LONDON - Now that the dust is beginning to settle from the debate on the patentability of gene sequences, and as clearer principles are beginning to emerge from European and US patent regulators, a new hurdle for patent applicants has appeared: enforcement.
With the advent of in silico research, traditional approaches to infringement are beginning to look outdated. In this installment of TrendSpotter, we look at the difficulties of using patents to restrict the in silico manipulation of genetic information.
Patentability of gene sequences
The enormous number of patent applications filed by companies like Incyte, Celera Genomics, and Human Genome Sciences for their sequence data is beginning to trigger a steady flow of patents claiming ESTs and complete-gene sequences. The US Patent and Trademark Office has responded by introducing more stringent utility requirements to the point that applicants must now show a substantial, credible, and specific utility.
In Europe, similar issues were intended to be tackled by the European Biotechnology Directive, which eventually was passed in 1998 after a 10-year ethical debate. Unfortunately only a handful of the European member states have to date incorporated the Directive into national law, while others have challenged outright the validity of the Directive. To an extent, these legal wrangles have been side-stepped by the European Patent Office, which, though not bound by the Directive, has chosen to incorporate it into its implementing regulations.
Under the Directive, applications for gene-sequence patents may be approved even if the sequence is identical to a naturally occurring sequence, provided that an industrial application of the sequence is disclosed in the patent application.
The EPO will object to biotech inventions because it feels they lack an inventive step. In this respect, in which an application relates to a gene sequence, attention is often given to the method by which a gene fragment is isolated: If, for example, the method is routine and the goal of isolating the relevant gene is known, patentability is likely to be denied.
Patent applications for bare gene sequences or ESTs, therefore, will likely fail under current EPO practice. On the other hand, there is an increasing number of successful European patents that have satisfied the EPO's requirements, and several of these have withstood validity challenges brought before the courts of the member states of the European Patent Convention.
Infringing gene-sequence patents by in silico manipulation
Patents claims can be broadly categorised as either 'process' or 'product' claims. While infringement is a matter for the laws of each of the EPO member states, the acts that amount to a patent infringement are similar in the majority of jurisdictions.
A process claim will be infringed by anyone who uses that process, or offers it for use or disposes of offers to dispose of, uses, keeps, or imports a product obtained directly by means of the process. A claim to a method for producing a deletion mutant by recombinant means would therefore be infringed by a person who carries out that method or who offers for sale the deletion mutant produced by that method.
Clearly, each of these acts of infringement is a physical act in relation to the invention. On the other hand, the drug-discovery process is becoming more dependent upon computer-assisted prediction and manipulation of sequences and whole molecules.
Increasingly, the discovery process is carried out in silico without the creation of any physical compounds or the use of "physical" processes in the traditional sense. Although the in silico process may well use sequence information or "virtually" create compounds that are covered by the patent claims, it is difficult to see that there can be any infringement of the patent in a traditional sense until a physical product is made or used.
In ordinary English terms the simulated production of a patented molecule within a computer is not the same as making or using that product. The contrary view would be tantamount in order to provide protection to the information disclosed in a patent.
It seems that in Europe for the time being, patent holders can take action against only those who carry out acts in relation to physical products covered by their patents rather than simulations of that product. Providing that the patented sequences or products are discarded in the in silico process, and the ultimate product is not covered by a patent, the patent owners may be largely powerless.
It has been proposed that claims may get around this problem if they are directed at using novel gene or protein sequences in specified in silico studies. For example, a claim may be drafted to the use of a novel protein sequence in a computer model that seeks to generate a structural model of the protein and potentially to identify likely inhibitors. However, there are additional hurdles to patent claims of this nature, particularly in Europe, where they may be disallowed if they are believed to relate to a computer program as such.
(The limit of patentability of computer-related inventions in Europe will be the subject of a future TrendSpotter column.)
Tim Powell is a partner at the law firm Bristows (www.bristows.com) in London. He has an MA in Natural Sciences (Biochemistry) and specialises in advising on IP rights in the biotech field. Alex Wilson, an assistant at Bristows, has a BSC in Biochemistry and worked at the EPO as an examiner before qualifying as a solicitor.