Whole-Genome Patenting Patenting Criteria Although a Key Tenet of Patent Law Is That Naturally Occurring Substances Cannot Be Maureen A

PERSPECTIVES

material, and indeed, what the study of SCIENCE AND SOCIETY genomes means for our understanding of biological entities.

Whole-genome patenting Patenting criteria Although a key tenet of patent law is that naturally occurring substances cannot be Maureen A. O’Malley, Adam Bostanci and Jane Calvert patented, substances that have been isolated and purified — such as DNA — can AbstractEPIGENETICS | Gene patenting is now a familiar to be treating genome patents as if they were be patented as long as they fulfil the criteria commercial practice, but there is little nothing other than standard DNA patents. for patentability6. In the United States, the awareness that several patents claim However, further analysis reveals that patent basic criteria are novelty, non-obviousness absractownership absract of the absract complete absract genome absract specifications describing whole-genome inven- and utility; in Europe, the equivalent crite- absractsequence absract of a prokaryote absract absract or virus. absract When tions use arguments that imply that genomes ria are novelty, inventive step and industrial absractthese patents absract are absract analysed absract and absract compared are qualitatively different from individual applicability. absractto those absract for other absract biological entities, it genes. Whole-genome patents also use differ- An invention is novel if it has not previ- (leavebecomes line spaceclear thathere) genome patents seek ent arguments from microorganism patents, ously been made public. Even if some gene textto exploit text text the text genome text text as text an texttext information text text which might be thought of as a similar cate- sequences in a genome have already been texttextbase and text are text part texttext of a broader text text shifttexttext text gory of ‘whole’ biological patent. These dis- published, genome sequences could be texttowards texttext intangible text text intellectual texttext text property text texttext tinctions are further complicated by the way argued to be novel because not all features of textin genomics. text text text text text text. in which the European Patent Office (EPO) the invention have previously been disclosed has dealt with genome patent applications — in a single publication7. Non-obviousness or News of genome patenting is often met with a treatment that leads our exploration of inventive step means that the invention would surprise, disbelief or dismissal. Nevertheless, genome patenting to the key issue of how require more than a routine procedure by an several whole-genome patents have been arguments for the utility of DNA fragments individual who is “skilled in the art”. issued by the US Patent and Trademark apply in genome patents. DNA patents have been subjected to Office (USPTO) and further applications are Genome patenting has emerged as an heavy criticism from lawyers, scientists and pending. Although gene patenting has been expression of the recent informational shift the public for inadequately fulfilling the util- challenged on ethical grounds and in regard in genomics and patenting. This shift is of ity requirement. Some of the strongest objec- to data access and criteria for patent- potential interest to several groups of inter- tions have been against attempts to patent ability1,2,3, whole-genome patenting has so ested parties and observers. For patent pro- ESTs for their use as probes in gene discov- far gone almost unnoticed. Even the recent fessionals, genome patenting gives an indi- ery. Following public consultation, the controversy surrounding patent applications cation of how developments in genomics USPTO has recently tightened its assessment for the genome sequence of the SARS-asso- and bioinformatics might be changing the criteria. Rather than just being generally use- ciated CORONAVIRUS4,5 (see also Online links nature of patenting. For scientists, genome ful, applications must now show “specific, box) is primarily concerned with whether patents blur the supposed line between substantial, and credible” utility8. Once the patenting is an appropriate and effective way research and its applications, with implica- function of a gene is disclosed (producing a to control access to data and stimulate tions for how research is financed and data specific protein, for example), it is considered research. The SARS discussion does not shared. For social scientists, the interactions to have such a use. The EPO has adopted address the implications of patenting a between genomics and the patent system similar standards9. whole genome instead of the more common are of great interest for understanding how Furthermore, EPO patent applications