Managing innovation: university-industry partnerships and the licensing of the Harvard mouse
Sasha Blaug1,Colleen Chien1 & Michael J Shuster
DuPont’s Oncomouse patent licensing program continues to cause a stir in academia and industry.
ver the last several decades, technology restructuring in the 1980s resulting in optimized for mutual near- and long-term Oand technological innovation have reduced industry R&D spending and scarcer benefits to the parties? Recently this dialog gradually replaced manufacturing and agri- federal R&D funding. Increased technology has been focused on DuPont’s licensing of culture as the main drivers of the US econ- transfer has sparked a debate on univ- the transgenic ‘Harvard mouse,’ subse- omy. The unparalleled system of American ersities’ roles in the national economy: on quently trademarked as the Oncomouse6. research universities1 and their association the extent to which these relationships affect http://www.nature.com/naturebiotechnology with industry are important drivers of the the mission of universities to carry out The Oncomouse patents new economy. The relationship between and disseminate the results of basic DuPont’s patent licensing program for universities and industry is multifaceted, research, and on how universities can man- ‘Oncomouse technology’ has caused a stir in encompassing exchanges of knowledge, age their partnerships, collaborations and academia and industry for over a decade7. expertise, working culture and money. technology transfer without compromising These patents broadly claim transgenic Whereas the transfer of technology from their mission. nonhuman mammals expressing cancer- universities to industry has been going on In the basic research paradigm, inves- promoting oncogenes, a basic research tool for more than a century, ties between uni- tigators’ inquiries are directed toward dev- widely used in the fight against cancer. The versity and industry have grown during the eloping an understanding of scientific Oncomouse technology provides an impor- past three decades, coinciding with the phenomena. Results from basic research tant tool useful for the initial testing of growth in biotechnology2. sometimes lay a foundation for advance- many new and promising cancer treatments. These university-industry interactions ments of enormous commercial signifi- Balancing academic access to this tool with can generally be categorized as collab- cance. However, as technology-based ind- DuPont’s rights to the technology has © 2004 Nature Publishing Group orations and research partnerships, with ustries have grown and become more com- proved to be a challenging undertaking that industry supplying financial support to plex, so has the relationship between univer- presents an interesting backdrop against universities in exchange for options on sities and industry. Issues arising related to which policy considerations implicated by developed technologies and inventions. The intellectual property, rights to publish research tool licensing can be viewed. most successful interactions result in dis- research results and academic freedom have DuPont’s increasingly aggressive approach coveries being licensed to and developed by caused many to reexamine university- to licensing Oncomouse technology ulti- industry in exchange for upfront and down- industry partnerships in relationship to mately made academic institutions and the stream monies3.Indeed, the current para- their value to public health, education and US National Institutes of Health (NIH) cry digm for biotech development is based on the economy. foul in response to concerns that compli- such partnering4.These partnerships have Much of the discourse concerns how and ance with onerous license terms would seri- been encouraged and stimulated by several whether the partnerships and subsequent ously affect the basic research mission of factors: passage of legislation, most notably licensing can be strategically designed and universities and hinder the search for new the Bayh-Dole Act of 1980, corporate managed to promote innovation, technol- cancer therapies. In 1999, after four years of ogy development, proper risk-rewards negotiations8, the NIH and DuPont arrived incentives and public health5. One aspect of at a memorandum of understanding Sasha Blaug is at Fenwick & West LLP, this partnering that has come under (MOU) that allowed NIH researchers and Silicon Valley Center, 801 California Street, scrutiny is deals that give the funding corpo- grantees to use the Oncomouse technology Mountain View, California 94041, USA. ration ‘first rights’ to develop and commer- for basic research without charge, provided Colleen Chien and Michael J Shuster are at cialize results from sponsored university that the research did not directly benefit Fenwick & West LLP, 275 Battery Street, research discoveries. commercial interests or a for-profit institu- San Francisco, California 94111, USA. Because substantial federal support is tion9.However, this did not stop DuPont 1Indicates joint first-authorship. often used to develop academic research from continuing to make aggressive licens- e-mail: [email protected] or before industrial partnering, how can the ing demands on academic and research [email protected] agreements governing these partnerships be institutions.
