Patents on individual genes are an important vehicle to enable the commercial exploitation of genetic engineering. But the extension of patents to genetic material and living organisms remains controversial. Regine Kollek brings forward some fundamental questions about the definition of genes on which patents are based.

The ongoing debate on patenting genes focuses on two complex issues. The first issue comprises the ethical, social and economic consequences that result from the extension of patent protection to genetic material and living organisms. The second issue concerns the definition of the object to be patented. What kind of patents for what genes should be granted? Are the descriptions of the full structure, the function and the potential use of the concerned DNA sequence prerequisites for patent protection? And, can genes be patented as such, or only when they are part of an invention?

The discussion related to the second issue reflects specific problems, which are at least partially due to the ambiguities of the understanding of the gene itself. Genes are not only continuous stretches of DNA which code for a protein, but can be rather complicated structures. In eukaryotes, most coding sequences are interrupted by so­called introns, which are excised prior to translation into gene products. Furthermore, matching strands of the double helical structure of a gene can code for different products. For example, one strand of a particular region in the DNA of a rat codes for a hormone produced in the brain, while the corresponding strand codes for a gene product found in its heart.

Genes are therefore often understood as "continuous stretches of one of the two opposing strands of a duplex DNA molecule which may include meaningless stretches (which are dutifully removed from the transcribe messenger RNA molecules by processing in the nuclei of eukaryotic cells)" (compare Nature 372:33, 1994). Other definitions describe the gene as a coding sequence, which may or may not include introns or regulatory sequences. These definitions take into account that genes can occur in the cell in different structures and at different stages of processing. Therefore, there is no common structure, which can unanimously be identified with "the gene".

The situation does not differ much if one tries to identify the function of a gene. Although the biochemical properties of a gene product can be described precisely as soon as it is identified and isolated, its physiological function quite often cannot. For example, the gene for a particular isomerase has been discovered in bacteria, yeasts, insects and mammals. In spite of their almost identical structural and biochemical properties, the proteins are involved in different processes in the various species. In the fruit fly, the specific isomerase is involved in vision, while in mammals it seems to regulate the maturation of immune cells. Therefore, it is not just a gene's sequence which determines its functions in the organism, but also its location in a specific chromosomal, cellular, physiological and evolutionary context, may play a role.

These findings are pointing out some of the problems of the current gene concept. Although it must be very clear that no normative conclusions can be drawn from empirical evidence, this evidence is of specific relevance to some aspects of the debate on patenting:

• Firstly, any definition of a gene upon which a patent would be based is a convention, which includes or excludes specific findings and interests;
• Secondly, there can be no patent on "genes" as physical entities only, since the function of a gene is not only determined by its sequence but also by its genetic environment. When a coding sequence is transplanted from its original host into a new one, it may still be responsible for the synthesis of a protein with specific biochemical properties, but its function in the new organism may be different or missing.
• Thirdly, since "genes" cannot be defined and hence not patented as such, the question arises in which way they can be part of an invention. When a transgenic organism is manufactured, one particular coding sequence is selected and combined with other genetic elements. None of these elements (coding sequences, vectors, regulatory elements, etc.) are invented by humans, since their structure and function has evolved over time in the context of a specific molecular and general ecology. The crucial step in constructing a biological system able to express specific traits or to produce a protein in commercially exploitable quantities, however, is to find a specific combination of existing genetic components and host cells. Therefore, novelty or non­obviousness of an "invention" does not lie in the DNA­sequence or biochemical properties of a protein, but in the new order in which these elements are combined.

From this perspective, patent protection could only apply to a specific combination of genetic elements and host cells used to manufacture a gene product, and not to genes, sequences, physiological processes or organisms involved. Broad patents granted to procedures of genetic intervention are questionable as well, since they do protect the obvious, and not the inventive part of a researchers work.
Regine Kollek

Regine Kollek got her PhD in molecular biology. Currently she is Project Director at the Hamburger Insitut für Sozialforschung (Germany).