RegenxBio v. Sarepta: a friendlier Section 101 landscape for engineered biologyFebruary 23, 2026

On February 20, 2026, the Federal Circuit reversed a Delaware district court and held that claims directed to a cultured host cell containing a recombinant nucleic acid molecule that encodes an AAV capsid protein sequence and includes a heterologous non AAV sequence are patent eligible subject matter.

While this specific case speaks to gene therapy, the lesson is more broadly applicable. Agricultural biotech runs on recombination, including assembling genetic parts from different sources, inserting them into plant cells or microbes, and letting biology do the manufacturing. Eligibility disputes often start when someone reframes that work as merely finding or using a natural sequence. RegenxBio provides a clean rebuttal, grounded in science.

Using the Chakrabarty “markedly different characteristics” framework, the court focused on what the claim required as a whole. The representative claim was not written as “an isolated AAV sequence,” the framing that caused issues in Myriad. It required a cultured host cell and, inside that cell, a recombinant (engineered by humans) nucleic acid molecule. The claim also required a heterologous sequence, meaning genetic material from a different species. The panel emphasized that this recombinant molecule does not and cannot exist in nature, and the host cell containing it likewise does not exist in nature. That structural reality is key because it places the claim on the Chakrabarty side of the line: a product of human ingenuity with a distinctive character and use.

The opinion also rejects an overreach from Funk Brothers. The district court treated combining two natural sequences in a host cell as no different from mixing naturally occurring bacterial strains together in one package. The Federal Circuit held that this analogy breaks down. A mixture of unchanged organisms is packaging, however, engineering sequences from separate species into a single molecule and transforming a cell so it incorporates that molecule is construction. For ag biotech, for example, that is the difference between putting ingredients in the same jar and building a pathway the cell can execute.

From a patent strategy standpoint, the court also refused to ignore “conventional” claim limitations when evaluating whether the claimed composition is markedly different from what occurs in nature. Sarepta urged the court to focus only on the natural AAV sequence and treat the rest as routine context. The panel declined, reminding everyone that eligibility is not novelty and it is inappropriate to strip a claim down to a single natural component. If the other elements are not novel, that will be handled via 102/103 arguments.

Key Takeaways

For patent prosecution this validates practice that many are actively utilizing and in claim drafting, including claims that make human intervention unavoidable. Continued emphasis on recombinant constructs, heterologous cassettes, and transformed cells, seeds, or microbes rather than a naturally occurring sequence in isolation is best practice. Use the specification to explain, step by step, why the claimed composition cannot arise without deliberate assembly and introduction into a host cell. During prosecution, when Section 101 comes up, the arguments should be centered on the claimed composition as a whole.

If your innovation truly constructs a non-naturally occurring biological composition, RegenxBio is a strong, scientifically literate precedent to keep handy.

Cassie J. Edgar is a Partner and Patent Attorney. She is Chair of the Licensing and Regulatory Law practice groups and advises clients in IP, regulatory law, and licensing including matters with USDA, FDA, and EPA. Cassie is also Co-Chair of the Data Privacy and Cybersecurity practice group. For additional information, please contact Cassie directly via e-mail at cassie.edgar@ipmvs.com. Please seek consultation for specific inquiries as this publication provides overview data only and does not provide legal advice.