In re Gleave, a recent decision by the Federal Circuit, affirmed a PTO anticipation rejection under 35 USC 102(b) of claims reciting antisense oligonucleotides directed against Insulin-Dependent Growth Factor Binding Protein (“IGFBP”). The facts of the case, and some implications of the decision, are discussed in a blog post on Patent Docs, available here. I'd like to follow up on that insightful post with a few comments of my own.
Basically, in Gleave the Federal Circuit held that a laundry list of oligonucleotides appearing in a printed publication, in this case a published PCT application, anticipates all of the listed oligonucleotides, even oligonucleotides that have never actually been made and for which there is no disclosed utility. All that is required is an enabling method of making the oligonucleotides, such that one of skill in the art would be able to arrive at a method for synthesizing the molecules without having to engage in undue experimentation. Today, in an era when full-length genes are routinely synthesized chemically, and even whole viral genomes have been artificially synthesized, it would seem that the mere listing of any conventional oligonucleotide, or even long polynucleotide comprising hundreds or even thousands of bases will anticipate and thus render unpatentable the molecule. For more on the synthesis of synthetic genes and genomes, see, for example, Synthetic Genomics: Options for Governance, or websites of gene synthesis companies such as DNA 2.0.
In Gleave, the prior art PCT application listed all of the possible 15 base oligonucleotides (15-mers) that appear in the full-length IGFBP gene sequence (see the Patent Doc post for a more detailed description of the disclosure). But according to the rationale of the Federal Circuit decision, which by the way is entirely consistent with earlier precedent, a prior art reference listing all possible 15-mers (i.e., the 4^15 possible combinations of the four nucleotides - adenine, thymine, cytosine and guanine (A, T, C and G)), would appear to bar the patenting of any and all 15 base oligonucleotides. And why stop there? In principle, a laundry list of all possible polynucleotides made up of the four naturally occurring nucleotides up to a certain length, say 1000 bases, should create patent invalidating prior art with respect to all the sequences, assuming that one of skill in the art would be able to synthesize all the sequences. The number of sequences would be huge, and I would guess there are not enough trees on the planet to provide sufficient paper, but electronic media, such as a reasonably accessible internet posting, can constitute prior art, and perhaps there are enough electrons to create this blockbuster publication which would prevent the patenting of all polynucleotides up to this length.
Professor Andrew Chin, of the University of North Carolina School of Law, specifically addressed this issue in a law review article entitled “Artful Prior Art and the Quality of DNA Patents,” available here. In his article, Professor Chin describes his creation and publication of a CD-ROM containing a list of more than 11 million oligonucleotides (apparently all 12-mers and shorter) that he felt were clearly enabled because they were “the easiest to make and most versatile to use.” For example, “[o]ut of an abundance of caution, the list was restricted to the oligonucleotides that are least likely to form secondary structures.” To my knowledge, the technology has advanced to the point where one of skill in the art could make essentially any oligonucleotide without engaging in undue extermination so, I think in this regard, Professor Chin was probably more conservative than he had to be. The CD-ROM was shelved in the North Carolina School of Law library’s non-circulating reference collection, was indexed under a call number and added to the university's online catalog. The document was subsequently listed in the International Online Computer Library Center. Thus, the CD-ROM likely constitutes a printed publication under sections 102(a) and 102(b).
Gleave and Professor Chin’s article raises questions with respect to the validity of composition of matter claims directed to oligonucleotides, and even longer polynucleotides. For example, some patents claim any fragment of a given polynucleotide sequence up to a certain length, e.g. every 15 base segment. Would such a claim be barred by an electronic publication listing all possible 15-mer oligonucleotides? Under Federal Circuit precedent, and the logic of Gleave, it would seem so. And by extension, any claim to a longer polynucleotide would be anticipated by a reference providing a laundry list of longer polynucleotides. Gleave provides no indication of any upper length limit on patent invalidating prophetic descriptions of polynucleotides.
But does this interpretation of Section 102 make any sense? Professor Chin’s CD-ROM provides no technical contribution to the prior art, it is the equivalent of the disclosure of “all possible oligonucleotides of 12 bases or less.” Similarly, a laundry list of all possible polynucleotides up to 1000 bases in length provides no additional quantum of disclosure over a publication that simply discloses “all possible polynucleotides of 1000 bases or less.” For those opposed to the patenting of DNA sequences, Gleave and Chin’s article provides a roadmap for creating invalidating prior art that would appear to block the patenting of these molecules. But if that is the case, it implies that a publication that teaches absolutely nothing to one of skill in the art can dramatically impact the patentability of a large class of important chemical compounds. It also exemplifies the sometimes illogical outcomes when the courts applies conventional chemical patent law to polynucleotides and proteins