patenting of DNA fragments. society might benefit from the genomics must satisfy a “unity of invention” require- Perhaps genome patents have escaped revolution and how commercial interests ment, the implementation of which is cur- scrutiny because, at least superficially, they might shape the future development of this rently being considered at the USPTO. This seem to be no more than simple extensions of science. Finally, for philosophers of biology, standard allows several sub-inventions to the DNA patenting that has flourished with the genome patenting raises issues about the be linked together by a common “general increased ease of entire genome sequencing. consequences of conceptualizing genomes inventive concept”,but prevents unrelated At first glance, patent offices certainly do seem as sequence information or biochemical inventions from succeeding as a single

Table 1 | Claim-specific whole-genome patents granted by the USPTO, ordered according to the date of filing Organism Genome Utility Assignee Inventors Sequence Filing Issue US argument publication date date date patent (reference) number Adult T-cell Clarifying Diagnosis; Juridicial Yoshida & June, 1983 (33) 5 October, 18 October, 4,778,756 leukaemia genome prevention; Foundation, Sugano 1987 1988 virus structure therapy Japanese Foundation for Cancer Research (Japan) Non-A non-B Distinguishing Detection; Immuno Okamoto & December, 7 August, 27 June, 5,428,145 hepatitis subtypes; prevention; (Japan) Nakamura 1990 (34) 1992 (cont.* 1995 virus better treatment 1992) diagnosis and vaccines Bacteriophage Genome Thermostable Prokaria Hjorleifsdottir Published only in 1 June, 10 December, 6,492,161 RM378 modification enzymes and (Iceland) et al. patent application 2000 2002 proteins (cont.* 1999) Haemorrhagic Isolating and Diagnostic; ABIC Ltd (Israel) Pitcovski et al. 30 September, 20 November, 16 December, 6,663,872 enteritis identifying vaccine 1998 (35) 2000 (cont.* 2003 virus genes; (turkeys); 1999) genome gene therapy manipulation (general) Buchnera sp. Genetic Pesticidal; Riken (Japan) Shigenobu 7 September, 23 February, 14 October, 6,632,935 APS information metabolic et al. 2001 (36) 2001 2003 mechanisms *This is a continuation of an earlier filing from the date shown. USPTO, US Patent and Trademark Office. application. If, for example, a group of genomic fragments but embed these claims in general advantages of having a whole DNA fragments or sequences can be linked a broader argument that the whole genome genome as the invention: the “clarification of together by an overarching concept, they constitutes a useful invention (TABLE 2). It is the structure” of the genome (for example, can then be covered by one patent. It seems not the legal status of any of these patents that adult T-cell leukaemia virus in TABLE 1), reasonable to think of the genome as a con- concerns us (in fact, they are still untested in identification of constituent genes (for cept with the potential for serving that unify- court), but the ways in which they make their example, haemorrhagic enteritis virus ing function, and patent documents provide arguments and show how genomes have been (HEV) in TABLE 1) and the capacity to dis- a good basis for examining the extent to used in recent patenting. tinguish similar genomes (for example, which this suggestion is supported in patent NANB). practice. Claim-specific whole-genome patents. The first category of genome patents, claim-specific Contextual whole-genome patents. Genome Genome patents and applications whole-genome patents, places the whole patents in the second category direct their A few publications mention the existence or sequence of a specified genome in the claims claims to specified open reading frames prospect of whole-genome patenting10,11,12, section as the primary invention. For exam- (ORFs) or polynucleotides, but do so in but do not discuss the actual patents. To pur- ple, the patent for bacteriophage RM378 the context of a broader specification of the sue the cross-disciplinary implications of claims the isolated and sequenced genome of invention that argues for the whole sequ- such a practice, we searched the online data- the phage, as well as any recombinantly pro- enced genome as an integral part of the bases of the USPTO and the esp@cenet duced DNA. Although the Buchnera sp. strain invention. The patents for Haemophilus worldwide database of the EPO with terms APS patent has only one claim — the isolated influenzae and Mycoplasma genitalium were such as ‘whole genome’ or ‘complete nucleo- genome as represented by its sequence first filed as claim-specific genome patent tide sequence’.We discarded all search results description — the other claim-specific applications, but during examination, their that only asserted specific DNA fragments genome patents