The limitation at issue is the length of the guide. The General Hospital Corporation's grant US10119133B2, "Using truncated guide RNAs (tru-gRNAs) to increase specificity for RNA-guided genome editing" (issued November 6, 2018; CPC C12N 15/102), claims a counterintuitive move: shorten the guide RNA's complementary region and the editing becomes more specific, not less. The independent claim's recital of a truncated guide — typically shortened at the 5' end — is the inventive core.
Why this works mechanistically, and why the claim is constructed around it: a CRISPR guide RNA directs the nuclease to a matching DNA sequence. Intuition says a longer match would be more precise. The Joung-lab inventors found the opposite — that beyond a certain length the extra base-pairing tolerates mismatches, so a slightly shorter guide is less forgiving of off-target sites. The claim therefore recites a guide of reduced length relative to the standard, and that length reduction is the limitation a competitor either practices or avoids.
Construe the truncation precisely. The claim's value depends on how the shortening is defined — by absolute nucleotide count, by relation to a reference guide, or by the complementary-region length. A competitor using a full-length guide does not infringe; one using a guide within the claimed truncated range does. This is a claim where a single structural parameter (guide length) is the entire infringement question, which makes precise construction unusually decisive.
The family context reinforces the point. The same specificity-improvement concept appears across related General Hospital grants (the earlier US9567604B2 and US10415059B2 carry the same tru-gRNA title), a pattern of continuation filings building a small family around one inventive insight. For a freedom-to-operate read, that means clearing the concept requires checking the family, not just the single grant.
Place it in the landscape: this is an applied-layer specificity improvement sitting on top of the foundational CRISPR machinery, not a claim on CRISPR itself. It is exactly the kind of narrow, mechanism-specific grant that populates the modern gene-editing thicket — valuable precisely because off-target editing is the field's central safety liability, and a claimed method to reduce it is commercially meaningful.
The teardown conclusion: read this patent as a single-parameter claim. The invention is the truncation; the scope is the claimed guide length; infringement turns on whether a competitor's guide RNA falls within that length limitation. It is a clean illustration of how a precise, counterintuitive structural limitation — make the guide shorter — can be the whole of an inventive claim in the CRISPR specificity space.