery claim like this reaches anyone making or using the nuclease architecture, regardless of the disease indication, which is a broader net than a method-of-treatment claim but a more crowded one to defend against prior dimeric-nuclease art such as zinc-finger nucleases.
The genus-versus-species question here is unusually clean. The genus is 'paired nickase / paired nuclease' strategies for specificity; the species the claim stakes out is the RNA-guided FokI fusion specifically. A challenger's natural move is to argue that pairing FokI with an RNA-guided scaffold was obvious over the long-established FokI-zinc-finger pairing - which is exactly why the enablement detail in the specification, showing the RNA-guided version actually works at the claimed specificity, carries the weight.
What this early grant signals for the landscape is that the foundational CRISPR estate was never just the Cas9 composition fights everyone watched. Specificity-engineering patents like this one created a parallel layer of claims that any clinical editing program eventually has to navigate, independent of which institution won the headline Cas9 interference.