extend their claims to cDNA, potentially far-reaching claims were without significant reference to the entire proteins, vectors and host cells (as would restricted to specific ORFs. This restriction genome and those in which the genomes had most patents for DNA fragments). was probably the result of objections by been modified before sequencing. We ended Utilities for genome patents in this cate- USPTO examiners. The specifications of up with 10 whole-genome patents that were gory range from disease diagnosis and ther- these patents, which do not normally issued (making no claims to exhaustiveness), apy to the development of thermostable change after filing, still persistently refer to 6 of which were for viral genomes and the enzymes and pesticides (TABLE 1). Although the whole-genome sequence (and any remainder for prokaryotic genomes. these utilities are often elaborated in relation sequence that is 99.9% similar) as compris- The patents fell into two categories. In the to particular DNA fragments or encoded ing or providing the basis of the invention. first, the genome is the primary object of polypeptides (for example, non-A non-B Likewise, in the Methanococcus jannaschii the patent’s claims section (TABLE 1). This sec- hepatitis virus (NANB) in TABLE 1), the patent, the summary of the invention tion is legally the most significant in defining descriptions of the inventions avoid exclud- begins with the whole-genome sequence, the protection that the patent provides. ing other genomic fragments from the over- after which the invention is further directed Patents in the second category, which we call all invention by regularly invoking the rest of to the ORFs in the claims. The claims of the ‘contextual whole-genome patents’, list the genome. Arguments are also made for the two virus genome patents in this category

Table 2 | Contextual whole-genome patents granted by USPTO, ordered according to the date of filing Organism Genome Utility Assignee Inventors Sequence Filing Issue US argument publication date date date patent (reference) number Maize chlorotic Distinguishing Diagnostic; Novartis Law et al. May, 1997 (37) 4 April, 2 February, 5,866,780 dwarf virus strains; resistance Financial 1995 1999 understanding engineering Corporation genome structure and organization Tospovirus No explicit Diagnostic; Novartis Goldbach December, 1990 (38) 26 November, 21 November, 6,150,585 (Tomato argument resistance Financial et al. September, 1991 (39) 1996 2000 spotted engineering Corporation November, 1992 (40) (cont.* 1989) wilt virus) Mycoplasma Understanding Diagnostic; The Institute Fraser et al. 20 October, 1995 (41) 19 October, 25 March, 6,537,773 genitalium chromosomal therapeutic; of Genome 1995 2003 organization; pharmaceuticals; Research (TIGR); (cont.* 1995) comparative industrial Johns Hopkins genomics; fermentation University; minimal University of genome North Carolina knowledge Haemophilus Comparative Pharmaceutical; Human Fleischmann 28 July, 1995 (42) 23 August, 4 March, 6,528,289 influenzae genomics; industrial Genome et al. 2000 2003 chromosome fermentation Sciences; (cont.* 1995) structure Johns and function Hopkins Methanococcus Comparative Probes; primers; TIGR; Johns Bult et al. 23 August, 1996 (43) 20 October, 28 September, 6,797,466 jannaschii genomics; general function Hopkins 2000 2004 identifying of any fragment (cont.* 1996) open reading frames *This is a continuation of an earlier filing from the date shown. USPTO, US Patent and Trademark Office. similarly focus on particular sequence frag- Microorganism versus genome patents patents are usually based on the genomes of ments but base their inventions on the Genomes are often thought of in a holistic well-known but previously unsequenced whole genome. The inventions of all of way, which makes it logical to compare them organisms. these patents are described very broadly, to whole organisms. Patenting microbes is a The main difference between whole- covering an extensive range of related bio- long-established practice, not only for modi- genome and organism patents is the extent logical material and its demonstrated and fied microorganisms13, but also for naturally to which the patent attempts to cover fur- postulated uses. occurring strains that have been isolated and ther biological material. Both categories of The three TIGR/HGS (The Institute of cultured14. All the genome patents listed in genome patent describe their inventions in Genomic Research and Human Genome TABLES 1,2 are for microorganismal genomes terms that stretch to any and every nucleotide Sciences) patents also originally claimed the (we include viral genomes in this category). and polypeptide implicated by the sequence, computer-readable sequence as the invention. The tradition of allowing microorganism as well as to vectors