NATURE BIOTECHNOLOGY VOLUME 22 NUMBER 6 JUNE 2004 761 PATENTS
In response to such demands, the the patents arising out of Leder’s work were On July 1, 1999, DuPont and NIH signed University of California asserted that it is assigned to Harvard, DuPont was entitled to the memorandum of understanding allow- entitled to the same royalty-free use of the an exclusive license (with a right to subli- ing “public health service” scientists to use Oncomouse technology accorded to the cense) to all inventions in consideration of Oncomouse-related patent rights in bio- NIH. Citing specific provisions of the 1999 its financial support. Initially, DuPont’s medical research free of charge provided MOU, the University of California coun- license fees for use of the mouse in basic that such use was not “for any commercial tered DuPont’s demands with an offer to research were nominal, and the license car- purpose or for the direct benefit of any for- take a license to the Oncomouse technology ried with it very few restrictions16.Use of profit institution”20.The MOU went on to on the same no-cost terms and conditions the technology proliferated quickly. DuPont enumerate two specific activities that would provided to the NIH as described in the collaborated with Charles River Labora- fall into this exception if undertaken for a MOU. On October 28, 2003, NIH informed tories to further develop the Oncomouse commercial purpose or the direct benefit of DuPont that it felt that the University of and other transgenic cancer-prone mice. a for-profit institution: testing compounds California’s position was “consistent with But in the mid-1990s, DuPont began negoti- or using the mice to produce products. The the intent, terms and conditions” of the ating licenses with substantial monetary fees terms and conditions of the MOU were to 1999 MOU. With this pronouncement, the from commercial entities and licenses from apply to research conducted by the public University of California and NIH have put academics that were unusually restrictive17, health service or “recipient institutions,” into serious question DuPont’s ability to col- for example, requiring universities and their defined as organizations that received NIH lect royalties for the use of the Oncomouse researchers to file frequent reports on their funds to conduct research. technology by universities and nonprofit use of the technology. It also reportedly institutions conducting cancer research. The University of California solution Although this agreement apparently solved Development of the Oncomouse On July 1, 1999, DuPont and the problem for the NIH, DuPont contin- at Harvard NIH signed the memorandum of ued to make demands on universities21.The Building on the original transgenic work of importance of the mice as a basic testing http://www.nature.com/naturebiotechnology understanding allowing “public Gordon and Ruddle at Yale10, the tool created an uncomfortable situation for Oncomouse (or Harvard mouse) was devel- health service” scientists to use academic institutions, creating huge pres- oped in the laboratory of Harvard professor Oncomouse-related patent rights sures for some to take a license from Philip Leder in the early 1980s (ref. 11). DuPont rather than jeopardize ongoing This transgenic mouse expressed the mouse in biomedical research free of research. mammary tumor virus (MMTV)-myc charge provided that such use Universities that DuPont approached oncogene, and so was prone to developing a received NIH funds, and therefore were variety of different tumors12.In 1988, the was not “for any commercial properly considered “recipient institutions” Harvard mouse became the basis for the purpose or for the direct benefit covered by the MOU. But DuPont argued first US patent granted to a higher organ- that university research sponsored by com- ism13.Subsequent patents broadly claimed of any for-profit institution.” mercial entities, or “sponsored research,” an entire class of transgenic nonhuman should be subject to license terms beyond mammals and methods to produce them14. the MOU. DuPont asserted that such © 2004 Nature Publishing Group These patents claim not only transgenic asked researchers to seek DuPont’s approval research provided a “direct benefit” to mice expressing c-myc, but essentially any before sharing the results of research per- sponsors that were commercial, for-profit nonhuman mammal that expresses any formed using the mice with commercial or entities. On this basis, DuPont refused to cancer-causing transgene. Thus, a trans- nonprofit entities18. acknowledge the applicability of the NIH genic mouse created at a university that MOU terms to sponsored research projects. expresses a different oncogene and develops The NIH-DuPont MOU The University of California disagreed. different types of cancers would likely The cancer research community, by now Sponsored research conducted at the uni- infringe the broader claims of the Harvard accustomed to using transgenic test subjects versity’s campuses is for the “direct benefit” patent portfolio. without significant