By Kevin E. Noonan --
Much of the gene patenting debate has appeared, to experienced patent practitioners, to be akin to a conversation with Kevin Kline's character in A Fish Called Wanda, or perhaps a passage from P.G. Wodehouse's The Luck of the Bodkins. In short, it has been the tale of doctors, scientists, and legal academics interpreting claims to isolated DNA in ways both bizarre and unlikely to anyone with any experience in actually claiming DNA. However, as Dr. Mason's response illustrates those of us on the patent side of the debate have contributed to the confusion.
Reviewing Dr. Mason's assessment of the patents and applications he identified as encompassing almost the entirety of the human genome reveals two sources of misunderstanding. The first is between the disclosure of patent specifications and the claims: while sometimes in their zeal to provide the broadest disclosure possible, patent practitioners have been overly expansive in defining the scope of the isolated nucleic acids that are the subject of their clients' inventions, the claims have typically been more circumspect. Most claims to isolated DNA have the canonical structure:
An isolated nucleic acid/human DNA molecule, encoding an amino acid sequence identified by SEQ ID NO. X.
Occasionally the word "comprising" is used in such claims:
An isolated nucleic acid/human DNA molecule, comprising a nucleotide sequence encoding an amino acid sequence identified by SEQ ID NO. X.
And there are claims to oligonucleotides, usually dependent on the claimed sequence:
An oligonucleotide comprising about 15 nucleotides of SEQ ID NO: X.
Considered in a vacuum, in the context of definitions as illustrated in Dr. Mason's response, it could be thought that such claims encompass a greater scope than they actually do. What is missing is context, both historical and legal, that provides the insights needed to properly understand the scope of these claims.
The base canonical form above was developed about thirty years ago, during the age when "cloning" a DNA molecule that encoded a protein of interest (erythropoietin, tissue plasminogen activator, blood clotting Factor VIII, interferon, etc.) was a Herculean feat, and such claims were expressly directed towards full-length versions of these genes. In the days before the Human Genome Project and the vast amount of genetic information it produced, such cloned genes had but one utility: to permit the production of recombinant cells that expressed the gene product, typically to be used as one of the first "biologic" drugs. In order for the claimed DNA molecule to have that utility, however, what was required almost always was that the full-length gene be isolated, because truncated proteins could not be expected to have the desired biological properties (and could have deleterious ones, such as increased immunogenicity or reduced half-life). Accordingly, such claims do not have the scope Dr. Mason ascribes to them, because any such claim would be invalidated for failing to have the necessary utility.
Moreover, the scope of such claims are extremely narrow, being limited to isolated DNA molecules that encode the specifically recited amino acid sequence identified as SEQ ID NO: X. An isolated human DNA comprising any change in this encoded amino acid sequence is not exclusive to the patentee and can be performed freely without literal infringement liability by anyone. (The availability of the doctrine of equivalents for isolated human DNA claims is severely limited by U.S. Patent and Trademark Office practice in interpreting the application of 35 U.S.C. § 112(a) to such claims, and by ensuing prosecution history estoppel.)
The significance of this limitation in scope of isolated DNA claims can be understood by the following example. If the single change in the amino acid sequence expressly recited in the claim is from a valine (Val) residue to an isoleucine (Ile) residue, the number of atoms in the protein would be increased by a mere 3 atoms (the difference in structure between these two amino acids is a methylene group,-CH2-) out a total of 50-100,000 atoms in the protein. Yet, even such a molecule changed to such a small degree would not literally infringe the generic claim. Similarly, insertions or deletions of the sequence, mutations and sequence variants (naturally occurring or man-made) do not fall within the scope of claims to a specific, particular isolated human DNA, due to this restricted scope. Such claims reciting such full-length "genes" are simply not preclusive of any other gene in the genome.
Turning to the oligonucleotide claims, there are two issues. First, the work of Bob Cook-Deegan and his collaborators have shown (Kepler et al., "Metastasizing patent claims on BRCA1") that oligonucleotides on the order of 15-16 nucleotide residues in length are much more ubiquitous in the genome than was once thought. This stems from the simplistic and incorrect assumption that any sequence of n nucleotides will occur once per 4n times in the genome. Thus, the random 15-mer was expected to arise once every 1,073,741,824 basepairs, and the random 16-mer once every 4,294,967,296 basepairs. The (false) conclusion was that in a haploid genome comprising 3 x 109 basepairs, the 15-mer and 16-mer (and larger species) would be unique. The source of the error is that human DNA is a historical molecule, and that many sequence motifs would be expected to arise many more times than random statistics would predict; the Human Genome Project (and basic biology) illustrates the fact that the human genome evolved by gene duplication and recombination, so that certain motifs (the ATP-binding cassette sequences, for example) sharing certain amino acid sequence and thus the nucleotide sequences encoding them, are much more common that was predicted. The consequence of this biological fact is not that claims to such oligonucleotide encompass the entirety of the genome. Rather, it is that claims of this scope are invalid, since the prior art is much more replete with these sequences than was expected.
The other distinction that should be appreciated is that what is disclosed (and even claimed) in patent applications does not raise infringement issues. Indeed, filing overbroad claims in International Applications can be a strategic decision, to produce prior art against competitors (this stratagem became available in the U.S. with publication of U.S. applications under the American Inventor Protection Act of 1999). Mere inspection of the published patent application database can find patents to a variety of "inventions" unlikely to ever grant with patentable claims, and their existence is thus not relevant to a serious debate over the scope of patents claiming isolated human DNA.
It is difficult to avoid the conclusion that a failure to appreciate these distinctions informs much of the angst enunciated by Dr. Mason and other physicians and researchers confronted by the prospect of infringement liability (or at least the fear that threats from patent holders will inhibit progress in genetics research). But the patent community is not entirely blameless either, as illustrated by Dr. Mason in his Rebuttal. There exists a certain amount of loose language in patent specifications that creates understandable confusion as to the claim scope the patentee believes she is entitled to, and thus uncertainty as to whether claims will be asserted as broadly as these specifications can be interpreted to support. (Of course, some have interpreted these claims to encompass isolated DNA itself, an interpretation that would render the claims unpatentable since the time of Friedrich Miescher's work in the 19th Century.) Some comfort should be taken by the fact that no gene patent holder has ever asserted claims to this extent, and that any attempt to do so would likely result in invalidation of the claims. And the patent case law has provided more and more illustrations of claims (including Myriad's method claims) being interpreted in view of these overbroad specifications and invalidated based on broad claim construction. In short, while there may be specifications suffering from symptoms of overbroad disclosure, the disease is self-correcting (even more so with the post-grant review and inter partes review provisions of the Leahy-Smith America Invents Act).
What is clearly needed is more dialog (and perhaps fewer assumptions) from both physicians and the patent community. This debate comes thirty years too late to be particularly meaningful to claims to isolated nucleic acids. But as the complexities of "modern" gene-based inventions increase (and their capacity to be ever more resistant to reverse engineering), the potential to eschew patent protection will increase, with the concomitant reduction in disclosure of biologically relevant inventions. Such a trend, and Supreme Court decisions that promote that trend, particularly if based on unrealistic apprehensions over the (vanishingly small) possibility that claims to isolated human DNA will negatively impact developing genetic technologies, could paradoxically promote the outcome that opponents of gene patenting purportedly are committed to prevent.