and host cells. The only These claims now exist only in the specifica- patents might partly explain the absence of organism patents that follow this strategy are tions, which describe the “computer-related any patented eukaryote genomes and also the few that refer to the DNA of the specified embodiments” of the genome as “a contigu- why whole-genome patenting has not given organism, either to extend the coverage to ous string of primary sequence information” rise to particular public concern. further biological material18,19 or to encom- suitable for storage and analysis by computer. But what is the difference between whole- pass all organisms in the same genus with a Having the genome in this in silico form, genome patents and patents that have been certain percentage of sequence similarity15. argue these patents, allows scientists to move issued on whole microorganisms? Not sur- In these cases, it seems that the DNA and its beyond a gene-by-gene approach towards prisingly, the utility arguments made for potential uses are called on to reinforce the larger discoveries of genomic structure, func- microorganism patents15–21 are generally sim- organism patent and expand the protection tion and evolutionary history, as well as the ilar to those for genome patents (TABLES 1,2). it provides. Genome patents take this strat- identification of “commercially important Microorganism patents for Archaea, for egy one step further by claiming the com- fragments”.Although the virus patents do not example, argue for uilities that are related to plete sequence. The obvious question is specify the in silico or machine-readable enzyme production in harsh environmental whether a genome patent achieves more nature of the invention, they too rely on the conditions, and bacteria patents claim appli- protection for the inventor than does a knowledge provided by the whole-genome cations that range from human health to patent on a collection of DNA fragments. sequence for purposes such as engineering BIOREMEDIATION. However, novelty is estab- plant resistance to the virus or for establishing lished in subtly different ways. Organism USPTO versus EPO perspectives boundaries between strains (for example, patents are based on previously unknown Archived EPO examiners’ reports (on the EPO maize chlorotic dwarf virus (MCDV-Tn) in organisms that might have been isolated in Online Public File Inspection page) of past TABLE 2). unusual circumstances, whereas genome and present applications for whole-genome

504 | JUNE 2005 | VOLUME 6 www.nature.com/reviews/genetics © 2005 Nature Publishing Group PERSPECTIVES patents take one step towards answering this Conclusion applications for the SARS genome provoked question. Most of these reports agree that a Once the surprise that whole-genome patents worries that a privately held patent would genome sequence is novel even when parts of exist has dissipated, it might be tempting to function as a gatekeeper to all SARS-related the genome have already been sequenced, conclude that such patents are either so few or research and inhibit drug development. It is although some dispute the novelty of newly so weak that their existence does not matter. likely that similar fears would arise with any sequenced genomes from closely related strains. However, the characteristics of whole-genome increase in the numbers and awareness of One examiner, for example, objects that a sub- patenting indicate an important movement in issued genome patents. Although empirical mitted genome sequence (Chlamydia pneumo- DNA patenting from biochemical tools and work on the impact of conventional DNA niae) is merely a definition of a particular strain products to information resources12,23,24.Had patents shows that they do not always have from the many isolates available, thereby ques- the computer embodiments remained in the negative effects on research access27,28, infor- tioning the patentability of genomic variation. TIGR/HGS claims, these patents would have mational patents could be much more In another case (Influenza A), the report notes attempted to control information in a way not restrictive23. that the identified genes are well known from yet realized in patent practice. As bioinformat- Anticipations of genome patenting could other related strains and that their presence ics and in silico modelling gain deeper and extend to questions about whether prokary- should therefore be expected. Both these more extensive purchase on every aspect of otic genome patents set precedents for reports also argue that the sequencing of a genomic science25, a full shift to allow claims eukaryotic genome patents (our non- genome is routine and does not in itself entail on intangible informational property seems exhaustive search found none of these, nor an inventive step. inevitable. did we find