How closely do we have to look at an organism before we can say it is new? Under a close enough lens, nothing about a heavily edited organism is new, because most edits on their own are already established mutations, found either in nature or in a lab. But what if we zoom out and see that these edits have been combined, in a background genome, in a way that has never been done before? And what if we zoom out further to observe those genes interacting in ways that produce phenotypes nobody has previously discovered? Patent law seems to leave a grey area in determining the standard for newness. Should patent law assess whether a genetically modified organism is new through its individual genetic components or through the characteristics of the organism as a whole?
This article examines the patent eligibility of genetically modified organisms under 35 U.S.C. § 101, concerning the question of whether something is the kind of invention that can be patented at all. The statute provides: “Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.”[1] Courts have interpreted this statute to exclude products of nature, so an organism must be distinct from nature to qualify. This definition differs from that of novelty under § 102, which asks whether an invention already exists in the prior art. I propose that the USPTO adopt a clearer standard for which characteristic, and at what scale, determines whether an organism is patent eligible. If we don’t clarify this, a fundamental debate remains: is newness something an organism innately possesses, or something we subjectively decide it has depending on how closely we look?
Funk Brothers Seed Co. v. Kalo Inoculant Co. (1948) established that combining natural things does not make them patent eligible when the combination changes nothing about how each part behaves. Funk predates § 101, which was enacted in the 1952 Patent Act, but Funk utilized the product-of-nature principle that courts now apply under § 101. An inventor discovered that certain strains of root-nodule bacteria do not inhibit one another’s function when combined, and claimed the mixture. The Court held that this was only the discovery of a phenomenon of nature, and not an invention. The mixture produced no new bacteria, changed none of the species, and enlarged none of their uses.[2] Putting natural parts side by side was not enough, because each part behaved exactly as it was already discovered to behave.
By contrast, Diamond v. Chakrabarty (1980) reached the opposite ruling although every part of the invention came from nature as well.[3] The plasmids, each carrying a separate pathway for breaking down a different component of crude oil, had been found in natural Pseudomonas strains. Chakrabarty brought these plasmids together in a single host that could not degrade oil on its own, which in turn gave it more plasmids than any natural strain carries and a function natural Pseudomonas lacked. The Court held by a five-to-four vote that the bacterium was patent eligible because the patentee “has produced a new bacterium with markedly different characteristics from any found in nature and one having the potential for significant utility.”[3] Even though the genetic components were natural, the new functionality that came from combining them was enough to decide the case in Chakrabarty’s favor. The two rulings therefore did not turn on the parts, which were natural in both cases, but on whether combining them produced new functionality: Chakrabarty’s bacterium could degrade oil in a way no natural Pseudomonas could, while Funk’s mixture did nothing its bacteria could not already do alone.
Association for Molecular Pathology v. Myriad Genetics (2013) asks a different question of natural material: whether the claimed part differs in structure from what exists in nature.[4] A genetic testing company, Myriad Genetics, had discovered the location and sequence of the BRCA1 and BRCA2 genes. It claimed both the isolated genes and cDNA, a laboratory-made copy that contains only the protein-coding parts of a gene. A unanimous Court held that a naturally occurring DNA segment is not patent eligible merely because it has been isolated, but that cDNA is eligible because it is not naturally occurring. Isolation had severed the chemical bonds connecting each gene to the rest of the genome, yet the Court gave this no weight, because the claims were concerned with the information in the sequence. cDNA carries the same protein-coding information as the natural gene, but its structure differs from anything in the genome, and that structural difference alone made it new, with no change in function required.
Where Funk and Chakrabarty asked whether a combination as a whole had acquired a new function, Myriad focused on a single component and asked, under § 101, whether it was distinct from nature in structure alone. The law is therefore hard to predict. It recognizes more than one criterion for eligibility: function when a combination is at issue, and structure when a single part is. An edited organism is both, a combination built from parts, and nothing says which criterion should govern. On the logic of these cases, an organism whose natural edits interact in a new way is new under the function criterion but not under the structure criterion. An organism with edits that are structurally different from nature but add no new function is the reverse. For a researcher deciding which edited strain is worth building, this uncertainty makes it difficult to predict whether their inventions will qualify as patent eligible. The gap, then, is not a missing definition of “natural.” It is, instead, the absence of any rule saying which test of newness governs an edited organism, and at what scale it should be applied. The United States Patent and Trademark Office (USPTO) is the guidepost for the rules regarding patent eligibility in practice, because its examiners apply its guidance to every application. This means its guidance is the place to look for the missing rule, and the place where a fix would have to be made.
