As the methodology of genomics finds broader applications, it influences discoveries in many different fields. One such field is structural genomics, or the analysis of protein structure and function.
Until recently, scientists attempted to study functional similarities by identifying proteins responsible for particular functions, presumably by producing analogous structures within those proteins. Structural similarities were confirmed or disaffirmed later after years of research.
However, proteins with very similar functions and presumably similar structures exhibited little amino-acid sequence homology. And on occasion, proteins having an amino-acid sequence that usually corresponds with a particular structure lacked that structure entirely.
Using recent advances in informatics and investigational techniques, researchers are now beginning to circumvent the amino-acid comparison step to look directly at the structures of proteins with similar functions in order to determine which structures are responsible for these functions. These powerful techniques, led by 3-D structural research, allow scientists to identify vastly different amino-acid sequences that give rise to similar structures and functions. Information gleaned from this kind of research may currently be used to further research, produce designer proteins, and allow more precise drug design.
A valuable tool misunderstood and underused
Given the many potential uses of 3-D structural information, it is wise to protect that value in a patent. However, there are currently no issued patents claiming only 3-D structural information. This may be a by-product of the non-commercial nature of 3-D structural research until quite recently. It may also reflect an opinion both among researchers and officials in the Patent and Trademark Office that structural information is a natural phenomenon, simply nothing more than the depiction of the protein in its natural state, or a law of nature or abstract idea. However, just as natural phenomena, laws of nature, and abstract ideas are certainly not patentable, 3-D structural information is certainly none of these things.
If a 3-D structure could be obtained for a protein in a living organism, one might argue that it was a depiction of the protein in a natural state, but present 3-D structures are a far cry from this. In order to obtain a 3-D structure of a protein using any of the currently available methodologies, the protein must be isolated and purified and analyzed in this state or coupled with one or a few known additional proteins or other molecules.
X-ray crystallography, which determines the 3-D structure of the protein, requires that the protein be crystallizable. Other manipulations are required before protein structure can be observed via other techniques, such as UMR analysis. In fact, these manipulations change the natural state of the protein so vastly that in some cases it may actually impair the usefulness of structural information.
The Supreme Court has held that "anything under the sun that is made by man" is patentable. Accordingly, no one argues that an isolated purified protein alone or in a composition of known components is not patentable. By the same logic through which one might attempt to characterize the 3-D information of a protein as the representation of it in its natural state, one is obligated to recognize that the 3-D information of a protein in an unnatural and therefore patentable state is not a phenomenon of nature.
Similarly, the data obtained in 3-D structural analysis is not simply a law of nature or abstract idea. While this may or may not be true of the raw data, such raw data have no meaning unless they are manipulated and interpreted by an individual to arrive at a final representation of the 3-D structure of a molecule. These interpreted data, therefore, have felt the hand of man and so are not barred from patentability.
Indeed, some uses of 3-D structure have already been deemed patentable. These include using the structure in rational drug design, using it to define a new class of proteins with a new protein conformation, and using it to define the function of a novel protein whose structure indicated a particular function. Additionally, 3-D structure may be used to define a class of compounds, such as inhibitor compounds, that are defined by their ability to bind to a particular structure of a protein, such as the binding pocket.
However, more than mere uses of 3-D structures should be patentable. The structures themselves would appear to be eligible for patent protection. As with any new type of patent claim, the Patent and Trademark Office would likely examine claims to 3-D structure very closely to ensure that they meet all patentability requirements.
Therefore, a patent application claiming 3-D structure should provide all coordinates and variations with specificity. It also should include a detailed description of how the structural analysis was accomplished. If a function is known to correspond to the structure or is likely to do so, this function should also be described.
The claim to the structural information might read something like the following: " A set of coordinates having x, y, and z positions (insert coordinates), wherein said coordinates define a three-dimensional structure of Protein X." Finally, any real-world uses should also be claimed as a method using the structural information to provide protection should the Patent and Trademark Office refuse to issue claims to the structure itself or in case a court might later find those claims invalid.
As with all new technologies, obtaining patent protection for 3-D structure claim may require some struggles, but the myriad uses of structural information will make any time pursuing such claims well spent.
Rochelle K. Seide is a partner at the law firm of Baker Botts. She is experienced in biotechnology, intellectual property, and patent issues, and also has a Ph.D. in human genetics. Michelle LeCointe is an associate at Baker Botts and a registered patent agent. They can be reached at [email protected] and [email protected] .
TrendSpotter is a weekly column that focuses on how trends in politics, patent law, and the US and European markets will affect the genomics industry. The column appears every Friday. To access previous columns just enter the word "Trendspotter" in the archive search window on the homepage.