any in a search of pending Most tellingly, some of the reports argue whole-genome patent applications). We see that genomes can only be considered as a sin- two general factors that inhibit this potential gle invention if they share a uniting feature trend: tradition and genomic organization. that is novel, inventive and technically rele- “Is the genome just … used As we noted earlier, prokaryotic micro- vant. According to the examiners’ comments, to unite several nucleotide organism patenting is well established, but the application examples of Influenza A and unmodified multicellular organisms are far C. pneumoniae do not constitute a unified sequences into a single less commonly the objects of patents. We invention that would meet EPO standards. invention, or … a causally believe that resistance to the patenting of Our initial hypothesis that the idea of a ‘higher life forms’ — including genetically genome is sufficient to unite several genes efficacious phenomenon modified ones such as the ONCOMOUSE — is and their functions into one invention is that does something more likely to similarly discourage patenting the not, therefore, supported by these examples. genomes of these organisms. At the human A genome, at least in these cases, does not have than an aggregation of level, ethical arguments have been made stat- the taken-for-granted unifying capacity of an genes can do?” ing that patenting a whole human genome organism: it is seen as merely a collection of violates the integrity of an individual in a fragments of DNA. way that patenting parts of the genome does not29. Differences between the USPTO and Gene versus genome utility So far, there seem to have been no obvious other patent offices on the patentability of If there are any qualitative differences between commercial benefits from whole-genome life forms30 will no doubt continue to be patents for whole genomes and those for DNA patents, but industry has a long way to go reflected in the international treatment of fragments, it seems likely that they will be before it catches up with all the DNA patent- future whole-genome applications. found in the utility arguments — the most ing of the past two decades. Harbingers of Moreover, the genomes of most eukary- contested feature of recent gene patenting. Are how this trend might develop can be seen in otes have proportionately less protein-coding any special uses attributed to genomes that are new patents for computer programs and DNA, meaning that it is more difficult to not attributed to isolated fragments of DNA? business methods. In discussions of how assign function to large amounts of sequence. Both claim-specific and contextual genome commercial protection is increasingly being patents rely on the utility of the information sought for the information that is produced provided by the whole sequence. Such infor- by bioinformatics24,26, unannotated genome Glossary mation is considered to be valuable because it sequences (that is, primary information) are BIOREMEDIATION allows better understanding of the organism, considered less patentable than secondary The use of microorganisms to degrade hazardous of specific genes, of chromosome structure information about how gene products might contaminants in soil and water to environmentally safe levels. and function, and of relationships with other interact in a cell. genomes. These genomic utilities seem to be As yet, there have been no high-profile CORONAVIRUS primarily research-orientated, in contrast to cases in which genome patents have been A genus of virus named after the projections that create commercial applications that might arise publicly or legally challenged, although the a crown-effect around the outside of each virus directly out of the more specific biochemical EPO examiners’ reports give an indication particle. They infect various mammals and birds, causing respiratory and enteric illness. The functions attributed to genes and gene prod- that future genome patents might be treated SARS-associated coronavirus is a previously ucts. By exploiting the whole genome as the sceptically in Europe with respect to unity of unrecognized member of the genus with no close informational basis of the invention, these invention and novelty. Because the validity genetic relationship to known coronavirus sequences. patents distinguish themselves from stan- of genome patents has yet to be tested in ONCOMOUSE dard DNA fragment patents that articulate court, the extent to which they will restrict (Also known as the Harvard mouse.) A type of laboratory their inventions as compositions of matter, research on the patented genomes is still only mouse that is genetically modified to carry genes that analogous to chemical patents22. a matter for informed speculation. Patent increase susceptibility to cancer (oncogenes).