The United States Patent and Trademark Office (USPTO)’s current test asks whether a nature-based product has “markedly different characteristics,” such as differences in structure or function, from its “closest naturally occurring counterpart,” and where no single counterpart exists, the examiner selects one “based on the examiner’s expertise in the particular art.”[5] This principle is reasonable, but it leaves three decisions open for an edited organism, each of which can change the outcome: which characteristic to assess, which scale to examine, and which natural counterpart to use. It lists structure and function without saying which counts (structure, as for cDNA, or function, as for Chakrabarty’s bacterium). It says the counterpart for a combination “may be” its “individual nature-based components,” without saying when the whole organism is the right comparison instead. And it does not say which counterpart to choose among the wild type, a wild carrier of the same allele, or each edit on its own.
Consider a hypothetical worm strain with two edits that each already exist in wild worms, but that extend lifespan when combined. An examiner who compares the whole strain to a wild-type worm, and asks whether it has a new function, finds that it does, and the strain is eligible. An examiner who compares each edit to the wild isolate carrying it finds that each is natural, and the strain is not. The same claim receives opposite outcomes, and nothing in the guidance says which examiner is right. Wherever the guidance is silent, subjective discretion replaces principle.
I propose that the USPTO should replace that discretion with a concrete, objective rule. Where any edit in the claimed organism is non-natural, a structural difference should suffice, as it did for cDNA. Where every edit is natural, the claim should have to show a new function in the whole organism, as in Chakrabarty. In either case, the claim should recite the edits, the background genome, and the resulting phenotype, since differences a claim does not recite cannot count toward eligibility.[6] Without clearer guidance, researchers may face increasing uncertainty over how the patent eligibility of their inventions will be assessed.
One weakness of this proposal is that distinguishing between natural and non-natural edits could create another arbitrary standard. An edit is natural if the same genetic change has been found in wild organisms of that species, and non-natural if it has only been created or observed in the laboratory. This distinction sorts strains by where each edit has been found, not by how different the organism is from anything in nature phenotypically. A strain with one non-natural edit that has no effect on how the worm behaves would be eligible on structure alone, even though it behaves exactly like a wild worm. A strain with dozens of natural edits that together change the worm substantially would have to prove a new function, even though no wild worm resembles it. The strain that differs less from nature gets the easier path.
Because an edit is natural only if it has been found in the wild, its classification depends on how many wild worms have been sampled and sequenced. An edit that looks non-natural today could be found in a wild isolate next year, which would change the path the same strain must take. The outcome then depends on how thoroughly nature has been surveyed, not on what the organism is. That repeats the problem the proposal is meant to solve, where eligibility depends on how closely someone looks.
The rule could narrow this problem by fixing when and how an edit is judged. An examiner could decide whether an edit is natural by checking the recorded wild isolates available on the filing date, and the claim would state which records were checked. A researcher would then know the answer to the natural or non-natural question at filing. This does not remove the dependence on how much of nature has been sampled, but it turns an open-ended question into one with a stated date and stated evidence.
Whether an organism counts as new depends, under current guidance, on which characteristic an examiner decides to compare, to which counterpart, and at what scale, which means newness is partly a function of where the examiner chooses to stand. At the scale of a single edit, a heavily edited strain may be old, and at the scale of the genome, the same strain may be new. A clear rule would not settle whether newness is something an organism innately possesses. But it would make the law clarify what it is measuring and let researchers know before building a strain whether anyone will consider it patent eligible. Until the USPTO says which scale and which test governs, eligibility will remain less a fact about the organism than a judgment made about it, and the unpredictability of the law will burden researchers who must commit years to building a strain before they know whether it will be patent eligible.
Edited by Lauren Kim
Endnotes
[1] 35 U.S.C. § 101.
[2] Funk Bros. Seed Co. v. Kalo Inoculant Co., 333 U.S. 127 (1948).
[3] Diamond v. Chakrabarty, 447 U.S. 303 (1980).
[4] Association for Molecular Pathology v. Myriad Genetics, Inc., 569 U.S. 576 (2013).
[5] United States Patent and Trademark Office, Manual of Patent Examining Procedure. § 2106.04(c): The Markedly Different Characteristics Analysis, online at https://www.uspto.gov/web/offices/pac/mpep/s2106.html.
[6] In re Roslin Institute (Edinburgh), 750 F.3d 1333 (Fed. Cir. 2014).