charge or limitation, of the university and its research mission, it DuPont’s contribution to the discovery of viewed DuPont’s activities as a significant argued. Indeed, the university has a long- the Oncomouse was in the form of a $6 mil- encroachment. Indeed, the licensing terms standing official policy of insisting that uni- lion donation it made to Harvard to support proposed by DuPont, one commentator versity research be conducted in an open Leder’s research. The grant was provided in noted, effectively allowed the company to academic environment, and in a manner 1981, when Leder left the NIH to join the “leverage its proprietary position in up- consistent with the freedom to publish22.In faculty of Harvard Medical School. In 1981 stream research tools into a broad veto right addition, intellectual property derived from and the subsequent year, Leder also received over downstream research and product sponsored research is consistently assigned NIH funding to study the regulated expres- development”19. One of the most promi- to the university. These policies, as well as sion of “genes of man and the mouse”15.By nent voices among those decrying DuPont’s the outcomes of collaborations, the univer- 1983, Leder and his colleagues had made the actions belonged to the NIH. High-level sity felt, support that the 1999 MOU, and initial c-myc transgenic mice, characterized negotiations ensued during which NIH nothing more, should govern the univer- their propensity for developing cancer and emphasized the importance to public health sity’s use of the Oncomouse technology23. filed their initial patent applications. The in having unimpeded access to the research In the university’s view, the up-front licens- first patent, entitled “Transgenic nonhuman tool, and it was rumored, the use of public ing fees that DuPont requested in many cir- mammals,” was awarded in 1988. Although funds in the initial discovery. cumstances dwarfed the overall value of the
762 VOLUME 22 NUMBER 6 JUNE 2004 NATURE BIOTECHNOLOGY PATENTS
sponsorship, and severely restricted the abil- MOU terms to sponsored research at UC, cess met by the licensing scheme set by ity of universities to participate in spon- however, so the issue presently remains another seminal set of research tools, sored research projects with companies unresolved. recombinant DNA technologies, could be interested in using mice that are arguably The current situation highlights many of instructive in this regard. Patent-holders covered by Harvard’s patent claims. the issues raised by the growing ties Stanford University and the University of Industry-sponsored research represents a between university and industry. These ties California licensed these technologies on vital source of funding for the University of have come under increased scrutiny in the terms that included a small upfront pay- California and for universities in general. As United States and abroad, as institutions ment and reasonable royalties. It is thanks to an example, such funding at the University and governments try to maximize innova- schemes such as this that the biotech indus- of California San Francisco increased nearly tion, advance public health and stimulate try was allowed to develop in its early years, fivefold from 1992 to 1999 (ref. 24). Nation- the economy. Although the outcomes of producing numerous life-saving and health wide, sponsored research represented 9% of these closer relations have generally been enhancements and growing into the multi- overall research funding at universities in viewed positively, one byproduct has been billion dollar industry it is today. 2000 (ref. 25). increasing scrutiny of, and as the Onco- Against this backdrop, and on behalf of its mouse case shows, threats to academic 1. Study on research universities funded by the European Commission (http://ed.sjtu.edu.cn/rank/ ten campuses and five medical centers, the research environments. As the narrow scope rank-2003.mht). University of California addressed the issue of the common law research exemption has 2. Association of University Technology Managers survey with DuPont head on. The University of recently been clarified27, many academic (2000) and Working Together, Creating Knowledge, The University-Industry Research Collaboration California argued that the university, as a institutions are finding the task of fulfilling Initiative (2001). recipient of NIH funding, was entitled to their basic research mission vulnerable to 3. Campbell, E. et al. Health Affairs 23, 64–76 (2004). take a license under the same terms 1999 legal problems. As the Oncomouse case 4. Hall, Z.W. & Scott, C. Science 291, 553 (2001). 5. Edwards, M.G. et al., Nat. Biotechnol. 