Therefore, genomic organization militates Maureen A. O’Malley, Adam Bostanci and 27. Walsh, J. P., Arora, A. & Cohen, W. M. in Patents in the Knowledge-Based Economy (eds Cohen, W. M. & against the success of eukaryote genome Jane Calvert are at Egenis, the ESRC Centre for Genomics in Society, University of Exeter, Merrill, S. A.) 285–340 (National Academies Press, patenting. If the complete genomes of com- Washington DC, 2003). Amory Building, Rennes Drive, Exeter EX4 4RJ, 28. Resnik, D. B. DNA patents and scientific discovery and plex multicellular organisms are ever to be United Kingdom. innovation: assessing benefits and risks. Sci. Eng. Ethics commonly patented, it will probably be as Correspondence to M.A.O’M. 7, 29–62 (2001). informational components that are incorpo- e-mail: [email protected] 29. Resnik, D. B. DNA patents and human dignity. J. Law doi:10.1038/nrg1618 Med. Ethics 29, 152–165 (2001). rated into system models that have diagnostic 30. Perry, M. Lifeform patents: the high and the low. J. Int. Published online 10 May 2005 and other purposes. Biotechnol. Law 1, 20–27 (2004). 1. Nuffield Council on Bioethics The Ethics of Patenting 31. Lawrence, J. G., Hatfull, G. F. & Hendrix, R. W. Imbroglios Overall, the aspect of whole-genome DNA (Nuffield Council on Bioethics, London, 2002). of viral taxonomy: genetic exchange and failings of patenting that lends itself most readily to 2. Heller, M. A. & Eisenberg, R. S. Can patents deter phenetic approaches. J. Bacteriol. 184, 4891–4905 investigation is conceptual. All the patents we innovation? The anticommons in biomedical research. (2002). Science 280, 698–701 (1998). 32. Gogarten, J. P., Doolittle, W. F. & Lawrence, J. G. have identified raise important questions 3. Organisation for Economic Co-operation and Prokaryotic evolution in the light of gene transfer. Mol. about how genomes are conceptualized, espe- Development Genetic Inventions, Intellectual Property Biol. Evol. 19, 2226–2238 (2002). Rights, and Licensing Practices: Evidence and Policies 33. Seiki, M., Hattori, S., Hirayama, Y. & Yoshida M. cially in regard to how the utility of a genome (OECD, Paris, 2002). Human adult T-cell leukemia virus: complete nucleotide can be specified. Is the genome just a concept 4. Gold, E. R. SARS genome patent: symptom or disease? sequence of the provirus genome integrated in that is used to unite several nucleotide Lancet 361, 2002–2003 (2003). leukemia cell DNA. Proc. Natl Acad. Sci. USA 80, 5. Gene patents and the public good [Editorial]. Nature 423, 3618–3622 (1983). sequences into a single invention, or is it a 207 (2003). 34. Kato, N. et al. Molecular cloning of the human hepatitis C causally efficacious phenomenon that does 6. Demaine, L. J. & Fellmeth, A. X. Natural substances virus genome from Japanese patients with non-A, non-B and patentable inventions. Science 300, 1375–1376 hepatitis. Proc. Natl Acad. Sci. USA 87, 9524–9528 something more than an aggregation of genes (2003). (1990). can do? What is the relationship between the 7. Barton, J. H. Reforming the patent system. Science 287, 35. Pitcovski, J. et al. The complete DNA sequence and utility of a part (a gene) and any utility associ- 1933–1934 (2000). genome organization of the avian adenovirus, 8. US Patent and Trademarks Office Utility examination hemorrhagic enteris virus. Virology 249, 307–315 ated with the whole (the genome)? The guidelines. Fed. Regist. 66, 1092–1099 (2001). (1998). answers to these questions will be different 9. Cornish, W. R., Llewellyn M. & Adcock, M. Intellectual 36. Shigenobu, S., Watanabe, H., Hattori, M., Sakaki, Y. & Property Rights and Genetics: a Study into the Impact Ishikawa, H. Genome sequence of the endocellular depending on whether the genomes are and Management of Intellectual Property Rights Within the bacterial symbiont of aphids Buchnera sp. APS. Nature thought of in terms of biochemistry or Healthcare System (Public Health Genet. Unit, 407, 81–86 (2000). Cambridge, 2003). 