In the field of biotechnology — as the achievements of genetic engineering and gene sequencing have broadened the range of potentially patentable subject-matter — the dividing line between invention and discovery has become increasingly difficult to determine.
Patents on ‘life’
Patentable inventions must — under conventional patent law — be new, useful and involve an ‘inventive step’. In contrast, it is generally accepted that utilising something that already exists in nature is a ‘discovery’, and is therefore not patentable.
Or is it? In the field of biotechnology, as the achievements of genetic engineering and gene sequencing have broadened the range of potentially patentable subject-matter, the dividing line between invention and discovery has become increasingly difficult to determine.
Since the 1980s, it has become increasingly common for micro-organisms, plants and animals to be patented, as have cell lines and even DNA sequences. Of these, it is the patenting of genes and gene fragments that appears to challenge one of the most fundamental tenets of patent law: the novelty requirement.
According to Giles Stokes of Derwent Information, a company that provides patent and scientific information, “[DNA] sequences first began appearing in patents in 1980, just 16 sequences all year. By 1990 that figure had risen to over 6,000 sequences. Throughout the 1990s the growth in the patenting of sequences expanded exponentially, and this looks set to continue. In 2000 over 355,000 sequences were published in patents, a 5000 per cent increase over 1990”. (1)
Many scientists and others have asked how, if patents apply to inventions but not discoveries, naturally occurring genes isolated in a laboratory can be patented?
The case for gene patenting
Companies and researchers defending such patenting argue that locating, isolating and describing molecular biological matter requires considerable ingenuity. If in doing so they have revealed to the world something that was previously unknown, and are able to explain its function and its possible industrial application (for example, as a potential therapy or diagnostic technique), then they have an invention that should be patentable.
Moreover, many claims are not actually for DNA in its raw state but for complementary DNA (cDNA). cDNA sequences are produced in the laboratory and differ from their naturally occurring counterparts in that certain sections of the molecule are ‘edited out’. Therefore many companies argue that, as with any other synthetic chemicals, cDNA sequences should be patentable provided that they fulfil the normal criteria of inventiveness and that the application discloses a credible function.
A more pragmatic argument for such an expansive interpretation of the term invention is that biotechnology research and development is risky and expensive. Patent ownership is therefore essential, as it enables new companies to attract investment even before they have products on the market. The argument here is that if these companies are to continue producing new drugs and therapies for treating disease, the public may be best served by allowing life forms and their structural components to be patented as inventions.
The biotechnology industry and its supporters (including those in political and legal circles) say that for this to be possible, the patent offices and courts should adopt flexible interpretations of novelty, usefulness and inventive step that give the benefit of the doubt to the applicants.
DNA: a complicated molecule
Many critics are sceptical that the deletion of ‘junk DNA’ is inventive enough to deserve the reward of a patent — a claimed cDNA molecule is likely to be obvious to somebody ‘skilled in the art’ who knew the sequence of its naturally occurring equivalent. This is because techniques for isolating and purifying DNA sequences are well known and no longer require a great deal of skill to use. But even if nobody knew about the naturally occurring equivalent, such a claim should still arguably fail due to the lack of an inventive step, on the basis that the techniques employed are routine.
But the problem goes further even than this. It is now understood better than ever before that the so-called Central Dogma — that DNA makes RNA makes protein — is not always reliable. All genes have a function but not all of them are involved in protein-making processes. To make matters even more complicated, different genes may occupy the same strand of DNA so that it may be extremely difficult to determine where one gene begins and another ends. Further, the whole protein-making process is complex and still to some extent a mystery. What is becoming apparent, however, is that successful protein manufacture requires the involvement of more than one gene.
So granting patents on a gene on the basis that it performs a single function such as coding for a particular protein or that it is associated with a particular disease is problematic. This is because it simplistically assumes that genes have independent functions. In fact, genomes should be seen as consisting of multiple intersecting "mini-ecosystems" rather than as a single collection of separately functioning "Lego bricks".
Treating genes as patentable inventions on the basis of a single function is therefore more a reflection of ignorance than of insight. It is also essentially anti-innovation, since it potentially hinders opportunities for follow-on researchers to carry out further investigations on genes that had previously patented for one out of possibly numerous functions.
Do patents restrict research?
The leads to one of the main arguments against patenting DNA: that the patenting of gene fragments used in basic research may be placing undesirable restrictions on the ability of others to scientists to use such gene fragments in their own research. Michael Heller and Rebecca Eisenberg of the University of Michigan describe this situation as a "tragedy of the anticommons"  — a reference to an earlier description of the degradation of the natural environment as the "tragedy of the commons".
Such critics warn that the increased patenting of scientific results used in pre-market or 'upstream' (i.e. basic) research may, in the words of Heller and Eisenberg, "be stifling life-saving innovations further downstream in the course of research and product development".
This possibility could arise because the development of future commercial products — such as therapeutic proteins or genetic diagnostic tests — often requires the use of multiple gene fragments, an increasing number of which are being patented by private corporations. For example, one US biotechnology firm, Incyte, has applied for patents on 500 such fragments. Even some pharmaceutical company researchers are concerned about the aggressive patenting of gene fragments.
Critics also claim that the increasing patent protection of scientific results will raise the cost of research and development. This is because a company that wants to use these results to develop new products will need to acquire licences from other patent holders, and will incur large (and possibly prohibitive) costs in doing so. Furthermore, some hospitals and laboratories have expressed concern about having to pay excessive license fees to companies that have taken out patents on disease-related gene sequences, in order to perform diagnostic tests for such diseases.
As for agricultural biotechnology, similar concerns have arisen. For example, 'Golden Rice', a genetically-engineered rice enriched with beta-carotene, is protected by 70 patents held by 32 companies and institutions.  Even though some of the patent-holding companies have agreed to co-operate, completing the complex negotiations necessary for the rice to be developed and disseminated will take a long time.
Because of these concerns — and also because patents can encourage the development of technologies that do not serve the public interest — some NGOs have been campaigning to abolish the patenting not only of genes but also of all life-forms and their structural and functional components.
A question of policy
Such issues have raised a number of important policy questions. For example, should the application of patent systems be modified to accommodate new areas of research that are risky and expensive even if the resulting discoveries appear to lack genuine novelty or inventiveness, as traditionally defined under patent law? Conversely, to what extent should the design and application of rules that allow the patenting of basic research tools take into account the possible damaging effects of such patenting on downstream innovation?
While these are largely issues for legislators to tackle, bodies such as the European Patent Office and the United States Patent and Trademark Office (USPTO) are in a position to amend the guidelines for their examiners. For example, the USPTO has tightened up its utility requirement, so that patent applications disclosing DNA sequences must now provide "a specific, substantial, and credible utility for the claimed isolated and purified gene". 
At present, there are no clear-cut answers to these questions. But their resolution will determine where the shifting line between patentable inventions and non-patentable discoveries is legitimately placed in future.
The author is based at the Queen Mary Intellectual Property Research Institute, University of London.
 Giles Stokes (2001) Patent applications of gene sequences
 Michael A Heller & Rebecca S Eisenberg (1998) Can patents deter innovation? The anticommons in biomedical research Science 280, 698
 Justin Gillis (2000) Monsanto Offers Patent Waiver. Washington Post
 USPTO (2001) Utility Examination Guidelines Federal Register Vol 66, No 4
Nuffield Council on Bioethics (2002) The ethics of patenting DNA