21, 618–624 DuPont/NIH agreement. As the university illustrates, nowhere is this problem more (2003). pointed out, the MOU specifically provided pointed than in the case of research tools, 6. Associated Press. Uproar over mighty mouse (8 November, 2003). that “DuPont agrees that it shall make Har- which are often much more difficult to 7. Marshall, E. Science 296, 1212–1213 (2002). http://www.nature.com/naturebiotechnology vard Patent Rights available for use by non- “work around.” The potential for society to 8. Smaglik, P. Nature 403, 350 (2000). profit recipient institutions under separate lose out when access to these tools, which 9. Memorandum of Understanding between E.I. Du- Pont de Nemours and Company and Public Health written agreements in accordance with the themselves enable further advances, is Service (MOU) (http://ott.od.nih.gov/textonly/onco- terms and conditions outlined above.” Du- restricted is arguably much greater than in mous.htm). Pont asked the University of California to the case of more conventional inventions 10. Gordon, J.W. & Ruddle, F.H. Science 214, 1244–1246 (1981). solicit NIH’s opinion on this argument. The that have limited applications. 11. Palmiter and Brinster reported in the scientific litera- director of the office of technology transfer The ramifications of limiting access to ture the first cancer prone transgenic mouse 4 months earlier than Leder and days before the filing at the NIH responded to the University of research tools reach far beyond the aca- of Harvard’s patent. Brinster R.L. et al. Cell 37, California’s request in late October 2003, demic laboratory. Fears that the hold-up 367–379 (1984). concluded that the university’s interpreta- costs to society of limiting access to 12. Stewart, T.A., Pattengale, P.K. & Leder, P. Cell 38, 627–637 (1984). tion of the MOU was correct and that, “The Oncomouse technology would be too great 13. US Patent 4,736,866. PHS-DuPont MOU clearly provides for originally motivated the formation of the 14. US Patent 5,925,803. NIH nonprofit recipient institutions to be NIH MOU to make the technology widely 15. Grants 1Z01HD000074-4 through 1Z01HD000074- 11, awarded by the National Institute of Child Health © 2004 Nature Publishing Group offered the same terms and conditions as available at no cost to public health and Human Development (http://crisp.cit.nih.gov/). those in the PHS agreement”26. researchers and their affiliates. Now in this 16. Grossblatt, N. (ed.). Sharing Laboratory Resources: Genetically Altered Mice. Summary of a Workshop latest chapter, similar concerns have led the Held at the National Academy of Sciences, March A solution for all academic institutions? University of California to position the aca- 23–24, 1993 by National Research Council, Ch. 4, Although other institutions may have demic community, as affiliates of the NIH, 21–23 (National Academy Press, Washington, DC, 1994). already signed licenses with DuPont, the and the general public, as beneficiaries of its 17. Melbourne Herald Sun, Australia. p. 52 (8 November MOU expressly provides that its terms work,to reap the benefits of the MOU. 2003). supercede those of any other agreement: In the case of the Oncomouse, the 18. Lamb, C. Sacramento Business Journal (3 November 2003). “DuPont agrees that any nonprofit recipient progress of cancer research is advanced by 19. Heller, A. & Eisenberg, R.S. Science 280, 698–701 institutions currently licensed under the making the technology widely available to (1998). 20. See MOU, supra note 9, at Section (1). Harvard Patent Rights may amend its all researchers, including those working in 21. See, for example, Cook, G. The Boston Globe (31 May license, in a separate written agreement, in industrial settings, on commercially reason- 2002). accordance with the terms and conditions able terms. Providing significantly disparate 22. See, for example, Policy Guidelines Governing Openness and Freedom to Publish (http://www.spo. outlined above.” The NIH has essentially access to the technology as between research berkeley.edu/Policy/ucmemos/publish.html), and Pri- stated that the University of California insti- institutions, academic campuses and indus- nciples Regarding Rights to Future Research Results tution can license the Oncomouse technol- try laboratories could prevent the develop- in University Agreements with External Parties (http://www.ucop.edu/ott/preamble.pdf). ogy on the same terms and conditions as the ment of a ‘gold standard’ mouse to be used 23. See Memo Operating Guidance on Industry- NIH, as provided for in the MOU. Applying in cancer testing across settings. Licensing University Relationships, No. 89-20 (6 June 1989) (http://www.ucop.edu/raohome/cgmemos/89-20.html). this same logic, other “recipient institu- agreements should be structured to encour- 24. Lawler, A. Science 299, 330–333 (2003). tions” may also avail themselves of these age, rather than suppress, widespread use of 25. AUTM 2000, 2001 and 2002 studies (http://www. license terms and even apply them in place such technologies while also providing just autm.net/). 26. See NIH letter to UC (http://206.151.87.67/docs/ of terms that may already exist. DuPont has rewards to DuPont and others taking risks OncoMouseDuPont.doc). not conceded to the applicability of the in the ventures that produce them. The suc- 27. Madey v. Duke, 307 F.3d 1351 (Fed. Cir. 2002).
NATURE BIOTECHNOLOGY VOLUME 22 NUMBER 6 JUNE 2004 763