37. Reddick, B. B., Habera, L. F. & Law, M. D. Nucleotide bioinformatics. 10. Human Genome Sciences 1996 Annual Report (Human sequence and taxonomy of maize chlorotic dwarf virus When the relationship between organ- Genome Sciences, Rockville, Maryland, within the Sequiviridae. J. Gen. Virol. 78, 1165–1174 ism and genome patents is examined, fur- 1997). (1997). 11. Shreeve, J. The Genome War (Knopf, New York, 38. de Haan, P., Wagemakers, L., Peters, D. & Goldbach, R. ther conceptual questions arise, especially in 2004). The S RNA segment of tomato spotted wilt virus has an terms of classification. Is the genome 12. Barton, J. H. Patents, genomics, research and ambisense character. J. Gen. Virol. 71, 1001–1007 other diagnostics. Acad. Med. 77, 1339–1347 (1990). sequence the representative of the organ- (2002). 39. de Haan, P. et al. Tomato spotted wilt virus encodes a ism? The genomic mosaicism of many 13. US Supreme Court. Diamond v. Chakrbarty 447, US 303 putative RNA polymerase. J. Gen. Virol. 72, 2207–2216 viruses and microbes makes the construc- (US Reports, 1980). (1991). 14. World Trade Organization. The Results of the 40. Kormelink, R., de Haan, P., Meurs, C., Peters, D. & tion of taxonomic relationships very com- Uruguay Round of Multilateral Trade Negotiations: Goldbach, R. The nucleotide sequence of the M RNA plex, and reducing this complexity to single the Legal Texts (Cambridge Univ. Press, Cambridge, segment of tomato spotted wilt virus, a bunyavirus with 1999). two ambisense RNA segements. J. Gen. Virol. 73, measures of overall relatedness is likely to 15. Stetter, K. O. Thermococcus AV4 and enzymes produced 2795–2804 (1992). obscure biologically meaningful connec- by the same. USPTO 5,714,373 (US Patent and 41. Fraser, C. M. et al. The minimal gene complement of Trademark Office, 1998). tions31,32. Existing whole-genome patents not Mycoplasma genitalium. Science 270, 397–403 16. Triplett, E. W. Microorganisms. USPTO 5,908,758 ( US (1995). only settle for simple measures of genomic Patent and Trademark Office, 1999). 42. Fleischmann, R. D. et al. Whole-genome random relatedness, but do so inconsistently. Some use 17. Bottone, E., J. & Peluso, R. Bacillus pumilus strain. sequencing and assembly of Haemophilus influenzae Rd. USPTO 6,090,613 (US Patent and Trademark Office, Science 269, 496–512 (1995). sequence differences between strains as the 2000). 43. Bult, C. J. et al. Complete genome sequence of the basis of their genome-patent claim (for exam- 18. Jones, B. E. et al. Thermopallium bacteria and enzymes methanogenic archaeon, Methanococcus jannaschii. obtainable therefrom. USPTO 6,218,164 (US Patent and Science 273, 1058–1073 (1996). ple, MCDV-Tn), whereas others discount Trademark Office, 2001). such variations between strains by arguing 19. Jones, B. E. & Grant, W. D. (Genencor International, Inc.). Acknowledgements Haloalkaliphilic microorganisms. USPTO 6,420,147 (US We would like to thank W.F. Doolittle and two anonymous referees that the patent covers other sequences within Patent and Trademark Office, 2002). for several useful suggestions. The research for this paper was a certain range of similarity (for example, 20. Bradfisch, G. A., Schenpf, H. E. & Kim, L. Bacillus supported by the Economic and Social Research Council (ESRC), thuringiensis isolates active against weevils. USPTO UK, as part of the programme of the ESRC’s Centre for Genomics M. genitalium, H. influenzae and HEV). The 6,605,462 (US Patent and Trademark Office, in Society (Egenis). A.B. also acknowledges a travel grant from the informational patenting of genomic variation 2003). Chemical Heritage Foundation. could have the benefit of bringing about bet- 21. Banfield, J. F. et al. Acidophile archaeal organism. USPTO 6,589,772 ( US Patent and Trademark Office 2003). Competing interests statement ter patent recognition of the complexity of 22. Kelves, D. J. A History of Patenting Life in The authors declare no competing financial interests. genomic relationships. the United States with Comparative Attention to Europe and Canada (citation of Judge G.S. Rich’s Our overview of current practices of published legal opinion therein) 20 (Office for Official Online links whole-genome patenting shows how these Publications of the European Communities, patents raise fundamental questions about Luxembourg, 2002). FURTHER INFORMATION 23. Eisenberg, R. S. Re-examining the role of patents in Egenis web site: www.ex.ac.uk/egenis genome utility, classification and the owner- appropriating the value of DNA sequences. Emory Law J. esp@cenet: http://ep.espacenet.com ship of intangible biological information. All 49, 783–800 (2000). European Patent Office: http://www.european-patent- 24. Maschio, T. & Kowalski, T. Bioinformatics — a patenting office.org EPO Online Public File Inspection web page: these issues mean that the future of genome view. Trends Biotechnol. 19, 334–339 (2001). http://ofi.epoline.org/view/GetDossier 25. Butler, D. Are you ready for the revolution? Nature 409, patenting should be carefully watched by sci- US Patent and Trademark Office: http://www.uspto.gov 758–760 (2001). World Health Organization — Patent Applications for entists, as much as by legal theorists, social 26. Bostyn, S. J. R. Living in an (imm)material world: SARS Virus and Genes: bioinformatics and intellectual property protection. J. Int. scientists and philosophers of biology — not www.who.int/ethics/topics/sars_patents/en to mention the patent owners themselves. Biotechnol. Law 1, 2–10 (2004). Access to this interactive links box is free online.

Author biographies Online links

Maureen O’Malley is a research fellow at Egenis, the Economic and Egenis web site Social Research Council Centre for Genomics in Society, at the www.ex.ac.uk/egenis University of Exeter, UK. There, she carries out philosophical, historical and sociological research into emerging postgenomic science, especially esp@cenet: systems biology and metagenomics. Her earlier work included social http://ep.espacenet.com and epistemological analyses of comparative microbial genomics in Ford Doolittle’s laboratory at Dalhousie University, Canada. European Patent Office: http://www.european-patent-office.org Adam Bostanci is a doctoral research student at Egenis. Previously, he worked as a science writer and journalist, most recently for Science.His EPO Online Public File Inspection web page: doctoral research focuses on the social and conceptual dimensions of http://ofi.epoline.org/view/GetDossier contemporary human genomics. US Patent and Trademark Office: Jane Calvert is a research fellow at Egenis. Her background is in science http://www.uspto.gov policy and science and technology studies. Before coming to Exeter, she worked at the University of Sussex, UK, on biotechnology policy, links World Health Organization — Patent Applications for SARS Virus and between university and industry, and on the concept of ‘basic research’. Genes: At Egenis, she is working in the area of ‘genosemantics’ — the study of www.who.int/ethics/topics/sars_patents/en the meaning and use of genomic terminology. She is particularly inter- ested in intellectual property and the use of genomic knowledge.

Further details of the authors’ work can be found on the Egenis web site: www.ex.ac.uk/egenis